EPA-600/2-76-135a
May 1976
Environmental Protection Technology Series
                 DEVELOPMENT OF  THE WESTYACO
                 ACTIVATED  CARBON  PROCESS  FOR
            SOX RECOVERY AS ELEMENTAL  SULFUR
                                             Volume!
                               Industrial Environmental Research Laboratory
                                    Office of Research and Development
                                   U.S. Environmental Protection Agency
                              Research Triangle Park, North Carolina 27711

-------
                RESEARCH REPORTING SERIES

Research reports of the Office of Research and Development, U.S. Environmental
Protection  Agency,  have  been grouped  into five  series. These five broad
categories  were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:

     1.    Environmental Health Effects Research
     2.    Environmental Protection Technology
     3.    Ecological Research                           v
     4.    Environmental Monitoring
     5.    Socioeconomic Environmental Studies

This report has  been  assigned  to the  ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to develop and
demonstrate instrumentation, equipment, and methodology to repair or prevent
environmental degradation from point and  non-point sources of pollution. This
work provides the new  or improved technology required for the control and
treatment of pollution sources to meet environmental quality standards.
                    EPA REVIEW NOTICE

This report has been reviewed by  the U.S.  Environmental
Protection Agency, and approved for publication.  Approval
does not signify that the contents necessarily reflect the
views and policy of the Agency, nor does mention of trade
names or commercial products constitute endorsement or
recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

-------
                                         EPA-600/2-76-135a

                                         May 1976


      DEVELOPMENT OF THE WESTVACO

         ACTIVATED CARBON  PROCESS

FOR  SOX RECOVERY AS ELEMENTAL  SULFUR

                     VOLUME  I
                          by

     G. Nelson Brown, Carl M. Reed, Albert J. Repik,
       Robert L. Stallings,  and Samuel L.  Torrence

                       Westvaco
                       Box 5207
         North Charleston, South Carolina 29406
                Contract No. 68-02-0003
                 ROAPNo. 21ACX-085
              Program Element No. 1AB013


        EPA Project Officer: Douglas A. Kemnitz

       Industrial Environmental Research Laboratory
         Office of Energy, Minerals, and Industry
            Research Triangle Park, NC  27711


                     Prepared for

      U.S. ENVIRONMENTAL PROTECTION AGENCY
            Office of Research and Development
                 Washington, DC 20460

-------
                      TABLE OF CONTENTS
Table of Contents                                         iii

List of Figures                                            v

List of Tables                                            ix

Acknowledgements                                          xiii
Management Summary                                        xiv

Sections

  1       CONCLUSIONS                                     1
  1.1       Overall Process                               1
  1.2       Integral Pilot Plant                          1
  1.3       Bench Scale                                   2
  1.4       Pilot Scale                                   3
  1.5       Prototype                                     4
  1.6       Commercial (1,000 MW)                         4

  2       RECOMMENDATIONS                                 5

  3       INTRODUCTION               _                    6
  3.1       Process Concept                               7
  3.2       Methodology of Contract                       9
  3.3       Chronological Sequence of Development        10
  3.3.1       S02 Sorption                               11
  3.3.2       Sulfur Generation                          11
  3.3.3       Sulfur Recovery and H2S Generation         11
  3.3.4       Integration                                12

  4       INTEGRATED PILOT PLANT EQUIPMENT AND RESULTS   13
  4.1       Pilot Plant Description                      13
  4.1.1       Introduction                               13
  4.1.2       Detailed Pilot Plant Description           16
  4.2       Integral Pilot Plant Results                 26
  4.2.1       Overall Integral Results                   28
  4.2.2       Detailed Integral Results                  36
  4.2.3       Material Balances                          53
  4.2.4       Process Control                            56
  4.2.5       Process Concept Modifications              59

  5       PRE-INTEGRAL PROCESS DEVELOPMENT               61
  5.1       Apparatus and Procedure                      61
  5.1.1       Thermogravimetric Reactor                  61
  5.1.2       Fixed Bed                                  66
  5.1.3       Moving Bed                                 66
  5.1.4       Batch Fluid Bed                            69
  5.1.5       Multistage Fluid Bed Reactor               71
                                ill

-------
              TABLE OF CONTENTS  (Continued)
5.1.6       Sulfur-Carbon Thermal Equilibrium           71
5.1.7       Solvent Extraction of Sulfur Procedures     73
5.1.8       Procedures in Bench Scale H2S               77
               Generation Studies
5.2       Pre-Integral Results                          81
5.2.1       S02 Sorption                                81
5.2.2       Sulfuric Acid Conversion to Sulfur         114
5.2.3       Sulfur Removal                             154
5.2.4       H2S Generation                             184
5.2.5       Combined S Stripping/H2S Generation        206
5.2.6       Elemental Sulfur Recovery                  220
5.2.7       Fluidizing Mechanics                       226

6       1,000 MW UTILITY BOILER FLUE GAS CLEAN-UP      236
6.1       Introduction                                 236
6.2       General Design Basis                         238
6.2.1       Scope                                      238
6.2.2       Boiler Operating Characteristics           239
6.2.3       Product                                    240
6.2.4       Process Conditions                         240
6.2.5       Activated Carbon Characteristics           242
6.3       Conceptual Design                            242
6.3.1       Process Description                        242
6.4       Heat and Material Balances                   245
6.5       Costs of 1,000 MW Conceptual Design          257
             Installation
6.5.1       Cost Summary                               257
6.5.2       Capital Costs                              257
6.5.3       Equipment Costs                            257
6.5.4       Indirect Costs                             258
6.6       Operating Costs of 1,000 MW Conceptual       258
             Design Installation

7       15 MW DESIGN AND COST                          261
7.1       Introduction                                 261
7.2       Scope of the Prototype Program               261
7.3       Description of Prototype Plant and           262
             Operation
7.3.1       General                                    262
7.3.2       General Design Basis                        263
7.3.3       Process Description - Prototype Plant      265
7.3.4       Heat and Material Balance                  267
7.3.5       Start-up and Initial Operation             268
7.3.6       Demonstration Operation                    268
7.3.7       Technical and Economic Review of           268
               Operation
                            iv

-------
              TABLE OF CONTENTS (Continued)

                                                      Page

7.4       Technical Approach                           269
7.4.1       General                                    269
7.4.2       Description of Program Elements            269

8       Bibliography              '                     273

9       Nomenclature                                   275

-------
                        LIST  OF  FIGURES

 No.                                                     page

  1    Chemistry  of  Westvaco S02  Process with  Reactants      8
         and Products  Shown
  2    Westvaco Process Integral  Pilot  Plant        .        14
  3    Mechanical Integration  - 20,000  CFH  S02 Pilot        17-A
         Plant - Process Flowsheet
  4    Continuous 18" Dia., 5  Stage  S02 and S03            19
         Adsorber Operating on Flue Gas from  a 50 MW
         Oil Fired  Boiler
  5    Schematic  of  Integral Westvaco S0£ Removal Pilot     27
         Plant
  6    S02 Removal Efficiency  During Integral  Pilot       <  31
         Tests
  7    Carbon Burn-off  During  Integral  Pilot Tests          32
  8    Activated  Carbon Attrition Rate  During  Integral      33
         Pilot Tests
  9    Activated  Carbon Performance  During  Westvaco S02     41
         Recovery Integral Pilot Runs
 10    S02 Activity  as  a Function of Carbon Cycle Time      42
         as  Determined by Bench  Scale  Apparatus
 11    Carbon Attrition,  Mean  Particle  Diameter, and Ash    44
         Content as a  Function of Carbon Cycle Time
 12    Carbon Dioxide Evolution as a Function  of Carbon     46
         Cycle Time
 13    Pore Volume and  Surface Area  of  Recycled Carbon      48
 14    Sulfur Generator Performance                         49
 15    H2S Generator/Sulfur Stripper Performance            51
 16    Sulfur Condenser Performance                         52
 17    Sulfur Balance for IR-2 Run During Operation         57
         under Process H2S
 18    Thermogravimetric Apparatus                          62
 19    Detail of  the Thermogravimetric  Reactor Sample       64
         Bucket  Envelope
 20    Fixed  Bed  Reactor System                             67
 21    Moving Bed Reactor System                            68
 22 •   Batch  Fluid Bed  Reactor                             70
 23    Multistage Fluidized Bed Reactor                    72
 24    Sulfur Adsorption Apparatus                          74
 25    Flow Schematic of Recycle  Extraction Apparatus       76
 26    H2S  Generation Kinetics Apparatus                    78
 27    Sulfur Vapor  Generator                               79
 28    Comparison of Westvaco  Model  to  Sorption Data at     85
         200°F
 29    Effect  of  02  on  S02 sorption  at  200°F with NO        88
        Present
30    Effect  of  H20 Concentration on S02 Sorption at       90
        200°F with NO Present
31   Effect  of  NO  Concentration on S02 Sorption at        91
        200°F

-------
                 LIST OF FIGURES  (Continued)

No.

32   Effect of Temperature on the Westvaco Model with
        Constant Order of Reaction for S02
33   Rate Constant as a Function of Temperature for
        S02 Sorption
34   Comparison of the Westvaco Model A to Experi-
        mental S02 Sorption on Activated Carbon in
        Differential Rate Apparatus
35   Comparison of the Westvaco Model B to Experi-       103
        mental S02 Sorption on Activated Carbon in
        Differential Rate Apparatus
36   Comparison of the Westvaco Model C to Experi-       104
        mental S02 Sorption on Activated Carbon in
        Differential Rate Apparatus
37   Differential S02 Sorption Rate vs. H2S04 Loading    106
        for a S02 Concentration of 2500 ppm at 150°,
        200°, and 300°F
38   6" Sorber Data - Plot of Corrected Sorber Rate      109
        using Stagewise Westvaco Model A vs. S02
        Cone. Showing Curve Predicted from
        Differential Bed Studies
39   Summary for Flue Gas Run (Run SA-34) - 18" Dia.     113
        S02 Sorber - Water Sprays To Control Temp.
40   Effect of Linear Gas Velocity on Rate of            116
        Sulfuric Acid Decomposition
41   Effect of Temperature on the Rate of Conversion     120
        of Sorbed Sulfuric Acid to Elemental Sulfur
42   Comparison of the Sulfur Generation Rate Model      124
        to the Experimental Data for 250° to 325°F
43   Effect of H20 Cone, on Rate of Sulfur Generation    125
44   Effect of Inlet H2S Cone, on Per Cent Conversion    130
        to Sulfur in Simulation Experiments Using a
        6" Diameter Fluid Bed Unit for Integrated
        Operation with an 18" Diameter S02 Sorber
45   Effect of H2S Cone, and Carbon Residue Time on - .   131
        Acid Evolved as S02 in Simulation Experiments
        Using a 6" Dia. Fluid Bed Unit for Integrated
        Operation with an 18" Dia. S02 Sorber
46   Moving Bed Sulfur Generator                         139
47   Effect of Inlet Carbon Temp, on the Evaluation      144
        of H2S04 as S02 in an 8" Dia. Moving Bed
        Reactor
48   Effect of- Inlet Carbon Temp, on H2S Utilization     145
        in an 8" Dia. Moving Bed Reactor
49   Effect of the Inlet Carbon Temp, on the Per Cent    146
        Conversion to Sulfur
50   Effect of Steam Heater on Improved Carbon           148
        Heating Capabilities
                              vii

-------
                 LIST OF FIGURES  (Continued)

No.                                                     Page

51   Effect of Heat Exchanger System  (1-1/2" Pipe        149
        x 4") on Carbon Flow in an 8" Dia. Moving
        Bed Reactor - Tracer Feed Composition)
52   Experimental Adsorption Isotherm Points for         157
        Sulfur on Activated Carbon at 800° and 1000°F
53   Polanyi-Dubinin Plot of Sulfur Adsorption Data      158
54   Equilibrium Lines for Concentration vs.             160
        Temperature at Various Loadings
55   4" Dia. Batch Fluidized Bed S Stripping and         163
        Hydrogen Desulfurization Runs
56   Extraction of Sulfur Loaded Activated Carbon        166
        with 15 Wt. % (NH4)2S Solution at 40°C
57   Extraction of Sulfur Loaded Activated Carbon        168
        with CS2 at 25°C
58   Extraction of Sulfur Loaded Activated Carbon        169
        with Xylene at 105°C
59   Effect of Percent Sulfur on Carbon on the S02       175
        Activity
60   Effect of Recycle on S02 Ability for Isothermal     178
        and Thermal/Reductive Regenerations
61   Effect of Treatment Time Using Hydrogen Post        182
        Treatment of (NH4)2S Extracted Sample of
        Sixth Cycle
62   Effect of Temperature of Hydrogen Post Treatment    183
        of (NH4)2S Extracted Sample of Sixth Cycle
63   Equipment Schematic for H2 Chemisorption            188
        Experiments on Virgin Carbon
64   Arrhenius Plots for Experimental and Literature     197
        Data
65   Test of Integral Rate Equation, Z as Function of    203
        Tube Volume
66   Test of Integral Rate Equation, Z as Function of    204
        Tube Area
67   Variation of Conversion with Residence Time for     207
        Three Bed Volumes
68   Effect of Flow Rate on Conversion at Constant       208
        Residence Time
69   Effect of Temperature on Sulfur Stripping with      214
        H2 Percent
70   Effect of H2 Concentration and Gas/Solid Contact    215
        Time on Sulfur Stripping at 1200°F
71   Sulfur Removal from Activated Carbon in an 8        217
        Stage,  4" Dia.  Regenerator
72   Effect of Temperature on the Conversion of Sulfur   218
        to Hydrogen Sulfide
                             viii

-------
                 LIST OF FIGURES (Continued)

No.                                                     Page

73   Operating Conditions for Sulfur Condenser           225
74   Experimental Determination of Minimum Fluidizing    228
        Velocity for Westvaco Granular Carbon
75   Pressure Drop Characteristics of Distributor        230
        Plates To Be Used in an 18" Dia. S02 Sorber
76   Westvaco S02 Recovery Process Schematic             237
        Flowsheet (Dwg. 2563)
77   Westvaco S02, Process Flowsheet for 1,000 MW Unit    243-A
        (250 MW Typical Module Shown) (Dwg. 2572)
78   Westvaco S02 Process Flowsheet for 15 MW            266
        Prototype Unit (Dwg. 2573)
79   Prototype Program Schedule                          270
                               ix

-------
                       LIST OF TABLES

No.

  1   Range of Operating Conditions for  Integral Pilot     29
        Plant Run
  2   Properties of  Sulfur Product                         31
  3   Effect  of Hydrogen Input on By-product Recovery      35
  4   Integrated Operating Conditions and Results          37
  5   Process Operating Performance                        38
  6   Carbon  Balance for IR-2 Integral Run                 53
  7   Hydrogen Balance for H2S Generator/S Stripper        54
  8   Sulfur  Balance for Integral Runs                     55
  9   Rate Expressions To Approximate S02 Sorption         82
        Data
10   Experimental Conditions for S02 Sorption in a        83
        Differential Rate Apparatus
11   Standard Deviation for Westvaco Equation for         84
        Sorption Data at 200°F with NO Present for
        Acid Loading above 0.01 gm. Acid/gm. Carbon
12   Deviation of Westvaco Equation from Experimental     86
        S02  Sorption Rate Data at 200°F with NO for
        Loadings above 0.01 gm. Acid/gm. Carbon
13   Standard Deviation for Modified Westvaco Equation    86
        for  Sorption Data at 200°F with NO for Acid
        Loading above 0.01 gm. Acid/gm. Carbon
14   Experiments To Determine 02 Dependency               87
15   Experiments To Determine H20 Dependency              89
16   Experiments To Determine NO Dependency               89
17   Experiments To Determine Effect of Temperature on    93
        S02 Sorption
18   Rate Constants for the Westvaco Model                93
19   Deviation of Westvaco Model from Differential        97
        Rate Data for Acid Loadings above 0.01 gm.
        Acid/gm. Carbon
20   Deviation of Multiple Regression Models from        102
        Differential Rate Data for Acid Loadings
        above 0.01 gm. Acid/gm. Carbon
21   Comparison of Rates from 6" Dia. Sorber to Rates    107
        Calculated from the Westvaco Model
22   Comparison of Predicted Number of Stages to         108
        Actual Number for 6" Sorber Runs
23   Water Spray Cooling Tests Made in Pilot Fluid Bed   111
        Reactors with Simulated and Actual Flue Gas
24   Experimental Conditions and Results for Sulfur      118
        Generation Experiments in an 8 Stage, 4" Dia.
        Fluidized Bed Regenerator
25   Overall Rates of Acid Decomposition and Conversion  119
        to Sulfur for the Reaction
        3 H2S + H2S04 -»• 4 S + 4 H20 in an 8 Stage, 4"
        Dia.  Regenerator
                               x

-------
                 LIST OF TABLES  (Continued)

No.                                                     Page

26   Experimental Conditions for Differential Sulfur     122
        Generation Runs
27   Summary of Sulfur Generation Results                129
28   Comparison of the Fluid Bed Design Model with       133
        Experimental Sulfur Generation Fluid Bed Data
29   Fixed Bed Sulfur Generation Experiments             134
30   Comparison of Fluid Bed and Moving Bed Sulfur       136
        Generation Tests
31   Data Summary - 1-1/2" Dia. Moving Bed Sulfur        137
        Generation Tests
32   Design Conditions - Moving Bed Sulfur Generator     138
33   Summary of Sulfur Generation Results                141
34   Effect of Vol. % H20 and Temperature on the Cone.    151
        of Acid Solution Sorbed on Carbon for Carbon
        Preheater
35   Comparison of the Moving Bed Design Model with      153
        Experimental Sulfur Generator Moving Bed Data
36   Experimental Results of Equilibrium Sulfur          156
        Adsorption Measurements
37   Isosteric Heats of Adsorption of Sulfur Vapor on    161
        Carbon
38   Sulfur Stripping in a Continuous 8-Stage Fluid      162
        Bed
39   Effect of Solvent on Virgin Carbon                  165
40   Comparison of S02 Activity and Surface Area and     171
        Pore Volume Measurements
41   Effect of Recycle on S02 Activity for Isothermal    176
        and Thermal/Reductive Regenerations
42   Pore Volume Distribution Results Using Engelhard    179
        Isorpta Apparatus
43   S02 Activities Integral Rate Determined Using       179
        Differential Rate Apparatus vs.  Using Fixed
        Bed
44   Effect of Post Treatments of (NH4)2S Extracted      181
        Sixth Cycle Carbon on S02 Activity
45   Planned Experimental Program for Studying Hydrogen  186
        Hydrogen Chemisorption on Activated Carbon
        During Regeneration of C
46   H2 Chemisorption on Virgin Carbon                   187
47   Experimental Conditions for H2S Generation Rate     190
        Experiments
48   Comparison of H2S Formation Rates from Literature   191
        and Experimental Data  (Homogeneous)
49   Comparison of Homogeneous and Heterogeneous         192
        Reaction Rates for Similar Inlet Concentrations
50   Experimental Conditions for Series HS-2             194
                              XI

-------
                 LIST OF TABLES  (Continued)

No.                                                     Page

51   Comparison of Rate Constants at Different           195
        Temperatures
52   Comparison of Rates and Rate Constants Based on     198
        Reactor Surface Area from Series HS-2
53   Experimental Conditions for Runs HS-4 to HS-7       205
54   Experimental Conditions for Evaluation of           211
        Combined S Stripping/H2S Generation
55   Experimental Conditions for Evaluation of           212
        Combined S Stripping/H2S Generation
56   Experimental Results from Evaluation of Combined    213
        S Stripping/Has Generation
57   Operating Conditions for Sulfur Condenser           221
        Testing System
58   Sulfur Condenser Test Runs                          222
59   Sulfur Condenser Operation                          223
60   Calculated Values for Minimum Fluidizing            227
        Velocity and Entrainment Velocity for Westvaco
        Granular Activated Carbon
61   Operating Characteristics Distributor Plates To     232
        Be Used in an 18" Dia. S02 Sorber
62   Gas Distributor Plate Characteristics Evaluated     233
        for Carbon Weeepage During Fluidization
63   Gas Distributor Plate Specifications Designed for   235
        Minimizing Carbon Attrition in the 18"0 S02
        Sorber
64   Overall Sulfur Balance for 1,000 MW Power Plant     245
65   Overall Energy Balance for 1,000 MW Power Plant     246
66   Stream Conditions                                   247
66   Stream Conditions                                   248
68   Stream Conditions                                   249
69   Stream Conditions                                   250
70   Stream Conditions                                   251
71   Stream Conditions                                   252
72   Stream Conditions                                   253
73   Stream Conditions                                   254
74   Stream Conditions                                   255
75   Stream Conditions                                   256
76   Cost Summary                                        257
77   Capital Cost Summary                                258
78   Annual Operating Costs                              260
79   Cost of Prototype Program                           272
                             xii

-------
                      ACKNOWLEDGEMENTS


The authors wish to acknowledge the assistance of Doug Kemnitz
and Leon Stankus, Project Officers, whose guidance has been a
significant factor in carrying out this demonstration program.
We also would like to thank all the other individuals at EPA
who have contributed their time and effort to this project.

We would like to acknowledge the contributions made by
Dr. Frank J. Ball, Associate Corporate Research Director for
Westvaco.  His guidance, suggestions and enthusiasm have
played a significant role in the development of the process.

We would like to also acknowledge the numerous contributions
made by many individuals at Westvaco.  Listed alphabetically:
E. A. Ankersen, E. C. Arms, N. L. Davis, R. Deleon, R. C. Flowe,
B. J. Gooch, C. E. Grooms, W. D. Heape, H. R. Johnson,
E. B. Lipscomb, G. F. McAllister, C. C. Matthews, M. Nelson,
W. C. Rhodes, R. L. Roquemore, C. B. Rowe, C. Smith,
H. E. Sparks, R. A. Stanton, I. A. Stine, S. R. Thompson,
E. D. Tolles, B. H. Van Dyke, W. A. Wier, R. M. Wise.  Many
other individuals at Westvaco were involved and thanks are
also extended to them.

Lastly the authors wish to acknowledge the efforts of
L. K. Hallex who was primarily responsible for typing and
editing this report.  Her continuing effort was also signifi-
cant in the many reports and pieces of communication prepared
during the project.
                             xiii

-------
                     MANAGEMENT SUMMARY
An all dry  fluidized bed S02 recovery process using activated
carbon has  been developed by the Westvaco Corporation to
recover  sulfur oxides from waste gases with elemental sulfur
as a product.  The process was developed for five years by
Westvaco before joint development with the Environmental
Protection  Agency  (EPA) began in January 1971.  During a
pause in the contract from December 1971 to September 1972,
Westvaco continued development of the process, and then from
September 1972 to June 1974 joint development continued with
EPA.  The process has been shown to be technically feasible
and economically attractive compared to other processes cur-
rently being developed.  Future plans are, therefore, to
demonstrate the process at the next stage of development,
presently anticipated as the equivalent of a 10 to 20 mega-
watt coal fired power plant.

In the process the activated carbon catalyzes the oxidation
of the sulfur oxides in the flue gas and adsorbs these con-
stituents as sulfuric acid.  The carbon is regenerated and
the sulfuric acid is reduced to elemental sulfur in unit
operations  separated from the flue gas clean-up.  These two
steps of regeneration use a hydrogen containing gas to regene-
rate the carbon and reduce the sulfuric acid to elemental
sulfur by way of an intermediate reductant, hydrogen sulfide,
which is produced and recycled within the regeneration system.
The regenerated carbon is reused for flue gas clean-up and the
elemental sulfur is recovered to be stored or sold.

The demonstration of the activated carbon process for recover-
ing sulfur  oxides as sulfur has been completed at an integral
pilot plant stage during which a slipstream of flue gas from
a 50 MW  oil fired boiler was treated.  The sulfur oxides were
recovered as elemental sulfur of over 99.77o purity for over
300 hours.  During the integral run, the activated carbon was
continuously regenerated and reused more than 20 times for
flue gas clean-up and the carbon maintained its activity with
no decrease observed, as indicated by the average removal of
94% of the  2000 ppm sulfur oxides.

The hydrogen requirement of the process was found to be 3.9
moles of hydrogen per mole of sulfur oxide recovered.  This
hydrogen requirement is 30% above that required by stoichi-
ometry   for the reduction of sulfuric acid to elemental sulfur.
The recovery process was operated successfully as a closed
loop system, as anticipated in commercial operation.
                             xiv

-------
A considerable amount of process development preceded
the integral demonstration of the process.   The major
technical developments were that a granular activated
carbon was chosen, the process chemistry established,
and the variables affecting the process specified.  In
addition, the SOo removal process step was demonstrated
in a multistage fluid bed reactor treating a simulated
flue gas and an actual flue gas stream from an oil-fired
boiler.  A satisfactory design procedure was developed
for this process step.  The acid conversion to elemental
sulfur was also demonstrated in multistage fluid bed
equipment and a design procedure was developed.
Mechanical constraint, however, necessitated that a
moving bed reactor be used in the integral pilot plant
run.  Several methods of sulfur recovery were assessed,
but the stripping of the sulfur with a ^-containing gas
appeared to be most suitable to sustain a high level of
S02 activity of the regenerated carbon.  The sulfur re-
moval and internal H2S generation steps which finally
evolved were demonstrated as a combined operation in one
multistage fluid bed reactor.  Sufficient information was
developed on the carbon regeneration step to, hopefully,
insure satisfactory scale-up to the next level of
anticipated development.

All of this process and design information which was
developed has permitted scale-up to a prototype unit of 15
MW.  The preliminary 15 MW demonstration program includes
installation on a coal fired boiler and the use of a coal
fed gas producer to supply the necessary reducing gas.
The installation is anticipated to cost about $2.4 million
to install.  The demonstration of the process at the proto-
type scale is anticipated to require 3 years at a total
operating cost of $1.4 million, or about $0.5 million/
year.

An economic assessment was also made for the recovery
process installed at a 1000 MW power plant.  The capital
investment was estimated to be $35/KW and the annual
operating cost was estimated to be 2.0 mills/KWH.  Even
though the costs of the process may increase after
additional process development information is obtained
at the prototype scale, these costs indicate that the
process is worthy of consideration for continued process
demonstration.

It is concluded that the technical viability and economic
competitiveness of the sulfur oxide process developed by
Westvaco with partial funding from EPA has been demonstrated.

It is recommended that the process be demonstrated at a
prototype stage on a coal fired boiler.
                            xv

-------
                         SECTION 1
                        CONCLUSIONS


       The general objectives of the contract were to develop
       further information on each process step of the
       Westvaco S02 Recovery Process, to demonstrate the tech-
       nical feasibility of the entire process, to evaluate
       the performance of the carbon under extended recycling
       conditions in an integrated pilot plant using flue gas
       from an oil fired boiler, and to scale up the process
       to a larger power plant installation.


1.1    CONCLUSIONS - OVERALL PROCESS

       1.  The process has shown to be technically feasible
           and economically competitive.

       2.  The process can use activated carbon to
           effectively remove flue gas S02 and 803 and
           recover elemental sulfur by-product.

       3.  Sufficient information has been generated on the
           performance of the activated carbon, process
           chemistry and pilot operation to proceed to the
           next stage of development.


1.2    CONCLUSIONS - INTEGRAL PILOT PLANT

       The basic conclusions reached in the integral pilot
       plant are:

       1.  Granular activated carbon of the type used can
           effectively remove S02 from flue gas and can be
           regenerated satisfactorily over a repeated number
           of cycles without reduction in activity or an
           unacceptable physical loss through chemical reac-
           'tion or mechanical attrition.

       2.  Information has been developed on' each of the
           three unit process steps, S02 sorption, sulfur
           generation, and S stripping/H£S generation, to
           define the principal variables affecting the
           process chemistry and their correlations in regard
           to rate of reaction.

       3.  An acceptable sulfur product can be produced by the
           process using hydrogen as a reducing gas with H2S
           as an internally generated intermediate reductant.

-------
       4.   Carbon burn-off is minimized,  in fact almost  elimi-
           nated, in the S02 recovery process  as proposed.

       5.   Mechanical attrition of the carbon  was acceptable
           with the improved carbons used in these tests and
           further reduction by a factor  of 3  or more is
           indicated.

       6.   Complete reduction of sulfuric acid is not required
           in the sulfur generator,  but part of the acid can
           be effectively reduced to elemental sulfur in the
           sulfur stripper/H2S generator  without undesirable
           side effects, such as an increase in carbon burn-off.

       7.   The hydrogen requirement of the process was 3.9
           moles of H£ per mole of sulfur oxide recovered for
           the pilot plant conditions tested.

       8.   Use of fluidized beds present  a viable and attractive
           method of gas-solids contacting,  although other
           contacting means are also applicable.

       9.   Operation of the integral pilot plant over the
           limited time did not appear to present any problems
           in regard to control of the process.


1.3    CONCLUSIONS - BENCH SCALE

       1.   The S02 removal kinetics are a function of tempera-
           ture, of gas concentrations of 02,  H20, S02 and  NO
           and of the acid loading on the carbon.

       2.   The S02 kinetics developed from a bench scale differ-
           ential reactor could be modeled by  an empirical
           expression and incorporated into a  procedure  for
           reactor design.

       3.   Sulfuric acid adsorbed on the  carbon can be con-
           verted to elemental sulfur by  reaction with H2S.

       4.   The kinetics of sulfur generation are functions  of
           acid concentration adsorbed on carbon, of tempera-
           ture, and of gas concentrations of  H2S and H20.

       5.   The kinetic data of sulfur generation can be  repre-
           sented by an empirical expression and can be  incor-
           porated into a procedure for reactor design.

-------
       6.   Sulfur adsorbed on carbon can be recovered by
           solvent extraction, but the S02 activity of the
           regenerated carbon can only be maintained at a high
           level if it is treated further by thermal means.

       7.   Sulfur adsorbed on carbon can be removed by vapori-
           zation and the S02 activity is maintained,  especially
           when exposed also to a hydrogen containing gas at
           the temperature the sulfur is vaporized from the
           carbon.

       8.   Equilibrium data of sulfur adsorbed on carbon can
           be represented by an empirical expression known as
           the Polanyi-Dubinin adsorption equation.
                                  *
       9.   The kinetics of H2S generation over an activated
           carbon catalyst can be represented by an empirical
           expression.


1.4    CONCLUSIONS - PILOT SCALE

       1.   Multistage fluid bed S02 sorber can be effectively
           used with activated carbon to recover S02 and 803
           from flue gas of an oil fired boiler.

       2.   Direct flue gas cooling by water spray injection
           after S03 removal can be accomplished in a fluid
           bed reactor to improve S02 removal efficiencies and
           minimize gas cooling costs.

       3.   A moving bed reactor for acid conversion to sulfur
           used in the integral pilot plant because of mechani-
           cal limitations associated with a fluid bed reactor
           was satisfactory for acid conversion.

       4.   The process steps of sulfur stripping and H2S genera-
           tion can be combined into a single operation.

       5.   Operating temperatures for the existing sulfur
           stripper/H2S generator near 1000 to 1200°F are indi-
           cated based on sulfur removal from the carbon.

       6.   A sulfur condenser utilizing recirculating sulfur
           for a scrubbing fluid is suitable for use in inte-
           gral operation to recover sulfur from the H2S
           recycle gas to the acid converter.

       7.   A shell and tube cooler is suitable to cool regene-
           rated carbon to be reused in the S02 sorber.

-------
1.5    CONCLUSIONS - PROTOTYPE

       1.  The equivalent of a 15 MW power plant appears to be
           a suitable size to which to scale up the process.

       2.  The process information developed to date was suf-
           ficient to design an installation of the S02 recovery
           process at a 15 MW power plant.


1.6    CONCLUSIONS - COMMERCIAL (1,000 MW)

       1.  The S02 Recovery Process appears to be economically
           competitive with other processes now being developed
           for S02 recovery.

-------
                  SECTION 2
               RECOMMENDATIONS
1.   Based on conclusions from the integral pilot run
    and process development to date,  it is recommended
    that scale-up to a larger prototype plant be
    pursued as the next step toward a commercial
    plant.

2.   The prototype unit should be installed on a coal
    fired boiler.

3.   Provisions should be made to test the prototype
    installation for use both upstream and downstream
    of the precipitator.

-------
                  SECTION 3
                INTRODUCTION
The use of activated carbon in dry regenerable S02
recovery processes avoids the critical control of chemi-
cal reactions necessary in wet processes and the costs
and problems involved in separating water from
by-product, either for recovery or disposal.  Addition-
ally, in the wet processes, flue gas reheating may be
necessary for fan protection and plume control.  Carbon
processes which have or are being used all depend upon
the catalytic and sorptive character of the carbon for
conversion of the S02 to sulfuric acid within the carbon
granules.  These'processes generally differ in the mode
of removal and recovery of the sulfuric acid from the
carbon.  In thermal regeneration, the acid reacts
chemically with the carbon to produce a S02 rich
by-product off-gas.  In wet regeneration, the acid
loaded carbon is washed with water to produce a weak
sulfuric acid.  Further differences exist in the addi-
tional methods of upgrading the by-product streams
through add-on steps for conversion of the S02 gas
stream to elemental sulfur or concentrated sulfuric
acid.  The method of contacting flue gas with granular
carbon also varies in that fixed beds or moving beds
with an upflow or crossflow gas pattern are used.
Particle size and characteristics of the carbon granules
with respect to the rate of S02 removal may differ,
affecting pressure requirements and equipment size.

Westvaco, as a major producer of activated carbon,
embarked on a program in which carbon, with a high S02
pickup rate capability, is recycled with regeneration
of the carbon achieved by reducing the sulfuric acid
chemically within the process to elemental sulfur and
without the carbon being consumed.  Furthermore, the
fluidized bed was selected for gas-solid contacting as
a stagewise approach to permit handling relatively
large volume rates of gases in contact with recirculat-
ing carbon solids.  The effectiveness of fluidized
carbon bed systems has been demonstrated in large
commercial units in existence,  handling gas rates up
to 540,000 cfm.   The feasibility of using such a carbon
system was confirmed in bench scale and small pilot
equipment whereby H2S in contact with the sulfuric acid
on the carbon resulted in conversion to elemental
sulfur which was then stripped off the carbon by heating
                        6

-------
       and then recovered as a 99.770 pure product.  An outside
       source of hydrogen was reacted with sulfur within the
       regeneration system to produce the needed H2S.

       The selection of a granular activated carbon and
       identifying the major variables in the process chemistry
       served as a basis for the joint work under this EPA con-
       tract which essentially involved scaling up the SC>2 re-
       moval sorption and the regeneration-sulfur recovery
       steps to a 20,000 cfh pilot plant.   The objectives of the
       contract were initially to develop further information
       on each process step and finally to demonstrate the
       technical feasibility of the entire process and to
       evaluate the performance of the carbon under extended
       recycling conditions in an integrated pilot plant using
       flue gas from an oil fired boiler.


3.1    PROCESS CONCEPT

       In the Westvaco Process, dry granular activated carbon is
       contacted with flue gas at stack gas temperatures.   The
       SO? is removed through catalyzed oxidation to 803 and
       subsequent hydrolysis to sulfuric acid which remains
       sorbed in the carbon granules, i.e.
                 S02 + 1/2 02 + H20 .   arko     H2S04 (Sorbed)        (1)
       Sufficient water vapor and oxygen are present normally
       in the flue gas for the reaction.  This reaction takes
       place in a staged fluidized bed vessel with provisions
       for adjusting the temperature for optimum S02 removal
       rates .

       The sulfuric acid loaded carbon is transported mechanically
       to a second fluidized bed reactor wherein the acid comes
       in contact with hydrogen sulfide to produce elemental
       sulfur, which remains in the carbon granules, and water
       vapor which is exhausted.  Temperatures near 300°F are
       required for the reaction, i.e.
                H2S04 + 3 H2S  Activated^ 4 s + 4 H 0                (2)
                 *•  *     *•    Carbon           /

-------
Generation of  the  required hydrogen sulfide  and  the
removal of the elemental sulfur for recovery are accom-
plished in a third fluidized bed reactor according to:
     3 H2 + 4 S  Activated^  3 H2S + S (Product)
                L* 3. IT DOXl
                                                     (3)
The thermal  stripping of the sulfur and the reaction to
produce H2S  requires temperatures near 1000°F.
S02 + 1/2 02 + H20 + 3 H2
                                        4 H2°
(4)
           S02
                      S02
                     SORBER
         1/2 02, H20
           3 H2
                         H2S04
                       S
                     GEN.
                          4 H20
                          3 S
                  SSTKEV
                     ;S
                             3 H2S
Figure 1.   Chemistry of Westvaco S02  Process with
            reactants and products shown
                       8

-------
       That is, the sulfur dioxide or sulfur trioxide from the
       flue gas is reduced with hydrogen in the form of
       recycled H2S to form water and elemental sulfur.

       The hydrogen may be supplied through a number of commer-
       cially available gasifiers utilizing coal or other
       fossil fuels.  Heating of the regenerating reactors may
       be provided by conventional fuel burning units.

       The carbon stream, not shown in Figure 1, serves as a
       carrier and catalyst for promoting the reactions effi-
       ciently, but does not directly take part.  It is
       recycled from the S02 removal vessel to the regeneration
       vessels where its activity is restored before returning
       to the S02 sorber.


3.2    METHODOLOGY OF CONTRACT

       The main objective of this contract was to achieve inte-
       grated operation for a sufficient time to determine the
       effect on carbon and process performance while treating
       flue gas from an oil fired boiler.  Work was conducted
       in two parts:  from January to December 1971 and from
       September 1972 to June 1974.

       In the first part of the contract, the technical feasi-
       bility and design information for the integral pilot
       plant were to be determined, operating the S02 sorber on
       real flue gas.  These objectives were basically com-
       pleted, with particular emphasis on the flue gas desul-
       furization step.  The kinetics of S02 removal were deter-
       mined with bench scale equipment.  The kinetic data were
       modeled by an empirical expression,  which was used in
       the development of a design procedure for the S02 sorber.
       Concurrent with the bench scale work on S02 sorption, a
       variable study was made with actual flue gas from an
       oil fired boiler.  It was shown that the design proce-
       dure developed from bench scale work satisfactorily
       represented the pilot results.  Acid conversion to
       elemental sulfur was studied extensively in multistage
       pilot fluid bed equipment.  Sulfur recovery from acti-
       vated carbon was assessed by both thermal and isothermal
       (solvent extraction) schemes.  The evaluation of
       solvent extraction included testing of a number of
       solvents.  Ammonium sulfide was judged to be the most
       suitable.  The thermal process of sulfur vaporization
       from the carbon was then compared to the solvent extrac-
       tion process.  The thermal process evolved as the most

-------
       suitable since the carbon regenerated by extraction
       required additional thermal treatment to maintain the
       S02 activity, whereas in thermal regeneration, the S02
       activity was maintained without additional carbon
       treatment.  This process development in the first
       phase of the contract provided grounds for extended
       development to be culminated by operation of an inte-
       gral pilot plant run for an extended period of time.

       During the second part of the contract, additional data
       was to be developed on the regeneration process steps
       and an integral pilot plant was to be operated.  The
       design information developed was to be used to design
       and cost a prototype installation for 10 to 15 MW
       boiler.  Also, the cost of a 1,000 MW installation was
       to be projected.  To expedite the completion of these
       objectives, detailed program plans were drawn,up.  In
       the pilot development leading to the integral pilot
       plant, the two key points were 1) could existing equip-
       ment be used for acid conversion to sulfur during
       integral operation, and 2) could the process steps of
       sulfur stripping and H2S generation be carried out in
       one process unit.  It was shown that existing fluid
       bed equipment could not be used for acid conversion
       during integral operation.  Subsequently a moving bed
       sulfur generator was designed, installed and operated
       as part of the preparation for integral operation.  It
       was also shown that the two process steps of sulfur
       stripping and H2S generation could be carried out in
       one reactor.  All of the pilot equipment was integrated
       and operated mechanically on an oil fired boiler, then
       operated for an extended period to demonstrate long
       term effects on process and carbon performance.

       All performance information was used to design and cost
       a prototype unit for installation on a 15 MW coal fired
       boiler. Also the economics of the process for a 1,000
       MW installation were assessed.
3.3    CHRONOLOGICAL SEQUENCE OF DEVELOPMENT

       Development of the main steps of the Westvaco S02
       Recovery Process occurred in several phases using vari-
       ous types of equipment.  The chronological sequence of
       development is described for the main steps of the
       process.
                             10

-------
3.3.1   S02 Sorption

        Pre-contract experimentation with S02 sorption was carried
        out initially in a 1" diameter fixed bed,  and later in 4"
        and 6" diameter multistage fluidized beds.   Under the con-
        tract, additional work was done in the 6" diameter unit
        operating on flue gas from an oil fired boiler, and an
        extensive kinetic study was carried out in a bench scale
        differential reactor to obtain SO? sorption rate data,
        which were modeled by an empirical expression.   An 18"
        diameter sorber was built and used initially to produce
        the large quantities of acid loaded carbon needed to study
        the other steps of the process.   The 18" unit eventually
        was used as the sorber for the integral pilot plant.

3.3.2   Sulfur Generation
        Sulfur generation work was begun in a 1" diameter fixed
        bed reactor but this proved unsatisfactory due to certain
        inherent disadvantages of fixed beds.   Experiments were
        then carried out in a 4" diameter,  8 stage fluid bed
        glass reactor and information was obtained on the reaction
        rate and effects of important variables.  A more
        accurate kinetic study was conducted on a batch differential
        basis, and this yielded a rate model.   After construction
        of the 4" diameter, 8 stage multipurpose reactor at the
        powerhouse location, additional sulfur generation tests were
        made in that reactor to determine whether the 6" diameter
        reactor could serve as the sulfur generator for the inte-
        grated pilot plant.  These tests showed that the 6" unit
        was unsatisfactory for the intended application and efforts
        were subsequently directed toward development of an alter-
        native moving bed sulfur generator.  Bench scale work was
        conducted in a 1.5" diameter moving bed and the results were
        used to design an 8" diameter unit, which was installed
        at the  powerhouse location.  Sulfur generation studies
        were made with the 8" diameter moving bed to determine whether
        its performance was adequate and to optimize operating con-
        ditions.   Satisfactory performance was obtained from the
        unit and it became a component of the integrated pilot plant.

3.3.3   Sulfur Recovery and H2S Generation

        In the early stages of the project, two different approaches
        were considered for the removal of sulfur from carbon
        solvent extraction and thermal stripping.  Solvent extrac-
        tion was determined to be unsuitable,  primarily because the
        S02 sorption activity of the carbon was reduced excessively
        by the operation.   After abandonment of the solvent extrac-
        tion approach,  efforts were directed toward high temperature
        thermal stripping.   Laboratory data was obtained on the
                               11

-------
       equilibrium adsorption of sulfur on carbon at high
       temperatures.   In addition,  a laboratory rate study was
       made of the H2S generation steps in a single reactor.
       The combined operation was studied in the 4" diameter,
       8 stage fluidized bed reactor at the powerhouse,  and
       feasibility was demonstrated.  The 4" diameter unit
       functioned as sulfur stripper/I^S generator in the
       integral pilot plant runs.

       Recovery of sulfur product from the off-gas of the 4"
       unit required development of a sulfur condensing system.
       A condenser was designed and fabricated,  and tests were
       conducted.  Satisfactory performance eventually was
       obtained and the sulfur condenser was installed in the
       integral pilot plant.
3.3.4  Integration
       After each of the pilot scale processing units had been
       operated successfully on an individual basis,  they were
       tied together to form what was termed "the integrated
       pilot plant".  Integration required:  1) closing the
       carbon flow loop to permit continuous recycling of carbon
       through the three reactors in the system, and 2) con-
       necting the off-gas line from the 4"  diameter sulfur
       stripper/H2S generator to the gas inlet of the 8" diameter
       sulfur generator, permitting operation with internally
       generated H2S.
                               12

-------
                             SECTION 4
          INTEGRATED PILOT PLANT EQUIPMENT AND RESULTS


4.1    PILOT PLANT DESCRIPTION

4.1.1  'Introduction

       The integrated pilot  plant  for continuous removal of S02
       from flue gas and reduction of S02  to elemental sulfur
       consists of three main reaction vessels plus various other
       processing units and  auxiliary equipment, as shown in
       Figure 2.  The three  main reactors  and the reactions that
       occur in each are:
        S02 Adsorber               SO  + 1/2 0, + EUO 	•• H?S04       (5)
          18" Dia., Fluid Bed        L

        Sulfur Generator            H?S04 + 3 H«S  	» 4 S + 4 H20      (6)
          8" Dia., Moving Bed

        S Stripper/H2S Generator     3 H2 + 4 S —*. 3H2S + S           (7)
          4" Dia., Fluid Bed      	

               NET REACTION        S02 + 1/2 02 + 3 H2  	» S + 3 H20   (8)
       Other processing units  include a sulfur condenser, fluid
       bed carbon preconditioner,  carbon cooler,  and carbon
       preheater.  Auxiliary equipment includes carbon con-
       veyors and flow controls,  gas  flow controls,  electrical
       heaters,  temperature controllers and recorders, gas
       analyzers, gas sample lines, pressure lines,  blowers,
       and dust  collection devices.

       The pilot plant processes  a 20,000 acfh flue gas slip-
       stream from a 50 MW oil fired  boiler.  Granular activated
       carbon circulates by gravity flow downward through each
       reactor at a rate of about 30  lbs./hr., with reactant
       gases passing countercurrently upward through the vessels.
       The activated carbon has  a catalytic effect in all three
       reactions.  S02 is removed from the flue gas in the
       adsorber.  Regeneration of the sulfuric acid laden carbon
                                13

-------
    Figure  2.   Westvaco Process  integral  pilot  plant
SOg  2000 PPM
 20,000 CFH
                      S0g SOUSE*   (

                     H"OI* X  17.5 FT  (
                         ITS *F
SULFUR GENERATOR
     8" DIA. X  SFT
        300 "F
                    S.  STRIPPER
                    H2 S  OENERATOR

                      4" DIA. X 19 FT.
                       8  STA8E8
                         IOOO°F
                                                           CAMON

                                                             I    M<
                                                                      FIMST STAftE
                                                                      FLUID BED
                                                                      OVERFLOW
                                                                      WKIN
                                                                      KCONO STAGE
                                                                      FLUID KO


                                                                      OA>
                                                                      OI3TRIIUTOR
                                              FLUIDIZIN9 8AS

                                        FLUID BED DETAIL
                                                           iULFUR
                                                        REGENERATED CARBON
                                                        RECYCLE, APROX 30Lb/HR
                                    BOLIDS
                                    NATE
                                    CONTROLLER
                                     14

-------
is accomplished in the other two reactors,  In the sulfur
generator the H2S04 is reduced to elemental sulfur by reac-
tion with H2S.  The sulfur remains sorbed on the carbon. In
the combined sulfur stripper/H2S generator, hydrogen gas
reacts with about 75% of the sulfur to generate the H^S
for the sulfur generator.  The remaining sulfur is stripped
from the carbon thermally and leaves the reactor as a vapor
in the off-gas.  The sulfur vapor is removed from the off-
gas in a condenser and recovered in molten form.  The
sulfur free off-gas  containing HoS then passes to the
inlet of the  sulfur generator.  The regenerated carbon is
recycled back to the adsorber.  The hydrogen rich gas used
to generate H2S for the  sulfur generation step  is  the  only
raw material besides activated carbon that is required by
the process.

Instrumentation and Control -

Sufficient instrumentation is available to maintain the
desired operating conditions during steady state conditions
and to collect the data necessary for performance
evaluation.  All input gas flow rates are monitored through
meters and checked by gas analysis instruments.  Tempera-
tures and pressures at appropriate points within the system
and reactors are either indicated or recorded.
   i
Sample Points and Analysis -

Ports were positioned on the inlet and outlet of each of
the three reactors for sampling the granular carbon to
determine the amount and form of sulfur and moisture content.
Gas sample ports were also positioned so that various inlet
and outlet points in the system were analyzed chromato-
graphically for H2, 02, H2S, S02, N2, C02, CO and H20 at the
desired time.  Samples of the carbon were analyzed using
standard tests for measuring the physical and adsorption
properties.

Granular Carbon -

The carbon used in the integral pilot operation is a commer-
cially producible coal-based carbon with a nominal 12x40
mesh size and bulk density of about 40 Ibs./cu. ft.  The
S02 number and attrition number were 60 minimum and 97 maxi-
mum,  respectively, as determined by specially designed
tests.  This carbon showed satisfactory attrition resistance
in pilot testing.  Carbons with improved resistance are
being developed.                                   v
                          15

-------
       Flue Gas Characteristics -

       Flue gas used in the integral testing was from the stack of
       a 50 MW oil fired boiler having a mechanical dust collector.
       The sulfur content of the oil was about 1.8-2.0%, which
       produces about 1100 ppm S02 in the flue gas.   In order to
       avoid variability at this stage of operation, provisions
       were made for injecting additional S02 into the flue gas to
       maintain a uniform level to the pilot plant.   The tempera-
       ture of the flue gas was kept at stack temperature (300°F)
       before introducing into the S02 sorber.

       Reducing Gas Composition -

       The reducing gas was a mixture of hydrogen and nitrogen
       from gas cylinders.   Hydrogen content was varied from 40 to
       48% of the total flow to establish process requirements.

       General Operating Procedure -

       In starting up the integral system, a known quantity of
       carbon, about 500 pounds, was placed in the system and
       recirculated while preheating with a start-up heater to
       approach the desired operating temperatures.   The switch to
       flue gas was then made.   The temperatures, carbon flow rate,
       and gas flows and compositions were adjusted to the desired
       conditions.   Manual adjustments were made to the S02 added
       to the flue gas above the actual oil produced S02 to main-
       tain a constant level.  The amount of carbon placed in the
       system was sufficient to minimize adding fresh carbon
       during the integral run and represents about 407o above that
       needed to fill the reactors and conveying system.

4.1.2  Detailed Pilot Plant Description

       A detailed flowsheet of the pilot plant is shown in
       Drawing 2507 (Figure 3).   The following description of the
       pilot plant equipment is organized to follow the path of
       the activated carbon as  it moves through the system, begin-
       ning with the S02 sorber.  The operation of each processing
       unit is described also,  including important operating condi-
       tions and procedures.   Equipment identifications are based
       on the process flowsheet, Figure 3.  Instrumentation iden-
       tifications  are based on the instrumentation flowsheet,
       Drawing 2528A in Appendix B-3,

       S02 Sorber [RV-10]  -

       The sorber [RV-10]  is an 18" dia. x 17.5 ft.  high vessel
       with five fluidized beds of carbon, each bed having an
       expanded depth of 12 inches except for the bottom bed which
                                  16

-------
17

-------
17-A

-------
           Figure  3.   Mechanical integration -  20,000 cfh  S02
                          pilot plant  -  process  flowsheet
a

b
                  > "****• **"•

                     /«" */=>
               	lf....s.-.y*?.y yr_>_
                                                   t"* Doc T "


i^r^
bj

-


IT ettv f-i
* IT t't i*



i

'
fw

. M
t ^
'

,
'3w.y-to-.MV
• IftovjrjhrfT
T
» J«.

« $"«.•./«> (
•f* -6f> 5(Z* (Hi
                                                 + J-* CrHSIOOe'r

                                                  + '• MtO
     A/I /z&fff*
     Gfiuseat\
     ft, I-Iff CfM
                               M^
                     ,	    Jii «**. '/-Z,
                   - ^   j—-^ uoo'F
                     go *•     77 »»» W,
                    H/wen
                   Z'fft. iu> «i
                                  y,|^r,(
                         .'ttfzr..*^/ ^ / '^^ >
                                                           Westvaco
                                                           CHARLESTON RESEARCH CENTER
                                                           I*. O. «0» M07 NORTH CHARLESTON. S C. 2t«>
                                                                            rLeitn
                                                                    omi, it /ICHKD   o»n
                                                                      ---
                                                             car  rararTS-B-iio-i^ ----- vr -------
                                                             ••'*  \: 4iTn\    I DWG No. 2507
                                         17-B

-------
has 8 inches.  A photograph of the sorber is shown in
Figure 4.  Assuming a carbon density of 38 Ibs./ft.-^, the
weight of carbon in the vessel is about 113 Ibs.  The down-
comers are 3" dia. and extend to within 1-1/2 inches of the
distributor plates below.  The gas distributor plates have
8% open area with 0.125" dia. holes drilled on 0.42" tri-
angular centers.
                	                                _\
Carbon is fed to the sorber at a rate of about 30 Ibs./hr.
A gravimeter feeder [CFC-1] feeds the carbon from an
inventory hopper  [RV-90] into a bucket elevator [CV-3],
which raises the carbon to the necessary elevation and
empties it onto the top stage of the sorber.  Carbon flows
by gravity downward through the column and passes out of
the unit into a seal leg between the sorber and the 6" dia.
carbon preconditioner [RV-300].  The carbon level in the
seal leg is controlled automatically by a level probe type
controller [LIC-301] which actuates a vibrating feeder
[CV-5] to feed carbon into the 6" unit.

The 22,000 acfh flue gas slipstream passes through a booster
fan [E-l] prior to entering the sorber.  Flue gas flow rate
is measured by the pressure drop across an orifice.  The
flow rate is controlled by adjusting a trunnion valve
located at the entrance to the sorber.  The trunnion valve
has a pneumatic positioner which is operated mannually from
the panelboard.  The sorber off-gas passes through a cyclone
dust collector  [C-l] followed by a bag filter [F-l], and
then it is ejected to the atmosphere through an exhaust
blower [E-2].  A small negative pressure, about -0.5" W.C.,
is maintained at the top of the reactor to facilitate carbon
feed into the unit from the bucket elevator.  The pressure
is controlled manually by adjusting a slide valve in the
off-gas duct.

Flue gas enters the sorber at 300°F with a total sulfur
oxides concentration of about 2,000 ppm, of which 50 ppm
is 803 and the remainder is S0£.  The bottom bed of the
reactor, operating at 300°F, is used to remove the 863 in
order to avoid corrosion problems.  The temperature  at
Stage 2 then is lowered to 175°F to increase the rate  of
SC-2 removal.  Cooling is accomplished by injection of water
into the fluidized bed at a rate of 10 to 50 #/hr.
The rate of water injection into Stage 2 is automatically
adjusted to control the temperature on Stage 2.  A tempera-
ture controller [TIC-102], linked to a pneumatic flow  con-
trol valve through an electro-pneumatic transducer,  regu-
lates the water flow rate.  Temperatures in the upper
stages of the reactor are allowed to seek their own  level,
which is determined by the balance between heat losses and
heat generated by the exothermic adsorption reaction.
These temperatures typically fall in the 180-190°F range.


                         18

-------
Figure 4.   Continuous 18" dia.,  5 stage S02 and 863
           adsorber operating on flue gas from a 50 mw
           oil fired boiler
                         19

-------
 Carbon  Conditioner  [RV-3QO]  -

 The  carbon  conditioner  [RV-300]  is a 6" dia. x 2 ft. high
 single  stage  fluid  bed  unit  which is used to adjust the
 temperature and moisture content of the carbon as it passes
 from the  S02  sorber into the sulfur generator.  The unit is
 operated  at a carbon bed temperature of 320°F, with pure
 steam or  an air-steam mixture as the fluidizing gas.  With
 pure steam, a carbon moisture loading of about 0.10 Ib.
 H20/lb. carbon is obtained.  The steam flow rate is
 measured  by a rotameter.  A  temperature controller  [TIC-302]
 controls  an electrical  heater to maintain the desired inlet
 gas  temperature.

 Carbon  flow into the 6" dia. conditioner is controlled by
 the  level probe controller  [LIC-301] which maintains the
 proper  carbon level in  the seal  leg between the S(>2 sorber
 and  the conditioner.  Carbon passes out of the unit by
 gravity flow  into a seal leg between the conditioner and
 the  8"  dia. moving  bed  sulfur generator [RV-200].  There
 is no direct  control of the  carbon flow out of the
 conditioner.   Instead,  the carbon level in the seal leg is
 maintained  by a system  which controls the carbon flow out
 of RV-200.  This system is described in the following section
 on the  sulfur generator.

 Sulfur  Generator [RV-200] -

 The  sulfur  generator [RV-200] is a moving bed reactor, 8"
 dia.  x  10 ft.  high,  with an  actual carbon bed depth of 6
 feet.   Carbon enters the unit from an upper seal leg between
 it and  the  carbon conditioner [RV-300] and flows downward by
 gravity in  essentially  a plug flow distribution.  A vibrat-
 ing  conveyor  [CV-2]  feeds the carbon out of the unit into a
 bucket  elevator [CV-1] which transfers it to the next
 reactor.  A pneumatic controller [LIC-201] controls the
 vibrating feeder so  that a constant carbon level is main-
 tained  in the upper  seal between RV-200 and RV-300.  The
 input signal  to LIC-201 is simply the AP across the purged
 seal  leg, which is  proportional  to the height of carbon in
 the  leg.
Reactant gas containing  about  30% H2S  enters  at the bottom
of the vessel and passes upward  through the bed,  counter-
current to the carbon flow.  Total gas  flow rate  is 185-
260 scfh.  The source of the reactant  gas mixture is either
cylinder gases or else the recycled off-gas stream from the
4" dia. S stripper/H2S generator [RV-100].  When  the 4" unit
off-gas is used, the pilot plant is said to be  operating
with "process I^S".  Cylinder  gas rates  are controlled at
the panel board with rotameters.  An electrical heater and
                          20

-------
temperature controller maintain the desired inlet gas
temperature of about 300°F when cylinder gases are used.
When the pilot plant is operating with process H2S, the gas
is hot already and requires no further heating.

The off-gas from the sulfur generator is exhausted to the
boiler house stack on the negative pressure side of an
ID fan.  A valve in the off-gas line permits adjustment
of the pressure inside the sulfur generator at 1 - 5 inches
W.C. at the top of the unit to facilitate gas sampling for
chromatographic analysis.

The temperature of the carbon bed typically ranges from
260°F at the bottom to 300°F at the top.  Ideally the
temperature would be near 300°F throughout, but as the
reaction approaches completion in the lower section of
the bed, less heat is generated by the reaction and the
temperature drops off.  Temperature is a function of 1)
inlet carbon temperature and moisture level, 2) reactant
feed rates, and 3) heat input through the walls of the
vessel.  Three electrical heating mantles,  individually
controlled with power-stats, are used to regulate heat
input through the walls.  Heat input by this means is
limited by the S(>2 evolution problem at temperatures over
300°F, so that the wall heaters are operated to maintain
a wall temperature only slightly over 300°F.  Maintenance
of proper bed temperature is strongly dependent on the
uninterrupted feed of reactants to the unit.  Bed tempera-
tures drop rapidly if the feed of either reactant is
interrupted.

Sulfur Stripper/H2S Generator -

The sulfur stripper/H2S generator [RV-1001  is a 4" dia. x
24 ft. high vessel with 8 fluidized beds of carbon, each bed
having an expanded depth of 5 inches.  Electrical heaters
around the unit maintain the temperature at 1000-1200°F.
The heaters are controlled automatically by 5 temperature
controllers, with one controller for every two stages and
the fifth controller for the inlet plenum heater.  A sixth
controller handles the inlet gas heater.

The unit operates at a linear gas velocity of 2 - 3 ft./sec.
with a total gas flow rate of 200-300 cfh at 70°F.
Hydrogen and nitrogen are supplied from cylinders and
metered through rotameters.  Flow rates are set manually.
The inlet hydrogen concentration was typically 38-4870
(shown incorrectly as 27-30% in Figure 3).  The  inlet  gas
is preheated before entering the reactor.   The reactor
off-gas passes through a cyclone dust separator  [C-2]
before passing on into the sulfur condenser system, which
is described in the next section.
                          21

-------
 The  carbon feed for  RV-100  is  the  sulfur  loaded  carbon
 product  from RV-200.  A bucket elevator  [CV-1] raises the
 carbon above RV-100  so  that it can feed by  gravity flow
 into the unit,.   The  carbon  flows downward through ,the
 reactor  and passes into a seal leg connecting RV-100 to
 the  carbon cooler [HX-2].   A level probe  controller
 [LIC-102]  maintains  the proper carbon level in the purged
 seal leg by controlling a vibrating conveyor [CV-4] which
 feeds  carbon from HX-2  back into the inventory hopper
 [RV-90].

 There  are two carbon feed systems  for the sulfur stripper/
 H2S  generator.   In one, the carbon passes from the bucket
 elevator into the upper seal leg and then is fed into the
 reactor  through a valve controlled by a pneumatic control-
 ler  [LIC-101] which maintains  a constant  carbon level in
 the  seal leg.   An 8" section of the seal  leg is a carbon
 preheater [HX-1] with electrical heating and a temperature
 controller.   The other  system  consists of a pair of ball
 valves controlled by a  set  of  electronic  timers so that
 the  valves  open and  close alternately, with one valve
 always closed to maintain a gas seal.  This system
 bypasses the carbon preheater  and  feeds the carbon
 directly from the bucket elevator  into the reactor.  The
 two  feed systems provide a  number  of options in the opera-
 tion of"the reactor.  They  can be  used simultaneously or
 separately,  and carbon  can  be  fed  to various stages in
 the  reactor.  During integral  operation,  the double ball
 valve  arrangement became the primary system and the feed
 was  placed  on the fifth stage  from the bottom during most
 of the run.

 Sulfur Condensing System [RN-400]  -

 The  sulfur  condensing system [RN-400] is shown in detail
 in Drawing  2537 in Appendix B-3.   The sulfur condenser is
 a combination gas cooler and scrubber in which the hot
 sulfur-laden off-gas from the HoS  generator/S stripper is
quenched from 1000 to 300°F, and the sulfur vapor is
 removed by  scrubbing with liquid sulfur.  The condenser
 is a jacketed baffle tray column,  3" I.D. x 4' long, with
 overlapping baffles  (1/2" overlap)  spaced 3/4" apart and
 occupying a  2'  long section  of the column.  Regenerator
off-gas enters below the baffle section and passes upward
 through the baffles,  counter-current to the flow of recir-
culating liquid sulfur.   The gas exits near the top and
 is passed through a 6"  thick wire  mesh mist eliminator to
remove sulfur mist.   The sulfur-free off-gas, containing
about 30% H2S,  then goes to  the sulfur generator [RV-200].
                         22

-------
Liquid sulfur is pumped from the collector pot [C-402]
beneath the condenser and introduced above the baffle
section.  The sulfur flows downward through the baffles by
gravity flow and collects in C-402.  Recovered sulfur
condensed from the gas spills over into a collection pot
[C-403] so that a constant inventory is maintained in C-402
Cooling is provided by water at 2506F parsed through" the
condenser jacket under 40 psig at a flow rate of 1.5-2.2
gpm.  The cooling water is pumped through a heat exchanger
to heat it to the desired temperature.

Pilot Plant Instrumentation -

Pilot plant instrumentation falls into three categories:
control, general, and analytical instrumentation.  All of
the instrumentation is indicated on Figure 3, or on
Drawing 2528A in Appendix B-3.  Each category of instru-
mentation is discussed below.

Control Instrumentation -

Instrumentation for control of the pilot plant falls into
three main categories which are temperature control, carbon
flow control, and gas flow control.   Time proportional
temperature controllers are used with the wall heaters and
gas heaters on the H2S generator/S stripper and the 6"
carbon preconditioner [RV-300]t and with the gas heater
for the sulfur generator.  The wall heaters for the sulfur
generator are controlled manually with powerstats.
Temperature control instrumentation in the S0£ sorber con-
sists of a time proportional temperature controller linked
to an electro-pneumatic transducer control valve, which in
turn regulates the water spray rate.
Carbon flow instrumentation consists of a gravimetric
feeder, two sets of electronic level probes which operate
vibrating feeders, a set of electronic timers which operate
a double ball valve feeder, and two pneumatic controllers.
One of these operates a carbon flow metering valve, and the
other controls a vibrating feeder, through an improvised
mechanical linkage of pneumatic and electrical signals.

Gas flow rates are set manually by means of rotameters and
metering valves, except in the S02 sorber where an orifice
is used to measure the flow rate, and a piston-actuated
trunnion valve equipped with a positioner is used to
control the flow rate by means of a manual loading station.

The rate of liquid sulfur recirculation in the sulfur  -
condenser system is controlled by manual adjustment of a


                          23

-------
 variable  speed  SCR drive on the positive displacement
 sulfur  pump.  Temperature  in the condenser is controlled
 by varying either the cooling water flow rate or its
 temperature.  Flow rate is set manually with a valve, and
 temperature  is  controlled  by adjustment of the steam pres-
 sure  on the  heat exchanger used to heat the water to the
 desired inlet temperature.

 General Instrumentation -

 The primary  functions of non-control type instrumentation
 are to  provide  temperature, pressure, and flow rate data.
 Three instruments are used for the temperature data,
 including an Acromag 10 point digital indicator, a
 Honeywell 24-point strip chart recorder, and a Doric 40-
 point data logger.  The temperatures considered most
 indicative of pilot plant  operation and most important
 to follow appear on the Acromag digital indicator.  The
 same  ten  temperatures and  a few others appear on the
 Honeywell strip-chart recorder, which provides a graph
 showing variation with time.  The same ten and thirty
 more, representing all of  the important process tempera-
 tures,  are recorded by the Doric-data logger in numerical
 form, either automatically at 30 or 60 minute intervals
 or else manually whenever  a printed record is desired.
 The indicated,  logged and recorded temperatures are
 depicted  by  TI, TL and TR, respectively, on Drawing 2528A
 in Appendix  B-3.

 Pressure measurements are useful in diagnosing the operating
 condition of the pilot plant.   Differential pressure gauges
were  used to measure pressure drop across each unit and
 also  across each stage within the fluid bed units.  This
provided  information on whether or not carbon flow is
 satisfactory and reveals any buildup in the system.  Pressure
 drop  across an  orifice is used to measure gas flow rates
 at four places  in the system.

Analytical Instrumentation -

Analytical instrumentation provides gas composition data
 for the inlet and outlet gas streams of the S02 sorber
and regenerator units.   The S02 sorber gas streams are
analyzed  for S02 concentration using  an EnviroMetrics,
 Inc.,  Series NS-200 S02 analyzer from which output is
recorded on a strip chart recorder.  Gas streams in the
regenerator units are analyzed by a Bendix process
chromatograph for seven components:  H2S, H20, S02, CO, H2,
C02 and N2-   The chromatograph can analyze five samples
per hour.   Peak heights for each component are read from
recorder,  modified for bar chart display.
                        24

-------
Carbon Dust Collection -

Carbon dust is removed from the off-gases of the H2S
generator/sulfur stripper and the S02 sorber.  Equipment
also was installed for dust removal from the 6" carbon
preconditioner off-gas, but it was necessary to bypass
this equipment due to operational difficulties.  The S02
sorber is responsible for the major portion of the carbon
dust generated in the process.  The sorber off-gas first
passes through a cyclone and then through a bag filter
to remove the finer particles.  The combined dust removal
efficiency is greater than 997o by weight.

Dust is removed from the off-gas of the H£S generator/
sulfur stripper by a cyclone followed by a specially
designed filter containing a 4" depth of Fiberfrax Long
Staple Fine ceramic fiber material.   The Fiberfrax filter
was bypassed, however, after it was found that the small
amount of carbon dust did not create problems in sulfur
condenser operation.

Sulfur Collection -

Sulfur condensed from the regenerator off-gas is recovered
as a liquid by draining the collection pot into a 6-liter
metal beaker at 8-hour intervals or more frequently if
desired.  The sulfur is then weighed and transferred to a
larger container for storage.

The recovered sulfur is analyzed for purity either by a
combustion analysis in which the carbon impurity is measured
as C02, or by a sulfur vaporization analysis in which the
impurity is measured after vaporizing away the sulfur.  Ash
content is also determined by a standard ash analysis.
                         25

-------
4.2    INTEGRAL PILOT PLANT RESULTS

       The broad goals of the Westvaco Process  are  as  follows:

            1)   High S02 removal
            2)   Minimum hydrogen use
            3)   Maximum S02 recovery as elemental sulfur
            4)   Minimum carbon burn-off.

       With respect to these broad goals  for process operation,
       an original set of goals were defined based  on  studies  of
       each of  the process steps separately.  Another  set  of
       goal bases which are directed to these broad goals  and  the
       main process streams shown in Figure  5 are:

            1)   S02 Sorber
                a.   S02 removal

            2)   Sulfur Generator
                a.   Sulfur compounds evolved/S02 removed

            3)   H2S Generator/S Stripper
                a.   H2 utilized/S02 removed
                b.   Carbon burn-off/S02 removed

            4)   Sulfur Condenser
                a.   Sulfur recovery/S02 sorbed.

       The basic difference compared to the  original set of goals
       is that  in the latter case only streams  that cross  the
       integral process boundary are chosen  and in  the former  case
       some of  the  goals were based on internal recycle streams.

       The integral run,  which totalled approximately  20 days,
       was made with pilot equipment installed  at an oil fired
       boiler.   A mechanical problem,  which  is  described in
       Appendix B-2,  was developed after 11 days or  about 18
       carbon cycles;  it was corrected and did  not  reoccur
       during subsequent integral operation.  The second part
       of the integral run was voluntarily stopped  after an addi-
       tional 8 days or about 11 carbon cycles.

       To allow continued evaluation of the  carbon  performance
       for additional  carbon cycles,  the  same carbon was used  in
       the entire run.   In total,  the carbon was circulated
       integrally for  approximately 29 cycles to allow evalu-
       ation of carbon attrition for this particular carbon.   Of
       these 29 cycles,  the carbon was exposed  to flue gas/regenera-
       tion gas conditions for 21 cycles  of  which 14 were  with
       process  H2S.  Since the source of  the H2S, i.e. process
       or  cylinder,  had no apparent effect on S02 sorption, 21
       cycles were  taken as  the process exposure time  in analyzing
       the S02  removal  performance of the carbon.


                                26

-------
               Figure  5.  Schematic of integral Westvaco S02
                           removal pilot plant
FLUE  GAS-
   S02
HYDROGEN-
                           S02
                         SORBER
                                                                   S02 SORBER
                                                                   OUTLET GAS
                          SULFUR
                         GENERATOR
                                                                   SULFUR GENERATOR
                                                                      OUTLET GAS
                                           PROCESS H2S
                                        SULFUR
                                       CONDENSER
                           H2S
                         GEN./S
                        feTRIPPER
ELEMENTAL SULFUR
     PRODUCT
                           T
                        REGENERATED
                         RECYCLE
                         CARBON
                       	I
                                         27

-------
       The development of the Westvaco S02 Recovery Process
       encompassed bench scale studies,  pre-integral pilot scale
       testing, and integral pilot plant operation.  Many experi-
       ments were performed in each of the phases of process
       development.  For clarity these results have been presented
       here in two sections — integral and pre-integral work.  In
       Section 5 both the bench scale studies and the individual
       pilot unit studies are discussed.  Since the principal goal
       of this program was to design and operate an integral pilot
       plant of the process, the integral work is presented first
       in the remainder of Section 4.

4.2.1  Overall Integral Results

       Operating Conditions -

       The general intent of the integral run was to maintain
       constant conditions over an extended period in which the
       granular carbon would be exposed  to repeated sorption and
       regeneration with H2S produced in the process.   In earlier
       studies, carbon had been exposed  to flue gas during sorp-
       tion,  but in the sulfur generator step only cylinder H2S
       had been used.   An arbitrary time of 30 cycles was initi-
       ally selected for the integral run,  during which time any
       trend would be detectable and indicative of longer term
       effects.  In addition,  a 90% S02  removal efficiency was
       to be maintained with a sulfuric  acid loading on the carbon
       of at least 18 Ibs.  acid/100 Ibs.  carbon.   Other limits on
       sulfur generation (acid conversion),  sulfur recovery and
       operating conditions were selected based on pre-integral
       pilot and bench scale test results.

       The operating conditions in Table 1 were selected to meet
       the target goal of 90% S02 removal with an acid loading
       of at  least 18 Ibs.  acid/100 Ibs.  carbon.   The inlet flue
       gas was controlled at the rate  of 22,000 scfh to the S02
       sorber.   The S02 content was adjusted as necessary to
       maintain 1900 to 2000 ppm,  and the inlet temperature was
       maintained at 300°F on the first  stage for S03 removal,
       with the next stage cooled by water spray to 175°F for S02
       removal.   The temperatures of the remaining sorber stages
       were not controlled and were allowed to rise due to heat
       of reaction during S02 removal.   Carbon bed depths of 3.5
       inches  on the bottom stage and 6  inches on each of the
       remaining stages were set for a total of 56 inches in an
       expanded state.

       The recycle rate of carbon was  set to achieve the desired
       acid loading based on the previous relationships developed
       between the operating parameters  in the sorber.
                               28

-------
        Table  1.  RANGE OF  OPERATING CONDITIONS FOR
                   INTEGRAL  PILOT  PLANT  RUN
INLET FLUE  GAS
   Gas Rate:
   Temperature:
   Composition,
                S03:
                NO:
                02:
                H20:
                Inert Gas (C02,4 N2):
           CONDITION

          22,000 SCFH
             300°F
        '1900-2000  PPM
                50  PPM
               150  PPM
               4.5  Vol.
              13   Vol.
            Balance
S02  SORBER
   Temperature,  Stage 1 (Bottom):
                Stage 2 (H20 Spray);
   Carbon Bed Depth  (Expanded):
   Fluidizing Velocity:
   Space Velocity:
             300°F
             175°F
      Stage  1       -  8"
      Stages  2  to 5 - 12"
      3.5 Ft./Sec. (<> 300°F
  3400 SCF Gas/CF Carbon-Hr.
REGENERATORS
   Acid Converter
      Temperature:
      H2S Inlet Rate:
                        •
      Space  Velocity:
   S Stripper/H2S Generator
      Temperature:
      H2 Inlet Rate:

      Space  Velocity, Stripper:
                     H2S Generator:
      Fluidizing Gas Velocity:
         290°F  (Avg.)
Output from H2S Gen. Range =
   2.5-2.9 mol/moles
   100 SCF Gas/CF Carbon-Hr.
          1000-1100°F
3.4, 3.9,  4.3 moles H2/mole
   S02 Sorbed
2000-3100  SCF Gas/CF Carbon-Hr.
6200-9300  SCF Gas/CF Carbon-Hr.
 1.8-2.7  Ft./Sec. @ 1000°F
CARBON RECYCLE RATE:
       29-30  Lbs.  C/Hr.
                                29

-------
Temperature and space velocity conditions for the sulfur
generator and sulfur stripper/H2S generator were based
on earlier process unit test results.  The hydrogen flow
to the generator was varied above the stoichiometric
requirement of 3 moles/mole of S02 sorbed on the carbon.
The amount of H2S entering the sulfur generator was pre-
determined by the hydrogen input, with no further attempt
to control this rate.

The pilot plant was started and operated for 315 hours
under these conditions using previously described
procedures.

Summary of Results -

The overall results of the integral run were as follows:

        No. of Process Cycles:   21
        No. of Operating Hours:  315
        Outlet S02:              120 ppm
        Avg. S02 Removal:        94%
        Product Sulfur:          99.7% Pure
        Carbon Attrition:        0.26 Ib./hr.

Discussion of Major Response Parameters -

The main factors observed in the pilot operation were the
degree of S02 removal during cycling of the carbon, the
requirements for hydrogen for H2S generation, and the con-
version of S02 to elemental sulfur.  Other factors were
the loss of carbon by chemical and/or mechanical means
and the disposition of any excess hydrogen.

S02 Removal Efficiencies -

The removal of S02 during the 300 hour period of the
integral run is given in Figure 6.  The flue gas, contain-
ing 1900 to 2000 ppm S02, was desulfurized to well above
90% with a maximum of 97%, corresponding to 60 ppm remain-
ing in the effluent gas.  By inspection of the plot,
there does not appear to be any trend toward reduction  in
S02 removal efficiency.  This has also been substantiated
by laboratory analysis of the recycled carbon.  During
the integral run,  the carbon was cycled through the
system some 21 times, based on a calculated carbon resi-
dence time of 15 hours in the integral system.  The amount
of S02 picked up by the carbon in terms of sulfuric acid
averaged 24 Ibs./lOO Ibs. carbon, substantially above
target.
                         30

-------
        Figure 6.  S02 removal efficiency during
                   integral  pilot tests
  100 .-
   95 ..
g  90
                                                 .-GOAL - 90T
S
*   J_                                      INLET S02 ' 1900-2000 PPM
                             RUN TIME, HOURS

         30   60    90    120   150   180   210   240   270   300  330   360
      -i	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1

         2     4    6    8    10    12    14    16   18   20   ??    24

                          NUMBER Of CARBON CYCLES
  No corrosion or dew point  problems were noted in operation
  at 175°F, since the 30 - 50 ppm gaseous 863 in the flue gas
  is adsorbed on the carbon.   This removal of 803 with
  carbon was demonstrated  in previous studies.

  The 150 ppm NO in the flue gas is not directly affected  by
  the carbon and as such remains in the flue gas.  The
  initial effect of the NO is to suppress the 802 pickup.
  This effect appears up to  a NO concentration of about  150
  ppm.  This aspect is covered more fully in Section 5.21.

  Sulfur Product -

  It is important that the sulfur by-product from the regene-
  ration system be a salable commodity.  The elemental
  sulfur recovered from the  pilot tests had characteristics
  as shown in Table 2.
         Table 2.  PROPERTIES  OF SULFUR PRODUCT

                  Sulfur            99.7%
                  Ash               380 ppm
                  Carbon          2500 ppm

                  Acidity  .          2 ppm
                  Chloride           <2 ppm
                            31

-------
These properties, measured for Westvaco by a sulfur
producer, classify  the  sulfur collected as a commercial
grade.

The small amounts of carbon in the sulfur, a result of fines
carryover from the  regenerator, gave the sulfur a greenish
cast.  It was  demonstrated that these fines could be readily
filtered out to give a  bright sulfur product of 99.9% purity.

Carbon Chemical Consumption -

In passing through  the  regeneration sequence, the activated
carbon is exposed to temperatures progressively increasing
from 300°F to  1000°F.   To prevent chemical consumption at
1000°F, the sulfuric acid is reduced to elemental sulfur at
300°F.  In addition to  production of elemental sulfur, a
goal of the Westvaco Process is to minimize the amount of
carbon reaction to  produce C02.  Measurements were made on
the C02 content of  the  regeneration off-gases to estimate
the amount of  chemical  consumption or "burn-off" that could
be occurring by this means.

As shown in Figure  7, the carbon burn-off,   calculated
from CC-2 evolution  reached a stable value of about 10 - 12
Ibs.  C/ton   of S02 sorbed from the flue gas.  As shown by
the dotted line this compares to a "burn-off" of 187
Ibs./ton if the carbon  were consumed by reacting with all
the sorbed acid under thermal regeneration conditions.  This
    Figure 7.  Carbon burn-off during integral
               pilot tests
   190 ••
 8 '85

 1 180
              MAXIMUM THEORETICAL BURN-OFF WITH THERMAL REGENERATION
 &  25.
20-
15.
10-
0.
°\° \
N® Hi
X«2_ o o0 oo o ^
uw O Q0
»
1 II 1 1 1 H — l-4—l — 1— 1 — 1 1 I 1 I
*? ° °n
v o Q O t>
1 — ! — U- 1 — 1 — 1 — I— i
                           10
14
16
                                            18
                                                 20
                                                      22
                         NUMBER OF CARBON CYCLES
                                                          24
                         32

-------
reduction in burn-off of about 95%  shows  that  the  original
objectives were achieved.  By inspection  of  the  data there
was no apparent effect on burn-off  when the  hydrogen input
was varied in the range of 3.9-4.6 moles/mole acid  dis-
cussed earlier.

Complete conversion of the acid to  sulfur was  not  required
to prevent burn-off.  Earlier experiments on the bench
scale verified this fact, in that the addition of  sulfur
by various means considerably reduced the chemical consump-
tion of the activated carbon during regeneration.

As discussed earlier there was some thermal  decomposition
in the  sulfur generator which would probably explain the
small amount of burn-off measured.

If all of the C02 measured is a result of burn-off, the low
values measured here would correspond to  a complete  replace-
ment of the inventory only about once every  two  years.

Carbon Mechanical Consumption  -

The attrition rate experienced in the integral pilot opera-
tion, Figure 8, showed an initial decrease,  probably due
to a rounding off of rough edges, and then a stabilization
at a rate of 0.26 Ibs./hr.
    Figure 8.  Activated carbon attrition rate
               during integral pilot tests
                    8   12   16    20

                      NUMBER Of CARBON CYCLES
28
    32
                          33

-------
The data indicated nearly all of this attrition occurred
in the fluidized beds of the S02 sorber.  Additional work
has shown that the combination of larger particle sizes and
carbons with improved hardness will reduce the attrition
rates to about 10% of the values measured here.  These
improvements should be incorporated in future scale-up work.

It is significant that there is no apparent increase in
the attrition rate as the carbon was recycled thermally
and chemically, as has been observed with other solid
adsorbents.  The nature of S02 recovery with carbon,
providing a surface for catalysis and adsorption
rather than actually chemically participating in the reac-
tions as is done with metal oxides, probably results in
the maintenance of structural integrity and strength of
the carbon.

Regeneration Results -

In the integral runs, the intent was to demonstrate that
the carbon could be repeatedly regenerated for reuse and
to maximize the amount of elemental sulfur produced within
the limitations of the present pilot equipment.  The only
deliberate change in regeneration conditions was in the
hydrogen input.   Other conditions were pre-set based on
prior work.

As discussed in the preceding section, the activated carbon
retained its adsorptive capabilities throughout the run,
attesting to the suitability of regeneration under all
hydrogen input conditions.
•
Three levels of hydrogen input were evaluated during the
integral runs.   Analyses of all the process streams were
used in preparing the material balance presented in
Table 3.
                         34

-------
       Table 3.   EFFECT OF HYDROGEN INPUT  ON
                 BY-PRODUCT RECOVERY
                                           Condition
                                       A
 TOTAL HYDROGEN INPUT                   4.6
    (moles/mole available acid)
 HYDROGEN USAGE
    (moles/mole available acid)

    1.  Formation of by-product          0 Qn    0 00    , _
         sulfur                       z'90    2'88    3-°

    2.  Reaction with by-product        •   or    ft ."    .
         sulfur to form H2S             °'Zj    °'n    °
    3.  Reaction with chemisorbed         n Qn    , n
         oxygen to form H20             u-au    '•"_    u-a

      TOTAL MEASURED HZ OUTPUT          4.05    3.99    3.90
Condition C essentially represents the process hydrogen
input necessary for conversion of the available acid  to
elemental sulfur product.   The hydrogen input above the
stoichiometric ratio of 3  reacted with chemisorbed oxygen
to  form water and H2S did not appear in the sulfur
generator vent gas.   The reaction of a part of the inlet
hydrogen  with chemisorbed oxygen had been observed in
previous  work and is apparently instrumental in retaining
the activated carbon's activity upon cycling.  As the
hydrogen  was increased from 4.3 to 4.6 a part of the
product sulfur reacted to  form H2S which appeared in  the
vent gas,  while formation of water essentially remained
constant.   The difference between the measured hydrogen
input and output amounts to about 1270 and could be the
result  of analysis error or, possibly, chemisorption  of
these small amounts of hydrogen on the carbon itself.

If  the  hydrogen ratio were lowered below  that of Condition
C,  S02  formation would be expected at the expense of  part
of  the  sulfur product.  This would be the desired
direction if the process is slightly out of balance since
S02 can be readily recycled to the sorber.

During  these  integral tests and with prior work, there was
limited temperature  control in the moving bed sulfur
generator.  As  a result of higher than desired temperatures,
a part  of  the  sorbed acid  decomposed to S02 in the upper
                         35

-------
       part of the unit and was not readily available for
       conversion to sulfur.  Thus a maximum of 857o conversion of
       the sorbed acid to sulfur was obtained with this equipment.
       Prior testing had shown that with proper temperature con-
       trol essentially 100% conversion to sulfur is possible and
       this should be readily attainable in larger equipment where
       fluid beds will be used.

       The gas residence time in the regenerators is only about 15
       seconds;  therefore,  response of the system to hydrogen
       input is  very rapid.   Thus,  control should be readily
       achieved  by monitoring regeneration off-gases and adjust-
       ing the hydrogen input accordingly.

4.2.2  Detailed Integral Results

       Detailed Operating Conditions and Results -

       A detailed summary of the integral run conditions and
       results for both IR-1 and IR-2 is presented in Table 4.
       As can be seen, five steady state operating periods
       occurred  during IR-1 and three during IR-2.  All these
       periods represent changes in the regeneration conditions
       only.  The S02 sorber conditions were maintained constant
       throughout.  In periods 4,  5, 7 and 8,  the ratio of
       hydrogen  input to S02 sorbed (moles H2/mole S02 sorbed)
       was deliberately changed to optimize the process for
       hydrogen  consumption.  Although the ratio of H2/S02
       sorbed is the same for Periods 7 and 8, Period 8 simu-
       lated the use of shifted producer gas for sulfur stripper/
       H2S generator feed to evaluate its effect on carbon
       regeneration.

       To evaluate the operating performance of the integral
       pilot plant,  five performance variables--S02 removal, H2$
       utilization,  acid conversion to sulfur, H2 utilization,
       H2 utilization to H2S-- were initially chosen as indicators.
       In the analysis of Runs 6,  7 and 8 where some of the acid
       was converted to sulfur in the sulfur stripper/H2S genera-
       tor,  it became apparent that several other process responses
       might be  more suitable for assessing integral operation.

       Since the overall process response time to changes in
       process variables was not known, the attainment of steady
       state was taken to be when no variations about a statisti-
       cal average were measured in the process streams.  The
       data indicated that approximately 30 hours were required
       while using process  H2S before the total integral pilot
       plant reaches  steady state after a change.  In view of
       this,  only Periods 3, 4 and 7 in Table 4 represent steady
       state conditions.   Period 5 was very close to steady state,
                                36

-------
    Table  4.   INTEGRATED OPERATING CONDITIONS AND RESULTS

STEADY STAIR 1'V.RIOD
CYL1XEW.R OR I'RiXESS H'S
TIME PERIOD (Hrs.)
CARBON RECYCLE RATE (Lb» . /llr . ) •
SOJ SORSER*
Condi t ions
Cas Rate. CFH
If rap. (Avg.). 0F
Outlet Gat Tcr.p.. °F
Inlet SO' Cone., ppm (Dry Bos Is)
Rate. CFH
Temp. (Range), °F
Results
Outlet S02 Cone., ppn (Dry Basil)
Rate. CFH
Outlet Acid Loading. Ib./lb. C
Pressure Drop, in. 1120
S02 Removal, "i Inlet
Carbon Attrition Rate, Ib./hr.
CARBON CONDITIONER
Conditions
**Steam Cone., vol. *
Carbon Bed Temp. (Avg.), °F
Results
Carbon Hoist. Load.. Ib./lb. C
SULFUR GENERATOR
Conditions
Temperature C Bed (Range) , °F
(Avg.). op
*—1 - *. M.. /»T"»*
*tt*e*. »*£ , W* (1
Inlet C02, CFH
Inlet Acid Load., Ib./lb. C
Results
Outlet Acid Load., Ib./lb. C
Outlet S Load.. Ib./lb. C
Outlet H2S, CFH
Outlet SO;. CFH
OutioL C02. c~!
Outlet HiO. CFH
H2S Utilization. %
Acid Conversion to S, 7.
S Evol., Ratio to 502 Sorbed
~ , tas H2S)
(as S02)
H2S GENERATOR/SULFUR STRIPPER
Conditions
Temp- C &ed (Range), °F
(AVR.). OF
**Inlet N2, CFH
"Inlet H2, CFH
**Inlct C02. CFH
Inlet S Load. . f S as S/# C
* S as H2SO&/* C
Results
Outlet S Load., t/f C
Outlet H2. CFH
Outlet H2S, CFH
Outlet S02. CFH
Outlet HjO. CFH
Outlet COj. CHI
H2 Utll., 7. of Inlet
Coin, to H;iS. 7.
Ratio to S<>2 Sorbed
SULFUR COMDr.llSER
Outlet Can Tcrpcrature (Avp, .), °F
S Kecov. , Ibn./lir.
1 of Stripped
T, of SO? Sorted
(by S Kt-cov.)
rhv S Kvrj !•/..
II 4
8
I'l OCOSS
H
29
,

15,400
185
184
?,030
;9.o
16S-197

80
1.1
0.23
26.3
96
0.28


100
324

0.09


262-306
287
•-Of
*.R
4o'g
0.23

0.04
0.21
1.3
4.4
40.9
126
98
62

0.05
0.16


500-1-350
1027-
66
95
40
0.21
0.04

0.047
0
58
0
45
40.9
100
61
•J.4

2/0
'/.O
'.J5

»li
M)
H 1
   *A11 sorber conditions steady for whole run period of about 8 days or 13 cycles.
  **0nly condition changes during run.
 ***Gas sampling error found and rectified.
****At end of period H2 leak was found.
                                 37

-------
 and Period 8 was close enough  to  steady state to yield
 data indicative of overall process  operation at the
 corresponding process conditions.

 The data in Table 5 indicate a number  of results obtained
 during the integral run.  The S02 removal goal of 90% was
 exceeded with 93 to 97% obtained  for all the carbon cycles
        Table 5.  PROCESS OPERATING PERFORMANCE
Performance of Process Unit
S0£ Sorber
S02 Removal , % of Input
Sulfur Generator
H2S Utilization, % of input
H2S Evolution, % of SOz removed
in S02 sorber
S02 Evolution, % of S0£ removed
in S02 sorber
H2S04 Conversion to Sulfur, %
H2S Generator/S Stripper
Ratio H2/S02
H2 Utilization, % of input
S02 Evolution, % of S02 removed
in S02 sorber
Carbon Burn-off, Ibs. C/T S02
sorbed
Sulfur Condenser
Sulfur Recovery, % -of S02
removed in SOg sorber
Goal
90
95
99
90
IR-1 Run
4
94
92
21
7
87
4.3
100
0
10
67
5
94
96
11
10
90
3.9
100
0
10
77
IR-2 Run
7
95
100
0.3
14
70
3.4
100
0
13
84
8
95
98
5
16
60
3.4
100
0
11
83
      *CTose approach to steady state in these regeneration periods.  In
       Task IV-A, Period 3 simulated producer gas used.
in both  tasks.   The H2S utilization was increased to 1007<,
by lowering  the H£ to S02 sorbed ratio to  3.4,  but as a
result,  the  sulfuric acid conversion to sulfur  was decreased
in the sulfur  generator.  The overall sulfuric  acid conver-
sion to  elemental sulfur, however, increased  as the desired
effect,  as shown by the sulfur recovery increase.  This
means that the  effective acid conversion was  completed in the
H2S generator/S stripper.  The H2 utilization was 100%, and
                         38

-------
H2 utilization to H2S decreased which would result from the
increased acid conversion in the H2S generator/S stripper
or decreased recycle process H2S.  The decreased regenerant
H2S gas recycle is viewed as a benefit, since the H2 utiliza-
tion is maximized.  In summary then, the overall intended
goals were markedly improved in the IR-2-C run as reflected
by the increased flue gas S02 recovery as elemental sulfur.
Most of the additional S02 was evolved in the sulfur
generator outlet and could be recycled back to sorber to
effectively close the loop, with a resultant increase in
S02 recovery as elemental sulfur.

It is felt that S02 evolution is a thermal effect
which can be reduced by better temperature control in fluid
bed sulfur generators planned in future development.  In
any case, to put this S02 evolution in perspective, the
sulfur generator off-gas contains S02, H20, C02 and N2, all
typical components of flue gas.   The total gas flow rate
of this stream is about 250 to 300 cfh @ 70°F in this pilot
equipment.  The'stream would be recycled back to the S02
sorber resulting in an increase of less than 2% in the gas
volume.

An unexpected results, mentioned above, that was found
during the integral runs was the distribution of the acid
conversion to elemental sulfur reaction between jthe
sulfur generator and sulfur stripper/H2S generator.  With
the decrease in H2 input, the acid conversion to sulfur
decreased in the sulfur generator as shown in Table 5.
But, the overall acid conversion to sulfur, as indicated by
sulfur recovery, increased as a desired effect.  This meant
that the effective acid conversion was completed in the
sulfur stripper/H2S generator.  More importantly, this
acid conversion in the sulfur stripper/H2S generator was
accomplished without an attendant increase in carbon
burn-off as might be expected at the 1000°F reactor
temperature and no S02 was evolved indicating conversion of
the unconverted sulfuric acid to elemental sulfur.  Also
accompanying this shift in acid conversion was a decrease
in H2 utilization to H2S while maintaining 100% H2
utilization.  Such a decrease in the H2S recycle is con-
sidered beneficial since the H2S utilization is maximized.

A short run was made, Period 8,  in which the effect of a
simulated shifted producer gas or reformer gas on regenera-
tion was evaluated.  Although the run lasted only 8 hours
and "true steady state" results were not obtained, the
apparent general effects were:

     1)  H2S production in sulfur stripper/H2S
         generator decreased
                        39

-------
      2)  Acid  conversion  to elemental sulfur in the
         sulfur generator was further decreased.

      3)  No  S02 evolution from sulfur stripper/H2S
         generator

      4)  No  effect on H£  reduction of flue gas S0£
         to  elemental sulfur.

      5)  Carbon burn-off  constant.

 Overall, the shifted producer gas had no deleterious
 effects on process performance.

 Activated Carbon Performance -

 One of the main intents of the integral runs was to evalu-
 ate the activated carbon  performance over extended carbon
 recycle conditions.  The  carbon was recycled for
 about 29 cycles, of which 21 were under flue gas/regenera-
 tion  gas conditions.  Some of the indicators of carbon
 performance  are S0£ activity, carbon attrition, carbon
 burn-off, mean particle diameter, pore volume, surface area,
 and ash content.  Throughout the 29 cycles, there was no
 indication of any adverse effects on the carbon performance.

 S0£ Removal  -

 Some  of the  indicators of S02 activity of the carbon are
 S02 removal  efficiency,  acid loading of the carbon, sulfur
 loading of the regenerated carbon, and relative S02
 activity as  determined by an independent bench scale S02
 sorption apparatus.  The  S02 removal efficiency, acid load-
 ing of the sorber product and residual sulfur loading of
 the regenerated carbon are shown in Figure 9  for the
 IR-1 and IR-2 runs.  For  all 21 cycles,the SOo removal was
 above 90% and actually ranged from 93 to 9770 for an inlet
 S02 concentration of 1850 to 2050 ppm S02 to yield a carbon
 loading of 22 to 26 Ibs.  acid/100 Ibs. carbon from the
 sorber.   In  IR-2 the S02  removal average was higher but
 the inlet S02 concentration was slightly lower.  The
 residual sulfur loading of the regenerated carbon was about
 0.035 Ib. sulfur/lb. carbon for IR-1 run and about 0.04 to
 0.045 Ib. S/lb. C for the IR-2 run.

 It would be  expected that this slightly higher residual
 sulfur loading might reduce the relative S02 activity.  As
 seen by Figure 10, in which the relative activity is given,
 the activity was decreased slightly because of the slightly
higher residual loading, but was relatively constant for
the duration of IR-2 run.  For the virgin precursor,the S02
                         40

-------
Figure 9.  Activated carbon performance during Westvaco S02 recovery integral pilot runs
1
« 100-
i
1 „„.
ce 90
80
(
I 0.3-
g
O
o a n 9-
< __..f «•«.
t- -5T
LU O
1- CM
Z> 1C
o
5 3 0.1
ca
o
CM
co 0
1
O
z
| o.io-
CO <_>
5 ffl 0.05'
T ^ o-
CQ
a — J
LU
m
c
CO
CM
O r
CO v
IR-1 -»
AVERAGE INLET S02 CONC. • 2000 PPM
fv
START-UP .,„ • 	 \ .J > 	 v -J^
C RECYCLE V-* 	 *r* **
NUMBER OF CARBON CYCLES
2 46 8 10 12 14 16
] 40 8*0 120 160 200 240
ELAPSED TIME,
y^^^^v^vj
•
.
NUMBER OF CARBON CYCLES
2 4 6 8 10 12 "14 16
3 40 80 120 160 200 240
ELAPSED TIME, F
>
NUMBER OF CARBON CYCLES
2 4 6 8 10 12 14 16
40 80 120 160 200 2'!Q
ELAPSED TIME, H
p»-IR-2
AVERAGE INLET S02 CONC. - 1900 PPM
V\^/V-^\
»^ RE-START.^. •yw> •
C RECYCLE - «,„,
•
18 20 22 24 - 26 28 30
280 320 360 400 440 480
k>URS
"^^ ?*v*^» / *y
18 20 22 24 26 28 30
280 320 360 400 440 480
OURS
1
1
18 20 22 24 26 28 30
280 320 350 400 440 480^
OURS

-------
            Figure 10.   S02 activity as a  function  of carbon  cycle time  as determined
                          by bench  scale apparatus
N5
K
UJ
cc
J
*s_
£

•*
i-
E
£
CO
'
• AM
100-

t\f\
yj

80-
70-
60






A •
	 • 	 e=-=^T.ri^
AVERAGE FOR FLUE GAS PERIOD
NUMBER OF
2 1 6 8 10 12
.1 t t I I 1
,
IR-l — ».|« — IR-2
|




•^-^-C/1^9
CARBON CYCLES ~
11 15 18 20
_! 	 j . . , i , 	 : 	




M P* ^* * ^ .
us FLUE GAS 	 *•*


— ~~°~-<^_j: — • — "*^*^T^-* 	 AVERAGE FOR FLUE
22 24 26 28 30 GAS PERIOD
.—; 	 1 	 ! — , 	 1 	 . — : 	 > 	 f
80
120 *    160
                                                 2CO     210    280
                                                   ELAPSED TIME/ HOURS
320
350
400
180

-------
number was 85 for comparative purposes.  As shown  in
Table 4, the average temperature of the H2S generator/
sulfur stripper was about 100 to 200°F lower than  in the
IR-1 run even  though the intention was to operate under
the same conditions as in IR-1.  The reason was that the
temperature controllers were set at a slightly lower tempera-
ture to prevent possible temperature overshoots experienced
in the IR-1 run.  It is expected, therefore, that  the
residual loading might be increased or the relative
activity decrease, as was the case.

The switch from cylinder hydrogen to a simulated shifted
producer gas had no apparent effect on the carbon's per-
formance, as determined by the above measures of S02
activity.

Attrition -

The carbon attrition rate was measured as a function of
time, as shown in Figure.  About 3070 of the carbon charge
for the IR-2 run was virgin carbon on the same original
batch of virgin carbon used in IR-1.  The need for this
added virgin carbon was due to accidental carbon spillage
at the end of the IR-1 run and explains why the carbon
attrition rate was slightly higher at the start of the IR-2
run, than 0.27 Ib. C/hr. at the end of IR-1.

Problems in IR-2 with the S02 sorber dust filter precluded
use of that equipment,  The rate of carbon collection was
relatively constant in the IR-1 run at 0.12 Ib. C/hr, so
that rate of collection is also taken for the IR-2 run.
This leads to a final attrition rate of about 0.25 Ib. C/
hr., compared to 0.3 Ib. C/hr. predicted by bench scale
measurements.  This data should be sufficient to allow
estimates of the carbon attrition on economics of the SC-2
removal process.

Also a direct result of carbon attrition, the mean particle
diameter would be expected to decrease with no carbon
make-up.  As shown in Figure 11, the mean particle diameter
did decrease from about 1.16 mm to about 1.0!) mm.  This
decrease caused no apparent problems in process control or
other operations of the pilot plant.  In an actual con-
tinuous operation with carbon make-up,the mean particle
diameter would be expected to stabilize.
                          43

-------
Figure  11.    Carbon  attrition, mean  particle  diameter,  and ash  content
                 as  a  function of carbon cycle time
                                     IR-l
                   TOTAL AVERAGE CARBON
                     ATTRITION RATE
                       IR-2
                         TOTAL AVERAGE CARBON
                           ATTRITION RATE
                  8    10
NUMBER OF CARBON CYCLES
 12    11     16    <:i8
  10      80      120     160-     200      210     280
                                   ELAPSED TIME, Houas
20    22    21    26
                                                           180
                                         DUST COLLECTED WITH
                                          S02 SORBER CYCLONE
                                         DUST COLLECTED «:TH
                                          S02 SORSER DUST
                                          FILTER
                            NUMBER OF CARBON CYCLES I
J 0.9
c f
2.
) 10
f P
80
8
120
10 12
160
11 •
1
200
16
210
"i
280
20
1

22
320
21
360
26
100
28
1

30
110

180
                                    ELAPSED TIME, HOURS
6-
5'
1-
3-
2
C
•
-^
2168
) 10 80 120
i
•
1
1
i
i
DUMBER OF CARBON CYCLES'
10 12 11 16 18- 20 22
i i i | i i i
160 200 210 280 320
»- TT tt

-« 	
21
	 1_ -
360

-, 	 •-— '
26 28 30
i , .. -i— i
100 110 llo
                                   ELAPSED TIME/ HCUFIS

-------
Ash Content  -

The carbon is exposed to flue gas each carbon  cycle  during
which the carbon could pick up fly ash.  Exposure  for  21
cycles did not affect the ash content as shown by  the
relatively constant value in Figure 11.  The ash content
varied continuously between 4.4 to 5.2%, indicating  no
significant  fly ash pickup.

Carbon Burn-off -

Throughout the runs in both IR-1 and IR-2, the inlet and
outlet gas streams of the sulfur generator and H2S generator/
sulfur stripper are analyzed for H20, C02, CO, H2S,  H2, N2
and S02 with a process gas chromatograph.  Presumably  the
difference between the C02 and CO content of the inlet and
outlet gas streams is a measure of the carbon  burn-off in
these vessels.  The detailed gas analyses are  given  in
Appendix A-12.  No carbon monoxide was detected in the
off-gas of either vessel.  Carbon dioxide, however,  was
detected in  both off-gas streams during process H2S  gas
recycle as shown in Figure 12 for the IR-1 and IR-2  runs.

As seen from Figure 12, the C02 determined in  the sulfur
generator was about equal to that determined from the
H2S generator/S stripper, which necessarily implies  that
the burn-off occurs in the H2S generator/S stripper.   The
carbon dioxide evolved in the H2S generator/S  stripper
leveled off  at about the same rate in the IR-2 run as  the
IR-1 run.  The C02 evolution rate of about 0.026 Ib. C as
C02/hr.  (or  about 10 Ibs. C/Ton S02) is apparently indica-
tive of what might be typically expected.  This burn-off
rate agrees with that predicted from previous bench  scale
results of about 4 to 10 Ibs. C/Ton S02 sorbed.

The maximum  carbon burn-off that would be expected from
reaction with sorbed sulfuric acid, Equation (9), is
about 187 Ibs. C/Ton S02 sorbed.
           H2S04 + 1/2 C 	»•  1/2 C02 + S02 + H£0            (9)
The decrease in burn-off from 187 to 10 Ibs. C/Ton S02
sorbed corresponds to a reduction of about  95%.
                         45

-------
Figure  12.   Carbon dioxide  evolution as a  function  of  carbon
               cycle  time
=i
PsJ
S
V!
£ 1

_j
^
3
-j
o
UJ
CM
CD
f >

1
0.05-
0.05 •

0.04 '

0.03 •

0.02 •
0.01-
0
1
m-i -*l— iR-2
• I f
V . i \
^""v^ \ o A
^"Ssv»^ 1 vW *
^^S*~—— *t ** 1 ^^ * 0*

** ** *
1
NUMBER OF CARBON CYCLES'
2 4' 6 8 10 12 14 16 18 20 22 24 26 28 30
• t1ifttltlill.ll —
0 40 80 120 160 200 240 280 320 360 400 440 480
                                                                                As DETERMINES FROM ^S
                                                                                 GENERATOR/S STRJPPES


                                                                                As DETERMINED FROM
                                                                                 SULFUR GENERATOR
                            ELAPSED TIME, HOURS

-------
Pore Volume and Surface Area  -

The effect of repeated cycling of  the  carbon under process
conditions on total pore volume and surface area was
measured.  Total pore volume  and surface area which were
determined for samples of regenerated  carbon taken periodic-
ally during the two  integral  runs  are  shown in Figure  13.
Because of the scatter in the data, which is within
experimental error limits,  the apparent slightly increasing
nature of both the pore volume and surface area plots may
not be considered statistically significant.  Therefore,
it appears that both the carbon's  total pore volume and
surface area were not appreciably  changed with repeated
exposure to sorption/regeneration  conditions.

Regeneration -

Sulfur Generator -

The major indicators of the performance of the moving bed
sulfur- generator are the acid conversion to sulfur and H2S
utilization.  The results obtained for these responses in
this run are included in Figure 14.  In comparing the two
runs, IR-1 and IR-2, the H2S utilization improved toward
100% utilization as  the sulfur recovery increased.   On
the other hand, the  acid conversion decreased to a steady
value of about 70% since the  increased H£S utilization
corresponded to a decrease in the  stoichiometric quantity
of H£S recycled back to the sulfur generator.   This        '
decrease in acid conversion was accompanied by a slight
increase in S0£ evolution from the sulfur generator.
However,  it is felt  that the S02 evolution could be
reduced in fluid bed sulfur generators planned in future
development since the fluid bed would offer more uniform
temperature control.  Moreover, the evolved S02 could be
recycled to the S02  sorber with a minimal increase in
load on the sorber.   More importantly for the IR-2 run,
the increased amount of acid entering the H2S generator
did not result in any S02 evolution from that reactor,
indicating that the  effective conversion of acid to
elemental sulfur was completed'in  the H2S generator.

As pointed out earlier in this report,the various perform-
ance goals were based partially on past performance of
the various units operated separately.  These goals may
not be particularly representative of the performance of
the integral process.  Therefore,   it is necessary to view
the effect of these recent H2S utilization and acid conver-
sion to sulfur responses in terms  of the overall process
objectives, i.e.  S02 removal, sulfur recovery, and H2
                         47

-------
Figure 13.  Pore volume and surface area of recycled
            carbon
     0.07
                  100       200       300
                      Elapsed Time, hours
400
      620
                  100        200        300
                      Elapsed Time, hours
400
                           48

-------
Figure 14.  Sulfur generator performance
100-
s-e
2 9°'
i—
<
ISI
_ 1
= 80-
GO
cvj
re
70
C
& 100 i
Q
u
CO
o:
o
CO
t 80 •
CO
o
1—
z
i 60-
LLJ
>
z
o
CJ
o
0 40
0
1
•
m
CYLINDER H2S ~*
2468
) W 80 120
m
m
'— 1
'
.
2468
^0 80 120
IK-l — *•
_
1 '
1
"*~ PROCESS H2S
NUMBER OF CARBON CYCLES
110 12 I'l 16
160 200 240
ELAPSEH TIME, H
'
1 . i' 	 '

NUMBER OF CARBON CYCLES
10 12 1H 16
till
"•-— 1K-Z
1
CYLINDER _».
H2S
18 20 22
III
280 320
OURS
1,
1
IS 20 22
i i i
160 200 210 28H 320
ELAPSED TIME/ HOURS

l— i
— <
— P1gSS
24 26 28 30
350 400 440 480
-
1 »
i — I
24 26 28 30
360 400 440~" 480

-------
 input/S02  sorbed ratio.  As can be  seen  from Table 5, the
 reduction  in acid conversion to sulfur in  the sulfur
 generator  has not reduced the overall process performance.
 In  fact, the three above mentioned  responses have changed
 in  the  desired direction of increased S02  recovery as
 elemental  sulfur.

 Sulfur  Stripper/H£S Generator -

 As  shown by Figure 9  (inlet sulfur  loading vs. elapsed
 time),  there was successful regeneration of the carbon to
 the same conditions throughout the  runs,  Although the
 inlet sulfur loading  on the carbon  was slightly higher in
 the IR-2 run due to lower average temperatures in the H2S
 generator/S stripper, the S02 removal remained steady and
 well above the target goal of 90% throughout the run.

 As  shown in Figure 15, the lower H2 feed in IR-2, i.e. a
 H2/S02  sorbed ratio of 3.4 compared to 3.9 for IR-1,
 yielded 100% H2 utilization in the H2S generator.  The H2
 utilization to H2S, however, steadily decreased during the
 IR-2-A run after the switch from cylinder H2S from 81%, to a
 steady  value of approximately 61% which was about the same
 in  the  IR-2-C run.  This decrease in the production of H2S
 during  the run was probably due to  the shift in the distri-
 bution  of  the acid conversion reaction between the sulfur
 generator and H2S generator where the acid entering the H2S
 generator was effectively converted to sulfur.  As the
 distribution of this conversion between the reactors stabi-
 lizes,  the H2S production reached steady state.

 Sulfur  Condenser -

 Previously, sulfur recovery was reported on the basis
 of  the  percent of stripped sulfur that was recovered.  But
 in  the  IR-2 run with more acid being converted to sulfur
 in  the H2S generator and the subsequent reduction in H2S
 production, the percent of the stripped sulfur recovered
 exceeded the 25% goal.  To provide  a more representative
 evaluation of the sulfur condenser performance, the percent
 sulfur recovered based on the amount sorbed as S02 is pre-
 sented  in Figure  16  for both the  IR-1 and IR-2 runs.  The
 lower H2S production or lower H2S breakthrough from the
 sulfur generator resulted in a higher sulfur recovery.  For
 IR-2, this response variable fluctuated, but steadied out
 to approximately 88% as the H2S production stabilized com-
pared to a goal of 10070.  This was  significantly higher
than the 75% achieved for the IR-1  run.
                         50

-------
                             Figure  15.   H2S  generator/sulfur  stripper  performance
                                                                  IR-2
Ul
           100 T
            90-
         CM
            70
        <
        rsl
           100 T
            80 ••
            60 ••
1
CYLINDER H2S ""**
•
2 1 6 8 .
0 40 80 120
1KB

' 1 ,
t
2468
0 40 SO 120
	 — 	 - 	 • i 	 1 i
1
|
- PROCESS H2S 1 ^P -
1
1
1
1
NUMBER OF CARBON CYCLES
10 12 14 16 18 20 22
160 200 240 280 320
ELAPSED TIME, HOURS
1
|
I , ,
1
1

""I
NUMBER OF CARBON CYCLES |
10 12 14 16 18 20 22
II 1 f I t i
160 200 240 280 320
— pT2sss
24 26 28 30
360 400 440 480
^^^^j
r^™""i •
, a
24 2(6 23 50
360 400 440 tsn
                                                     ELAPSED TIME, HOURS

-------
                                       Figure  16.   Sulfur condenser performance
m
         o
         UJ
         CQ
         o:
         o
         oo

         CM
         O
         0
         u
100-
90 •
80-
70-
60 •
50 •
40 •
30
IR-1 »!•» IR-2 , 	 ,
i -H!-_r-' "*"'
•""i
1- §
— * , 	 _ 	 _j
i
i
i
NUMBER OF CARBCN CYCLES
2 4 6 8 10 12 14 16 18 23 2.7 24 26 28 30
iti Tiii i I t i ifit
                       40       80     120     160      200      2*;0      280

                                                        ELAPSE! TIME/ HOURS
320
360
403
440
480

-------
4.2.3  Material  Balances

       Carbon  Balance -

       An attempt was made to perform a carbon balance  in  each of
       the  integral  runs;  but, due to excess carbon  spillage in
       the  IR-1  run,  only an overall carbon balance  was made for
       the  IR-2  run.

       For  the carbon balance that was determined for Run  IR-2,
       the  carbon  loop was closed.  At the start of  the run  the
       carbon  fed  to the reactors and inventory hopper was weighed
       carefully to  determine the initial input to the  system.
       Carbon  leaving the system during the run as dust, leakage,
       or C02  was measured; and after completion of  the run  all
       material  was  removed from the system and weighed.   Material
       weight  was  corrected for sulfur and moisture  content  to
       obtain  the  true weight of carbon.  The carbon balance is
       presented in  Table 6.
                 Table 6.   CARBON BALANCE FOR
                           IR-2 INTEGRAL RUN
IN
SYSTEM CHARGE = 428 IBS. C



.

TOTAL IN =428 Lbs. C
OUT
CARBON REMOVED FROM
INTEGRAL PILOT PLANT
S02 SORBER CYCLONE
S02 SORBER DUST FILTER* =
CARBON SAMPLES
C02 EVOLUTION
LEAKAGE AT BUCKET
ELEVATOR
TOTAL OUT
359
31
20
3.
3.
6
423
Lbs
Lbs
Lbs
5 Lbs
6 Lbs
Lbs
Lbs.
. C
. C
. C
. C
. C
. C
C
              *Based on constant rate of collection for the last steady
              state period in IR-1 run of 0.12 Ib. C/hr.
                                53

-------
 Hydrogen  Balance  -

 The  hydrogen balance  around  the H2S generator/S  stripper
 is given  in Table 7.  The  sources of hydrogen  input were
 hydrogen  in the inlet gas  and hydrogen as adsorbed water
 and  sulfuric acid on  the incoming carbon.  Hydrogen output
 consisted of H2S, H2  and H20 in the outlet gas.
        Table  7.  HYDROGEN BALANCE FOR H2S
                  GENERATOR/SULFUR STRIPPER
Run
IR-1
AVG.
IR-2
AVG.
Period
1
2
3
4
5
158 Mrs.
6
7
8
113 Mrs.
H2 Input,
Lbs./Hr.
Gas - H2
0.451
0.591
0.591
0.663
0.601
0.587
0.492
0.492
0.492
0.492
Carbon - H20
0.043
0.048
0.040
0.015
0.014
0.029
0.094
0.140
0.163
0.132
H2 Output,
Lbs./Hr.
Gas - H2S
0.321
0.409
0.409
0.480
0.417
0.416
0.397
0.332
0.300
0.343
Gas - H2
0
0.002
0.005
0.01
0
0.005
0
0
0
0
Gas - 1120
0.143
0.214
0.229
0.194
0.217
0.196
0.103
0.233
0.234
0.190
Total ,
Lbs./Hr.
In
0.494
0.639
0.631
0.678
0.615
0.616
0.586
0.632
0.655
0.624
Out
0.464
0.625
0.643
0.684
0.634
0.617
0.500
0.565
0.534
0.533
For the IR-1 run, the difference between input and output
averaged less than 170.  For IR-2, the hydrogen output
averages 11% below the input.  This 11/4 discrepancy could
reasonably be attributed to experimental error in gas
analyses, in flow rate measurements and in the inlet carbon
moisture analysis.  The measured carbon moisture content for
IR-2 is higher than the theoretical value predicted by
assuming that the water associated with the sulfuric acid
on the carbon is at equilibrium between the gas and carbon
phases on discharge of the carbon from the sulfur generator.
Therefore, the 11% error in the H2 material balance is
considered within experimental error.

Sulfur Balance -

During the integral runs, the carbon and gas streams into
and out of each vessel were monitored to obtain a total
sulfur balance for each unit and for the total pilot plant.
For the IR-1 integral run, there were five steady state
periods.   Two of the periods with cylinder H2S gave an
                         54

-------
overall sulfur balance of IN/OUT of 8.65/8.93 and 8.31/
6.41.  Some recirculated sulfur was carried over into the
sulfur product due to flooding of the mist eliminator in the
sulfur condenser causing uncertainties in these balances.
This was rectified and the overall sulfur balances with
process H2S were very good.  The sulfur balances for the
other periods are given  in Table 8.    Prior to the last
two steady state periods, a problem with the sulfur
    Table  8.   SULFUR BALANCE FOR INTEGRAL RUNS
Integral
Run
IR-1
IR-2
Overall
Period
3
4
5
7
8
87 Mrs.
Input
Sulfur
S02
Sorber,
Ibs./hr.
2.54
2.68
2.68
2.43
2.40 '
2.42
Sulfur Outputs,
Ibs./hr.
S02
Sorber
Off -Gas
0.19
0.20
0.20
0.13
0.16
0.14
Sulfur
Gen.
Off-Gas
0.92
0.55
0.46
0.38
0.31
Sulfur
Cond.
Product
1.58
1.61
1.86
1.94
1.94
1.85
Total
Output,
Ibs./hr.
2.73
2.61
2.53
2.48
2.30
Input/Output
Ratio
--
0.96
0.99
1.05
analysis of the sulfur generator off-gas was rectified so
the last two sulfur material balances should be very
reliable.  The sulfur balance IN/OUT ratios were 2.54/2.43,
2.70/2.64, and 2.78/2.57 as given in Table  8  for  opera-
tion with process H2S.   This gave an average overall sulfur
balance while on process H2S of 2.64 Ibs. S/hr. in and
                         55

-------
        2.56  Ibs. S/hr. out tor an average discrepancy of about 37o,
        or  9770 of the  sulfur in or out was accounted for.  This
        indicates that especially for the Periods A and B, the
        conversion calculations and indicators of carbon perform-
        ance  are accurate.

        An  overall sulfur balance around the pilot plant system was
        calculated for the total period of process H2S utilization
        during the IR-2 runs.  The results of the sulfur balance
        are given  in  Figure 17.   The amount of sulfur entering
        the sorber was 211 pounds and the amount leaving the sorber
        was 12 pounds, so that the amount sorbed which entered the
        regeneration system was 199 pounds.  Of this, 27 pounds
        passed out of  the system in the sulfur generator off-gas
        and 161 pounds was recovered in the sulfur condenser.  This
        accounts for all but 11 pounds, or 5%, of the input sulfur,
        and this is probably within experimental error.

        Twenty-eight pounds of sulfur which leaked out of the sulfur
        pump were collected after the run was over and included in
        the material balance as part of the 161 pounds of recovered
        sulfur, but the 28 pounds may not have accounted for the
        entire leak.

        Sulfur balances for periods of operation with process H2S
        during both the IR-1 and IR-2 runs are shown in Table 8.
        The results are presented as rates of sulfur input and
        output during each interval.  Input consists of S02
        adsorbed from the flue gas, and output is divided into
        three categories: ,1)  sulfur contained in the S02 sorber
        off-gas as S02 or sorbed on the carbon dust, 2) sulfur in
        the sulfur generator off-gas as S02 or H2S, and 3) sulfur
        product collected in the condenser.  The values for sulfur
        collection rate in the condenser have not been corrected
        for the sulfur leak,  because of uncertainty in distribution
        the leak over the various periods.  It is clear from
        Table 8,  however,  that   the leak was worst during the
        periods from 81 to 126 hours, when the sulfur collection
        rate fell significantly.

4.2.4  Process Control

       Overall process control of the integral pilot plant was
       achieved by a combination of automatic and manual control
       of critical process variables.   Of these variables, only
       the temperatures  of the S02 sorber's second stage and the
       sulfur stripper/H2S generator,  the overall carbon recycle
       rate,  and the carbon levels in the seal legs of the
       reactors  were under automatic control.  The other vari-
       ables, mainly gas rates and compositions, were controlled
                               56

-------
           Figure  17.   Sulfur balance  for IR-2 run during operation
                        under  process H2S
211  IBS.
                     REGENERATED
                       CARBON
                                                     DUST
                                        CYCLONE      FILTER
                                                          OFF-GAS
                     S02 SORBER
                                                          OFF-GAS
                               SULFUR
                               GENERATOR

-------
manually on the basis of material balances and the process
performance responses.

Manual control of the integral plant was accomplished by
maintaining the flue gas rate and S02 concentration con-
stant to the S02 sorber and adjusting the gas rates and
compositions to the other reactors.  During operation with
process H2S, the hydrogen/nitrogen rate to the sulfur
stripper/H2S generator was the only gas flow varied.

The overall carbon recirculation rate was controlled auto-
matically at 30 Ibs./hr. by a gravimetric solids rate feeder
to the S02 sorber.  However, at several points in the
carbon loop, the carbon flow was under local automatic con-
trol to maintain the desired levels in the carbon seal legs
to and from the reactor vessels.  The control variable
for these seal legs was the pressure drop of N2 purge gas
in the legs.

In the S02 sorber, the flue gas entered at 300°F and was
cooled to about 175°F at the second stage by direct water
injection into the second stage carbon bed.  This tempera-
ture was maintained by automatic control of the water
injection rate.  The temperatures of the remaining three
stages were not controlled, but allowed to reach their
equilibrium temperatures.

The temperatures of the sulfur stripper/H2S generator,
sulfur generator, and sulfur stripper were maintained
by electrical heating.    In  the  sulfur stripper/
H2S generator,  the temperature was automatically controlled
by regulating the current to electrical resistance heaters
on the reactor walls.  Temperature in the acid converter on
the other hand were controlled primarily by manually setting
the voltage rheostats to the electrical resistance heaters.
Response to changes in temperature in this reactor was slow
because of the heat transfer characteristics of moving bed
reactors.  Also, due to the exothermic reaction occurring
in this moving bed reactor, localized over-temperatures were
somewhat compensated for by adjusting the water content on
the carbon before it entered the reactor.  The temperature
in the sulfur condenser was also controlled manually be
regulating steam pressures to the exchanger and tracing.
                         58

-------
i.2.5  Process Concept Modifications

       Based on findings which arose at various stages in this
       development program,  several modifications of the original
       process concept have evolved that not only have simplified
       process design and control requirements but also have
       reduced the capital and operating costs of the system.
       Briefly, these modifications involve:

            1)  The method in which the flue gas temperature
                is controlled within the S02 sorber

            2)  The combination of sulfur stripping and H2S
                generation into one process step

            3)  The fact that some of the acid can be reduced
                in the sulfur stripper/H2S generator without
                deleterious  effects on the process.

       In the S02 sorber,  the first stage was operated at the
       stack gas temperature of approximately 300°F to remove  the
       30 to 50 ppm S03,  and the second stage was operated at
       175°F to facilitate SC>2 removal.   The temperatures of the
       remaining fluid bed stages were not controlled,  but were
       allowed to reach their equilibrium temperatures.   To cool
       the flue gas between  the first and second stages,  an
       external shell and tube heat exchanger was installed and
       evaluated initially.   The off-gas from the first  stage  was
       withdrawn from the sorber, circulated through the  exchanger
       and returned to the sorber below the second carbon bed.
       Since certain design  and economic constraints made this
       heat exchanger method of gas cooling particularly  unattrac-
       tive for large process units,  direct water injection into
       the second stage fluid bed was installed.   After  the posi-
       tion of this water spray nozzle and the proper nozzle
       pressure were determined,  this latter method was used
       satisfactorily during the integral runs and has  been
       incorporated into the designs for the larger systems.   As
       a result, the capital investment was reduced.

       In the original process concept,  the thermal stripping  of
       the elemental sulfur  from the carbon and subsequent partial
       reaction of this sulfur with hydrogen to H2S were  to be
       performed in separate process vessels.   However,  bench
       scale and pre-integral pilot plant testing indicated that
       these two process  steps could be carried out satisfactorily
       in one reaction vessel, thereby appreciably simplifying
       the overall process.   This process change was subsequently
       incorporated into the integral pilot plant as a fluid bed
       sulfur stripper/H2S generator.
                               59

-------
The design and operating conditions for the sulfur generator
were selected initially to achieve 997o conversion of acid
to elemental sulfur with maximum H2S utilization.  It was
felt that this high level of acid conversion was necessary
because the introduction of acid loaded carbon into the
high-temperature H£S generator could increase carbon burn-
off and S02 evolution.  During the integral runs, however,
it was found that part of the acid on the carbon was reduced
to sulfur in the sulfur stripper/H2S generator without
deleterious effects of S02 evolution or enhanced carbon
burn-off.  As a result the integral process can be operated
with lower H2S recycle and, consequently, without H2S break-
through from the sulfur generator.  Moreover, reducing the
amount of acid conversion in the acid converter should
allow greater flexibility in process control.
                        60

-------
                            SECTION 5
                 PRE-INTEGRAL PROCESS DEVELOPMENT


5.1    APPARATUS AND PROCEDURE

       There were numerous types of reactors and procedures that
       were used in the process development.  In general the
       studies can be classified by the type of reactor which
       included thermogravimetric,  fixed bed, moving bed, batch
       fluid bed, and multistage fluid bed reactors.  For each
       specific reactor type, the procedures are similar for
       studying the separate process steps and many of the
       procedures will thus be grouped as to reactor type.

5.1.1  Thermogravimetric Reactor

       In many adsorption or desorption processes using activated
       carbon, the change in carbon weight is directly proportional
       to the amount of a constituent adsorbed.  Therefore, by sus-
       pending a small container with a carbon sample from a gravi
       metric balance and placing the suspended sample in a therm-
       ally controlled environment, the carbon can be exposed to
       a gas containing some constituent to be removed and the
       amount of pickup be followed directly as a function of time.
       Such an apparatus was assembled as described below, but
       since it was only used under this contract for studying the
       kinetics of S02 sorption, frequent mention will be made to
       S02 even though the equipment can be used more generally.

       The apparatus is shown schematically in Figure 18.
       Simulated flue gas or some other suitable gas mixture is
       mixed by metering sulfur dioxide, nitric oxide, and nitrogen
       through individually calibrated Brooks "Sho-Rate" rotameters
       equipped with dual floats (steel and glass) and highly
       stable "ELF" nonrising stem needle valves.  Use of these
       valves together with ordinary cylinder gas pressure
       regulators allows accurately reproducible setting of flow
       rates.  In clean rotameter tubes, drift in flow indication
       has never exceeded 0.5% of the tube scale length over the
       course of any experiment.  Flow rates in a convenient range
       are obtained by using cylinder gases containing approximately
       1% S0£ in N£ or He and 0.3% NO in N2 or He.  Actual cylinder
       concentrations for the S02 tanks are determined by wet
       chemical analysis.  The NO tanks are calibrated by the
       manufacturer.

       Water vapor is added by means of a bubbler apparatus con-
       sisting of two 14" x 2-1/2" OD gas scrubbers connected in
       series and completely submerged in a thermostat bath.  A
       head of about 6-10 inches of water was maintained above
                               61

-------
                                      Figure  18.   Thermogravimetric apparatus
to
       A - Magnehelic Pressure Gauge
       B - Humidifier Thermostat
       C - Sample Thermostat
       0 - Bucket Envelope
E - Fiber Envelope
F - Electrobalance
G - Oven Temperature Regulator
H'- Vacuum Pump
J - Thermocouple
K - Vent
L - Inlet Toggle Valve
M - Exit Toggle Valve

-------
the fritted dispersion tube in the first bubbler and about
2 inches in the second.  Analysis of the gas effluent from
the humidifier assembly using a Cambridge Systems dew point
hygrometer has shown that saturation is achieved at all
thermostat temperatures and carrier gas flow rates tested
in the S02 sorption rate measurements.

The dry and humidified gas streams are mixed and then pass
either to the sample or to a vent.  This enables the
simulated flue gas mixture to be completely mixed before
introducing to the sample.

A needle valve in the vent line allows the pressure drop
in the vent and sorption lines to be equalized so that flow
may be switched at the start of a run without disturbance
of pressures in the humidifier or the rotameters.

In the main part of the reactor, shown in detail in Figure
19, the carbon sample is held in a cylindrical fused silica
bucket suspended from the beam of a Cahn RG electrobalance.
The gas mixture is admitted to the sample through a nozzle
positioned inside the bucket about 3 to 5 mm above the
surface of the carbon layer.  The nozzle was made of Teflon
stock drilled to give dispersion of gas over the whole
sample.  Gas is conducted to the nozzle through a 1/16"
stainless steel tube which passes through a Teflon sealed
thermocouple gland attached by Swagelok fittings to a
standard glass ball and socket joint which is sealed to the
envelope in which the sample is suspended.  Use of the gland
and joint allows the lateral, rotary and angular motion
required to position the dispersion nozzle within the bucket.
The gas injection tube is connected to the gas mixing
system by a coil of 15 feet of 1/16" SS tubing which allows
preheating of the influent gas and provides sufficient
flexibility to avoid undue strain on the glass-to-tnetal
connections.

Also connected to the sample bucket envelope is a separate
nitrogen purge line which is used to prevent air from
entering the sorption apparatus after outgassing the sample
and during the time that gas mixtures are being set up
prior to a run.

The sample bucket envelope is removable so that samples may
be changed and it is attached by a ball and socket joint
to an upper tube which surrounds the bucket suspension
fiber.  The fiber envelope is about 14 inches long by 25 mm
in diameter and contains a 3" section of 10 mm tubing at
the upper end in order to prevent back diffusion of air
into the sample section.  Just above this construction, a
side arm is attached which is connected to a Cast vacuum
                       63

-------
Figure  19.  Detail of  the thermogravimetric reactor
             sample bucket envelope
                          SJ 35/25  Pyrex Rail  Joint
                          SJ 18/9 Pyrex Ball Joint
                          Swagelok  Fittings and Thermocouple Gland
                          Gas'Injection Tube
                          Purge Gas Line
                          Sample Bucket
                          Dispersion Nozzle
                          Thermocouple
                          Suspension Fiber
                          Swagelok  Connection to Preheat Coil
                        64

-------
pump through a needle valve and rotameter.  During a run,
the flow through the vacuum system is maintained at such
a rate that no corrosive or toxic gases are allowed to
enter the atmosphere or the electrobalance mechanism case.
Mixture with sufficient room air entering through the open
top of the fiber envelope also prevents condensation of
water from the gas mixture on the upper end of the fiber
or in the vacuum system.

Reaction temperature is maintained by an air bath oven
which is placed up around the reactor assembly and encloses
the sample bucket envelope and the preheat coils attached
to the gas injection and purge lines.  The oven, which is
very well insulated but weighs only about 10 pounds, was
constructed in two parts.  The bottom section was made
from a 12" x 6" x 7" aluminum chassis box, insulated on the
inside with 1/2" Fiberfrax ceramic fiber blanket.  This
section contains two 660 watt exposed coil heating elements
and a centrifugal blower with an external fan cooled motor,
The upper section, which formed the air bath, was constructed
from 1/2" Fiberfrax blanket sandwiched between 9 mil
corrugated aluminum sheeting.  In operation hot air is blown
out of the lower section, circulates through the air bath
and returns through a 2-1/2" hole in the top of the heater
chamber.  Adequate temperature control is obtained using a
Honeywell time proportioning controller with a thermistor
sensor mounted in the outlet of the blower.  Reaction tempe-
rature is measured*by a thermocouple placed just above the
sample bucket.

All lines coming into contact with water vapor are traced
with electrical heating tapes to prevent condensation.

Weight pickup of SC>2 as sulfuric acid is measured by a Cahn
RG electrobalance which is connected to a 10" Texas
Instruments recorder to give a continuous readout of weight
versus time.  The sensitivity of the balance system is such
that a full chart width deflection corresponds to a pickup
of about 5 Ibs. S02/100 Ibs. carbon.  In operation the flow
of gases into the sample bucket causes a weight deflection
corresponding to the flow rate.  At the rate used in these
experiments (1,000 cc/min.) the deflection is about 10
milligrams.  Because the flow rates are very stable and
reproducible, the deflection is always constant over the
course of any experiment and the flow deflection can simply
be subtracted from the observed weight changes due to S02
sorption.  Gas flow around the bucket and suspension fiber
also introduces a small amount of vibration into the weighing
system.  However the resultant noise band is less than 0.1 mg
wide in most cases and had no effect on the accuracy of  the
measurements.


                         65

-------
       The weight pickup vs. time curves are differentiated by
       finding straight line slopes across closely spaced time
       intervals and plotting these rates vs. the mean loading
       in the corresponding interval.
5.1.2  Fixed Bed
       The basic components of a typical fixed bed apparatus are
       shown schematically in Figure 20,  The reactor, containing
       a known weight of carbon vertically oriented, typically
       has been 1" diameter.  The reactor is equipped with a
       heating system consisting of electrical heaters wrapped
       around the pipe and a rheostat to control the heat input.
       Carbon temperature is measured with a thermocouple inserted
       into the carbon bed and gas temperature is also measured
       with a thermocouple.  A gas mixing system provides control
       of reactant gas flows.  Rotameters are used to measure the
       gas flow rates.  The direction of gas flow is generally
       downward through the bed to avoid any possibility of
       fluidizing the bed.  A gas chromatograph was used to moni-
       tor inlet and outlet concentrations of the gas constituents.

       In operating a fixed bed system, the general procedure is
       to establish the desired inlet gas conditions with the gas
       bypassing the reactor, and to establish the desired carbon
       bed temperature with the heaters.  When all conditions are
       set, the gas flow is switched to enter the reactor.  The
       outlet gas is analyzed to give concentration vs. time data.
       The data from this type of experiment is transient rather
       than steady state.  Outlet gas concentrations change con-
       tinuously and eventually approach the inlet concentrations.
       Following the run the carbon is removed and analyzed for
       sulfur compounds.

       The data can be analyzed to provide screening information
       on overall degree and rates of conversion and type type of
       reactor was used to study each of the process reaction
       steps.
5.1.3  Moving Bed
       The main features of a moving bed system are shown schematic-
       ally in Figure 21.  There are many similarities to the fixed
       bed, including the systems for gas mixing, reactor heating,
       and gas analysis, the vertical reactor orientation, and
       the fact that the carbon beds are not fluidized.  The main
       differences are that the moving bed reactor is a continuous
       flow system which operates under steady state conditions
       and the direction of gas flow is upward through the bed in
                              66

-------
Figure  20.   Fixed  bed reactor system
                         Thermocouple
     Gas
   Mixing
   System
             RheostaT Heater
            ..r
           Vent
            or
         Gas
         Anal.
                                   1"0  Fixed Bed
                                    Carbon Sample
                 67

-------
Figure  21.  Moving bed  reactor  system
       tarbon
       Feeder
\
                            Vent




Tae
OdS
Mixing
System



Heater C
or (
Heating /
System ?





i





i
iMHMI
••••
*

/
TGas


Moving Bed
Reactor
Thermocouples (Typical)
__i
— Carbon

_. .Cone Discharge Angled To
Insure Plug Flow of
Carbon and Carbon Forms
                        Gas Distributor
                  £1
                Carbon
               Discharge
                Feeder
                                  Carbon
                  68

-------
       order to provide countercurrent contact with the carbon
       phase.   In addition,  the moving bed requires a carbon
       flow control system to maintain the desired carbon flow
       rate through the reactor.  Carbon feeders are used at both
       the inlet and outlet  of the reactor.   The carbon flow
       distribution through the reactor is close to theoretical
       plug flow.

       The moving bed reactor is subject to temperature control
       problems as the vessel size increases due to poor heat
       transfer characteristics.  Its main advantage is the
       capability for a long carbon residence time and low gas space
       velocity, two parameters which are important in obtaining
       high conversion of reactants.

       Low gas space velocities can be obtained in a moving bed
       because the minimum fluidizing velocity constraint which
       applies to a fluid bed reactor is not applicable in a moving
       bed.

       A bench scale (1-1/2" dia.) and pilot (8" dia.) moving bed
       reactor were used to study the conversion of sulfuria acid
       to sulfur by reaction with H2S.  Other reaction steps were
       not studied in the moving bed equipment.

5.1.4  Batch Fluid Bed

       The batch fluid bed reactor shown in Figure 22 is a differ-
       ential reactor useful primarily for studying kinetics of
       a reaction.  The approach is to fluidize a batch carbon
       sample with a reactant gas at a space velocity sufficiently
       high that the concentration of the gas reactant remains
       essentially unchanged as it passes through the carbon bed.
       Carbon samples are taken from the bed at closely spaced
       time intervals and are analyzed to generate data that is
       convenient for a kinetic analysis.

       The equipment consists of a cylindrical vessel with a gas
       distributor plate to support the carbon bed.  Electrical
       heaters provide heat input through the walls and are con-
       trolled manually with a rheostat.  Temperature can be con-
       trolled accurately and is uniform throughout the bed
       because of the excellent heat transfer and mixing charac-
       teristics of the fluidized bed.  A gas mixing system pro-
       vides the desired gas flow rate and constituent concentra-
       tions,  and the gas is introduced below the distributor plate.
                       s
       A 4 inch diameter reactor of this type was used to develop
       a model for the reaction of acid with H2S to form elemental
       sulfur and to study the thermal stripping of sulfur.
                                69

-------
     Figure 22.   Batch fluid bed  reactor
      Thermocouple
  Gas
Mixing
System
               Heater
                        To Vent or -
                            i
            Gas
           Anal.
_ Manual Carbon Sampler
   4"0 Batch Fluid
   Bed Reactor


    Fluid Bed of Carbon
                                     Drilled Gas Distributor
                                       Plate
                         Manual
                         Carbon
                        Discharge
                       70

-------
5.1.5  Multistage Fluid Bed Reactor

       A multistage fluid bed reactor system is shown in Figure 23.
       Carbon is fed at a constant rate onto the top stage of the
       reactor and flows by gravity from one stage to the next.
       Gas distributor plates support the carbon beds.   The plates
       are drilled to a specified percent open area, typically 8%
       with 0.125" dia. holes.   The expanded bed depth,  determined
       by the height of the overflow weirs,  is generally limited
       to a value equal to the vessel diameter in order  to
       minimize slugging of the bed.   Downcomers extend  into the
       carbon beds to within about 1-1/2" of the distributor
       plates.  The downcomers contain baffling to prevent
       slugging and improve the carbon flow characteristics.
       Carbon passes out of the reactor into a seal leg  designed
       to prevent the escape of any reactant gas.   The carbon level
       in the seal leg is controlled automatically.

       Reactant gas enters below the bottom distributor  plate and
       passes upward through the reactor, countercurrent to the
       carbon flow.  The gas is sampled for analysis at  the inlet
       and outlet of the reactor, and may also be sampled at each
       stage of the reactor.

       Temperature is controlled by electrical heaters wrapped on
       the outside of the reactor.  Separate temperature control
       is frequently provided for each stage or for every two
       stages.

       The multistage fluid bed reactor operates at steady state
       conditions.  The time needed to reach steady state after
       start-up or after a change in operating conditions may be
       fairly long; however, so this type1of reactor is  not
        fenerally used in extensive kinetic studies.   Multistage
        luid bed units ranging from 4-18 inches in diameter were
       used to study the separate reaction steps and for thermal
       sulfur stripping.

5.1.6  Sulfur-Carbon Thermal Equilibrium

       Adsorption isotherm points were obtained by saturating an
       inert gas stream with sulfur vapor at particular  temper-
       atures in order to produce a known partial pressure of
       sulfur, and passing this stream over carbon held at the
       desired adsorption temperature for a length of time suffi-
       cient to establish equilibrium.  Sulfur loading on the
       carbon was determined for each set of equilibrium conditions
       by means of combustion analysis of the individual samples.
                                   71

-------
          Figure  23.  Multistage fluidized bed  reactor
  Gas Anal.
  (Typical)

Carbon Anal.
  (Typical)
                     Carbon
                    Feed System
                  Heater;
                (Typical)
     Thermocouple  (Typi
            Pressure
              Drop   ©
           (Typical)
       Gas  Source
        or  Mixing
         System
I
                     Gas
                  Analysis
                                                   1
                            Gas
                          Analysis

                       Multistage Fluidized
                       Bed Reactor
                        Overflow Weirs
                        Baffling

                        •Downcomer

                        .Fluidized Carbon Bed

                        Drilled Gas
                           Distributor Plate
                      Level  Control
                                72

-------
       The  apparatus  used is  shown diagramatically in Figure 24.
       In operation about 0.5 gm.  of carbon sample was introduced,
       and  after purging with nitrogen,  the vapor generator and
       adsorption tubes  were  heated to the desired temperatures by
       manual control of Variac transformers.   Nitrogen flow
       through the apparatus  was maintained at a rate (10 cc/min.)
       low  enough to  attain saturation in the  relatively inefficient
       generator.   Connecting tubing between the generator and
       adsorber was heated to a temperature above that of the
       sulfur in order to prevent  condensation.   The exit tube
       at the bottom  of  the adsorber was also  heated down to a
       condensation trap which was used to prevent sulfur from
       plugging the exit line.   Nitrogen was injected into the
       line at this point in  order to seal against entry of air
       from the exit.

       After an equilibrium time of 9-1/2 hours  which was set  based
       on observations during the  first experiment in which the
       highest sulfur pickup  was achieved,  gas flow through the
       generator was  shut off.   The stopcock between generator
       and  adsorber was  closed, and the entire apparatus was
       removed from the  furnaces and allowed to  cool.   The carbon
       was  then weighed, avoiding  contact with atmospheric moisture,
       and  the sulfur loading of the entire sample determined  using
       a Dietert Sulfur  Analyzer.

       The  carbon starting material used in these experiments  had
       initially been loaded  with  about 24% sulfur by complete
       reduction of HoS04 in  sulfur generation experiments.  Carbon
       having an initial sulfur load was used  in order to reproduce
       adsorbent properties due to S02 pickup  and reduction in
       accordance with our intent  to use these data in connection
       with thermal stripping work.

       Since the existence of chemisorbed sulfur on carbon has been
       suggested by previous  thermal stripping studies in which a
       residual loading  remained after extended  purging with inert
       gas,  additional runs were made in which samples were purged
       for  9.5 hours  with pure nitrogen at several temperatures
       and  then analyzed as noted  above.


5.1.7   Solvent Extraction of  Sulfur Procedures

       Screening evaluations  were  made of sulfur extraction with
       carbon disulfide, ammonium  sulfide and  xylene by successive
       stages of slurry  contact in stirred beakers.  Since the
       procedures differed slightly for each solvent they will be
       described separately.
                                 73

-------
                      Figure 24.   Sulfur adsorption  apparatus
                          Thermocouples
i 1
Tapered Joint LJ
"^-C/^
Variac < — l //

Vycor
Adsorption
Tube
' -^^


Carbon __^_^^
^ M •* ^ -4» A j4 n^r-i^
Fn uted DISK — -
Variac < — °








r






VL

i
A
^
,i





i*v
!U*

Vr/
i
Hoa-Hnn Tanp 	 1 ...

i

;




j^^






	 _,



<
Heating Tape — ^
	 ^j f
i / ) ' ' "'
Stopcock '



Tube Furnace ~^_^
^^^«


"" Thenrocouple
Shield












t/\





L
—
i i
— »_

_^
' */
\ ' Taoered \\
\ \ 	 '°H Variac
Sulfur
Variac
  Vent  <
                Condenser
                                                                                =:—NZ
                                                                                    N2

-------
Carbon Bisulfide Extraction -

Twenty grams of a sulfur loaded carbon were slurried in a
vertical stoppered flask for about 0.5 hr. at room temper-
ature.  The solvent was then decanted and evaporated to
recover the dissolved sulfur for weighing.  A fresh
quantity of solvent was added and the procedure repeated
to the desired number of extraction stages.  After the
final extraction the carbon was freed of excess solvent on
a steam bath and then oven dried.  Samples of the carbon
were also taken at the end of each extraction stage, dried
and analyzed for sulfur.

Ammonium Sulfide Extraction -

Forty grams of sulfur loaded carbon were slurried in a
beaker at room temperature with 100 ml. of 20% ammonium
sulfide for 15 minutes.  The ammonium sulfide solution was
then filtered off and the slurrying repeated with 100 ml.  of
fresh solution.  After each stage of extraction a carbon
sample was taken, slurry washed with water to remove excess
ammonium sulfide, oven dried and analyzed for sulfur.  The
extraction solutions after each stage were acidified and
boiled to liberate recovered sulfur which was then reclaimed
through carbon disulfide extraction and dried.

Xylene Extraction -

Twenty grams of a loaded carbon were slurried with 50 ml. of
xylene, in a flask with an extended neck and thermometer,
at a temperature of 105-110°C (220-230°F) for 20 to 30
minutes.  The solvent was decanted and a fresh quantity of
solvent added and the procedure repeated.  The carbon was
sampled between extraction stages.  The final carbon was
freed of xylene by boiling in water and then oven dried
overnight.  The xylene extracts after each stage were
evaporated on a steam bath and residues oven dried for an
hour before weighing to determine sulfur extracted.

Bench Scale Extractor -

A schematic of a recirculating bench scale apparatus used
in a study of sulfur extraction from carbon is shown in
Figure 25.  The carbon to be extracted was placed in a
quartz tube inside the Hoskins furnace.  The furnace was
then heated to the extraction temperature and the system
was purged with pure nitrogen or helium metered from a gas
cylinder to expel air from the system.  The extractant
was then added to the system from a separatory funnel and
recirculated through the system by a stainless steel
centrifugal pump at a rate of 35 ml./min. for 30 minutes.
                          75

-------
                Figure 25.   Flow schematic of  recycle
                              extraction apparatus
Syringe
 Pump
            Distilled
             Flash
           Vaporizer
                                                                        N2
                                                                     Cylinder
Flowmeter
                Furnace with
                 Quartz Tube
                Carbon
                Sample
                                                         Pump
                           Drain    Exhaust  Drain
                                  76

-------
       The solvent  was  then drained from the system,  fresh solvent
       added and the procedure repeated for the desired number of
       stages.   Samples of carbon were taken after each extraction
       stage,  dried and analyzed for sulfur.   The extractant  was
       saved for sulfur analysis.   After the final extraction the
       carbon was steamed for eight hours at a known  rate by
       metering water with a syringe pump to a flash  vaporizer to
       produce the  steam.   The carbon was then cooled under a
       nitrogen purge,  oven dried and analyzed for sulfur.  The
       system contained a water leg for pressure relief,  a
       manometer for pressure measurement and a trap  for solvent
       surges.

5.1.8  Procedures in Bench Scale H2S Generation Studies

       Apparatus for investigating reaction rates of  hydrogen
       and sulfur vapor is shown schematically in Figure  26.
       Component testing was performed as follows:

       1.   Sulfur Vapor Generator - The generator,  shown  in
           Figure 27, operates by saturation of a stream of
           nitrogen which is bubbled through molten sulfur.
           This stream  is  mixed with the main reaction gas
           stream and manipulation of the sulfur temperature  and
           the relative flow rates allows adjustment  of the final
           sulfur vapor concentration.   Heat is provided  by a
           beaded heater wound on the body of the vessel  in such
           a way as to  provide a greater watt density in  the
           upper portions  above the liquid level to prevent con-
           densation in this zone.   Splashing of the  liquid
           against  the  higher temperature walls in this zone  was
           to be prevented by an internal baffle.   Visual
           observations through the range of expected operating
           conditions confirmed that the baffle was operating
           properly.  Tests were also made to determine that
           the unit could  be brought back to room temperature
           and reheated without difficulties.   Maintenance
           of a low flow of nitrogen during cooling prevents  loss
           of liquid sulfur into the gas dispersion tube  which
           would lead to plugging.   It was also found that
           channeling in the solidifying melt occurs  so as to
           maintain free gas passage through the vessel after
           cooling,  thus allowing inert purging during the next
           heating  cycle.

       2.   Leak tests showed that the 2" pipe union cap on the
           sulfur generator would not seal properly under the
           closing  torque  which could be applied to the assembled
           apparatus.   Lapping the sealing surfaces allowed tight
           closure  but  if  another such unit were constructed, the
           cap should probably be flanged and gasketed.


                               77

-------
                                 Figure  26.   H2S  generation kinetics apparatus
                   Vent
00
                 GC Valve
           GC
                                                                               Vent
                                               Heaters
                                               A - Sulfur Generator
                                               B - Gas Preheat
                                               C - Manifold Tracing
                                               D - Tube  Furnace
                                               E - Tube  Furnace
                                               F - Manifold Tracing
                                               G - Condenser Inlet Heat
                             Sulfur
                            Condenser
                            and Filter
r
                                                                     N2     H2
                                                            Low Press.
                                                            Regulator

-------
           Figure 27.   Sulfur vapor generator
Vapor Out
                                       Normal  Sulfur Liquid Level
                                       Internal Baffle
                                       Sintered SS Gas Dispersion Tube
                                     N2 In
                            79

-------
3.  Heating tests were made to establish power loads
    required for heatup and maintenance of design tempera
    tures in sevel independent zones.  It was found that
    good temperature control could be obtained in all
    zones by manual adjustment of auto-transf ormers .

4.  All gas rotameters were calibrated.
5.  Chromatograph system tests using H2$ at concentrations
    in the 1-10% range showed good response and linearity.
                        80

-------
5.2    PRE-INTEGRAL RESULTS

5.2.1  S02 Sorption

       Development of S02 Sorption Rate Model  -

       A large part of the pre-integral S02  sorption  work was
       directed toward measurement of the  S02  kinetics  to be used
       in designing fluid bed adsorbers. A rate model was found
       to mathematically describe the reaction kinetics and was
       useful in developing the reactor design procedure.   S0£
       sorption is shown in Equation  (5):


            S02 + 1/2  02 + H20   Activated    H2S04  (Sorbed)            (5)
                               Carbon


       The reaction rate is defined as the time rate  change of
       acid loading,  dXv/dt,
          Rate  =    L =  f(T, Xv, Xg, X(), y, y, y, y^y*       (10)
       If all variables except acid loading  and  S02  concentration
       are held constant, the rate equation  reduces  to the form
                           =  K g(Xv)  h(ySQ2)                      (11)


       Several forms of this equation are given in Table 9.
       These were evaluated by determining how well each could
       fit  the experimental data.  The model which best fit
       the  data was  then expanded to include the effects of
       the  other important variables.
           *Terms  are  defined in Section 9, Nomenclature.


                              81

-------
      Table  9.   RATE EXPRESSIONS TO APPROXIMATE
                S02 SORPTION DATA
      Designation


      Westvaco


   Modified Westvaco

           1*
     Amundson


      Jost


      AVCO L
      Differential Equation
     Tracer
dXy_
 dt

dXv
 dt

dXy
 dt

dXv
~dT
                       dt
dXv
"IT
___
2 aXv + b
              AxvsYS02
The term Xvs  in Table 9 is the saturation acid  loading,
for which  a value of 0.38 gin. H2S04/gm. carbon  was  obtained
by extrapolation of experimental rate data.

Kinetic data  on S02 sorption was generated in 21  differen-
tial rate  experiments in which the weight gain  of a carbon
sample was monitored as a function of time while  the sample
was exposed continuously to a constant concentration of
S02-  The  apparatus and procedure were described  in
Section 5.1.1 and the experimental conditions are shown  in
Table 10.

Raw data in the form of total sample weight versus  time
was converted to a curve of acid loading versus time.  This
curve was  differentiated by calculating its slope at
numerous points to obtain reaction rate values  at different
acid loadings.   The data in this form is included in
Appendix A-l  for the 21 differential rate experiments.
     *References  listed in Section  8, Bibliography.

                        82

-------
Table 10.  EXPERIMENTAL CONDITIONS  FOR S02 SORPTION
           IN A DIFFERENTIAL RATE APPARATUS
Run
Number
1W, 153
ll*9
150
11*0
162
163
138
160
139
157
lM*
161
I1* 5
11*1
ll*2, l6k
' 1U3
151
152
158
Temperature
"F
150
150
150
200
200
200
200
200
200
200
200
200
200*
200
200
200
300
300
300
Gas Composition
S02
PPM
2500
1500
500
2500
2000
2000
2000
2000
2000
2000
2000
2000
2000
1500
1000
500
2500
1500
500
NO
PPM
150
150
150
150
0
50
150
150
150
150
150
300
150
150
150
150
150
150
150
°2
2
2
2
2
2
2
0.7
2
2
2
3.5
2
•<>
2
2
2
2
2
2
H20
10
10
10
10
10
10
10
5
10
15
10
10
JO
10
10
10
10
•10
10
C02
i
11.3
11.3
11.3
11.3
11.3
11.3
11.3
11.3
11.3
11.3
11.3
11.3
0
11.3
11.3
11.3
11.3
11.3
11.3
Inerts
I
Balance
Balance
Balance
Balance
Balance
Balance
Balance
Balance
Balance
Balance
Balance
Balance
Balance
Balance
Balance
Balance
Balance
Balance
Balance
      and He were used as the inert gas. Experiments using He versus
       and Ng showed a small effect on the rate which was within the
    experimental error.
The initial  rate model evaluation  was based on Runs  139,
140, 141,  143  and 164.  These five runs were conducted
at 200°F with  all conditions held  constant, except for
    concentration,  which was varied from 500 to 2500 ppm.
The reaction  rate vs .  acid loading  data given in Appendix
A-l was applied to each rate equation by the following
procedure.  The Westvaco equation in Table 9, which  eventu-
ally was selected as the best model, is used as an example.
The Westvaco  equation
                   dXy
                   dt
(12)
                        83

-------
can be rearranged to the form
         In  [
  dXy/dt
1 -
In K + mln jr   = Z
         S02
(13)
At a constant S02 concentration, the right side of the  equa-
tion is constant, and the values of  In  [^—_ ^ yx—] or  Z,
calculated from the data, should also be constant if the
model is valid.  A graph of Z versus In 7302 snould give  a
straight line with a slope of m and an intercept of In  K.
Values of Z are tabulated for each run in Appendix A-2.
The falues are reasonably constant for acid loadings
above 0.01 gm H2S04/gm carbon.  The average Z value for
each run, omitting data at acid loadings below the 0.01
level, is given in Table 11 with the standard deviation
     Table 11.  STANDARD DEVIATION FOR WESTVACO
                EQUATION FOR SORPTION DATA AT
                200°F WITH NITRIC OXIDE PRESENT
                FOR ACID LOADING ABOVE 0.01 GM
                ACID/GM CARBON
Run
140
139
141
164
143
S02 Cone. ,
ppm
2500
2000
1500
1000
500
Average z
-6.99
-7.06
-7.18
-7.30
-7.60
Standard
Deviation
0.056
0.069
0.023
0.019
0.038
of the data.  A plot of Z versus In Ygo? ^s snown  in
Figure 28.  After determination of the constants from  the
slope and intercept, the resultant rate equation is:
                   0.0X05  (1 -
                                            (14)
                       84

-------
           Figure 28.   Comparison of Westvaco model  to
                       sorption data at 200°F
-6.0
                                                  Represents Error of Two
                                                   Standard Deviations
                                Ln
      Equation  (14) was  used to calculate reaction rate values
      for comparison with the experimental data.  The experi-
      mental and  calculated sets of data are tabulated in
      Appendix  A-3  and  the average percent difference between
      them for  each run  is given in Table 12.  The percentage
      varies from 1.7 to 6.3% with an overall percent deviation
      of 4.3%,  which is  considered very good agreement.
                              85

-------
        Table  12.   DEVIATION OF WESTVACO EQUATION FROM
                    EXPERIMENTAL S02  SORPTION RATE DATA
                    AT 200°F WITH NITRIC  OXIDE FOR
                    LOADINGS ABOVE 0.01 GM ACID/GM CARBON
S02 Concentration,
ppm
2500
2000
1500
1000
500
Average Per Cent
Deviation
5.0
6.3
1.7
4.8
3.7
       A second  Westvaco rate  equation was  evaluated in a
       similar fashion, but  with poorer results.   The equation was
       arranged  in the form
(4-7)
                     dXy/dt
(l-Xv/Xvs)2
]  =  In K + m In ygo = Z
                                             (15)
       and values  of  Z were  calculated as  before.   Table 13
       shows  the average  Z values  and the  standard deviation
       values, which  are  much  greater than for the first Westvaco
       model.  The second model  clearly does not represent the
       Table 13.  STANDARD DEVIATION  FOR MODIFIED WESTVACO
                  EQUATION FOR  SORPTION  DATA AT 200°F WITH
                  NITRIC OXIDE  FOR ACID  LOADING ABOVE
                  0.01 GM ACID/GM CARBON
S02 Concentration,
ppm
2500
2000
1500
1000
500
Average z
-6.77
-6.82
-6.94
-7.09
-7.39
Standard
Deviation
0.200
0.238
0.176
0.161
0.162
                             86

-------
experimental data as well as the first model does.  The
other models were also'evaluated by similar techniques,
but none represented the data as well as the first
Westvaco model.  The evaluation of these other models is
included in Appendix J-l.

After selecting the Westvaco model as the basic form of
the rate expression and determining the order of the reac-
tion with respect to S02 concentration, further experiments
were conducted in order to expand the equation to include
effects of 02, H20, and NO concentrations and of temperature
The simplifying assumption was made that there is no first
order interaction of variables.  This allowed a stepwise
determination of each variable's effect.

For example, to determine the dependency on 02 concentra-
tion, Equation (14) was modified to yield
          dt
              =  K
                        °'40 v n
                            y02
                                          (16)
                                     vs
which reduces to Equation  (14) if K
Equation  (16) is arranged  in the form

                                     =  0.0105.   If
ln
       dXy/dt
      1 -
       _
] =  In [K
                        0.40
+ n
= Z
(17)
it is seen that K and n can be determined by analyzing  the
experimental data in the same manner described before.
Data from the three experiments used to  find the  02
concentration dependency are given  in Table  14 and the
graph of Z versus In yQ  is shown in Figure  29.
  Table 14.  EXPERIMENTS TO DETERMINE 02 DEPENDENCY
Run
138
139
144
Temp.
°F
200
200
200
S02
ppm
2000
2000
2000
NO
ppm
150
150
150
H20
%
10
10
10
02
%
0.7
2.0
3.7
Average
Z
-7.67
-7.06
-6.67
Standard
Deviation
0,053
0.080
0.072
                       87

-------
   T3
00  ^
00
                  Figure 29.  Effect of C>2 on  SC>2 sorption at  200°F with NO present
         -1
-2
                                                       -3
-4
-5

-------
The values of K and n are  obtained graphically, and the
order of the reaction with respect to oxygen is found to
be 0.63.  The new K value  is  0.1245 so that the rate
equation now becomes
    dX
    = 0.1245 y  °'40 y °'63
dt           S02    02
                                 XVs
                                          for
T  = 200°F
NO = 150 ppm (18)
H20 = 10%
The H£0 dependency is found  in the same manner.   Data from
the pertinent experiments  is given in Table 15,  and the
graph of Z versus  In ys02  ^s shown in Figure 30.
 Table 15.  EXPERIMENTS  TO  DETERMINE H20 DEPENDENCY
Run
160
139
144
Temp.
oF
200
200
200
S02
ppm
2000
2000
2000
NO
ppm
150
150
150
02
%
2
2
2
H20
%
5
10
15
Average
Z
-7.56
-7.06
-6.76
Standard
Deviation
0.164
0.080
0.031
The graph is a straight  line,  and  the order of the reac-
tion with respect to H20 is  0.73.   The new K is 0.667 and
the rate equation becomes
                      0.63    0.73
               S02
                                Xvs
                                       -)   for
[~T •
[NO =
= 200°F
  150 ppm
                                                     (19)
The NO dependency was treated  in  a similar manner, but in
this case the results showed a zero order dependency at
NO concentrations above  100 ppm.   The data is presented
in Table 16 and Figure 31.
        Table 16.  EXPERIMENTS  TO DETERMINE
                   NITRIC OXIDE DEPENDENCY
Run
162
163
139
161
Temp.
°F
200
200
200
200
S02
ppm
2000
2000
2000
2000
02
%
2
2
2
2
H20
%
10
10
10
10
NO
ppm
0
50
150
300
Average
Z
-5.97
-6.88
-r.oe
-7.07
Standard
Deviation
0.214
0.048
0.080
0.086
                        89

-------
         Figure  30.   Effect of H20 concentration  on S02 sorption at  200°F with NO present
VO
    X
    •o
         -1.5
-2.0
                                                       JL
 -2.5
Ln
-3.0
-3.5

-------
                      Figure  31.  Effect of NO concentration on  S0£ sorption at 200°F
          0.25
I-1    >
                                    100
          200
yNf) x.!0°, Vol. Fraction
300
400

-------
Because flue gas  NO  concentrations are typically higher
than 100 ppm, modification of the rate equation to extend
the valid range below 100  ppm was considered unnecessary,
so that the rate  equation  now is  expressed properly as
                    °'40   °'63   °'73 /i  XVx  f   r  =200°F
                 ys02    y02    yH20    (1 -j-g  f°rrc> > 100 pp»
Temperature Dependence  -

In determining  the  temperature  dependence of the reaction ,
the rate constant was assumed to  be  the only temperature
dependent  term, and the Arrhenius equation was assumed to
give a satisfactory representation of the temperature
effect.  The rate constant  K is defined by the Arrhenius
equation as
                    K =  k0 e"E/RT                            (21)
          where  ko  =   frequency factor
                 E   =   activation energy
                 RT  =   product  of gas  law constant and
                           temperature.
By taking the  logarithm  of  both  sides,  one obtains
                 InK  =  In k0 -E/RT                          (22)
from which it is seen  that  a  graph of In k versus 1/T
should yield a straight  line  with a slope of -E/RT and an
intercept of In ko.

The value of the rate  constant  was already specified at
200°F by rate equation (20).  The rate donstant was also deter-
mined at 150 and 300°F from the experiments listed in
Table 17.
                           92

-------
      Table 17.  EXPERIMENTS TO DETERMINE THE
                 EFFECT OF TEMPERATURE ON S02
                 SORPTION
Run
148-G
153-G
149-G
150-G
140-G
139-G
141-G
142-G
164-G
143-G
151-G
152-G
158-G
Temperature
°F
150
150
150
150
200
200
200
200
200
200
300
300
300
1
Gas Composition
S02
ppm
2500
2500
1500
500
2500
2000
1500
1000
1000
500
2500
1500
500

NO
ppm
150
150
150
150
150
150
150
150
150
150
150
150
150

02
o/
JO
2
2
2
2
2
2
2
2
2
2
2
2
2

HgO
%
10
10
10
10
10
10
10
10
10
10
10
10
10

Average Z
-6.01
-6.25
-6.22
-6.76
-6.99
-7.06
-7.18
-7.39
-7.30
-7.60
-7.49
-8.08
-8.76

Standard
Deviation
0.688
0.569
0.659
0.469
0.056
0.080
0.025
0.086
0.019
0.038
0.149
0.211
0.323

Values of Z and the standard deviation of the data are
also included in the table.   To calculate the rate con-
stants at 150 and 300°F, the Z values were plotted versus
In yso? as seen i-n Figure 32.   Straight lines were drawn
through the points subject to the constraint that the order
of the reaction with respect to S02 was not a function of
temperature so that the slope was constant.  Once the lines
were drawn, the rate constants could be found by inserting
the Z and In YS02 coordinates of any point on a line into
equation
                                    In
(23)
and then solving for k.  Table 18 gives the rate con-
stants at three different temperatures.  Figure 33 is
 Table 18.  RATE CONSTANTS FOR THE WESTVACO MODEL
Temperature ,
op
150
200
300
Rate Constant,
gm acid/gms carbon/min.
1.476
0.667
0.244
                         93

-------
Figure 32.  Effect  of  temperature on the Westvaco model with constant order
            of reaction  for  S02
                                                                                  -9

-------
    Figure  33.   Rate  constant as a  function  of temperature
                 for S02 sorption
2.0
1.0


0..8


0.6




0.4
0.2
0.1
  1.1
1.2
1.3
1.4
                                1(103).
 1.5

'R"
                            00-1
1.6
1.7
1.8
                              95

-------
the graph of  In K versus  1/T from which the  frequency
factor and  the activation energy  were  obtained.   The
results are
                  k0 =  1.59xlO~4                         (24)
             E  =  -1.079xl04 BTU/lb. mole                   (25)
or
                    E/R  = -5520.                          (26)
The final form of the Westvaco rate equation  is then


                                                         (27)
    .           e      ^0.40 ^0.63 ^0.73
One additional qualification is placed on Equation  (27) ;
it is strictly valid only for acid loadings above
0.01 gm H2S04/gm carbon, because rate data at  lower acid
loadings was excluded from the calculations.   Equation  (27)
predicts a lower reaction rate than measured experimentally
at loadings below  0.01  gm H2S04/gm C.

Success of Westvaco Equation in Fitting
Differential Rate  Data

Analysis of the Westvaco rate equation showed  that  it
represents the experimental data very well at  200°F but
less satisfactorily at  150 and 300°F.  The results  of cal-
culations comparing the Westvaco equation predictions with
the differential rate data are tabulated in Appendix A-3.
Table 19   is a summary of these results showing  the
average percent difference between the predicted  and
experimental values.   The overall average difference for
the twelve 200°F runs is only 7%, but the deviation is  3770
at 150°F and 23% at 300°F.
ro
                          96

-------
    Table 19.   DEVIATION OF WESTVACO MODEL FROM
                DIFFERENTIAL RATE DATA FOR ACID
                LOADINGS ABOVE 0.01 GM ACID/
                GM CARBON
Run
138
139
140
141
142
143
144
145
157
160
161
164
Temperature
OF
200
ii
ti
ii
ii
ii
it
ti
ii
ii
ii
ti
AVERAGE 200
148
149
150
153
AVER
151
152
158
AVER
150
ii
ii •
ii
AGE 150
300
ii
it
AGE 300
Average 70
Deviation
4.5
"1.7
8.1
5.6
1.9
3.9
6.6
5.0
4.8
13.9
8.0
12.3
6.9
47.7
40.0
20.6
39.2
36.9
36.8
15.9
17^4
23.4
Another comparison  between predicted and experimental
values is presented in Figure 34,  which is a plot of  the
average Z value for each differential rate experiment
versus the term
    5520/T + 0.40 In ys02 + 0.63 In y02 +0.73 In yH20
                                             (28)
This term is the natural  logarithm of the variable part
of the right side of  the  Westvaco equation in the form
    dXv/dt
   l-Xv/0.38
P-40   0.63
-  1.59(10-4) e5520/T y   -0 ^
.73
          (29)
                           97

-------
  Figure 34.
-4.5
 -5.0-
 -5.5-
 -6.0-
 -6.5-
,-7.0-
 -7.5-
-8.0-
      Comparison of the Westvaco Model A  to experimental
      S0£ sorption  on activated carbon in differential
      rate apparatus
•  150°F


•  200°F


A  300°F
    Represents
    Error or
    Two Standard
    Deviations
-8.5-
                              Westvaco Equation, Model A;

                                        5520  •
                        ^ » (1.59)(10-4)e
                                                    NO Concentration i 100 PPM
                                            dXdt
-9.5-
-9.5
              0.4
               0.8
1.2

 98
1.6
2.0
2.4
2.8

-------
The straight line  in Figure   34  represents the Westvaco
equation as the natural  log  of both sides  of the above
expression.  The experimental Z  values  are shown with an
error band representing  twice the  standard deviation of
the data.  The Z values  and  standard deviations are in
Table 17.

The standard deviation provides  a  useful measure of the
data's probable accuracy.  There is a 95%  probability
that the true Z value lies within  an error band'that is
two standard deviations  wide, so that a small standard
deviation indicates high accuracy  of the data.   It  is
seen from Table 17  and  Figure 34   that the standard
deviation of the data is much smaller at 200°F than at
150 or 300°F.  As  a percentage of  the average Z value,
the standard deviation averages  0.7% at 200°F,  2.8% at
300°F, and 9.4% at 150°F, so that  statistically,  the
200°F data is excellent  but  the  150 and 300°F data  is
not as good.

Sorbet Design Study - Effects on Size -

Based on the Westvaco S02 sorption rate expression  and two
important reactor model  assumptions,  a  procedure was
developed for calculating the required  size of a multistage
fluid bed S02 sorption reactor.  Using  this procedure a
sorber design study was  made encompassing  the effects of
seven important variables.   The  complete derivation of the
design equations is in Appendix  C.

The two important reactor model  assumptions concerned the
flow characteristics of  the  gas  and carbon phases within
the reactor.  It was assumed that  the carbon phase  was well-
mixed on each stage, so  that the carbon beds represented a
series of backmix reactors.   Plug  flow  was assumed  for the
gas phase.

The design equations used in the reactor size calculation
procedure are:
                                                         (30)
      n.6    ...  .0.6   . 	,,  ^  0^8'  X02    1H20
' (YS02)j'6  - 3.687(10-4)
                       "    XVi   0.63   0.73   2
                       (l-TT^o") Y09    YHOO   hD
                                                          5520/T
                                               qT
      XVj+1  = XVj - (3^) (g) l(Yqno).  -  (YSn,)<4.il            <31)
                          99

-------
The first equation (30) was the acid loading and S02 concen-
tration on a given stage to calculate the S02 concentration
on the next stage.  The second equation (31) is a material
balance relationship which is used to calculate the acid
loading on the new stage from the other variables.  Using
these equations, plate-to-plate calculations are made
beginning with the bottom stage of the reactor and pro-
ceeding until the end conditions at the top of the reactor
are reached.  This determines the required number of stages
and hence the reactor size.

A computer program was written to facilitate the calcula-
tional procedure.  A listing of the program is an Appendix
E.  The effects of the following parameters on reactor
size were studied:

     1)  Inlet S02 concentration (2,000 ppm)

     2)  Inlet 02 concentration (0.046 mole fraction)

     3)  Inlet H20 concentration (0.13 mole fraction)

     4)  Outlet acid loading on carbon (0.184 Ib. acid/
         Ib. carbon)

     5)  Temperature (200°F)

     6)  Linear Gas Velocity (4 ft./sec.)

     7)  Carbon bed depth/stage (9 inches).

Each parameter was varied separately while holding all
others constant at the base conditions given in parentheses
in the preceding list.   The results of the design study are
in Appendix A-9.


Attempts To Further Improve Model
by Multiple Regression Analyses

Multiple regression analyses were carried out in an attempt
to improve the Westvaco model's representation of the  data
at 150 and 300°F.  Based on the following form of the  rate
equation,
         dt
             =  k e
                  -E/RT   m
                       100

-------
The multiple regression analyses were used to determine
the best values of the constants k, E/R, m, p, and q for
the two cases where the order of'reaction relative to  the
acid concentration in the carbon phase, constant A, was
allowed to vary in one case and equal to 1.0 in the other
case.  Both cases were processed by computer using a
program written by IBM.  A listing of the program is in
Appendix E-l.  The Westvaco equation, (27),  was designated
Model A, and the multiple regression variations were desig-
nated Models B and C.  For Model B with A = 1.0 the multi-
ple regression analysis yielded the following expression:
      -  3.32(10-*) e»"/T y°-53 y°-62 ,°'73 U->      (33)

Allowing the constant A  to  seek  its  statistical value
resulted in the following expression for Model C:
     .  3.42(10-4) e5732/T ^0.53 ^0.62 ^0.73 ^_J^I.B3    (34)

                                              U * jo
Calculations were carried out on  the  computer  to compare
Models B and C with the differential  rate data and the
results are included in Appendix  A-7.  Table  20  is a
summary of these results showing  the  average percent
difference between the predicted  and  experimental values.

Figures 34, 35, and 36  provide a graphical  comparison
of the three models.-  The results in  Table   20  show  that
Model B fits the data slightly better than Model A at
300°F but not as well at 150°F, while Model  C  provides  a
significant improvement at both 150 and  300°F.  At 200°F
Models B and C are both worse than Model A,  as expected.
Clearly, Model B does not offer any real improvement  at
150°F to warrant its substitution for Model  A  near this
temperature and perhaps over the  150  to  175°F  range,  but
Model A remains the best choice over  the remainder of the
temperature-range from 175 to 300°F,  because of  its
superior performance at 200°F.
                        101

-------
Table 20.  DEVIATION OF MULTIPLE REGRESSION MODELS
           FROM DIFFERENTIAL RATE DATA FOR ACID
           LOADINGS ABOVE 0.01 GM ACID/GM CARBON
Run
138
139
140
141
142
143
144
145
157
160
163
164
AVER
148
149
150
153
' AVER/
151
152
158
AVER;
Temperature
'oF
200
it
ii
ii
it
ii
n
ii
n
n
n
n
\GE 200
150
n
u
n
\GE 150
300
u
n
\GE 300
Averayo Percent: Dew 1. 'it ion
Model A
4.5
7.7
8.1
5.6
1.9
3.9
6.6
5.0'
4.8
13.9
8.0
12.3
6.9
47.7
40.0
20.6
39.2
36.9
36.8
15.9
17.4
23.4
Model H | Model C
13.6
17.6
23.1
16.3
8.1
4.6
13.1
8.0
18.1
24.1
20.5 :
18.4
15.5
51.3
41.0
20.7
47.6
40.2
25.9
16.1
19 . 4
20.5
22.8
25.9
30.7 •
19.5
13.9
11.1
24.5
14.6
21.3
26.8
24.5
26.1
21.8
33.5
25.8
13.2
22.1
23 . 6
27.6
8.2
5.0
13.6
                       102

-------
 Figure 35
-4.5
   Comparison  of the Westvaco  Model B to experimental
   S02 sorption on  activated carbon in differential
   rate  apparatus
-5.0-
-5.5-
-6.0-
-6.5-
-7.0-
-7.5-
-8.0-
-8.5-
-9.0-
-9.5
150°F


200°F


300° F
Represents
Error of .
Two Standard
Deviations
.0 I
             -0.4
                          dX
                     Westvaco Equation, Model B:

                              5416

                 = (3.32)(10'4)e "~ (1
                 Xv v .,  0.48 .. 0.62 .. 0.73
                   M
                   H


                   NO Concentration > 100 PPM
                                               dXv/dt
            0
 I
0.4
0.8
 i
1.2
 i
1.6
                                                                           2.f
                                   103

-------
 Figure  36.
  Comparison of  the  Westvaco Model  C to  experimental
  S02  sorption on activated carbon  in differential
  rate apparatus	
 -4.5-
 -5.0 -
 -5.5 -
 -6.0 -
 -6.5 -
-7.0 -
-7.5  -
-8.0  -
-8.5  -
 -9.0-
150°F


200°F


300°F
              Represents
              Error of
              Two Standard
              Deviations
                               (3.42) (10-
                       Westvaco Equation, Model C:

                            5732
                                        1.83 „  0.53 „ 0.62 „  0.73
                                                       NO Concentration i 100 PPM
                                     Y = Ln
                                               dXy/dt
                                                   "
                                                     .83
                           5732         0.53   0.62   0.73
                       Z = -j- * Ln [YS02    Y02    YH20   ]
           I
         -0.4
                 0.4
 i
0.8
  Z
1.2
 i
1.6
2.0
                                   104

-------
In the Westvaco equation (Model A) the rate is proportional
to (l-Xv/0.38) or equivalently, to (0.38-XV) where
0 < Xv < 0.38.  This model is valid if the plot of dXv/dt
versus Xv is a straight line.  Examination of the rate
curves in Figure  28  reveals that dXv/dt versus Xv
approaches a straight line at both 200 and 300°F, but not
at 150°F.  Model A should, therefore, give satisfactory
results at the higher temperatures, but a different model
may necessarily be needed at 150°F.  In Model C the rate
is proportional to (1 - Xv/0.38)1-83 which provides a better
fit at 150°F.

From Figure  37  it is seen that the 200 and 300°F curves
are not linear until the acid loading is above 1 gin/100
gms carbon.  This explains why the data at acid loadings
less than 1.0 gm/100 gms carbon was excluded from the rate
model computations.  It also indicates that application of
the model at low acid loadings would cause a significant
underestimation of the sorption rate.

Fluid Bed S02 Sorption Experiments -

This section covers the pre-integral fluid bed sorption
work that was done in the 6" and 18" diameter S02 sorbers.
Objectives included demonstration of satisfactory mechani-
cal operation and process performance, gathering of addi-
tional information on reaction rate and process character-
istics, and completion of necessary developmental work in
preparation for integral pilot plant operation.

6" -Diameter Sorber -

Operation of a 6" diameter, eight stage fluid bed sorber
provided the first test of the S02 sorption step under
actual flue gas conditions of a slipstream from a 50 MW
oil fired boiler.  Successful demonstration of the 6"
unit was an important .achievement in development of the
process.  Operation of the 6" sorber provided valuable
reaction rate data for comparison with the Westvaco rate
model.

The 6" sorber runs offered the first opportunities to test
the rate model in predicting the performance of an actual
flue gas system.  The data from six runs was processed to
give average values of S02 concentration, acid loading and
reaction rate on each stage of the reactor.  These average
values are given with the run conditions in Appendix A-5.
The yso2 an<* xv values were then used in the Westvaco rate
equation to calculate predicted rates for comparison with
the experimental rates.   The results of the comparison are
presented in Table  21.    The predicted rates average
about 147o below the experimental rates.


                        105

-------
               Figure  37
            Differential  S02 sorption rate versus H2S04  loading for  an
            S02 concentration of  2500 ppm at  150, 200, and  300°F
                   4.0
o
cr>
                e
                o
                60
                o
                CO
                CM
                EC
                C
                O
o

cs


<4-l
O

V
                   3.0
   2.0
                   1.0'
                                      Sulfuric Acid Loading, gms. H2S04/100 gms.  C

-------
Table  21.   COMPARISON OF  RATES FROM 6"  DIAMETER SORBER
              TO RATES  CALCULATED FROM THE WESTVACO MODEL
Run*

SA-21






SA-23






SA-26






SA-22


„



SA-25






SA-24






Stage

1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3,
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
Temperature,
°F

200






200






200






200






200






200






gms Acid/gm Carbon

0.0213
0.0426
0.0662
0.0920
0.1178
0.1480
0.1817
0.0087
0.0260
0.4777
0.0703
0.0954
0.1241
0.1562
0.0170
0.0377
0.0583
0.0817
0.1068
0.1346
0.1660
0.0308
0.0607
0.0888
0.1214
0.1531
0.1875
0.2274
0.0140
0.0307
0.0464
0.0657
0.0832
0.1043
0.1279
0.0241
0.0482
0.0738
0.1002
0.1282
0.1530
0.1981
S02
Concentration,
ppm
251
425
607
808
1,018
1,247
1,507
46
183
388
621
872
1,155
1,475
132
324
534
758
1 ,005 •
1,274
1,576
1,306
1,612
1,904
2,210
2,535
2,868
3,242
484
644
813
996
1,187
1,388
1,621
1,219
1,502
1,795
2,101
2,420
2,731
3,142
Sorption Rate x 10J,
#Acid/#Carbon-Min.
6 Inch**
0.76
0.76
0.84
0.92
, 0.92
1.08
1.20
0.35
0.71
0.89
0.92
1.03
1.17
1.31
0.72
0.88
0.88
0.99
1.07
1.18
1.33
1.24
1.21
1.13
1.32
1.28
1.39
1.61
0.56
0.67
0.63
0.77
0.70
0.84
0.95
0.96
0.96
1.02
1.05
1.11
0.99
1.79
Model ***
0.72
0.84
0.90
0.93
0.93
0.89
0.82
0.38
0.63
0.80
0.90
0.95
0.95
0.92
0.58
0.78
0.89
0.95
0.98
0.96
0.91
1.36
1.36
1.32
1.25
1.16
1.03
0.86
0.52
0.56
0.59
0.60
0.61
0.60
0.59
0.79
0.80
0.80
0.78
0.74
0.70
0.59
   *Inlet gas compositions for the runs are given in Appendix B.
  **Rate data from 6" diameter sorber.
  ***Rate data calculated from Westvaco Model, Equation (27).
                               107

-------
A graphical comparison of the rate model to the fluid bed
rate data is given in Figure 38.    Average experimental
reaction rates are plotted versus S02 concentration, with
the rate equation superimposed  for comparison.  The reac-
tion rates in Figure 38   were obtained from the actual
rates by dividing by all variable terms in the equation
except yso?-  This mathematical comparison assumes that
all differences then arise with the S02 concentration
term in the model, which is not necessarily true but if
large discrepancies had occurred, then additional analysis
would have been required.  The graph shows a reasonably
good fit of the data by the model.

The rate model was also tested by using it in a design
procedure to calculate the theoretical number of stages
required for'each run.  The detailed derivation of the
design procedure is given in Appendix C-l.  The results
of this comparison are shown in Table 22.  The average
predicted number of stages is 9.9 compared to the actual
8, which is an over-estimation of 2570.  The rate model,
therefore, can be expected to yield conservative esti-
mates of reactor size.
     Table 22.  COMPARISON OF PREDICTED NUMBER OF
                STAGES TO ACTUAL NUMBER FOR 6"
                SORBER RUNS
Run

SA-21
SA-23
SA-26
SA-22
SA-24
SA-25
Temperature ,
°F

200
200
200
200
250
250
Inlet S02
Cone . ,
ppm
2000
2000
2000
4000
2000
4000
Actual
No. of
Stages
8
8
8
8
8
8
AVERAGE : 8
Predicted
No. of Stages
Model A
9.2
8.5
10.1
9.3
11.6
10.4
9.9
\
                         108

-------
   Figure 38
6"  sorber data - plot of corrected  sorber  rate
using stagewise Westvaco Model A vs.  S02
concentration showing curve  predicted from
differential bed studies
  100+
 : .10-
 CM
 X
U-l


LO
 •
r~i

 O
M
e>
f>
 ai
                       Prediction by Stagewise
                         Westvaco Model A
                                 O  200°F, 2000 PPM

                                 O  250°F, 2000 PPM

                                 •  200°F, 4000 PPM S02

                                 •  250°F, 4000 PPM S02
    100
                       1,000

         S02 Concentration,  Volume Fraction
100,000
                                   109

-------
 18" Diameter  Sorber  -

 Operation of  an  18"  diameter  sorber was the next step in
 development and  provided additional process information.
 In the operation of  an S02  sorber, the flue gas typically
 enters at300°Fto350°F.   Before  the gas was cooled, the
 863 at a concentration of about  50 ppm was removed to pre-
 vent  corrosion which would  result if the 803 were allowed
 to condense.  803 removal is  accomplished in the first
 stage of the  reactor, and then the gas is cooled to
 increase the  rate of 802 removal in the remaining stages.

 Based on prior development  work of using water spray cool-
 ing in the production of carbon  in fluidized bed reactors,
 a process change was made in  802 sorption, namely, the
 substitution  of  direct flue gas  cooling by means of water
 injection in  place of indirect heat exchange. .In the 6"
 diameter sorber  and  in the  initial design of the 18"
 diameter unit, the flue gas after 803 removal in the
 bottom stage  of  the  reactor was  cooled by indirect heat
 exchange before  carbon/flue gas  contact in the second and
 subsequent stages.   The heat  exchanger presented tempera-
 variation problems and was  also an expensive item in the
 capital investment estimates  of  the Westvaco Process.  These
 reasons provide^ the impetus  for development of a direct
 cooling method.  The success  of  the new cooling method,
 water spray injection, was  a  significant break-
 through which simplified operation and substantially
 reduced capital  cost estimates.  The initial direct cool-
 ing tests, Table  23  were made in a one-stage, 18"
 diameter unit.   The  gas temperature was lowered from 300
 to 150°F with no apparent operating problems.

 After demonstrating  satisfactory water spray operation
 in the one stage unit, modifications to the existing 5
 stage, 18" diameter  fluid bed sorber at No. 6 power boiler
 were  made to  replace the indirect heat exchanger with a
 direct water  spray cooling  system.  Prior to operation
 with  flue gas, test  runs SC-7, -8 and -11 were made to
 check out the equipment using air and an air-steam mixture
 which simulated  the  flue gas  moisture conditions.  No
.operating problems were encountered in these tests with
 cooling down  to  150°F.

 Direct water  spray cooling  of actual flue gas was first
 demonstrated  for use in the Westvaco Process in Runs  '
 SC-14, -16, -16A and -16B.  These runs lasted 6 to 10
 hours each with  no difficulties  in equipment operation.
 In an extended demonstration  run, SA-32, mechanical
 problems were encountered that were unrelated to  the
water spray system,  but despite  the problems a total
                          110

-------
   Table  23.  WATER SPRAY COOLING TESTS MADE IN PILOT FLUID BED REACTORS
                WITH  SIMULATED  AND  ACTUAL FLUE GAS
RUN
NUMBER
SC-1
SC-2
SC-3
SC-4
SC-5
SC-7
SC-8
SC-11
SC-14
SC-15
SA-16A
SC-16B
SA-32
UNIT USED
18" Dia.
Fluid Bed
One-Stage
Unit
18" Dia.
Fluid Bed
S02 Sorber
NUMBER
OF
STAGES
1
1
1
1
1
4
4
4
4
4
4
4
4
TOTAL
CARBON
BED
HEIGHT
(SETTLED)
INCHES
6
6
6
7.5
9
15.5**
15.5**
15.5**
15.5**
15.5**
15.5**
15.5**
15.5**
INLET
GAS
TEMP.
»F
300
it
II
380
330
327
310
300
310
310
315
	 5TBGT 	
TEMPERATURE
°F
11
200
178
149
152
149
350
317
310
295
280
290
290
295
n


...


242
140
190
175
140
150
150
175
#3


---


250
150
183
180
155
160
160
180
#4


...


240
150
175
175
170
160
160
180
LINEAR
GAS
VELOC.
FT/SEC
2.9
3.1
3.1
3.1
3.1
3.1
2.6
2.9
3.0*
3.1*
3.1*
3.1*
3.1*
H20
SPRAY
RATE
I/HR.
23.4
31.2
43.0
33.2
27.4
10 .
37.2
44.7
29.8
45
40.2
40.2
34
TOTAL
GAS
FLOW
RAT£,
CFH ?
70°F
14,070
14,500
16,500
16,500
16,500
14,830
14,430
1 5,2'80
15,250
16,230
16,200
16,200
16,767
INLET
02
ANAL,
VOL.
%
10
It
II
Air
II
H
3.1
3.1
3.1
3.1
3.1

INLET
H20
ANAI
VOL.
%
10
II
it
~4
-4
-12
12.7
12.7
12.7
12.7
12.9

SULFUR DIOXIDE
ANALYSIS, PPM
INLET









995
880
1,125
1,100
2,200

STAGE
1









925
1,070
2,088

STAGE
2









540
560
1,511

STAGE
3









360
460
1,008

OUTLET









70
170
160
55
522
CARBON
PRECURSOR
SKID NO.
AND
CHAS. NO.
WV-W
II
II
II
It
96257, C-71-98
II
II
H
II
96324, C-71-23
96283, C-71 -100
CARBON ACID
10ADING
LBS. SOz/
100 LBS. C
INLET
...
0
0
0
0
4.1
8.7
0
0
OUTLET
...
0
0
0
10.3
13.7
20.9
14.2
11.9
CARBON
DATT
TC/HR.

0
0
0
0
0
23
23
23
23.2
19.2
20.2
19.9
32.9
 *Based on average temperature of Stages 2, 3 and 4.
"Excluding carbon on bottom stage at about 300°F (about 3 inches carbon).

-------
operating time of 29 hours was achieved, and the water
spray cooling system subsequently was declared a success.

The advantages of using water spray cooling instead of
indirect heat exchange are summarized as follows:

     1)  A savings on the order of 2570 in the capital
         investment estimated for the Westvaco Process
                         f

     2)  Temperatures as low as 150°F can be achieved
         which would be difficult using a heat exchanger
         because of moisture condensation at cooling
         surfaces or very large surface areas for heat
         transfer.

     3)  The small amount of moisture from the water
         spray increases the rate of S02 removal.

     4)  Better control of the column temperature has
         been realized with water spray cooling.

Additional runs were made in the 18" diameter sorber in
order to produce acid loaded carbon needed for the sulfur
generation and sulfur stripping experiments.  These runs
contributed valuable information about the unit's operat-
ing characteristics and they also led to early recognition
of problem areas, thereby providing a basis for necessary
corrective measures and improvements to increase opera-
tional reliability.

The summary for a typical sorption run, SA-32, is given
in Figure  39.    Temperature and S02 concentration pro-
files through the column are shown.  Average column
temperature for Stages 2 to 5 was 168°F.  The S02 removal
was 94%, and the material balance agreed within 67o.

Effect of Fly Ash on Carbon S02 Activity -

Application of the Westvaco S02 Process in the flue gas
desulfurization area would involve both coal and oil
fired boilers.   One difference between the two is the
significantly higher fly ash concentration in the flue
gas from coal fired boilers.  Pilot plant S02 sorption
work has been conducted using flue gas from an oil fired
boiler only, so that the results are not necessarily
applicable under coal fired boiler conditions.

Of greatest concern is the possibility that prolonged
exposure of activated carbon to high fly ash concentra-
tions may reduce the carbon's S02 activity.  This was
                          112

-------
Figure  39.   Summary for flue  gas  run  (run SA-34);
               18" diameter S02  sorber; water  sprays
               to control  temperature
38.1 Ibs. C/hr. 	

165°F

174°F

171°F
H20 Pxate,

(AUTO Control) 162op
263°F
290°F
Flue Gas + S02
2,050 PPM of S02
14,800 CFH 9 70° F
I

4'.^:+&-:ili:&tl:
*
^^(fS^^ri

^Tfj:-!-':^' '••• '•• \ ' Jv?-' '•.

_ .*•
J^-iv^^^i^
m^m

"
i
	 j3^

Stage No. 5

Stage No. 4

Stage No 3


Stage No. 2
Stage No. 1



                                         ^•Outlet S02  Cone.  = 125 PPM
                                                    Cone.
                                                         =   725 PPM
                                                    -^  = 1,125 PPM
                                                    Ccr.c.
        S02
       Cone.
                                                    Cone.
                                                           1,550 PPM
                                                           1,950 PPM
                         0.135 Ib. S02/lb. C
                          37.3 Ibs. C/hr.
AVG. TEMP.:     168°F
% S02 REMOVAL:  94

MAT'L. BALANCE
   S02 IN:      5.0 Ibs./hr.
   S02 OUT:     5.3 Ibs./hr.
                            113

-------
       investigated in a bench scale experiment in which recycled
       carbon was exposed to fly ash laden air for a period of
       5 days.   The ash content of the carbon was measured each
       day and the S02 activity was measured before and after.
       The ash content did not increase above the initial level
       corresponding to the inherent ash content of the carbon,
       which is typically in the range of 4-1/2 to 5% by weight.
       S02 activity measurements showed only a 6% drop in
       activity which is within experimental error and, therefore,
       not really significant.  This result is encouraging and
       indicates that the carbon should perform satisfactorily
       under coal fired boiler conditions,   Additional details
       of the fly ash exposure experiment are included in
       Appendix J-2.

5.2.2  Sulfuric Acid Conversion to Sulfur

       Sulfuric acid conversion to sulfur is the second step in
       the Westvaco Process,  and involves the reaction of sorbed
       H2S04 with H2S to form sorbed elemental sulfur and water
       vapor.  The overall reaction is:


               H2S04 + 3 H2S   Activated ^  4 S + 4 H20 +              ( 6 )
                             Carbon


       The main objective of sulfur generation studies were:
       1)  to obtain enough information to properly design and
       construct an acid conversion reactor and optimize its
       operating conditions,  and 2) to demonstrate satisfactory
       operation of a pilot acid converter incorporated into
       the integrated S02 pilot plant.   An important requirement
       in satisfactory demonstration of a pilot reactor was high
       degree of conversion of both reactants.

       The reaction was studied extensively in various bench and
       pilot scale equipment, including:

            1)   Fixed bed reactors - 1" and 1.6" diameters

            2)   One stage,  4" diameter fluid bed batch reactor

            .3)   8 stage,  4" diameter fluid bed glass column
                counter current reactor

            4)   8 stage,  4" diameter fluid bed reactor (eventu-
                ally used as sulfur stripper/H2S generator in
                integral pilot plant)

            5)   Moving bed reactors - 1.5" and 8" diameters
                (8"0 eventually used as sulfur generator in
                integral pilot plant).


                               114

-------
A rate expression was found to represent kinetic data
measured in a one stage fluid bed reactor.  The kinetic
data^was incorporated into a reactor design procedure.
The important variables in the reaction rate are
temperature, acid concentration on carbon, and the gas
concentrations of H2S and H20.

Experimental studies indicated that the required high
conversions of both H2S and H2S04 were not possible at the
space velocities obtainable in a 6"0 fluid bed reactor
that initially was considered for use in the integrated
pilot plant.  Subsequent tests in a 1.5" diameter bench
scale moving bed reactor indicated that a moving bed
was better suited for integral operation of the pilot
plant.  A pilot scale 8" diameter moving bed unit was then
designed and tested; and satisfactory performance was
demonstrated.

It should be recognized that the use of a moving bed
sulfur generator in the integrated pilot plant did not
alter plans to specify fluid bed reactors in larger scale
applications.  The rate expression was developed from fluid
bed data obtained in a set of batch differential experi-
ments, and to the extent that the model accurately repre-
sents the reaction, it is entirely valid for design of
either reactor type.

Fixed Bed Acid Conversion Experiments -

Fixed bed studies were carried out in 1" and 1.6" diameter
reactors.  The more important results are presented here,
and the complete results are given in Appendix J-3.  The
purpose of experiments in the 1" fixed bed was to determine
the effect of linear gas velocity on the reaction rate.
Runs were made at linear velocities from 0.03 to 0.14
ft./sec., as shown in Figure 40.  It was found that the
reaction rate increased with increasing linear velocity
up to a velocity of about 0.12 ft./sec,  Further
increases above 0.12 ft./sec. did not raise the reaction
rate.   This result indicates that external diffusion does
not limit the reaction at linear velocities above 0.12
ft./sec.  The data point shown above the line on Figure
40 represents a temperature of 395°F, about 75°F above
the other data points.

The objective of experiments in the 1.6" diameter fixed bed
was to study the effects of space velocity on reactor
performance.   Runs were made at constant reactor volume at
space velocities from 100 to 1,000 hr". ~i.  The inlet gas
composition was held constant throughout each run, and
the outlet H2S concentration was monitored continuously to
determine the time at which H2S breakthrough occurred.
The results showed almost immediate H2S breakthrough at


                          115

-------
       Figure  40.  Effect of  linear  gas velocity on  rate of
                    sulfuric acid decomposition
    -1.2
  I
  o

  §>
X
•o
 I
 •>

 o
 in
 O
 o



 TJ
 to
 a:
               0.02       0.04      0.06       0.08       0.10


                              Linear Gas Velocity, ft./sec.
0.12
0.14
                                   116

-------
space velocities above 300 hr.'1, but at 100 hr."1 break-
through did not occur until 130 minutes.  Overshadowing
these results, however, were the implications of the
temperature rise that occurred in the bed due to the
exothermic (-65 kcal/mole acid) heat of reaction.  The
same problem occurred to a lesser degree in the 1" fixed
bed runs.  A temperature rise of 135°F was recorded in 3
out of 8 runs in the 1.6" diameter bed.  The most import-
ant result of these experiments, therefore, was to
clearly demonstrate the unsuitability of a fixed bed
reactor in kinetic studies of the sulfur generation
reaction.  The temperature control which is necessary"to
obtain useful data could not be achieved in a fixed bed
unit.

Fluid Bed Rate Studies -

The effects of temperature and H2S concentration on the
rate of reaction were studied in an 8 stage, 4" diameter
fluidized bed reactor constructed from flanged sections of
glass pipe.  Each stage had 4 inch overflow weirs for a
typical carbon bed depth of 2.5 inches/stage, so that the
total settled bed depth was typically 20 inches of carbon.
Experiments were run at temperatures of 250, 275, 300,  325
and 350°F, and at inlet H2S concentrations from 127o to
42%.  Average steady state conditions for each run are
shown in Table 24.

The overall rate of sulfuric acid decomposition was calcu-
lated for each run based on both gas and carbon analyses
for sulfur compounds.  A comparison of the results given
in Table  25  shows that reaction rates calculated from
carbon analyses are about 2070 lower than those determined
from gas analyses.  In the following discussion of results,
the rates based on carbon analyses are used because they
are considered more reliable and are more conservative.

The rate of acid decomposition increases steadily with
increasing temperature, rising from 0.007 Ib. acid/lb.
C-min. at 250°F to 0.20 at 325°F.  The rate levels out
around 325°F and is about the same at 350°F as at 325°F.
Figure  41  shows a plot of rate versus temperature.

An important temperature dependent effect begins to appear
at 300°F which effectively places an upper limit of 325 to
350°F on the practical operating temperature range.  This
refers to the evolution of reactant acid in the form of
S02 in the outlet gas.  Although the exact mechanism is
uncertain, the net result is that acid converts to S02 at
the expense of complete acid reduction to elemental sulfur.
                          117

-------
Table  24.   EXPERIMENTAL  CONDITIONS AND RESULTS  FOR  SULFUR GENERATION EXPERIMENTS
              IN AN EIGHT STAGE,  4"  DIAMETER  FLUIDIZED BED REGENERATOR
Run Number
and Purpose
Estimation of
Requirements
f cr Total Acid
Conversion



Effect of 4
Temperature

Effect of
TT ft
HgS i
Cone.
fsG-27
SG-28
/SG-29
1 SG-30
[sG-31****
/^G-33
1 SG-3^
) SG-35
{ SG-36
JSG-37
(SG-3&
SG-39
SG-^0
SQ-kl
Temperature
°P
300
300
300
300
300
300
250
275
325
350
350
•325
325
325
Inlet Gas
Flow Rate,
N2
285
302
302
33^
320
283
SOU
29!+
2?6
267
267
. 276
236
362
H2S
96
108
108
72
108
130
131
126
118
Ilk
Ilk
119
158
37
Gas Composition*,
Volume %
Inlet
H2S
2k
27
.27.2
17.0
25.6
27-7
30. k
31.5
30.8
30
32.0
31.8
42.0
12.0
Outlet
H2S
10
17.8
11.2
7.6
13.5
15.7
2k. k
20
6.0
8.8
6.0
12.0
3.1
SO?
0
0
0.12
0.25
0.12
0
0
0
0.35
2.6
1.3
2.0
1.6
1.6
Total Solid Flow**
Rate, RTI
Ibs./hr.
53
61
58
56
61
61
6k
6k
56
5k
53
55
76
18
Sulfur Analysis
on Inlet Solid, Wt. -":
Inlet***
Psi
5.6
5.6
5.0
5.1
14.5
6.1
6.0
6.1
6.1
6.2
6.k
6.0
5.5
5.7
Outlet
Pso
13.0
11.5
12.7
11.0
22.4
14.2
9-k
11.3
18.7
18.0
18.3
18.0
17.0
ik.k
            *As determined by gas chromatograph.
           **Total solid flow rate of the carbon plus its sorbed sulfur compounds.
          ***Except for SG-31 the per cent sulfur is present only as sorbed acid.
         ****Inlet material for this run was a blend of the partially converted loaded carbon from
     Runs SG-28 and -29 which originally had an average acid loading of approximately .19 # HgSOl^ C.

-------
     Table  25.  OVERALL RATES OF ACID DECOMPOSITION  AND
                 CONVERSION TO SULFUR FOR THE  REACTION
                 3  H2S + H2S04 —* 4 S + 4 H20 IN AN
                 EIGHT STAGE,  4" DIAMETER REGENERATOR
Run
No.
SG-27
S6-28
SG-29
SG-30
SG-31
SG-33
SG-34
SG-35
SG-36
SG-37
SG-38
SG-39
SG-40
SG-41
Temp.
°F
300
300
300
300
300
300
250
275
325
350 '
350
325
325
325
Inlet
Acid
Loading
#H2SOzi/#C
0.21
0.21
0.18
0.19
****
0.24
0.23
0.24
0.24
0.24
0.25
0.23
0.21
0.22
Inlet
H2S
%
24
27
27.2
17
25.6
27.7
30.4
31.5
30.8
30
32
31.8
42
12
Outlet
S02
%
- 	
0
0
0.12
0.25
0.12
0
0
0
0.35
2.6
1.3
2.0
1.6
1.6
Conv.
to
S***,
%
I • r- 1 i ii
43
34
51
39
115
46
18
24
62
66
61
66
65
50
Overall Rate of Reac.,
# H2S04/#C-min.
Gas
Analysis*
•••^" [••^•••^•^^•^•^•^^^^^^.^^^^
1. 52(10-2)
1-05(10-2)
1.87(10-2}
1.08(10-2)
1.40(10-2)
1.39(10-2)
0.71(10-2)
1.33(10-2)
2.83(10-2)
	
2.54(10-2)
2.95(10-2)
3.24(10 )
1.04(10 )
Carbon
Analysis**
1.2(10-2)
1.1(10-2)
1.34(10-2)
1.06(10-2)
	
1.62(10-2)
0.66(10-2)
0.88(10-2)
2.0(10-2)
2.08(10-2)
1.96(10-2)
2.04(10-2)
2.56(10"2)
0.48(10-2)
   *Based on gas analysis using a chromatograph.
  **Based on carbon analysis using a combustion analysis technique.
 ***Calculated from experimental data.
****Inlet material for this run was a blend of partially converted
      loaded carbon from Runs SG-28 and  -29 which originally had an
      average acid loading of approximately .19 Ib. acid/lb. C.
                               119

-------
Figure 41.   Effect of temperature on  the rate  of conversion
              of sorbed sulfuric acid to elemental sulfur
                                      Inlet HgS = 31 vol % (AVG)
                                      Inlet Acid = 0.24 Ib. acid/lb.
                                      Loading     (AVG)
           250
275          300          325

    Reactor Tempera tvre, °F
                               120

-------
The following  two-step  reaction  sequence  is believed  to  be
the most probable mechanism for  conversion of  acid  to
elemental  sulfur:
           H2S04  +  H2S 	*» S02 + S + 2 H20                 (35)


            S02 + 2 H2S 	+.  3 S  +  2 H20                 (36)
           H2S04 + 3 H2S 	+•  4 S  +  2 H20                 (37)
Because  SC>2  is  an  intermediate  in  these reactions, its
presence in  the outlet gas  could be expected under at
least some reaction  conditions.

The results  in  Table  25  show  that the outlet gas S02
concentration increases above 300°F, rising from about
0.15% at 300°F  to  27, at 350°F.  The 2% figure represents
about a  20%  conversion of acid  to  S02-  The results indi-
cate that operating  temperature above 325°F should be
avoided  in order to  minimize S02 formation.

The reaction rate  also increases with increasing H2S
concentration,  as  seen most clearly by inspecting the
results  of SG-39,  -40 and -41.  Further discussion of
the H2S  concentration effect is postponed to the next
section  on rate model development.

These experiments  also provided an average reaction rate
for anticipated process conditions to be used in prelimi-
nary design  work.  To obtain the average rate over a
wide set of  operating conditions it was necessary to
obtain complete acid conversion, and this was accom-
plished  by passing the carbon through the reactor twice.
The partially reacted carbon product from Runs SG-28 and
-29 was  combined and fed to the reactor again in SG-31.
The results  of  these experiments indicate an overall
average  reaction rate to complete  conversion under antici-
pated conditions of  about 0.01  Ib. acid/lb. C-min.

Rate Model Development -

A series of  differential batch  experiments was carried
out in a 4"  diameter fluidized  bed reactor in order to
obtain data  to  develop a rate expression.  In each run  a
small batch  of  acid  loaded  carbon  was fluidized with
reactant gas at constant conditions of temperature and


                          121

-------
gas concentration.  Samples  of carbon were  removed at
intervals  for analysis,  thereby providing reaction data
as a function of time for  total sulfur content and acid
loading.   Space velocity was sufficient  that  the inlet
and outlet reactant gas  concentrations were essentially
identical.   This justified the differential reactor
assumption.

The variables studied were temperature,  acid  loading,
sulfur  loading, and the  gas  concentrations  of H2S and H20
The conditions for each  run  are presented in  Table 26.
Temperatures of 250, 300 and 325°F were  tested,  and H2S
 Table  26.   EXPERIMENTAL CONDITIONS FOR  DIFFERENTIAL
             SULFUR GENERATION RUNS
           Carbon Used:
           Virgin Procursor:
           Residual S on Virgin Carbon:
           Carbon Density:
           Linear Gas Velocity:
           Diluent Gas:
SA-27-A (335-365 min.)
Skid 96298
0.6 wt. $
1*9.7 lbs./ft.3
2 ft./sec.
Nitrogen
Run
Humber
DSG-2U
DSG--25
DSG-26
DSG-lil
DSG-2T
BSG-23
DSC -29
DSG-30
DSG-31
DSG-32-
DSG-33
DfiG-3'i
DSG-35
DSG--36
DRG-3Y
DSG-30
DSG-39
Gas Concentrations ,
Volume %
H_2S
J».»i
lit.l
25. U
lull
5.3
10.0
15.0
20.8
31.6
5.3
5.3
5.3
5.0
9.1
18.2
26. 1>
37.6
H?0
0
0
0
10
0
0
0
0
0
30
20
10
0
0
0
0
0
Nominal Bed
Temperature ,
°F
250
250
250
250
300
300
300
300
300
300
300
300
325
3?5
325
325
325 *
                           122

-------
concentration was varied over a range of  4  to  37%.  The
H20 concentration was varied from 0  to  30%.  The  data
from each run is given in Appendix A-ll.

To evaluate the data the rate expression  was assumed to
be of the form:
                     g2(Xv) g2(XS) g3(y)  g(y)            (38)
The data was analyzed by stepwise and multiple regression
methods and the following expression was obtained for
experiments of an inlet water concentration of zero.
              23.6 a'2644/1 v  °'58 X?-67    Ibs. S/lb. C-min.     (39)
The sulfur loading  (Xs) was found to have no observable
effect on the rate, or at least the effect could not be
isolated from the effect of the acid loading (Xv), so
that g(XS) = 1.  The function of water, gCyuorp' was not
determined explicitly due to the nature of this variable's
effect on the reaction, and consequently Equation  (39)
applies only for yjjoQ = ^-  Equation (39) fits the differ-
ential data reasonably well as shown in Figure 42.  A
statistical analysis showed the average difference between
the model and the differential data to be 17%, with a
maximum difference  of 50%.

Extension of the rate model to situations where water
vapor is present required some provision for the  effect
of this variable.   Figure  43  shows the effect that water
vapor has on the reaction rate.  The data indicate that
the rate decreases  sharply as yH2o increases from 0 to 20%,
but that there is no further decrease above 20%,   There  is
a factor of about 6 between the rates at 0 and
20% water vapor.

The rate model was  tested against the results of  sulfur
generation runs in  the 4" diameter fluid bed pilot regene-
rator (subsequently described in this section) by using  it
in a design equation to calculate predicted reactor volumes
In conducting the evaluation,various representations  of
                          123

-------
Figure 42
Comparison  of the sulfur  generation rate model
to the experimental data  for 250 to 325°F
                       6    8   10         20

                         HgS Concentration, Volume %
                                     40
60    80
                              124

-------
  Figure 43.  Effect of H20 concentration on rate  of sulfur  generation
  ID

  O
  oo
  IT)
OJ
4J
to
  *
  *1-
  UD
                            10                20


                                    H20  Cone., volume %

-------
 g(yH20)  were tried in an attempt to obtain the best corre
 lation between actual and predicted reactor volumes.  The
 approach which proved most successful was to take a con-*
 servative rate constant adjusted by a factor of 6 for H20
 concentrations above 2070.  More sophisticated methods did
 not significantly improve the correlation.  The simple
 approach is considered justifiable with respect to the
 available data on the effect of y^O' anc* it: yie^s con~
 servative design estimates.  The adjusted rate equation
 proposed as the best available expression for designing
 sulfur generators is, therefore,
                                                          (40)
                   3.8 e-         0-58 ^0.67
            dt                H2S
Design  equations were  derived from this rate expression
based on  the  same multistage fluid bed reactor model used
for the S02 sorber  design equations.   The important assump-
tions in  the  reactor model were that  the carbon phase was
well-mixed and  the  gas phase was plug flow.   The following
design  equations were  derived.   The complete derivation
is in Appendix  A-9.
        N0 42     ,    0-42       -2644/T   9    0.67
    (YH2s)j+l   *  
-------
Fluid Bed Studies in 4" Diameter Pilot
Regenerator To Determine Adequacy of Existing
6" Diameter Reactor as Pilot Sulfur Generator -

Experiments were carried out in the 8 stage, 4" diameter
fluid bed pilot regenerator to determine whether the
existing 6" diameter S02 sorber would make a satis-
factory sulfur generator for the integrated pilot plant.
The primary question was whether  high  conversion of
both reactants could be obtained in the 6" diameter reactor.
The goal was set at 99% acid conversion to sulfur at the
maximum H2S utilization at the space velocities possible
in the existing 6" diameter reactor.  The goal of 997o conver-
sion of acid to sulfur was considered necessary during this
period in development in order to avoid disruptions caused
by unreacted acid entering the next step of regeneration.
Later results from integrated operation showed that
unreacted acid could be tolerated, but  this was not known
at the time decisions were made concerning adequacy of the
6" unit as a sulfur generator.

Achievement of 99% conversion of acid to sulfur was expected
to be a difficult goal because the inlet M2S concentration
would have to be kept fairly low due to fluidization
requirements in the 6" diameter unit and in order to
obtain acceptable H2S utilization.  In fact, the stoichi-
ometric H2S concentration  required for complete acid con-
version at 1007o H2S utilization was only about 6% for the
6" diameter unit.  This concentration was based on the  i
anticipated sulfuric acid feed rate for integral operation,
which determined the stoichiometric H2S feed requirement,
and on the volumetric gas flow rate necessary for proper
fluidization in the 6" reactor, which determined the
degree of dilution.

The experiments in the 4" diameter unit were designed to
simulate the anticipated conditions in the 6" unit as
closely as possible, but the conditions could not be simu-
lated with complete accuracy due to a difference in total
carbon bed depth between the two reactors.  The 6" unit's
planned bed depth was greater by a factor of 1.7 to 2.3
depending on the length of the overflow weirs, so that
simulation of carbon residence time in the 4" unit required
a reduction in the carbon feed rate per unit area to
attempt to compensate for  lower  bed depth.  As a conse-
quence of the reduced carbon feed rate per unit area, the
stoichiometric H2S concentration was lower for the simu-
lation runs.  The  lower bed depth in the 4" unit also
caused a higher space velocity, since linear gas velocity
was held constant.  These two inaccuracies in simulation--
higher space velocity and lower stoichiometric H2S
concentration--had a negative effect on conversion of


                         127

-------
reactants.  Therefore, the results of the simulation experi-
ments were conservative with respect to the likely per-
formance of the 6" unit.

The conditions and results of the experiments are presented
in Table  27.   The runs were made at carbon feed rates of
9.8 and 6.3 Ibs./hr. to provide carbon residence times of
32 and 50 minutes, respectively.  Inlet acid loading was
about 0.21 Ib. acid/lb. C.  Average column temperature
was about 290°F in most experiments.  Lower temperatures,
about 260°F, were tried in a few runs.  Linear gas velocity
was 2.0 ft./sec.  The inlet H2S concentration ranged from
3.6% to 29.8% and the H2S feed rate ranged from 1 to 5 times
the stoichiometric requirement for complete acid conversion.
Space velocity was 2800 to 3000 hr.'1.

The main conclusion of the experiments was that the goal
of 99% acid conversion to sulfur with high utilization of
H2S was not attainable under the 6" diameter unit
conditions.  The highest H2S utilization, 72%, was achieved
in Run SG-55 at a H2S feed rate of 1.05 times the stoichi-
ometric requirement, with an acid conversion to sulfur of
63%.  At higher H2S feed rates the H2S utilization was
lower, falling to about 50% at a ratio of 1.6, 40% at 2.2,
and 30% at 4 with acid conversions to sulfur of 73, 85,
and 90%, respectively.

In discussing the conversion of H2S04, it is important to
note that in addition to forming the desired elemental
sulfur product, a fraction of the acid can react to form
S02 that is evolved in the outlet gas of the acid conver
converter.  This reaction was observed in the experiments.
The results show that the percent acid conversion was
about the same for all runs, falling between 92% and 9770.
The conversion to sulfur, however, ranged from 61 to 9070;
and the amount of acid converted to S02 varied from 0 to
43%.  Inlet H2S concentration appeared to be the main
determinant of the product distribution between sulfur
and S02-  As H2S concentration was increased, conversion
to sulfur also increased, and evolution of S02 decreased.
The effects of inlet H2S concentration on acid conversion
to sulfur and S02 evolution are shown in Figures 44 and
45, respectively.

In order to obtain 90% acid conversion to sulfur, an inlet
H2S concentration of 20 to 30% was required, which repre-
sented a H2S excess of 3 to 5 time's the stoichiometric
requirement and yielded a H2S utilization below 30%.
Experiments in which higher H2S utilization was sought
                       128

-------
                  Table  27.    SUMMARY OF SULFUR GENERATION  RESULTS
RUN
HO.
SG-55
SG-57
SG-64
SG-56
SG-62
SG-65
SG-58
SG-59
SG-60
SG-61
SG-6fi
SG-67
SG-68
SG-69
SG-70
COLUMN
TEMP.,
"f
AVG
294
29E
2S5
294
291
2S3
286
290
280
284
280
289
290
260
262
RANGE
275-312
276-315
270-303
275-324
271-316
242-270
270-313
270-311
250-303
263-297
236-297
237-303
242-320
223-273
225-281
CARBON
• INLET STREAM
C
RATE,
*C/HR
9.8
It
n
ii
it
M
6.3
if
N
«
9.8
M
n
ii
u
RES.
TIME
KIN.
32
II
"
M
(1
H
50
II
y
H
32
U
II
H
II
ACID
LOAD.,
*AC1D/
*C
0.215
• •
t
Ii
•>
II
II
•
II
»
0.200
0.209
0.208
0.207
0.211
GAS INLET STREAM
GAS*
FLOW
CFH?
70°F
438
439
439
436
435
435
438
438
437
435
464
443
435
456
461
H2S CONC.
VOL.
6.1
9.6
9.6
13.6
12.8
13.4
3.6
6.3
10.4
13.1
10.7
21.7
29.8
22.6
8.4
TIMES
STOICI!
1.05
1.66
1.66
2.36
2.22
2.32
0.92
1.62
2.67
3.45
2.15
3.74
5.13
4.29
1.58
CARBON OUTLET STRFAH
ACID
LOAD.,
#ACID/
K
0.015
•0.012
0.011
0.011
0.011
0.017
0.013
0.012
0.013
0.012
0.009
0.005
0.008
0.010
0.014
S LOAD.,
K/K
THEO.
0.282

"
•
II
11
11
II
II
II
0.261
0.272
0.271
0.269
0.275
ACT.
0.178
0.206
0.222
0.217
0.239
0.236
0.171
0.197
0.230
0.240
0.207
0.240
0.245
0.239
0.205
WT.
FINES
"C/HR
...
—
...
0.007
—
—
0.006

—
...
...
...
—
...
—
GAS OUTLET STREAM
COMPOSITION**,
VOLUME I
H?S
1.8
4.4
4.5
8.0
7.3
8.0
1.5
2.6
5.0
8.0
4.8
15.2
22.0
15.3
3.8
SO?
0.79
0.49
0.44
0.36
0.30
0.16
0.50
0.35
0.17
0
0.46
0.09
0
0.07
0.17
H?0
7.0
7.3
8.6
7.9
8.9
6.6
6.2
7.8
10.0
8.0
6.4
8.0
7.6
8.6
5.5
N?
90.6
87.7
86.5
83.7
83,5
85.2
91.8
89.2
84.8
84.0
88.3
76.7
70.4
76.0
90.5
~so7
EVOL.
",. IN
43.0
27.5
23.9
19.1
16.4
8.7
43.0
30.0
14.6
0
30.9
5.2
0
4.2
9.6
H2S
UTIL.
% OF
72
52
52
40
41
39
57
56
49
37
54
30
26
29
55
S BALANCE,
LBS.
IN
2.96
4.28
4.25
5.
5.37
5.62
1.78
2.79
4.25
5.20
4.83
8.90
11.5
9.26
8.94
OUT
2.87
4.12
4.26
5.45
5.22
5.49
2.00
2.50
3.60
4.63
4.34
8.4
10.5
6.61
3.36
ACID
DECOMP
X
98.0
94.4
95.0
95.0
95.0
92.0
93.0
95.0
94.0
94.0
95.5
97.6
96.3
95'. o
93.2
CONV
TO S
V
«B
63
73
79
77
85
84
62
70
82
85
79
88
90
89
75
ACID
DECOMP.
RATE,
#ACID/
#C-MIN.
*. 27 (ID"3)
4. 94(10-3)
5.33(10-3]
5. 2K10-3)
5.74(10-3)
5.66(10-3)
2.64(10-3)
3.04(10-3)
3.55(10-3)
3.70(10-3)
4.97(10-3)
5.76(10-3)
5.88(10-3)
5.73(10-3)
4.92(10-3)
SPACE
YELOC.
KR.'1
2,ss:
2.S60
2,860
2.840
2,530
2,830
2,830
2,850
2,840
2.83Q
3,020
2,880
2,830
2,970
3,000
 *Linear gas velocity is 1.9-2 ft./sec. at average column temperature.
**C02 and CO not detected in outlet gas.

-------
   Figure  44,
Effect of  inlet H2S concentration on  the

per  cent conversion to  sulfur  in simulation

experiments  using  a 6"  diameter fluid bed

unit for integrated operation  with an 18"

diameter S02 sorber
o

-------
Figure 45
Effect of H2S concentration and carbon residue
time  on acid evolved as  S02 in  simulation
experiments  using  a 6" diameter fluid  bed unit
for  integrated operation with an 18" diameter
S02  sorber
-a
0)
•o
•r—
O
                                              C Residence Time, 32 Min
                                               Avg. Temp. Near 285°F
                                              C Residence Time, 50 Min.
                                               Avg. Temp. Near 285°F
                                              C Residence Time, 32 Min.
                                               Low Avg. Temp. Near 255
                       10       15       20       25
                         Inlet H2$ Concentration, Volume %
                                131

-------
yielded  lower  acid  conversions,  in  the  60 - 707o range.  The
results  indicated convincingly,  therefore,  that reactant
conversion goals could not be met in the 6" unit.  It was
estimated ba.sed on  the experimental data that an increase
in  carbon residence time  in  the  6"0 unit to about 90
minutes  would  be required to achieve the goal of 99% acid
conversion to  sulfu*.  An inlet  H2S concentration of 30%
would  also be  necessary,  resulting  in a H2S utilization of
about  20%.  To obtain the 90 minute residence time would
require  a carbon bed depth 3.25  times that  employed in the
4"0 unit simulation experiments.  This  would mean an
increase in the number of stages in the 6"0 unit from 9
to  13, with 8" overflow weirs.   The increased bed depth
would  reduce the space velocity  to  about 1,000 hr.~l.

Comparison of  Fluid Bed Design
Model  with Experimental Results  -

The fluid bed  acid  converter design model was compared to
the 4" diameter fluid bed runs by using the model to pre-
dict the theoretical reactor size and number of stages
under  the conditions for  each run.  The results of the
comparison are presented  in  Table 28.    The predicted
reactor  volume averages about 307o higher than the actual
volume.  The predicted number of stages ranges from 8 to
13  with  an average  of 10.5,  compared to the actual 8 stages
in  the reactor.  The model,  therefore,  conservatively pre-
dicts  reactor  volume, and it is  considered adequate for
prediction and scale-up to the next development stage.

Fixed  Bed Experiments -

After  finding  that  acid conversion  is limited to a maximum
of  90% in the  simulation  experiments, tests were made in
a 1" diameter  fixed bed apparatus to ascertain whether
the 9970  conversion  goal was  obtainable  for  the particular
batch  of carbon in  question.  The tests were made at a
H2S concentration of 30%.  Runs were conducted using fresh
acid loaded carbon  and partially reacted product material
from Runs SG-65 and -68.  The results are presented in
Table  29.

For the  acid loaded sample,  acid decomposition was measured
at  97.5% and conversion to sulfur was 119%.  The SG-65
product  material yielded  98.1%, acid decomposition and 9970
conversion to  sulfur, while  the  yields  for  the SG-68 sample
were 97.3% acid decomposition and 94%, conversion to sulfur.
These  results  indicate that  99%  acid conversion to sulfur
definitely is  possible with  the  carbon  in question.
                        132

-------
                  Table  28.   COMPARISON OF THE FLUID  BED DESIGN MODEL WITH  EXPERIMENTAL
                                SULFUR GENERATION FLUID  BED DATA
u>
RUN
NO.
S6-27
SS-55
SG-56
SS-57
SG-58
SG-59
SG-60
S6-61
S6-62
S6-64
S6-65
S6-66
S6-67
SG-6S
SG-69
AVG.
TEMP.
°F
300
294
294
295
286
290
280
284
291
286
253
280
289
299
260
CARBON
RATE,
*/HR.
42.8
9.8
9.8
9.8
6.3
6.3
6.3
6.3
9.8
9.9
9.8
9.8
9.8
9.8
9.8
INLET
GAS
RATE,
CFH 1? 70°F
410
438
436
439
438
.438-
437
435
435
439
435
464
443
435
456
INLET
*H2s
0.24
0.061
0.136
0.096
0.036
0.063
0.104
0.131
0.128
0.096
0.134
0.107
0.217
0.298
0.226
OUTLET yH2$
ACTUAL
0.100
0.016
0.080
0.044
0.015
0.026
0.050
0.080
0.073
0.045
0.080
0.048
0.152
0.220
0.153
CALC.*
0.128
0.032
0.092
0.057
0.017
0.039
0.075
0.096
0.082
0.055
0.086
0.073
0.166
0.245
0.178
INLET ACID
LOADING,
#/#C
ACTUAL
0.210
0.216
0.216
0.216
0.216
0.216
0.216
0.216
0.216
0.216
0.216
0.200
0.209
0.208
0.207
CALC.**
0.210
0.125
0.175
0.159
0.127
0.153
0.186
0.216
0.182
0.165
0.198
0.144
0.198
0.208
0.199
S02
EVOL.,
% INLET
ACID
0
42.1
19.2
26.2
41.1
29.0
14.1
0
15.9
23.6
8.5
28.2
5.0
0
4.1
rOTJTtET
ACID
LOAD.,
*/#C
0.1197
0.0150
0.0110
0.0120
0.0130
0.0120
0.0130
0.0120
0.0110
0.0110
0.0170
0.0090
0.0050
0.0080
0.0100
' MATERIAL
BALANCE ,
LBS. SULFUR
IN
11.10
2.91
5.61
4.18
1.75
2.73
4.21
5.17
5.31
4.18
5.52
4.76
8.64
11.4
• 9.21
OUT
10.30
2.67
5.23
3.86
1.83
2.35
3.38
4.46
5.16
4.03
5.35
4.11
8.08
10.50
8.30
NUHBER
OF
STAGES
ACTUAL
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
CALC.
9.5
12.6
10.3
11.3
12.4
9.8
8.3
8.0
11.1
12.8
12.2
11.7
9.9
7.9
9.7
SPACE
VELOCITY,
HR.'1
ACTUAL
2,725
2,850
2,840
2,860
2,850
2,850
2,840
2,830
2,830
2,860
2,830
3,020
2,880
2,830
2,970
CALC.
2,290
1,840
2,250
1,980
1,880
2,330,
2,800
2,900
2,080
1,820
1,900
2,110
2,370
2,940
2,490
TOTAL
VOLUME,
CU. FT.
ACTUAL
0.1 39C
0.1396
0.1396
0.1396
0.1396
0.1396
0.1396
0.1396
0.1396
0.1396
0.1396
0.1396
0.1396
0.1396
0.1396
CALC.
0.1663
0.2203
0.1796
0.2057
0.2165
0.1704
0.1448
0.1390
0.1937
0.2240
0.2123
0.2045
0.1735
0.1372
0.1698
                  *Higher outlet H2S concentration used by computer to
                 **Inlet acid loading corrected for S02 evolution.
                                                      obtain correct material balance.

-------
                                   Table  29.    FIXED BED  SULFUR GENERATION  EXPERIMENTS
CO
Run
Nunber
2
3***
4***-*
5*****
Gas Flow
cc/mi n .
N2
125
tl
It
II
H2S
55
n
it
M
Average
Temp.
°F
286
285
289
282
Space
Velocity
hr.-1
1,000
II
«
(1
Carbon Anal
Inlet
% S
5.56
18.1
0.56
19.3
Acid Load, 1 acid/# C
From
% S
0.194**
___
—
—
SO?
Anal .
0.218
__.
—

Acid
Titr.
0.181
0.020
0
0.011
/ses
Outlet
# S/*-C*
0.30
0.28
0.016
0.25
* acid/I C
0.005
0.004
0.001
0.006
Acid
Decomposition
%
97.5
98.1

97.3
Conversion
to Sulfur
%
119
99
...
94
                              *Less residual sulfur on virgin carbon.



                             **Used as inlet acid loading, since have  standard for total  sulfur analysis.



                            ***Feed sample to fixed bed from product of Run SG-65 (18.1%  S, 0.0164 Ib. acid/lb. material).



                           ****Feed is virgin carbon.



                          *****peecl sample to fixed bed from product of Run SG-68 (19.3%  S, 0.00882 Ib. acid/lb. material).

-------
The 11970 conversion to sulfur for the acid loaded sample
was an indication of a separate phenomenon, believed to be
reaction of H2S with chemisorbed oxygen.  This would cause
additional formation of elemental sulfur on the carbon.  To
investigate this possibility, a sample of virgin carbon was
tested in the fixed bed apparatus, and the results showed
enough sulfur formation to account for TL of the 19%
excess conversion.  This was considered valid qualitative
evidence that reaction of H2S with chemisorbed oxygen does
occur.  The other 12% was attributed to experimental and
analytical errors.

Development of Moving Bed Reactor
for Integrated Pilot Plant-

The results of the simulation experiments in the 4" diameter
fluid bed reactor indicated that the 6" diameter unit was
unsuitable for use as a sulfur generator in the integrated
pilot plant because it was oversized for the application.
A sulfur generator properly sized to match the existing
18" diameter S02 sorber would be only 2.5" diameter,  which
would not be feasible from a mechanical standpoint.

Fixed bed tests showed that the desired degree of acid con-
version was obtainable.  This led to the suggestion of a
moving bed reactor as an attractive alternative for the
fluid bed reactor initially intended for the pilot plant.
Exploratory experiments were then carried out in a 1.5"
diameter moving bed reactor in order to evaluate the
potential of this approach.  The results were favorable,
indicating acid conversion of 98-100% and H2S utilization
over 99%, at space velocities in the 100-330 hr.-l range
and.carbon residence times between 70 and 185 minutes.
The conditions and results of the moving bed runs are
compared to fluid bed data in Tables 30 and 31.

The success of the 1.5" diameter moving bed experiments  led
to the decision to proceed with development of a pilot scale
moving bed sulfur generator.  Design calculations were made
for reactors of 6" and 8" diameters.  The 8" diameter  size
was chosen because in a 6" diameter unit there was a possi-
bility that the minimum fluidizing velocity could be
approached in the reactor and seal leg.   (A moving bed of
this size does not have the  fluidization requirements, but
in fact fluidization should be avoided.)  The  design  speci-
fications for the 8" diameter moving bed sulfur generator
are given in Table 32,   and a drawing of the  reactor  system
is shown in Figure  46.    Although the overall reactor
length is 10 feet, the carbon bed depth is 6 feet  in  the
                       135

-------
                TABLE 30.   COMPARISON OF  FLUID  BED AND MOVING  BED SULFUR  GENERATION TESTS
Equipment
4" Dia.
Fluid
Bed
Fluid
Run
No.

SG-68


Bed*
Modified
6" Dia.

Moving
Bed**




SG-73

Column
Ten?., °F
Avg.

290


290



285
Range

242-320






270-300

Carbon
Inlet Acid
Loading
# Acid/* C

0.208


0.184



0.165

Residence
Time
minutes'

32


~90 -



-180

Rate
!/hr.

12.1


35



1.1
1

Hydrogen
Cone.

29.8

Stoich.
Ratio

5.1

I
30 i 5




,
32.3

1.!

Sulf-;ce
Outlet! Util.
1
22.0 ! 26
!

24



-0

~20



>90
i
Acid Lest as
SOz in Off Gas
Inlet

0


0



0
Cone.
in Gas

—0


0



~0

Acid
Decorap.

96


98



94
Sulfur
8aUr.ce
In/Out

1.1


—


Outlet Sulfur
Loea, ? S/- C
Theor.

0.271


0.240


!
—
0.212
Actual

0.248


0.238



i
Conversion j y*?»«
Sulfur ; hr-

90


99



0.216 : 102

2.850


1,000



100
' 1
OJ
                        *These are projected results ir. 6" diameter unit if codified based on all results to operate at a space velocity
                  of 1,000 hr.~l, i.e. increase niniber of stages from 9 to 13 and bet height to 8 inches.
                       **Actua1 moving bed data.

-------
Table 31.  DATA SUMMARY - 1-1/2" DIAMETER MOVING BED SULFUR GENERATION TESTS
Run
No.
SG-74
SG-79
Column
Temp.
°f
Avg.
270
268
Range
260 -
280
255 -
'280
Carbon Inlet Stream
Reac.
Inv.
# C
3.09
0.92
Carbon
Rate
?C/hr.
1.02
0.81
Res.
Time
min.
185
68
Acid
Load
#Acid/#C
0.19
0.19
Gas Inlet Stream
Gas
Flow
CFH ?
70° F
9.14
6.2
H2S
Cone.
Vol.
a
IB
29.5
30.6
Times
Stoich.
1.1
1.1
Linear
Gas
Velocity
ft. /sec.
0.3
0.2
Carbon Outlet Stream
Sulfur Load
Ib. S/lb. C
Theoreti cal
0.248
0.252
Actual
0.248
0.247
Gas Outlet Stream
Composition,
Vol. % (Dry)
12$
~0
.18
S02
-0
.10
C02

—
CO

—
H20
N.D.
N.D.
N2
N.D.
N.D.
S02
Evol.
% In.
Acid
0
<1%
H2S
Util.
% of
Inlet
>99
>99
S Balance
S
In
Ib.
0.28
0.22
S
Out
Ib.
0.27
0.18
%
Conv.
to
S
100
98
Acid
Decompositior
Rate,
#Acid/#C/min.
t
IxlO'3
3xlO"3
Space
Veloc.
hr.'1
100
335

-------
         Table 32.   DESIGN CONDITIONS  - MOVING
                      BED  SULFUR  GENERATOR
      Unit Size
          Nominal, NPS                                  8
          I.D., inches                                8.33

      Temperature, °F                                  300

      Carbon Rate, Ibs./hr.                            27.7

      Acid Load, Ib./hr.                               0.184

      Gas Rate, SCFH                                   220

      Gas Velocity, ft./sec.  at  300°F                  0.25

      Carbon Residence Time at S = 100 hr."1, hrs.      2.85

      Carbon Inventory at  S = 100 hr."1, Ibs.           79

      Carbon Bed Height, ft.
          S = 100 hr.'1                                  6
          S = 200 hr."]                                  3
          S = 500 hr."1                                 1.2

      Pressure Drop* at S  = 100  hr."1, in. H20          31

      Reactor Length, ft.                               10

      Seal Leg Length, ft.
          Total  Length                                  8
          Purged Length                                  6

      Seal Leg Velocity, ft./sec.
          S = 100 hr."1                                0.25
          S = 200 hr.'1                                 <.l
          S = 500 hr.'1                                 <.l
*Based on previously measured pressure drop.
                            138

-------
                                      Figure  46.   Moving  bed sulfur  generator
VO
                                                                              DIA « I STA5E FLUID 6EO
                                                 e"c« » " ic' oss
                                                  T--2SO-5ZS *F
NOTES
                                                                                               I. ALL OIANAETER OMANMlEi BO
                                                                                                CARBOK1 Uie 2S% CONJB.S
                                                                                               2.THCRivioce>up'..ss Ti-3 4 Ti-4
                                                                                                  £AS SAMPLE PROBE 5.S-1
                                                                                                ARE  RETRACTABLE.
                                                                                             TO BUCKET E.I.E.VA70R.
                                                                          COKjVt'N'OK

-------
 8"  diameter unit  to give a space velocity of about 100 hr."-*-
 for a gas  flow rate of 220 SCFH.  Carbon residence time is
 about 2.85 hours  for a carbon feed rate of 27.7 Ibs./hr.

 An  experimental program was  carried out to evaluate the
 performnace of the 8" diameter moving bed reactor and to
 determine  the optimum operating conditions.  The experi-
 ments were divided chronologically into three series of
 runs.  An  initial series of  runs was made primarily to
 study the  effect  of temperature.  The results indicated a
 need for additional heating  in the lower section of the
 reactor.   The reactor was modified by installing a
 special finned steam heater, and a second series of runs   L
 was carried out.  These runs were not too useful because
 the finned heater caused a carbon flow problem which led
 to  poorer  reactor performance.  The finned heater was
 removed, and a final series  of runs was made to study
 other methods of  improving reactor performance.  The
 experimental conditions and  results of all three series of
 runs are summarized in Table 33.

 A carbon feed rate of 28 Ibs./hr. was used in most runs,
 with an acid loading between 0.14 and 0.24 Ib. H2S04/lb.
 carbon.  The H2S  feed rate was 0.94 to 1.4 times the
 stoichiometric requirement for complete acid conversion.
 The H2S concentration was 27.6 to 34.67o by volume, and
 total gas  flow rate was 200  to 280 cfh at 70°F.  Tempera-
 ture was an important variable with average bed tempera-
 tures ranging from 233 to 319°F.  An average of around 270-
 280°F was  typical in many runs.  Temperatures at the top
 of  the reactor were higher than at the bottom, with vari-
 ations ranging from 20 to 100°F.

 Initial Runs -

 The initial eight runs were  primarily intended to study the
 effect of  temperature.  These runs were made at average
 carbon bed temperatures from 230 to 320°F and at inlet H2S
 concentrations near 30 volume 7o.  The total gas flow rate
was adjusted as necessary to provide the 110% of stoichi-
 ometric H2S reactant needed  to convert the sorbed sulfuric
 acid which varied from about 0.17 to 0.23 Ib. acid/lb. C.
 The results generally indicate that the higher temperatures
 favor higher H2S  utilization, but lower temperatures favor
 lower amounts of  sorbed acid evolved as S02 in the off-gas.

Effect of Temperature on Evolution of S02 -

The effect of the inlet carbon temperature, which is the
maximum measured  carbon temperature in the present experi-
                          140

-------
                       Table 33.    SUMMARY  OF  SULFUR  GENERATION  RESULTS

RUN
NO.

SG-86

SG-87

56-88

SG-88R


SG-89

SG-90


SR-91



SG-92


COLUMN
TEMP.,
AVG

319

290

253

262


243

242


?33



244


RANGE

326

260-
305
91Q
265

236-
301

206-
296

198-
300

192-
236


203-
289


CARBON INLET STREAM
REAC
INV.

75



H

H


n

tt


n

,

it


c
RATE

28

28

27

28


28

28


?8



28


RES.
TIME
KIN.

162

162

166

162


162

162


16?



162


ACID
LOAD.
#ACIO/
#C -

0.167

0.185

0.176

0.227


0.204

0.227


0.220



0.212


GAS INLET STREAM
*
GAS
FLOW
CFH@
70°F

204



II

280


280

280


280



280


H2S
CONC.**
VOL.

30.5

28.8

27.6

30.4


32.0

29.5


31.8



32.1


TIMES
STOICH

1.1

0.94

0.96

1.1


1.3

1.1


1.2



1.3


LINEAR
GAS
VELOC.
R/SEC

0.24

0.23

0.22

0.30


0.30

0.30


0.29



0.30


CARBON OUTLET
STREAM
ACID
LOAD.
fACID/
1C

0.019

0.023

0.015

0.018


0.010

0.007


0.012



0.007


SULIUR
LOADING,
THEO.

0.218

0.241

0.230

0.296


0.26E

0.29f,


0.287



0.277


ACT.

0.204

0.221

0.205

0.273


0.264

0.285


0.281



0.284


GAS OUTLET STREAM
COMPOSITION, VOL. %
H2S

1.2

1.6

5.9

2.0


4.3

1.3


4.7



3.5


S02

2.2

0.8

0.09

0.90


0.8

0.80


0.43



0.61


C02

0.3

0.18

0.09

0.0


0.1

0.0


0.09



0.09


CO

0

0

0

0


0

0


0



0


H20

37

35,5

34.8

39.4


37.1

34.8


38.1



40.6


N2

61.5

61.6

62.2

56.8


57.0

59.7


59.3



59.8


S02
EVOL.
% OF
INLET
ACID

27

6

1

12


12

10


6



11


H2S
UTIL.
% OF
INLET

95

97

74

92


84

95


82



83


SULFUR
BALANCE
IN

6.9

6.7

6.3

9.3


9.7

9.0


9.8



9.8


OUT

6.5

7.0

7.0

8.9


9.4

8.5


9.7



9.6



ACID
DECOMP

88

88

91

92


95

97


94



97



CONV
TO S
at

90

92
***

90

92


99

94


98



102



SPACE
VELOC
HR.-l

90

90

90

123


123

123


123



123



PURPOSE
OF RUN
Effect of
Temp. -
320° F
Effect of
Temp. -
290°F
Effect of
Tercp. -
250DF
C Prehtr.

& lO'.R.H.
C Prehtr.
(3 230CF
i 5SR-.H.
C Prehtr.
g 300" F
I KK.H.
C Prehtr.
9 230°F
f. 10IR.H.
C Prehtr.
P 28CCF i
Add'l r.tg
from Wai 1
at Bottom
of Reactor
  *Frora rotameters
 **As determined ay gas chromatograph.
***An error analysis indicated the maximum uncertainty in this numter is ±10%.
     calculated to be tO.3 for inlet t S in.
The error in the sulfur material balance was
                                                   (continued)

-------
                         Table  33  (continued).   SUMMARY OF  SULFUR GENERATION RESULTS
t"0

RUN
HO.
S6-94


SS-95




SG-93


SG-96

SG-99

SG-100


SG-101

SG-103

SG-102

COLUWi
TEMP. ,
°F
AVG
280


283




270


276

274

282


277

255



PA'IGE
250-
305


247-
312



238-
?i n


2J7-
311

254-
285

256-
230


259-
232

251-
273

280-
316

CARBON INLET STREAI!
9EAC
ICV.
i C
70


II




tl




75

»


I

•

«

C
RATE
aC/HR
28


24




24


24

28

23


23

28

28

RES.
TIME
MIN.
151


176




176


175

162

162


162

162

162

ACID
LOAD.
*ACID/
#C
0.239


0.229




0.224


0.222

0.163

0.166


0.168

0.143

0.227

GAS INLET STREAM
GAS
FLOW
CFH 0
70°F
214


231




214


214

206

206


206

200

206

H2S
CCNC."
VOL.
w
A
32.4


34.6




32.6


32.2

32.8

33.0


33.0

31.0

32.9

TIMES
STOICH
1.0


1.3




1.3


1.3

1.3

1.3


1.3

1.4

1.0

LINEAR
GAS
VELOC.
FT/SEC
0.25


0.27




0.25


0.25

0.26

0.26


0.26

0.25

0.26

CARBwmrfLET
STREAM
ACID
LOAD.
SAC ID/
*C
0.046


0.039




0.047


0.052

0.007

0.009


0.007

0.011

0.014

SULFUR
LOADING,
«S/IC
THEO.
0.287


0.299




0.293


0.290

0.212

0.217


0.220

0.187

0.296

ACT.
0.226


0.251




0.227


0.221

0.154

0.2CO


0.202

0.199

0.190

GAS OUTLET STREAM
COMPOSITION, VOL. 1.
H2S
3.0


s.s




7.0


7.2

4.4

5.0


5.4

4.0

4.1

S02
0.95


1.0?




0.37


0.31

0.35

0.38


0.33

0.22

1.34

CO?
0.12


0.07




0.08


0.07

0

0


0

0

0

CO
0


0




0


0

0

0


0

0

0

H20
37.4


36.3




34.6


33.7

36.3

36.6


34.1

36.2

33.9

N2
56.0


54 6




53.5


52.7

59.1

59.0


59.2

59.8

60.7

532
EVOL.
; OF
INLET
ACID
10


14




5


4

5

5


4

3

13

H2S
UTIL.
? OF
INLET
89


82




80


74

85

83


81

85

86

STFlTut)
BALANCE
IN
' S
8.8


9,1




3.6


8.7

7.5

7.5


7.6

6.9

8.0

OUT
*S
8.1


8,3




7.7


8.0

6.7

6.9


7.1

6.7

6.8


ACID
DECOMP
V
81


83




79


77

96

95


96

92

94


CONV
TO S
tt
/a
74


84




77


76

92

92


92

106

64


SPACE
VELOC
HR.-l
100


no




100


100

9C

90


90

90

90


PURPOSE
OF RUN
Cir. 18"
to 8'
Dir. 18"
to S"
(Higher
H->S
Stoich.
Ratio)
C Pre"itr
r: 2SO F
& lO'-R.H.
CFrehtr.
P 270'F
& 1C K.I-'.
C Preitr.
C27C:~r
•: IJTR.a.
Increas.
C Tero.
with
InlstGas
Ircreas.
C lenp.
witn C
Frecore.
Decreas.
i2S Cone
Dir. 18"
to S";
Moisture
Le»el
                   *From rotameters
                  **As determined by gas chromatograph.

-------
in Figure 47.  As can be seen, as the inlet carbon tempera-
ture was decreased from 325 to 265°F, the evolution of S02
decreased from about 30% to 1% of the inlet acid.  The
data indicates that the "best" temperature for the inlet
carbon would be in the range of 250 to 270°F.
Effect of Temperature on H2S Utilization-

The effect of the inlet carbon temperature (maximum
reactor temperature) on the utilization of H2S is given
in Figure 48.    The data indicate that as the temperature
increased from 260°F to 325°F, H2S utilization increased
from 75% to 97%.  The condition favoring highest H2S
utilization is the higher reactor temperature.

Effect of Temperature on Per Cent Conversion to Sulfur -

The effect of the inlet carbon temperature on the per cent
conversion to sulfur is shown in Figure 49.    As can be
seen the percent conversion passes through a maximum at
about 285 to 295°F due to a decreased acid evolution as
S02 and then decreases after 285°F because of a decreased
utilization of H2S.

Vertical Temperature Profile -

In all of the runs, the temperature was higher at the top
of the carbon bed and decreased toward the bottom of the
reactor.  For Run SG-90, the carbon temperature varied from
about 190°F at the bottom of the reactor to about 300°F at
the top, for an average of about 240°F.  For Run SG-91, a
decrease of about 15°F at the top of the reactor resulted
in an average decrease in temperature of about 10°F, but
cut the S02 evolution almost in half (from 10 to 6% of
inlet acid).  On the other hand, the utilization of inlet
H2S decreased from 95% for SG-90 to about 82% for SG-91
for the 15 degree decrease in temperature at the top of
the reactor.
Gas Concentration Profiles for H2S and S02 -

The H2S concentration typically decreased uniformly from
the bottom to the top of the bed.  The S02 concentration,
however, was essentially zero until near the top of the
reactor.  The formation of S02 apparently occurred almost
entirely in the top six inches of the carbon bed.  This is
believed to be attributable mainly to the significantly
higher temperature in the upper part of the bed.
                        143

-------
    Figure 47.   Effect of inlet  carbon temperature* on the  evolution  of H2S04
                  as S02 in an 8"  diameter moving bed reactor
   32 ••

-------
      GO
       CVJ
          70 -
          65 -
          60 -
                    Figure  48
                           Effect  of inlet  carbon  temperature on H£S  utilization
                           in an 8"  diameter moving  bed reactor
         100
-P-
Ui
95 -
          90 -•
          85 -•
       -  80 -•
       
10
N
             Carbon:
               Rate
               Type


             Gas:
               Rate •
               Type •
                            28 #  C/hr.
                            Acid  Loaded
                  204  - 280 CFH @ 70°F
                  30%  H2S/Bal. N2
          75 -•
          55
             250
                 260
                        270
280        290        300


Inlet  Carbon Temperature, °F
310
320
330

-------
   104
oo

o
   1C2--



   100



    98--



    96-
 -   Od --
 5   " • ^
'.o
i.
o
a
    92--
=   90-
Q-   83 -r
    86
      250
            Figure 49.   Effect of  the inlet carbon temperature  on the per  cent
                         conversion to sulfur
260
270
 280        250        SCO


inlet Carbon Temperature, CF
                                                                       310
                                                                 320
330

-------
Strategy for Improving Performance
of 8  Diameter Moving Bed

The results of the initial eight runs indicated that a sub-
stantial improvement in performance could be obtained if
the vertical temperature profile was modified so as to
raise the temperature in the bottom of the reactor and
lower it at the top.  The anticipated effects were a
reduction in formation of S02 and an increase in H2S
utilization.  Modifying the temperature profile in this
way required some means of providing additional heat input
to the bed.  To obtain this heat, a finned steam-operated
heat exchanger was installed vertically inside the
reactor.

Finned Steam Heater Experiments -

A series of four runs, SG-93, -94, -95 and -96, was made
with the finned steam heater installed.  The finned heater
was successful in raising the temperature, as
shown in Figure  50.   Reactor performance did not
improve, however, due to carbon flow stagnation on one side
of the reactor, which was caused by the presence of the
heater.  A residence time distribution study clearly showed
flow stagnation and a large deviation from plug flow.  This
is shown in Figure 51.

In these runs the 18" diameter 862 sorber was run con-
currently with the 8" diameter moving bed acid converter.
The four runs included two runs with the carbon from
the sorber routed directly to the 8" diameter unit, and
two runs with the carbon passing from the sorber to the 6"
diameter carbon conditioner and then into the 8" diameter
moving bed.

As seen in Table 33, the results of these runs are signifi-
cantly poorer than the results of the first series of runs.
The H2S utilization is consistently less than 90% and acid
conversion to sulfur averages only 8070.  Formation of S02
is about the same as before.  Further discussion of the
results would not be meaningful in view of the overall
performance reduction which resulted from installation of
the finned heater.

Final Series of Runs -

After determining that the flow problem associated with the
finned heater could not be corrected, the heater was removed
and a final series of-runs was made.  The objective in these
runs was to study the effects of 1) additional heat input
from sources other than an internal heat exchanger, 2) H2S
concentration, and 3) moisture content of the entering
carbon.

                          147

-------
               Figure 50.   Effect  of steam heater on improved carbon heating capabilities
00
         350
      
-------
               Figure 51.
Effect of  heat  exchanger  system  (1-1/2" pipe x 4")  on  carbon
flow in  an 8" diameter moving bed reactor  -  tracer  feed
composition
vo
17

16

15

14

13

12

11

10

 9

 8

 7

 6

 5

 4

 3

 2

 1

 0
                                                Predicted Profile for
                                                Theoretical Plug Flow
                  Virgin Carbon
                               Actual Measured Profile
                               with Heat Exchanger and
                               Exit Pipe Present
                                         *_O	
                             ••« mmm ^^^ ^^ .   •
                              i   i    i   i
  0
                         12  18 22  24  28 32  36  40  44  48  52  56  60 64  68  72  76  80  84

                                                  Time, minutes
                                                               92  96

-------
 Heat Input through Inlet Gas -

 In Run SG-100, an attempt was made to raise the temperature
 in the bottom of the reactor by increasing the inlet gas
 temperature from 300 to 400°F.   This produced a temperature
 rise of about 10 to 20°F in the lower part of the carbon
 bed, but there was no noticeable improvement in acid con-
 version or H2S utilization.

 Heat Input through Carbon Conditioner -

 In Run SG-101,the  carbon conditioner was  operated at 325°F
 in order to raise  the temperature of the  carbon entering
 the 8" diameter unit.   This did not  produce  the expected
 increase in carbon bed temperature,  however,  apparently
 because of heat losses in the transfer piping between the'
 carbon conditioner and acid converter.  Operating the
 carbon conditioner at 325°F required the  use  of pure steam
 as the fluidizing  gas in order  to obtain  the  desired 1570
 relative humidity.   Higher carbon conditioner temperatures
 would not be practical because  it would then  be necessary
 to operate above atmospheric pressure in  order to obtain
 the desired relative humidity.

 Effect of Inlet Carbon Moisture Content -

 In all the experiments in the 8" diameter moving bed,  the
 inlet carbon moisture content was found to be an important
 parameter in acid  conversion.   Variation  of  the moisture
 level had a large  effect on the temperature  at the top of
 the carbon bed,  which apparently was directly related to
 the extent of S02  formation.  At a low moisture level,
 the temperature at  the top of the bed was high and forma-
 tion of S02 was also high.   By  increasing the moisture
 level,  temperature  was lowered  and formation  of S02 was
 reduced.

 The probable explanation is that adsorption  of water vapor
 from the gas phase  onto the carbon occurs unless the carbon
 moisture content is above a certain  level.   Adsorption
 of water is exothermic and causes the observed temperature
 rise.   Formation of SC-2 is the  end result.

 The  results in  Tables  33 and 34 support this  explanation.
 In Table  34,  the carbon conditioner  operating conditions
 are  shown for all the  moving bed runs.  Carbon moisture
 level  is  presented  in  terms of  the sulfuric  acid solution
 concentration,  Ibs.  acid/(lbs.  acid  + Ibs. H20).   A
high acid concentration means a low  moisture  content.
The effect  of moisture level  on temperature  in the top
of the bed  and  on S0£  formation is seen by comparing
Runs SG-94  and  -95 with Runs  SG-100  and -101.   These
runs are  chosen for  comparison  because the average bed


                        150

-------
   Table 34.   EFFECT  OF VOLUME  % H20 AND TEMPERATURE
               ON THE  CONCENTRATION OF ACID  SOLUTION
               SORBED  ON CARBON  FOR CARBON PREHEATER
       Total
     Gas Flow
       Rate,
    CFH @ 70°F
      cid Cone.,
# Acid/(# Acid + #
 Steam
Cone.,
Vol.  %
Relative
Humidity,
      Predicted  Meas
*Carbon direct from 18" diameter S02 sorber to 8" diameter sulfur
 generator.
                            151

-------
 temperatures are all in the 277-283°F range.   In SG-94
 and -95 the relative humidity in the carbon conditioner  was
 3.3% compared to about 14.5% in SG-100 and -101.   These
 conditions produced a low moisture level in SG-94 and -95,
 and a high moisture level in SG-100 and -101.   The results
 in Table 33 show an average S02 formation of 12% in
 SG-94 and -95 compared to 5% in SG-100 and -101,  which is
 a significant difference.

Overall  Reactor  Performance  Summary  -


The results  of the  final moving bed runs showed that  perform-
ance  of  the  8" diameter reactor was adequate for  the  intended
application  as the  acid converter for  the integrated
pilot plant.  Conservatively, the results indicated that
the reactor  could be  expected to perform at the following
levels:

         Acid Decomposition           >95%
         Acid Conversion to  Sulfur    >92%
         Acid Conversion to  S02       < 5%
         H2S Utilization              >85%


Although better  results for  each individual response were
obtained in  some runs, the above performance levels repre-
sent conditions  that  can be  expected with a fair degree
of certainty.

Comparison of Moving  Bed Runs with Design Model -

Based on the rate expression for the  acid decomposition
reaction, a  design  model was derived  for a moving bed
reactor, assuming plug flow  of both the carbon and gas
phases.  The design model was used to  predict teactor
volume for the conditions in the moving bed experiments.
The results  of the  comparison are shown in Table  35.

It is seen that  the predicted reactor  volume in all of
the 8" diameter moving bed runs is less than the  actual
volume by a  factor  of 6.  In view of  the possible  gas
channelling  and  solids flow  problems  inherent in  the
moving bed used, these discrepancies might be expected.

The design model comparison  demonstrates,  therefore,  that
the results  of the  8" diameter moving bed  experiments
unsuitable for purposes of reactor modeling.  This  had
no effect, however, on the adequacy of the  8" diameter
moving bed reactor  to perform the acid conversion step
in the integral pilot plant  runs.
                          152

-------
Table  35.   COMPARISON OF THE MOVING  BED  DESIGN MODEL  WITH EXPERIMENTAL
              SULFUR GENERATOR MOVING BED DATA
RUN
NO.
SG-74
SS-79
SG-99
SG-100
SG-101
SG-102
SG-103
1NT-7
AVG.
TEMP.
°F
270
270
274
282
277
300
265
289
CARBON
RATE,
*/HR.
1.02
0.81
28.0
28.0
28.0
28.0
28.0
29.0
INLET
GAS
RATE,
CFH 8 70°F
9.14
6.20
206
206
206
206
200
228
INLET
yH2S
0.295
0.306
0.328
0.330
0.330
0.329
0.310'
0.272
OUTLET yH2S
ACTUAL
0.0005
0.0018
0.044
0.050
0.054
0.041
0.040
0.0003
CALC.*
0.0565
0.028
0.090
0.091
0.083
0.041
0.098
0.039
INLET ACID
LOADING, f/tC
ACTUAL
0.190
0.190
0.163
0.166
0,163
0.227
0.143
0.220
CALC.**
0.190
0.190
0.155
0.158
0.161
0.198
0.139
0.189
S02
EVOL.,
% INLET
ACID
0
0
4.9
4.8
4.2
12.8
2.8
14.1
OUTLET
ACID
LOAD.,
#/*C
0.010
0.010
0.007
0.009
0.007
0.018
0.011
0.034
MAT'L. BAL.,
LBS. SULFUR
IN
0.286
0.208
7.092
7.153
7.172
7.694
6.447
7.225
OUT
0.257
0.204
6.381
6.683
6.784
6.238
6.432
6.719
BED HEIGHT.
INCHES
ACTUAL
82.9
16.8
72.0
72.0
72.0
72.0
72.0
72.0
CALC.
15.46
13.03
12.39
11.67
12.76
11.54
11.53
11.11
TOTAL VOLUME,
CU. FT.
ACTUAL
O.C848
0.0172
2.094
2.094
2.094
2.094
2.094
2.094
CALC.
0.0158
0.0133
0.366
0.339
0.371
0.336
0.335
0.323
      *Higher outlet H2S concentration used by computer to obtain correct material balance.
     **Inlet acid loading corrected for 502 evolution.

-------
5.2.3  Sulfur Removal

       The sulfur sorbed on the activated carbon has to be
       removed to recover the sulfur values and to regenerate the
       carbon for reuse.  Two basic methods were used to effect
       this sulfur removal.   The first was to vaporize the
       sulfur from the carbon, and other was to extract the
       sulfur.  The bench scale results of each of these methods
       are discussed, followed by a comparison of the two methods.

       Under thermal sulfur recovery, equilibrium data of sulfur
       adsorbed on activated carbon is presented.   The sulfur
       stripping runs made in fluid bed reactors are presented,
       but since it was demonstrated that the unit operations of
       sulfur stripping and of H2S generation could be combined
       into one reactor the operational data on sulfur stripping
       only is more limited than combined operation discussed in
       Section 5.2.4 (H2S Generation).

       Sulfur recovery by solvent extraction is also presented.

       The two methods of sulfur recovery were compared by
       recycle experiments of six carbon cycles.   The data indi-
       cated that sulfur vaporization from carbon was preferred
       because the S02 activity of the carbon was maintained,
       whereas the solvent extracted carbon required further
       treatment.

       As back-up information, a pilot plant was designed for
       sulfur recovery from carbon by solvent extraction.   The
       design and economics  are presented.

       Thermal Stripping Studies -

       Equilibrium adsorption data for sulfur vapor on activated
       carbons were obtained by contacting carbon with a known
       partial pressure of sulfur in a stream of nitrogen followed
       by combustion analysis of the carbon to determine sulfur
       loading.

       The sorbed sulfur was assumed to exist in both physically
       adsorbed and chemisorbed states.   The amount chemisorbed
       was taken as the residual loading after extended purging
       with inert gas,  and assumed to be constant below 1000°F.
       The remaining physically adsorbed portion was found to be
       characterized by a form of the Polanyi-Dubinin adsorption
       equation with respect to equilibrium vapor pressure and
       temperature:
                                154

-------
            In(L-Lc)  -  A - K(T log ^)2                   (43)


    where   L    = total  equilibrium sulfur loading
            Lc  = the amount  chemisorbed
            T    = adsorption  temperature
            PS  = saturation  vapor  pressure of sulfur at  T
            P    = equilibrium sulfur pressure
            A,K  = constants.


Data which fit this equation were obtained over ranges of
temperature, pressure, and loading which are of interest
in the analysis of thermal stripping operations associated
with S02 recovery.

A search of the literature shows that previous activities
in the field have been primarily concerned with sulfur
chemisorption on carbon.  An exception is the work of
Juza and Blanke2 who measured sulfur vapor isotherms
manometrically under static conditions.   Data obtained for
two activated carbons showed evidence of chemisorption,
physical adsorption, and capillary condensation over various
ranges of sulfur loading at temperatures of about 700°F.

In the present work, adsorption measurements were made
dynamically rather than under static conditions in order
to avoid complications due to the production of gaseous
reaction products.

Experimental Results for Equilibrium of Sulfur over Carbon -

Table 36 lists the equilibrium sulfur loadings found  on
carbon at the various experimental carbon temperatures
and sulfur vapor pressures.  These pressures were taken
from the corresponding sulfur temperatures  according  to
data presented in the SULFUR DATA BOOK3.


Inspection of Table  36  shows that at the higher sulfur
loadings, the amount sorbed depends upon both  temperature
and vapor pressure as would be expected in physical
adsorption.  It is seen from the purge data, however,  that
                       155

-------
      Table 36.  EXPERIMENTAL RESULTS OF EQUILIBRIUM
                 SULFUR ADSORPTION MEASUREMENTS
Carbon
Adsorption
Temperature ,
op-
Sulfur
Generator
. Temperature,
»F
Sulfur
Vapor
Pressure,
torr
Equilibrium
Sulfur
Loading,
gms S/100 gins C
Adsorption Data
650
650
800
800
1000
800
1000
800
550
450
550
450
550
380
450
325
38
7.1
38
7.1
38
1.55
7.1
0.36
63.0
47.7
46.6
31.8
24.3
16.2
12.7
10.6
Purge Data
1000
1200
1400
	
0
0
0
7.3
7.1
6.6
there exists at a given temperature a minimum sulfur loading
Figure 52,showing experimental adsorption isotherm points
for 1000° and 800°F, illustrates this further.  As the
equilibrium pressure approaches zero, a residual loading is
retained which probably represents chemisorbed material.
Attempts to describe equilibrium adsorption of sulfur in
terms of sulfur loading, vapor- concentration and carbon
temperatures must, therefore, consider both physically and
chemically bound sulfur.

It has now been found that the temperature dependency of
the experimental sulfur sorption data is well represented
by a form of the Polanyi-Dubinin adsorption, Equation 43.
Figure 53 shows the data of Table 36 plotted in this
form as ln(L - Lc)"vs. (T log PS/P)Z where Lc was assumed
to be constant below 1000°F, the maximum temperature of
the equilibrium data.  Taking Slope K and intercept A
from the straight line plot, the adsorption equation
becomes:
       ln(L-7.3)  =  4.10-0.179(1 log  )  x 10
                                          ~6
(44)
     where  L = gm S/100 gms C
            T = °R.
                           156

-------
Figure 52
Experimental adsorption isotherm points for
sulfur on activated carbon at 800° and 1000°F
                        .p
                103
                        157

-------
Figure 53.  Polanyi-Dubinin plot of sulfur adsorption  data
0.8
                            158

-------
 Using^this equation the partial pressure of sulfur in
 equilibrium with carbon having a given loading can be
 calculated at any temperature up to about 1000°F.   At
 higher temperatures, up to 1400° for example,  better
 results might be obtained by substituting correct  values
 for Lc as determined by extended purging.  Such values are
 noted in Table 36.    In this case it would be necessary to
 recalculate the proper values for constants A and  K.

 For the purposes of analyzing thermal stripping data, it
 may be more convenient to express vapor pressures  in terms
 of sulfur concentration.  Since sulfur exists  in the vapor
 phase as polyatomic molecules in which molecular weight
 depends on temperature, the relationship between pressure
 and concentration is:
     where  % S   = concentration expressed as monatomic
                      sulfur at 1 atm. total pressure
            P     = vapor pressure in torr
= average number of sulfur atoms per
    molecule at the temperature of
                T
                      interest.
A table of values for this latter term vs. temperature is
given in the SULFUR DATA BOOK3.

Figure  54  shows a plot of sulfur concentrations as Si vs.
temperature for various sulfur loadings as calculated from
Equations (45)  and (46).   Such equilibrium lines may be
used, for example, to calculate the minimum number of stages
required to reach a certain residual sulfur loading on
carbon by means of thermal stripping.

From the temperature and pressure dependency of sulfur
adsorption it is possible to calculate the isosteric heat
of adsorption according to the relation:
                       159

-------
   Figure 54.   Equilibrium  lines  for concentration vs.
                 temperature  at various loadings
0.02
0.01
                                                     30 gms S/100 gms C
                                                     24 gms S/100 gms C
                                                     18 gms S/100 gms C
                                                     12 gms S/100 gms C
       600
700        800       900

                •
     Temperature, °F
1000
                              160

-------
     where  PI
            P2
            q
equilibrium pressure at temperature TI
pressure at T2
differential heat of adsorption evaluated
  at a particular loading.
The calculated heats of physical adsorption are given in
Table  37  for various total loadings.
       Table 37.  ISOSTERIC HEATS OF ADSORPTION
                  OF SULFUR VAPOR ON CARBON
Total Load
^ms S/100 gms C
12
18
24
q
Kcal/mol
24.7
23.0
21.9
q
BTU/lb.
223
208
198
These heats of adsorption which must be supplied during
stripping may be  compared to the heat of vaporization of
bulk sulfur of 134 BTU/lb. at 1000°F.  The relative magni-
tudes of these heats are reasonable for systems involving
physical adsorption.

Comparison of the experimental sulfur adsorption isotherm
results obtained  here with those of Juza and Blanke noted
previously was made by means of Polanyi-Dubinin type plots
Their data obtained near 700°F showed similarities to the
present data within the limits which might be expected to
result from variations  in carbon type.   However,  the
temperature dependency was not properly described by the
P-D relation since the few data reported for higher
temperatures did  not fall on a single curve with those at
700°F.  The reason for disagreement is not clear although
it is possible that small amounts  of gaseous reaction
products formed at the higher temperatures could have
caused errors in  the static sulfur pressure measurements
used in this work.
                        161

-------
Fluid Bed Sulfur Stripping  -

A trial  sulfur  stripping run was made at an average column
temperature of  approximately 1050°F in an 8 stage 4"0
fluid bed regenerator.  The results are" summarized in
Table 38 below.
    Table  38.  SULFUR STRIPPING IN A CONTINUOUS
               8-STAGE FLUID BED
Run
No.
SIS-1
Carbon
Rate
Ibs./hr.
- 31
Temp.
OF
(Avg.)
1048
S Loading
Ib./lb. C
In
0.307
Out
.094
Sulfur
Removal
%
66*
      *The removal of 66% of the adsorbed  sulfur
       in the 11 minute  residence time  compared to
       60% obtained previously  in a  three  stage
       fluid bed unit.
A McCabe-Thiele analysis, assuming 100% stage efficiency,
indicated six and a fraction theoretical stages were
required to achieve the observed results from the trial
run in the 8 stage column.  To facilitate the stage effici-
ency calculation a computer program was written.  The
program executes an iterative search procedure which con-
verges to an average value of the Murphree tray efficiency
over the entire column.  For the trial stripping run an
efficiency of 79% was calculated.  Assuming the data from
future sulfur stripping runs yield similar values for the
stage efficiency, then the value that is obtained should
be a valuable piece of information in designing a sulfur
stripper for a particular application.

Bench Scale Sulfur Stripping/H2S Generation -

Two stripping runs were made at 1000°F in the batch 4"
electrically heated fluid bed to determine the effect of
contact time (space velocity) on the approach to equilib-
rium in sulfur stripping.  Following the stripping of the
physically sorbed sulfur, removal of chemisorbed sulfur
was investigated as a function of the concentration of
the hydrogen reductant.  The results of these runs are
shown graphically in Figure 55.  It appears that the
                       162

-------
                Figure  55,
                   4"  diameter  batch fluidized bed sulfur  stripping and hydrogen
                   desulfurization  runs  (Runs JFC-1  and  -2)
              24
u>
              20
              16
c
o
£
3 12
c
o
s-
            3
            CO
                           JFC-2
                        "26,500 hr.'1 S. V,
                   
-------
stripping is most likely equilibrium limited and
independent of space velocity over the range investigated.
Conditions run were 2 ft./sec. with 4 inch and 6 inch
expanded bed depths at 1000°F.  The space velocities corre-
sponding to the corresponding settled bed depths were
38,400 v/v/hr. (4" expanded bed) and 26,500 v/v/hr. (6"
expanded bed). If the stripping were equilibrium limited,
the rate of sulfur removal at the lower space velocity
would be equal to that at the higher space velocity.
Since total sulfur removed is equal to the product of the
bed weight and the change in sulfur content, sulfur removal
as a function of time can be calculated at each space
velocity.  Over the straight line portion of the curve,
the total amount of sulfur removed is the same for the two
runs.

The rate of removal of chemisorbed sulfur appeared to be
zero order with respect to hydrogen concentration within
the accuracy of the analyses.  Run JFC-1 was made with 33%
H2 and Run JFC-2 was made with 17% H2.   A constant sulfur
content of 3.5% was attained in both cases after 45
minutes' exposure to hydrogen.

Solvent Extraction Studies -

An experimental program was completed to obtain design
information for the evaluation of extraction systems.
Measurements of sorption rate, pore volume, surface area,
and residual sulfur content were made on carbons previously
loaded with 14 wt. % sulfur and extracted with (NH4)2C, /
CS2, xylene or ether.  The dependence of sulfur loading on
sulfur removal was investigated by batch extractions of
carbon loaded with 14 wt. % sulfur.  This work progressed
to the point of carrying out ten-stage extractions with
CS2 at 25°C, 15 wt.  % (NH4)£S at 40°C, and xylene'at 105°C.
These experiments were to determine the extraction
behavior of the sulfur deposited on the carbon.
Physically sorbed sulfur would be expected to be removed
quantitatively but with increasing difficulty for sulfur
in smaller pores.  Chemisorbed sulfur removal would not
be expected under extraction conditions.

It was also necessary to measure the S02 sorption rates
of these extracted carbons and of hydrogen treated  carbons.
It might be expected that relative activity will be more
dependent on the removal of chemisorbed sulfur than physic-
ally adsorbed sulfur.  Several pairs of extracted samples
have been subjected to high temperature (1000°F) purges,
one sample with pure helium, the other with 30% hydrogen
in the helium.  The high temperature purge was an attempt
                        164

-------
to restore the activity of extracted carbon to the level
of virgin carbon.  Ether extraction was performed to
determine if a sulfur-free extraction solvent, which could
be removed from the carbon at low temperatures,  would
produce an extracted carbon retaining more of its initial
S02 sorption activity than the other solvents tested.
This was primarily aimed at gaining insight into tempera-
ture and solvent deactivation effects, not at any potential
commercial use of ether as an extraction solvent.

A literature survey on the removal of sulfur by extraction
was made.  The results of the survey are given in detail
in Appendix A-17.  Included are sulfur-solvent equilibrium
data for ammonium sulfide, carbon disulfide, xylene,
benzene, toluene, and pitch oil.

Ammonium Sulfide Extractions -

After a series of bench scale extractions, it was decided
to run the ammonium sulfide extraction in the closed
circulating system (see Section 5.1.7) under an inert
atmosphere to avoid any possibility of oxidation of
(NH4)2S to sulfur.  It was found that both carbon disul-
fide and ammonium sulfide decompose to some extent to pro-
duce sulfur when contacted with virgin activated carbon.
This effect was observed even when the carbon had been
pretreated with N2 at 1800°F to remove oxygen, Table 39.
   Table  39.  EFFECT OF SOLVENT ON VIRGIN CARBON
Virgin Carbon C-70-77
Inert
Gas
NO
NO
He
N2
Carbon
Degassed
No
No
No
1800°F
1 hr.
Solvent
Wash
IxCSz
30 min.
lx(NH4)2S
30 min.
lx(NH4)2S
30 min.
lx(NH4)2S
60 min.
Steamed
No
No
1 .5 hrs.
200°F
8 hrs.
250-275°F
0.4% S
2.2% S
4.0% S
1.7% S
1 .8% S
                        165

-------
  A ten stage extraction was  run in the closed, circulating
  system.  The sulfur  level was reduced to 2.08% after  the
  final extraction, washing and steaming.  The equilibrium
  for this system is shown in Figure 56.
   Figure 56.
Extraction of  sulfur  loaded activated
carbon with 15 weight % (NH4)2S
solution at 40°C
                                  Numbers refer to
                                  successive contacts  with
                                  fresh 15% (NH4)2S
.001
                  9 s/g Carbon (Solvent Free)
                         166

-------
Carbon Bisulfide Extraction -

Carbon disulfide extractions were carried out to determine
if carbon disulfide would be superior to ammonium sulfide
as an extractant.   As shown in" Table 39, carbon disulfide
contact with virgin carbon was found to have a higher
level of residual sulfur (2.2% sulfur) on the carbon than
did ammonium sulfide (1.8% sulfur).  The decomposition
of CS2 on activated carbon has been enountered previously
in the operating experience of textile companies.
These companies have, for many years, used large scale
activated carbon units for carbon disulfide recovery and air
purification.  The adsorbed carbon disulfide tends to be
hydrolyzed to a small degree, releasing hydrogen sulfide
which is oxidized by the air to sulfur.  Since this sulfur
builds up on the carbon and reduces sorption efficiency,
steps are taken to remove it.  In the early carbon processes
recovering carbon disulfide, this sulfur contaminant was
removed by extraction with aqueous sodium sulfide; but
present fluid bed recovery plants remove this by continu-
ously stripping a slip stream of the circulating carbon
with hot inert gases.

The carbon disulfide extractions were all run on the bench
and closely paralleled the results of the ammonium sulfide
extraction.  The ten stage extraction resulted in a
residual sulfur level of 2.6% on the carbon.  The equilib-
rium data for this system are shown in Figure 57.

Xylene Extractions -

A series of bench scale extractions with xylene were
carried out in an attempt to obtain lower sulfur levels.
The extraction was carried out in five stages, the carbon
oven dried and the extraction continued for five addi-
tional stages.  The sulfur level was found to be nearly
independent of sulfur concentration in the solvent.  It
may be suspected that xylene would be strongly sorbed in
the smallest pores and that the residual sulfur would
also be concentrated there.  A carbon loaded with 14%
sulfur was taken to 3.7 g. S/g. carbon after a single
contact with xylene at 105°C.  Four subsequent contacts
for a total of five reduced it to 2,9 g. S/g. carbon.  At
this point, the carbon was dried in an attempt to bring
to the carbon surface a portion of the sulfur contained
in  the  xylene trapped in the smaller pores.  This was
indicated by both an increase in sulfur concentration
in the xylene and a sharp drop  in sulfur concentration
on the carbon in the sixth extraction.  Subsequent extrac-
tions followed the pattern of the first five stages.
After the tenth stage, the residual sulfur  level was
reduced to 1.5 g. S/g. carbon.  The equilibrium  for this
system are shown in Figure 58.


                       167

-------
  Figure 57.
Extraction of  sulfur loaded activated
carbon with CS2 at 25°C
 .03
 .02
.0001
                      .04       .06        .08

                   g. Sulfur/g. Solvent Free Carbon
                                    .10
.12
                             168

-------
Figure 58
                Extraction of sulfur  loaded activated

                carbon with xylene at 105 °C
I
r"
O
to
cr.

O>
• vv
.04
.02
.01 •
.008
.006
.004
.003
/\f\O
• \)\)c*
.001
.0008
.0005
.0004
.0003
.000?









S
'£
to
c
•r-
C
*l —
r—

1








	 _




O 6



t



D°
8
1
i
9
J,.









c
?'
1
1
r
1
i
i
1
1
' ^


5








3



' NOTTS:
1. Numbers refer to succes-
siv
xyl
2. Car
and
3. Sta
FHS
Rof
e contacts w
ene.
bon dried be
6th stages.
rting carbon
-9 - [112 - 1
. I AS 3175:1
ith rresh
two en Gth
sample
«', Sulfur ~
0-11
              '.01        .02        .03      .01


                g. Sulfur/g. Solvent-Free Carbon
                                                        .05
                            169

-------
Consideration must be given also to the implications of
xylene extraction on the regeneration sequence in a stack
gas treating process.  Prior applications of xylene extrac-
tion have been in applications where trace amounts of
xylene in the treated gas were not a serious drawback.  Our
experience in solvent emission control indicates that a
final stripping of xylene extracted carbon at 500 to 600°F
would likely be required to minimize xylene loss into the
circulating carbon stream and subsequent transfer to the
flue gas.

Comparative S0£ Sorption Rate of
Solvent Extracted Carbon Samples -

Measurements of the relative S02 adsorption rates were
made on various extracted carbon samples and carbon samples
of related interest.  These measurements were made in the
differential rate apparatus as described in Section 5.1.1

It appears that the activity of carbon for S02 adsorption
is not a simple function of residual sulfur.  One would
conclude that there are differences between treat-
ments required to restore activity for H2S pickup and that
required for S02 pickup.  Literature data indicate that
extractions with (NH4)2S and CS2 as performed here would
have restored most of the carbon's activity for H2S
oxidation.  In H2S oxidation, the carbon may also be loaded
with an equal weight of sulfur and still pick up H2S at a
satisfactory rate.

Solvent extraction of sulfur from carbon decreases the
carbon's activity for S02 sorption.  The relative activi-
ties of sulfur loaded carbons extracted in 10 stages with
xylene, (NH4)2S and CS2 and with ether in the Soxhlet
Extractor were measured on the sorption apparatus.
Further treatment of these samples included thermal strip-
ping under helium and under a mixture of 3070 hydrogen and
70% helium.  The results of these measurements are given
in Table 40  along with surface area and pore volume
distribution comparisons.

The various solvent extracted samples were thermally
stripped with an inert gas when it became apparent that  the
residual xylene on the carbon was interfering with the S02
activity.   Treatment at 555°F did not remove all the xylene
and it was necessary to go to 1000 F to remove all the
solvent.  With this increase in activity, the  (NH4>2S and
CS2 extracted samples were similarly stripped and these
activities also improved.
                       170

-------
             Table  40.   COMPARISON OF  S02 ACTIVITY AND  SURFACE AREA
                         AND PORE VOLUME MEASUREMENTS
1
Sa=ple
3C7-1
ETH-1
SXA-1
iXA-2
ZZA-3
ZXA-1.
ZZA-5
2X3-3
EX^i
122-5
j ZXC-1
EXC-2
J3C-3
i
I 3C-1
i
Material
Virgin Carbon
Sulfur Loaded Carbon
Sulfur Loaded Carbon
Sulfur Loaded Careen
Sulfur Loaded Carbon
Sulfur Loaded Carbon
Sulfur Loaded Carbon
Sulfur Loaded Carbon
Sulfur Loaded Carbon
Sulfur Leaded Carbon
Sulfur Leaded Carbon
Sulfur Leaded Carbon
Sulfur Loaded Carbon
Sulfur Loaded Carbon
L_ 	 .
Treatment
Solvent
None
Kone
IC-Stage Xylene
10-Stage Xylene
10-Stage Xylene
10-Stage Xylene
(2nd Sun)
10-Stage Xylene
(2nd Run)
10-Stage (NHj,)2S
10-Stage (Iffii^gS
10-Etage (1IH^)2S
10-Stage CS2
10-Stage CS2
10-Stage CS2
Staer (Soxhlet)
Stripping
Hone
None
—
555°F
1000°F
1000°F
1000°F
—
1000°F
1000°F
1000°F
1000°F
—
Stripping
Gas
Kone
Ncne
—
Helium
Eelium
EeliuB
30? Eydrogen
70? Eelium
—
Helium
30? Hydrogen
70? Helium
Helium
30? Eydrogen
70? Keliua
—
? Activity '•% Surface Area
Coapared to '• Compared to
Virgin Carbon Vireir Carbon
100.0
	
7.5
9.1*
1.8.0
1.6.0
1.1.9
56.1
78.6
77.6
1*5.1;
63.0
57.2
1*6.1
100.0
1<5.3
It It. 8*
20.8
80.9
98.9
93.0
96.6
81*. 5
90.9
101.9
91-6
93.6
19-7
% Pore Volume Coarared to Virgin Carbon
?otal Pores
r<500^- Radius
100.0
1.7.7
1*7.9
21.. 3
81.0
'99-7
91*. o
98.1
83.3
91.1
102.2
91.2
92.8
23.3
Pores
100
-------
 From earlier work  on  hydrogen  treatment  of  regenerated
 samples, we found  a return  of  activity for  SO? when  the
 sample was stripped at  1000°F  with  33/0 H2 in helium.
 In  addition, a  similar  sample  treated at 1500°F  increased
 its activity 65% over the virgin material.  In view  of
 these earlier results,  we decided to run parallel extracted
 samples  in which 30 volume  70 H2 was added to the helium
 purge gas at 1000°F.  The activities remained essentially
 unchanged with  H2  added to  the purge gas. However, the pore
 volume,  surface area  and residual sulfur measurements indi-
 cate that the bench scale stripping conditions chosen were
 not sufficient  to  give  good sulfur removal  from  the  smallest
 pores.   The hydrogen  flow was  much  lower than used in earlier
 work on  hydrogen regeneration.  Replicate runs will  be made
 at  higher purge rates to allow further evaluation of the
 solvent  extracted  carbon's  ability to be restored to its
 original activity  by  hydrogen  treatment.

 Conclusions which  may be drawn from the  data at this
 point are as follows.   Xylene  appears to be an inferior
 extraction solvent for  sulfur, due to the difficulty in
 removing it from carbon and some residual solvent deacti-
 vation effect which persists,  even through  a high tempera-
 ture hydrogen treatment.  Ammonium sulfide  and carbon
 disulfide appear to be  close in effectiveness.  However,
 the ammonium sulfide  appears to have a lesser residual
 solvent  deactivation  effect when the carbon is given
 either a high temperature treatment with inert gas or
 hydrogen.

 Indications are that  ether  has no less of a solvent  deacti-
vation effect on carbon  than does carbon disulfide,  and
 the solvent deactivation effect is not strongly  influenced
 by  the presence of sulfur in the solvent structure.

While it appears that a  return of surface area or pore
volume will not guarantee a return of activity for S02, a
significant loss of either  will indicate a  loss  in activity
for  S02-  Similarly,  high residual  sulfur content will
tend  to cause activity  loss, but poor activity can persist
even  at low residual  sulfur levels.

 The indication  that iso-thermal regeneration is  as effective
 as  high  temperature regeneration is somewhat clouded by  the
mild  regeneration  conditions.  As one attempts to return
carbon to its original  activity, chemisorbed oxygen  prevents
this unless higher temperatures are used.   Hydrogen  is
believed to remove the  chemisorbed  oxygen at lower
temperatures.   Chemisorbed  oxygen also has  a cumulative
effect under recycle  conditions.
                       172

-------
Recycle Work - Extraction Regeneration
vs.  Thermal/Reduction Regeneration     -

Two samples of carbon were recycled through the adsorp-
tion and regeneration steps to allow a direct compari-
son of the isothermal (extraction) and thermal regeneration
sequences.  The experiments also showed the effect of
recycle on the S02 activity of the carbon for each of the
two regeneration sequences.  The results of the study
were:  1) that the S02 activity of thermal/reductive
regenerated carbon leveled off at about 92% of the activity
of the virgin carbon, and 2) that the S02 activity of the
isothermally [(NH4)2S] regenerated carbon decreased gradu-
ally with each cycle and after 6 cycles was at about 34%
of the activity of the virgin carbon.  Pore volume meas-
urements showed a decrease in the total pore volume and
surface area of the isothermally regenerated sample and
an increase in total pore volume and surface area of the
thermally regenerated sample.  The S02 sorption differen-
tial rate apparatus was shown to be useful in assessing
various treatments of the (NH4)2S extracted sample after
six cycles.

Details of the recycle experiments are given in Appendix
A-15.

Experimental Results and Discussion -

A recycle series was run to compare the isothermal
(extraction) and thermal/reductive regeneration sequences.
Sample A will be referred to as  the isothermally regene-
rated carbon and Sample B referred to as the thermally
regenerated carbon.  Samples A and B were loaded with acid,
reacted with H2S to form sulfur  on the carbon, and then
regenerated by (NH4)2S extraction or thermally with H2,
respectively.  The sequence in the cycling was as follows:
        A.   Isothermal           B.   Thermal/Reduct ive

   1.   S02  sorption                1.   S02  sorption

   2.   Sulfur generation          2.   Sulfur generation

   3.   Sulfur extraction          3.   High temperature
       with  (NH4)2S  followed          H2~S reaction on
       by  steaming                     carbon
                        173

-------
Each  sample was then passed through successive  complete
cycles.   The S02 sorption step on each cycle  provided a
means of comparing S02 activities to the virgin material.
As  a  basis for comparing S02 activity, the  acid  loading at
210 minutes relative to the virgin material was defined
as  the S02 activity.
 SOo Activitv = Acid Load, at 210 Min. for Sample A or B for Cycle i
  2       y       Acid Load, at 210 Min. for Virgin Carbon
 The rate of S02 sorption is a function  of temperature.
 Variation in the average sorption  temperature for
 any cycle was corrected back to average temperature  tor
 loading  the virgin material with acid.  The Westvaco equa-
 tion was used as a guide to make this correction,
TF, Temperature Correction Factor,  = 552°-(TY*rgin ~ Tcycle i)     (43)
                                    Tvirgin TCyCie i

     where   Tvirgin  = sorption temperature at which
                          virgin material  is loaded, °R
             Tcycle i = sorption temperature at which
                          sample is being  loaded in  cycle  i
 The  sulfur remaining on the carbon after  isothermal or thermal
 treatment varied between 2 and 470 S.   In  order  to compare
 S02  activities  at comparable sulfur loadings,the  S02
 activity  was  corrected to a common sulfur level of  270.
 Previous  work,  in which varying amounts of sulfur were
 sorbed on activated carbon, was used to make this sulfur
 correction.   The results of the work are  shown  in Figure
 59.    The sulfur correction factor, Sp,  is given as
     SF, S Correction Factor to 2% S Basis, =  Activity at % Si
                                         Activity at 2% S

                                         Activity at % Sj
                                              0.909
The recycle work  with the respective correction  factors is
summarized in Table   41.    The final S02 activity is calcu
lated by  multiplying the uncorrected  S02  activity by
                             174

-------
     Figure 59
Effect of percent sulfur on carbon
on  the S02 activity
   1.0
   0.95.
r:
o
o

c
   0.85- •
•r-
4->
to
•£
•r-
2  0.75--
O
10
   0.65
                               3        4


                              % S on Carbon
                               175

-------
     Table  41.   EFFECT OF  RECYCLE ON S02 ACTIVITY  FOR  ISOTHERMAL

                   AND THERMAL/REDUCTIVE REGENERATIONS

•v-

'•~~er

Sul
A
~ r* o -*
210 :•:
g=s Hasc
furic
eid
ir.£ at
i-utes,
I;/— Car con

Un corrected
SCj
Act^v-ty






3XT-S-IA"
-2A
-3A**
,

-5A
-6A

i
0.155 i 1
0.1^-0
0.1^5

0.1^^
0.136
0.140

0.903
0-935
^
0.929
0.877
0.903

2C
C.'^.
1C
«. s

2C
2C
IS

Arr:oni^- S'^lfide Sxtre
5XT-7-1A***
-2^.
-3A
-UA
-5A
-6A

0.167
0.077
0.090- /
0.072
0.05^
0.062
J.
0.461
•0.539
0.431
0.323
0.371
i
2C
2C
2C
2C
2C
2C

Avera-o 5"-? • I
C|r--v— s- -
-
-.-
j. CIT."0 S T 2,'w "J^T S


,-^__=^^.-,_
'-0-s-
*-*^t_ A <4 • J
Factor
Sulfur
Correction
Factor to
£'" S
Bisis

SO
Activity


riegensrc-vxcns
^
o
* _>
.8
.2
.5
. 2

|
|
i (^ '.
-L j C.-
1.0233 i 2.6
- ^ _ ^
0.9^39
0-9766
0.5515

? •?
2.6
3.2
3.6

0-909
l.C^l
. 1.0-32
1.0~5
1.077
1.181

0.909
0-962
0.905
0.915
0.922
0.9^0

:ion Regenerations
.0
i
, Lf
^ T
^
* s *
.5
.2
1
0-9679
0-9535'
1.0063
0.9956
0.9775
0.4
U.I
3.1
2.1
2.9
2.0
0.909
1.143
1.072
1.053
1.057
1
0.909
0.510
0.563
0.457
0.339
0.363
       *Virgin carbon is -!A;and -6- signifies therzal regeneraticr..


      **H\sidifier teipsrature vas low, correction factor for lover moisture content using Westvaco

Bc-uation = 1.0712.
             »    "^

     ss*'/ir£in carbon is -1A; and -7- signifies isothermal regeneration.

-------
the two correction factors for temperature and sulfur
level.  The S02 activity is shown as a function of the
cycle number in Figure 60.  As can be seen from Figure 60
and Table 41, for the thermal/reductive sequence the S02
activity of the carbon levels off somewhere near 92?0 of
the virgin carbon activity.  The virgin carbon activity
was corrected to the 2% sulfur level.  The S02 activity
of the carbon, which was regenerated isothermally using
(NH4)2S, decreased from about 6070 to 34% of the original
carbon activity after 6 cycles.  It is obvious that this
rapid decrease in S02 activity is an undesirable aspect
of extraction.  Therefore after 6 cycles the recycle
experiments were halted to attempt to find an appropriate
treatment of the solvent extracted carbon which would
return its activity.

Pore volume distributions were run on sixth cycle Samples
A and B to gain a possible insight into the deactivation
process.  The results are given in Table 42.  It can be
seen that there is a definite decrease in the total pore
volume and surface area of the sixth cycle isothermally
regenerated (extracted) carbon (RXT-7-6C) and an increase
in the total pore volume and surface area of the sixth
cycle thermally regenerated carbon (RXT-6-6C).  The pore
volume and surface area measurements have seemed to indi-
cate only gross effects in S02 activity reduction.  That
is, a significant drop in pore volume or surface area
definitely results in a decrease in S02 activity.  Small
changes, however, do not necessarily indicate a decrease
in S02 activity.

As mentioned above, the recycle experiments were stopped
after 6 cycles to assess an appropriate treatment of the
(NH4)2S extracted carbon to restore its activity.  Since
there was only about 30 grams of this sixth cycle carbon,
it was desirable to use as small a quantity of the sample
as possible, but a large enough sample to assess each
treatment experiment.  It was decided to use the S02
sorption differential rate apparatus discussed in previous
sections to measure S02 activity, as only 0.1 g. samples
are required.  As a check on the fixed bed S02 sorber,
two each of the thermally regenerated (KXT-6-) and iso-
thermally regenerated (RXT-7-) carbon cycle samples were
used.  The results of the S02 activity found for each
of the corresponding runs in Table 41 are given  in
Table 43.  As can be ssen from the table, reasonably  good
agreement of the S02 activity was obtained from  the
integral rate determined in either apparatus.
                        177

-------
Figure 60.   Effect of recycle on S02  ability for  isothermal
              and thermal/reductive regenerations*
        c
        o
        CJ

        c
        >

        4->
        TO
        >
       /•i—
        4->
        U
        
-------
Table 42.  PORE VOLUME DISTRIBUTION RESULTS USING ENGELHARD
           ISORPTA APPARATUS
Sample
EXV-1
(Virgin Carbon)
RXT-6-6C
(Thermal)
RXT-7-6C
(Extracted)
7o Pore Vol. Compared to Virgin Carbon
Total Pores
<500A
100
118
93.8
Pores
100-500A
100
77.1
67.0
/
Pores
10-100A
100
114
95.4
Pores
<10A
100
121.0
94.2
70 Surface Area
Compared to
Virgin Carbon
100
119.0
92.6
   Table 43.  S02 ACTIVITIES INTEGRAL RATE DETERMINED
              USING DIFFERENTIAL RATE APPARATUS VERSUS
              USING FIXED BED
Run
^•^^^^^^•-^••••••••M—l^— •M*«v^«MMM«
RXT-6-4A
RXT-6-6A
RXT-7-4A
RXT-7-6A
S02 Activity for
Differential Bed*
^^^^B^^^mmrH**^^^^***—*^^^*^^'^^**^******'^*^^^"*******^*"1**'^^"^^*'^^***
0.834
0.818
0.426
0.397
S02 Activity**
Differential Bed
0.866
0.966
0.449
0.397
Fixed Bed
0.915
0.940
0.457
0.363
   *Uncorrected to 2% S basis.
  **Corrected to 2% S basis.
                           179

-------
 Treatments  To Restore  S02  Activity
 of Solvent  Extracted Carbon	 -

 The  sixth cycle  carbon sample  from the  isothermal regene-
 ration  recycle studies which had  about  34% of  its original
 activity was  put through a number of post treatments, which
 were assessed as to  their  success in restoring  the S02  sorp-
 tion activity.   The  SC>2 sorption  differential  rate apparatus
 was  used to make the evaluation of the  success  of each  of
 the  post treatments.   The  experiments included both iso-
 thermal and thermal  treatments.   The runs, respective post
 treatments, and  activity measurements are summarized in
 Table 44.

 From the data given  in Table   44   it can be seen that the
 post treatment experiments included the determination of:

      1)  Isothermal  NH40H  treatment to remove  the
            sulfate or  other possible deactivation
            species
      2)  Effect  of treatment time with H2
      3)  Effect  of treatment temperature

      4)  Effect  of using heat  only

      5)  Effect  of using CO as the reductant.

 The  isothermal treatment with  NH40H was unsuccessful in
 returning the S02 activity of  the solvent extracted carbon.
 In fact, some further  deactivation was noted.   For compari-
 son  purposes  virgin  carbon showed a drop in activity to
 60%  of  its  original  value  upon being treated with the
 NH40H experimental step.

 The  effect  of treatment time using H2 at 1000°F is also
 shown in Figure  61.     The increase in the SC>2  activity is
 less  rapid with  increasing treatment time.  It  appears  that
 the  optimum time in  a  fixed bed would be 4 to  6 hours.

 The  effect  of temperature  using !!£ for a 4 hour treatment
 is also shown in Figure 62.   The S02 activity of the
 (NH4>2S extracted sample increases rapidly with treatment
 temperatures  from 800  to 1200°F.   SO? activities of more
 than  130% of  the original  value can be obtained at 1200°F
with  a  4 hour treatment time.  It appears that the lower
 limit of temperature for H2 treatment might be near 800°F.

 It is seen from  the  experiment using N2 at 1000°F that
although heating is  giving some increase in S02 activity,
 the H2  is playing a  definite added role of reactivation.
The N2  treated sample  at 1000°F had an  S02 activity of
                           180

-------
                 Table 44.   EFFECT  OF POST TREATMENTS OF  (NH4)2S EXTRACTED SIXTH CYCLE
                             CARBON  ON S02 ACTIVITY
Eun
Kister
16C-G
197-G
191-G
189-G
193-G
192-G
191-G
195-G
196-G
198-G
Sample
Virgin
RXT-7-6C
HXT-7-6C-U
HXT-7-6C-1
RXT-7-6C-6
RXT-7-6C-5
RXT-7-6C-7
RXT-7-6C-3
KXT-7-6C-9
Virgin
Treatment
Gas
„

305? H2
305? H2
30/5 K2
30/1 K2
K2
30? CO
°F '


•800
1000
1000
1200
1000
1000
Tine,
hrs. .


U
2
U
h
h
k
2% NH1,OH;
Purged l.-hr. with N2 at UOO°F
. 2% KKj-OK;-
Purged 1 hr. vith N2 at UOO°F
H2£0!4
Loading**
nss. Acid/em . C
173-90
63-17
116.69
16C.37
I81t.95
23«..'*9
88.37
131.29
58.56
119-32
Uncorrected
SC2"
Activity
l.CCC
0.363
0.671
0.925
1.06U
1.3.3
0.508
0.755
0.337
0.636
^r
o.u
2-1
0.8
0.9
0.8
0.8
0.6
1.0
l.U
0-7
Sulfur
Correction
Factor*
0.988
1.076
1.000
1.005
1.000
1.000
0.988
1.005
,.03,
0.995
502 Activity
Corrected to
0.8:* S Basis
0.983
0.391
0,671
0-929
1.06U
1.31:3
0.502
0.759
..*.
0.655
00
               *Using relationship developed of effect of sulfur on S02 activity.


              **To 210 minutes sorption time.

-------
Figure 61
Effect of treatment  time using hydrogen post
treatment of  (NH4)2S extracted sample of
sixth cycle
                                Temperature:
                                Hydrogen Concentration:
                                Carbon:
                                       iono°F
                                       30 volume %
                                       RXT-7-6C
   0.2
                           Treatment Time, hours
                             182

-------
Figure 62
    /; ?f  temPerature  of hydrogen post treatment
ot  (NH4)2S extracted sample of sixth cycle
                                  Treatment Time:        4 Hours
                                  Hydrogen Concentration:  30 volume
                                  Carbon:               RXT-7-6C
                                          I
                          600             800
                             Treatment Temperature, °F
                                         1000
1200
                                 183

-------
       about  50%,  whereas  the H£  treated sample  at  1000°F  had
       about  90%, of its  original  activity of the virgin  carbon.
       The  CO treated sample  had  an activity of  about  75%  which
       lies somewhere between the two  extremes of thermal  treat-
       ment only and thermal/reductive H2-   This gives even more
       support to the definite beneficial effect H£ has  on
       returning the activity of  activated carbon.

5.2.4  H2S  Generation

       Hydrogen is  the basic  reductant that  is used to reduce the
       sulfuric acid sorbed on the  carbon.   It was  found,  however,
       that a secondary  reductant produced  from  the hydrogen,
       hydrogen sulfide, is a better reductant of the  sulfuric
       acid.   Sulfuric acid reduction  was  found  to  occur in the
       temperature  range of 200 to  300°F,  as  discussed in  a
       previous section, so process development  proceeded  toward
       producing this  recycled reductant within  the process.  The
       initial  process concept was  to  vaporize the  sulfur, origi-
       nating from  the S02 in the flue gas,  from the carbon.
       The  remaining recycled sulfur on the  carbon  was then
       reacted  with hydrogen  to form the hydrogen sulfide  needed
       to reduce the sulfuric acid  to  elemental  sulfur.  The
       major  deviation from that  initial concept is  that the two
       unit operations of sulfur  stripping  and H2S  generation
       have been combined successfully into  a single fluid bed
       reactor  as discussed in a  later section.

       In this  section,the bench  scale data  that was developed on
       the  kinetics  of H2S formation from hydrogen  and sulfur vapor
       over a catalyst of activated carbon  is discussed.    The pilot
       work that was  done on  this unit operation is  discussed in the
       later  section on  combined  sulfur stripping/H2S  generation.

       In addition  to  the use  of  hydrogen for sulfuric acid reduc-
       tion by way  of  hydrogen sulfide, other possibilities of
       hydrogen  requirements  are  hydrogen to  remove chemisorbed
       oxygen from  the reused activated carbon and,  hydrogen
       chemisorption by  the regenerated activated carbon.
       The  chemisorbed 02 is  discussed under  the integral  run
       results.  The  studies  of H2  chemisorption on activated carbon
       are  discussed in  this  section.

       Fixed  Bed Studies H2 Chemisorption -

       The  objectives  of the  study were to  ascertain the extent of
       H2 chemisorption on carbon and  the  effect of cycling the
       carbon.   The  general approach was  to evaluate H2  chemisorp-
       tion by  following the  gas  phase analysis  of  H2  during
       exposure  of  virgin carbon  in a  fixed bed.
                             184

-------
The next step, not initiated, was to start a recycle bench
scale experiment  in which virgin carbon is loaded with S02,
reacted to form S, then exposed to H2 to convert the
sulrur to H2S and to measure the H£ chemisorbed.  During
the runs the off-gas from each process step would be
analyzed by gas chromatograph.  The carbon at the end of
each cycle would be analyzed for total sulfur content and
tor S02 activity  in a differential S02 sorber as a check
on the S02 sorption step.  It was anticipated initially to
complete six cycles.  This second part of the study was not
completed because of the delay that would follow in complet-
ing the integral  run.

The major response was the amount of H2 chemisorbed either
on virgin carbon  or on recycled carbon.  Secondary responses
were the S02 activity of recycled carbon and the effect of
recycle on conversion of sorbed H2SC-4 to S and of the sulfur
to H2S, and on H2 chemisorption.

H2 Chemisorption  on Virgin Carbon -

The virgin carbon was heated to about 1850°F in an inert
atmosphere to purge off any chemisorbed oxygen.  The sample
was then cooled to the indicated temperature under an inert
purge and H2 introduced to the reactor at the indicated
concentration.  During all phases of treatment the off-gas
was analyzed using a gas chromatograph.

Recycle Experiments - H2 Chemisorption on Recycled Carbon -

Virgin carbon was to be loaded with H2S04, reacted with
H2S to form elemental sulfur, and then treated at 1200°F
with 30 vol. 70 H2 to form H2S and measure chemisorbed H2-
During each process step the off-gas was to be analyzed
with the gas chromatograph or S02 analyzer.  The carbon
was then to be analyzed for S02 activity and recycled
through each process step again.

Run Conditions -

The run conditions initially anticipated are listed in
Table 45.

The first H2 Chemisorption studies on the bench scale
were planned using virgin carbon, but the original equip-
ment and procedures used were unsatisfactory for detecting
H2 Chemisorption.  Five experiments indicated  that either
no H2 was chemisorbed or that the process was  so fast that
the Chemisorption could not be detected with the restraint
that a gas analysis could only be made every 12 minutes.
                          185

-------
Table 45.  PLANNED  EXPERIMENTAL PROGRAM  FOR STUDYING HYDROGEN
            CHEMISORPTION ON ACTIVATED CARBON DURING
            REGENERATION OF  THE CARBON
Run
HCS-1
HCS-2
IICS- 3
HCS-ll
HCS-5
Run
Inlet
Hg Cone.
Vol. %
38
38
38
20
10


*
HCS-u-A-_
Run
HCS-6-B-_
Run
HCS-6-C-_
Grams
Carbon
15-2
16.2
17.3
15.2
15.2
Inlet
S02
PPM
2500
HO
PPM
150
Lin. Gas
Velocity
ft. /sec.
0.20
0.20
0.20
0.20
0.20
Gas Cone.
Vol. %
N2
Bal.
Inlet
HgS Cone.
Vol. %
30
Inlet
H2 Cone.
Vol. %
30
02 H20
3.5 10
Grams
Carbon

Space
Velocity
1300
1300
1300
1300
1300
Grams
Carbon
15

lit
Grams
Carbon
15
Temp.
1200
1000
800
1200
1200
Effect
Temperature
M
tt
Inlet H2 Cone.
11
Lin . Gas Space
Velocity Velocity
ft. /sec. hr."1
0.18 3,000
Lin. Gas
Velocity
ft. /sec.
O.Oll*
Lin. Gas
Velocity
ft. /sec.
0.20
Space
Velocity
hr.-1
200
Space
Velocity
hr.-1
1300
OF
200
Temp.
300
Temp.
op
1200
Purpose
H2
Chemisorption
on Virgin
Carbon
Purpose
Effect of
Recycle
on H2
Chemisorption
            *Run HCS-6- -
                       Denotes cycle number
                      A denotes S02 sorption.
                    *• B denotes sulfur generation.
                      C denotes H2S generation.
                              186

-------
Therefore a different equipment setup was used which was
composed of a thermal conductivity cell as used on a gas
chromatograph.   A schematic of the equipment used is shown
in Figure  63.    A major difference in this setup is that
the off-gas from contact with the carbon is analyzed
continuously.  The experimental procedure was as follows:

     1)  The virgin carbon was heat treated at 1800°F
           with 120 scc/min. of N2,(R-l) for 3 hours.

     2)  The sample was cooled to room temperature at
           the same N2 flow rate.

     3)  The detector was calibrated.  The H2/N2 mixture
           (about 30%) was passed through the reference
           side of the detector and known mixtures were
           passed through the other side of the detector.
           The resultant responses were plotted and
           served as a calibration curve.

     4)  At time zero, the N2 (R-l) was cut off and the
           H2/N2 mixture (R-2, R-3) was passed thrqugh the
           bed when the response from the detector was
           zero, the sample was heated to 1200°F.   The
           resultant deflection is proportional to the
           amount of H2 chemisorbed.

Two experiments were run by this experimental procedure.
The results of the experiments are summarized in Table 46.
The experiments differed in the heat-up rate from 100 to
1200°F, namely 20 and 75 minutes.  At the inlet hydrogen
concentration of about 30 volume % at a total gas flow rate
    Table 46.  H2 CHEMISORPTION ON VIRGIN CARBON
Run
HCS-8
HCS-9
Heat-up
Time to
1200°F,
mi nutes
20
75
Initial
Carbon
Charge,
gms.
15.2
15.2
Total Gas
Flow
Rate,
scc/min.
49
50
Inlet
H2
Cone.,
Vol. %
30.6
31.7
Total
H2
Pick-up,
sec
85
101
Total H2
Chemisorbed,
moles H2/100 gms C
0.026
0.030
                          187

-------
   Figure  63.
Equipment schematic  for H2 chemisorption
experiments on virgin carbon
                                  Recorder
                                              Power Supply
 Temperature
Potentiometer
                                       Thermal
                                     Conductivity
                                       Detector
                     T    T
                                   Toggle Valves
                             188

-------
of 50  scc/min.  the  amount  of H£  chemisorbed was  about
0.028  moles H2/100  gms  carbon.   In  earlier experiments,
the amount of H2 unaccounted for was of  the order  of 0.1
moles  H2/100 gms carbon.   Therefore only about 30% of
the necessary H2 is accounted  for if the H2 chemisorption
phenomena is similar on virgin carbon and sulfur loaded
carbon.

There  are a number of factors  which make determination of
H2 chemisorption on sulfur loaded carbon difficult to
determine on the bench  scale.  These include the following:

     1)  The bench  scale is at unsteady  state.

     2)  The phenomena  of  H2 chemisorption is apparently
         fast at the temperatures of H2S generation.

     3)  The system for continuous  analysis (Figure 63)
         is only suitable  for  a  two component system
         (inert of N2 + one other component such as H2),
         but during H2S generation,H2S is also formed,as
         well as a number  of other  possible compounds.

     4)  All of the problems mentioned above would prob-
         ably lead  to difficulties  in setting up an ana-
         lytical system with available equipment for
         studying H2 chemisorption  as a  function of
         cycling on the bench  scale.

     5)  Because of the difficulties mentioned in  Step 4
         a delay in achieving  integral operation was
         anticipated.

Further analysis of the impact of the possible increased
H2 requirements  due to H2 chemisorption was made  as to the
effect on overall process  economics.  The process  economics
were assessed on the basis of  20% additional H2 required.
The results based on a  previous  economic analysis  by
Westvaco, which should  at  least  indicate the relative cost
increases even  if the economic bases may have changed
somewhat.  There is a projected  increase in costs  of
about  1.7% in the capital  costs  and about 3% in the
operating costs for a 1,000 MW installation.

At this time a program  that  lead to integration and opera-
tion of the pilot plant for repetitive cycling of  the carbon
was being pursued.  It was felt  and shown that this opera-
tion gave a good indication of the  hydrogen requirements on
a long term basis.  Preliminary bench scale tests  indicated
                       189

-------
 that complete definition  of the hydrogen use  in  an
 unsteady state system would require significant  effort
 and instrumentation not available at this time.

 Any further effort on bench scale experiments would have
 delayed the integration of  the pilot plant and it  was felt
 that further consideration  of bench scale work was not
 justified.  This was based  on the minimal impact of the
 increased hydrogen use on economics and anticipated data
 that was obtained in the  pilot plant.

 H2S Generation Kinetic Studies -

 The first runs were made  to find the effects of  H£ concen-
 tration, and sulfur vapor concentration on the formation
 of H2S with and without activated carbon as a catalyst.
 Conditions for these runs are listed in Table  47.
      Table 47.  EXPERIMENTAL  CONDITIONS FOR H2S
                 GENERATION  RATE  EXPERIMENTS
                                Run I
                   Run II
    Inlet H£ Cone.
    Inlet S Cone, as Si
    Total Flow Rate
    Reaction Temperature
    Linear Veloc. @ 1000°F
    Bed Depth
    Space Veloc. @ 1000°F
    Carbon Weight

    Carbon Type
    Empty Reactor Volume

    Carbon Bed Volume
    28.5%           18.9%
     6.36%            7.13%
645 cc/min. STP   625 cc/min. STP
   1000°F          1000°F
 0.22 ft./sec.    0.21  ft./sec.
  0.2 inch         0.2 inch
 47,800 hr."1     46,200 hr.'1
           1.65  gm

 Virgin WV-W 12x40, Log  C-70-30
         0.00273  ft.3

       0.0000908 ft.3
Homogeneous  Reaction -

The reactor  system was constructed with two identical
reactor  tubes,  one of which contains  a carbon sample and
another  which is empty.  Reference to the homogeneous  reac-
tion indicates  reaction in the empty  tube although, of
                           190

-------
course, there may be some influence of the tube wall.  Rate
determinations were made by measuring the steady state
concentrations of H2S in the effluent gas and calculating
rate of formation using the total gas flow rate.  Based on
the reactor volume inside the reactor furnace held at
1000°F, the rates of formation were found to be 0.054 Ib.
mol/hr./ft.3 and 0.038 Ib. mol/hr./ft.3 for Runs I and II,
respectively, under the conditions listed in Table 47.

These results may be compared to the rates derived from
kinetic data of Aynsley, Pearson and Robinson^ and of
Norrish and Rideal5,   In each case these authors' data
yielded good straight line plots according to the
expression
                   dt
- k[S]1/2[H2S]
                                                        (50)
     where  k  =

when the  calculated  values  of  In k  from their  data were
plotted against  1/T  over  their experimental  range of  550-
650°F.  Using  values of k so extrapolated  to 1000°F,  reac-
tion rates were  calculated  for the  present experimental
conditions and are compared in Table  48.
   Table  48.   COMPARISON  OF  H2S  FORMATION  RATES
               FROM LITERATURE AND EXPERIMENTAL
               DATA (HOMOGENEOUS)
Source
Any s ley
Norrish
Experimental
Rate at Inlet Conditions at
at 1000°F (Ib. mol/hr./ft. 3)
Run I
0.00024
50.6
*
0.054
Run II
0.00017
35.5
0.038
•^•••••••^^•^^^••^•^•^^^^-^••••i • •• .*«^^fc^«j^^»Maa
k
23
4,800,000
5,024 (I)
5,270 (II)
•MK*^^— ^^^wi*B*a^bv^^^^^^^^^^^M
                        191

-------
As seen, the present experimental data fell between the
widely separated literature data although somewhat closer
to that of Aynsley.  No adjustment of the present experi-
mental data for the difference between inlet and outlet
concentrations or the residence time in zones below 1000 F
in the reactor can account for the differences observed.
In the latter case, the inlet and outlet lines for the
reactor were heated to below 700°F, and the temperature
coefficients for reaction determined from the literature
indicate that reaction in these zones should be 500 to
1,000 times slower than at 1000°F and, therefore, negligible

Effect of Temperature, H2S .and S Concentrations
on the Heterogeneous Reaction To Form H2S with
a Catalyst of Activated Carbon

The heterogeneous reaction was carried out in an identical
reactor tube containing a small bed of activated carbon
sized to approximate the gas residence time used in 4"
pilot H2S generation work.  The rate of formation of
H2S in the bed was taken as the difference between the
total H2S production from the catalytic reactor and that
found for the homogeneous reaction.  At 1000°F where a
considerable amount of H2S was formed by homogeneous
reaction, this treatment may not be completely accurate
but the present results are at least illustrative.

The rates of H2S formation found were 1.5 Ib. mole/hr. /ft. •*
for Run I and 1.9 Ib. mol/hr. /ft.J for Run II under the
conditions listed in Table 47.
   Table 49.  COMPARISON OF HOMOGENEOUS AND
              HETEROGENEOUS REACTION RATES FOR
              SIMILAR INLET CONCENTRATIONS

Homogeneous
Heterogeneous
Rate H2S Formation
(Ib. , mol/hr. /ft. 3)
Run I
0.029
1.5
Run II
0.025
1.9
                           192

-------
A comparison to the homogeneous reaction may be made by
calculating the homogeneous rate at the reactant concen-
trations corresponding to those at the inlet to the carbon
bed which is located near the exit of the reactor tube
assuming rate = 5100[S]^[H2].

It is seen, therefore, that the H2S production rate
increased 50 - 75 times in the presence of carbon.  However,
from these data it is still not clear whether the carbon
has any particular catalytic influence.  For example, the
surface area of the reactor tube is approximately equal to
the external area of the carbon particles in the reactor,
so that on the basis of H£S formed per unit of exposed
surface, the "catalytic" rate is no more than twice the
"homogeneous" rate.  In spite of this, it still will
be possible to obtain rates applicable to reactions
within a carbon bed as employed in practice.  It should
be noted that the sulfur concentrations influent to the
carbon in these experiments are equivalent to the equilib-
rium concentration over carbon containing adsorbed sulfur
in the 10-11 Ibs. S/100 Ibs.  C loading range.  Such
concentrations should be typical of vapor phase sulfur
obtained during H2S generation after most of the sulfur
has been stripped off thermally.

Following the initial experiments a second set of experi-
ments was performed to further compare the homogeneous
and heterogeneous reactions, test the applicability of
the sample rate expression rj^S = k[H2][S]l'2 and deter-
mine the effect of temperature on reaction rates.
Table  50   lists the experimental conditions used.  It
is noted that in the case of the heterogeneous reaction,
reactant concentrations influent to the carbon bed were
assumed to be equal to the effluent concentration from
the empty reactor.  These were calculated based on the
observed conversion in the homogeneous reaction.

The experimental data was treated by calculating, for each
run, the values of the rate constant k from the observed     .
rate of production of H2S, the rate equation r^S = k[H][S]_]
and the calculated averages of -the hydrogen and sulfur
concentrations between inlet and outlet.  This implies
differential conditions at the average concentrations,
which is not entirely accurate.

If the assumed rate equation is valid, values of k should
be equal at a given temperature and In k should change
linearly with 1/T.  Table  51   shows the values of average
                       193

-------
    Table 50.  EXPERIMENTAL CONDITIONS FOR SERIES HS-2
                    Uniform Conditions
Carbon Type:
Bed Depth:
Linear Veloc.  at Exper.  Temp
Empty Reactor Volume:
Carbon Bed Volume:
Carbon Weight:
WV-W Loaded to 24% S Run H
0.2 inches
0.22 ft./sec.
0.00273 ft.3
0.0000908 ft.3
2.255 grams
              Temperatures and Concentrations
Conditions
Temperature, °F
Run
1
900
2
900
3
1000
4
1000
5
1000
6
1100
7
1100
HOMOGENEOUS REACTION
% H2 Inlet Cone.
% Si Inlet Cone.
30.1
10.0
20.0
10.0
30.1
10.1
20.1
10.1
30.2
14.5
30.0
9.93
19.9
9.93
HETEROGENEOUS REACTION
% H2 Inlet Cone.
% Si Inlet Cone.
28.7
8.29
19.0
8.84
26.6
5.73
17.4
7.07
25.7
9.07
23.2
1.20
14.0
3.35
                           194

-------
Table 51.  COMPARISON OF RATE CONSTANTS
           AT DIFFERENT TEMPERATURES
Run
No.
1
2
3
4
5
6
7
Temperature ,
°F
900
900
1000
1000
1000
1100
1100
kavg (Horn . )
xlO-3
2.07
2.05
5.40
5.45
5.68
12.2
15.1 '
kavg (Het.)
xlO-3
124
92
309
340
340
1,040
810
                 195

-------
k calculated for each of the runs made in the homogeneous
and heterogeneous reactors.  Agreement between values of
k tends to support the assumption of the rate equation for
both homogeneous and heterogeneous cases.

Figure  64  shows plots of In k vs. 1/T for these data and
compares the results of these experiments with those
extrapolated from calculations based on the work of
Norrish and Rideal, and Aynsley, et al.  The experimental
data is seen to yield straight line plots, the slopes of
which are approximately equal for both the homogeneous
and heterogeneous reactions.  Activation energies calcu-
lated from these slopes are 40.9 Kcal for the homogeneous
case and 44.7 Kcal for the heterogeneous reaction.  The
similarity of these activation energies indicates that while
the presence of carbon apparently increases the reaction
rate per unit volume of reactor, carbon does not have a
specific catalytic effect, except perhaps by virtue of its
surface area.  Table  52   compares experimental reaction
rates and rate constants based on the reactor tube surface
area in the homogeneous case and the external carbon
surface area in the heterogeneous case.  This latter area
was estimated as the surface area of spheres having the
mean particle diameter of the carbon sample.  On this basis
the rate constant for the "heterogeneous" reaction was
only about twice that for the "homogeneous" reaction.  This
is probably within the error of the external particle
surface area estimate.

It may also be reactor surface area which accounts for
the wide difference in results for the two literature
investigations noted.  In any case it is apparent that one
aspect of future investigations must be to establish the
proper reactor size basis for rate constant calculations.

Effect of Mass Transfer on H2S Generation Kinetics -

Following H2S generation experiments reported previously,
the experimental program has continued to investigate:
1) "homogeneous" reaction kinetics with integral analysis
of data, 2) gas film resistance and the effect of linear
velocity on reaction over activated carbon.

Homogeneous Reaction - Empty Reactor -

Analysis of earlier experimental data was made assuming
that differential reaction conditions were present but it
was realized at the time that this could not be true since
conversion of sulfur was in some cases almost complete.
These data did,  however,  indicate that the rate expression
                          196

-------
       Figure  64.  Arrhenius  plots  for experimental and
                    literature data
  ixio6 -J
  ixio
id
OL
  1x10
  IxlO3
               .50
.60       .70      .80

        1/T, ("R)"1 x 103
.90       1.0
                               197

-------
                        Table 52.   COMPARISON OF RATES AND RATE CONSTANTS BASED
                                   ON REACTOR SURFACE AREA FROM SERIES HS-2
Run
No.
1
2
3
4
5
6
7
1
Rate (Horn.) x 1(P
# mol/hr. meter 2
6.14
4:10
14.3
10.0
18.2
25.5
22.1

avg k (Horn . )
470
466
1,230
1,240
1,290
2,770
3,450

Rate (Het.) x 10 J
# mol/hr. meter 2
11.2
5.91
21.2
16.1
27.2
26.6
21.8

avg k (Het.)
1,020
761
2,550
2,810
2,810
8,560
6,690

k(Hom)/kHet
2.17
1.63
2.07
2.27
2.18
3.09
1.94

00

-------
                 rH2S =  k[H2][S]                         (51)
was  likely  candidate for further tests.  To this end, the
plug flow equation was integrated using an analogous rate
expression  as  follows.

Derivation  of  Integrated Rate Equation -

Assuming  reaction
                  H2 + 1/2 82 —*H2S                       (52)
 then:
                                                         (53)
     where   N  = moles  reactant
             No  = initial moles reactant
              NS2  -



     where   X = conversion of reaction.



Assuming  the rate  expression



                 -r   =  K C  C 1/r2                       (55)


     where   C = concentration.



Dividing  (54) and (55)  by V,  and substituting:



                            199

-------
                 -rH  = K[CH(l-XH)]-[Cs-l/2 CHXH]1/2 *              (56)
       Assuming the plug  flow equation


                                    -XH
V -   /   **
H>o    J0  -'H
 /
                                V V U
            where  u = a + bx,  a = Cs, b = -1/2  CH
                   v = C + dx,  c = CH, d = -CH
                   k = ad - be  = Cn2/2 -
                         (FH)
            where  (F)o =  initial molar flow rate
                  V    =  reactor volume
       and substituting  for  rjj



               V    _  1^  C   	dXfl	             (58)
             (FH)O     K JQ   (cH - CHXH) (cs -1/2 cHxH)1/2


       This is integral  of form
                                                                 (59)
       Because for present  data CH > 2 GS, kd  <  o,  therefore,
       proper integrated form is



              I    =;           arctan [	"  ]                   (60)
             y  v /u    =  / -kd          •  -kd
                            -?— arctan -CH/Cs-1/2 CHXH
                    	XH
      Ut *. Vr I.U1.1.    "	"	 ' "•

   kCH
*Subscript  molecular numbers  are dropped, H =  H2 etc.


                                200

-------
                          [ (-arctan CH^Cs-1/2 CHXH}
                        -  (-arctan
                          (arctan B - arctan A)
     where  B = CH
           A = cH 'CS/ -1/2 CHXR/ /kci
Then
                         arctan (, ~AB)  -  =               (61)
     where  K = rate constant

           k = CH2/2 - CHCS
           A = CH /Cs-l/2 CHXH/ /kCH

           B = CH
Confirmation  of  the integrated rate equation,  (61) ,  is
obtained if the  experimental data plotted  as V(FH2)o vs•
the right hand expression (Z) gives a straight  line  pass-
ing through the  origin.   The reaction rate constant  can
then be calculated from the slope of this  line.

Experimental  -

Experiments using  the empty reactor tubes  have  now shown
that the equation  does properly describe the experimental
results.  In  the course of these runs, however,  it was
                        201

-------
discovered that the two matched reactor tubes, one of which
was to be used for the reaction over carbon and the other
which was to be left empty for reference did not produce
the same conversion when both were run empty.  Nevertheless
the data obtained allowed testing of the proposed integrated
rate equation.

Experimental conditions were:

        Temperature:     1000°F

        Sulfur Cone.:    15% as Si

        Hydrogen Cone.:   30% H2

        Reactor Volume:   0.00273 ft 3 without inserts
                         0.00176 ft.3 with inserts.

The results are shown in Figure 65.    Run HS-3 was for the
empty reference reactor and Run HS-4 was made using the
empty reference reactor and the empty carbon reactor, both
with Vycor inserts.  This figure illustrates three findings:

     1)  The lines are straight and pass through the
         origin, indicating a correct rate equation.

     2)  There is a pronounced difference in the rates
         obtained using the reference reactor as com-
         pared to the empty carbon reactor although
         they have the same internal dimensions.

     3)  There is almost coincidence between the data
         taken with the reference reactor with and
         without the Vycor insert.

It is noted that this coincidence is related to using
reactor volume at the reactor size factor in the term
V/(FH2)O-   Figure  66  shows that agreement is not as
good if the reactor size factor is in terms of internal
surface area, A, instead of volume, V.

With regard to the results of experiments in the reference
reactors it would be tempting to conclude that there was
relatively little contribution to reaction due to wall
area and that a simple homogeneous rate constant could be
best derived based on reactor volume.  Yet the other
reactor tube used in Run HS-4 which was dimensionally
identical to the reference tube produced a significantly
lower rate of reaction.   Comparative reaction rate constants
based on volume for the tubes are as follows:
                        202

-------
Figure 65.  Test of integral rate equation  (61), Z as
            function of tube volume   (Runs HS-3 and -4
            reference reactor)
                                                       /   /
                                                  •/
                                        HS-3 (Ref) -v'
    /

      4
     /
                                                 / ,^HS-4 (Ref)

                                        x  X

                                 /  /
                                 /  '
                           '/
                  .4*
5-4 (React)
                              _j	

                               3

                               V/F
                          203

-------
  Figure 66.   Test of integral rate  equation (61), Z as
              function of tube area  (Runs  HS-3 and -4
              reference reactor)
,-y
                                    X
                                   X
                                  X
                                 x








                          '         X
                            ,-t
                  *   •

         X
        X

x    — x
                            204
                                                          / HS-3

                                                         X
                                                         X
                                                        X
                                                       X
                                                      X

                                                    X
                                                   X
                                                   X
                                                  X
                                                 X

                                                                 '1
                                                               X
                                                              X
                                                          '  HS-4
        40         80         120        160        200        240


                            A/(FH2)0

-------
        Run   Reference Tube   Carbon  Tube
       HS-3
       HS-4
ll.OxlO3,
10.4xl03
4.60xl03
Whatever other  conclusions are drawn  from this,  it is
obvious that  in studies of the reaction  over carbon, it is
not possible  to account for the "homogeneous" contribution
to total conversion using parallel data  taken with this
reference tube.   Thus,  the only alternative is to account
for homogeneous reaction by means of  the empty tube data
for the reactor to be used in experiments with carbon.
                                  /
investigation of Mass Transfer Effects  -

In the heterogenous reaction over carbon,  the rate constant
derived from  bench scale data is to be applied to the pilot
fluid bed case.   Since there is a significant difference
in gas linear velocity in the two cases  it becomes necessary
to determine  the extent of velocity dependence, since such a
dependence  indicates a rate limiting  resistance  due to the
diffusion rate  of reactants and products through the gas
film surrounding the carbon particles.   An investigation
was made by measuring conversion of hydrogen as  a function
of gas flow rate for three different  carbon bed  sizes.
Experimental  conditions are as noted  in  Table 53.
         Table 53.   EXPERIMENTAL  CONDITIONS
                    FOR RUNS HS-4 TO  HS-7
  Carbon Type:

  Temperature:

  Inlet H2 Concentration:

  Inlet Si Concentration:

  Carbon Bed Weights:

  Carbon Bed Volumes:

  Total Gas Flow Rates:

  Space Velocity Range:

  Linear Velocity Range:
       SG-32, 24% S Loading

       1000°F

       30%

       15%

       0,  1, 2, 4 gms

       0,  4.58xlO~5, 9.16xlO~5, 18.3xlO~5 ft.3

       1.96 to 7.84 ft.3/hr. @ 1000°F

       10,700 to 171,000 hr."1 @ 1000°F

       0.173 to 0.692 ft./sec.
                        205

-------
       The results of these experiments are shown in Figures
       67   and 68.    Figure  67  shows conversion of H£(XH2) as
       a function of bed volume/H2 feed rate [V/(FH2)ol-  If Sas
       film diffusion were unimportant the data should all lie on
       a single curve.  It is obvious that this is not the case
       here and that this resistance affected reaction rates at
       all velocities used.  Figure  68  shows the conversion as
       a function of gas flow rate for nearly constant inlet condi-
       tions and space velocity.   In the absence of film resistance,
       conversion would be constant with increasing linear velocity.
       Since the conversion continually increases,it is clear that
       these data are affected by mass transfer effects.

       H2S Generation Kinetic Model -

       The kinetics of H£S generation from the reaction of hydrogen
       and sulfur vapor over a catalyst of activated carbon have
       been studied.  The effects on the reaction rate of H2S
       formation,  of temperature, of hydrogen concentration, of
       sulfur vapor concentration, and of mass transfer effects
       have been measured.  This  has led to a combination of all
       the variables into a kinetic model relating these variables
       to the rate of H2S formation as given by Equation  (62) :
                   rH2S  =  k e-^^/1 (Si)J-/^ (H2)                 (62)


            where  rjj£g =  rate °f ^2S  formation,  moles/hr.-m^
                            of C surface area
                               2.08(ig9)(y)0-5
                   k    =  (299)(T)(e-30645/T) +(v)0.5 '

                            moles/hr./m^ of C surface area.
5.2.5  Combined Sulfur Stripping/H2S Generation

       The existing 8 stage,  4" diameter reactor was shown to
       be suitable for integral operation as a combined H2S
       generator/sulfur stripper.   The conclusion is based on
       the facts that 86% of the stripped sulfur was converted
       to H2S (75% required),  81% of the H2 was converted to
       H2S (90% required),  and the S02 activity of the carbon
       was 106% of the virgin precursor.  Although 817o of the
       H2 was converted to H2S, 94% of the inlet H2 was
       utilized.   The unaccounted for H2 was postulated as
       being chemisorbed onto the carbon or reacted with chetni-
       sorbed oxygen.
                              206

-------
 Figure 67.  Variation of conversion with residence
             time  for three bed volumes
0.4,
                                         Run HS-7
                    Gas Velocity Increase
0.1
0.2
0.3
                                             0.4
0.5
                            207

-------
   0.4
   0.3
            Figure  68.   Effect of  flow rate on  conversion
                         at constant  residence time
                                                              HS-7
CM
   0.2
                                    HS-6
   0.1
                    5-5
                                     At V/FH2 = 0.087
    0
                            8
  12


(FH2),
16
20
24
                                  208

-------
Preliminary experiments suggested that  the gas/solid  con-
tact time for the H2/sulfur reaction  should be  increased.
This equipment change was made, problems encountered  in
the initial runs were rectified, and  the above  mentioned
objectives were met.

The overall results  indicated  that  for  sufficient sulfur
removal, for carbon  regeneration for  S02 pickup,  and  for
H2/sulfur conversion to H2S, the operating temperature
would be 1000 to 1200°F,  the inlet  H£ concentration would
be 20 to 40 volume %, and the  carbon  residence  times  of
6 to 13 minutes were sufficient.

It was recommended that the 8  stage,  4" diameter  regenerator
be used in integral  operation  as a  combined H2S generator/
sulfur stripper and  be integrated with  the 18"0 S02 sorber.
It was further recommended that runs  be made  on bench scale
to verify the phenomena of H2  chemisorption and that  addi-
tional runs in the 4"0 unit be considered to  evaluate lower
operating temperatures.   The unit was used as a combined
reactor in integrated operation with  the 18"0 S02 sorber.
Also, some preliminary bench scale  work on H2 chemisorption
was completed, as discussed in  a previous section.

Combined Regeneration Process  Concept -

In the Westvaco Process, S02 is removed  from the flue  gas
by activated carbon.  During the removal, S02  is converted
to sulfuric acid which remains as the sorbed  species  on
the outlet carbon.   The sorbed acid is  subsequently con-
verted to elemental  sulfur by  the reaction with hydrogen
sulfide.  The next step is to  recover one-fourth  of the
sulfur as elemental  sulfur and the  remaining  three-fourths
as hydrogen sulfide  by reaction with  hydrogen by  Reactions
63 and 64:
       4 S  (Sorbed)   Activated^   s + 3 S  (Sorbed)            (63)
                     Carbon
         3 H2 + 3 S (Sorbed)   Activated».  3 H2S              (64)
                            Carbon
The hydrogen sulfide produced  is  used in the  first step of
regeneration.  The reactivated carbon is recycled to the
S02 sorber.
                       209

-------
 Results  and Discussion  -

 For  integral operation, both regeneration  steps of  sulfur
 stripping and H2S generation can be carried out  in the
 existing 8 stage, 4" diameter regenerator.  The  objectives
 of the task were met with 86% of the stripped sulfur con-
 verted to H2S, 947<> utilization of the inlet hydrogen, and  :•
 the  SC>2  activity of the product was 6% higher than the
 virgin precursor.

 The  conversion of H2 and sulfur to H2S increases with
 increasing temperature  and with decreasing space velocity
 (increasing gas/solid contact time).  The S02 activity of
 the  carbon and removal  of sulfur from carbon increases
 with increased residence time of the carbon in the reactor
 and  with increased temperature.

 The  experimental results for all runs made to evaluate
 combined sulfur stripping/H2S generation are summarized
 in Tables 54, 55 and 56 with detailed data given in
 Appendix A-18.  Tables  54 and 55 give the average  measure-
 ments for the carbon and gas phase, respectively.   Table
 56 summarizes the major response variables calculated
 from the experimental data.


 Effect of Temperature on Sulfur Removal from Carbon -

 As the temperature increases, the vapor pressure of sulfur
 over carbon increases.  The presence of hydrogen in the gas
 phase enhances the driving force for the sulfur removal,
 because  of the reaction of the hydrogen with sulfur in the
 gas phase and in the pores of carbon.  The effect  of tempera-
 ture on  the removal of  the sulfur from the carbon  with H2
 present  is given in Figure  69.   As expected, as  the
 temperature increased,  the percentage of sulfur stripped
 from the carbon in the  8 stage reactor increased.   An
 operating temperature of 1000 to 1200°F is indicated from
 the data to provide sufficient removal rates of sulfur.

 Effect of H2 concentration and Carbon Residence
 Time on  Sulfur Removal  from Carbdn

 The effect of H2 concentration and carbon residence time
 on sulfur removal from  carbon at 1200°F is given in Figure
 70.    As the hydrogen  concentration increases the sulfur
 removal  increases as expected.  Also as the carbon residence
 time increases the sulfur removal increases.  Carbon resi-
 dence times of 6 to 13  minutes for an inlet H2 concentra-
 tion of  20 to 40 vol. % appears sufficient at 1200°F to
provide  sufficient removal of sulfur.
                       210

-------
                    Table  54.   EXPERIMENTAL  CONDITIONS  FOR EVALUATION  OF  COMBINED
                               SULFUR STRIPPING/H2S  GENERATION
Expt.
Number
SHG-1
SHG-2
SHG-3
SHG-5
SHG-7
SHG-8
SHG-9
SHG-10
SHG-11
Carbon
Inlet
% S
20.70
20.30
20.35
19.78
20.16
5.08
20.12
19.81
19.81
Mat'l
Rate
#/hr.
40.5
35.1
30.9
30.9
30.9
25.2
30.9
35.0
35.0
C
Rate
#/hr.
32.2
28.0
24.6
24.8
24.7
23.9
24.7
28.1
28.1
Carbon
Residence
Time
minutes
10
11
13
13
13
13
13
10
10
Outlet
% S
15.79
4.85
2.82
4.91
2.80
2,90
2.64
4.21
8.27
Mat'l
Rate
#/hr.
37.3
29.1
23.85
25.1
25.0
24.9
24.6
28.9
30.8
C
Rate
#/hr.
31.4
27.8
23.2
23.9
24.3
24.2
24.0
27.7
28.2
Cyclone
% S
28.9
42.7



25.4
31.2
Mat'l
Rate
#/hr.
0.0055
0.0036
0.0033



0.015
0.021
Temperature, °F
Stage Number
# 1
768
942
1130
1144
1135
1132
1140
1132
1070
# 2
771
970
1168
1169
1175
1168
1173
1135
1035
# 3
781
982
1178
1179
1180
1178
1186
1072
1040
# 4
778
967
1147
1140
1192
1150
1165
899
925
# 5
801
1001
1148
1170
1198
1178
1187
1120
1085
# 6
799
970
1115
1139
1025
1178
1164
1219
1207
# 7
1024
1170
# 8
734
840
914
951
910
938
918
N>

-------
                    Table 55.   EXPERIMENTAL CONDITIONS FOR EVALUATION OF COMBINED
                                SULFUR STRIPPING/H2S  GENERATION



Expt.
Number

SHG-1
SHG-2
SHS-3
SHG-5
SH6-7
SH6-8
SHG-9
SHG-10
SHG-11
Gas
Inlet

Linear
Gas
Velocity
ft./sec.
1.9
2.0
1.9
2.1
1.8
1.8
2.0
2.0
2.0
Tntal
Gas
Flow,
CFH A
70° F
254
229
191
214
176
176
202
206
202

H2
Flow,
CFH @
70°F
66
67
52
22
63
63
53
' 79
0

H2
Cone. ,
Vol.
%
26.8
32.0
29.8
5.4
37.7
36.7
27.3
36.1
0
Outlet

Gas Concentration, Vol. %

H2
16.2
10.0
3.1
0
5.2
27.7*
3.06
2.0
0
H2S
4.1
13.4
15.2
3.0
18.6
2.6
13.0
27.3
1.32
S02
0.20
0
0
0.07
0
0
0.007
0
0.82
H20
6.30
8.50
5.90
5.40
6.50
0.42
5.73
9.0
9.0
CO
0
0
0
0
0
0.01
0
0
0
C02
1.0
1.25
1.80
1.40
1.70
0
1.40
1.80
2.00
N2
68.3
68.5
70.7
88.0
64.7
69.4
71.7
60.0
88.0

Sulfur
Cone.
# Si/hr.









Temoerature. °F

Stage Number

# 1
800
997
1203
1212
1210
1210
1210
1215
1200
# 2
	
	
	
	
	
	
	
1182
T150
9 3
800
997
1200
1204
1207
1200
1195
1198
1150
# 4
	
	
	
	
	
	
	
992
1020
# 5
806
997
1205
1208
1222
1210
1202
1204
1135
1 6
	
	
	
	
	
	
	
1234
1290
* 7
801
990
1192
1201
1192
1208
1197
1200
1215
# 8
	
	
	
	
	
	
	
1248
1130


Avg.
802
995
1200
1206
1208
1207
1201
1184
1161
to
       *Used  Stage 7 analysis (stage below gas outlet).

-------
                        Table 56.   EXPERIMENTAL RESULTS  FROM EVALUATION OF  COMBINED
                                    SULFUR STRIPPING/H2S  GENERATION
Expt.
Number
^fft^fm^mf^m^^f^m^mm
SHG-1
SHG-2
SHG-3
SHG-5
SHG-7
SHG-8
SHG-9
SHG-10
SHG-1 1
Space
Velocity,
Vol. Gas/
Vol. C/
hr.
1,650
1,490
1,240
1,390
1,150
1,150
1,320
1 ,530***
1,530
Hydrogen
Balance
In/Out
1.31
1.34
1.72
1.72
1.73
1.32
1.80
1.15
	
Sulfur
Balance
In/Out







1.03

# mole S
Stripped/
hr.
^g^gggMgggMMM,,^^
0.079
0.179
0.176
0.152
0.173
0.017
0.174
t
0.179
0.129
# mole H2
Avail.(H2+S)/
hr.
mti^f^mitiiimiiitttimtitartimimimmttmmmmitmmiimiimMmi^
0.178
0.198
0.153
0.028
0.176
0.170
0.145
0.186
—
# mole S
Stripped/
# mole He
Available
•••{•••^•^••^••••••••••••••••••••••••••i
0.44
0.90
1.15
5.43
0.98
0.10
1.20
0.96
—
Sulfur
Stripped
% of
Inlet
mftntumfmiiimiiiii^mimi^^^miffffimm
30
80
89
80
89
44
90
82
62
*
H2
Utiliz.
% of
Inlet
39
68
90
100
87
18
89
94
—
**
H£ Coriv.
to H2S,
% of
Inlet H2
^••^•.•••.••.^•••••••HWIIIIIIIIIIM
16
43
48
58
45
7
45
81
—
**
S Conv.
to H2S,
% of S
Stri pped
••••••••••••••••••••aiiiiiHHA^Bv-"*
35
47
42
10
55
64
38
86
—
S02
Activity
Rel. to
Vi rgi n
Carbon

0.98
—
—
1 . 1 5****
—
—
1.06
—
to
         *Goal £
        **Goal *

       ***Effective space velocity  for H2 + S  reaction in vapor phase «3,000 hr.'
      ****S02 activity of carbon sample from Stages 4 and 5 was about 0.94.

-------
             Figure 69.   Effect of temperature on sulfur
                           stripping with H2 present

-------
  Figure 70.   Effect  of H£  concentration and gas/solid
                contact time  on sulfur stripping at
                1200°F

-------
 Effect  of Carbon  Residence  Time  on S02  Activity -

 From previous  work, it  is  known that the S02  activity  is  a
 function of  the final  sulfur  loading, regeneration
 temperature, and  time  of  exposure  to a  reducing atmosphere.
 The S02 activity  increases  with  increasing temperature and
 exposure time  to  a  reducing atmosphere.  At  a  fixed tempera-
 ture the activity increases with decreasing  sulfur loading
 on the  regenerated  carbon.  In the present experiments at
 1200°F,  in particular  Runs  SHG-7 and -10, the  carbon  resi-
 dence time of  13  and 6 minutes,  respectively,  is sufficient
 to provide a product carbon more active than the precursor.
 The S02 activity  for Run  SHG-7 is  1.15  relative to the
 virgin  precursor  and for  Run  SHG-10 is  1.06  relative  to
 the virgin carbon.

 Sulfur  Removal Profile -

 The sulfur removal  stage  profile for SHG-7 (a  typical run)
 is given in  Figure  71.   As  seen  in this  figure,
 made at 1200°F, almost all of the  sulfur is  stripped
 off in  two stages.  This  means that even though the space
 velocity for sulfur removal from the carbon  is  about  1,150
 hr.~l,  the effective gas/solid contact  time  for the hydrogen
 and sulfur vapor  to form  H2S  is  about 2 stages  (space
 velocity about 4600 hr.~l).   This  suggested  that the  gas/
 solid contact  time  for the  H2 sulfur reaction  over carbon
 should  be increased,as was  done  in Runs SHG-10  and -11.

 Effect  of Temperature  on  Sulfur  Conversion to  H2S  -

 The effect of  temperature (SHG-1,  -2 and -3) is given in
 Figure  72.    The sulfur  conversion to  H2S increases  with
 increasing temperature, as does the percent sulfur  removed.
 Assuming the unaccounted  H2 is from inleakage  of air, then
 the curve given in  the figure results.   The  results indi-
 cate that higher  temperatures near 1200°F are  necessary  in
 the present  8  stage reactor at the space velocity  of  about
 5,000 hr.-l  for the H2/sulfur reaction.  In  Run SHG-10 the
 space velocity for  the H2/sulfur reaction was  decreased  to
 about 3,000  hr."1 by feeding  most  of the carbon at the
 middle  of the  column.  This increased the conversion  of  the
 sulfur  to H2S  and indicates that the temperature may  be
 decreased below 1200°F and  still allow  the goals to be
met.

 Carbon  Burn-off Rate   -

 The  CO?  produced  during the runs was necessarily indicative
 of  carbon burn-off.  Previous experiments indicated that
 first cycle burn-off could  be as high as 92  Ibs. C/Ton
 sorbed S02-  Burn-off  then  appeared to  decrease to as low


                          216

-------
         Figure  71.   Sulfur removal from activated carbon  in an

                        8 stage,  4" diameter regenerator
    0.28
u>
c
TJ


-------
        Figure 72.   Effect of temperature on the conversion
                      of sulfur to  hydrogen sulfide
    100
     80-
TJ

OO
OJ
1C
HI
O


S_

H-


l/>
     60-
    40-
     20-i
         [Effective Space Velocity  for H2 + S
          Reaction, 6,600 to 13,200 hr.-']
                             O Experimental
                                                      i

                             X Corrected Assuming Leak
                                  in Outlet Gas Sample
                                  Line
      700
800
900       1000         1100

  Average Temperature, °F
                                                              1200
1300
                                   218

-------
as 2 Ibs. C/Ton S02 after 8 cycles.  The data from Run
SHG-10 (1.8 vol. % C02) was equivalent to a burn-off of 67
Ibs. C/Ton sorbed S02 for a carbon loss of 0.42 wt. %.
Recent experimental and literature studies indicate that
oxygen may be chemisorbed on the carbon.  This is then
evolved as CO and C02 at elevated temperatures in the
presence of oxygen-free gas.  Since chemisorbed oxygen
would be depleted as C02 is evolved, burn-off should
decrease with cycling, as shown by the previous data.

Runs SHG-1 to -9  -

After Runs SHG-1 to -9 had been made a number of equipment
problems were found and subsequently corrected.  The first
problem was a leak in the gas sample lines.  A second
problem was that the chromatographic gas analysis of
hydrogen had been specified by the manufacturer to be
linear, but the calibration was subsequently found to be
non-linear.  The hydrogen concentrations were adjusted
accordingly with the resultant values given in Table 55.

For a typical run, SHG-7, to see the effect of the leak in
the outlet gas line, if the assumption is made that all
of the unaccounted H2 was due to the inleakage of air into
the outlet gas sample line, then the outlet H2 and H2S
concentrations would be increased accordingly to 7.3 and
26.2 volume %, respectively, from 5.2 and 18.6 (see
Table  55) .   This would then mean 81% conversion of the
stripped sulfur to H2S and 79% H2 utilization; however,
as discussed later, there is another factor affecting this
assumption which increases the H2 utilization to form HgS.
Also the source of the H20 in the vapor phase will be dis-
cussed later.

Regardless of the problems encountered in these first
runs, the data  suggested that additional runs should be
made to increase the gas/solid contact time (decrease
space velocity) for the reaction of H2 with sulfur and
that 4 stages would be sufficient for the necessary sulfur
removal.

Runs SHG-10 and -11 -

Therefore, the equipment was modified to allow the carbon
to be fed both at the inlet and at Stage 4 (number from
bottom to top).  These runs (SHG-10 and -11) were made at
1200°F with 85% of the carbon fed to Stage 4 and 15% fed
to Stage  7.  This increased the effective number of stages
for H2 sulfur contact by three,.but decreased the number
for reactivation for S02 pickup by about four.
                        219

-------
       Run  SHG-11, with only  N£  in  the  inlet  gas, showed  that  the
       water  in  the  gas phase came from water  sorbed on the  carbon.
       Subsequent moisture determinations  in bench  scale equipment
       verified  the  findings in this run.  The water on the  carbon
       probably  came from sorption of  H20  from the  air  during
       handling.

       In Run SHG-10,  86% of the stripped  sulfur was converted to
       H2S, 95%  of the H£ was utilized (81%  converted to H£S),
       and  the outlet  carbon had an S02 activity 6% higher than
       the  virgin precursor.  The  discrepancy  of H£ utilization
       (94  vs. 81%)  is tied  up  in  the  hydrogen material balance,
       namely 0.37 Ib.  H2/hr. into the reactor and  0.32 Ib.  H2/
       hr.  out of the  reactor.   Based  on literature data on  02
       and  H2 chemisorption^, 7, 8 and work  Westvaco  has  com-
       pleted on 02  chemisorption,  it  is felt  that  the
       unaccounted hydrogen  is  chemisorbed.  Calculations show
       that only about 0.1 mole  of H2/100  gms.  C would  have  to
       chemisorb on  the carbon  to  account  for  the H2.   The
       literature data taken on  similar activated carbonaceous
       material  showed the chemisorption of  02  and  H2 are similar
       on a mole/gram  basis.  This  fact,combined with Westvaco
       data on 02 chemisorption  at  530°F of  about 0.03  mole/100
       gms. C with actual plant  produced carbon containing about
       0.1 mole  02/100 gms.  C, makes a  strong case for the H2
       chemisorption as  a possible mechanism for accounting  for
       the H2-   The  literature  data indicates  that  once the H2
       is chemisorbed,  then  it  is  irreversible  unless the tempera-
       ture is increased to  about  1800°F.  This  means the phenom-
       ena might be  expected only  in the first  few  cycles.   If
       the H2 chemisorption mechanism  is accepted then  the effect-
       ive H2 conversion for the formation of H2S is  increased
       to 94%.   Loss of hydrogen by leakage  in these runs is
       discounted, based on  the  fact that  the  sample lines were
       tested with known gases  at  the  reaction temperature.

5.2.6  Elemental Sulfur Recovery

       A sulfur  condenser was required for operation of the
       integral  pilot plant  on  a closed loop cycle using H2S pro-
       duced  internally with the process.   A condenser was
       designed, installed,  and tested prior to use in the integral
       run.   This pilot development of the condenser resulted in
       smooth operation of  sulfur  condensation and recovery during
       the  integral  operation.   The pilot  development and results
       for  the sulfur condenser prior  to the integral run are given
       below.
                              220

-------
Sulfur Condenser Operation  -

Initial Sulfur Condenser Testing -

Initial sulfur condenser tests were made without a recir-
culating  stream of liquid sulfur for scrubbing purposes.
These runs,  discussed below,  provided a basis  for improved
condenser design and operation.  The run conditions for
three initial runs that were  made are summarized in
Table 57.
    Table 57.   OPERATING  CONDITIONS FOR SULFUR
                CONDENSER  TESTING SYSTEM
    Liquid Sulfur Reservoir Temperature

    Liquid Sulfur Reservoir Pressure

    Liquid Sulfur Flow Rate to Vaporizer

    N£  Carrier  Flow Rate to Condenser

    Sulfur Vaporizer Temperature
        /
    Condenser Inlet Gas Temperature

    Condenser Outlet Gas Temperature

    Condenser Steam Jacket Temperature

    Condenser Exit Gas Sample Flow
     Rate Through Trap
-  290-300°F

-  2-5 PSIG

-  6 Ibs./hr.

-  225 SCFH

-  1000-1200°F

-  1000°F

-  250°F

-  250°F

-  2-3 SCFH
The results of running  at these conditions  are summarized
in Table 58.   The first run showed a  61%  recovery of
sulfur,  i.e., of the  sulfur which left  the  reservoir, 61%
was recovered from the  condenser as a liquid.   The rest
was carried out with  the nitrogen off-gas.   For the second
and third runs a mist eliminator, consisting of a roll of
                         221

-------
        Table  58.   SULFUR CONDENSER TEST RUNS
Run
1
2
3
Duration
minutes
260
180
104
N2 Flow
Rate
SCFH
230
230
230
Avg. Sulfur
Flow Rate
Ibs./hr.
2.2
3.0
3.0
Flow Rate
Range
Ibs./hr.
1 - 5
1 - 7
2.5 - 3.5
Avg. Sulfur
Vapor Cone.
Ibs. S/lb. N2
0.13
0.18
0.18
Recovery
%
61
88
88
 tightly wound wire mesh, was placed inside the condenser,
 increasing recovery to 88% in both subsequent runs.
 However, there is some question as to the validity of
 these percentages, because the possibility exists that some
 liquid sulfur may have bypassed the vaporizer; i.e., it
 may have passed through unvaporized and reached the con-
 denser as a liquid.  If this condition did exist, then the
 actual condenser efficiency would be lower than it appears.

 In order to achieve 99% recovery of the sulfur, it was
 necessary to provide some degree of scrubbing action.
 One approach would be to circulate a stream of liquid
 sulfur through the condenser, as was originally intended
 when the condenser was designed.  Another method would be
 to further process the off-gas by bubbling it through a
 column of liquid sulfur.

 Sulfur Condenser Operating Results -

 All of the experiments performed prior to integral testing
 on the recovery of sulfur from regeneration off-gases by
 condensation are summarized in Table 59.   As can be seen
 from the data, the original baffled exchanger without sulfur
 recirculation or a mist eliminator removed about 61% of the
 desired sulfur.  Addition of a mist eliminator increased
 recovery to 88%; however, it was found that the scrubbing
 action of recirculating liquid sulfur was necessary to
 raise sulfur recovery to the desired 99+%.  Addition of 29%
H2S to the gas contacting the recirculating sulfur caused
no apparent changes in sulfur viscosity, as indicated by the
 constant pump electrical load over a six hour period.  The
jacket cooling fluid was changed from steam to hot water
 to facilitate temperature control and improve heat transfer
rates.  The recirculating sulfur condenser system was
                          222

-------
                                    Table 59.   SULFUR CONDENSER OPERATION
Ron
Ha.
1
2
3








4




S






Comments
Much S Vapor/Hist
in Off Gas
Installed Mist
Eliminator
Tried Unsuccess-
fully to analyze
off gas S cone.
Achieved better
, control of S flow
to vaporizer.
Exoerienced prob-
lem controlling
jacket temp.
Liquid S recircu-
lation installed.
Also new steam-
life cooling
system.
Pump enclosed in
heated box. Pump
motor current
constant ? 7.5
amps. Observed no
change in liquid
S viscosity.
Gas
Comoosition
Sulfur*
18
13
15








13




0






H2S
0
0
0








0




29






j
»2
82
87
85








87




71






Total**
Gas
Flow,
CFH 9 70°F
164
236
226








225




231






Sulfur
Recirculation
Rate,
GPM
0
0
0


•




•
0.5




0.5






Jacket
Coolant
Steam
Steam
Steam








Water




Water






Temperatures, °F
Jacket
Inlet
280*10
280110
290110








27015




26015






Jacket
Outlet
290*10
295110
305110








27015




265i5






Cond.
Inlet
1000*50
1000+50
1000150








890130




875






Cond.
Outlet
—
--
31615








280HO



.-
290tlO






Run
Time,
hours
4.3
3.0
1.7








4.0




6.0






Sulfur
Recovered,
X
61
88
88








99.6









.

Off-Stl
Sulfur
Cone..
mole I
—
—
_



-




0.06




<*»•






ro
N>
U>
                     *As Si-

                     "Based on Sulfur as

-------
tested for 14 hours of intermittent operation without
serious problems and was used as the basis for design of
a system for installation in the pilot plant.  Because
of the unusual viscosity characteristics of sulfur,
temperature control will be a major design consideration
in order to keep all parts carrying liquid sulfur in the
range 260-315°F.

The feasibility was shown experimentally of a sulfur con-
denser system using recirculation of liquid sulfur as a
scrubbing media.  The anticipated improvement in the con-
densing efficiency to 99.6 weight % was realized.

The run conditions for the best sulfur performance are
schematically shown in Figure  73.    The sulfur concentra-
tion in the outlet gas from the condenser was determined
by weight pickup by passing a slipstream through a tube
packed with glass wool at room temperature.  The slip-
stream volume was measured with a wet test meter.  The
sulfur on the glass wool was determined by extraction
with carbon disulfide.  The run, which lasted four hours,
operated with an outlet gas concentration near equilibrium.
From equilibrium the sulfur rate in the off-gas from the
condenser is predicted to be 0.0088 and 0.0132 Ib./hr. at
260 and 270°F, respectively.  In the run made, the actual
temperature was 275 * 5°F with a sulfur rate of 0.011
Ib. S/hr. measured.  Even with a predicted accuracy of
5070 for the off-gas sulfur concentration analysis, the
condenser is operating near equilibrium conditions.  The
success of the run is strengthened by the material balance
of 11.3 Ibs. sulfur vaporized into the system over the
four hour period and 11.7 Ibs. sulfur recovered from the
system in the outlet streams.

The safe lower operating outlet gas temperature is about
260°F because the liquid sulfur solidifies at 238°F and a
safety margin of about 20°F is recommended.  Although
there were problems with the sulfur circulating pump during
start-up, it functioned well during the run.

It has been demonstrated that the sulfur condenser can
achieve the desired condensing efficiency of greater than
99%.   The improved performance is the direct result of the
liquid sulfur recirculation system, which provides the
scrubbing action needed to condense the sulfur   mist.
                        224

-------
          Figure 73.   Operating conditions for  sulfur condenser
Cooling HgO In
    270° F
                  Scrubbed Gas
                  0.011 Ib. S/hr.
                    275 * 5°F
                                 Liquid ST  Circulation Rate =0.5 GPM
                        d
                 Sulfur Laden Gas
                  221 CFH @ 70°F
                  2.82 Ibs. S/hr.
                    890 t 30°F
                                        Cooling H20 Out
                                            265°F
                                                                      Recovered
                                                                   -*•  Liquid
                                                                       Sulfur
                                                                   2.9 Ibs.  S/hr.
                          Circulation
                             Pump
                           SULFUR MATERIAL BALANCE
                   Sulfur In

                   Sulfur Out
-  TOTAL  IN

-  Liquid
-  Gas

-  TOTAL  OUT
11.3 Ibs.

11.7 Ibs.
  .04 Ibs,

11.7 Ibs.
                                      225

-------
5.2.7  Fluidizing Mechanics

       There are numerous types of reactors for contacting acti-
       vated carbon with gases.  From the standpoint of smaller
       equipment, of the resulting decreased investment, of high
       heat transfer rates, and good solids flow characteristics,
       multistage fluidized bed reactors were chosen as the type
       of gas/solid contactor.  There are many important parameters
       to be considered in design of the fluid bed reactor.
       The gas/solid residence time requirements have been dis-
       cussed in previous sections as related to the process
       chemistry.

       Westvaco has developed considerable technology in fluid bed
       design and operation.   In addition to the basic process
       development of fluid bed design data, which had been
       developed prior to the present contract, some aspects
       relating to the fluidizing mechanics have been studied
       under the present contract.  Those points are discussed
       below.

       Carbon Fluidization Requirements  -

       As presently conceived,S02 sorption and sulfur generation
       steps will be run in fluidized beds of activated carbon.
       These fluidized beds operate at linear gas velocities
       between the minimum fluidizing (UMf) and entrainment
       velocities (Ut).   Kunii and Levelspiel^ present equations
       for calculating these bed  characteristics, i.e.
           Minimum fluidizing velocity, UMf


             UMf  = -- [(33.7)2 + 4.08(10-2) W^'" P8)g]l/2 . 33.7     (65)
           Terminal velocity,
                         (Ps-Pg)dp]1/2      500 < Rep < 200, 000          (66)
                         Pg
           where  Rep
                                 226

-------
The calculation of minimum fluidizing velocity and entrain-
ment velocity for Westvaco granular carbon, which has an
average particle size of 0.1 cm and a particle density of
1.0 g/enP, is shown in Table 60.
Table 60.  CALCULATED VALUES FOR MINIMUM FLUIDIZING
           VELOCITY AND ENTRAINMENT VELOCITY FOR
           WESTVACO GRANULAR ACTIVATED CARBON
Gas
Air
Air
Air
Hydrogen Sulfide
Flue Gas**
Temp . ,
°F
70
200
300
300
200
Minimum
Fluidizing
Velocity,
ft. /sec .
0.84
0.87
0.75
0.86
0.82
Entrainment
Velocity*,
ft. /sec.
10.7
—
_ _
--
w •_
    *For particle size of 0.042 cm.
   **For gas composition:  76.6% N2, 3.4% 02,
                           6.0% H20, 14% C02
As can be seen from the table, there is very little differ-
ence between the calculated values for air over the range
70-300°F.  In addition there is little difference between
the minimum fluidizing velocities calculated for air, H2S
and flue gas.  Thus fluidization characteristics with air
at 70°F can be used to closely approximate fluidization
parameters for S02 sorption and sulfur generation.  Actual
experimental measurement of the minimum, fluidizing velocity
with Westvaco granular carbon is shown in Figure 74.  The
measured value, taken as the maximum in the curve as
suggested by Levenspiel9 is approximately 0.5 ft./sec.
or about 40% below the calculated value.  This is slightly
outside the range of ±34% in which Levenspiely   says the
values normally fall.  The difference is probably due to
the deviation of the irregular carbon particles from a true
spherical shape.
                       227

-------
       Figure  74.
Experimental determination of minimum
fluidizing velocity for Westvaco
granular  carbon
o
CM
CL
o

o
3
in
t/1

t
a.

•o
01
CO
!.*»
1.2
1.0
0.8
0.6
0.4
0.2
0.1











*














































?






_.


/


~






/









/











*>






y-
/
/
/
/


•



/






Gas
Tern
Avg
Car









Minimum Fluidizing Velocity
A
\
\





\
~»j







•~~.







€






— -



-








oW£







: Air
perature: 70°F
. Carbon Size-: 1.0 mm
ion Particle Density = 1.
.... ..Ill 1
~}J$i5







.--







0 g/cc
                             0.1

                            Air Velocity, ft./sec.
                                1.0
                              228

-------
Also shown in Table   60  is  the  calculated  entrainment
velocity of 10.7 ft./sec.  for a  0.42 mm particle which  is
the smallest particle of  significant percentage in  the
carbon sample.  Thus, for  the Westvaco granular carbon used
in S02 removal, the range  of  operating velocities is  the
0.5 ft./sec.'(experimental value) minimum  fluidizing
velocity and the approximately 10.7 ft./sec.  entrainment
velocity.  In the S02 sorber, it  is desirable  to run  as  high
a velocity as practical in order to keep  the  adsorber
cross-section as small as  possible for treating the  large
volumes of flue gas.  Since  some smaller  particles will be
generated by the bed action, a value of about  8 times the
minimum fluidizing velocity, or  about 4 ft./sec., is used
for the sorber.  Since the sulfur generator will be much
smaller than the sorber,  a value of 4 - 6  times the minimum
fluidizing velocity  (^2.5-3.0  ft./sec.) should be  suffi-
cient to maintain good contact and minimize elutriation.

S02 Sorber Gas Distributor Plate Design -

One of the equipment modifications specified  before  integral
operation was the installation of new gas distributor
plates in the 18"0 S02 adsorber.  The goals for the plates
were to minimize carbon attrition while still maintaining
proper fluidization  for a  high S02 recovery efficiency.

A number of drilled  distributor  plates were evaluated in a
batch, one stage, 18"0 fluid  bed  unit.  The  purpose was  to
provide the operating characteristics of  the  plates to
achieve the goals given above.   The plates  originally in the
18"0 unit were 3.2%  open  area plates with l/8"0 orifices
on a 2/3" equilateral triangular pitch.  The  proposed plates
to reduce carbon attrition were  8% open area. In line with
this, a 2.1% open area plate  with l/8"0 orifices was used
to approximate the operating conditions of  the 3.270 open area
plates.  In addition, two  plates with 87<> open area and  1/4"
and 3/16" orifices were fabricated to test  the effects  of
the higher open area plates  on the S02 removal efficiency.

One operating parameter of prime importance is the pres-
sure drop characteristics  of the drilled  distributor
plates.  The pressure drop characteristics  of the 2.170and 870
open area plates were measured as a function  of the  super-
ficial linear gas velocity.  The pressure drop increases
with increasing velocity  as  shown in Figure 75.    As the
open area increases  the pressure drop decreases and  as  the
orifice diameter decreases for a particular open area the
pressure drop also decreases.  The experimental data compares
favorably with data  taken  from the literature.  In terms of
what can be expected in the  18"0 S02 sorber,  the pressure
drop across the 3.2% open  area plate is about 3" H20 at
                          229

-------
       Figure  75.   Pressure  drop  characteristics of distributor
                      plates  to be used  in  an  18"  diameter  S02  sorber
   100

    80
o
CXJ
1C
O)
-C
u
c
o.
o

o
1/1
VI
    40
    20
10

-------
3 ft./sec. and with an 8% open area plate  at  3  ft./sec.  a
pressure drop near 0.6" H20 would be realized or  a total
pressure drop reduction across the plates  of  a  factor  or 5
or more.

The S02 sorption characteristics of the plates  were deter-
mined by loading the bed with 35 Ibs. virgin  carbon, then
at time zero the simulated flue gas (2,000 ppm  S02,  200
ppm NO. 2 to 2.4% H20, balance air) was introduced into
the fluid bed at 3 ft./sec.  The total sulfur analysis of
the carbon was then followed as a function of time.  The
run was stopped when the acid loading was  above 0.22 Ib.
acid/lb. C.  The results of the runs are given  in
Table  61.   The S02 removal efficiency was 0.24,  0.22,
and 0.15 for the 1/8", 3/16", and 1/4" hole diameter
plates, respectively.  An unanticipated problem occurred,
however, with what is called carbon weepage during
fluidization.  This phenomena is carbon flow  through the
distributor plate, which in a continuous unit would mean
possible short-circuiting of the carbon which could
result in an inefficiency of the S02 sorber.  The phenomena
was only observed, at least to an appreciable extent, for
the high open area plates (870) , with larger orifices.
Nevertheless, on a qualitative basis it is felt that the
S02 sorption efficiency is comparable in light  of this
carbon removal phenomena occurring continuously as  a
function of time.

Since the carbon weepage appeared important in  the gas
distributor design for the 18"0 unit, the  above plates and
a number of other ones were evaluated for  carbon  weepage.
The carbon weepage rate was investigated as a function
of the open area, orifice diameter, superficial linear gas
velocity, and carbon bed loading.  The six plates which
were evaluated are given in Table 62.    The  plates varied
from 2.17. to 8.37o open area with 3/32"0 to  l/4"0 orifices.

The final carbon weepage rate data was subjected  to a
multiple regression analysis.  The data was found to
correlate well as given by Equation (67):
        A -  8.4(10~n) e23.2 do el-95 Ao(-i-)1.3 eQ.12 L       (67)
                                                     r\
     where  A  =  carbon weepage  rate,  Ib.  C/hr.-ft.
            do =  orifice  diameter,  in.
            Ao =  gas distributor open  area,  70
            v  =  superficial  linear gas  velocity,
                    ft./sec.
            L  =  carbon bed loading, Ibs.  C

                        231

-------
                              Table  61.    OPERATING CHARACTERISTICS DISTRIBUTOR PLATES TO  BE  USED
                                              IN  AN  18" DIAMETER S02  SORBER
Run
No.
SA-42
SA-43
SA-44
PLATE CHARACTERISTICS II COLUMN CONDITIONS
Overal 1
Dia.,
in.
18
18
18
Thick.
in.
1/8
1/8
1/8
Triang.
Pitch
in.
0.84
0.82
0.625
Hole
Dia.
in.
1/8
1/4
3/16
No.
of
Holes
440
415
750
Open
Area
%
2.1
^c^nkiri
8.0
8.2
Hole
Velocity
ft. /sec.
142
37
36
Gas
Flow,
CFH
9
70°F
278
278
271
Super-
ficial
Gas
Veloc.
FPS
3.0
3.0
2.9
Gas Concentration
S02
PPM
2,000
2,000
2,000
NO
PPM
200
200
200
H20
Vol. %
2.0
2.2
2.4
Air
Bal.
Bal.
Bal.
Carbon
Type
Virgin
Virgin
Virgin
Carbon
Loaded
Ibs.
35
35
35
	
Run
Time
rain.
240
240
180
Carbon
End
on
Plate,
Ibs.
30.5
10.5
10.3
C in*
Cyclone
Ibs.
2.2
0.6
0.7
C in**
Plenum
Ibs.
***
23.9
24.0
Final
Acid Load
#Acid/#C
0.235
0.253.
0.244
S02
Removal
Efficiency
0.24
0.15
0.22
1TT-W*
Csrsor
Weecsce ;.ate
*/hr.

6.0
8.0
*/hr.-ft.2
...
3.4
4.5
N>
W
N)
   *Cyclone dust given on total  weight basis; was not analyzed for moisture and acid content.
   **By difference since no analysis was made of carbon in plenum.
  ***Was not measured but discrepancy in material balance is believed due to start-up problems in which carbon may have bypassed cyclone.
 ****Taken as constant over time  period of run.                            ,      •
*****Used to approximate 3.21 open area plate.

-------
Table 62.  GAS DISTRIBUTOR PLATE CHARACTERISTICS
           EVALUATED FOR CARBON WEEPAGE DURING
           FLUIDIZATION
Plate
Diameter,
inches
18
18
18
18
18
18
Orifice
Diameter ,
inches
1/8
1/8
5/32
3/32
3/16
1/4
Triangular
Pitch,
inches
0.840
0.667
0.667
0.310
0.820
0.625
Plate
Open Area,
%
2.1
3.2
5.4
8.3
8.2
8.0
                         233

-------
The correlation coefficient (R) for the equation was 0.95,
which indicates a good fit of the model to the experi-
mental data.  Although larger orifice diameters up to
l/4"0 were considered, the operating difficulties during
start-up and shutdown make the 3/16"0 and l/4"0 orifices
unsuitable for the present pilot plant equipment.
Therefore for the present plates, the orifice diameter was
taken to be 1/8".

The carbon weepage rate for the present plates in the 18"0
unit  (L = 30 Ibs., Ao = 3.2%, do = 1.8", and v = 3 ft./sec.)
was predicted to be 6.7(10-6) Ibs. C/hr.-ft.2 or for the
18"0 unit, 1.1(10-5) Ibs. C/hr., which is effectively ;zero.
For the proposed 8% open area plates (1 = 30 Ibs., AQ =
8%, d0 = 1/8", and v = 3 ft./sec.) the predicted carbon
weepage rate is 0.14 Ib. C/hr.  This weepage rate is accept-
able in the upper stages of the 18"0 unit (about 0.5%
of the total carbon flow rate through the unit), but is
felt to be less desirable for the bottom stage in extended
integral runs.  This would correspond to 65 Ibs. C which
would be collected in the inlet gas plenum for the 18"0
unit over a 20 day operation.

To minimize this carbon handling a 6.15% open area plate
with l/8"0 orifices was considered and the predicted
carbon weepage rate was 0.004 Ib. C/hr. or about 2 Ibs. C
over a 20 day period.  This is an acceptable quantity of
carbon but is at the expense of a slightly higher carbon
attrition rate,since the carbon attrition is a direct
function of the per cent open area of the plate.

All of the experimental data led to the design of four
distributor plates in the upper stages of 87o open area
with l/8"0 orifices on a 0.42" equilateral triangular
pitch.  Two of the top uppermost stages are of aluminum
and the other two are of 316 stainless steel.  The dis-
tributor plate for the bottom stage was specified to be
6.15% open area with l/8"0 orifices on a 0.48" equilateral
triangular pitch.  The specifications for these distributor
plates are given in Table 63.
                       234

-------
Table 63.  GAS DISTRIBUTOR PLATES SPECIFICATIONS
           DESIGNED FOR MINIMIZING CARBON
           ATTRITION IN THE 18"0 S02 SORBER
Plate
Diameter ,
inches
18
18
Orifice
Diameter ,
inches
1/8
1/8
Triangular
Pitch,
inches
0.42
0.48
Plate
Open Area,
%
8.0
6.15
                       235

-------
                            SECTION 6
            1,000 MW UTILITY BOILER FLUE GAS  CLEAN-UP
6.1    INTRODUCTION
       The Westvaco Process has been developed over  the  past
       eight years, the last three under joint support from EPA.
       The process is designed to adsorb S02 out of  flue gas
       streams with activated carbon and to produce  elemental
       sulfur as a by-product.  Pilot tests of the complete
       process treating 20,000 cfh of flue gas from  an oil  fired
       boiler have recently been completed under this EPA
       contract, and results are contained in this report.

       With the Westvaco Process, sulfur dioxide is  removed from
       waste gases with activated carbon acting as a catalyst
       and adsorbent in the reaction:
         S02 + 1/2 02 + H20       e»•  H2S04 (Sorbed)     [150-300°F]   (5)
       The sorbed acid is then converted to elemental  sulfur
       by reaction with H2S,  with the carbon again acting  as  a
       catalyst and adsorbent:
         H2S04 -I- 3 H2S          * 4 S  + 4 H20    [200-300°F]   (6)
       The sorbed elemental sulfur is then thermally  stripped
       from the activated carbon at 800-1000°F and  condensed as
       product.  In utility applications where H2S  is unavail-
       able for conversion of sorbed acid to sulfur,  hydrogen
       is added to the gases during the thermal  stripping  step
       to produce the necessary amount of H2S by the  reaction:
        H2 + S  (Sorbed)          ^ ^               [800-1200°?]  (68)
       A schematic of the process is shown  in Dwg.  2563
       (Figure 76).
                              236

-------
                              Figure  76.    Westvaco  S02 Recovery  Process  schematic  flowsheet
N5
                 SOj
               SOj LADEN FLUE 6AS •
                                                   CLEAK TtlJE
                                                  GAS TO STACK
 so,
REHOVAL
                                                                               HYDROGEN
                                                                                            REGENERATED
                                                                                             CARBON
                                                                                                              5Hj3 * HgSC4 3°°*F-1 S I »
 SULFUR
CONDENSER
                                                                                                                                         SPENT GAS
                                                                                                                                         TO BOILER
                                                                                                                                         SULFUR
                                                                                                                                         PRODUCT
                                                                                                                            • Westvaco
                                                                                                                             CHARLESTON RESEARCH CENTER
                                                                                                                             f. 0. MX SM7 WMTH CM1UILESTOM. S. C.
                                                                                                                             WESTVACO SOz .RECOVERY .PROCESS
                                                                                                                                SCHEMATIC. .FLOWSHEET	
                                                                              I JM*> [WO NO
                                                                                                                                         '^oiro   vi
                                                                                                                                         lDv»G. No.

-------
6.2    GENERAL DESIGN BASIS

6.2.1  Scope


       The full scale 1,000 MW installation will consist of an
       integral closed loop system continuously recycling granular
       carbon for removal of S02 from approximately 1,750,000 cfm
       of actual power plant flue gas for a coal fired boiler and
       recovering elemental sulfur as a product.  The flue gas is
       treated after an electrostatic precipitator then returned
       to the boiler stack.

       The 1,000 MW installation consists of the following major
       processing steps:

            1)  S02 removal from flue gas
            2)  Sulfur production from recovered S02
            3)  Thermal stripping of sulfur product and
                  production of hydrogen sulfide
            4)  Condensation and recovery of sulfur product
            5)  Production of chemical reducing gas (hydrogen).

       Pilot plant experiments have established the technical
       feasibility of the process and supplied the design basis.
       Operation of a prototype unit 0-^15 MW) should finalize
       scale-up data needed prior to the full scale installation.
                              238

-------
6.2.2  Boiler  Operating Characteristics

       Coal  Composition -

            Component,  Ash:
                       Sulfur:
                       Hydrogen:
                       Carbon:
                       Nitrogen:
                       Oxygen:
                       H20 in Coal:
            Excess Air:

            Carbon Burned in Coal:
            Heating  Value in Coal:

            Sulfur to  S02:
            Sulfur to  803:

            S02  Removal:

            Flue Gas Temp.  Out
              of Air Preheater:

            Minimum  Gas Temp.  to
              Stack  after Treatment:

            Gas  Velocities in Ducts:
            Plant Size:
            Heat Rate:

            Flue Gas Quantity:
            Flue Gas Molecular Wt.:
            Flue Gas Weight:

       Flue  Gas  Composition -

            Component,  S02:
                       803:
           Nitrogen:
           Oxygen:
           Water:
           Fly Ash:

Water in Combustion
  Air (60% RH @ 80°F)
                            15.2% Dry Wt. %
                            3.5 Dry Wt. %
                            5.0 Dry Wt. %
                            67.2 Dry Wt. %
                            1.6 Dry Wt. %
                            7.5 Dry Wt. %
                            4.8 Ibs./lOO Ibs
                            207o
                                             Wet
                           100%
                           Wet - 11,980 BTU/lb.
                           Dry - 12,580 BTU/lb.
                           98%
                           2%

                           90%

                           300°F


                           200°F

                           6.0 ft./sec.
                           1,000 MW
                           9,000 BTU/KWHR.

                           293,000 moles/hr.
                           29.54
                           8,655,000 Ibs./hr.
                           0.00261
                           0.00005
                           0.13657
                           0.74087
                           0.03276
                           0.08714
                           77,900
 mole fraction
 mole fraction
 mole fraction
 mole fraction
 mole fraction
 mole fraction
Ibs./hr.
                           0.0212 mols/mol dry air
                           751,252 Ibs./hr.
Coal Consumption
  Rate (Wet):
Gas Ducts 300°F or Higher To Be Insulated
                   239

-------
6.2.3  Product
            Type:      Elemental  Sulfur
            Purity:    99.8% or Better  -  Bright  Yellow
            Form:      Liquid at  250°F
6.2.4  Process  Conditions
       SQ2 Sorber  -
            Reaction:    S02 + %  02  + H20  -»•   H2S04
            Heat of  Reaction:  Exothermic  =   -117,000 BTU/mole  S02
            Contact  Mode:    Stagewise Gas/Solid Fluidized Bed
            Fluidizing Gas  Velocity:  4  FPS  (At Base  Load of   . i
                                        1,000 MW)
            Fluidized  Bed Temperature:   150-300°F
            S02 Overall Recovery Efficiency:  907o
            S02 Rate (At Base Load), from  Coal    = 48,943 # S02/hr,
                                    from  Recycle =   7,782 # S02/hr,
                                          TOTAL  = 56,725 # S02/hr,
            Acid Loading on Carbon:  0.22  Ib. H2S04/lb.  C
            Number of  Stages:    5
            Space  Velocity:   2,350  hr.'1
            Carbon Bed Depth/Stage:  11  inches
            Carbon Rate:  43 Tons Carbon/Hr.  (At Base Load)
            S02 Concentration, Inlet (Coal)     = 2,610 PPM
                                     (Recycle)  =  415
                                      TOTAL   = 3,025 PPM
                             . Outlet:   245  PPM
            Boiler Feed H20 Spray Rate:  183 GPM
       Sulfur Production -
            Reaction:    H2S04 +  3 H2S  +  4  S + 4 H20
            Heat of  Reaction:  Exothermic  =  -41,107 BTU/mole H2S04
            Contact  Mode:   Stagewise Gas/Solid  Fluidized Bed
            Fluidizing Gas  Velocity:  3  FPS  (At Base  Load)
            Fluidized  Bed Temperature:   200-325°F
            Inlet  H2S  Concentration:  2  Vol.  %
            Conversions, H2S Utilization:    99.9% of  Inlet H2S
                        S02 Recycle:        15% of  Sorbed H2S04
                        Sulfur  Formation:  70% of  Sorbed H2S04

                              240

-------
     Number of Stages:  11

     Space Velocity:  460 hr.~^

     Carbon Bed Depth/Stage:  17 Inches

     Carbon Rate:  43 Tons/Hr. (At Base Load)

     Outlet H2S Concentration:  270 PPM

Sulfur Stripping/H2S Generation -

     Reaction:  a)  H2 + S  ->  H2S
                b)  H2S04 + 3 H2S  -»•  4 S + 4 H20

     Heat of Reaction:  a) Exothermic = -8,667 BTU/mole H2S04
                        b)  Exothermic = -41,107 BTU/mole H2S04

     Contact Mode:  Stagewise Gas/Solid Fluidized Bed
     Fluidizing Gas Velocity:  3 FPS (At Base Load)

     Fluidized Bed Temp.:  a) Carbon Preheater = 710°F
                           b) H2S Gen./S Strip. = 1000-1200°F

     Inlet H2 Concentration:  19.5%

     Inlet H2 Requirement:  3.3 moles H2/mole S02 recovered

     Carbon Residence Time:  21 minutes

     Design Rates       a) S Stripping   =   796 moles S/hr.
       (At Base Load):  b) H2S Formation = 1,910 moles H2S/hr.

     Number of Stages:  a) Carbon Preheater  =  1
                        b) H2S Formation  =     2
                        c) Sulfur Stripping  =  4

     Space Velocity:  1600 hr."1

Sulfur Recovery  -

     Type:  Shell and Tube Condenser

     Duty:  796 moles S/hr.
     Temperature:  a) Inlet Gas =      1040°F
                   b) Outlet Gas -      250°F
                   c) S Liquid Prod. =  250°F

     Efficiency:  99.9% of Inlet Sulfur

Carbon Cooler -

     Contact Mode:  Gas/Solid Fluidized Bed

     Fluidizing Gas Velocity:  3 FPS

     Fluidized Bed Temperature:  a) Inlet C =    1040°F
                                 b) Outlet C =    300°F
                                 c) Outlet Gas =  300°F


                           241

-------
            Gas Type:  a) Recycled Inert Gas
                       b) 5% Inert Gas Make-up/Cycle
                       c) 5% Boiler Feed H20 Make-up/Cycle

            Number of Stages:   1

       Gasifier -

            Type:  Bituminous  Coal Feed

            Product Gas:   28 cf (H2 + CO)/lb.  Coal
                          65 cf Total Gas/lb.  Coal

6.2.5  Activated Carbon Characteristics

            Type:            Coal Based

            Size:            8x30 M [Nominal;  1.5 MM (Avg.
                                       Particle Size)]

            Density:          40 - 43 Ibs./ft.3

            SC-2 Number:       75 (Minimum)

            Attrition No.:    30 (Maximum)
6.3    CONCEPTUAL DESIGN

       The conceptual design flowsheet for the  process  is  shown
       in Figure 77 (Dwg.  2572)  and described by the  process
       description below.

6.3.1  Process Description

       S02 Sorber - Boiler Flue Gas (FB-101-A,  B, C and D) -
                                                        i
       The boiler flue gas «^p> is desulfijrized  by counter-current
       contact with regenerated carbon <^> in stagewise fluid bed
       reactors (FB-101-A, B, C and D).  The total flue gas
       volume of 105 million SCFH is split among four 55  diameter
       mild steel sorbers  containing 5 stages of fluidized carbon
       for SOX removal.  The bottom stage removes 803 at 300°F to
       prevent acid condensation.  The gas temperature is then
       lowered to 170°F for more efficient S02  removal by direct
       evaporation of water sprayed into the second fluid bed
       stage.  The remaining 4 fluid bed stages, containing 13.5
       inches of activated carbon eachlower the S02 concentra-
       tion to 245 ppm at  the outlet ^£>.  Reaction heat liberated
       during S02 removal  reheats the gas temperature to 200°F.
       The fluidized carbon flows down the column by gravity
       through overflow weirs and downcomers becoming progres-
       sively loaded with  sulfuric acid and leaves the column
       at 300°F containing 22% of its weight in adsorbed acid.
                              242

-------
243

-------
feOiltLR. Ftr->  wATP.lt
 t4> i fiueL e»L
   . ,  ^H
Owm tn i ^r)
                                                                                                      d
                                              243-A

-------
Figure 77.   Westvaco S02  Process flowsheet  for 1,000 MW
              unit (250 MW  typical module  shown)
                                      M/ec.7
                                     TV Pt-tti 4 W
            t/.i'Di* « if. ii srff* flvioiut of a
            e*s. DKf nmiit fV.ua. '/t'Ufi.m.t
            • U t CIMTtXi,
           /«.« MB •MI &ncve 'Jwc 0*0
                             243-B

-------
A blower  (F-101) is included to overcome the gas pressure
drop in each adsorber.  A start-up heater  (A-101) is
included  for pre-heating the sorber.

Sulfur Generator (FB-102-A. B. C and D) -

The acid  laden carbon <^> from the S0£ sorber is contacted
with H£S  at 300°F in four, 11 stage fluidized bed reactors
(FB-201-A, B, C and D) for conversion of the acid
elemental sulfur.  The H2S required for
comes from H2S produced during the stripping step.  The
of f-gas <^> containing possibly S02 evolved or traces of
H2S is recycled to the boiler for conversion to 502 easily
removed in the S02 sorber.  The four reactors are each
11.3' diameter constructed of Alonized 304 stainless steel
throughout.  Activated carbon flows downward from stage to
stage through overflow weirs and downcomers and leaves the
reactor at 300°F <^£> containing approximately 24% of its
weight in elemental and about 3% of its weight in uncon-
verted acid.

H2S Generator/Sulfur Stripper (FB-103-A, B, C and D) -

The sulfur laden carbon <^> from the sulfur generator is
contacted with hydrogen rich gas <^>in the H2S generator/
sulfur stripper (FB-jU)3-A, B, C and D) to thermally strip
the sulfur product ^2^>and to produce a part of the H2S
feed <^>for the sulfur generator.  The reactor is 20'6"
diameter, contains 7 fluidized stages and is constructed
of Alonized 304 stainless steel throughout.  The top
stage of the reactor is used to preheat the carbon to
710°F and is 18'7" diameter.  The remaining 6 stages, at
1000-1050°F, are for sulfur stripping and H2S conversion.
The carbon flows down through the vessel from stage to
stage by gravity through overflow weirs and downcomers.
The regenerated carbon at 1050°F  leaves the vessel
containing 4.1% of its weight in residual adsorbed sulfur
which remains constant throughout the whole loop.
  j
The hydrogen rich gas <4£^is produced from a coal gasifier
[G-102] in series with a water-gas shift converter (CV-101-
A, B,  C and D) for conversion of CO to hydrogen.  Tars are
removed prior to the shift converter.  After CO conversion
the gas is heated to 1040°F in a fired heater [GH-102].

The sulfur product ^^>is separated from the recycle H2S
stream <3£>in a shell and tube condenser (SC-101) ,
filtered for traces of carbon, and sent to a sulfur
storage pit (SP-101).
                        244

-------
      Carbon Cooler  (FB-105-A,  B,  C and D)  -

      After regeneration  the  carbon is  cooled to 300°F by direct
      evaporation  of water  sprayed into the fluidized bed cooler
      (FB-105-A, B,  C  and D).   Recycle  inert  gas is  used as  the
      fluidized  gas.   The 14.5'  diameter cooler contains one
      stage and  is of  Alonized  304 stainless  steel  construction.
      The H20  is condensed  from the recycle gas and  both are
      reused.

      Carbon Handling  and Storage  -

      Activated  carbon is recirculated  through the process loop
      at a rate  of 172 tons per hour.   Storage for  6 hours of
      carbon recirculation  is provided  for  both the  acid laden
      and regeneration loop to  accommodate  variation in waste
      gas rates.   The  storage vessels  (S-101  and -102)  are
      coned bottom tanks  28*0x60'  tall of  mild steel
      construction.  The  total  carbon inventory is 2,000 tons
      and make-up  carbon  is added  at a  rate of 275 Ibs./hr.
      Horizontal and vertical conveying are accomplished by  belt
      conveyors  and  bucket  elevators, respectively.   These are
      not detailed on  the flow  sheet but are  described  in the
      equipment  lists  in  Appendix  K.
6.4   HEAT AND MATERIAL BALANCES

      The heat and material  balances were made  for the overall
      process just described.   The  overall  sulfur balance is
      shown  in Table  64.   The  overall  sulfur balance reiterates
      the 90% sulfur  recovery from  the  flue gas  taken as a basis
      for this particular  evaluation.
              Table 64.   OVERALL SULFUR BALANCE  FOR
                         1,000 MW POWER PLANT
SULFUR IN
Stream
Flue Gas
Gasifier Gas
TOTAL IN
Ibs. S/hr.
24,940
896
= 25,836
SULFUR OUT
Stream
Flue Gas
Elemental S Product
IDS. S/hr.
2,494
23,342
TOTAL OUT = 25.836
                              245

-------
 The overall energy balance is  given in Table 65.   About
 60% of the total heat  input originates with the flue gas,
 indicating the relative energy requirement of the regene-
 ration system.
        Table  65.
OVERALL ENERGY BALANCE FOR
1,000 MW POWER PLANT
ENERGY INPUT
Stream
Flue Gas
Spray H20
AHrxn - SOX Removal
AHSoln - H20-H2S04
AHrxn - S Gen.
AHrxn - H2S Gen,
Sulfur Cond.
Gasifier Gas
Steam for Shift Run
Reducing Gas Preheat
Reducing Reheat
Reducing H20 Cond.
Reducing Reheat
Carbon Cooler H20 Cond.
MM BTU/Hr.
554
57
96
17
24
31
4
1
12
126
60
98
14
74
%
48.7
2.4
8.4
1.5
2.1
2.7
0.4
0.1
1.1
11.1
5.3
8.6
1.2
6.5
TOTAL IN = 1,138 MM BTU/Hr.
ENERGY OUTPUT
Stream
Flue Gas
Spray H20 Vap.
Reducing Gas
AHsoln H20-H2S04
H20 Vap. - H20-H2S04
Sulfur Vap.
AHrxn " H2S Generator
Sulfur Product
Sulfur Condenser
Carbon Cooler - H20 Vap.
Carbon Cooler - H20 Rec.
H20 Condenser
H20 Condensed from R.G.
Heat Losses - Carbon Storage
MM BTU/Hr.
284
367
•13
17
16
4
6
2
103
83
96
112
3
32
%
25.0
32.2
1.1
1.5
1.4
0.4
0.5
0.2
9.1
7.3
8.4
9.8
0.3
2.8
TOTAL OUT = 1,138 MM BTU/Hr.
The detailed heat content and stream compositions are
given in Tables 66-75.
                       246

-------
                                     Table  66.   STREAM CONDITIONS
N>
-P-

S02
H2S
H2S04
Elemental S
02
H20
N2
C02
H2
S03
CO
CH4
C2H6
Residual S
Carbon
TOTAL

M SCFM
Temp., °F
Press., in. H20
Heat, MM BTU/Hr.


STREAM 1
Ibs./hr.

















1,753
330
+40
554


moles/hr.
844



9,612
25,520
217,112
40,140

16





293,284







STREAM 2
Ibs./hr.













10
258
268

1,850
200
+10
284


moles/hr.
78



9,208
43,388
217,112
40,140







309,926







STREAM 3
Ibs./hr.













13,664
341,556
355,220

—
. 77
—
0


moles/hr.























STREAM 4
Ibs./hr.


78,944


14,276







13,654
341 ,298
448,222

—
300
—
26


moles/hr.
























-------
                                      Table 67.  STREAM CONDITIONS
ro
-P-
oo

S02 	
~H2S ~
H2S04
Elemental S
02
H20
N2
C02
H2
S03
CO
CH4
C2H6
Residual S
Carbon
TOTAL

M SCFM
Temp., °F
Press., in. H20
Heat, MM BTU/Hr.


STREAM U-l
Ibs./hr.




364,960









364,960

—
150
___
27


moles/hr.
^ -^•••^^^^-^^•^h^V^bM





















STREAM 5
Ibs./hr.
^^MW^H^^»^H^^^»_M















11.7
325
59
15.9


moles/hr.
1,850



680
3,036
1,872


132
160
28


7,846







STREAM 6
Ibs./hr.
• ^iii















13.5
325
4
12.5


moles/hr.
102
2.1



3,688
3,036
1,872


132
160
28


9,037







STRE/'M 7
Ibs./hr.

11.916
77.769

2,264







13,654
341,287
446,890

—
300
—
26.4


nnles/hr.
























-------
                                   Table  68.   STREAM CONDITIONS
-P-
vo

S02
H2S
H2S04
Elemental S
02
H20
N2
C02
H2
S03
CO
CH4
C2H6
Residual S
Carbon
TOTAL

M SCFM
Temp., °F
Press., in. H20
Heat, MM BTU/Hr.


STREAM 8
Ibs./hr.



75.824









13,654
341 ,287
430,765

— _
710
—
76.4


moles/hr.























STREAM 9
Ibs./hr.













13,654
341,281
354.934

—
1040
—
112


moles/hr.























STREAM 10
Ibs./hr.

















79.9
1300
109
139


moles/hr.

28



5.496
3,036
1 ,872
2.608

132
160
28


13,360







STREAM 1-1
Ibs./hr.

















81.8
710
105
70.0


moles/hr.
182
28



5^3J__,
3.036
1,872
2,608

132
160
28


13,664








-------
Table 69.   STREAM CONDITIONS

S02
H2S
H2S04
Elemental S
02
H20
N2
C02
H?
$63
CO
CH4
C2H6
Residual S
Carbon
TOTAL

M SCFM
Temp., °F
Press. , in. H20
Heat, MM BTU/Hr.


STREAM 1-2
Ibs./hr.

















81.8
1200
91
130


moles/hr.
182
28



5,616
3,036
1,872
2.605

132
160
28


13,664







STREAM 1 1
Ibs./hr.



23.344









0.2
6


78.7
1040
67
119


moles/hr.

1.850



5,984
3,036
1,872


132
160
28


13.150







STREAM 12
Ibs./hr.

















78.7
250
63
18.4


moles/hr.

1,850



5,984
3,036
1,872


132
160
28


13,150







STREAM 13
Ibs./hr.













^_



46.9
110
61
2.1


moles/hr.

1,850



680
3,036
1,872


132
160
28


7,846








-------
                                  Table 70.  STREAM CONDITIONS
to

S02
H2S
H2S04
Elemental S
02
H20
N2
C02
H2
S03
CO
CH4
C2H6
Residual S
Carbon
TOTAL

M SCFM
Temp., °F
Press., in. H20
Heat, MM BTU/Hr.


STREAM 1-3
Ibs./hr.



23,344










1
25,457

___
250
—
1.6


moles/hr.























STREAM 14
Ibs./hr.



23,344











25.456

___
250
—
1.6


moles/hr.























STREAM 15
Ibs./hr.













13.653
341,281
354.934


300
—
19.4


moles/hr.























STREAM 16
Ibs./hr.













3
275
278

—
77
—
0


moles/hr.




















I
_ .

!
s

-------
                                     Table 71.   STREAM CONDITIONS
to
Ul

S02
H2S
H2S04
Elemental S
02
H20
N2
C02
H2
S03
CO
CH^
C2H6
Residual S
Carbon
TOTAL

M SCFM
Temp., °F
Press. , in. H20
Heat, MM BTU/Hr.


STREAM 17
lbs./hr.

















2.0
77
2
0


moles/hr.





12
258
64







334







STREAM 18
lbs./hr.





3.676









3.676

__ _
150
—
0.3


moles/hr.























STREAM 1-4
lbs./hr.

















77.0
150
9
7.1


moles/hr.





3,7fiQ
7,280
1,812







12.860







STRE/'M 19
Ibs./hr.

















101
300
4
29.5


molfis/hr.





7,855
7,280
1.812







16,947








-------
                                   Table 72.  STREAM  CONDITIONS
fo
Ul


SO 2
H2S
H2S04
Elemental S
02
H20
N2
C02
H2
S03
CO
CH4
C2H6
Residual S
Carbon
TOTAL

M SCFM
Temp., °F
Press., in. H20
Heat, MM BTU/Hr.

Tons Coal/Hour
STREAM 20
Ibs./hr.





69,864
..








69.864

—
160
_ __
5.6


moles/hr.























STREAM 21























17.7
























STREAM 22
Ibs./hr.

















23.5
—
___
—


moles/hr.




824
3,100









3,924







STREAM 23
Ibs./hr.

















38.3
100
2
1.0


moles/hr.

28



90
3,036
326
1,062

1,678
160
28


6,408








-------
                                      Table 73.   STREAM CONDITIONS
N>
Ln

S02
H2S
H2504
Elemental S
02
H20
N2
C02
H2
S03
CO
CH4
C2H6
Residual S
Carbon
TOTAL

M SCFM
Temp., °F
Press. , in. H20
Heat, MM BTU/Hr.


STREAM 24
Ibs./hr.





125,136









125,136

41.6
280
—
ll.fi


moles/hr.























STREAM 25
Ibs./hr.

















79.9
200
166
12.0


moles/hr.

28



7,042
3.036
326
1,062

1,678
160
28


13,360







STREAM 26
Ibs./hr.

















79.9
675
155
63.2


moles/hr .

28



7,042
3.036
326
1,062

1,678
160
28


13,360







STRE/'M 26
Ibs./hr.

















79.9
710
150
68.6


nn^es/hr.

28



5,496
3.036
1,872
2,608

132
160
128


13,360








-------
                                    Table 74.  STREAM CONDITIONS
to
Ul
m

S02
H2S
H2S04
Elemental S
02
H20
N2
C02
H2
S03 '
CO
CH4
C2H6
Residual S
Carbon
TOTAL

M SCFM
Temp., °F
Press., in. H20
Heat, MM BTU/Hr.

1 No. 2 Fuel Oil/Hr.
STREAM 28























1.027
























STREAM 29
Ibs./hr.

















4.0
77
2
0


moles/hr.





24
516
128







668







STREAM 30
Ibs./hr.






















4,146
moles/hr.























STREAM 31























3,592

























-------
Table 75.  STREAM CONDITIONS

S02
H2S
H2S04
Elemental S
02
H20
N2
C02
H2
S03
CO
CH4
C2H6
Residual S
Carbon
TOTAL

M SCFM
Temp., °F
Press., in. H20
Heat, MM BTU/Hr.

#No. 2 Fuel Oil/Hr.
STREAM 32
.















*






4,847
























STREAM
Ibs./nr.





















-

moles/hr.























STREAM
Ibs./hr.























moles/hr.























STREAM
!bs./hr.























molejj/Jir.
























-------
6.5    COSTS OF 1,000 MW CONCEPTUAL DESIGN INSTALLATION

6.5.1  Cost Summary

       Based on the conceptual design flowsheet, heat and
       material balances, the costs estimated for installation
       of the Westvaco Process on a 1,000 MW boiler are
       summarized below:
                     Table 76.  COST SUMMARY
Capital Cost
£ Million
35
S/KW
35
Operating Cost
' $ Million/Yr
14.6
Mil/KWH
2.0
       Details of the estimate are discussed in the following
       sections and back-up information is contained in
       Appendix K.

6.5.2  Capital Costs

       The conceptual design flowsheet was used to estimate
       the cost of installing a Westvaco Process in a 1,000 MW
       power boiler.  The summary is given in Table 77.

6.5.3  Equipment Costs

       Purchased equipment cost estimates were  based on actual
       vendor quotes when available.  When quotes were not
       available, standard engineering estimating procedures
       were used.  This led to a basic purchased equipment cost
       The total direct installation cost was then obtained by
       factors given by Miller10.   The detailed estimate is
       given in Appendix K.
                              257

-------
                  Table 77.   CAPITAL COST SUMMARY


         Basic Equipment Costs                    $ 7,278,100
         Equipment Installation Costs                7,350,900
         Cost of Additional Battery Limit Items        5,611,800
            TOTAL DIRECT COST - BATTERY LIMIT        20,240,800

         Auxiliary Costs (Storage, Auxil., Serv.)      1,619.200
            TOTAL BATTERY LIMIT + AUXILIARIES        21,860,000

         Catalyst Costs (Carbon + Shift Catalyst)      1.703,200
            TOTAL DIRECT COST                     $23.563.200
         Engineering &  Supervision Cost             $ 3,298,800
         Construction Cost                          2,356,300
         Contractor's Fee                             942,600
         Contingency                               4,712,600

            TOTAL INDIRECT COST                    $11,310,300
                 TOTAL  INSTALLED COST =
$34,873,500
6.5.4  Indirect  Costs

       The indirect costs are made up of engineering, construc-
       tion,  contractor's fee,  and an estimating contingency.
       Factors for obtaining these costs are  given by Peters  and
       TimmerhausH.   The factor used for contingency was more
       than  twice that suggested by Peters and Timmerhaus.

6.6    OPERATING COSTS OF 1,000 MW CONCEPTUAL DESIGN INSTALLATION

       The annual operating costs for the Westvaco Process  at a
       1,000 MW  installation is about $14.6 million/year  or 2.08
       mills/KWH.  The operating utilities estimates were based
       on operation of the installation on an 8070 yearly  basis.
       These operating basis and capital cost change factors were
       the same  as those used by M. W. Kellogg in an earlier
       comparison of the process to other S02 removal processes.
                                258

-------
The annual operating cost is given in Table 78.  The
total direct costs were $7.3 million/yr. if no sulfur
credit is taken and $5.5 million/yr. if minimal credit
were taken.  The indirect cost was about $1 million/yr.
and the fixed cost was $6.3 million/yr.  This led to gross
operating costs of $14.6 million/yr. or 2.08 mills/KWH if
no sulfur credit is taken.  The net operating cost is
$12.8 million/yr. or 1.83 mills/KWH if minimal sulfur
credit is taken.
                         259

-------
                    Table  78.   ANNUAL  OPERATING  COSTS
                     PLANT SIZE  (MW):                      1,000
                     FIXED CAPITAL  INVESTMENT  (FCI):     $34,873,500
                     STREAM TIME  (HRS./YR.):              7,000
  DIRECT  COST
                                                                               $/Year
                                                                           $40/Lb. Carbon
        1.  Operating  Labor (4 Men/Shift @ $5.50/Hr.)                               154,000
        2.  Supervision - 15% of Item  1                                             23,100
        3.  Maintenance, Labor & Materials - 4% of FCI                            1,394,900
        4.  Plant Supplies - 15% of Item 3                                         209,200
        5.  Utilities
           a.  Cooling Water     - 1,430 GPM  $0.1Z/Gal.                        1,898,400
           e.  Coal              - 17.6 TPH P $12/Ton                            1,478,400
        6.  Chemicals & Raw Materials
           a.  Activated Carbon -  275 Ibs./hr.  @ $0.40/lb.                         770,000

        7.  SUBTOTAL Direct Cost (Excl. Credits)                                  7,303,200

       8.  Credits
           a.  Sulfur - 305 TPD @  $20/Short Ton                                 -1,779.200

       9.  TOTAL  DIRECT COST                                                   5,524,000


 INDIRECT  COST

      10.  Payroll Overhead - 20%  of (1+2)                                          35,400
      11.  Plant  Overhead - 50% of (1+2+3+4)                                       890,600

      12.  TOTAL  INDIRECT COST                                                   926,000


 FIXED COST

      13.  Capital Charges  -  18.22% of FCI  (Includes Depreciation  Interim         6,354,000
             Replacements,  Insurance, Taxes  and  Cost of Capital)


TOTAL OPERATING COST
      14.  Net Production Cost - Items (9+12+13)                                12,804,000


UNIT  PRODUCTION COST
      15.  Gross  - Items  (7+12+13)                                             14,583,200
           a.  Mills/KWH                                                           2  08
           b.  $/Ton  Sulfur Not Emitted                                           163!g

      16.   Net -  Items (9+12+13)
           a.  Mills/KWH                                                           !  83
           b.  $/Ton  Sulfur Not Emitted                                           143*9
                                        260

-------
                            SECTION 7
                      15 MW DESIGN AND COST
7.1    INTRODUCTION
       The technical feasibility of the Westvaco Process has
       been demonstrated in a 20,000 cfh integral pilot unit.
       A^detailed assessment of the process indicated economic
       viability and competitiveness with other systems proposed
       for sulfur gas control.  The following section outlines
       the program and cost to test the process on a large scale
       in an actual utility.  The test program is designed to
       generate operating data for more detailed process assess-
       ment and to demonstrate reliability and compatability
       with utility operation.


7.2    SCOPE OF THE PROTOTYPE PROGRAM

       The technical feasibility of the Westvaco Process has been
       demonstrated in integral pilot plant tests.  In order to
       obtain engineering information for a detailed economic
       assessment and for process scale-up, a prototype program
       is'proposed for testing at the 15 MW (30,000 cfm) level.
       The scope of the program is to assess the process perform-
       ance and to obtain engineering data on the major questions
       related to:

            1.  Compatability with the boiler interface
                a.  turndown, upsets and outage in boiler
                b.  reliability
                c.  fuel feed variations
                d.  safety

            2.  Control of process chemistry
                a.  response to upsets and variations
                b.  long term stability

            3.  Performance of activated carbon
                a.  mechanical
                b.  chemical

            4.  Performance of large scale fluid bed vessels
                a.  solid distribution
                b.  gas distribution
                c.  ease of operation.
                              261

-------
       In attaining these objectives  the scope of this program
       includes:

            1.   Definition of boiler  operating characteristics
            2.   Preparation of detailed prototype test program

            3.   Preparation of prototype design specifications

            4-   Detailed engineering  design and bid evaluation

            5.   Construction

            6.   Start-up
            7.   Operation
            8.   Data evaluation and process technical  and
                economic assessment.

       As an adjunct to  the prototype program the scope of work
       will also include,  as necessary,  additional testing in  the
       current  pilot and other equipment to refine the prototype
       design as seems necessary.


7.3    DESCRIPTION OF PROTOTYPE PLANT AND OPERATION

7.3.1  General

       The prototype plant will consist of an integral closed  loop
       system continuously recycling  granular carbon for removal
       of S02 from approximately 30,000 cfm of actual  power plant
       flue gas from a coal fired boiler and recovering elemental
       sulfur as a product.

       The prototype size is equivalent to about 15 MW of the
       boiler capacity with design capabilities to handle 10 - 20 MW.
       The flue gas will be withdrawn between an electrostatic
       precipitator and  the boiler stack.   The desulfurized flue
       gas from the prototype plant will be returned to the stack.

       The prototype plant consists of the following major proces-
       sing steps:

            1.   S02 removal from flue gas.

            2.   Sulfur production from recovered S0£
            3.   Thermal  stripping of  sulfur product and produc-
                tion of  hydrogen sulfide

            4.   Condensation and recovery of sulfur product

            5.   Production of chemical reducing gas (hydrogen).
                                 262

-------
       Pilot plant  experiments  have established the technical
       feasibility  of the process and supplied the data base for
       the prototype  design basis.   In the general design  basis
       that follows,  specifications are given to accommodate the
       range of  conditions anticipated in the test program and
       demonstration  run.

7.3.2  General Design Basis

       Boiler Characteristics -

           Size Equivalent:     15  MW (30,000 scfm equivalent)
           Type of Fuel:    Coal

           Sulfur  Content:    1-4%
           Load Variation:    10 - 20 MW

           Flue Gas:    20,000-40,000 scfm; 800-3100 ppm S02

       Process Conditions -

       S02  Sorber -

           Reaction:   S02 + % 02 + H20  ->  H2S04

           Heat Release:    117,000 BTU/mol S02
           Contact Mode:    Gas/Solid Fluidized Bed,  Stagewise

           Fluidizing Gas Velocity:    2-4 ft./sec.

           Space Velocity:    2,350 hr."1 (Design Based on
                                 Experimental Rate Model)
           Fluidized Bed Temperature:  150 - 300°F

           Carbon  Feed Rate:    4.4-8.8 M Ibs./hr.
           SC-2  Rate:    250-1,000 Ibs./hr.

       Sulfur Production -

           Reaction:    3 H2S + H2S04  +  4 S + 4 H20

           Heat Release:    41,107 BTU/mol S

           Contact Mode:    Fluidized Bed, Stagewise

           Inlet Gas:    50,800 scfh with 22% H2S
           Gas  Velocity:    3 ft./sec.

           Space Velocity:    475 hr."1 (Design Based on
                                 Experimental Rate Model)

           Temperature:    250 - 325°F

           Carbon  Feed Rate:    6,600 Ibs./hr. (Average)
                               263

-------
     Carbon Residence Time:    40 minutes
     Sulfur Formation Rate:    1,328 Ibs./hr.
Sulfur Stripping/H2S Formation -
     Reaction:    a)  H2 + S  •>  H2S
                  b)  3 H2S + H2S04  •*  4 S + 4 H20   (Overall)
     Heat Release:    a)  8,667 BTU/mol S
                      b)  41,107 BTU/mol S
     Contact Mode:    Fluidized Bed, Stagewise
     Inlet Gas:    92,600 scfh with 19.4% H£
     Gas Velocity:    3 ft./sec. @ 1000°F
     Space Velocity:  1600 hr.'r for C preheat and S  stripping
                      3800 hr."1 for H2S formation
     Temperature:  715°F (Preheat);  1000°F (Stripping)
     Carbon Feed Rate:    6600 Ibs./hr. (Average)
     Carbon Residence Time:   21 minutes [3.5 min. (Preheat) +
                                17.5 minutes (Stripping)]
     Design Rates:     419 Ibs./hr. (S Stripping)
                       187 cfm (H2S Formation)
Sulfur Recovery -
     Type:    Condenser, Shell and Tube
     Duty:    500 Ibs. S/hr.;  2 MM BTU/hr.
     Temperature:    940°F, Inlet Gas
                     260°F, Outlet Gas
                     260°F, Liquid Product
     Efficiency:    99.5% of Inlet Sulfur
Gasifier -
     Type:    Coal Feed with Steam/Air Blast
     Feed:    716 Ibs./hr.  Bituminous Coal  (Maximum)
              716 Ibs./hr.  Anthracite Coal  (Maximum)
     Product Gas:    28 cf H2 + CO/lb. Coal
Activated Carbon Characteristics -
     Size:    8x30 Mesh (Nominal; 1.5 MM Avg. Particle  Size)
     Density:    40-43 Ibs./ft.3
     S02 Number:    75 (Minimum)
     Attrition No.:    30 (Maximum)
                          264

-------
       Instrumentation -

       Adequate instrumentation is included,  based on pilot plant
       experience,  to control and monitor temperature, pressure,
       gas  and carbon flow,  and gas composition.   Additional
       instrumentation is also included in the design to assess
       the  various  methods of overall process control under the
       varying modes of utility boiler operation.

       In addition  to the major equipment items,  auxiliary equip-
       ment such as blowers, heat exchangers, conveyors, dust
       collectors,  etc. are included in the final  design to
       accommodate  the range of operating conditions defined for
       the  major equipment items.  Details of these items are
       listed in Appendix L.

7.3.3  Process Description - Prototype Plant (Dwg.  2573,  Fig.  78)

       Flue gas from the precipitator at 300°F passes first
       through the  flue gas blowers (F-101) where  the pressure  is
       boosted and  then into the S0£ adsorber (FB-101).   The
       adsorber is  16 feet in diameter and contains five fluidized
       stages of activated carbon.  Sulfur trioxide is removed
       from the hot gases in the bottom stage of the adsorber.
       Water sprays above the second stage cool the gas to 170°F
       prior to completion of the 90% sulfur oxide removal in
       this and the remaining three stages.  Conversion of sulfur
       dioxide to sulfuric acid during sorption by the activated
       carbon reheats the flue gas <3> to 200°F.   A cyclone
       collector (M-101) removes entrained carbon dust from the
       flue gas prior to its return to the stack.   The normal
       dust loading of the clean flue gas from the cyclone would
       be about .01 gr/scf with a maximum expected of .02 gr/scf.
       A baghouse is included to measure the efficiency of the
       cyclone and  evaluate the total attrition rate.  Recycled
       activated carbon  is fed to
       the  sulfur generator  (FB-102) by a bucket elevator  (V-101)
       and feeder  (V-201).  In the sulfur generator  (FB-102), a
       3' diameter, 11 stage fluid bed, the  sorbed  sulfuric acid
       is converted to elemental  sulfur by reaction with hydrogen
       sulfide<|>at 300°F.  The heat of reaction  is dissipated
       by interstage water sprays as necessary.  The  activated
       carbon becomes progressively loaded with elemental  sultur
                               265

-------
                                  TTMf ^ Bottflt Tl/tM- \     /
                                  AevMtf  5l#HSLATt»ti    \    /
                                                  S  GENERATOR. I
                                                       STRIPPER
266

-------
         Figure 78.   Westvaco  S02 Process  flowsheet for 15 MW
                         prototype unit
n 10, me**t* •*
                                                            CHARLESTON RESEARCH CENTER
                                                            f. 0. BOX 5207 NORTH CHARLESTON. S. C.
                                                             IVfSrttVO SO, PROCESS FLOtvSHeCT
                                                             fcue  ifMH f>*oTOTrr>e UNIT
                                      26 6-A

-------
       as it flows by gravity through the sulfur generator and
       exits with a loading of 20 Ibs.  S/100 Ibs.  C<7>.  The
       additional sulfuric acid not converted to sulfur in this
       reactor is converted in the H2S  generator/sulfur stripper.

       The elemental sulfur product is  recovered by thermally
       stripping concurrently with H2S  formation in the H2S
       generator/sulfur stripper (FB-104).   The H2S generator/
       sulfur stripper is a 70" diameter, 7 stage  fluid bed with
       1 stage for carbon preheat,  2 stages for H2S formation,
       and 4 stages for stripping.   The carbon is  preheated with
       the reducing gas at 1300°F containing about 20 vol.  % H2-
       The carbon flows from the preheat stage at  715°F, bypas-
       sing two stages for H2S formation, to four  stages for
       sulfur removal at 940 - 1000°F.  The  off-gas from the
       carbon preheater is reheated to  1300°F and  passed to the
       bottom of the H2S generator/sulfur stripper where it con-
       tacts counter-currently the carbon to thermally strip the
       sulfur.  The hydrogen and sulfur subsequently  are con-
       verted to the H2S necessary for  sulfuric acid  conversion
       in the sulfur generator.  Final  conversion  of  any remain-
       ing acid to sulfur also occurs in the sulfur stripper/H2S
       generator.  The gas leaves the H2S generator/sulfur
       stripper at 940°F and passes to  a shell and tube sulfur
       condenser to recover the elemental sulfur.   The liquid
       sulfur at 270°F is filtered of any dust before being
       solidified by a sulfur flaker.  The  cooled  gas at 270°F is
       effectively free of sulfur.   The regeneration  gas is
       supplied by a gas producer capable of gasifying anthracite,
       coke, charcoal, or bituminous coal.   The gas from the
       gasifier passes through a shift  converter before being
       used in the carbon preheater-H2S generator/sulfur stripper.

       Regenerated carbon <^>  from the H2S generator at 1000°F
       is cooled to 300°F in the carbon booler (E-501).  Cooling
       is by evaporation of water sprayed over the single bed of
       fluidized carbon in the 70" diameter unit.   Superheated
       steam used as the fluidizing gas passes through a cyclone
       and is exhausted.

       The regenerated and cooled carbon 
-------
7.3.5  Start-up and Initial Operation

       The operating schedule for the prototype plant calls for a
       three month period of start-up, a three month test program
       and a six month demonstration run.

       The start-up period will be used to work through the plant
       putting all units into operation, checking their opera-
       bility over the specified temperature, flow and pressure
       ranges, and making any required adjustments, modifications
       or replacements.  At the conclusion of the start-up period,
       the plant will be capable of accepting flue gas and circu-
       lating carbon through the adsorption and regeneration
       equipment.

       The initial operating period will be devoted primarily to
       establishing the operating characteristics of the process.
       A material and heat balance will be obtained around the
       plant which will be checked against calculated values to
       determine if any significant deviations are occurring.
       The S02 removal capability on both a once-through basis and
       as a function of the number of cycles for a limited number
       of cycles will be determined.  The process control charac-
       teristics,  particularly stability and turndown capability,
       will be tested.  These data will be analyzed as they are
       obtained.  At the conclusion of the initial operating
       period, the data will be reviewed and any changes needed in
       the program for demonstration operation will be made.

7.3.6  Demonstration Operation

       The demonstration operating period is intended to show the
       capability of the Westvaco Process to operate reliably
       under actual industrial conditions.  The level of staffing
       of the plant will be reduced to that anticipated
       commercially.  Boiler operating personnel will be used to
       the greatest extent possible.  The primary responses being
       monitored during this period will be catalyst activity and
       attrition characteristics, process stability and control.
       Process operating information will be obtained for use in
       scale-up design.

       The equipment will be inspected and photographed.  The
       condition of all critical components will be noted.

7.3.7  Technical and Economic Review of Operation

       After the initial and demonstration runs in the prototype
       plant, a technical and economic review will be made of the
       process.  All pertinent information will be incorporated
       into a final report covering the technical data and present-
       ing an economic evaluation of the process.


                                268

-------
7.4    TECHNICAL APPROACH

7.4.1  General

       The  overall intent of the program is to assess the Westvaco
       Process  on a 15 MW slipstream of a coal fired boiler.
       Prior to the engineering of the unit, Westvaco will prepare
       a detailed program for the equipment operation and process
       design specifications for the unit.   These will be used by
       an applications-engineering firm to prepare the detailed
       design of the system and obtain bids for fabrication and
       erection.   After the acceptance of bids,both Westvaco  and
       the  engineering company will monitor fabrication and
       erection.   Prior to start-up, Westvaco will train technicians
       for  operation during the test program and operators to be
       supplied by the utility who will operate the unit during
       the  demonstration period.  Data reduction,evaluation and
       process  assessment will be performed jointly by Westvaco
       and  the  applications-engineering firm.

       During the construction of the prototype unit,  tests will
       be made  to define fully the operating characteristics  of
       the  boiler as they would affect process operation.

       As an adjunct to the prototype design,  tests will be made
       as necessary to evaluate proposed control modes,  operating
       ranges and design features of the prototype unit.   Input
       from these will be used to modify the prototype design as
       required.

       The  prototype program is conceived to operate the process
       under actual boiler conditions for sufficient cycles to
       obtain information for process assessment and design of a
       large scale unit.

7.4.2  Description of Program Elements

       Based on pilot plant data"and boiler operating  character-
       istics a detailed test program was prepared for prototype
       operation  as given in Figure 79.   The test program will
       include  the assessment of the effects of flue gas composition,
       reducing gas composition temperature,  boiler turndown,  etc. on
       the  operating characteristics of the unit.   Data obtained will
       be used  to modify design procedures as necessary for
       process  scale-up and assessment and to supply the input
       for  designing additional tests as deemed necessary.  The
       program  will also include a demonstration run to assess
       longer term reliability.
                                269

-------
                              Figure  79.   Prototype program schedule
c
1. PREPARE DETAILED TEST PROGRAM.
2. PREPARE PROCESS DESIGN SPECIFICATIONS.
3. DETAILED ENGINEERING DESIGN & BIDS
H. PROCUREMENT AND CONSTRUCTION
5. DEFINITION OF BUILER OPR. CHARACTERISTICS
.6. OPERATION
A, OPERATOR TRAINING
B. START-UP
c, TEST PROGRAM
D. DEMONSTRATION RUN
7. DATA EVALUATION AND PROCESS ASSESSMENT
MONTHS
1 2 4 6.8 10 12 14 16 18 20 22 24 26 28 30 32 34 36
1 1 I

N-

•

-
i i i i i

•4





i i i i i
•






i i i i i


, ...


M

i i i i i
	 —



•_
*B
*B
i i i i i




— '
P *







ro
-vj
o

-------
Based on*the proposed test program and the pilot plant data
available, detailed process design specifications for the
prototype unit will be prepared by the process developer.
This will include the general design basis, process descrip-
tion, heat and material balances, process equipment specifi-
cations, instrument list, proposed layout and utilities
requirements.  This will be assembled in report form for use
by the applications-engineering firm to prepare a detailed
design.

Using the process design specifications report from Westvaco,
the  applications-engineering company will prepare a detailed
design of the prototype unit and  secure bids for its fabri-
cation and construction.

The  engineering contractor will procure and construct the
prototype unit.

During the construction phase an  analytical program will be
conducted to define the characteristics of the boiler
operation.  This will include:  coal composition, flue gas
composition and temperature and load variation.  Tests will
be conducted at intervals in order to establish trends.

During the construction phase operating manuals will be pre-
pared and training programs conducted to familiarize
development support personnel and selected plant operators
with the operation of the recovery process.

The  start-up will include sequential testing of the proto-
type equipment to achieve design operating characteristics.
Modifications will be made as necessary.   Operators will be
given actual experience in equipment operation.

The  test program will evaluate steady state and transient
behavior of the system with variations in flue gas composi-
tion, reducing gas composition, temperature, boiler load,
etc.  An anticipated 100 - 125 cycles will be completed
during the test program.

A six month demonstration run will be conducted to define
the  long term reliability of the  process.  Over the six
month period 300 - 360 process cycles would be completed
and  there would be 1.25-1.5 complete inventory turn-overs.

Beginning with the test program there would be continuing
analysis of data.  This would be  incorporated in a
detailed assessment at the end of the test program which
would be updated after the demonstration run.
                        271

-------
7.5  COST OF PROTOTYPE PROGRAM

     The summary of costs for the prototype  program is  shown
     below in Table 79:
              Table 79.   COST OF PROTOTYPE  PROGRAM
         1.   Installed Equipment                $2,411,000
         2.   Design Engineering                    240,000
         3.   Manpower, R&D            $303,246
                      Operation       78,300
                      Maintenance     63.288      444 , 834

         4.   Overhead                              422,600
         5.   Consultants                            14,400
         6.   Travel                                 14,400

         7.   RawMat'l., Utilities
              &  Supplies
                  TOTAL COST                   $3,868,994
       The  total prototype cost was estimated as follows:

       Purchased equipment costs were estimated using general
       engineering methods and budget quotes from the equipment
       lists  contained  in Appendix  L  .  The total installed
       equipment cost was derived from the purchased equipment
       cost based on the factors given by Miller10.  The cost
       breakdown is shown in Appendix L.

       Engineering costs were estimated from factors presented
       in Peters and Timmerhaus11.

       The  required manhours for R&D, operators and maintenance
       were derived using the development schedule and assigning  •
       manhours based on experience in Westvaco's prior development
       work and plant experience.  The manhour breakdown and cost
       breakdown are shown in Appendix L.

       A provisional overhead on manpower costs was estimated at
       95%  of direct labor costs based on past experience.

       Raw  materials and utilities were determined from energy
       and  material balances for the prototype unit and applied
       to the number of operating hours at rates shown in
       Appendix L.
                              272

-------
                          SECTION  8
                        BIBLIOGRAPHY
 1.  Russel, J. H., Town, J. W., and Kelly, H. J.,  U.  S.
     Bureau of Mines Report of  Investigation  7415  (August
     1970) .

 2.  Juza, R. and Blanke, W., Z. Anorg. Chem. 210,  81  (1933).

 3.  Tuller, W. N., Ed., THE SULFUR DATA BOOK, McGraw-Hill
     Book  Co., New York, 1954.

 4.  Aynsley, E. D., Pearson, T. G., and Robinson,  P.  L.,
     J. Chem. Soc.. 58-68 (1935).

 5.  Norrish, R. G. W., and Rideal, E. K., J. Chem. Soc. 123,
     1668-1704 (1923).                     	

 6.  Hart, P. J. , Vastola, F. J. and Walker, P. L. , Jr'. ,
     Carbon 5, 363-371  (1967).

 7.  Bansal, R. C., Vastola, F. J. and Walker, P. L., Jr.,
     Carbon 5, 185-192  (1971).

 8.  Puri, B. R., Singh, D.  D., Nath, J. and Sharma, L. R.,
     Industrial and Engineering Chemistry 50(7):1071-1074
     (1958).                              ~~~

 9.  Kunii, D., and Levenspeil, 0., FLUIDIZATION ENGINEERING,
     John Wiley & Sons, Inc., New York (1969).

10.  Miller, C. A., Chemical Engineering 72, 21-29  (Sept. 13,
     1965).

11.  Peters, M. S., and Timmerhaus, K. D., PLANT DESIGN AND
     ECONOMICS FOR CHEMICAL ENGINEERS, McGraw-Hill Book Co.,
     New York, 1968.

12.  Brunauer, S., P. H. Emmett, and E. Teller, J. Amer.
     Chem. Soc. 60, 309-319 (1938).

13.  Roberts, B. F.,  J. Colloid and Interface Science 23,
     266-73 (1967).

14.  Prchlik, J., Hlinak, L. and Kartakova, C. D., "Recovery
     of Sulfur by Extraction from Spent Gas Works Purification
     Mass", Paliva (Czech. Journal^ 34, 298-305 (1954).

15.  Russell, J.  H.,  J. W. Town, and H. J.  Kelly, Bureau of
     Mines Report of Investigations 7415 (August 1970).


                            273

-------
17.  Smisek, M. , and Cerny, S., ACTIVE CARBON, American
     Elsevier Publishing Company, New York  (1970).

18.  Hildebrand-Jenks, J. Am. Chem. Soc. 43,2172-7 (1921).

19.  Jacek, Rocz. Chem. 6. 501-9  (1926).

20.  Delaplace, J. Pharm. Chim. 26, 139  (1922).

21.  Burden, F. A., and W. B. S. Newling, The Inst. of Gas
     Eng., No.  392, pg. 27-28 (Nov. 1951).

22.  Hammick, D. L. and S. E. Holt, J. Chem. Soc. 129,
     1995-2003  (1926).

23.  Levenspiel, 0., CHEMICAL REACTION ENGINEERING, John
     Wiley & Sons, Inc., New York (1962).
                             274

-------
                        SECTION 9
                      NOMENCLATURE
 A      -   Constant

 a      -   Constant

 Ao     -   Gas  Distributor  Open Area,  70

 B      -   Constant

 b      -   Constant

 C      -   Gas  Concentration,  Ib. moles/ft.3

 D      -   Reactor Diameter

 d0     -   Orifice Diameter

 E      -   Activation Energy

 F      -   Molar Flow Rate

 g      -   Function

 H2S     -   Hydrogen Sulfide Gas  Concentration

 h      -   Carbon Bed Height

 k      -   Rate Constant

 ko     -   Frequency Factor for  Arrhenium Equation

 L      -   Equilibrium Sulfur Loading Sorbed Carbon or
            Carbon Bed Loading

 Lc     -  Amount of Sulfur Chemisorbed by Carbon

 N      -  Moles

 N0      -   Initial Moles

 P       -  Equilibrium Vapor Pressure of Sulfur

 Ps      -   Saturation Vapor Pressure of Sulfur with Carbon

 q       -  Gas Flow Rate or Differential Heat of Adsorption
            of Sulfur by Carbon

R       -  Gas Law Constant or Carbon Flor Rate

Re      -  Reynolds Number

                         275

-------
     r      -  Rate of Reaction
     S      -  Sulfur Gas Concentration as S
     T      -  Temperature
     t      -  Time
     U      -  Gas Velocity
     V      -  Reactor Volume
     v      -  Superficial Linear Gas Velocity, ft./sec.
     X      -  Carbon Loading, weight of material/weight of carbon
            -  Average Number of Sulfur Atoms/Molecule

Greek Symbols
     p      -  Density
     y      -  Viscosity
     A      -  Carbon Weepage Rate

Subscripts
     c      -  Carbon
     g      -  Of Gas Phase
     H      -  Hydrogen
     H20    -  Water
     H2S    -  Hydrogen Sulfide
     j       -  Stage Number j
     j+1    -  Stage Number j+1
     Mf      -  Minimum Fluidizing
     NO      -  Nitric Oxide
     02      -  Oxygen
                              276

-------
    p      -  Of Carbon Phase
    S      -  Sulfur
    S0£    -  Sulfur Dioxide
    T      -  Total
    v      -  Sulfuric Acid
    vs     -  Sulfuric Acid Saturation
    1      -  At Condition 1
    2      -  At Condition 2

Superscripts
    A      -  Order of Reaction
    m      -  Order of Reaction
    n      -  Order of Reaction
    p      -  Order of Reaction
    q      -  Order of Reaction
                              277

-------
                                 TECHNICAL REPORT DATA
                           (Please read Inalnictiunf on the reverse before completing}
 1. REPORT NO.
                            2.
  EPA- 600/2 -76-135a
 4. TITLE AND SUBTITLE
  Development of the Westvaco Activated Carbon
   Process for SOx Recovery as Elemental Sulfur,
   Volume I	
                                                       3. RECIPIENT'S ACCESSION NO.
                                  5. REPORT DATE
                                  May 1976
                                  6. PERFORMING ORGANIZATION CODE
 7.AUTHOR(S,G.N.  Brown, C.M. Reed, A.J. Repik,
  R.L.  Stallings, andS.L.  Torrence
                                                       8. PERFORMING ORGANIZATION REPORT NO.
 9. PERFORMING ORGANIZATION NAME AND ADDRESS
  Westvaco
  Charleston Research Center
  Box 5207, North Charleston, SC  29406
                                                       10. PROGRAM ELEMENT NO.
                                  1AB013; ROAP 21ACX-085
                                  11. CONTRACT/GRANT NO.
                                  68-02-0003
  12. SPONSORING AGENCY NAME AND ADDRESS
  EPA, Office of Research and Development
  Industrial Environmental Research Laboratory
  Research Triangle Park, NC 27711
                                                       13. TYPE OF REPORT AND PERIOD COVERED
                                  13. TYPE OF REPORT AN!
                                  Final: 1/71-6/74
                                  14. SPONSORING AGENCY CODE
                                   EPA-ORD
 15. SUPPLEMENTARY NOTES pr()ject Officer fQr

  Ext 2557.
                     report is D.A. Kemnitz, Mail Drop 62,
 16. ABSTRACT The report gives results of a demonstration (in a 20,000-cfh integral pilot
  plant) of an all-dry, fluidized-bed process, using activated carbon for recovering
  SO2 as elemental sulfur. Granular carbon was recycled continuously more than 20
  times between contact with flue gas from  an oil-fired boiler and carbon regeneration
  to recover sulfur. During the 315-hour run, carbon performance remained high with
  essentially no chemical and low mechanical losses. Over 90% of the 2000 ppm SOx
  was removed from the flue gas as sulfuric acid by catalytic oxidation and subsequent
  hydrolysis within the  carbon granule. In the two-step regeneration: (1) the acid was
  converted to elemental sulfur at 300F with internally  produced H2S, and (2) an exter-
  nal source of hydrogen at 1000F was used to thermally strip the by-product sulfur
  from the carbon and produce the required H2S by reaction with the remaining sulfur
  on carbon. Sufficient  process and design information  was developed from data ob-
  tained in the integral  run and prior stepwise pilot equipment operation to permit
  scale-up to a 15-MW prototype for a coal-fired boiler.  In the preliminary design,
  reducing gas is produced in a coal gasifier. An economic assessment of a 1000-MW
  conceptual design for  the process indicates capital and  operating costs competitive
  with those of other regenerable systems.
 7.
                              KEY WORDS AND DOCUMENT ANALYSIS
                 DESCRIPTORS
                                           b.lDENTIFIERS/OPEN ENDED TERMS
                                                cos AT I Field/Group
 Air Pollution
 Flue Gases
 Activated Carbon
 Sulfur Oxides
 Fluidized Bed
    Processing
Regeneration
  (Engineering)
Fuel Oil
Sulfuric Acid
Catalysis
Oxidation
Air Pollution Control
Stationary Sources
Elemental Sulfur
Westvaco Process
Catalytic  Oxidation
13B
2 IB
11G      21D
07B
         07D
13H,07A  07C
 8. DISTRIBUTION STATEMENT

 Unlimited
                      19. SECURITY CLASS (This Report)
                      Unclassified
                         21. NO. OF PAGES
                           298*
                     20. SECURITY CLASS (Thispage)
                      Unclassified
                         22. PRICE
EPA Form 2220-1 (9-73)
                  278

-------