WATER POLLUTION CONTROL RESEARCH SERIES • 17020 DAO 07/70
OPTIMIZATION OF THE REGENERATION
     PROCEDURE FOR GRANULAR
        ACTIVATED CARBON
 U.S. ENVIRONMENTAL PROTECTION AGENCY

-------
           WATER POLLUTION CONTROL RESEARCH SERIES

The Water Pollution Control Research Series describes the
results and progress in the control and abatement of pollution
in our Nationrs waters.  They provide a central source of
information on the research, development,  and demonstration
activities in the Environmental Protection Agency, through
inhouse research and grants and contracts  with Federal, State,
and local agencies, research institutions, and industrial
organizations.

Inquiries pertaining to Water Pollution Control Research Reports
should be directed to the Head, Publications Branch (Water),
Research Information Division, R&M, Environmental Protection
Agency, Washington, D.C. 20460.

-------
          OPTIMIZATION OP THE REGENERATION PROCEDURE
                FOR GRANULAR ACTIVATED CARBON
          Mine Safety Appliances Research  Corporation
                Evans City, Pennsylvania   16033
                            for the
                  ENVIRONMENTAL PROTECTION AGENCY
                       Project  #17020 DAO
                       Contract #14-12-469
                           July  1970
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 • Price $1.25

-------
            EPA Review Notice
This report has been reviewed by the
Environmental Protection Agency, and
approved for publication.  Approval
does not signify that the contents
necessarily reflect the views and
policies of the Environmental Pro-
tection Agency, nor does mention of
trade names or commercial products
constitute endorsement or recom-
mendation for use.
                  ii

-------
                        ABSTRACT
Granular  activated  carbons,  spent in  tertiary treatment of
waste water,  are  thermally  regenerated in  regenerators such
as the multiple-hearth  furnace.   Wet  spent carbon is fed to
the regenerator  and undergoes  three  naturally occurring
steps, namely  (1)  drying  at  about 220°F,  (2)  pyrolysis of
the adsorbed  pollutants  at  500°  to 1550°F  and (3) activation
with flue  gas  and  steam  at  1600°  to  1700°F.   For each ad-
sorption-regeneration  cycle,  the  carbon volume loss varies
from 5% to  10%  and  the  activity  loss  is as high as 13% on
the first  cycle  but at  diminishing amounts on subsequent
cycles.

Laboratory  studies  of  parameters  controlling  regeneration
show that  alkaline  and  iron  nxide ash accumulation in the
carbon. during  the  use  cycles,  catalyzes  the  oxidation of
pores  in  the  18^  to 28A  diameter  range, the  pores most ef-
fective in  pollutant adsorption.   By  HC1  acid leach, to
yomflyo fho<^o  metallic  elements  from  th"e spent CaTPftn"!ETie
subsequenttherrria I  regeneration  prnrpprU  wit^ Tejj^de^""
struct!on  of  the  carbon  as  measured  by the iodine and
molasses  numbers.                       '—•	

Steam  as  activating gas  is  more  effective  than C02 in re-
covery of  the  initial  properties  of  the carbon.

Carbon volume  decrease  during  the baking  step averages out
around 2%  and  during the  activation  step  at  1.8%.  Baking
studies indicate  that  the 2%  is  not  a true carbon loss but
rather an  apparent  volume decrease due to  pyrolysis of
colloidal  pollutants on  the  carbon particle  exterior sur-
faces.  The true  loss  of  carbon  during the laboratory re-
generations is  then about 1.8%.   The  latter  appears to
proceed by  generation  of  submicron sized  fines from the
carbon particle  surfaces  rather  than  due  to  oxidation.
     drying  of  the  wet  carbon  is  more effective than fast
Attempted  regeneration  by  leaching  with solutions of NaOH,
NaOCl, ^02  and  CC1.  were  ineffective and/or uneconomical.

This report was submitted in fulfillment of  Project Number 17020DAO,
Contract 14-12-469, under the sponsorship  of the Environmental
Protection Agency-

Key Words:   Carbon  regeneration,  chemical regeneration,
             HC1  acid  treatment,  thermal regeneration,
             tertiary  treatment with carbon.


                           ill

-------
                  TABLE OF CONTENTS
                                                  Page No
  I.   CONCLUSIONS                                     l

 II.   RECOMMENDATIONS                                 5

III.   INTRODUCTION                                    7

 IV.   EQUIPMENT AND PROCEDURES                        9

        Rotary-Tube Thermal Regenerating Unit         9
        General Procedure for Thermal Regenerations  11
        Unit for Leaching Spent Carbons              12
        Carbon Test Procedures                       14

  V.   EXPERIMENTAL RESULTS AND DISCUSSIONS           17
        Task 1 - Study of Regeneration
            Operating Parameters                     17
        Task 2 - Adsorption-Regeneration
            Cycle Studies                            46
        Task 3 - Determine Feasibility of Low
            Grade Carbons as Make Up                 48
        Task 2A and 2B - Cyclic Adsorption-
            Regeneration Studies                     51
        Task 4 - Chemical Oxidation and Solvent
            Extraction                               52
        Task 5 - Engineering Studies on Furnaces     89
        Task 6 - Regeneration Control by Effluent
            Gas Analysis                             89
        Task 7 - Regeneration of 25 Ib Quantities
            of Spent Carbon                          97

 VI.   ACKNOWLEDGMENTS                               105

      APPENDIX A - INPUT AND OUTPUT GAS
            COMPOSITIONS OF ACTIVATION RUNS         107

      APPENDIX B - GAS ANALYSIS WITH NO CARBON
            IN REGENERATOR                          113

      REFERENCES                                    115
                           v

-------
                        FIGURES
                                                    Page
 1       Cross Sectional  View of Rotating Tube,
          End Caps and Heating Furnace               10

 2       Unit for Leaching  Carbon at Temperatures
          Up to 100°C                                13

 3       Unit for HC1  Leaching of Spent Carbons       15

 4       Temperature Profiles for Gas and Carbon
          in Multiple Hearth Furnace, Regeneration
          of Wet Spent Carbon                        21

 5       Effect of C0£ Activation on Iodine Number    30

 6       Pore Size Distribution, Darco Run 58 and
          Filtrasorb  400                             45

 7       OSF 400 Leached with Low Concentration
          Solutions of Caustic Soda and HC1  and
          Pure Water; Bulk Density as Function
          of Amount Liquid Passed Through Carbon     60

 8       OSF 400 Leached with High Concentration
          Caustic Soda Solution and HC1 and  Pure
          Water in Batch Type Process; Bulk
          Density as  Function of Number of
          Batch Treatments                           61

 9       Ash Content of HC1 Leached OSF 400           68

10       pH of Filtrate From HC1 Leach of OSF 400
          When Acid to Carbon Contact Time is 2 Hr   70

11       pH of Filtrate From HC1 Leach When 5 ml of
          Acid is Added to 500 cc OSF 400            73

12       pH of Filtrate From HC1 Leach When 20 ml
          of Acid is  Added to 500 cc OSF 400         74

13       Ash Content of Carbon at Lower End of
          Bed When Bed is  Treated with Different
          Quantities  of Acid                         77

14       Pore Size Distribution Measured by Water
          Adsorption  and by Mercury Penetration      81

                            vi

-------
                   FIGURES (Continued)
                                                   Page
15      Pore Size Distributions of Regenerated
          Carbons                                   82

16      Pore Size Distributions of Regenerated
          Carbons                                   83

17      Cumulative Surface Areas of Regenerated
          Carbons                                   84

18      Cumulative Surface Areas of Regenerated
          Carbons                                   85

19      Particle Volume Decrease During Activation
          as Function of Difference in C02
          Utilization as Determined by Direct
          Measurement and by CO-Ho Analysis         95

20      Iodine Numbers of Regenerated Darco, F 400
          and WVP&P Carbons as Function of Bulk
          Densities                                102
                            vi5

-------
                        TABLES
                                                     Page

   I    Regeneration Conditions, Effect of Oxygen      19
  II    Regeneration Results, Effect of Oxygen         20
 III    Regeneration Conditions, Wet and Dry OSF 400   23
  IV    Regeneration Results, Wet and Dry OSF 400      24
   V    Regeneration Conditions, Effect of
         Temperature                                  25
  VI    Regeneration Results, Effect of Temperature    26
 VII    Regeneration Conditions, Steam and C02
         Activation                                   28
VIII    Regeneration Results, Steam and C02
         Activation                                   29
  IX    Baking Conditions Fines Formation Study,
         OSF 400                                      32
   X    Baking Results Fines Formation Study, OSF 400  33
  XI    Weight Decrease Determined From Vent Gas
         Analyses                                     35
 XII    Volume Decrease During HC1 Leach and Fixed
         Bed Baking of OSF 400                        37
XIII    Volume Decrease During HC1 Leach and Rotary
         Tube Baking of OSF 400                       38
 XIV    Volume Decrease of Nonleached Carbons
         During Baking                                39
  XV    Volume Decrease During Activation              41
 XVI    Regeneration Conditions, Effect of Particle    42
         Size on Regeneration
XVII    Regeneration Results - Effect of Particle
         Size on Regeneration                         43
                          viii

-------
                    TABLES (Continued)
                                                       Page
 XVIII   Regenerating Conditions, Tasks 2 and 3         49

   XIX   Results of Regenerations, Tasks 2 and 3        50

    XX   Regeneration Conditions, Tasks 2A and 2B,
           First Cycle                                  53

   XXI   Regeneration Results, Tasks 2A and 2B,
           First Cycle                                  54

  XXII   Regeneration Conditions, Chemically
           Treated OSF 400                              55

 XXIII   Regeneration Results, Chemically Treated
           OSF 400                                      56

  XXIV   Caustic Solution and Methanol Leach of
           Once-Spent Filtrasorb 400, Run 18            58

   XXV   Regeneration Conditions, HC1 Leach             64

  XXVI   Regeneration Results, HC1 Leach                65

 XXVII   Ash Content of Regenerated Carbons Before
           and After Dilute HC1 Acid Leach              66

XXVIII   Decolorizing Test Results on HC1 Leached
           and Nonleached Carbons                       66

  XXIX   Ash Content of HC1 Leached OSF 400             71

   XXX   Analysis of Filtrate From Experiment 3         71

  XXXI   Analysis of Filtrate From Experiment 1         69

 XXXII   Ash Analysis of HC1 Leached Carbons            76

XXXIII   Ash Composition of Acid Pretreated and
           Nonpretreated Regenerated Carbons            79

 XXXIV   Iodine and Molasses Numbers as Determined
           by Test and Calculated From Surface Area     86

  XXXV   Surface Area Change at Different Pore
           Diameters During Regeneration                87

                             Ix

-------
                     TABLES (Continued)
  XXXVI   C02 Utilization
 XXXVII   Results of Gas Analysis During Activation
XXXVIII   Reactions Associated With Each Gas Analysis
  XXXIX   K for Tests 1  Through 5
     XL   Regeneration Conditions, WVP&P and Darco
    XLI   Regeneration Results, WVP&P and Darco
Page

 92
 93
 96
 97
 98
 99

-------
                      SECTION I

                     CONCLUSIONS
1.  Thermal regeneration of wet spent carbon from the
tertiary treatment of waste water proceeds by three
naturally occurring steps;  (1) drying at about 212°F,
(2) pyrolysis (baking) of the adsorbate at 500° to 1550°F
and (3) activation with steam flue gas mixture at 1650°F to
1700°F.  In the Pomona multiple-hearth furnace, Step 1 oc-
curs in 22 min, Step 2 in 9 min and Step 3 in 15 min.

2.  Slow drying of the as-received wet spent carbon, as in
an air-convection oven at 150°C for 48 hr, produced regen-
erated carbons of higher iodine number than when drying was
done rapidly in 10 to 20 min in the rotary-tube laboratory
regenerator.  Iodine numbers of the regenerated slow-dried
carbons were in the 950 to 970 mg/g range while those of
the fast-dried carbons were in the 900 to 935 mg/g range.

3.  Low temperature activation at 1600°F produces lower
activity regenerated carbons than when the activation is
carried out at!650° to 1700°F.  The iodine numbers for the
former were in the range 900 to 910 mg/g while for the
latter they were 950 mg/g and slightly higher.

4.  Thermal regeneration of graded mesh size fractions in
the 8 to 60 mesh size range showed no significant difference
in rate of regeneration.  Spent West Virginia Pulp and Paper
Company carbon of 8 to 20 mesh fraction regenerated at es-
sentially the same rate as the 30 to 60 mesh fraction.  The
finer mesh fractions, however, showed a tendency toward
larger carbon losses during the regeneration.

5.  The rate of thermal regeneration is more dependent on
the pore structure than on particle size.  A carbon such
as 14 to 30 mesh Darco, with 0.60 cc/g of pores in the 30A
to 1000A diameter range, regenerates faster and under much
milder regeneration conditions than Filtrasorb 400 with
0.20 cc/g of pores in the same diameter range.  When gran-
ular carbons are sized to smaller particles, pores in the
10.000A to 100,OOOA diameter range are destroyed, but since
these pores do not control the regeneration rate, decrease
in particle size does not increase the regeneration rate.
This generalization applies to processes involving gas dif-
fusion but not to liquid diffusion.

-------
6.   Attempts to regenerate spent carbons by leaching with
CC14 or with aqueous solutions of NaOH, NaOCl  and \\2®2 were
not effective or economical.

7.   HC1 acid leach of spent carbons had a beneficial effect
on  the subsequent thermal regeneration.  Iodine and molasses
numbers of the regenerated products were closer to the in-
itial values than for the nonleached regenerated products.

8.   The effectiveness of the HC1 acid leach was due to the
fact that the acid leached out metallic elements such as Fe,
Na, Ca and K which, if left in the carbon, catalyzed the
oxidation of the pores in which they happen to be located.

Pore structure studies showed that, in nonleached carbons,
enlargement of pores occurred in the ISA* to 28A diameter
range during activation.  This fact also implies that pol-
lutant adsorption is predominantly in pores of this size
range since the metallic elements that catalyze the pore
enlargement are initially a part of the pollutants.

9.   Studies to optimize the HC1 treatment showed that it
could be performed effectively at ambient temperature, at
acid to carbon contact time of about 2 to 3 hr, at 0.076 Ib
acid/lb carbon (38% HC1 assay) and 1.4% acid concentration
based on weight of HC1.  Estimated cost of acid is 0.145<£/lb
of regenerated carbon.

10.  Steam activation consistently produced regenerated
carbons of higher iodine number than C02 activation.  For
HC1 acid pretreated carbons, steam activation  produced car-
bons of 1040 mg/g iodine number while with C02 activation
the iodine number was 960 mg/g.  For nonleached carbons, the
results were 940 mg/g with steam activation while with
the iodine number was 880 mg/g.

11.  Since the baking is done under conditions that should
not cause oxidation of carbon structure, no carbon volume
loss should be expected.  About a 2% volume decrease con-
sistently appeared.  The evidence obtained appears to in-
dicate that this volume decrease is due to pyrolysis of
colloidal contaminants on the exterior surface of the carbon
particles and not a true loss of the activated carbon.

12.  Carbon volume decrease during the activation step
averages out around 1.8% and appears to be due to loss of
submicron fines broken off the carbon particle exterior
surface rather than oxidation by the steam.

13.  Cyclic adsorption-regeneration studies, where the
spent carbon was not acid leached, showed progressive de-

-------
crease of iodine number over two cycles.   On the first cycle
of the study, employing acid leach, the initial iodine num-
ber was attained.

14.  In direct-fired furnaces the flue gas usually contains
1  to 2% unreacted oxygen.  The laboratory studies showed
that oxygen in these quantities has no effect on the regen-
eration process.

15.  A vent gas analysis study was undertaken to determine
whether monitoring the vent gases from the activation step
could be used to control the process.  The results produced
too many inconsistencies to recommend the method.

16.  Vent gas analyses showed that the H20, CO, C02 and H2
react with each other and come to an equilibrium according
to the equation
                      co~

where the equilibrium constant K varies with temperature.
With excess steam, the utilization of C02 is suppressed and
in some cases C02 has been produced.

-------
                      SECTION II

                   RECOMMENDATIONS
The success of the HC1 acid
oratory scale recommends it
scale regeneration systems.
economical relative to acid
to be the difficult problem.
leach pretreatment on  the  lab-
for application to the large
 The pretreatment step is
cost, but corrosion is expected
Further studies are recommended on carbon pore structure
relative to ease of regeneration.  Such a study would lead
to a better understanding of the adsorption and regeneration
process and form a basis for the selection of carbons best
suited for tertiary treatment.

Studies of the effect of particle size on ease of regen-
eration demonstrated that regeneration of powdered carbons
could be done effectively in conventional equipment, with
modifications to equipment and procedures of operation.
Further study is recommended for this approach.

-------
                     SECTION III

                    INTRODUCTION
On the previous program, Contract No. 14-12-107, regen-
eration equipment was installed and regeneration procedures
worked out for study of the regeneration process.   Partial
answers were found for the regeneration process, but also,
new areas of study became apparent which were outside of
the scope of the contract.

One of the more important observations made was that the
overall regeneration, as practiced in the multiple hearth
furnace, consisted of three natural occurring steps; i.e.,
(1) drying, (2) pyrolysis of adsorbate (baking) and (3)
activation.  In the activation step, free carbon residue
from the baking step is removed by steam oxidation.  These
three steps were studied, but completely satisfactory
answers were not obtained regarding the parameters affect-
ing them.

The present program, Contract No. 14-12-469, was initially
contracted for 14 months, but promising areas of further
investigation were discovered that were outside the scope
of the contract.  To investigate these new areas, the con-
tract was extended another 6 months.  The continued studies
were divided into seven tasks as identified below.

     Task 1 - Study of Regeneration Operating Parameters

     Task 2 - Adsorption-Regeneration Cycle Studies

     Task 3 - Determine Feasibility of Low Grade Carbon
              as Make Up

     Task 4 - Chemical Oxidation and Solvent Extraction

     Task 5 - Engineering Studies on Furnaces

     Task 6 - Regeneration Control by Effluent Gas Analysis

     Task 7 - Regeneration of 25 Ib Quantities of Spent
              Carbon

After considerable study had been made on the hearth furnace
operating parameters as part of Task 1 studies, a set of
conditions were to be set for the regeneration of one con-
tactor  full  of carbon at the Pomona, California pilot plant,

-------
a project jointly funded by the County Sanitation Districts
of Los Angeles County and the Federal Water Quality Admin-
istration, U.S.  Department of the Interior.

Task 2 had a two-fold purpose.  The initial properties of
the spent carbon, as received from the Pomona plant, were
not known exactly, therefore there was some uncertainty in
determining the recovery of the initial properties.  This
uncertainty was removed in Task 2 ( and also Task 3) studies
since the properties of the starting material were deter-
mined and the same batch of carbon was carried throughout
the cyclic studies.  The other purpose was to study the type
and rate of carbon degeneration over a number of cycles of
regenerati on.

Task 4 involved study of liquid chemical  treatment of the
spent carbon as a primary regenerating process, an alterna-
tive to the thermal regeneration.  It also involved study
of chemical pretreatment of spent carbon  as an aid to the
thermal regeneration.

Task 4 opened up new areas of investigation in that HC1
leach of the spent carbon prior to thermal regeneration
greatly minimized loss of adsorptiye capacity during re-
generation.  To investigate the acid leach, and other areas
of promise, the program was extended another six months.

-------
                      SECTION IV

               EQUIPMENT AND PROCEDURES
Rotary-Tube Thermal Regenerating Unit

The thermal regenerations, involving the baking and activat-
ing steps, were conducted in a continuous feed, indirectly
heated rotary-tube regenerator.  If offers the widest range
of versatility with respect to operating parameters.  Acti-
vating gas composition and input rate can be varied to any
practical limit.  Heat input can be varied independently of
the activating gas input rate and the carbon residence time
can also be varied independently of the other parameters.

Figure 1 shows the cross sectional view of the rotating tube
and sections of the furnace.  Thermocouples TC I, TC II, and
TC III monitor the three sections of the furnace I, II, and
III, which can be temperature controlled independently of
each other.  Thermocouples TC (I), TC (2), TC (3), TC (4)
and TC (5) monitor the gas temperature over the carbon bed.

Other pertinent details of the unit are listed below:
     Tube size and material -
     Flights
     Heati ng
3.25 in.  dia by 65 in.
length, stainless steel

six 1/4 in.  deep flights
about the full length of
tube.  This  depth flight
gives carbon a rolling
action rather than a lift
and free  fal1.

the heated portion of tube
corresponding to furnace
length is 42 inches, carbon
residence times are reported
for the 42 inches rather than
the total 65 inches.
     Tube inclination
     Carbon residence time
has been varied
to 4 inches per
tube length.
from level
42 inches of
The carbon residence time
can be varied from 10 min
to 3.7 hr by varying in-

-------
      Carbon feed
Heating sections of furnace
t
End x , 	 L
"" li-TJ
Gas outlet
ube
Carbon Rotating
seal gas seal
/ '' n
^ m 1
1
I | II
1
1
TC I | TC
1 j l'l
i '"
I
II | TC III
i i
IM I !
Rotati ng
gas seal
/End cap
/
^t TC(1) TC(2) TC{3) TC(4) TC(5J 1 ,
	 ^> c
--/,
Rotary /
tube /
Drive
sprocket
1 1

— I 	 ~ — pp — — r^ ~i —
Coll type
electric heaters
I I
I I
A
Gas inlet \
1 i
                     10
                               20
                                          30
                                                     40
                                            Inches
                             50
                                        60     65

                                       Regeneratec
                                       carbon re-
                                       ceiver - 2  qt mason
FIGURE  1  -  CROSS SECTIONAL  VIEW  OF ROTATING TUBE,  END  CAPS AND HEATING FURNACE

-------
Gas input
clination and rotation rate.
The usual residence time for
baking is 30 min and for ac-
tivating 15 min.

The input gas composition and
input rate are varied by met-
ering N2, COp, 0? from com-
pressed gas cylinders and gen-
erating steam by means of a
calibrated boiler.   To simu-
late multiple hearth furnace
operations, gas mixtures ap-
proaching the flue  gas mix-
ture are usually used, i.e.,
                         72% N2, 10% C02 and 1
                        H20.
                be varied from
Carbon feed rate
Vent gas analyses
                              Total""input can __  . _
                              4  to  40  ft3/hr  (stp).

                             - An  auger  type volumetric feed-
                              er  is used having a  feed rate
                              range between 0.010  and 0.100
                              ft-vhr for granular  carbons.
                              The  usual feed  rate  is 0.016
                              ft3/hr,  453 cm3/hr.

                             - About 1.0 ft3/hr of  gas mix-
                              ture  is  withdrawn from the
                              main  vent gas stream during
                              activation, for C02, CO and
                              H2  analyses.  The side-stream
                              is  passed in succession through
                              a  (1 ) glass wool fi1ter, (2)
                              water freeze out trap at -80°C
                              and  (3)  MSA In-Line-Ultra par-
                              ticulate  filter, before reach-
                              ing  the  gas analyzers.  The CO
                              and  COo  concentrations are ana-
                              lyzed with MSA  Lira  Model 300
                              analyzers and the H2 with an
                              MSA  Thermatron.

General Procedure for Thermal Regenerations

Spent carbons, as received from  the tertiary  treatment pilot
plant at Pomoma are first dried  for 48  hours  in an air-con-
vection oven at 150°C, then  sieved  to  14 by 40 mesh, and
blended as preliminary preparation  for  the regeneration.
The amount of oversize and undersize particles was always
small and was discarded.  The drying is considered the first
                      11

-------
step in the overall  regeneration procedure.  Generally 2500
cm-* of the dried spent carbon are measured out for each re-
generation run.  The second step is baking,wherein the dried
spent carbon is passed through the regenerator at a rate to
give a 30 min residence time.  The gas input during baking
has been varied considerably.  On some runs a flue gas mix-
ture was used and on others, nitrogen.  The latter, of
course, is not attainable in a direct-fired hearth furnace,
but the runs were made to obtain data on extreme conditions.
The gas temperature on most runs was controlled to vary lin-
early from 800°F at the carbon influent end of the bed to
1550°F at the carbon effluent end.  The baking was also tried
with wet carbon feed, i.e., 40% by weight of water, to sim-
ulate conditions in the Pomona hearth furnace even closer.
With wet carbon feed, the baked carbon bulk density was
slightly less than with dried carbon feed.  However, with
wet carbon feed, vent line plugging occurred frequently,
hence, dried carbon feed was then standard practice for
most of the baking runs.

During baking, tne adsorbate is carbonized yielding a vola-
tile portion and a free carbon residue which is then removed
by steam oxidation during the activating step.  Fifty to
seventy percent of the adsorbate is removed as volatiles.
For the activating step,the gas temperature profile down the
tube is generally 1550°F for section I, 1650°F for section
II and 1700°F for section III.  The required activating gas
input rate is calculated from previous experimental results.
For this purpose the bulk density is used as the control
test, i.e., gas input is varied to bring the bulk density
to that of the virgin carbon.

Further details on the regenerator unit and operating pro-
cedures are presented in the final report for Contract 14-
12-1071.

Unit for Leaching Spent Carbons

Task 4 involved leaching experiments with chemical oxidants
and solvents to develop a primary regenerative process or,
if not successful in this effort, to find an oxidant or
solvent that would aid the thermal regeneration process.
Chemicals tried in aqueous solution were H^O?, NaOH, NaOCl
and HC1.  In some preliminary experiments, Buchner funnels
were used, but their use did not permit sufficient control
over temperature and flow rate of solution through the car-
bon bed.  To correct this problem, the unit shown in Figure
2 was constructed.
                            12

-------
      Glass tube
      54 mm 00
      47 in. tot,
      length
Solution
constant
head
   level at
   pressure\
Disk with center
hole to mln1- —'
mlze convection
current
     Heater
       &
   Insulation
   Filtrate
     flow
 Carbon bed
      Screen
-A
u
X
X*
^


/
X
X
\



*



< '.'X'
••V.';
/•'* •*
1 ."•••
•^ " ^ -
» '." *»
*'".**






^




F
V
v P




K
J
limp



12





1n. x 18 1n.
27 1 cap.
of leach
solution


FIGURE 2 - UNIT FOR LEACHING CARBON
       TEMPERATURES UP TO 100°C
                                          AT
                       13

-------
As shown in the figure, the carbon bed is contained in the
lower half of the glass tube with a constant head of solu-
tion above the bed.  The column was heated by resistance
wire wrapped around the tube to maintain temperature of
solution and carbon bed at temperatures near 95°C.  Flow
rate was maintained by adjusting the valve attached to
the filtrate tube.  Flows of 0.5 to 3.0 1/hr were experi-
mented with, but the flow generally was set at 1.0 1/hr.

The progress of the regeneration was monitored by removing
the carbon from the column at intervals for drying and
weighing.  Prior to removing carbon from the column, the
carbon bed was leached with pure water to remove the chem-
icals being used.  For HC1 acid leaching of larger quan-
tities of spent carbon at ambient temperature, the unit
shown in Figure 3 was used.

Carbon Test Procedures

To determine the effectiveness of the regenerating process,
various measurements and tests are made on the carbon be-
fore and after the three steps of regeneration.  These are
namely bulk density, particle density by mercury displace-
ment, real density by helium displacement, iodine number
and molasses number.  The bulk density is based on weight
per unit volume of the container occupied by the carbon, the
particle density is per unit volume of particles and real
density is per unit volume of solids in the carbon particles,
Knowledge of the particle and real densities permits calcu-
lation of the pore volume of the carbon.  The bulk densities
are measured during the run for control purpose since they
can be determined quickly.

The iodine number is related to the total surface area of
the carbon and molasses number to the area of pores larger
than 28A" in diameter.

Further details on making these measurements and their sig-
    cance are given in the final report of Contract 14-12-
nifjc
107T-
Tables reporting the test results present the bulk density,
percent weight decrease, percent bulk and particle volume
decreases, pore volume, and iodine and molasses numbers.
At the top of each table, property data are given on the
spent carbon, representing starting material, and on the
virgin carbon, representing the goals to be attained by the
regeneration.  A comparison of the bulk densities, pore
volumes and iodine and molasses numbers of regenerated car-
bons with those of the virgin carbon, measures the degree of
                             14

-------
HC1 acid
  or
 water
    Liquid
    level
    Spent
   Carbon _
   2500 cc
Glass container,
   4 1 cap.
Fine weave
nylon liner

Polyethylene
Contai ner,
 3.8 cap.
                                  SS fine mesh
                                  screen
                                  Flow controlled
                                  by stopcock
                                  Graduated
                                  cyli nder
FIGURE 3 - UNIT FOR HC1 LEACHING OF SPENT CARBONS
                        15

-------
success of the regeneration.  The percent decrease in bulk
and particle volumes measures the carbon loss.
                           16

-------
                      SECTION V

         EXPERIMENTAL RESULTS AND DISCUSSIONS



Task 1 - Study of Regeneration Operating Parameters

Effect of Oxygen - In the direct-fired multiple hearth fur-
nace, the carbon comes in contact with flue gas from com-
bustion of natural gas or flue gas mixed with product gases
from activation.  The flue gas can contain 1% to 2% of
oxygen by volume with compositions approximately as given
be!ow:


                 	Composition, %	
    Component    With 6% excess air    With 12% excess air

       N2              71.7                   71.5

       C02              9.2                    8.7

       H20             17.9                   17.7

       02               1.0                    2.1

                       99.8                  100.0

The composition  of the gas passing over the hearth changes
after coming in  contact with  the carbon.   For the Pomona
furnace, the calculated composition of gas entering the
fourth hearth, where  the baking step occurs, is approxi-
mately as follows:

     Component    ft3/(hr x  Ib carbon)     Composi tion,%"'

        N2                25.7                 46.2

        C02                3-8                  6-8

        H20               18.2                 32.7

        CO                 3.9                  7.0

        H2                 4.0                  7.2

                          55.6                 99.9

     (!)  This is a  calculated composition.  Due to a gas
          phase  reaction between the CO and H20, the  volumes
                            17

-------
          of CO and H20 are reduced while C02 and H~ volumes
          are increased.  This reaction is discusses later
          in the report.

In the Pomona furnace regeneration, steam is added to the
flue gas, but if steam were not added the calculated com-
position of gas entering the fourth hearth would be closer
to the percentagesgiven below:

    Component    ft3/(hr x Ib carbon)    Composition.%
       "2
       co2

       H20

       CO
24.1

 2.7

 3.8

 2.7

 2.2

 0.2

35.7
 67.5

  7.6

 10.6

  7.6

  6.2

  0.6

100.1
Two activations were run with flue gas mixture of 1% oxygen.
These are Runs 4 and 13, with conditions and results given
in Tables I and II.

In regard to the baking step, no attempt was made to exactly
duplicate the gas compositions of the hearth furnace.  It
was rationalized that at the 1550°F maximum temperature, the
presence of CO and H2 would have very little effect.  In the
baking steps, flue gas mixtures were used with 1% oxygen
content.  These are Runs 1  and 12 in Tables I and II.  A
comparison of the results,  with and without oxygen, tend to
show no significant difference.  Run 16 does not fit the
pattern of results generally obtained on nonoxygen acti-
vations, hence some doubt exists on the validity of the low
(920)  iodine number.

The possibility exists that the oxygen reacts with the H.
and CO in gas phase and never reaches the carbon, where
would  react exothermally.   On subsequent regeneration runs,
the addition of oxygen was  discontinued.

Baking Wet and Dry OSF 400  - Figure 4 shows the measured gas
temperature over each hearth of the Pomona furnace.  The car-
bon temperatures shown here were calculated from the thermo-
dynamic properties of the  system.   As indicated in the figure,
                            18

-------
               TABLE I - REGENERATION CONDITIONS, EFFECT OF OXYGEN


                                    Temperature, °F        Gas input, ft3/hr (stp)
Run no.    Regeneration step
64 (1)         baking
65 (1)     Run 64 activated

 1             baking
 4         Run 1 activated

 7             baking
13         Run 7 activated

12             baking
16         Run 12 activated

22             baking
23         Run 22 activated
    (1) Data on runs from contract 14-12-107.
1
1700
1700
550
1550
700
1550
750
1550
650
1545
2
1700
1700
1250
1650
1200
1650
1250
1650
1255
1650
3
1700
1600
1550
1720
1550
1700
1500
1700
1545
1700
h
4.8
3.6
6.8
1.09
7.9
2.7
6.8
2.7
7.0
3.7
CQ_2
0.00
0.50
0.90
0.14
1.0
0.34
0.90
0.34
0.90
0.47
02
0.00
0.00
0.10
0.015
0.00
0.02
0.05
0.00
0.00
0.00
H20
0.00
1.90
1.71
0.37
1.27
0.47
1.57
0.56
2.11
0.86

-------
           TABLE II - REGENERATION RESULTS.  EFFECT  OF OXYGEN



Run no.
virgin
spent ( ! i
64 (2)
65 (2)
1
4
virgin
spent {•*)
7
13
12
16
22
23

Bulk
density ,
q/cc
0.490
0.570
0.522
0.486
0.500
0.480
0.468
0.583
0.506
0.476
0.503
0.479
0.510
0.480
Bulk
Weight volume
decrease, decrease,
%°/
la

	 	
10.41 17 2
7.5J u'*
12.91 0.5|_ , o
5. If 17'4 1.3J 1>8
— _ _ _ _ _
	 	
13.51 ,g 5 O.Tl ! 6
6.9J 19'b 0.9J '
14.21 g 7 0.6] . ,
6.4_j 19'7 0.9f K5
13.31 l.ol
7 . 2 J ] y ' ' 1 . 2J 2 . 2
Particle
vol ume
decrease ,
o/
10

	
1.41 2 2
0.8_j L't-
]*81 i ft
2. Of 3'8
— - _
	
l-4l 1 9
0.5J '*y
?:i}i-8
°-5l 2.2
l.Tf

Pore
volume,
cc/cc

0.510
0.598
0.625
0.619
0.628
0.650
0.500
0.611
0.633
0.615
0.635
0.600
0.630

Iodine
number ,
mq/g
990
650
960
950
950
1090
630
940
950
940
920
950
950


Mol asses
number
220
210
225
210
220
250
190
211
258
224
253
_ — «
	
(1)   Runs 1,  4,  64 and 65 with  this  once-spent  Filtrasorb  400  (OSF 400)
(2)   Data from runs on contract 14-12-107
(3)   Runs below  with this OSF 400

-------
  3500
01
  3000
  2500
  2000
fO


S. 1500
E
O)
  1000
   500
                                              Calculated  flue gas  temperature,  3350°F
Calculated flue gas-steam
mixture to hearth 4,
2650°F
             Calculated  flue gas-steam
             mixture temperature to
             hearths 5 and 6, 2230°F
                   Mean gas temperature
                                                        Carbon temperature
             12345
                                       Hearth  number

       FIGURE  4  -  TEMPERATURE PROFILES FOR  GAS AND  CARBON  IN  MULTIPLE
               HEARTH FURNACE, REGENERATION  OF WET SPENT CARBON

-------
wet carbon fed into the furnace dries while traversing the
first three hearths and then goes through the baking step
on the fourth hearth.  Activation occurs over hearths 5
and 6.

To simulate the conditions in the hearth furnace, a number
of drying-baking runs were made in the laboratory rotary
furnace with rewetted OSF 400.  It is necessary to dry the
wet spent carbon, as received from Pomona, to permit mea-
surements of its properties.  Rewetting consisted of adding
water to the carbon just to the point where it was still
free flowing.  On a dry spent carbon basis this occurred at
about 38% water content.

Two regenerations of rewetted carbon were run.  The con-
ditions and results are given in Tables III and IV, Runs
2-3 and 8-14.  Data on nonwetted carbon runs, Runs 1-4,
and 22-23 are also given for comparison.  The bulk densi-
ties of wet-baked carbons were slightly lower than those of
the dry-baked carbons.  This would normally be regarded
favorably, except on activation the wet-baked carbons had
slightly lower iodine numbers, i.e., 900 and 935 mg/g com-
pared to 950 mg/g for the dry-baked carbons.  In Task 4,
regeneration Runs 9-15, Table XXII, the OSF 400 was wetted
with HpOo and in this case also the iodine number was down,
i.e., 920 mg/g.

Baking of rewetted carbons caused considerable difficulty
by way of vent line plugging.  To avoid this problem, sub-
sequent baking runs were made with oven dried OSF 400.

Effect of Temperature - As shown in Figure 4, the gas tem-
perature varies from 600 to 1700°F over hearths 1 to 4 and
holds steady at 1700°F over hearths 5 and 6.  Various com-
binations of temperatures on baking and activating were
tried.  Some were almost duplicates of the hearth furnace
conditions, some milder and some more severe.  Conditions
of regeneration and results which show the effect of tem-
perature are presented in Tables V and VI.

In the regeneration Run 22-23, the baking was carried to the
maximum temperature of 1545°F with relatively high input of
C02 and H^O.  Activation was then conducted at 1545°F to
1700°F.  The regenerated carbon iodine number was 950 mg/g.
When the baking temperature was increased to 1600°F, as in
regeneration Run 48-51, the regenerated carbon iodine number
dropped to 910 mg/g.  Run 50-52 is in part a repeat of Run
48-51 except that the activating temperature was increased
to the 1700°F level.  For this run the iodine number dropped
to 900 mg/g.
                           22

-------
TABLE III - REGENERATION CONDITIONS, WET AND DRY OSF 400
Run
no.
1
4
2
3

8
14
22
23
Regenerating step
and special treatment
dried and baked
Run 1 activated
dried, rewetted, baked
Run 2 activated
rewetted 550 ml H20/
2500 cc and baked
Run 8 activated
dried and baked
Run 22 activated
Temperature,
1
550
1550
680
1550

980
1550
650
1545
2
1250
1650
1250
1650

1270
1650
1250
1650
°F
3
1550
1720
1550
1720

1550
1700
1545
1700
Gas i
N
6
1
6
1

7
2
7
3
2
.8
.09
.8
.99

.9
.7
.0
.7
nput ,
C02
0
0
0
0

1
0
0
0
.90
.14
.90
.26

.00
.34
.90
.47
ffVhr
°2
0.10
0.015
0.10
0.027

0.00
0.00
0.00
0.00
(stp)
?
1.
0.
1.
0.

1.
0.
2.
0.
,(J
71
37
71
37

75
51
11
86

-------
                    TABLE IV - REGENERATION RESULTS, WET AND DRY OSF 400
N5
-P-
Bulk Particle
Bulk Weight volume volume Pore
density, decrease, decrease, decrease, volume,
Run no. g/cc % % % cc/cc
virgin 0.490
spent (1) 0.570
1 0.500 12.9
4 0.480 5.1
2 0.500 12.4
3 0.475 6.4
virgin, 0.468
spent (2) 0.583
8 0.507 13. l"
14 0.482 6.5_
22 0.510 13.3
23 0.480 7.2_
0.510
L 17 A °-5l i Q 1-s1?R 0.619
f 1.3J ' -° 2.0J" 0.628
-180 1 • ^ ;_ -i n 2 . 2 '_ /i i 0.613
IB' 1.3! 3'° 1.9; 4'] 0.631
0.650
0.500
18 R °-5L 1 fi °-9' 9 R 0.607
j 1.11 1.9i 0.629
- 19 7 T-0"! 9 9 °-5 oo °-600
1 .2f ^'2 1.7" 2.2 0>630
Iodine
number,
mg/g
990
650
950
950
923
900
1090
630
940
935
950
950
Mol asses
number
220
210
210
220
224
220
250
190
212
229
	
           (1)   Runs 1,  4,  2 and 3 with this OSF 400
           (2)   Runs 8,  14, 22 and 23 with this OSF 400

-------
Ul
                 TABLE  V  -  REGENERATION  CONDITIONS.  EFFECT  OF  TEMPERATURE


                                                                               o
                                            Temperature,  °F        Gas  input,  ft /hr  (stp)
      Run  no.      Regenerate ng  step
22           baking
23        Run 22 activated

48           baking
51        Run 48 activated

50           baking
52        Run 50 activated

10           baking
41        Run 10 activated

61           baking
66        Run 61 activated
1
650
1545
820
1550
800
1700
800
1560
1700
1700
1250
1650
1350
1650
1350
1700
1250
1650
1700
1700
1545
1700
1600
1700
1600
1600
1550
1700
1600
1600
7.0
3.7
7.0
3.33
7.0
3.24
3.7
8.3
3.7
4.0
0.90
0.47
0.93
0.41
0.93
0.44
0.00
1 .00
0.00
0.55
2.11
0.86
1.68
0.74
1.71
0.56
0.00
2.62
0.00
1.49

-------
                  TABLE  VI  -  REGENERATION  RESULTS, EFFECT OF TEMPERATURE
ro
Run
no.
virgin
spent
22
23
48
51
50
52
10
41
61
66
Bulk
density ,
g/cc
0.468
0.583
0.510
0.480
0.503
0.471
0.500
0.469
0.53
0.476
0.513
0.477
Weight
decrease ,
	
13.31 -,Q 7
7.2J
15. 2J. 21.4
15.8 01 o
7.2" 21'8
lO-f-20.5
"1:1 ^
Bulk
vol ume
decrease ,
	
r.2~ 2'2
1.8" or
0.7]
1.6 op
1.2 2'8
1.0 -, 5
0.5" *D
- 3 0
0.7; J'U
Particle
vol ume
decrease ,
%
	
0.5^ 0 0
1.7
1.9" , R
1.9 lj'b
2.0". 4 7
2.7 4>/
0.1 o o
2.2.~2'3
o!e ^2<5
Pore
vol ume ,
cc/cc
0.650
0.500
0.600
0.630
0.617
0.649
0.623
0.644
0.600
0.636
0.605
0.638
Iodine
number ,
mg/g
1090
630
950
950
910
910
910
900
980
950
880
940
Mol asses
number
250
190
	
260
245
260
250
204
263
220
240

-------
The reason for the difference in the results given above is
due to an appreciable amount of low temperature activation
during the baking step for Runs 48-51 and 50-52.  At tem-
peratures above 1500°F, the rate of activation increases
rapidly with temperature.  But increasing the temperature
at the end of the baking step by 50°F for Runs 48-51 and
50-52 caused a considerable amount of low temperature ac-
tivation to occur at 1600°F.  Low temperature activation
appears to be detrimental to the recovery of iodine number,
i.e., regeneration of micropores to original state.

To support the above interpretation, regeneration Run 60-61
was run at the 1700°F level but nitrogen sweep gas was used
during baking to avoid any activation.  In this case the
iodine number was 940 mg/g; up considerably from the 900
mg/g as obtained for Run 50, also activated at the 1700°F
level.

Regeneration Run 10-41 was performed under a similar tem-
perature schedule as Run 22-23 except that N~ sweep was used
during baking.  For this run, the iodine numBer of the re-
generated carbon was 950 mg/g, the same as for Run 22-23.
The conclusion is that at temperatures of 1550°F and below,
C02 and HoO vapor are relatively inert toward the basic car-
bon structure.

In regard to the temperature during the activating step, it
is desirable to maintain the temperature at or close to the
1700°F level.  In those runs where the temperature profile
was from 1500° to 1700°F, most of the activation occurs in
the third section of the regenerator where the temperature
is 1700°F.  When the gas reached sections II and III at the
lower temperatures, the COo and H20 contents are partially
depleted, hence reaction rates are considerably slower.

Steam and C02 Activations - Task 6 results indicated that
when C02 was utilized to any appreciable extent in the ac-
tivation step, the iodine number decreased.  To verify this
trend, regenerations were then conducted in which the acti-
vation step was solely with C02-N2 mixture or steam-N2 mix-
ture.  The C02-N2 activation was also carried out on a HC1
pretreated carbon.

The conditions of regeneration and results are presented in
Tables VII and VIII.  Included for comparison are data of
Run 61-66, which was activated with a steam-C02-N2 mixture
but the C02 did not contribute to the activation.  Figure 5
presents graphically the results of Table VIII and results
obtained in the Task 6 studies.  The iodine number of the
C02-activated, nonleached carbon is 840 mg/g compared to
                            27

-------
                TABLE VII - REGENERATION CONDITIONS, STEAM AND CO? ACTIVATION
N3
CO
      Run no.     Regenerating step
62           baking
67        Run 62 activated

64           baking
68        Run 64 activated

61           baking
66        Run 61 activated

          HC1 leach
95           baking
96        Run 95 activated
                                            Temperature, °F      Gas input, ft3/hr (stp)
1
1700
1700
1700
1700
1700
1700
850
1550
2_
1700
1700
1700
1700
1700
1700
1350
1650
3.
1600
1600
1600
1600
1600
1600
1550
1700
N.2
3.7
4.0
3.7
2.0
3.7
4.0
7.0
14.0
C0.2
0.00
0.00
0.00
2.91
0.00
0.55
0.93
4.00
H.20
0.00
2.81
0.00
0.00
0.00
1.49
1.71

-------
                TABLE VIII - REGENERATION  RESULTS,  STEAM  AND  C0   ACTIVATION
10
     Run no,
virgin
spent

 62
 67

 64
 68

 61
 66

1 each
 95
 96

Bulk Wei
Bulk Particle
ght volume volume Pore
density, decrease, decrease, decrease, volume,
g/cc %
0.468
0.583
0.518 13.3J
0.467 11. Ij
0.516 13.l!
0.478 9.1;
0.513 13. ll
0.477 8.0
0.555 6.3"
0.498 11.7
0.486 4.6
01 of r r- / r r
h io L L / L L
0.650
0.500
?o a 2.4 _ o 7 2.6 ,0 0.603
"'y 1.3 6ml 1.2, 0.647
21 0 2-2- 3.2 2-3- 3.7 °-604
1.0 1.4: 0.645
?n 1 2-3l 3 0 ]-9- ? 5 °'605
- 20.1 Q>7^ 3.0 Q>6j 2.5 Q>638
1.21 2.6] 0.528
-21.2 1 .7r 5.0 0.3- 4.8 0.624
2.1 1.9j 0.640

Iodine
number,
mg/g
1090
630
880
940
900
840
880
940
_ _ _
900
960
                                                                              Molasses
                                                                               number
250
190

220
260

220
250

220
240
                                                                                       250
                                                                                       240

-------
   1100
en

en
E
S-
01
JD
EE
T3
O
   1000
900
    800
              0
                        © MCI  leach, 1550° to 1700°F act.

                  Runs   Qnonleach, 1550° to 1700°F act.
                  36 &
                  79     Q nonleach, 1700° to 1600°F act.
             0.0                0.5                1.0
                   Mol C02/mol  carbon oxidized
    FIGURE 5 - EFFECT  OF C02 ACTIVATION ON IODINE  NUMBER
                          30

-------
940 mg/g for the steam activated.  For the HC1  pretreated
carbon, the iodine number of C02-activated carbon is 960
mg/g compared to 1040 mg/g for the steam activated.   A 100
unit drop in iodine number can be expected when activation
is solely by C02.

Study of Particle Volume Decrease During Baking - The carbon
loss during regeneration is measured by the particle volume
decrease.  It occurs during both the baking and activating
steps.  The volume decrease during the activating step is
understandable and expected since the activating gases would
also be expected to act on the basic carbon structures and,
thereby, reduce its volume.  The particle volume decrease,
however, was unexpected during the baking step  when  a nitro-
gen sweep gas is used.  A considerable effort was made to
determine whether the volume decrease was due to actual loss
of basic carbon structure or whether it was due to two other
possible effects, i.e., (1) expansion of carbon during ad-
sorption and contraction during  removal of adsorbate or (2)
decomposition and volatilization of  colloidal adsorbate
clinging to the exterior surface of  the carbon  particles.
A satisfactory answer to this question was not  found, al-
though some of the experimental  evidence appeared to favor
the colloidal adsorbate hypothesis.  Because this question
was not resolved, there was uncertainty as to the true car-
bon loss suffered during the regeneration.  Was it about 1.5%,
assuming true loss occurring only during the activating step,
or was it more like 3.0%, assuming true loss occurring dur-
ing both steps?

Vent gas analyses have indicated that, at least in part, the
particle volume decrease during  activation is due to loss of
fines of submicron diameter.  If fines formation is  the
mechanism by which particle volume decreases during baking,
it would be profitable to determine  whether there are con-
ditions by which it could be minimized.

One method that suggested itself was to carry out the baking
under extremely mild conditions.  This can be done in a fixed-
bed oven.  The carbon would not  be agitated as  in a rotary
tube to cause possible attrition  of weakened particle surface
layers.  The heatup time for the baking step can be very
gradual and over an extended time of at least 50 hr.  Evolu-
tion of gases would not be rapid enough to carry off any
fines.  To determine this point, four fixed-bed oven baking
runs were made.  The baking conditions and test results are
given in Tables IX and X.  Each  spent carbon was HC1 acid
leached prior to baking according to a standardized procedure
given in Task 4.
                           31

-------
TABLE IX - BAKING CONDITIONS  FINES  FORMATION  STUDY.  OSF  400





                      Temperature,  °F            Gas  input, ft  /hr  (stp)
Run no.

0)
_ — —
(2)
_ — —
(3)
80
— — —
(4)
92
93
94
97
Treatment
HC1 leach
Fixed bed
HC1 leach
Fixed bed
HC1 leach
Fixed bed
Rotary tube
HC1 leach
Fixed bed
Rotary tube
Rotary tube
Rotary tube
Rotary tube
1

60-^
« ~ *
60^
_ mm _
60 ->
850
_ — _
60-^
850
850
850
850
2

1550
_ — _
1550
_ _ _
1550
1350
_ — —
1550
1350
1350
1350
1350
3

in 48 h r
— — —
in 48 hr
_ — _
in 48 hr
1550
_ — _
in 48 hr
1550
1550
1550
1550
No COo HoO

1.25
_ _ .
1.25
„ .. _
1.25
3.7
— • —
1.25
3.7
3.7
3.7
3.7

-------
                  TABLE X -  BAKING  RESULTS  FINES  FORMATION  STUDY. OSF  400
Lo
Ru
n no.
OSF 400
HC1 leach




(1)

HC1 leach


HC


HC






(2)
1 leach
(3)
80
1 leach
(4)
92
93
94
97
Bulk
Bulk Weight volume
density, decrease, decrease,
q/CC % %
0
0

0

0

0
0
0
0
0
0
0
0

0
.583
.555 5.8]

.506 10. 8j
-i
.560 4.8

.507 11.1
.562 4.4~
.511 10.0
.510 0.8_
.552 6.1"
.505 9.9
.504 0.4_
.506 0.6
0.6
.502 0.1
i r r\ 'J •
- 1 K 1 1
1 O * \) n
t- •

ISA °-

IS. 4 ^
0.
-14.7 1.
0.
0.
-15.8 1.
0.
0.


8l, i
	 -f I
3JJ- '
"\
8 f\ r
— / n
8
6]
1 j-2.1
4
8
4 r 2.3
1 j
8

0.5
Particle
vol ume-
decrease ,
1 .

1.

1.

0.
1 .
0.
0.
1 .
0.
0.
0.
0.
1.
	

2j

6 in
— 1 LI
3
o"
2-1.6
4..
2~
3-1.9
4J
4
7
1
Po
vol
cc/
0.
0.

0.

0.

0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
re
ume ,
cc
501
527

617

520

614
521
598
608
519
622
618
621
616
617
Iodine
number,
mq/q
630

	

	

	
600
880
880
_ -. ..
930
930
850
920
900
Molasses
number
190

	

	

- — —
_ .. _
	
	
_ _ _
	
220
240
— — —
220

-------
In the fixed-bed Run (1), a considerable amount of ash was
found on the surface of the baked carbon particles, indi-
cating that oxygen from some source had reached the carbon.
This was unexpected since nitrogen was passed through the
oven during heat-up and cool-down.  On Run (2), the amount
of ash formed was considerably less, but still considered
too much for the purpose of the run.  On Runs (3) and (4),
panels of granular carbon, between stainless steel screens,
were placed around the OSF 400 carbon beds to shield out
external oxygen.  Also new stainless steel screens and trays
were used to eliminate oxidized surfaces which could re-
lease oxygen on  contact with hot  carbon.  On these two runs
no  ash appeared  on the baked carbon surface.

The  results of Runs (3) and (4), which are the most valid
ones, present conflicting evidence.  Particle volume de-
crease was  very  small, of the order of 0.3%, while the bulk
volume decrease  was much  larger,  i.e., 1.1% and 1.4%.  It
appears as  though the 0.3% decrease in volume of  the par-
ticles then permitted the particles to fit closer  together,
hence the larger decrease in bulk volume.

When the fixed-bed runs were then rebaked in the  rotary-
tube regenerator, again in an inert atmosphere, the bulk
volume and  particle volume decrease tended to be  equal and
of  the order of  0.4% over repeated baking runs, Runs 80  and
92  to 97.   A 0.4% to 0.8% weight  loss also occurred on each
successive  baking run.  Vent gas  analyses, Table  XI, in-
dicate that the  weight loss is essentially due to evolved
CO  and COo.  However, the amount  of oxygen in the ^ was not
sufficient  to account for the amount of CO and CO? evolved,
hence other sources of oxygen pick-up were investigated.

A possible  chemisorption  could occur when the baked carbon
is  exposed  to air during  handling and bulk volume measure-
ments.  On  each  baking run, the chemisorbed oxygen would
then be evolved  as CO and C02.  To test this hypothesis, the
product of  Run 93 was kept under  nitrogen and fed to Run 94.
In  this run, also, there  was the  usual weight and volume
loss, and evolution of CO and C02-  The conclusion arrived
at  was that the  oxidized  surface  of the rotary tube yielded
oxygen to the carbon.  This could mean that on every baking
run, whether in  inert atmosphere  or not, a 0.4% particle
volume decrease  occurs.   This loss is characteristic of  the
manner in which  the unit  is operated, i.e., on a  long con-
tinuous run this loss would not occur after the metal sur-
faces have  become reduced.

The  conclusion at this point of the study was that, on MCI
leached spent carbons, there is only a negligible loss of
carbon volume during baking.

                            34

-------
       TABLE XI - WEIGHT DECREASE DETERMINED FROM
                    VENT GAS ANALYSES
Run no.
92
93
94
97
Weight
decrease ,
g/hr
1 .05
1.38
1.44
0.82
Gas release rate,
mol /hr
CO
0.035
0.035
0.010
0.019
C02
0.010
0.051 (])
0.025
0.009
Weight of
gas released ,
g/hr
1.42
3.32 (])
1.38
0.93
(1)   This value appears to be in error
                          35

-------
The study was continued by considering the carbon losses in-
curred during the acid leach.  Table XII separates out the
pertinent data from Table X.  For Runs (2), (3) and (4), the
particle volume decrease is about 1.3% while the bulk volume
decrease is about 0.7%.  This is the reverse of the volume
decreases occurring during baking.  Apparently, the larger
particle volume decrease causes a very small change in pack-
ing efficiency of the particles.

If a further examination is made of HC1 leached-rotary tube
baked carbons (see Table XIII), the same pattern is observed
on bulk and particle volume decreases as in Table XII.  Here,
also, the particle volume decrease during baking of previous
HC1 leached spent carbons is of the order of 0.2%.  Acid
leach reduced the particle volume by about 2.0%.  These bak-
ing runs were made in a flue gas atmosphere, and since the
particle volume decrease is essentially the same as for the
fixed-oven runs, the flue-gas then must also act as an in-
ert gas.

A second conclusion, arrived at from this study, is that col-
loidal material from the waste water adheres to the exterior
surfaces of the carbon particles, thereby enlarging them.
The HC1 acid leach removed this colloidal material, hence
the particle volume decreases during leaching but with little
or no particle volume decrease during baking.

When nonleached spent carbon is baked, the colloidal mater-
ial is subject to heat treatment which will cause it to de-
compose and volatilize.  In reviewing the past runs, evidence
has been found which shows that the degree of decomposition
and volatilization varies with the conditions of baking.
These data are presented in Table XIV.

In the early part of the program, when runs up to and in-
cluding number 31 were made, Section I of the rotary tube
was heated to 750°F.  On these runs, the particle volume de-
crease was on the order of 0.5%.  At this temperature, liquid
water collected in the carbon influent end of the tube caus-
ing operational difficulties.  To correct this, the temper-
ature of Section I has since been operated at 800°F and
higher.  Except for Run 44, the particle volume decrease also
rose, now in the 1.4% to 3.3% range.  The explanation that
appears most reasonable is that when Section I is 750°F, the
carbon temperature rise is gradual as it proceeds down the
rotary tube into higher temperature sections.  Under a
gradual temperature rise, the colloidal matter stays fixed
to the carbon surface and then decomposes, leaving a deposit
of free carbon.  The free carbon deposit lessens the volume
loss from what would occur if all the colloid had been re-
moved.


                            36

-------
      TABLE XII - VOLUME DECREASE DURING HC1  LEACH
      	AND FIXED BED BAKING OF OSF 400
      Run no.
                                Volume decrease,^
Bulk
          Particle
Spent leached
Fixed bed baked, No.(1)

Spent leached
Fixed bed' baked, No.(2)

Spent leached
Fixed bed baked, No.(3)

Spent leached
Fixed bed baked, Mo.(4)

means, excluding No.(l)
2.3

0.8
1.8

o.el
     3
     J'
     2.6
i.i

0.8
1 .4
   r 1-7
2.2
1.6
0.3

i.o"
0.2
             - 1.9
             -1.2
                '  - 1  5
               0.3  ''^
     2.2
               1.5
                          37

-------
     TABLE XIII - VOLUME DECREASE DURING HC1  LEACH
           AND ROTARY TUBE BAKING OF OSF 400
                                 Volume decrease,
   Run no.                     Buik            Parti cTe


Spent leached                 O.ol 0 7           Loo
    45                        2-7J              J

Spent leached                 0.4] 9 r        1.9
    49                        2.1_j ^>;J       0.2r   '•1

Spent leached                 0.2
    63                        2.4j"

Spent leached                 O.ol , Q       1 . 7|_   -,  r
    78                        1.9J l>y      -0.27

Spent leached                 1.2  0 Q       2.6_   9  Q
    95                        1.7j'y       0.3^-y

        means                     2.5             2.4
                          38

-------
       TABLE XIV - VOLUME DECREASE OF NONLEACHED
       	    CARBONS DURING BAKING
Run no
   Section 1
Temperature,
. _y_glume decrease. %
BuikParticle
  12
  22
  29
  31

  37
  38
  44
  48
  50
  62
  64
  71
  72
     750
     650
     750
     750

     850
     850
     810
     820
     800
    1700
    1700
     800
     800
 0.6
 1 .0
 1.2
 0.8

 2.0
 2.0
 0.4
 1.8
 1.6
 2.4
 2.2
 1.8
 1.6
0.3
0.5
0.8
0.3

1.6
1.4
0.0
1.9
2.0
2.6
2.3
3.3
2.9
                          39

-------
When the temperature of Section I is raised to 800°F and high-
er, the increased temperature causes an increase in expulsion
rate of volatiles from within the pores and also in the rate
of volatilization of the colloidal matter.  The net result
is that a greater portion of the colloidal matter is removed
with the vent gases, and hence the larger decrease in particle
volume.

When these carbons are activated, the prior treatment does
not appear to greatly affect the volume decrease.  As shown
in Table XV, the mean particle volume decrease is about 1.8%.
If the particle volume decreases, during HC1 leach and bak-
ing is accepted as being due primarily to removal of col-
loidal matter from the exterior surface of the particles,
then the true carbon loss is closer to 1.8% rather than the
3.5% to 5.0% as previously reported.

Effect of Particle Size on Ease of Regeneration - A parameter
that needed to be investigated was the effect carbon mesh
size may have on the ease of regeneration.  It is known that
powdered carbons regenerate much more rapidly than coarse
granular carbons.  If the regeneration rate is appreciable
within the range of mesh sizes normal in commercial carbons,
it is evident that the small particles could be over-acti-
vated and thereby suffer excessive particle volume loss
while the large particles are still underactivated.  A large
difference in ease of activation can lead to a new investi-
gation in optimizing mesh size, contacting procedures and
regeneration.  The ease of regeneration of the open pore
structured Darco carbons has demonstrated that wide vari-
ations can exist amongst carbons.  On the Darco, complete
regeneration was always effected by the baking step alone.

This study was conducted primarily with West Virginia Pulp
and Paper Company spent carbon since it consisted of coarser
mesh sizes than the OSF 400.  The original 8 to 30 mesh was
sieved and sized to obtain four mesh size fractions, 8 to 12,
20 to 30, 30 to 60 and -60.  A powdered -100 mesh OSF 400
and a powdered -100 mesh Nuchar Aqua A were later added to
the study.  Prior to thermal regeneration, each carbon was
HC1 acid leached and dried according to procedure given in
the Task 4 section of this report.

The conditions of regeneration and results are given in
Tables XVI and XVII.  The results of the study indicate that
the size of the particle does not control the rate of re-
generation.  In the WVP&P series, no trend is indicated in
the iodine number.  The iodine number of the OSF 400, Run 85,
is in line with those of the baked acid leached granular
OSF 400 carbons.  However, decrease in particle size appears
                           40

-------
     TABLE XV - VOLUME DECREASE DURING ACTIVATION
                                     Volume decrease,  %
Run no.        Pretreatment         Bulk       Particle
29-30           HC1 leach            0.4          0.4
31-32           HC1 leach            1.4          2.6
37-39           HC1 leach            1.4          2.0
38-40           HC1 leach            1.3          1.5
48-51           HC1 leach            0.7          1.9
50-52           HC1 leach            1.2          2.7

                 means               1.1          1.8
12-16            none                0.9          1.5
22-23            none                1.2          1.7
29-30            none                0.4          0.4
31-32            none                1.4          2.6
37-39            none                1.4          2.0
38-40            none                1.3          1.5
44-46            none                1.4          2.5
45-47            none                1.1          1.8
49-53            none                0.7          1.7
63-69            none                0.7          1.9
78-79            none                0.9          2.0

                 means               1.0          1.8
                           41

-------
N3
                   TABLE XVI - REGENERATION CONDITIONS, EFFECT OF PARTICLE
                                  SIZE ON REGENERATION (1)
Run
no.
81
88
82
89
83
90
84
85
Carbon an
mes
WVP&P,
WVP&P ,
WVP&P,
WVP&P,
WVP&P,
WVP&P,
WVP&P,
OSF 400
h size
8 to 1
8 to 1
20 to
20 to
30 to
30 to
-60
d
Regenerating
step
2
2
30
30
60
60

, powdered
b a k i
Run 81
baki
Run 82
baki
Run 83
baki
baki
ng
act.
ng
act.
ng
act.
ng
ng
Temperature ,
1
800
1555
800
1555
800
1550
800
1700
2
1300
1650
1300
1650
1300
1650
1300
1700
°F G
3
1500
1700
1500
1700
1500
1700
1500
1600
as i
No
2.
3.
2.
3.
2.
3.
0.
3.
nput,ft3/hr (stp)
C02
0
5 0.43
0
5 0.43
0
5 0.43
5
0
H?U

1.13
_ _ _
1.14
_ _ _
1.25
	
_ _ _
                     -100

       87   Nuchar, powdered
                     -100
baking
830
1320
1500
2.0   0.18   0.13
            (1)  All spent carbons HC1 leached prior to regeneration

-------
TABLE XVII - REGENERATION RESULTS - EFFECT OF PARTICLE SIZE ON REGENERATION
Run
no.
Carbon
Spent WVP&P (8 to 30)
OSF 400
Spent Nuchar (-100)
Virgi
Virgi
Virgi
81
88
82
89
83
90
84
85
87
n WVP&P (8 to 30)
n F 400
n Nuchar (-100)
WVP&P
8 to 12
WVP&P
20 to 30
WVP&P
30 to 60
WVP&P -60
OSF 400 -60
Nuchar, aqua A
Bulk Particle
Bulk volume volume Weight Iodine
density, decrease, decrease, decrease, number,
g/cc % % % mg/g
0.597
0.583
0.304
0.504
0.469
0.320
0. 538 1 .6J. 2 3 0.4
0.519 0.7J ' 0.2.
630
390
1070
1090
760
-0.6 7.3l_n.7 1000
4.7J 1050, 1070
Molasses
number
180
190
250
290

0.532 2. it , ? 0.6~1 o i 8.2l 1? c 950
0.519 1.1J 1.5J 4.6J 1010, 1030
0.478 1 .8J_ 4 n 0.7
0.457 2.2J ' 1.3.
0.630 9.6
0.694 5.7
0.294 9.0
2 n 9-3"l 14 9 1Q00
6.1J ' 1060
13.1 1000
12.3 870, 860
15.1 770

230

270
       -100

-------
to accelerate the carbon loss as measured by bulk volume and
particle volume decreases.

A more important factor in  rate of regeneration may be the
pore size distribution.  Granular carbons such as the Darco,
of the type used in tertiary treatment, regenerates more
readily than WVP&P and Filtrasorb 400, granular or powdered.
Pore size distribution curves in Figure 6 of F 400 and the
Darco show differences that can explain the greater ease with
which the Darco can be regenerated.   The Darco has an abun-
dance of pores in the 30A to 1000& diameter range which offer
passage ways for the gases  to diffuse out of or into the
small pores.  Also a greater portion of the Darco surface
area is in larger pores than that of F 400.  In F 400, the
number of passage ways or pores of 30A to 1000A diameter are
considerably less, hence gas diffusion out of or into the
20^ pores is then slower.  The major portion of the surface
area of F 400 is in pores less than 20A diameter, which can
hold the adsorbate molecules more strongly.

The experimental evidence indicates that pores of 1000& to
100.000A diameter do not appreciably slow the rate of gas
diffusion while pores below 1000A, particularly in the 30A
region, may be rate controlling.  When granular F 400 or
Darco are pulverized, some  of 1000A to 100,OOOA diameter pore
volume becomes interparticulate volume.  Hence pulverizing
does not appreciably change the gas diffusion rates in the
particles.  Nuchar Aqua A is expected to have a pore size
distribution similar to Darco.  Nuchar Aqua A was easily
regenerated to the virgin carbon adsorptive capacity at 15%
weight decrease and 9.0% vulk volume decrease.

The evidence for the large  pores becoming interparticulate
volume is the bulk density  change that occurred when OSF 400
was sized.  In the 14 to 40 mesh granular form, the bulk den-
sity is about 0.560 g/cc, but after pulverizing, the bulk
density increased to 0.750  g/cc.  The bulk density of pow-
dered carbon was based on the minimum volume obtained by
tapping a graduated cylinder containing a weighed amount of
the carbon until the volume stopped decreasing.

General Cone! usions - Aside from the conclusions drawn fro;n
thevarious parameter studies, one important result was not
explicitly stated although  quite apparent, i.e.,, that the
iodine numbers of the regenerated carbons were all consider-
ably lower than the iodine  number of the virgin carbon given
at the top of each table.  The properties of the virgin car-
bon were assumed to be initially those of the OSF 400, hence,
were the goals to be attained when complete regeneration was
accomplished.  This assumption was based on the results at-
                           44

-------
            1.20
Ul
                10     30       100              1000      0      10,000        100,000
                                           Pore  diameter, A

                 FIGURE  6  -  PORE  SIZE  DISTRIBUTION, DARCO RUN 58 AND FILTRASORB 400

-------
tained on the OSF 400 when regenerated with caustic and then
baked in Task 4.   The parameter study of Task 1  definitely
indicates that manipulations of the temperature, gas compo-
sition and gas input rates during thermal regeneration do
not create the necessary conditions for a satisfactory re-
generation.

Task 2 - Adsorption-Regeneration Cycle Studies

After repeated regenerations, a gradual change in pore struc-
ture can be  expected to occur which will ultimately cause
disintegration of the carbon particles and cause the carbon
to lose its  adsorptive capacity.  Inorganic compounds are
also expected to accumulate in the pores and thereby further
diminish adsorptive capacity.

To study these possible effects, an adsorption-regeneration
cyclic study was carried out with the cooperation of the
Pilot Plant Pomona.  Starting with 2500 cc bulk  volume of
Filtrasorb 400, 14 by 40 mesh carbon, the adsorption phase
was carried out at the Pomona Plant, with secondary treated
waste water, and the regenerating and testing at the MSA Re-
search Corporation laboratory at Evans City, Pennsylvania.
Carbon loss  occurring on each cycle was made up  with virgin
Filtrasorb 400.

The fact that the properties of the initial carbon were
known at the outset of the cyclic studies avoided one of the
problems that had caused considerable uncertainty on all
previous regeneration runs.

It had been the original intent to perform ten cycles, but
after the third one it was apparent that the carbon was de-
generating much faster than had been expected.  This trend
is shown by the decrease in iodine number and increase in
molasses number, as below:

    Cycle     Iodine number, mq/g      Molasses  number

    initial            1090                   250

    1st               1040                   310

    2nd                935                   290

    3rd                940                   355

It was also noted that the activating step on the second and
third cycles were greatly accelerated.  To prevent over acti-
vation, it was necessary to greatly reduce the activating
gas input rate-  Below are input rates which show the large
                           46

-------
decrease necessary on the second and third cycles.

              Activating gas Input rate. ft3/hr (stp)
     C.y c 1 e         jjgCO^         F£o

      1st         3.6         0.36        0.99

      2nd         1.27        0.17        0.18

      3rd         1.27        0.17        0.12

For each cycle the other activating conditions were essen-
tially the same and, in each case, the final regenerated
carbon density was 0.468 g/cc, the bulk density of the
virgin carbon.

The overall carbon loss for the first cycle was 5.4%, 2.8%
for the second, and 1.2% for the third, as measured by de-
crease in particle volume.  The 5.4% loss is larger than
normally experienced while the 2.8% for the second cycle
i s normal.

In the light of improved results on thermally regenerating
HC1 leached spent carbon, it became apparent that leaching
could have greatly reduced the degeneration rate.  The
accelerated degeneration rate appears to be associated with
the increased ash contents on each successive cycle.  Data
on ash increase are as follows:

                   Cycle          Ash,%

                   virgin          5.7

                    1st            7.6

                    2nd            8.6

                    3rd            9.5

Estimates were made to determine whether the decrease in
iodine number was accompanied by a decrease in adsorption
of waste water contaminants.  The calculations were made
using the formula

(Density of spent carbon)-(denslty of reg carbon) Y lnn
 	(density of reg carbon)              A IUU =
                                        % adsorbate.
For the two completed cycles and the adsorption phase
third, the calculated results are as follows:
                            47

-------
                Cycle            %, Adsorbate

                 1st                 17.1

                 2nd                 14.1

                 3rd                 15.0

A decrease in adsorptive capacity may be indicated, although
the decrease may also be due to fluctuations in adsorbate
concentration of the secondary treated waste water.

The regenerations were carried out under conditions that can
be attained in a multiple hearth furnace.  Further details
on the cyclic studies are presented in Tables XVIII and XIX.

Task 3 - Determine Feasibility of Low Grade Carbons as
         Make Up

The procedure for this task was identical to Task 2 except
that a lower grade Filtrasorb 100 was used as make-up carbon
This study was based on the premise that a lower activity
carbon would be upgraded during regeneration and, because
of the price difference between Filtrasorb 400 and 100,
would lead to an economic advantage.  The 1969 prices were
28 l/2<£/lb for Fi 1 trasor b 400 and 20 l/2<£/lb for Filtrasorb
100.   On the two cycles, the pattern was the same as  in
Task 3.   Any benefit that might have accrued from use of
lower grade make-up carbon was obscured by the rapid  degen-
eration.  Test results showing the degeneration of the car-
bon are  given below:

     Cycle    Iodine number, mg/g     Molasses number

     Initial         1090                   250

      1st            1040                   310

      2nd             920                   310

As in Task 2, the activating gas input had to be decreased
considerably on the second cycle to prevent over activating,
as i ndi cated below:

                Activating gas input rate, ft3/hr (stp)
     Cycle           N_2         C£?         -2—

      1st            2.5        0.32        0.72

      2nd            1.27       0.17        0.21
                           48

-------
TABLE XVIII - REGENERATING CONDITIONS,  TASKS  2  AND  3
Run
no.
24
25
54
56
26
27
55
57
91
98
Task
2
2
2
2
3
3
3
3
2
2
Cycle
1
1
2
2
1
1
2
2
3
3
Regenerati ng
step
bake
activation
bake
activation
bake
activation
bake
activation
bake
activation
Temperature
1
770
1550
825
1550
760
1550
840
1550
850
1550
2
1350
1650
1350
1650
1340
1660
1350
1650
1350
1650
, °F
3
1560
1700
1550
1700
1560
1700
1550
1700
1550
1700
Gas input, ft3/h
N2
7.0
3.6
7.0
1.27
7.0
2.5
7.0
1.27
7.0
1.27
COo
0.90
0.36
0.93
0.17
0.90
0.32
0.93
0.17
0.90
0.17
r (stp
H20
1.47
0.99
1 .68
0.18
1.57
0.72
1.57
0.21
1 .43
0.12

-------
                 TABLE  XIX -  RESULTS  OF REGENERATIONS. TASKS  2 AND  3
Run
Task
virgin
spent
24
25
reg
spent
54
56
reg
spent
91
98
virgin
spent
26
27
reg
spent
55
57
reg
spent
2
2
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
Regenerating
step
to Pomona
adsorption
bake
acti vati on
to Pomona
adsorpti on
bake
acti vation
to Pomona
adsorpti on
bake
activation
to Pomona
adsorption
bake
activation
to Pomona
adsorption
bake
acti vati on
to Pomona
adsorpti on
Bulk
density,
g/cc
0.468
0.548
0.493
0.468
0.474
0.541
0.480
0.469
0.472
0.549
0.482
0.473
0.468
0.546
0.484
0.464
0.476
0.538
0.481
0.469
0.478
0.544
                                          + 13.0"
                                          -12.8
                                          - 4.2
                                                -4.0
                                          -5.6
                                          + 14.6
                                          -15.0 - -5.8
                                          - 3.3
                                          + 11.8
                                          -11 .7
                                          - 4.1
                                         h -5.1
Bulk Particle
volume volume Iodine
change, change, number,
% % mg/g
3.36J -1.12]
2.52T 7'° ---1
1.38.1 ---J
0.81"
1 .73
1.12.
o.r-
1.0
1.1 .
+ 0.84
- -3.6 -3.33
-0.33.
-oT
- -2.6 -0.0
-1.2 J
0.60J -0.72"
1.81k -3.9 -0.73
1.54J -0.86.
0.97
1.31
1 .51
-1.98"
- -3.8 -1 .24
-0.33
1090
- -5-4 III
1040

- -2.8 915
935

- -1 .9 834
940
1090
L -2.3 HI
1 1040
...
- -3.5
920
Molasses
number
250
310
- - -
270
290
	
280
355
250
310
- - -

310
                                             + 13.2
                                                 -0.40
-1 .48

-------
For_this task, the conclusions are the same as for Task 2.
It is felt that removal of part of the inorganic compounds
by HC1 leach would have greatly reduced the rate of de-
gradation.  The ash content increased in the same manner as
in the Task 2 carbons.
                   Cycle

                   Initial

                    1st

                    2nd
Ash,%

 5.7

 7.9

 8.7
Overall carbon  loss on the first cycle was 2.3% and on the
second 3.5%.

The estimated adsorbate on the carbon on each cycle was as
follows :
                   Cycle

                    1st
% Adsorbate

    16.7
                    2nd

                    3rd
    13.0

    13.8
Further details on Task 3 studies are also presented in
Tables XVIII and XIX.

Task 2A and 2B - Cyclic Adsorption-Regeneration Studies

In view of the improved thermal regeneration of HC1 leached
carbons, the cyclic studies were resumed during the six
months extension period.  For these studies, two new 2500 cc
batches of virgin Filtrasorb 400 were prepared.  Batch 2A
was as-received except for sieving out the -40 mesh fraction
and batch 2B was leached in succession by HC1 acid, pure
water, hot caustic solution and pure water.  This was done
to remove all of the F203, CaO, MgO, Na20 and K2Q and part
of the Si02 and A1203, see Task 4 on HC1 leach.  The final
ash content was 3.5% compared to 5.7% for the virgin carbon.
These carbons were spent at Pomona as was done in Task 2 and
then regenerated at MSA Research laboratories.  Prior to the
thermal regeneration, each batch of spent carbon was HC1
leached.
                             51

-------
The regeneration conditions and results of these two regen-
erations are presented in Tables XX and XXI.

For both  regenerated carbons, the iodine numbers are higher
than those of the original by a considerable margin although
the final bulk density in each case matched the original.
This is the first time this had occurred in the regeneration
program.

The molasses numbers did not change significantly which is
as desired.

The carbon losses are larger than had been expected.  The
actual  regeneration losses are 4.3% for 2A and 4.4% for 2B
as measured by particle volume change.  For 2B, the carbon
loss during the adsorption phase was  1.3%, by particle
volume  change, while there was an actual increase of 0.52%
in bulk volume.  For 2A, the carbon loss during the adsorp-
tion phase was very large, being 7.2% by bulk volume change
and 8.5% by particle volume change.

The bulk and particle volume changes  occurring during each
step of the regenerations, for these  two carbons, do not fit
the general pattern as established in Tables XIII and XV.
The bulk volume changes of 2B come closest to fitting the
general pattern.  In view of the studies made on particle
volume  decrease during baking in a previous section, most of
the apparent baking losses should not be included in the
overall regeneration loss.  If 2.4% is allowed for  the ap-
parent  baking loss, the overall adsorption-regeneration cycle
loss for 2B is 3.3%.  This percentage figure agrees well with
the 3.7% overall weight loss.

Ash analyses on 2A  and 2B after regeneration were 4.7% and
3.5%.

This was only the first adsorption-regeneration cycle but
the results look promising in that original activity is re-
covered without excess carbon loss.

Task 4  - Chemical Oxidation and Solvent Extraction

H20? Pretreatment - Two pretreatment  experiments were per-
formed  with 3% H202.  In Run 9, Tables XXII and XXIII, 2500 cc
of dried OSF 400 were wetted with 550 ml of 3% H202 and then
baked  in the regenerator.  This treatment produced  no ob-
servable effects on the baked carbon.  In Run 11, 2500 cc of
dried  OSF 400 were  leached with 4800  ml of 3% H20?  in Buchner
funnels and the leached carbon dried  and baked.  The leach-
ing reduced the bulk density from 0.583 g/cc to 0.569 g/cc.
                            52

-------
   TABLE XX - REGENERATION CONDITIONS, TASKS 2A AND 2B,  FIRST CYCLE


                                        Temperature,  °F     Gas input.  ft3/hr (stp)
Run             Regeneration step      1        2       3      N9     CO?     H_?£
t 2A
leached
03
04
it 2B
1 eached
02
05
dried
	
baked
activated
dried
	
baked
activated
	 — 	 — C. 	 C- —t- 	
	 	 	 	 	 	
860 1360 1550 7.00 0.90 1.58
1550 1650 1700 2.19 0.37 1.12
	 	 	 	 	 	
	 	 	 	 	 	
860 1360 1550 7.00 0.90 1.39
1550 1650 1700 2.19 0.37 0.81

-------
              TABLE  XXI  -  REGENERATION  RESULTS, TASKS  2A  AND  2B,  FIRST  CYCLE
Ln
Run
Initial 2 A
Spent 2A
HC1 leached
103
104
Initial 2B
Spent 2B
HC1 leached
102
105
Bulk
d e n s i ty ,
g/cc
0.477
0.540
0.527
0.492
0.474
0.473
0.546
0.533
0.494
0.475
Bulk Particle
Weight volume volume Pore
change, change, change, volumes
% % % cc/cc
0.659
+ 5.0] -7.2j -8.5l 0.568
-3.1_ n R -0.6L n 1 -0.5L_i? Q 0.592
-8.8f "'8 -2.3 ' -1.4^ ^ 0.631
-4.8J -1.3] -2.8 0.649
0.669
+ 15.9 +0.51 -1.3J 0.551
-3.7 -37 -1 .4i 41 -0.8_ c 7 0.574
-8.6 " -1.3 ' -3.0 ' 0.631
-5.6J -1.8J -0.7 0.649
Iodine
number>
mg/g
1105
1150
1065
1150
Molasses
number
290
290
235
300

-------
             TABLE  XXII  -  REGENERATION  CONDITIONS.  CHEMICALLY  TREATED  OSF  400
         Run
         no.
Ui
Ln
          8
         14
          9
         15
         11
         17
         18
         20
         21
Regenerating step,
special treatment
baker feed rewetted
with 550 ml H20 per
2500 cc carbon
Run 8 activated

baker feed wetted
with 550 ml 3% H202
per 2500 cc carbon
Run 9 activated
              Temperature,  °F
baker feed leached
with 4800 ml 3%
H202 per 2500 cc
carbon
Run 11 activated
500 cc carbon
with caustic,
Run 18 baked
Run 20, second
leached
HC1 ,  etc.

 bake
1
980
1550
1000
1550
860
1700
485
550
2
1270
1650
1250
1650
1250
1700
1230
1240
I
1550
1700
1550
1700
1550
1600
1550
1550
Gas
*2
7.9
2.7
7.9
2.7
7.9
2.1
2.0
2.0
input, •
cp2
1.0
0.34
1.0
0.34
1.0
0.27
0.00
0.00
Ft3/hr
02
0.0
0.00
0.0
0.02
0.0
0.00
0.00
0.00
(stp)
HzP-
1.60
0.47
1.32
0.72
1.50
0.39
0.00
0.00

-------
               TABLE XXIII  -  REGENERATION RESULTS,  CHEMICALLY TREATED OSF 400
Ul
Run no.
virgin
spent
8
14
9
15
11
17
18
20
21
Bulk
Bulk Weight volume
density, decrease, decrease,
g/cc % %
0.469
0.583
0.507 13. V
0.482 6.5_
0.507 13.5
0.477 7.1J
0.504 14.5]
0.480 5.9J
0.481 18.6"
0.467 5.5
0.460 0.9
	 	
- 18.8 0 • 5~L 1 . 6
l.lj
-i
-197 0* °\- 1 9
iy./ K1j i.y
1Q fi °'9l 1 7
- 19.6 Q>8J. 1.7
2.4"
- 21.9 1.8 - 4.9
0.7.
Parti cl e
volume
decrease ,
	
0.9] 2.8
1.9J
0.81 , 9
l.lj '
0.9"1 , ,
2 . 2f '
6.4'
	 8.9
2.7
Pore
vol ume ,
cc/cc
0.650
0.500
0.607
0.629
0.611
0.629
0.610
0.629
0.616
0.634
lodi ne
number,
mg/g
1090
630
940
935
940
920
960
950
910
1030
Mo 1 asses
number
250
190
212
229
211
237
213
251
260

-------
On baking, the bulk density reduced to 0.504 g/cc, which is
slightly lower than the 0.507 g/cc for untreated baked car-
bons.  The iodine numbers of both carbons after activation
were at the same level as for the unpretreated carbons, Run
14 as example, hence no benefit was derived.  Additional
Ho02 treatment would not create more favorable results since
trie \\2$2 cost already for the above leaching pretreatment is
7.2<£/Tb carbon.

Caustic Leach - Three caustic leach runs were performed to
determine the economic feasibility of caustic leaching as
an alternate method for spent carbon regeneration.

The  first leach run, designated as Run 18, was exploratory
to determine the critical parameters.  Concentration, flow
rate, and temperature were  varied.  Methanol, caustic-methanol
solution,, and periodic pure water and HC1 acid leach were
also tried.  Attempts were  made to follow the effectiveness
of the treatment by grading the color density of the filtrate,
but this method proved to be unreliable since the color kept
changing in tone.  The only trustworthy method of monitoring
was by periodic bulk density determinations.

Table XXIV gives a chronological record of the carbon treat-
ment.  The initial volume and weight were 500 cc and 292 g,
respectively.  About 350 1  of the various solutions were
passed through the column before the filtrate became essen-
tially colorless.  At this  point the bulk density was 0.481
g/cc, still above the 0.468 g/cc bulk density of virgin
Filtrasorb 400.

The leached carbon was then baked twice in a nitrogen at-
mosphere to volatilize as much of the adsorbate as possible.
The baking treatment brought the density down to 0.460 g/cc
and iodine and molasses numbers to high values of 1030 mg/g
and 260, respectively.  These test results are fairly close
to the properties of the virgin carbon received from Pomona
on December 9, 1968.  On this basis it was assumed that the
spent carbon used in most of the regeneration runs of this
program had initially the properties of this virgin carbon.

During the leaching, it was noted that if the carbon bed was
allowed to boil, even gently, by raising the temperature to
100°C, the filtrate started to carry carbon fines.  The
movement of the carbon bed  may have caused abrasion of the
carbon particles.  The particle volume decrease during the
leach was 6.4%.  During baking, it was 2.7%, which is high
compared to results on previous OSF 400 baking runs.  The
higher particle volume decrease here may be another indi-
cation of generation of fines during the leaching process.
                           57

-------
                    TABLE  XXIV  -  CAUSTIC  SOLUTION  AND  METHANOL  LEACH  OF
                             ONCE-SPENT  FILTRASORB 400.  RUN  18
CO
Vol ume
of
leach
liq, 1
0
46
9
3
6
6
9
6
3
12
3

Fl ow


rate* Type
1/hr liquid
2.5-3.
2.5
2.5
2.5
2.5
2.5
2.5
2.5
4.0
4.0
0 5.0% NaOH
1 .0% NaOH
Water
0.25% NaOH
0.50% NaOH
1.0% NaOH
5.0% NaOH
0.25% NaOH
0.10% NaOH
Water


Temp, oH of
°C filtrate
...
50-80

1






Above leaching done in Buchner funnel
Below leaching continued in. heated column, uni
1
4
11
50
3
4
12
4
1
10
4
1
12
1
32
2
2
1
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
2.0
2.0
2.0
2.0
1:8
0.10% NaOH
0.10% NaOH
0.05% NaOH
0.10% NaOH
1.0% NaOH
0.10% NaOH
Water
Methanol
1 .0% NaOH in
methanol
Methanol
Water
1.3% HC1
Water
0.10% NaOH
0.10% NaOH
0.10% NaOH
1.3% HC1
0.10% NaOH
0.10% NaOH
95-100
11


i

10.7
60-65 8.8
10.6
95-100
80
95-100 5.3
95-100


80
95-100
Bulk Bulk
den of volume
Color of carbon, decrease,
filtrate g/cc %
0.584
Dark yel low







Light yellow 0.529 0.0
t shown in Figure
100 color unit
60
17
3
4
2.5
0.513 1.8
Light pink
Light yellow
Light pink 0.509 1 .8
Light yellow
Green
Colorless 0.492 2.2
70
11
..
Col orless
30

-------
In the second caustic leach run, Run 19, the caustic con-
centration was kept very low, at 0.025% and 0.013%.  The
carbon bed was kept at near 100°C, but actual boiling or
movement of the bed was avoided as much as possible.  At
intervals, the carbon bed was leached with dilute HC1.
About 550 1 of caustic and HC1 solutions were passed through
the bed, but filtrate still had low level of color.  After
450 1 of solution had been passed through the bed, the car-
bon was taken out, dried and the bulk density measured.   No
change had occurred in bulk volume (still at 500 cc) but
the density was still high, 0.576 g/cc.

Figure 7 shows the decrease in bulk density as function  of
amount of solution passed through the carbon for each Run
18 and 19.  In general, the concentrations of caustic were
5- to 10-fold greater in Run 18 than Run 19.  The higher
concentration required less solution but reduction in water
required was considerably less than 5- to 10-fold.

A third run was made with 25% caustic soda and HC1 solution
treatments in a batchwise process.  Each batch treatment
consisted of a routine of caustic soda leach, water leach,
HC1 leach and water leach; the routine was done twice.  In
the caustic leach, 235 g (regenerated weight) of spent car-
bon was placed in a flask with a caustic solution and heated
at boiling point temperature for one hour.  The caustic
solution was filtered from the carbon in a Buchner funnel
and the carbon, on the funnel, then leached with six liters
of hot pure water, followed by one liter of hot 6% HC1
solution and finally with four liters of hot pure water.
Each batch treatment required overall 600 g caustic soda,
30 cc concentrated HC1 and 24 liters of water.

The results of this run are given in Figure 8, which show
the change in bulk density after each batch treatment.  At
the fifth treatment the effectiveness of the treatment was
very small.  An extrapolation of the curve indicates that
inordinately large quantities of caustic would be required
to bring the bulk density down to 0.468 g/cc.  At ten-batch
treatment, the amount of caustic required would cost 25<£/lb
of carbon.  If recovery of caustic, for reuse, were possible,
the recovery process would have to be better than 98% ef-
ficient to avoid excessive material cost.

On the basis of the three caustic soda and HC1 leach runs of
widely different agent concentrations, the conclusion is that
caustic leaching is not competitive with thermal regeneration,
Inordinately large amounts of water would have to be used to
regenerate the carbon to initial activity.  Over 8% of the
water purified per pound of carbon would be reused for leach-
                            59

-------
  0.60 i—
  0.55
o
o
"•-.
en

 #t

>>
j«:


3
03
  0.50
  0.46
               50
                       TOO
150      200      250
     Gal/lb carbon
300
350
                                                                              400
FIGURE 7 - OSF 400 LEACHED WITH LOW CONCENTRATION SOLUTIONS OF CAUSTIC  SODA  AND  HC1,
  AMD PURE WATER; BULK DENSITY AS FUNCTION OF AMOUNT LIQUID PASSED THROUGH CARBON

-------
  0.60
o
o
en
  0.55
O)
•a
_i^

3
   0.50
   0.46
1
0
1
1 2 3
1 ! 1
4
Number
of
5
times
1
6
each
ed
7 8

9 10
 FIGURE  8 -
OSF 400 LEACHED WITH HIGH CONCENTRATION CAUSTIC SODA SOLUTION
  AND PURE WATER IN BATCH TYPE PROCESS; BULK DENSITY AS
          FUNCTION OF NUMBER OF BATCH TREATMENTS
AND HC1

-------
ing purposes.   The caustic and HC1 acid costs would also be
excessive.   For Run 18, more than 0.84 Ib of 76% caustic
and 0.15 Ib of 37% HC1 acid would be required.   At 5.35<£/lb
for the caustic soda and 1.95<£/lb for the acid, the agent
costs would be more than 4.8<£/lb of regenerated carbon.  For
Run 19, the agent costs would be considerably over 3.3<£/lb
of carbon before regeneration could be effected to initial
properties.

Of academic interest, it was observed that the leaching ef-
ficiency of the caustic solution increases with increased
temperature.  Periodic HC1 acid treatments appear to accel-
erate the caustic leach.

NaOCl Leach - Two short leaching runs were made with NaOCl
solution in the leaching column with 500 cc quantities of
OSF 400.  The first run was with 35 1 of 1.0% NaOCl solution
at about 90°C and 1.0 1/hr flow rate.  The carbon density
decreased from 0.580 to 0.556 g/cc, but bulk volume decreased
from 500 to 452 cc.  The filtrate was black from finely sus-
pended carbon.

In the second run, 40 1 of 0.25% NaOCl solution were run
through the column.  In this case also, there was consider-
able disintegration of the carbon.  Density decreases from
0.580 to 0.550 g/cc and volume from 500 to 420 cc.

Because of the high carbon losses, NaOCl was judged an un-
satisfactory regenerating agent.

CC14 Extraction - Fifty cc of OSF 400 were extracted in a
300 ml Soxhlet extractor for 40 turnovers of the CCl^.  The
amount of CCl^ used in the extractor was 210 ml.  After the
extraction, the treated carbon was air-oven dried at 150°C
for several hours and again heated under vacuum at 110°C in
preparing samples for CC1* adsorptive capacity test.  The
CC1* adsorptive capacity, under static vacuum test, was 36%
compared to 38% for the untreated OSF 400.  Estimated iodine
numbers based on the CC1. capacity tests were 550 and 640 mg/g
for the treated and untreated carbons, respectively.

No benefit was derived from the CCl^  extraction.

HC1 Leach Pretreatment, Effect on Iodine Number - HC1 leach
was first conceived as a pretreatment to assist in the re-
moval of adsorbate during the baking step.  The rationali-
zation was  that organometal1ic compounds of Fe, Ca and Na
adsorbed from waste water would resist volatilization and
thereby leave a larger carbon deposit.  HC1 leached OSF 400
carbon, however, showed no greater adsorbate loss than un-
                            62

-------
leached carbons but a large increase in iodine number oc-
curred after the activating step.  It is now believed that
in_the unleached carbon, Fe203, CaO and Na20 catalyzed the
oxidation of the base carbon structure and thereby lowered
the iodine numbers and increased the molasses numbers of
the regenerated carbons from what they should have been.
HC1 leach removed these metals and the thermal regeneration
then restored the carbons closer to their initial properties.

Tables XXV and XXVI present regenerating conditions and re-
sults of a number of runs made under a variety of conditions
but in all the runs HC1 acid pretreatment was employed.  Re-
generation Runs 33-34, 35-36, 45-47, 49-53 and 78-79 were
performed under conditions that can be attained in a directly
heated multiple hearth furnace.  The iodine numbers of these
regenerated carbons varied between 1020 and 1050 mg/g com-
pared to 900 to 950 mg/g for thermal regenerated unleached
carbon.  Since the iodine number of the initial carbon is
believed to be 1090 mg/g, there is still room for improve-
ment.

Regeneration Runs 63-69 and 65-70 were performed under con-
ditions that cannot be duplicated in directly heated multiple
hearth furnaces because of the inert gas sweep used during
baking.  The high temperature over the baking step would also
be difficult to attain.  The inert gas sweep during baking
and the higher temperatures are now of academic interest
since they produced regenerated carbons of lower iodine
numbers.

The HC1 leach for these regenerations was done on Buchner
funnels with acid solution and leach water heated to near
boiling point before being poured over the carbon.  About
30 ml of 37% HC1 diluted to 500 ml water was used to treat
500 cc of carbon.  About 5 1 of water was then used to leach
out the chloride ion, i.e., until a silver nitrate test was
negative.  At this point, the pH was still 3.5.

HC1 Leach Posttreatment, Effect on Iodine Number - Prior to
the HC1 pretreatment studies, tests had been carried out to
determine the effect HC1 leach of regenerated carbons would
have on the iodine and molasses numbers.  The results showed
that the ash content had no effect on molasses number while
removal of ash lowered the iodine number in some cases.  The
results of these tests are summarized in Tables XXVII and
XXVIII,  This information confirms the fact that the in-
crease in iodine number of regenerated carbons, pretreated
with HC1, is due to the effect the pretreatment had on the
thermal regeneration rather than being due simply to removal
of the ash.
                             63

-------
                     TABLE XXV  -  REGENERATION CONDITIONS. HC1  LEACH



Run                                                  Temperature, °F    Gas input,ft3/hr (stp)
no.     Regenerating step, special  treatment        _]_        2_        3_     N_2    £0.2    —2—


        leach, 150 ml HC1  per 2500  cc  OSF  400,  hot  ---      ---      	
33      above carbon baked                         800      1300     1550   7.0   0.90   1.67
34      Run 33 activated                           1550      1650     1700   7.2   0.90   1.84

        leach, 200 ml HC1  per 2500  cc  OSF  400,  hot  ---      ---      	
35      above carbon baked                         800      1300     1550   7.0   0,90   1.29
36      Run 35 activated                           1575      1675     1700   8.3   1.00   2.25

        leach 100 ml HCL + 50 g  NaOH/2500  cc
        OSF 400                                    ---      ---      	
45      above carbon baked                         800      1350     1600   7.0   0.90   1.78
47      Run 45 activated                           1550      1650     1700   5.8   0.77   1.70

        leach, 150 ml HC1  per 2500  cc  OSF  400,  hot  ---
49      above carbon baked                         810      1350     1600   7.0   0.93   1.71
53      Run 49 activated                           1550      1650     1700   3.12  0.41   0.93

	     leach same as for  Run 45                   	      	      	    	   	    	
63      above carbon baked                         1700      1700     1600   3.7   0.00   0.00
69      Run 63 activated                           1700      1700     1600   9.2   1.10   2.10

65      OSF 400 baked                              1700      1700     1700   3.7   0.00   0.00
	     Run 65 HC1 leached                         	      	      	
70      above activated                             1700      1700     1600   9.2   1.10   2.69

	     leach same as for  Run 49                   	      	      	    	   	    	
78      above carbon baked                         800      1350     1550   7.0   0.93   1.96
79      Run 78 activated                           1550      1650     1700   6.0   0.80   1.94

-------
                            TABLE XXVI  - REGENERATION  RESULTS,  HC1 LEACH
Ul
Run  no.

virgin
spent

1 eached
  33
  34

1eached
  35
  36

leached
  45
  47

1 eached
  49
  53

1eached
  63
  69

  65
1eached
  70

leached
  78
  79

Bulk Uei
Bulk Particle
ght volume volume Pore
density» decrease, decrease, decrease, -volume,
g/cc
0.468
0.583
0.557 4.2'
0.508 10.2
0.478 7.7.
0.560 3.8'
0.510 10.7
0.484 6.9.
0.563 3.7"
0.500 13.7
0.468 7.7.
0.555 5.2'
0.495 12.5
0.472 5.6.
0.544 6.01
0.506 9.5
0.472 7.8.
0.517 13.1 "
0.507 2.1
0.475 7.7.
0.551 4.6"
% *, % cc/cc
0.650
0.500
0.1 1 0.0'
- 20.6 1 .7 !- 3.7 0.5
1.8J 1.3.
O.Ol O.l"
- 20.2 2.0 r 3.1 0.5
1.1 J 3.3.
O.O"1 0.0 "
• 23.4 2.7
1.1.
0.4 '
- 21.8 2.1
0.7.
0.2 1
- 21 .6 2.4
0.7 j
2.5'
-21.5 0.0
1.3 .
0.0 '
0.504 10.9 i- 20.0 1 .9
0.480 5.8
0.7 .
-3.8 2.8
1.8 .
0.3 '
-3.3 1.9
1.7 .
•
-3.3
1 .9 .
2.0 '
-3.8 1.0
2.1 j
0.527
-1.8 0.604
0.627
0.524
-2.7 0.611
0.639
- - -
-4.6 0.612
0.638
0.527
-3.9 0.629
0.645
_ _ _
-4.3 0.614
0.642
0.614
-5.0 0.615
0.642
1.7 1 0.519
-2.6 -0.4 h3.3
2.0J 0.632

lodi ne
number ,
mg/g
1090
630
660
910
1020
_ _ —
930
1040
...
940
1050
- - -
970
1020
- - -
900
990
820
	
970
— — —
1000
1040


Mol asses
number
250
190
219
213
234
_ - -
210
215
_ _ _
240
230
_ - _
210
250
_ _ _
230
210
210
	
220
.. _ _
230
230

-------
     TABLE XXVII - ASH CONTENT OF REGENERATED CARBONS
         BEFORE AND AFTER. DILUTE HC1 ACID LEACH	

                                                    Ash after
                                     Ash after     regeneration
  Run no.       Buik density,  g/cc  regeneration,  & dil HC1 leach,
(regenerated)   spent  regenerated  	%	  	%	

Virgin F 400    ---     (0.469)        (5.7)            (5.1)

  56 M)       0.615     0.488         8.4              4.5
  63 (1)       0.616     0.445         9.2              4.8
  65           0.574     0.486         7.5              5.8

  27           0.546     0.464         7.9              5.3
  32           0.582     0.467         7.3              4.4


     (1)  Runs from contract  14-12-107
     TABLE XXVIII- DECOLORIZING TEST RESULTS ON HC1
     	LEACHED AND NONLEACHED CARBONS	

              Iodine number, mg/g         ;1 glasses number
Run no.       no leach    HC1 leach     no leach    HC1  leach
Virgin F 400   1090        1060          240         250

  56            880         790          290         280
  63            850         840          320         320
  65            960         920          220         230

  27           1040        1040          310         280
  32            950         940          ---         270
                            66

-------
HC1 Leach Pretreatment. Effect of Temperature - The success
of the HC1 acid pretreatment in improving the thermal regen-
eration prompted  further study to optimize the pretreatment
conditions.  A rapid preliminary study was made, according
to the procedure  used  previously to determine whether an ef-
fective leaching  could be performed at ambient temperature.
The leaching was  done  by passing 500 ml HC1 solution through
500 ml of OSF 400  in a Buchner funnel.  The acid filtrate was
recycled  twice and  discarded.  Pure water was then passed
through the  bed until  the filtrate was essentially neutral
and free  of  chloride ion.  About 2 liters of water was used
in each case to leach  out the acid.

The results  of this study are presented graphically in
Figure 9.  Leaching was done at near boiling point and at
ambient temperature.   A good measure of contact time is not
obtainable with this procedure, hence, the data obtained is
not suitable for  scale-up purposes.  Ash analysis, by ig-
nition was made on  the original OSF 400 and each leached
batch.  In order  to put the ash analyses on a comparable
basis, the measured ash content was multiplied by the ratio,
0.468/density of  spent carbon.  This gives the would-be ash
analysis  after regeneration to 0.468 g/cc bulk density.

The curves in Figure 9 show that there is not a significant
difference between  cold and hot leach and that in the acid
to carbon contact  time employed, the ash content cannot
readily be reduced  below 4.5%.  Since it is considerably
better to operate  at ambient temperature, because of reduced
corrosion and avoidance of heating costs in large scale op-
erations, the hot  leach studies were discontinued.

HC1 Leach Pretreatment, Effect of Concentration - With know-
ledge obtained from the preliminary studies, another test
procedure was devised  so that reliable information could be
obtained  on  contact time versus amount of acid needed.  Five
glass tubes  were  set up to permit five simultaneous leaching
experiments.  Tube  dimensions were:  I.D. = 4.8 cm, length =
63 cm.  Five hundred cc of OSF 400 was placed in each tube,
giving carbon bed  depth of 34 to 36 cm.  The carbon in each
case was  then submerged in water so that the water level
came to the  top of  the carbon bed.  (This would simulate the
condition of a contactor at Pomona, after a backwash, in
case the  acid leaching could be dOne in the contactors.)
Five hundred ml acid were then added to each of four tubes
and pure water into the fifth.  Acid concentrations used
were: 5,  10, 20 and 30 ml concentrated HC1 diluted to 500 ml.
Flow through each  bed  was controlled by means of a stopcock
at lower end of tube and, for this set of experiments, flow
was set at 100 ml/hr.   This gives a contact time of about
                            67

-------
oo
             4.0
             2.0
                0
                     0
                                                                    E Ambient temp,
                                                                    0 90-95°C
                                j_
 40          80           120          160          200
        ml of  cone.  HC1/500 cc carbon

FIGURE 9 - ASH CONTENT  OF HC1  LEACHED OSF  400

-------
2 hr.  After the acid level reached the top of each carbon
bed, water was added to the tube and flow continued at prev-
ious rate.  The pH, measured on each 50 ml of filtrate, are
plotted in Figure 10 as function of the filtrate volume.
Elemental analyses were also made on the filtrate of the
20 ml acid experiments.  Qualitative chloride tests were
made on each filtrate with silver nitrate.

After the acid leach was completed, carbon samples were
taken from the top, midsection  and bottom for ash analyses.
The  results of the ash analyses are given in Table XXIX.
Also, in  calculating these results, the measured ash con-
tent was  multiplied by the ratio, 0.468/bulk density of
spent carbon, to put the results on a  comparable basis.

On analyzing the data obtained  thus far, several trends were
noted.  A part of metallic ions comes  off very easily.  Pure
water passed through the bed,  reduced  the ash content  from
8.2  to 6.8%.  The pH rose  from  about 5.0 to 6.0 at the same
ti me.

When 5 ml of acid were added,  the ash  content reduced  to
5.1% at the top part of the bed but was still 6.2% at  the
bottom.   In this experiment, also the  pH rose from 5.0 to
6.0  indicating that no free acid penetrated the bed.   The
filtrate  contained Ca and  Na ions, but no other metal  ions,
in concentrations given in Table XXX.  These facts signify
that the  5 ml of acid was  completely consumed within the
bed.  For the 500 cc carbon (235 g regenerated weight), 5 ml
of acid is definitely too  small to give uniform removal of
the  metallic elements.

When 10 ml or more of acid was  added,  the ash content
dropped to the 4.8% level  with  variation between 4.5 to
5.4%.  This variation appears  to be experimental error since
no logical trend is indicated.  The pH of the filtrates
passed through a minimum,  below 1.0, indicating penetration
of acid.

Elemental analyses were made on the filtrate of Experiment  1
(20 ml acid) at several filtrate volumes.  The results of
the analyses are given in  Table XXXI.  The %2®3 1S mainly

    TABLE XXXI - ANALYSIS  OF FILTRATE  FROM EXPERIMENT  1

     Filtrate         Element content of filtrate, mg/ml
    volume,  ml           Na       £a_         E.2^-3

       700              0.055     1.98       3.78
       800              0.045     1.77       2.18
      1000              0.020     trace      0.46
      2000              	     	       	°-11 m9/ml
                                               total solids

                           69

-------
3.0
2.0
1.0 —
0.0
                    1000
         2000
Filtrate volume,  ml
3000
4000
     FIGURE  10  -  pH  OF  FILTRATE  FROM  HC1  LEACH  OF  OSF  400  WHEN  ACID  TO
                        CARBON  CONTACT  TIME  IS  2 HR

-------
     TABLE XXIX - ASH CONTENT OF HC1 LEACHED QSF 400
Experiment
   no.
   5

   3

   2

   1

   4
Acid used,
Ash content, %
ml

0
5
10
20
30
top

6.8
5.1
4.6
4.7
4.8
middle

	
4.9
4.7
5.1
4.5
bottom
r
6.8
6.2
5.4
4.9
5.0
       Initial ash content of unleached OSF 400 = 8.2%
       Ash content of virgin F 400              = 5.7%
   TABLE XXX - ANALYSIS OF FILTRATE FROM EXPERIMENT 3
 Filtrate
volume, ml
    Element content of filtrate, mg/ml
                          Residue on
 Na.      C_§_      C_L      evap.  to dryness
   500

  1000
0.070   0.225

0.010   0.170    0.36
        2.14

        1.32
                            71

-------
Fe with smaller amounts of Cr, Mn and Al.   If the mean_
atomic weight of R is assumed to be 56 g/mol, that of iron,
then the metallic elemental composition of leached ash is
approximately 1% Na, 49% Ca and 50% R.  It comes out that
one percent of ash reduction is chemically equivalent to
2.6 g HC1.  Pure water reduces the ash content from 8.2% to
6.8%, while 10 ml of acid reduces it down to about 4.8%
hence, the water aids the acid considerably.  If it is as-
sumed the acid contributes to the ash reduction from 6.8 to
4.8%, i.e., 2.0%, then the amount of HC1 required is 5.2 g.
Ten ml of concentrated HC1 contains 4.5 g HC1, hence over
10 ml are required for the reaction.

HC1 Leach Pretreatment. Effect of Contact Time - A series of
leaching experiments were conducted at nominal contact times
of 1, 2 and 4 hr.  Since the contact time varies with void
volume in the carbon bed and the void volume in turn depends
on the manner the carbon was loaded into the contactor, the
actual contact time can vary for a given carbon loading.
Void volume can vary from about 40% to 55%, hence a contact
time of 2 hr calculated for 40% void could in reality be
2.9 hr for an expanded bed of 55% void.  To avoid this error,
the contact times are reported in terms of liquid flow rate
per unit weight of carbon, thus:

       ml liquid/        gal. liquid/      Contact time
    (g carbon x hr)    (Ib carbon x hr)     range, hr

          0.850             0.102            1 to 1.4

          0.426             0.051            2 to 2.7

          0.213             0.025            4 to 5.5

Progress of the leaching experiments was measured by deter-
mining pH of the filtrate and chloride ion concentration with
silver nitrate.  When chloride ion concentration became very
low, the water leach phase of the treatment was then stopped
and carbon then dried and analyzed.

Figures 11 and 12 show the effluent pH curves at the various
contact times for 5 and 20 ml acid dosages.  Figure 11 shows
that at the 5 ml acid dosage, the chloride ion does not pen-
etrate to any extent as a free acid even at the shortest
contact time.  Most of the acid is converted to salts.  At
the 10, 20 and 30 ml acid dosages, free acid penetrated the
carbon bed, indicating that more than the stoichiometric
amount of acid was added.  Curves of Figure 12 also show no
definite trends regarding contact time, which indicates that
the chemical reaction and acid diffusion are relatively fast.
                            72

-------
  7.0 —
  6.0
  5.0
= 4.0
OL
   3.0
   2.0
   1.0
   0.0
      0
                              0.426 ml/(g x hr)
1000
    2000            3000
Filtrate volume, ml
4000
                                                                            4500
         FIGURE 11  - pH OF FILTRATE FROM HC1 LEACH WHEN  5 ml  OF  ACID
                          IS ADDED TO 500 cc OSF 400

-------
                        0.852 ml/(g x hr)
            1000
                2000           3000
           Filtrate volume, ml
4000    4500
FIGURE 12 -
pH OF FILTRATE FROM HC1 LEACH WHEN 20 ml OF ACID
     IS ADDED TO 500 cc OSF 400

-------
On the 0.426 experiment of Figure 12, the carbon was ana-
lyzed for chlorine after the acid treatment.   The amount
of chlorine found was 0.2% as compared to 0.04% of the
virgin Filtrasorb 400.  Although the filtrate was free of
chloride ion, according to the silver nitrate test, some
chloride still remained in the carbon.  At present, it is
not known whether the chloride will be liberated from the
carbon during regeneration.  Ash analyses of the HC1 treated
carbons have shown only traces of chloride.

Table XXXII presents  ash analysis data on the leached car-
bons ^sampled from three positions in the carbon bed.  Figure
13 gives the ash analyses for the lower end of the carbon bed.
The results indicate  that the 10 ml dosage is not sufficient
to bring the ash content down to the 4.8% level, for the
235 g carbon being leached.  The optimum amount of acid may
be 14 to 15 ml for a  0.426 ml/(g x hr) flow rate, as indi-
cated by the broken  line curve.

HC1 Leach Pretreatment, Effect on Ash Content - There is a
logical reason why the acid  leach reduces the ash content to
a lower limit of about 4.8%  regardless of amount of acid
used and contact time.  Pittsburgh Carbon Company has found
that HC1 leach reduces the ash content of Filtrasorb 400
type carbon by about  1.0%, i.e., for initial ash content of
5.7% the ash content  after leach is about 4.7%.

The elemental analysis of ash from Filtrasorb 400 type car-
bon, as determined by Pittsburgh Carbon Company, is approxi-
mately as follows:

                  Component            %_

                     SiOo              41.4     85 4
                     A1263             44.0     Bb'4

                     Fe203              7.4
                     CaO                2.9
                     MgO                2.4     14.6
                     KoO                1.0
                     Na20               0.7
                                      99.8

In the analysis the  two relatively insoluble components, Si02
and A1203, constitute 85.4%  of the ash.  The others are quite
soluble in HC1.  By  taking the ratio of 85.4/99.8 times 5.7%
we get 4.9% as the nonsoluble portion of the ash on the car-
bon.  This means that the acid leach has removed all the in-
organic compounds retained by the carbon plus 15% of the ash
initially on the carbon and  leaving essentially an Si02-Al203
                            75

-------
TABLE XXXII - ASH ANALYSES OF HC1 LEACHED CARBONS
Experiment
no.
as received
5
• _ _
3
6
8
2
10
11
1
13
7
4
12
9(Run 79)
14(F-400)
15(F-400)
Acid ,
ml

0
5
5
5
10
10
10
20
20
20
30
30
30
0
15
30
Flow rate >
ml/hr

100
50
100
200
50
100
200
50
100
200
50
100
200
100
200
100

ml/(g x hr)

0.425
0.213
0.425
0.850
0.213
0.425
0.850
0.213
0.425
0.850
0.213
0.425
0.850
0.425
0.850
0.425
Ash
top
8.2
6.8
. _ _
5.0
4.9
4.7
4.7
4.7
4.8
4.7
4.7
5.0
4.9
4.8
•. « _
5.0
	
content
center
8.2
—
...
4.9
	
...
4.7
	
4.9
5.1
5.0
...
...
	
— _ _
4.9
	
, %
bottom
8.2
6.8
— — -•
6.2
6.4
5.2
5.4
5.5
4.6
4.8
5.0
4.6
4.8
4.7
4.6
4.9
4.5
                       76

-------
   7.0
 c
 Ol
 4->
 c
 o
 o
 to
   6.0
                                O 0.852 ml/(g x \\r)

                                0 0.426 ml/(g x hr)

                                A 0.213 ml/(g x hr)
                      10      15       20
                      ml concentrated  HC1
25
30
FIGURE  13  -  ASH CONTENT OF  CARBON  AT LOWER END OF  BED
WHEN BED  IS  TREATED WITH DIFFERENT QUANTITIES OF ACID
                        77

-------
residue.   To remove these compounds requires a caustic leach
in addition to the HC1 leach.  This investigation was started
in Task 2B but since this study progressed only through the
first cycle, no definite conclusion could be drawn.

A more complete study was made by determining the ash content
for Run 67, a nonpretreated regeneration run, and Run 79, an
acid pretreated run.  The analyses are given in Table XXXIII
with those of a virgin Filtrasorb 400.  Run 67 shows a con-
siderable increase in the oxides, Fe203, CaO, MgO, H20, Na20
and Cr203.  As is evident from Run 79, these oxides are con-
siderably reduced by the HCl acid leach, hence one or more
of these oxides contribute to the lowered recovery of iodine
number and over recovery of the molasses number.

HCl Leach Pretreatment. Effpr.t on Pore Structure - In pore
structure studies performed at Pittsburgh Carbon Company1^>ij
it was found that the iodine number was proportional to the
surface area of pores larger than 10A in diameter and the
molasses number was porportional to the surface area of pores
larger than 28A in diameter.  Equations 1 and 2 below show
these relationships.

         I2 No. = 17 + 1.07 x (s.a. of pores >loA)      (1)

         Molasses No. = 129 + (s.a. of pores />28A)      (2)

With reference to the iodine and molasses numbers given in
previous experimental results, these equations indicate that
acid pretreatment minimizes the decrease in surface area of
the smaller pores and also prevents increase in surface area
of pores larger than 28A diameter.  Ash build up, as occurs
on successive regenerations without acid leach, accelerates
these changes.

A further study was conducted to determine the manner in
which the surface area changes occurred.  Pore size distri-
bution curves were determined for selected carbons, using the
water adsorption method^ and mercury porosimetry.  In pre-
paration for the water isotherm determinations (and also por-
osimetry), the carbons were HCl acid and pure water leached
to remove hydrophilic compounds from the carbon surface.
The validity of the water adsorption method depends on the
water being adsorbed by capillary condensation, with negli-
gible monolayer adsorption.  The adsorption method is ap-
plicable to maximum pore diameter of 500A to 1000A\

Pores in the larger diameter range are measured by mercury
porosimetry5.  These measurements cover the diameter range
30A to lOO.OOOA, hence the two methods overlap in the range
                            78

-------
TABLE XXXIII - ASH COMPOSITION OF ACID PRETREATED AND
         NONPRETREATED REGENERATED CARBONS
Component
Si02
A1203
Fe203
CaO
MgO
K20
Na20
Ti02
Cr203
Total ash
in carbon

virgin
F 400
2.36
2.51
0.42
0.17
0.14
0.06
0.04
0.03
0.01
5.7
Ash composition, '
nonpretreated
Run 67
2.27
1.94
0.80
1.23
0.34
0.40
0.15
	
0.12
7.2
K
pretreated
Run 79
2.24
1.62
0.59
0.05
0.09
0.20
0.07
	
0.12
5.0
                          79

-------
30& to 500JL   For some of the pore size distributions, the
agreement in  the overlap range is good while for others,
some discrepancy exists.  The trend is for the mercury por-
osimeter to measure a larger pore diameter at a given volume.
Figure 14 presents two distribution curves, Run 68 showing
the best agreement and F 400 the poorest.  A possible explan-
ation for the discrepancy is that the carbons are slightly
compressed by the mercury at the higher pressures.  Maximum
pressure at 30A is 60,000 lb/in2.

Figures 15 and 16 present the complete pore size distribution
curves of the selected samples.  The overlap portion as
measured by mercury porosimetry was left out in each case.

Those in Figure 15 show the effects of acid leach, steam ac-
tivation and C02 activation.  The virgin F 400 is uppermost
in pore volume at the 28A* diameter, and then Runs 79, 67 and
68 in descending order.  The iodine numbers were respectively,
1090, 1040, 940 and 840 mg/g, while the molasses numbers were
quite close to each other.

Those in Figure 16 also show effect of acid leach; Run 36 be-
ing uppermost was pretreated, while Runs 46 and 67 were not.
Runs 46 and 67 are presented together because they have iodine
numbers close to each other, i.e., 920 and 940, respectively,
but have greatly different molasses numbers, 320 and 260, re-
spectively.

The relationship between the decolorizing tests and pore
structure is  not readily apparent by visual inspection of
the pore size distribution curve, but comes more discernable
when the cumulative surface areas at different pore sizes
are compared.  The cumulative surface area can be calculated
with the equation
                  AA =
                        4 AV
where AA is an increment of
increment of pore volume AV
summing up AA over the pore
pores larger or smaller than
calculated.
in Figure 17
This has been done
and for carbons of
                surface area associated with
                with mean diameter D.  By
                volume, the surface area of
                any specified D can then be
for the carbons of Figure
Figure 16 in Figure 18.
15
From Figure 17, the surface areas of pores larger than 10A
in diameter for the carbons F 400, Runs 79, 67 and 68 are
965, 925, 860 and 805 m2/g, respectively.   The surface areas
of pores larger than 28A* are respectively 120, 100, 110 and
100 m2/g.  Likewise from Figure 18, the surface areas of
pores larger than 10A" in diameter for the carbons Runs 36, 67
                            80

-------
         1 .00
00
                                                             5000  10,000
100,000
                                         Pore diameter, A
                   FIGURE  14 - PORE SIZE DISTRIBUTION MEASURED BY WATER ADSORPTION
                                     AND BY MERCURY PENETRATION

-------
           1 .00
00
          0.80
        cr.

        o

        ° 0.60
         *>
        
-------
           1.00
oo
           0.80 —
         CT>
           0.60 —
         O)
         QJ
         S-
         o
         O-
           0.40 —
           0.20 —
                10    20  30   50     100


                      FIGUPsE 16 - PORE SIZE DISTRIBUTIONS OF  REGENERATED  CARBONS
   500  1000   0     5000  10,000
Pore diameter, A
100,000

-------
                  I  1  1
         1000
00
       CD
       Ol
       J_
       O!
       
-------
           1000
00
Ln
            800
          en


         CVJ
          to
          o>
          03
          
-------
and 46 are 915, 860 and 840 m£/g, respectively.  Surface
areas of pores larger than 28A in diameter are respectively
85, 110 and 150 m2/g.  When these surface areas are sub-
stituted into equations 1 and 2, the calculated decolorizing
numbers agree with test results within * 5% in the iodine
numbers and within 13% in the molasses numbers.  These re-
sults are given in Table XXXIV.

      TABLE XXXIV - IODINE AND MOLASSES NUMBERS AS
   DETERMINED BY TEST AND CALCULATED FROM SURFACE AREA
   Carbon

Filtrasorb 400
Run 79 (HC1)
Run 67
Run 68 (C02)

Run 46
Run 36 (HC1)
(Run 58 [Darco]
                  Iodine number, mg/g
test

1090
1040
 940
 840

 920
1040
 580
calc

1040
1010
 940
 880

 920
1000
 620
                         Molasses number
test

 250
 230
 260
 250

 320
 230
 370
calc

 250
 230
 240
 230

 280
 210
 420
Further study of the surface area curves also indicates where
the major portion of the adsorption may be occurring.  By
inspection of the curves in Figure 17, it is observable that
mo§t of the change in core structure occurs in the pores from
14A to 28A.  Beyond 28A, there can be considerable change in
pore volumes, as is apparent from inspection of curves in
Figures 15 and 16.  The surface area of these larger pores
is, however, too small to be effective.  The same is true
for Runs 67 and 38, of Figure 18, but Run 46 is an exception.
For this carbon considerable change occurs at the 28A diameter
region.

To show the change in pore structure, the difference in sur-
face area between that of F 400 and each of the other car;
bons has been calculated at pore diameters from 10A to 50A.
These calcualted areas are given in Table XXXV.  For Runs 79,
67, 68 and 36, the surface area changes, as already noted,
occur mostly in the 14A to 28A diameter range.  The result;
indicate a shift in the pore sizes.  Pore volume in the I
to 28A range decreases while the volume in the 14A to 20A in-
creases.  It appears as though the pores that were originally
ISA to 22A diameter have decreased to 14A to 18A diameter.
The net effect, however, is that the cumulative surface area
of pores larger than 10A diameter decrease on regeneration.
Further study is desirable to clarify the exact changes that
do really occur.
                             86

-------
     TABLE XXXV - SURFACE AREA CHANGE AT DIFFERENT
          PORE DIAMETERS DURING REGENERATION
Surface area di

F
h
F
K
F

400
400
400
400
400

- 79
- 67
- 68
- 46
- 36

(HC1)

(C02)

(HC1)
10A
40
105
160
130
30
12)1
35
110
160
115
40
14ft
30
105
155
90
40
16A
80
160
165
155
79
fference,
ISA*
200
270
370
170
195
20fl
150
215
130
115
150
m2/q
28A
20
10
20
-30
30

soft
16
00
00
-10
15
The results definitely do show that HC1 pretreatment mini-
mizes pore structure changes.  The results also suggest that
the regeneration action takes place mostly in pores of 18A
to 28A diameter.  These pores must then also be the most ac-
tive in the adsorption process.

HC1 Leach Pretreatment. Standardized Procedure - Based on the
parameter studies, a standardized procedure was adopted for
treatment of carbons as for Task 2A, 2B and 7.  The procedure
was as follows carried out in the unit shown in Figure 3.

    Carbon Bed - 2500 cc of dried spent carbon are placed
    in the contactor.

    Wetting of Carbon Bed - Pure water is run upflow until
    water level is at top of carbon bed.

    HC1 Acid - 75 ml of 37.9% HC1 acid is diluted to 2500 ml
    and passed downflow through carbon bed at 500 ml/hr
    flow rate.

    Water Leach - After HC1 solution has passed through
    carbon bed, 16 1 of pure water are passed downflow
    through carbon bed at 500 ml/hr flow rate.

    Drying of Carbon - Treated carbon is removed from
    contactor and dNed in air convection oven at 150°C
    for about 48 hr.  The carbon is ready for thermal
    regenerati on.

Based on the preceding laboratory data, the scaled-up con-
ditions for acid leaching of carbon bed of size in the
Pomona contactor are given below.
                            87

-------
Tamperature                - Ambient

Carbon volume              - 269 ft3 (6 ft dia by 9.5 ft
                                      depth)

Carbon weight              - 6650 Ib

HC1 acid per unit carbon   - 0.076 Ib acid/lb carbon
                                     (37.9% assay)

Total HC1 acid             - 505 Ib (0.076 x 6650)

Flow rate per unit         - 0.426 lb/(lb x hr)
     carbon weight         - 0.051 gal/(lb x hr)

Flow rate of acid
solution and water         - 340 gal/hr (0.051 x 6650)

Leach acid solution
       volume              - 1700 gal (340 x [500 x 100])

Leach water volume         - 10,200 gal

pH of filtrate at end of
     leaching operation    - 3.0 to 3.5

Chloride ion concentration
   in fi1trate at end of
   leaching operation       - nil
Ash content of leached
  spent carbon
"- 4.1% based on spent carbon
    density of 0.548 g/cc
- 4.8% based on reg. carbon
    density of 0.468 g/cc
Overall leaching time      - 35 hr

Acid-carbon contact time   - 2 to 3 hr

To perform the acid leach, the contactor is given  its final
back-wash.  Then the water level in the bed is lowered so
that it is level with the carbon surface.  Acid solution is
then pumped through the bed at rate and quantity stated
above, followed by the water leach.  From this point on the
carbon is handled in the usual manner.

At the 0.076 Ib acid/lb carbon rate, the acid cost is about
0.145<£/lb of regenerated carbon when 37.9% assay technical
grade acid sells for 1.95<£/lb.

-------
Task 5 - Engineering Studies on Furnaces

The anticipated need for engineering studies was based on the
premise that the regeneration is controlled by the parameters;
temperature, gas composition, gas input rate and carbon
residence time as investigated in Task 1.   At the time the
program was outlined, difficulties were expected in attain-
ing the precise conditions in a direct-fired multiple hearth
furnace and that indirect-heated rotary furnaces may have to
be used to carry out the regeneration in whole or in part.
As the program progressed, it was found out that the main
obstacle in the way of satisfactory regenerations was the
metallic elements.  When these elements were leached out with
HC1, it became apparent that the multiple  hearth furnace
could perform an adequate regeneration and the need for en-
gineering studies of various other furnaces was no longer
necessary.

Task 6 - Regeneration Control by Effluent  Gas Analysis

The analysis of effluent gas from the activating step ap-
peared to offer an instant monitoring method for the ac-
tivation.  When the baking and activating  are run in suc-
cession as in the multiple hearth furnace, it is essential
that the activating effluent gas in withdrawn from the re-
generator before any appreciable mixing has occurred with
the gas released from the baking step.  In the pomona fur-
nace the point of withdrawal would be at the fourth hearth.

In order for the effluent gas analysis to  be meaningful,
an analysis of the input gas composition and also a good
control of the input rate are essential.

Vent gas analyses were conducted on 25 activation runs.
However, an answer as to the possibility of this method for
monitoring purposes cannot be given, as yet, because of a
number of unexplained deviations that have appeared in the
results.  These deviations, however, appear to be giving
information about the activating process.

The concept of vent gas analysis for activation monitoring
was based on the premise that the carbon entering the activat-
ing step is essentially free of combined oxygen, sulfur and
hydrogen, or at least present in quantities too small to in-
validate the results for monitoring purposes.  It was also
assumed that the gas entering the activating step is essen-
tially N2' CC>2» H20 and an insignificant amount of 02-  The
activating reactions should then be basically:
                            89

-------
                 C02
+ C

+ C
2CO

CO +
                                        H2
(1)

(2)
The input flow rates of N2, C02 and H20 are known  and  of  these
the N2 flow rate remains unchanged on passage  through  the  regen-
erator.  Since either reaction seldom goes to  completion,  the
vent gas is usually a mixture of N2, C02, H/>0,  CO  and  Ho.  A
partial stream of the vent gas was passed through  a  cold  trap
at -80°C to remove the water, hence the percentage compositions
of C02, CO and H2 in N2 were analyzed.  Since  N2 makes  up  the
balance of the gas and its flow rate is known,  it  is then  pos-
sible  to calculate the flow rate of each component.

The gas utilization is then a simple subtraction of  gas out-
put rate from the gas input rate, thus:

    C02 utilization

    (C02 ft3/hr input) —  (C02 ft3/hr output)  =

                           (C02 ft3/hr utilization)        (3)

    H20 utilization

    Is equal to H2 ft3/hr  output                          (4)

    H20 unreacted

    (H20 vapor ft3/hr input) — (H2 ft3/hr output) =

                                (H20 ft3/hr unreacted)    (5)

The C02 utilization can also be calculated from the  CO  and H2
output rates, thus

    (CO ft3/hr output) —  (H2 ft3/hr utilization)  =
                           2
                           (C02 ft3/hr utilization)        (6)

All gas measurements were  converted to standard temperature
and pressure (stp).
Input and Output Gas
Resul ts_ - For the
       summarized
                '2
    record, the input and
    in Appendix A.  In
    ratio was that of the
output gas compositions are
most activations, the input C02 and N
flue gas mixture.  The H20 vapor concentration was  varied
considerably relative to the N2-C02 mixture.   In  Run  67, no
C02 was added, only N2 and H20 vapors.   In  Run 68,  no  H20
vapor was added, only N2 and C02.  However,  for this  run the
C02 concentrations were above trie analyzer  capacity and,
therefore, no analyses are recorded for  this  run.
                             90

-------
 ^    Phase  Equilibria  -  According  to  thermodynamics,  a gas
mixture  consisting  of H20,  CO  and  C02  and  H«  can be  expected
to  react  at  the  activating  temperatures  according  to the
equation,

                 H20 + CO  <    > C02 + H2                   (7)

and  come  to  an equilibrium,  which  varies with  temperature.
The  equilibrium  equation  and values  of the  equilibrium  con-
stant  K  are  given below for several  temperatures of  interest

                       (C02)  (H2)

                       (H20)  (CO)   =  K

The  quantity  in  each  bracket is  the  mole percent of  the gas.

              Temperature °F            K_

                  1400                 1.36
                  1500                 1.04
                  1600                 0.86
                  1700                 0.73
                  1800                 0.63

If,  after  reaching  equilibrium at  the  elevated  temperature,
the  gas  mixture  is  cooled rapidly, it  can  retain the ele-
vated  temperature equilibrium  composition.  That this happens
to  some  degree during the activation runs  is  indicated  by
the  K  values  given  in the last column  of Table  XXXVII.  They
vary from  0.61 to 1.42, suggesting approach to  equilibria at
temperatures  from 1400°F  to 1800°F.  Since  the  temperature
of  zone  one  in the  regenerator was either  near  1550° or 1700°F.,
it  had been  expected  that K for the  1550°F  runs would stabi-
lize near  1.0 and for the 1700°F  runs  near  0.7.  However, the
temperature  of zone one had no bearing on  the  K value.  At
present  no explanation  can  be  offered  for  the  observed  vari-
ation  in  values of K.

COp  Uti1j zati on  - The C02 utilizations as  calculated from
the  COp  input and output  rates,  Equation 3, were for most
actiyation runs  less  than when calculated  from  the CO and H2
output rates  according  to Equation .6.  The  CO?  utilizations
and  the  differences are reported  in  Table  XXXVI.

In  the attempt to determine  the  significance  of the  difference
in  C02 utilization, it  was  found  that  a  low order  of correlation
exists  between this  difference  and the percentage particle
volume  decrease.   The percentage figures  for the particle
volume  decreases  are  given  in  the last column of Table XXXVI
                            91

-------
                                       TABLE  XXXVI  -  CO? UTILIZATION
                        C02 utilization,ft3/hr  (stp)
ro
Run
32
34 (HC1)
36 (HC1)
39
40
41
47 (HC1)
51
52
53 (HC1)
66
67 (HC1)
69 (HC1)
70
76 (WVP&P)
77
79 (HC1)
88 (HC1)
89 (HC1)
90 (HC1)
98 Task 2
104 Task 2A
(HC1)
105 Task 2B
(HC1)
106 (WVP&P)
(HC1)
108 (WVP&P
(HC1)
Direct
0.22
0.00
-0.13
-0.07
0.14
-0.12
-0.07
0.15
0.20
-0.02
0.11
-0.53
-0.09
-0.10
0.08
0.07
-0.08
-0.01
-0.01
-0.11
0.00

-0.14

-0.06

-0.03

-0.15
CO-H2 anal
0.23
0.02
-0.02
0.02
0.16
-0.01
0.01
0.18
0.30
0.06
0.07
-0.52
0.02
-0.05
0.11
0.03
-0.03
0.03
0.00
-0.04
+ 0.10

-0.03

+0.07

+0.01

-0.05
                                                     CO-
  Di fference
utilization,  ft3/hr
    -0.01
    -0.02
    -0.11
    -0.09
     0.02

    -0.11
    -0.08
    -0.03
    -0.10
    -0.08

    + 0.04
    -0.01
    -0.11
    -0.05
    -0.03

    +0.04
    -0.05
    -0.04
    -0.01
    -0.07

    -0.10

    -0.11

    -0.13

    -0.04

    -0.10
Particle  volume
  decrease,  %

      2.6
      1 .3
      3.3
      2.0
      1.5

      2.2
      1.8
      1.9
      2.7
      1.7

      0.6
      1.2
      1.9
      2.1
      1 .6

      0.5
      2.0
      0.2
      1.5
      1.3

      1.2

      2.8

      0.7

      1.7

      2.1

-------
        TABLE XXXVII  - RESULTS  OF  GAS ANALYSIS  DURING ACTIVATION
Gas utilization, %
Run no.
32
34 (HC1)
36 (HC1)
39
40
41
47 (HC1)
51
52
53 (HC1)
66
67
69 (HC1)
70 (HC1)
76
77
79 (HC1)
88 (HC1)
89 (HC1)
90 (HC1)
98
104 (HC1)
105 (HC1)
106 (HC1)
108 (HC1)
C02
47.0
2.5
-2.0
2.0
25.0
0.0
1.3
43.9
68.2
14.6
12.7
	
2.0
-4.1
26.8
3.5
-6.0
7.0
1.0
-9.3
58.8
-5.5
12.5
1 .8
-6.3
H20 C02+H20
74.8
40.0
42.7
63.0
74.0
59.0
43.5
73.0
82.5
65.5
69.8
73.0
46.0
35.2
61 .6
42.9
43.3
57.5
47.4
50.4
66.7
55.4
50.6
36.2
33.5
65.9
28.3
28.8
43.6
54.0
41 .6
31.3
62.6
76.2
50.0
54.4
54.0
31 .4
23.0
49.5
31.9
28.8
43.6
34.7
35.1
62.1
39.6
40.7
25.7
23.1
Rate carbon oxi
By gas anal
16.8
11 .4
15.3
20.8
13.0
23.3
11.4
10.9
11 .7
10.1
16.8
23.2
15.9
13.6
9.1
17.0
12.3
7.7
8.2
9.5
2.3
8.9
7.3
8.4
13.0
dation, q/hr
By weighi ng
21.
17.
16.
21 .
16.
26.
15.
14.
16.
11.
20.
27.
17.
17.
13.
24.
13.
12.
11.
13.
9.
13.
12.
11 .
15.
5
0
5
6
3
3
4
5
3
4
0
7
6
0
0
5
5
8
8
4
2
9
7
0
7
Di f f erence
4.
5.
1.
0.
3.
3.
4.
3.
4.
1.
3.
4.
1 .
3.
3.
7.
1 .
4.
3.
3.
6.
5.
5.
2.
2.
7
6
0
8
3
0
0
6
6
3
2
5
7
4
9
5
2
9
6
9
9
0
4
6
7
Weight 0)
decrease ,
g/h
5.
3.
7.
3.
3.
5.
a.
3.
4.
3.
1.
4.
4.
4.
3.
0.
4.
0.
3.
2.
3.
8.
1 .
4.
5.
r
6
7
4
3
2
1
5
8
7
2
3
6
0
2
4
6
9
4
9
8
6
2
5
2
3
K
0.86
0.67
0.93
1.32
1.36
1.14
0.85
0.78
0.87
1.12
0.86
1.42
0.95
0.85
0.79
0.61
0.86
0.60
0.72
1 .00
1 .21
1.15
0.79
0.70
0.74
                                                                 3.9



(1)  Weight decrease calculated  from particle volume decrease

-------
and the correlation is shown graphically in Figure 19.  No
logical explanation can be offered for this correlation.

Rate of Carbon Oxidation - The rate of carbon oxidation was
calculated from the rates of COo and steam utilizations,
with COo utilization rate calculated according to Equation 6.
The oxidation rates based on gas analysis were then compared
to oxidation rates based on direct weight of carbon before
and after the activating step.  It was found that the oxi-
dation rates based on weight measurements were consistently
larger and that the differences were approximately equal to
the weight decreases associated with particle volume decreases.
The results of these calculations are summarized in Table
XXXVII.  The mean weight of carbon oxidized, as obtained by
difference, was 3.7 g/hr while the mean weight based on
particle volume decrease was 3.9 g/hr.  This fact suggests
that particle volume decreases may be due to formation of sub-
micron size fines which are then carried out of the regen-
erator by the vent gas.

For Runs 36, 70 and 79, the calculated C02 utilizations were
negative, indicating that COo was actually produced rather
than consumed.  This is possible if steam input rate is large,
in view of Equation 7 and equilibrium constants K.

Gas Analysis Without Carbon in Regenerator - Because of some
of the apparent discrepancies in the gas analyses, a number
of tests were conducted in which gas mixtures were passed
through the regenerator under activating conditions but with-
out carbon.  These tests were primarily a check on the ac-
curacy of the gas concentration measurements.

The results of the measurements are reported in Appendix B,
and Table XXVIII presents an analysis of the results.  Ex-
cept for Test 5, these show a self consistency of 0.01 to
0.02 ft^/hr in the consumed and produced gas volumes, hence
the apparent discrepancy observed in the gas analyses of
the activation runs are not due to errors in measurements.

Tests 3, 4 and 5 show formation of C02 when CO is added to
the gas mixture, indicating that with high steam and CO con-
centration, C02 can be formed as per Equation 7.

Table XXXIX presents the K values for these tests.  Except
for Test 1, they fall in the range 0.65 to 1.36.  Because of
the very small CO concentration in Test 1, the high value
of K = 1.69 is due to lack of instrumental accuracy.
                            94

-------
VD
Ul
                4.0i—
                3.0
              Ol
              to
              ro
              
-------
        TABLE XXXVIII - REACTIONS ASSOCIATED WITH
                    EACH GAS ANALYSIS
Test 1
     COp + metal—-metal oxide + CO
           C02 consumed 0.92 - 0.93 = -0.01 ft3/hr
           CO produced  0.01 - 0.00 =  0.01 ft3/hr

     H20 + metal—> metal oxide + H2
           H2 produced  0.02 - 0.00 =  0.03 ft3/hr
           \\2 consumed              = (-.03 ft3/hr)
              (from H2 prod)
Test 2
     H2  + C02—* CO + HoO
           Hp consumed  0.58 - 0.79 = -0.21 ft3/hr
           C02 consumed 0.72 - 0.93 = -0.21 ft3/hr
           CO produced  0.20 - 0.00 =  0.20 ft3/hr
           H20 produced
           (from H2 prod)0.20 - 0.00 = (0.20 ft3/hr)

Test 3

     CO + H20 —> C02 + Ho
           CO consumed  0.49 - 0.90 = -0.41 ft3/hr
           H20 consumed             = (-0.41 ft3/hr)
               (from CO cons)
           COo produced 1.35 - 0.93 =  0.42 ft3/hr
           H20 produced 0.38 - 0.00 =  0.38 ft3/hr

Test 4

     CO + H20 •--> COp + H?
           CO consumed  6.23 - 0.50 = -0.27 ft3/hr
           H20 consumed             = (-0.27 ft3/hr)
               (from CO cons)
           C02 produced 1.21 - 0.93 =  0.28 ft3/hr
           H2 produced  0.25 - 0.00 =  0.25 ft3/hr
Test 5
     CO + HoO — > C02 + H2
           CO consumed  6.10 - 0.20 = -0.10 ft3/hr
           H20 consumed             = (-0.10 ft3/hr)
               (from CO cons)
           C02 produced 1.00 - 0.93 =  0.07 ft3/hr
           H2 produced  0.17 - 0.00 =  0.17 ft3/hr
                            96

-------
        TABLE XXXIX - K FOR TESTS 1 THROUGH 5

               Test            K

                 1            1.69
                 2            0.98
                 3            0.72
                 4            0.97
                 5            1.06

Task 7 - Regeneration of 25 Ib Quantities of Spent Carbon

The object of this task was to carry out larger quantity re-
generations after the optimum operating conditions had been
worked out in Task 1.  Regenerations of Darco and WVP&P were
to be carried out.  Darco is a 14 x 30 mesh granular carbon
manufactured by Atlas Powder Company and the WVP&P is an
8 x 30 mesh granular carbon manufactured by West Virginia
Pulp and Paper Company.

As has already been learned, the work of Task 1 during the
first 12 months of the program did not yield optimum con-
ditions which would reproduce the initial properties in the
regenerated carbon, hence confirming runs of larger duration
could not be carried out during that period.

Exploratory regeneration runs were, however, carried out
with the WVP&P and Darco carbons.  The Darco regeneration
followed a drastically different pattern from that of Fil-
trasorb 400; it regenerated to the virgin bulk density or
lower during the baking step, although the iodine number
did not always come up to the initial value-  No activating
step appeared necessary, but variations were tried in baking
condi ti ons.

The regeneration conditions and results for these two car-
bons are given in Tables XL and XLI.

For the Darco carbon, in Run 58, a gas mixture approximating
a flue gas was used as sweep gas and carbon residence time
was 34 min.  This brought the bulk density below the virgin
carbon bulk density and the iodine number came up to the
level of the virgin carbon.  On Run 59, the residence time
was shortened to 21 min in an attempt to raise the bulk
density.   In Run 60, N2 sweep gas was used and the baking
temperature was raised.  In this case, the bulk density
matched the virgin carbon density but the iodine number was
now below that of the virgin carbon.  Run 75 is similar to
Run 59 but with lower temperature; the results were not sat-
isfactory.  Run 73 was HC1 pretreated and produced a regen-
                            97

-------
                        TABLE XL - REGENERATION CONDITIONS. WVP&P AND DARCO
VO
00


Run no .













1

1
1


_
1
1
58
59
60
. _

73
75
74
76

99

_ _ _
00

01
06


_ _
07
08
Regenerati ng
step
Darco, baked
Darco, baked
Darco, baked
spent Darco HC1
1 eached
above, baked
Darco, baked
WVP&P, baked
Run 74 act
spent Darco HC1
thermal reg
Large Darco
spent Darco HC1
thermal reg
spent WVP&P, HC1
above baked
Run 101 act
Large WVP&P

spent WVP&P, HC1
above baked
Run 107 act
Tempera
1
850
850
170C

	
800
850
800
1550

850
run -
	
880

880
1550
run -

	
880
1550

1
1
1

-
1
1
1
1

1
10
-
1

1
1
15

-
1
1
ture ,
2
350
350
700

--
350
250
350
650

360
.9 Ib
--
400

350
650
.5 Ib

--
350
650
°F

1
1
1

-
1
1
1
1

1
Gas input, ft3/hr (stp)
3
550
550
600

--
550
450
550
700

600
feed, 21.7
-
1

1
1
--
600

550
700
feed, 48.3
(i
-
1
1
ncl udes
—
550
700
N2
7.0
7.0
3.7

	
7.0
7.0
7.0
3.3

7.0
hr
	
7.0

7.0
4.0
hr
baki
	
2.0
7.6
co2
0.90
0.90
none

—
0.93
0.93
0.93
0.41
...
0.93
thermal
	
0.93

0.93
0.67
thermal
H?0
1.49
1.71
none

—
1.50
2.05
1 .96
0.78
...
2.14
regenerati
	
1 .76

1.67
1 .49
regenerati
Residence
time, min
34
21
35

	
38
34
30
16
...
27
on time
	
34

26
9
on
ng and activating)
	
0.93
1 .00
	
1.74
2.68
	



-------
                    TABLE XLI - REGENERATION RESULTS, WVP&P AND DARCO
v£>



Run no.
spent Darco
spent WVP&P
virgin Darco
virgin WVP&P
58
59
60
Darco HC1
leach
73
75
74
76
spent Darco
HC1 leach
99
spent Darco
HC1 leach
100
spent WVP&P,
lot 1
HC1 leached
101
106
spent WVP&P,
lots 2 to 6
HC1 leached
107
108

Bulk
densi ty ,
g/cc
0.450
0.597
0.408
0.504
0.380
0.390
0.408

0.433
0.387
0.402
0.531
0.509
0.502
0.472
0.419
0,496
0.471
0.415

0.582
0.568
0.535
0.517

0.582
0.568
0.531
0.512

Weight
decrease ,
of
...
	
	
	
17.2
16.0
13.1

H^} 18'°
13.7
12.41
5.5-f

6 . 3 1 on 3
14.9J
	
l^}20"2

_ _ _
2.3]
7.8k 14.1
4.6J

	
2.81
9.5h 17.5
6.lJ
Bulk
vol ume
decrease ,
%
...
—
	
	
3.2
3.2
4.2

to}3'9
2.0
1.3l
o.ej

0 . 4J_ A n
3.6J
_ - _
0.21.4.4
4.2f

	
0.0"!
2.2H 2.9
0.7J

	
0.2]
3. IF 4.7
1.4J
Particle
volume
decrease ,
%
...
	
---
	
3.7
7.0
6.8

4-1"L 8.0
6.0
2 • 9~L 4
i.ej-

0 . 3|_ 2 1
1 .8J
	
I:?}2-1

	
-0.3J
2.2h 3.6
1.7J

	
-0.5]
3.9k 5.5
2.1J

Pore
volume,
cc/cc
0.585
0.522
	
	
0.686
0.669
0.664

0.592
0.677
0.662
0.603
0.619
0.562
0.581
0.667
0.562
0.585
0.665

0.554
0.574
0.617
0.628

0.554
0.576
0.618
0.637

Iodine
number ,
mg/g
310

570
1070
580
540
510

630
530
590
970
330
360
660
330
360
700

	
800
950
1070


805
1120


Mol asses
number
. . .
180
	
	
366
330
300

310
320
210
250
260
270
305
250
270
310

230



250
280

-------
erated carbon with a low bulk density but also a high iodine
number.  This is a favorable position since it will now be
possible to adjust baking conditions to increase bulk density
while decreasing the iodine number to match those of the
virgin carbon.

After the effectiveness of the HC1 leach procedure had been
established, regeneration of these two carbons was resumed
during the extension period, but on new batches of carbons
received from Pomona.  The conditions of regeneration and
results are included in the lower halves of Tables XL and
XLI, Runs 99, 100, 101, 106, and 107 and 108.

About 13.7 Ib of (dried) spent Darco were available for the
extended regenerations which now consisted of (1) an HC1
pretreatment according to the standardized procedure and
(2) thermal regeneration normally designated as the baking
step.  Of the 13.7 Ib, 2.8 Ib was used in Run  99 to estab-
lish the regeneration conditions.  The remaining 10.9 Ib
were then regenerated in a continuous run over a 22 hr
period.

Other conditions of operation not given in Table XL are:

      Feed rate          - 0.016 ft3/hr; 450 cc/hr

      Tube rotati on      -1.1 rpm

      Tube inclination   - 3.75 in. per 42 in. length

No problems that could not be easily corrected were encount-
ered during the run.  The carbon losses, 4.4% by bulk volume
and 2.1% by particle volume decrease, are lower than had been
expected as based on results of the previous runs. Runs 59,
60 and 73.  The iodine number is higher than the value re-
ported for the virgin carbon (of lower density).  Some of
these differences may be due to different initial properties
of the carbons, since these results are now on a different
batch from Pomona.

The great ease with which the Darco can be regenerated is due
to its open pore structure, i.e., a large portion of the
total pore volume is in pores greater than 30A diameter.

Carbons with a large portion of large pores also tend to be
softer and therefore subject to greater attrition and this
is reflected in the generally larger decrease in particle
volume during the regeneration.  For Darco the particle volume
decrease ranged from 3.7% to 8.0% for the first batch, while
on the extended regeneration run of the second batch, the
particle volume decrease was only 2.1%.
                          100

-------
The regeneration of WVP&P carbon was expected to follow a
pattern very similar to that of Filtrasorb 400 since they
are both activated by a similar method from bituminous coal.
In the spent WVP&P regenerations, a baking and an activating
step are necessary and also the required conditions are very
similar to that for Filtrasorb 400.  Run 74-76 of Tables XL
and XLI shows the regeneration results when no acid pretreat-
ment was employed.  The iodine number was 970 relative to
1070 mg/g for the reported virgin carbon.  Table XVII, Runs
88, 89 and 90, shows the effect of HC1 acid pretreatment.
For these regenerated carbons the iodine numbers were in the
1010 to 1070 mg/g range.  On the new batch acquired for the
extended run, Runs 107 and 108, the spent carbon was HC1 pre-
treated as routine procedure.  Iodine numbers of 1070 and
1120 were attained.  A large excess of steam was used on Run
108 in the attempt to bring the bulk density down to 0.504
g/cc, the density reported for the virgin carbon.  The excess
steam did not accomplish this.  However, the steam utilization
was low, about 23.5%, and particle volume decrease was ab-
normally large, 5.5%.

Other conditions not given in Table XL for WVP&P runs are:

     Feed rate          - 0.016 ft3/hr

     Tube rotation      - 1.2 rpm for bake; 24 rpm for
                                 activation

     Tube inclination   - 3.75 in. per 42 in. length

Figure 20 shows the various carbons studied in perspective.
The Darco carbons have much lower densities and iodine num-
bers than Filtrasorb 400 and WVP&P carbons.  There is some
indication that absorbate pick up may be slightly less for
the Darco since the weight loss of the Darco during regen-
eration ranged from 13% to 20% while the weight loss of
Filtrasorb 400 varied between 18% and 24%.  This may not be
an entirely valid comparison since the absorbate concentration
in the waste water varies and the Darco may have been spent
at a lower concentration.

HC1 leach of the spent Darco, prior to baking, increased the
iodine number from 560 to 630 mg/g for the first batch re-
ceived from Pomona.  This increase is proportionately the
same as observed with acid leached Filtrasorb carbons.  The
iodine numbers of the regenerated second batch are much high-
er.

In the Filtrasorb 400 regenerations, the attempt has been to
regain the 0.469 g/cc bulk density and 1090 mg/g iodine num-
                            101

-------
  1200


  1100


  1000


   900
   800
0>
^
01
OJ
T3
O
   700
   600
   500
   400
   300
   200
   100
     0
       —   Darco
                              WVPRP
       F11trasorb
          400
                      X

                      0  00
                       0 0
       O
       0

       O
                                    El
                                          C02
                                          act.
                          0
       2nd hatch

       ©
                         1st hatch
   0 • IIC1  leached
   HB Monlpached
   X  V I rgln carhons

   Filled  In synhols, bake and act.
   at 1700" to 1600°F
   Open symbols, bake and act. at
   1550" to 1700"F
                                         I   1   I    I
                                    I   I
     0.35
0.40            0.45
   Bulk  density,  g/cc
0.50
0.53
 FIGURE  20  - IODINE  NUMBERS  OF REGENERATED  DARCO,  F 400 AND
       WVP&P CARBONS AS  FUNCTION  OF BULK  DENSITIES
                             102

-------
ber.  Densities in the 0.460 to 0.484 g/cc range were  at-
tained but the high iodine number of the virgin carbon was
never reached.  A straight line through the nonleached car-
bon iodine number points indicates that the virgin carbon
iodine number cannot be attained by extrapolation of den-
sities in either direction.  It is felt that the parameters;
temperature, residence time, gas composition and input rate,
both during baking and activating have been adequately cov-
ered.  It is beginning to appear as an impossibility to re-
gain the original iodine number simultaneously with all
other initial properties of the carbon when the metallic
elements, brought in with the adsorbate, are in the carbon
during baking and activating.  When these were removed by
HC1 leach, the iodine number increased from the 930 to the
1030 mg/g level, which is still below the virgin carbon
iodine number.
                            103

-------
                       SECTION  VI

                    ACKNOWLEDGMENTS
This program was sponsored by the Environmental Protection Agency,
Water Quality Office,  Washington, D.C.; Project Officer was Mr. Arthur
N.  Masse, Chief, Municipal Treatment Research Program, Advanced Waste
Treatment Research Laboratory, Cincinnati, Ohio.  Negotiator was Mr.
John H. Blake, Water Quality Office, Washington, D.C.

The cooperative phase of  the  program at Pomona,  was  performed
by Messrs  John English  and Jay Pitkin.

The major  part of the investigation was performed  by MSA Re-
search  Corporation, Evans City, Pennsylvania, where  Dr.  J.W.
Mausteller, Associate Director of Research, is  responsible
for all  research activities at the  Evans City laboratories.
Mr. Frederick  Tepper was  Head of Physical Section  (to  Septem-
ber 1,  1969) in which this program  was carried  out.   The
principal  investigator  on the program was Dr. A.J.  Juhola.
Performance of the direct experimental work was  done by Mr.
Edward  Krieger.
                               105

-------
                       APPENDIX A
   INPUT AND OUTPUT GAS COMPOSITIONS OF ACTIVATION RUNS
Run
 32
 34
 36
 39
 40

Gas
N2
CO
C02
H2
H20

N2
cO
C02
"2
H20

N2
CO
i C02
H2
H20
£•
No
C§
1 C02
H2
H20
c.
No
cS
) C02
t.
H2
£
H20
N2

1 C02
H2
H20
Input
ft3/hr (stp)
4.00
	
0.51
	
1.11
5.62
7.16

0.81
	
1.84
978T
8.30
	
1.10
	
2.41
11.81
8.26

1.00
2.14
11.40
5.36

0.64
_ — -
0.95
679T
8.3

1.10
2.63
12.03

% Comp
71.2

9.1
	
19.7
100.0
73.0

8.2
	
18.8
100.0
70.3
	
9.3
	
20.4
100.0
72.5
	
8.8
18.8
100.1
77.1
	
9.2
	
13.7
100.0
68.9
_ - -
9.1
21.9
99T9
Output
ft3/hr (stp)
4.00
1.36
0.29
0.89
0.22
6776
7.16
0.78
0.81
0.73
1.11
10.59
8.30
0.99
1.23
1 .03
1.38
12.93
8.26
1.39
1.07
1.35
0.79
12.86
5.36
1.03
0.50
0.70
0.25
7.84
8.3
1.53
1.22
1.55
1.08
13.68

% Comp
59.2
20.1
4.3
13.2
3.3
100.1
67.6
7.4
7.6
6.9
10.5
100.0
64.1
7.6
9.5
8.0
10.7
99.9
64.2
10.8
8.3
10.5
6.1
99.9
68.3
13.1
6.4
8.9
3.2
99.9
60.7
11.2
8.9
11.3
7.9
100.0
                            107

-------
                 APPENDIX A (Continued)

                      Input
Run
 47
 51
 52
 53
 66
 67
Gas
 *
CO
COo
H2
H20
N-
Cl
C02
H2
H20
N2
CO
COo
Ho
H20
C
COo
H2
H20
 ?
CO
COo
H2
H20
N
C
CO
       H20
ft3/hr (stp)
5.80
0.77
1.70
8.27
3.33
0.41
0.74
4.48
3.24
0.44
0.57
4.25
3.12
0.41
0.93
4.46
4.00
0.55
1.49
6.04
4.00
2.81
6.81
% Comp
70.1
9.3
20.6
100.6
74.3
9.2
16.5
100.0
76.2
10.3
13.4
99.9
70.0
9.2
20.8
100.0
66.2
9.1
24.7
100.0
58.7
41.3
100.0
                                       Output
ft3/hr (stp)
5.80
0.76
0.84
0.74
0.96
9TTO"
3.33
0.90
0.26
0.54
0.20
5.23
3.24
1 .08
0.20
0.47
0.10
5.09
3.12
0.73
0.43
0.61
0.32
5.21
4.00
1.18
0.44
1.04
0.45
7.11
4.00
1.00
0.53
2.05*
0.76
8.34
% Comp
63.7
8.3
9.2
8.1
10.5
9~978
63.7
17.2
5.0
10.3
3.8
100.0
63.7
21.4
3.9
9.2
2.0
100.2
59.9
14.0
8.2
11.7
6.1
99.9
56.3
16.6
6.2
14.6
6.3
100.0
48.0
12.0
6.3
24.6
9.1
100.0
 *Estimated  from CO  and  C02  analyses, H2 concentration off
  scale  on  analyses
                           108

-------
                APPENDIX A (Continued)

             	Input	   	Output	
             ft3/hr (stp)   % Comp   ft3/hr (stja)   % Comp


                 9.22        73.3        9.22        67.2
                 ---         ---         1.07         7.8
69    C02        1.10         8.8        1.19         8.7
      H2         -"         ---         1.03         7.5
      H2°        2.24        17.8        1.21         8.8
                12.56        99.9       13.72       100.0

      No        10.14        72.2       10.14        67.7
      CO         ---         ---         0.84         5.6
70    C02        1.21         8.6        1.31         8.7
      H2         ---         ---         0.95         6.3
      H2°        2.69        19.1        1.74        11.6
                14.04        99.9       14.98        99.9

      No         3.3         73.5        3.3         64.4
      CO         ---         ---         0.71        13.9
76    C02        0.41         9.1        0.33         6.4
      H2         ---         ---         0.49         9.6
      H20        0.78        17.4        0.29         5.7
                 '  9       100.0        5TT2       100.0
      N2         7.6         68.2        7.6         62.1
      CO         ---         ---         1.16         9-5
77    C02        1.01         9.1        0.94         7.7
      H2         ---         ---         1.09         8.9
      H20        2.53        22.7        1.44        11.8
                11.14       100.0       12.23       100.0

      No         6.00        68.7        6.00        62.5
      CO         ---         ---         0.78         8.1
79    C02        0.80         9.1        0.88         9.2
      H2         ---         ---         0.84         8.7
                 1.94        22.2        1.10        11.5
                 8.74       100.0        9.60       100.0

      No         3.50        69.1        3.50        62.4
      CO         ---         ---         0.54         9.6
88    C02        0.43         8.5        0.44         7.7
WVP&P Ho         ---         ---         0.48         8.6
      HoO        1.13        22.3        0.65        11.6
       *         5.06        99.9        576T        99.9
                         109

-------
Run
89
WVP&P
90
98
Task
  2
104
Task
  2A
105
Task
  2B
Gas
 ?
CO
C02
H2
H20
CO
C02
H2
H20
N2
CO
C02
H2
H20
C
C02
H2
H20
C
C02
H2
H20
                 APPENDIX A (Continued)

                      Input
ft3/hr (stp)
3.50
0.43
1.14
T70T
3.50
0.43
1.25
5.18
1.27
0.17
0.12
1.56
2.19
0.37
1.12
3.68
2.19
0.37
0.81
3.37
% Comp
69.0
8.5
22.5
100.0
67.5
8.3
24.1
99.9
81.4
10.9
7.8
100.1
59.5
10.0
30.4
99.9
65.0
11.0
24.0
100.0
                                       Output
ft^/hr (stp)
3.50
0.55
0.44
0.54
0.60
5.63
3.50
0.55
0.54
0.63
0.62
5784
1.27
0.28
0.17
0.08
0.04
1.84
2.19
0.55
0.51
0.62
0.50
4.37
2.19
0.56
0.43
0.41
0.40
3.99
% Comp
62.2
9.8
7.8
9.6
10.7
100.1
59.9
9.4
9.2
10.8
10.6
99.9
69.0
15.2
9.2
4.3
2.2
99.9
50.1
12.6
11.7
14.2
11.4
100.0
54.9
14.0
10.8
10.3
10.0
100.0
                          110

-------
Run   Gas
106
WVP&P
N2
CO
C02
H2
H20
108   C(
WVP&P C02
      H.
                 APPENDIX A (Continued)

                      Input
ft3/hr (stp)
4.00
0.67
1.49
67T16
7.60
1.00
2.68
11 .28
% Comp
65.0
10.9
24.2
100.1
67.4
8.9
23.7
100.0
                                        Output
ft-Vhr (stp)
4.00
0.57
0.70
0.54
0.95
6.76
7.60
0.79
1.15
0.90
1.78
12.22
% Comp
59.1
8.4
10.4
8.0
14.0
9979
62.1
6.5
9.4
7.4
14.6
100.0
                          111

-------
                       APPENDIX B
       GAS ANALYSES WITH NO CARBON IN REGENERATOR
          TEST 1 - APPROXIMATE FLUE GAS MIXTURE
               THROUGH HEATED REGENERATOR

           	Ijiput	      	Output	
Gas     ft3/hr (stp)% Comp      ft3/hr (stp)     % Comp

No          7.0            73.0          7.0            73.0
CO          ---            ---           0.01             0.1
C02         0.93            9.7          0.92            9.6
H2          ---            ---           0.03            0.3
H?0         1.66           17.3          1.63           17.0
            9.59          100.0          9.59          100.0
      TEST 2 - APPROXIMATE FLUE GAS MIXTURE THROUGH
         HEATED REGENERATOR WITH HYDROGEN ADDED

No          7.0            65.7          7.00
CO          ---            ---           0.20
COo         0.93            8.7          0.72
Ho          0.79            7.4          0.58
HoO         1.93           18.1          2.14
 *         10.65           99.9         10.64


      TEST 3 - APPROXIMATE FLUE GAS MIXTURE THROUGH
      HEATED REGENERATOR WITH  0.90 ft3/hr CO ADDED

No          7.0            65.7          7.0            65.7
CO          0.90            8.4          0.49            4.6
COo         0.93            8.7          1.35           12.7
H *         .--            ---           0.38            3.6
HoO         1.83           17.2          1.45           13.6
 2         10.66           100.0         TOT          T0672
                           113

-------
                 APPENDIX  B (Continued)
      TEST 4 -  APPROXIMATE  FLUE  GAS  MIXTURE  THROUGH
      HEATED REGENERATOR WITH 0.50  ft3/hr CO ADDED

           	Input	       	Output	
Gas     ft3/hr  (stp)%  Comp       ft3/hr  (stp)      %  Comp

N2          7.0            69.7           7.0            69.7
CO          0.50            5.0           0.23            2.3
CO?         0.93            9.3           1.21           12.0
Ho          ---            ---            0.25            2.5
HoO         1.61           16.0           1.35           13.4
           10.04          100.0          10.04           99.9
      TEST 5 -  APPROXIMATE  FLUE  GAS  MIXTURE  THROUGH
      HEATED REGENERATOR  WITH  0.20  ft3/hr  CO ADDED

N2          7.0            70.8           7.0            70.9
CO          0.20            2.0           0.10            1.0
C02         0.93            9.4           1.00           10.1
Hp          ---            ---            0.17            1.7
H20         1.77           17.9           1.60           16.2
            "  "          100.1           9.87           99.9
                           114

-------
                      REFERENCES
1.   Juhola, A.J. and Tepper, F., "Regeneration  of Spent  Gran-
    ular Activated Carbon," Robert A.  Taft Water Research
    Center Report No. TWRC-7, February,  1969.

2.   Juhola, A.J., Matz, W.M., and labor,  J.W.,  paper pre-
    sented at the American Chemical Society meeting, Division
    of Sugar Chemistry, April 1, 1951, "Adsorptive Properties
    of Activated Carbons."

3.   Grant, R.J., "Basic Concept of Adsorption  on Activated
    Carbon", Pittsburgh Activated Carbon  Company.

4.   Wiig, E.O., and Juhola, A.J., JACS 71., 2069, 2078 (1949).

5.   Measurements made by American Instrument Company, Incor-
    porated, Silver Springs, Maryland.
                            115

-------
1

5
Accession Number
2

Subject Field & Group
10 A
SELECTED WATER RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
Organization
           Evans City, Pennsylvania
     Title
	       OPTIMIZATION OF THE REGENERATION PROCEDURE
          FOR GRANULAR ACTIVATED CARBON
1 Q Authors)
Juhola, A, J0
16

21
Project Designation
Program #17020
DAO
Note
                                           Report NOo  TWRC-7
                                           Contract  No.  14-12-107
 22
     Citation
           Contract Report, 106 pages, 41  tables,  20  figures
 23
Descriptors (Starred First)

      *Wastewater purification,  Tertiary treatment
 25
Identifiers (Starred First)
 *Thermal regeneration, *Carbon  regeneration losses,  Carbon regeneration,  Steam
 activation, C09 activation,  Carbon  pore  structure8  Chemical regeneration, Iodine
 number,.Molasses numbera Rotary tube  regenerator^Caustic regeneration, Carbonization,
 Activation-gas reactions,  Cyclic regenerations,  HCI  treatment,  Spent carbon drying
 27
     Abstract
            Spent granular activated carbons  from  tertiary water  treatment,  on multiple-
   hearth furnace regenerations, suffer a volume loss  of  5%  to  107,  per  regeneration.   On
   the first regeneration, activity loss is as high  as  137o but  diminishes  on subsequent
   cycles.   Laboratory studies to improve regeneration  have  demonstrated  that on regenera-
   tion of  wet spent carbon three steps occur;  (1) drying at about  220°F,  (2) pyrolysis of
   adsorbed pollutants at 500° to 1500°F(baking step)  and (3) activation with flue gas and
   steam at 1600° to 1700°F<,  Alkaline and iron oxides  accumulate in the  carbon and catalyze
                               Q       o
   oxidation of pores in the ISA to 28 A diameter  range.  When  metallic elements are  leached
   from the carbon, prior to regeneration, less carbon  and activity loss  occurs.  Steam
   regeneration is more effective than that with CC>2.   Carbon volume decreases during
   laboratory baking and activation average 2% and I087o,  respectively.  The  apparent
   volume decrease during baking is due to pyrolysis of colloidal pollutants on the
   particle surfaces; true carbon loss is then Io87>o   The latter  proceeds  as submicron
   fines  formation from particle surfaces.  Regeneration  attempts by leaching with
   solutions of NaOH, H_0_ and CC1.  were ineffective and/or  uneconomical.  Report contains
   41  tables and 20 figures.
Abstractor
  Juhola_j_A»  Jo
                          Institution
                           Mine  Safety  Appliances
Research Corporation
 WR:102  (REV. JULY 1969)
 WRSIC
                                          SEND TO: WATER RESOURCES SCIENTIFIC INFORMATION CENTER
                                                 U.S. DEPARTMENT OF THE INTERIOR
                                                 WASHINGTON, D. C, 20240
                                                                               * CPO: 1969-359-339

-------