INSTRUMENTAL ANALYSES
FOR WET SCRUBBING PROCESSES
CONTRACT 68-02-0007
by
E.A. BURNS
A. GRANT
D.F. CARROLL
M.P. GARDNER
J.C. GRAY
INTERIM REPORT
Prepared for
ENVIRONMENTAL PROTECTION AGENCY
Office of Air Programs
Research Triangle Park
North Carolina
15 JANUARY 1972
TRW
-------
INSTRUMENTAL ANALYSES
FOR WET SCRUBBING PROCESSES
CONTRACT 68-02-0007
by
E.A. BURNS
A. GRANT
D.F. CARROLL
M.P. GARDNER
J.C. GRAY
INTERIM REPORT
Prepared for
ENVIRONMENTAL PROTECTION AGENCY
Office of Air Programs
Research Triangle Park
North Carolina
15 JANUARY 1972
TRW
trtTIHS SHOUT
ONE SPACE PARK • REOOMOO BEACH. CALIFORNIA
-------
17551-6012-RO-OO
FOREWORD
This document constitutes the interim report for the work accomplished
between 16 November 1970 to 31 December 1971 by TRW Systems for the
Environmental Protection Agency, Office of Air Programs, Durham, North
Carolina, under Contract 68-07-0007 on Instrumental Analysis for Wet
Scrubbing Processes. This work was conducted under the direction of Dr.
Robert M. Statnick of the Office of Air Program, Durham, North Carolina.
The Applied Chemistry Department and Chemical Engineering Department
of the Chemistry and Chemical Engineering Laboratory, Applied Technology
Division, were responsible for the work performed under this program.
Mr. B. Dubrow, Manager, Chemistry and Chemical Engineering Laboratory pro-
vided the overall program supervision and Dr. E. A. Burns, Manager of the
Applied Chemistry Department was Program Manager. The Principal Investigator
was Mr. A. Grant. Major technical contributions throughout the program were
provided by Messrs., D. F. Carroll, M. P. Gardner and J. C. Gray. Acknowl-
edgment is made of technical assistance provided during the program by the
following TRW Systems personnel:
Members of the Professional Staff:
Chemistry Department
Chemistry Department
Chemistry Department
Chemistry Department
Engineering Department
Engineering Department
Engineering Department
Engineering Department
Staff
Chemistry Department
J.
C.
F.
M.
J.
W.
J.
R.
S.
P.
F. Clausen
A. Flegal
K. Harpt
L. Kraft
L. Lewis
D. Lusk
R. Ogren
S. Ottinger
Srinvisan
E. Testerman
Applied
Appl i ed
Applied
Applied
Chemical
Chemical
Chemical
Chemical
Science
Applied
Technical Support:
J.
D.
H.
D.
A. Buehner
B. Kilday
D. Lindewall
J. Luciani
Applied
Appl i ed
Applied
Applied
Applied Chemistry Department
-------
17551-6012-RO-OO
ABSTRACT
The development of control methodology for sulfur oxide and particu-
lates from power plant emissions by wet scrubbing processes requires accur-
ate and reliable measurements of process variables. Planned EPA process
demonstration studies will result in a requirement for a large number of
chemical analyses requiring 1) automatic instrumental methods and 2) asso-
ciated data acquisition and processing capabilities which exceed current
instrumental capabilities. This report describes activities undertaken at
TRU Systems under Contract 68-02-0007 toward the development of methods
suitable for optimization and control of the wet limestone and dolomite
scrubbing processes by continuous onstream analytical methods. Emphasis
was placed on development of continuous on-line methods for slurry sampling
and separation that do not disturb the chemical steady state condition.
Establishment of sampling requirements and an effective means for total
phase separation in a period less than thirty seconds were accomplished.
Analytical instrumental methods having capability of continuous or
slug flow analysis within two minutes were identified for characterization
of the separated solid matter and liquor. Analytical methods were identi-
fied which permit continuous X-ray analyses of solid constituents for
sulfur, calcium, magnesium and iron contents. Liquid phase analyses methods
were established for instrumental analysis of acidity, sulfite, sulfate,
calcium, magnesium and carbonate contents. A new method for rapid analysis
of sulfite content based on furfural bleaching is being carried to a state
of prototype analytical instrument development. In addition, approaches
for total complete on-line analysis of other wet limestone scrubber
constituents have been identified.
iii
-------
17551-6012-RO-OO
INSTRUMENTAL ANALYSES FOR WET SCRUBBING PROCESSES
by
E. A. Burns, A. Grant, D. F. Carroll, M. P. Gardner and J. C. Gray
SUMMARY
This interim document presents the work accomplished by TRW Systems
during the period 16 November 1970 to 31 December 1971 for the Environmental
Protection Agency, Office of Air Programs, under Contract 68-02-0007. The
objective of this program was to identify instrumental analysis methodology
suitable for laboratory and especially on-line analysis of selected species
in and properties of wet scrubber process streams from pollutant emission
control systems. The emphasis during this program has been placed on the
development of on-line sampling and separation techniques to provide separa-
tion of the slurry system into solid and optically clear liquid phases and
analysis and measurement methodology for pH, sulfite, sulfate, calcium, mag-
nesium, carbonate, nitrite and nitrate. This phase of the program was or-
ganized in three tasks in order to accomplish the stated objectives:
t TASK I - Development of Laboratory Instruments and Analysis
• TASK II - Development of Process Instrumentation
• TASK III - Data Acquisition and Processing
Upon review of analysis requirements and process characteristics, ef-
fective sampling, separation and quenching of reactants was identified as a
major prerequisite to development of analysis methodology. Several centrif-
ugation and filtration techniques were evaluated in terms of rapid phase
separation at the point of process sampling for a sampling criterion of 30
grab samples per hour minimum. The system designed and developed in the
laboratory utilizing actual and simulated slurries (3% nominal loading) and
recommended for field evaluation is comprised of a one gpm cyclone (cone)
stage followed by a dual-parallel filter stage for polishing the liquid
stream. This system may also permit continuous operation, if necessary, as
the dual in-line filters are of high capacity-quick interchange design.
-------
17551-6012-RO-OO
One Instrumental analysis technique, namely, x-ray fluorescence (XRF)
was shown to provide on-stream or laboratory elemental analysis capability
for unseparated slurry as well as separated liquid and solid phases. The
unit recommended for the on-stream application is an ARL Model PCXQ which
can accommodate up to 15 slurry streams and with nine spectrometers to de-
termine nine elements from magnesium and heavier. Sensitivity for total
sulfur, especially important in the analysis of slurry was found to be
0.03% for solid samples and considerably better than 0.25% for the mixed
slurry. For solids analysis, XRF was shown to be the optimum method for
calcium, magnesium, iron, silicon and other metals of concern, offering a
significant saving in time and cost per analysis compared to other candi-
date techniques.
Atomic absorption spectrometry (AA) was evaluated in the laboratory
for determination of those species very slightly soluble or present in
trace quantities in the liquid or solid phase (after dissolution). AA is
recommended specifically for dissolved calcium, magnesium, iron, potassium,
sodium and other trace elements. This method is especially valuable where
secondary pollution problems from mercury, arsenic, chromium, lead, etc.,
must be monitored and controlled.
During the review of candidate analytical methods for the continuous
determination of dissolved sulfur dioxide (HS03~ and SO.,") it was deter-
mined that no satisfactory methods existed for determining concentrations
in the range to be found in the limestone slurry mixture. A new method
based on bisulfite bleaching of the furfural UV absorption was developed to
facilitate this analysis. This method is based on the chemical bleaching of
the 276 nM absorption of furfural by reaction with bisulfite. Detailed
studies of the effect of pH, diverse ions, temperature and time to constant
color development has resulted in the selection of a single reagent addition
consisting of furfural, phosphate buffer and sulfamic acid (to eliminate in-
terference from trace concentrations of nitrite). The reproducibility of
the method has been determined to be better than 2% relative or 0.2 nM abso-
lute whichever is higher. A preliminary design of a prototype plug flow
automated analyzer to accomodate up to ^100 samples per eight-hour shift
has been completed.
VI
-------
17551-6012-RO-OO
Methods for the analysis of sulfate, carbonate, nitrite and nitrate
were evaluated theoretically and experimentally less extensively than those
above utilizing simulated slurry liquid samples. Two state of the art
methods for sulfate determination were found to warrant further considera-
tion and development, i.e., sulfate precipitation by barium with the measure-
ment being accomplished by turbidimetry or by AA, and 2) ion exchange with
barium chloranilate and colorimetric measurement of free chloroanilic acid.
Feasibility tests demonstrated the potential utility of a pyrolysis/acidi-
metric carbonate determination. Released CCL could be measured readily by
NDIR in a batch automated or continous analyzer. In this phase of the pro-
gram nitrite/nitrate literature was reviewed and several were selected for
further consideration and future experimental evaluation.
A basic modular designed bench scale test loop wet scrubber unit was
fabricated to permit evaluation of the recommended methods under simulated
use conditions. A loop system was selected because of the necessity of:
1) closely approximating the full scale operating unit, 2) accurate con-
trol, and 3) producing stable (equilibrium) and unstable (non-equilibrium)
conditions for evaluating candidate instruments under known, controllable
conditions with realistic compositions. The system consists of a bench
scale Venturi scrubber with a second stage packed bed, fitted with a re-
circulating gas stream and slurry pumping capability to approximate actual
L/6 ratios to be tested on pilot and full scale scrubbers. Equipped with
ports for acquiring appropriate gas and slurry samples and a continuous pH
monitor in the scrubber downcomer, the scrubber experiments 1) demonstrated
the adequacy of the proposed methodology, and 2) contributed significantly
to the elucidation of the chemistry of the scrubbing process. Process var-
iables that were studied and found to impact process chemistry were oxygen
content, fly ash loading and composition, and temperature. Sulfite oxida-
tion and rate of sulfite precipitation were two such important effects that
were isolated in these studies.
Recommendations were provided for data acquisition and process equip-
ment suitable for use with recommended analytical instrumentation and capa-
bility for interfacing with the engineering computational system to be oper-
ational at the TVA wet scrubber plant site. The primary candidate system
vii
-------
17551-6012-RO-OO
was the Hewlitt Packard 2411C computer because of its proven application with
the recommended ARL PCXQ 4400 X-ray fluorescence unit. An alternative com-
puter data process for the analytical instrumentation was the Digital
Equipment Corporation POP 12/LDP.
viii
-------
17551-6012-RO-OO
CONTENTS
Page
1.0 INTRODUCTION 1
2.0 TASK I - DEVELOPMENT OF LABORATORY INSTRUMENTS AND
ANALYSES 5
2.1 Literature Review 6
2.1.1 Theoretical Modeling of the Wet Scrubbing
Process 6
2.1.2 Review of Analysis Methodology and
Instrumentation 8
2.2 Generalized Instrumentation and Slurry Sampler
Specifications 12
2.2.1 General Instrument Specifications 12
2.2.2 Operating Specifications 13
2.2.3 Requirements for Slurry Sampling and Separation. 13
2.3 Review of Limestone Wet Scrubbing Process Operation . . 15
2.3.1 Review of the Shawnee Process Demonstration
Operation 15
2.3.2 Inspection of the Zurn Engineering Wet Limestone
Scrubbing Operation at the Key West Electric
Company 16
2.3.3 Inspection of the Wet Limestone Scrubbing
Operation of the Kansas Power and Light
Company 19
2.4 Instrumental Methodology for Cations and Elements ... 19
2.4.1 Laboratory X-Ray Spectroscopic Methodology ... 21
2.4.1.1 XRF Applications in Literature 21
2.4.1.2 XRF Analysis of Simulated and Field
Samples 22
2.4.1.3 X-Ray Equipment Vendor Contacts .... 23
2.4.1.4 Experimental Evaluation at Vendor
Application Laboratories 25
2.4.2 Atomic Absorption Spectrophotometric
Methodology 27
2.5 Spectrophotometric Analysis of Sulfite and Bisulfite. . 29
2.5.1 Effect of pH 32
2.5.2 Effect of Diverse Ions and Total Ionic
Strength 39
2.5.2.1 Elimination of Nitrite as an
Interferent 41
IX
-------
17551-6012-RO-OO
CONTENTS (CONTINUED)
Page
2,5.2.2 Iron as a Potential Interferent .... 42
2.5.3 Time Dependence of the Furfural-Bisulfite
Complex Formation 45
2,5.4 The Effect of Temperature on the Furfural-
B1sulf1te Analysis 47
2.6 Instrumental Analysis of Sulfate 48
2.6.1 Theoretical Evaluation of Sulfate Methods. ... 50
2.6.1.1 Titrimetric Methods 50
2.6.1.2 Turbidimetric Method 50
2.6.1.3 Barium Chloranilate Colorimetric
Method 51
2.6.1.4 Benzidine Colorimetric Method 51
2.6.1.5 Infrared Spectrophotometric 51
2.6.1.6 Atomic Absorption Method 51
2.6.1.7 Specific Ion Electrode 52
2.6.1.8 Sulfate by Difference from Total
Sulfur 52
2.6.2 Experimental Screening of Candidate Sulfate
Methods 53
2.6.2.1 Turbidimetric Method 54
2.6.2.2 Infrared Spectrophotometric Method. . . 54
2.6.2.3 High Frequency Titration Method .... 55
2.7 Instrumental Analysis of Carbonate 56
2.7.1 Precipitation Reactions 56
2.7.2 Precipitation Reactions 56
2.7.3 Thermal or Acidimetric Removal of C02 58
2.8 Survey of Nitrite/Nitrate Instrumental Analysis
Methodology 59
2.8.1 Brucine Colorimetric Method 59
2.8.2 Diazotization After Reduction 60
2.8.3 Ion Specific Electrodes 60
3.0 TASK II - DEVELOPMENT OF PROCESS INSTRUMENTATION 61
3.1 Slurry Sampling, Separation and Quenching 62
3.1.1 Continuous Cyclone Separation/Filtration .... 64
3.1.2 Solids Discharging Methods 69
-------
17551-6012-RO-OO
CONTENTS (CONTINUED)
Page
3.2 Continuous On-Line X-Ray Fluorescence (XRF)
Methodology 74
3.2.1 Limit of Detection 75
3.2.2 Repeatability 75
3.2.3 Matrix Effects 78
3.2.4 Quantitative Interpretation 81
3.2.5 Particle Size Effects 81
3.2.6 Dilution Effects 82
3.2.7 Analysis of Liquid Samples 82
3.2.8 Advantages of the ARL System 85
3.3 Continuous On-Stream Carbonate Analysis 86
3.4 Development Plan for Wet Scrubber Bisulfite Analyzer
(WSBA) Prototype Fabrication and Evaluation 87
3.5 Bench Scale Wet Scrubbing Process Simulator 89
3.5.1 Bench Scale Scrubber Tendency for Slurry
Oxidation 94
3.5.2 Evaluation of Recommended Methods for
Characterization of the Limestone Scrubber
Process 96
3.5.2.1 Standard Operating Procedure for
Bench Scale Scrubber 96
3.5.2.2 Results of Characterization of Scrubber
Slurries 97
3.5.3 Study of Wet Scrubber Process Variables Using
the TRW Bench Scale System 98
3.5.3.1 Experimental 98
3.5.3.2 Results 100
3.5.3.3 Recommendations 104
3.6 Process Monitoring for pH 104
4.0 TASK III - DATA ACQUISITION AND PROCESSING 107
4.1 Computer Data System for XRF On-Line Process
Instrumentation 107
4.2 Alternative Computer Data Processor for General
Analytical Instrumentation 109
4.3 Non-Computer Data Acquisition HO
XI
-------
17551-6012-RO-OO
CONTENTS (CONTINUED)
Page
5.0 CONCLUSIONS AND RECOMMENDATIONS Ill
6.0 NEW TECHNOLOGY 115
6.1 Wet Scrubber Bisulfite Analyzer 115
6.2 Total Sulfur Analyzer for Process Streams 115
6.3 X-Ray Fluorescence Analysis of Elements 116
6.4 Pyrolysis/Acidimetric Carbonate Method 116
6.5 Continuous Slurry Phase Separator 116
APPENDIX A - SURVEY OF STANDARD ANALYTICAL METHODS FOR SLURRY
COMPONENTS 119
APPENDIX B - SURVEY OF ELECTROCHEMICAL METHODS FOR ANALYSIS OF
DISSOLVED OXYGEN AND SULFUR DIOXIDE 125
APPENDIX C - INITIAL SHAWNEE PROCESS DEMONSTRATION OPERATIONAL
MODES 135
APPENDIX D - DETAILED ASSESSMENT OF X-RAY ANALYTICAL METHODS. . . 169
APPENDIX E - TENTATIVE METHOD FOR ANALYSIS FOR SULFITE AND
BISULFITE ION BY FURFURAL BLEACHING 185
APPENDIX F - DIRECTORY OF SEPARATOR MANUFACTURERS AND VENDORS . . 189
APPENDIX G - PROCESS pH MONITORING SYSTEM 191
REFERENCES 195
-------
17551-6012-RO-OO
LIST OF TABLES
Page
I Liquid Analysis Requirements 8
II Literature Recommended Analytical Procedures for
Major Species in Limestone Slurries 9
III Characteristics of Zurn Limestone Slurry 16
IV Sulfite Fixation Experiments on Filtered Weir Bottom
Limestone Slurry Liquor 17
V Possible Limestone/Dolomite Slurry Component
Distribution 20
VI XRF Analytical Data for Limestone 21
VII Comparison of XRF and Chemical Values for Total Sulfur
in Standards and Mineralized Samples 22
VIII XRF Standard Samples for Instrument Evaluation .... 26
IX Effect of Flame Oxidizer, pH and Inhibitor on Atomic
Absorption Determination of Calcium in Limestone
Solutions 28
X Atomic Absorption Analysis of Selected Samples
Obtained from Operational Limestone Wet Scrubber
Units 28
XI pH Dependence of Furfural Absorption Spectrum at
276 nM 33
XII Determination of Bisulfite by Furfural Bleaching in
Unbuffered Media 34
XIII Determination of Sulfur (IV) by Furfural Bleaching . . 35
XIV Evaluation of pH Effect in Determination of S (IV) in
Phosphate Buffer 36
XV Experimental and Computed pH versus 1/A(HS03") .... 39
XVI Interference Screening Tests for Bisulfite-Furfural
Bleaching 40
XVII Effect of Sulfamic Acid on N02" Interference 42
XVIII Interference of Fe+3 in the Furfural-Bisulfite
Method 43
XIX Time Dependence on the Furfural-Bisulfite Equilibrium. 46
XX Thermal Dependence of the Bisulfite Calibration Curve. 48
XXI Limestone Slurry Sampling Requirements 61
XXII Summary of Laboratory Evaluation of Separation
Methods 54
XXIII Effect of Filter Cake Buildup on Simulated Wet
Scrubber Filtrate Composition 68
xm
-------
17551-6012-RO-OO
LIST OF TABLES (CONTINUED)
Page
XXIV Signal Data from ARL Quantometer 72000 76
XXV ARL-72000 Vacuum Quantometer Repeatability Data. ... 78
XXVI Effect of Kapton Window on X-Ray Spectrometer
Performance: 0.25-Mil Sheet 84
XXVII Bisulfite Analysis for Bisulfite Oxidation Bench
Scrubber Experiment 104 95
XXVIII Bisulfite Analysis for the Limestone Slurry 97
XXIX Limestone Slurry Liquor Composition as a Function of
Operating Conditions and Time 102
XXX Recommended Computer System for XRF Data Acquisition
and Reduction 108
xiv
-------
17551-6012-RO-OO
ILLUSTRATIONS
Page
Figure 1 Sampling System for Zurn Wet Scrubber 18
Figure 2 Sulfur IV Species as a Function o pH 32
Figure 3 Bisulfite-Furfural Absorption as a Function of pH. . 38
Figure 4 Change In Furfural Bisulfite Complex Equilibrium
Constant with Temperature 49
Figure 5 Carbonate Species as a Function of pH 57
Figure 6 Demco Centrifugal Separator 65
Figure 7 Schematic of Continuous Slurry Separation
Approaches 66
Figure 8 Schematic Design of Continuous Solid Separation
and Analysis Apparatus Concept 70
Figure 9 Flow Rate versus Clarity for Sharpies
Super-D-Canter 71
Figure 10 Zurn Slurry Separation by Demco Cone 73
Figure 11 Working Curves for Sulfur Analysis 79
Figure 12 Calcium Working Curves 80
Figure 13 Conceptual Concentration • Loading - Signal
Map for Sulfur 83
Figure 14 WSBA Flow Diagram 88
Figure 15 Bench Scale Scrubber Analysis Loop 90
Figure 16 Photograph of Instrumental Bench Scale Process
Scrubber Simulator 91
Figure 17 Effect of Fly Ash and Temperature on Soluble S(IV)
in Bench Scale Wet Scrubber 101
xv
-------
17551-6012-RO-OO
1.0 INTRODUCTION
This interim report presents the work accomplished by TRW Systems for
the Environmental Protection Agency, Office of Air Programs, under Contract
68-02-0007 during the period 16 November 1970 to 31 December 1971. This
program consisted of analytical and experimental studies aimed toward the
identification of instrumental methods suitable for on-line analysis of
selected chemical species in the Shawnee Plant limestone wet scrubbing pro-
cesses. The underlying motivation for conducting this program was to de-
velop methods suitable for continuous monitoring of the key chemical species
to facilitate 1) an understanding of the wet limestone scrubbing process,
and 2) provide mass balance information related to the effectiveness of the
process for the abatement of sulfur dioxide emissions from stationary power
sources.
The development of control methodology for sulfur oxide and particu-
lates from power plant emissions by limestone/dolomite wet scrubbing re-
quires accurate and reliable measurements of process variables. Efficient,
proven methods for many of these measurements have not yet been developed.
The monitoring of the complex chemistry involved in this scrubbing process
and associated sampling of representative samples in quiescent and dynamic
mixtures of liquors, slurry and solids are in themselves challenging analyt-
ical problems. In addition, planned OAP process demonstration studies at
the Shawnee Power Plant, Paducah, Kentucky, will result in a requirement
for a large number of chemical analyses requiring 1) automatic instrumental
methods and 2) associated data acquisition and processing capabilities which
exceed current instrumental capabilities.
The chemistry of the process is not sufficiently understood at the
present time because of the lack of definitive mass balance information in-
volving the chemical species existing in the scrubbing solution. The de-
velopment of suitable on-stream analysis methods will provide a means to
fill this gap through detailed characterization of the process. High ana-
lytical accuracy (e.g., 0.1% relative) is not a requisite of the needed
methods but rather they must be adaptable to instrumental techniques that
will be reliable, reproducible, cost effective and employ hardware requir-
ing little maintenance.
-1-
-------
17551-6012-RO-OO
The objective of the program reported here was to carry out systematic
analytical and experimental studies for development of laboratory instru-
ment and analysis methods, development of process instrumentation and out-
line data acquisition and processing requirements suitable to handle the
on-line information to be generated in the process demonstration studies.
Task I of this program consisted of identification and/or development
of analysis methods for critical chemical species in the liquids, slurry and
solid materials resulting from the aqueous limestone scrubbing of fossil
fuel combustion gases. The intent of this effort was to identify instru-
mental methods based on their specificity, reproducibility, potential lag
time, and reliability of operation. Specifically, the task involved a re-
view of available literature information on the primary candidate wet scrub-
bing process (limestone), identification of applicable analysis methodology
and instrumentation, preparation of generalized instrument specifications
and experimental evaluation of the methods developed using laboratory
samples.
The Task II efforts consisted of the evaluation of the applicability
of the methods identified in Task I to continuous on-line instrumentation
for monitoring selected species in the liquid, slurry and/or solid phase at
several locations in the scrubber system. As part of this evaluation, the
methods were tested in a laboratory bench scale simulated process which in-
volved both stable and unstable slurry systems. Again, the reliability,
sensitivity, reproducibility, specificity and accuracy of the measurements
were used as criteria for evaluating candidate instruments, breadboard pro-
totypes and modified apparati.
Task III efforts consisted of identification of applicable data han-
dling systems for collecting and reducing all information produced by the
analysis scheme in a form which lends itself to computer input. From the
studies conducted during this program, recommendations for future needed
research and development activities in the area of characterizing process
streams for mass balance purposes have been generated.
This report is divided into three principle sections covering the pro-
gram tasks: 1) development of laboratory instruments and analysis, 2) de-
velopment of process instrumentation and 3) data acquisition and processing.
-2-
-------
17551-6012-RO-OO
The significant conclusions reached from evaluation and assessment of the
results are listed together with recommendations for activities that war-
rant further investigation. This report identifies in a special section
the new technology originating from the program. The information presented
in the main body of this report is supplemented by appendices covering de-
tailed descriptions of procedures, assessment of candidate analytical pro-
cedures, and lists of vendors and manufacturers which offer suitable equip-
ment.
-3-
-------
17551-6012-RO-OO
2.0 TASK I - DEVELOPMENT OF LABORATORY INSTRUMENTS AND ANALYSES
This task was aimed at identification and/or development of analysis
methods for critical chemical species in the liquids, slurry and solid mater-
ials resulting from the aqueous scrubbing of fossil fuel combustion gases.
In accordance with EPA direction, TRW concentrated its activities on the
measurement of the following chemical species and characteristics of the
wet scrubbing process:
• Calcium concentration
0 Magnesium concentration
• Sulfite concentration
• Sulfate concentration
ff pH
• Ionic strength
During the conduct of this effort it was determined that many of the
analytical methods suitable for laboratory characterization of the limestone
scrubber constituents were not applicable to the planned future on-line
characterization. On the other hand, most of the analytical methods suita-
ble for on-line use are also acceptable as laboratory methods. This obser-
vation, together with the subsequent initiation of a parallel program to
the Radian Corporation, (Reference 1) for development of a laboratory analy-
sis scheme resulted in the relative de-emphasis of Task I activities rela-
tive to those of the on-line instrumental analytical methods. An assess-
ment of the early clans by Bechtel Corporation for sampling points and anal-
ysis requirements indicated that after an initial evaluation period, on-line
process control measurements were required to be cost effective.
However, the comprehensive literature survey for laboratory experimenta-
tion conducted in Task I provided the basis for 1) identification of analy-
sis and characterization criteria, 2) establishment of on-line or automated
process instrumentation specifications, and 3) development of acceptable
methodology. The following section describes the
1. Literature review of available information on the primary
candidate wet scrubbing process,
2. Analysis methodology and instrumentation,
-5-
-------
17551-6012-RO-OO
3. Generalized instrumentation specifications, and
4. The methods developed and evaluation pertaining to
laboratory analysis.
2.1 LITERATURE REVIEW
The literature review was initiated in order to update and maintain
currency of our compilation of data on instrumental and automated manual
analytical methodology for wet scrubbing processes. The literature review
was aimed at obtaining data on:
• Solubility of slurry constituents
• Methods of analyses outlined in laboratory studies, and
• Applicability of current instrumentation to the labora-
tory study
2.1.1 Theoretical Modeling of the Met Scrubbing Process
Key related reports generated under EPA sponsorship have been reviewed
to provide background information. Throughout the review there appeared to
be a lack of applicable information concerning the chemical species which
exist under the actual temperature and chemical conditions of the wet scrub-
bing process. Because of the lack of empirical data, EPA has sponsored
several efforts aimed at providing a theoretical description of the process
suitable for identification of future process design improvement. The lead-
ing process model is that generated by the Radian Corporation (Reference 2)
which describes the process through chemical compositional computation using
equilibrium conditions. This model has proven to be very useful as a start-
ing point in identifying the chemical composition of the wet limestone slurry
at various points in the process under fixed conditions. It is generally
recognized that there are shortcomings to the equilibrium assumption, how-
ever, modification of the model to provide a "real-life" process model have
not been undertaken primarily because of the lack of empirical information
needed for confirmation.
As part of this task, TRW reviewed the Radian model in detail to deter-
mine its applicability to describe the scrubber chemistry. It is the in-
tention of this review to identify some of its limitations and point out how
it can be further improved. These shortcomings are delineated below:
-6-
-------
17551-6012-RO-OO
• The most significant shortcoming is the assumption of
equilibrium conditions. Recent experiments at TRW (see
Section 3.5.3) have clearly identified that the solids
do not precipitate to their equilibrium value in a time
period to be consistent with either the scrubber or ef-
fluent hold tank.
• Several of the chemical species used in the theory
have not been substantiated and are inferred by cal-
culations aimed at making an internally consistent
set of data (without cross check).
t The constants used in many cases were determined by
experimental techniques which are open to technical
question [use of specific ion electrodes which tend
to have errors in response, particularly at moderate
(lO"^ M) concentrations].
• The model requires the input of many variables which
might be determined theoretically. These include the
quantity of NOX dissolved (which has been shown to be
considerably lower than that used in the calculations),
the amount of CaO and MgO converted to hydrate and the
amount of S02 oxidized to S03.
0 Several chemical species which are known to exist in
partially oxidized sulfitg solutions, namely, the
thionates, S206~ and S306~ have not been considered in
the theoretical treatment.
• The thermochemical equilibria in solutions are quite
sensitive to properties, such as ionic strength, but
most of the data utilized in the formulation of the
model were taken from pure compound solubilities and
assuming the validity of the extended Debye-Huckel Law.
• Another assertion which has a strong bearing on the
chemical composition is that use of Fuoss equation is
valid for determining the temperature dependence of
the constants.
Lacking more definitive information during the early phases of the
program, analysis error requirements for the key chemical constituents pre-
sent in the limestone/dolomite scrubber were based on estimates provided by
the EPA Project Monitor of concentration range and relative error of the
methods required for 20% sulfur mass balance closure as determined by the
Bechtel Corporation. These data are presented in Table I and were used to
guide the direction of the program pending updating of these requirements
in concurrent programs by the Radian Corporation and Bechtel Corporation.
-7-
-------
17551-6012-RO-OO
TABLE I
LIQUID ANALYSIS REQUIREMENTS
Concentration
Range, mM
Mg++
Ca++
S03=
so4=
co3=
Na+
K+
Cl"
1
1
1
1
1
1
1
1
- 1000
250
150
500
20
500
500
500
Maximum Allowable
Relative Error*
3
3
3
5
15
15
15
15
*For 20% sulfur mass balance closure
2.1.2 Review of Analysis Methodology and Instrumentation
Available literature on analytical procedures used to characterize the
major chemical constituents in limestone slurries were reviewed. Table II
provides a summary of the methods that have been recommended and information
concerning the technique and its applicability to slurry solid, liquor or
mixed phases together with appropriate comments relating to end point and/or
alternative detection procedures. In general, the prior investigations were
conducted over a short period of time and little, or no information was pro-
vided concerning the utility of the methods for continuous on-line measure-
ment. The methods were used without consideration of potential interferences
and their effect on the accuracy and reproducibility of the methods.
Concurrently, analytical techniques were compiled for characterization
of cations and anions that were expected to be present in the slurry liquor.
Appendix A delineates the wet and some standard instrumental methods for
sixteen of the key chemical species. Tabulated are the components to be
measured, the principle of method, sensitivity, interferences and related
comments. In general, utilization of these methods requires separation of
phases prior to analysis.
-8-
-------
TABLE II
LITERATURE RECOMMENDED ANALYTICAL PROCEDURES FOR MAJOR SPECIES IN LIMESTONE SLURRIES
Component
Ca, Mg, Na,
K, Fe, other
metals
Total Ca
and Mg
N02, NO;
N03
S03=
SO/
Cl"
C03=
Total C03=
Trace
Metals
pH
Conductivity
Suspended
solids
Referee
AA
Titri metric
(EDTA)
Photometric
Gravimetric
Titri metric
Gravimetric
Titri metric
Titri metric
Manometric
Emission
spec.
Glass
electrode
-
Gravimetric
Rapid
AA
-
Photometric
-
Titri metric
Colorimetric
Titrimetric
-
-
AA
Glass
electrode
Probe
Radiometer
Phase*
(S & L)
(L)
(L)
(L)
U&S)
(L)
(L&S)
(S)
(S)
(TS)
(TS)
(TS)
Comment
~
-
-
Weigh as C12 H22T1N03
I2-phenyl-arsine oxide
(amperometric)
Turbidimetric
AgN03- potenti ometri c
Ba(OH)2-titn with HC1
-
-
-
™
Alternate References 1
Na+K flame photometric; EDTA for 3 1
Ca and for hardness. Glyoxal 2- 1
hydroxyanil (color)
2
N03 Brucine; NO^ Phenoldisulfonic 3
acid. N03 polarographic; NO 2
diazotization
2 •
HBr-electrolytic titration** 3
Barium chloroanilate-colori metric 3
titn. Ba(C10j2-thorin 3
Mercuric nitrate 3
3
2
Polarography 3
3
3
(Ca.,, Mg, Na, K, Cl", NO^, NO^, , 1
SOi," and trace metals ion- 1
selective electrodes) 1
-vl
en
en
cn
o
ro
i
O
O
*S = solids; L = Liquids; TS = total slurry
Considered by EPA
-------
17551-6012-RO-OO
Of the standard manual laboratory methods for sulfite and sulfate, the
1odometr1c redox methods are still generally preferred for the lower oxida-
tion states, while for sulfate 1on the most widely accepted methods are
based on the barium tltratlons with a visual end points and barium precipi-
tation with a gravimetric determination (carbonate and sulfite interfers
and must be removed). The strontium-EDTA method (see Appendix A, Reference 4)
for differentiating carbonate and bicarbonate, which appears to offer a dis-
tinct advantage over other methods, remains to be evaluated. Cation analysis
is commonly performed by the complexemetrie EDTA and colorimetric methods,
however, atomic absorption spectrophotometry and specific ion electrode
methods provide distinct advantages over the classical methods in simplicity,
specificity and economy.
Electrochemical methods appear to be the prime approach for determina-
tion of dissolved oxygen in process streams which is considered to be highly
significant in the liquid phase oxidation of sulfite and bisulfite. Addi-
tionally, the sulfite ion concentration in the liquid phase of the wet
scrubber process stream may be suited to electroanalytic measurements [e.g.,
in the millimolar (mM) concentration range]; the actual dissolved oxygen
content in these solutions is somewhat in question, however, dissolved
oxygen from air (in non-reactive media) at wet scrubber process stream temper-
atures is ^0.5 mM. The process stream at a pH of near 6 will possess major
components of HC03", S04S, HS03", H2C03, Na+, Mg+2, Ca+2 and CaS04. These
major cations do not interfere with the electrochemical process and any
other reducible metals originating from the fly ash will be at a concentra-
tion of at least two orders of magnitude less, hence, will be insignificant
relative to the sulfite and dissolved oxygen content. One requirement the
electrochemical method has is that the liquid should be free of fine suspen-
sions.
A critical review of current methods used for analysis of dissolved
oxygen and dissolved sulfur dioxide by polarographic, chronopotentiometric,
conductometric, coulometric and other methods were reviewed. Criteria
used for evaluation of the various methods included:
-10-
-------
17551-6012-RO-OO
• The present state of development (laboratory technique,
availability of laboratory or process stream instruments)
• Sampling mode
« Specificity and interferences
• Pretreatment of sample for analysis (concentration,
filtration, adjustment of pH)
• Useful concentration range, sensitivity, accuracy,
precision
• Temperature requirements
t Analysis time
t Data reduction capability
I Requirements for further development
t Maintenance requirements
t Cost
f Life time and cycle life
• Commercial instruments
The detailed compilation of this information is in tabular form in
Appendix B for possible future implementation of an experimental investi-
gation.
Numberous candidate state of the art instrumental techniques for wet
scrubber constituents were evaluated experimentally as well as theoretically,
including some of the standard photometric (colorimetry, turbidimetry,
atomic absorption, etc.) techniques listed in Appendix A. The following
specific techniques are discussed in detail in subsequent sections of this
report.
Electrochemical Methods Electron Diffraction
Atomic Absorption (AA) Colorimetry
X-ray Fluorescence (XRF) X-ray Diffraction (XRD)
UV-VIS-NIR and IR Flame Emission
Spectrophotometry
-11-
-------
17551-6012-RO-OO
Quantitative automated laboratory or instrumental techniques were
successfully developed and demonstrated for:
f Cations and elements by AA (in solution) and XRF
(solids and liquids), and
• Sulfite by a new UV spectrophotometric method.
The feasibility of laboratory automatable methods for the following
species was also demonstrated:
• Sulfate by turbidimetry
• Sulfate by high frequency titrimetry
• Carbonate by C02 liberation and NDIR detection
2.2 GENERALIZED INSTRUMENTATION AND SLURRY SAMPLER SPECIFICATIONS
As a result of the review of the sampling and characterization require-
ments of the limestone dolomite wet scrubbing S02 abatement process through
both review of literature and visits to operating facilities, generalized
specifications for instruments to monitor the wet scrubbing process and
sampler systems have been generated. These general specifications are pre-
sented below for use as guidelines in purchasing various types of instru-
ments which will be required. Specific types of instrumentation are recom-
mended later in this document.
2.2.1 General Instrument Specifications
• Selectivity - Measure S04=, HS03", H+, Ca++, Mg++ and
the other identified species of concern and be relatable
to the content of the original wet scrubber stream at
the time of sampling.
• Calibration - Capable of calibration and standardization
by a chemical and/or instrument technician or a person
of equivalent training and experience.
• Routine Operation - Capable of attended operation with
no more than four simple calibrations per day as accur-
acy checks.
• Maintenance/Repair - Capable of being maintained by a
qualified chemist or chemical technician with a maxi-
mum of two days special training. Capable of repair by
a chemical instrumentation technician who has received
reasonable maintenance and repair training.
-12-
-------
17551-6012-RO-OO
I Operating Voltage and Voltage Variations - Capable of
operation at 120 or 240 volts A.C. with line voltage
variations which may exceed +30%.
t Accuracy and Reproducibility - It is desired (but not
always possible) that all instruments will be relatable
after calibration to +5% of the true value for the
species being measured".
0 Output Signals - All output signals where practical should
be 100 mv or compatible with a specific readout system if
so specified.
2.2.2 Operating Specification
The operating specifications for each instrument will depend on prin-
ciples of measurement, sampling methods and end use requirements. Specifi-
cations defined at this time include:
t Ambient Conditions - Instruments must be capable of opera-
tion in the presence of or protected from the coal dust,
dirt, temperature extremes, vibrations, etc., encountered
in and around an enclosed coal burning power station.
• Sampling Time - A sample rate of 30 samples per hour per
species being measured is desirable.
• Response Time - Direct instream analysis is desirable,
however, when not practical a time delay between sample
collection and data presentation is acceptable if such
data is directly relatable to the process sample point
at the time of sampling. A maximum analysis time of
two minutes per sample per species is desirable.
2.2.3 Requirements for Slurry Sampling and Separation
In order to achieve separation between the solids and liquids of a
non-equilibrium slurry a rapid separation is required to "freeze" the com-
ponents of the non-equilibrium system.
t Slurry Sample Rate - In order to minimize perturbation
of the scrubber system not more than 1% of the stream
v/v should be sampled.
System Flow Rate Max. Sample Flow Rate
150 gal/min 1.5 gal/min
600 gal/min 6.0 gal/min
1200 gal/min 12.0 gal/min
-13-
-------
17551-6012-RO-OO
t In-Stream Sampling Velocity - Because of the nature of
wet limestone slurries and the broad distribution of
participate sizes, conditions approaching isoklnetlc
sampling are desirable. The sample flow velocity at
input point should be held within two times that of
the actual stream velocity at that point. This devi-
ation will depend on particle size, particle distribu-
tion, turbulence, etc.
• Sample Probes - Removal probes should be of such de-
sign and construction so as to remove a representative
sample.
• Lag Time - From time of sampling, complete separation
to pure liquid and solid phases should not take more
than <30 seconds. It can be assumed that if "all
other things are held constant" the resulting samples
will be representative for elements at least in the
system at an equal flow time down stream of the sample
time.
t Materials of Construction - In all cases care must be
exercised to ensure that all surfaces subject to sample
contact be an inert non-reactive material.
• Sample Transfer Lines - All sample transfer lines should
be constructed of inert material, e.g., Teflon, main-
tained at the sampling point temperature and as short as
practical to minimize dead volume.
• Particulate Size Solids Sample - The separation device
must be capable of providing an optically clear liquor
from a slurry containing particles ranging from less
than 0.5 micron to over several hundred microns.
• Solids Content - The solids separation system must be
able to handle solids loadings ranging from less than
3% to over 15% (w/w).
• Drying of Solids - A non-reactive solvent should be
utilized to wash the solids before final drying in
order to remove the liquid phase components present.
• Maintainability. Cleaning - The solids sampling system
must be easily cleaned. Backflush components or throw-
away cartridge filters are recommended.
t Power Requirements - Pumps and other electrical devices
should operate at 120 V AC +30V.
-14-
-------
17551-6012-RO-OO
2.3 REVIEW OF LIMESTONE WET SCRUBBING PROCESS OPERATION
As an integral part of the survey of the state of the art for lime-
stone wet scrubbing processes and for the purpose of identifying realistic
process characterization needs, several meetings were held with process de-
velopment engineering firms. Technical interchange (T.I.) meetings were
held with key engineering and chemistry personnel at Bechtel Corporation,
at their San Francisco facility, Zurn Environmental Engineers at their pilot
scrubbing installation at Key West Electric Company, and Kansas Power and
Light (KPL) to review the Combustion Engineering Company process. Field
samples of scrubber slurry were obtained and phase separation techniques
evaluated at the latter facilities. At a T.I. meeting held at TRW with
EPA and TVA personnel, tentative recommendations of separation and analysis
methods were discussed and numerous TVA synthesized samples of slurry com-
ponents were transmitted for the purpose of methods evaluation. The fol-
lowing paragraphs present the highlights of the process reviews and the
sampling that was accomplished. The resultant experimental effort are
described under the appropriate methodology headings.
2.3.1 Review of the Shawnee Process Demonstration Operation
The operation procedures planned by Bechtel Corporation for implementa-
tion at the Shawnee Power Plant wet scrubbing process demonstration were
reviewed. In addition to detailed discussion of the three types of scrubber
designs, namely, Venturi, TCA plastic ball and flooded marble bed designs,
information was provided concerning both process and instrument diagrams
and the planned operational modes for investigation of several different
configurations. After examination of the planned sample port locations,
it was concluded that sufficient sample ports will be available in the
planned facility. Because Bechtel had a cut-off time of January 28, 1971
for freezing the design of the plant hardware, any future recommendations
were not compatible with this time frame. If it should be necessary to
add new sample ports, this can be implemented by modifications of flanged
sections of the process stream piping hardware. A complete package of
Bechtel's planned operational mode sheets is given in Appendix C.
-15-
-------
17551-6012-RO-OO
2.3.2 Inspection of the Zurn Engineering Wet Limestone Scrubbing Operation
at the Key West Electric Company
As indicated above, the field trip to operational wet scrubbing faci-
lities had the express purpose of determining methodology and equipment
currently employed for process characterization, to evaluate sampling
schemes and to acquire real samples for methods evaluation and development.
With the cooperation of Zurn Environmental Engineering the pilot limestone
and coral wet scrubbing process unit operated at the Key West Electric
Company was inspected and sampled by A. Grant and E. A. Burns on January 13,
1971. This unit was scrubbing 1/8 of the flue gas from a 20 megawatt
boiler burning a fuel oil containing 0.6% sulfur. The scrubber apparatus
consisted of a modified particulate removal unit which provided turbulent
mixing and a high level of liquid/gas contact. The inlet gas had a temper-
ature of 250°F and a flow rate of 935 SCFM. The scrubbing solution studied
during the run inspected by TRW representatives consisted of a 3% w/v lime-
stone slurry. Table III lists the temperature and pH of the scrubbing unit
TABLE III
CHARACTERISTICS OF ZURN LIMESTONE SLURRY
Location
Inlet
Hopper
Weir Bottom
Spent Slurry
Filter Slurry
Temperature
(°F)
78
135
125
124
PH3
7.76
6.18
6.18
6.14
6.05
Obtained using a Welch Sargent pH Meter equipped with
automatic temperature compensation
slurry at various locations in the system. Instrumentation used by Zurn
consisted strictly of monitoring the sulfur dioxide conent in and out with
a Whittaker Dynascience instrument. To assist in defining problems asso-
ciated with phase separation and species isolation, some exploratory exper-
iments were conducted with the assistance of Dr. John Craig of Zurn
Environmental Engineering.
-16-
-------
17551-6012-RO-OO
The apparatus shown schematically in Figure 1 was used to provide fil-
tration separation of one liter within one minute. This apparatus 1) con-
sists of valving, manifold, filter and sample bottle, and 2) provides a
uniform, representative flow through the manifold in the by-pass mode of
operation. In the by-pass mode the remainder of the manifold was purged
with gaseous nitrogen (to eliminate residual air which could possibly oxi-
dize sulfite species in the slurry during subsequent handling operations).
Activation of the by-pass valve and vacuum pump initiated flow of slurry
from the bottom of the scrubber weir through the in-line filter unit (3 mi-
cron nominal size). The filtered liquid was optically clear and showed no
evidence of particulate matter. In separate experiments the separated
solids were 1) flushed with an inert halocarbon liquid (trichloroethylene)
or alternatively, 2) flushed with dry nitrogen to quench any possible post
sampling reactions of the solids.
Post sampling treatment of the liquids to inhibit subsequent reaction
of sulfite to sulfate was accomplished as shown in Table IV. In additi:n
TABLE IV
SULFITE FIXATION EXPERIMENTS ON FILTERED WEIR BOTTOM
LIMESTONE SLURRY LIQUOR
Treatment
Purpose
Resulting pH
Observations
1:10 v/v 8-hydroxy
quinoline solution
1:10 v/v glycerin
1:10 v/v glycerin
plus 1:5 v/v
formaldehyde solu-
tion (37% w/w)
To complex metals
and inhibit cata-
lytic reactions.
To retard sulfite
oxidation
To retard sulfite
oxidation and form
stable aldehyde-
bisulfite complex
6.05 Copious yellow
precipitate
6.22 Clear liquid
9.11 Clear liquid
to obtaining liquid and solid samples from the scrubber unit, samples were
also obtained at both the make up water and starting limestone, as well as
55 gallons of spent scrubber slurry which was used in the evaluation and
development of candidate separation techniques.
-17-
-------
ZURN
ENGINEERING
SCRUBBER
00
i
SAMPLE PORTS
(SEE NOTE)
3-WAY VALVES
1/4" BALL VALVE
WHITEY S.S.
NOTES
SAMPLING PORTS IDENTIFIED:
1. SPENT OUTLET
2. INLET
3. WEIR
4. HOPPER
INERT FLUSHING LIQUID
OR N2 PURGE GAS
ro
i
50
O
I
O
O
WASH
FILTER
SAMPLE
BY-PASS AND
UNFILTERED
SAMPLE
*J
4
A
^
SAMPLE
G)VACUUM
PUMP
IN-LINE
ACROFLOW
FILTER
BOTTLE
Figure 1. Sampling System for Zurn Wet Scrubber
-------
17551-6012-RO-OO
These studies demonstrated that rapid filter separation of the slurry
can be accomplished under anaerobic conditions in short periods of time and
laid the foundation for the recommended sampling/filtration procedure for
continuous sampling (See Section 3.1).
2.3.3 inspection of the Wet Limestone Scrubbing Operation of the Kansas
Power and Light Company
On February 10, 1971, the limestone scrubbing process unit at the
Kansas Power and Light Company (KPL) was inspected by E. A. Burns. Dis-
cussions with Lee Brunton of KPL revealed this unit consists of marble bed
scrubber of the effluent from the 115 megawatt coal fed (3.5% sulfur) com-
bustion unit. In this unit limestone is injected dry into the combustor
and the effluent particulate consists of 1.5% w/w going into the effluent
pond of which half of that (0.75% w/w) consists of fly ash from the coal.
Combustion Engineering, Windsor, Connecticut, is responsible for the sam-
pling of this unit and grabs control samples at both the marble bed and
the effluent pond. At one time they had experienced a serious calcium sul-
fate/sulfite scaling problem but this no longer is occurring; this problem
was attributed to operational conditions at a relatively high pH (7.5-8.0).
When the unit is operated in a recycle mode they have demonstrated 85% sul-
fur dioxide removal compared to a 60-70% removal without recycling. After
discussions by phone with Jim Martin (Combustion Engineering), authoriza-
tion was given to TRW to obtain samples of 1) the scrubber effluent going
into the pond, and 2) the make-up water (that coming out of the pond after
many days of settling). These samples were returned to TRW for character-
ization.
2.4 INSTRUMENTAL METHODOLOGY FOR CATIONS AND ELEMENTS
The list of elements in solution presented in Table I is of signifi-
cance for determining the required process mass balance about a 20% sulfur
closure. However, laboratory analysis of numerous other elements will un-
doubtedly be required in the forthcoming demonstration tests. The origin
of these species are in the trace component variations in the coal, scrub-
bing agent and make-up water while the importance of their determination
lies in understanding catalytic reactions or other spurious side effects.
Table V delineates the elements of concern in their probable ion or compound
forms.
-19-
-------
17551-6012-RO-OO
TABLE V
POSSIBLE LIMESTONE/DOLOMITE SLURRY COMPONENT DISTRIBUTION3
Major Components
Liquid Phase Solid Phase
Ca2+ CaO
Mg2+ MgO
HSO ' Ca(OH)9
0 ^.
SO/' Mg(OH)2
H9CO, CaSO,
c.
-------
17551-6012-RO-OO
2.4.1 Laboratory X-Ray Spectroscopic Methodology
The X-Ray fluorescence (XRF) technique for quantitative laboratory
analysis of elemental distribution, notably total sulfur, calcium, magne-
sium, as well as other metals, in slurry liquor and solids was evaluated
primarily as an adjunct to process instrumentation development. The results
of this study have, in addition, definitely identified XRF as one of the two
recommended techniques for laboratory instrumental analysis. A concise re-
view and recommendation cf dispersive and non-dispersive X-Ray equipment is
presented in Appendix D. Because of the emphasis on process instrumentation
only those companies offering units amenable to process control were included
in this effort. Consequently, the findings are not all inclusive for labora-
tory instruments because there are several other instrument manufacturers
that may offer acceptable instruments.
The evaluation and development of the technique was accomplished through
a review of pertinent literature, in-house experimentation, vendor contacts
and experimental evaluation at vendor laboratories.
2.4.1.1 XRF Applications in Literature - Results of typical laboratory X-Ray
analysis of limestone are listed in Table VI showing that excellent precision
can be achieved with this instrumental technique.
TABLE VI
XRF ANALYTICAL DATA FOR LIMESTONE (REFERENCE 5)
Element
Ca
A12°3
Si02
Fe203
MgO
Ko Energy
kev
3.69
1.49
1.74
6.4
1.25
Range
% w/w
35-52
0.5-3
0.5-30
0.5-1.5
1-6
Cone. Found
% w/w
41.2
2.0
10.8
0.95
1.8
Avg. Dev.*
% w/w
0.36
0.06
0.26
0.01
0.08
Precision
% w/w
+0.055
±0.057
+0.15
+0.003
+0.08
*Average Duplicate Analysis - ARL VPXQ
Comparison of X-ray fluorescence to standard wet chemical methods have
been reported by Fabbi (Reference 6) (Table VII) for the single most im-
portant element that must be determined in the slurry solids - sulfur. Much
additional statistical data are available that indicates the adequacy of the
techniques in terms of sensitivity, accuracy and precision for sulfur and the
other cations and metals of interest (Ca, Mg, Fe, Mn, V, Co, etc.).
-21-
-------
17551-6012-RO-OO
TABLE VII
COMPARISON OF XRF AND CHEMICAL VALUES FOR TOTAL
SULFUR IN STANDARDS AND MINERALIZED SAMPLES
No. Rock Type
Sulfur (% w/w) '
Total
Sulfur
by XRF
1 Tuff or ash 0.01
Total
Sulfur by
Na2C03
Fusion
0.00
> i
2 Diabase j 0.20
3 Tuff
4 Crush rocks adit
5 Altered quartz
6 Sulfidic phyllite
7 Sulfidic phyllite
8 Altered augite
biotite monzonite
9 Augite biotite
monzonite
10 Hybrid aplite
11 Hybrid pyroxenite
12 Hybrid aplite
13 Hybrid monzonite
14 Hybrid monzonite
15 Hybrid aplite
16 Hybrid aplite
1.03
1.97
2.42
3.35
4.97
1.62
0.07
1.06
0.11
0.79
0.64
0.08
0.42
1.10
17 Hybrid pyroxenite ; 0.18
0.19
1.16
1.85
2.52
3.10
4.97
• • *
• • •
• • •
• • •
• • •
• • •
• • •
• • •
* * •
Sulfur by
Aqua Regia
Solution15
• » •
• • •
• • •
• • •
• • •
• • t
1.60
0.07
1.00
0.10
0.80
0.64
0.16
0.44
1.20
0.16
aAnalyses by conventional gravimetric methods. U.S. Geological Survey
Analytical Laboratories under the direction of L.C. Peck
bAnalyses by rapid methods. U.S. Geological Survey, Analytical
Laboratories under the direction of L. Shapiro
2.4.1.2 XRF Analysis of Simulated and Field Samples - Prior to and concur-
rently with experimentation at vendor application laboratories (discussed be-
low), numerous samples of simulated and actual field samples of starting
limestone, fly ash, scrubber liquor, and scrubber solids were analyzed by XRF
to demonstrate the applicability, accuracy and cost effectiveness of this
technique. The instrument utilized at TRW for all in-house analyses was a
-22-
-------
17551-6012-RO-OO
General Electric Model XRD-5 spectrometer. Samples were received directly
fron J. Barkley of TVA and consisted of simulated liquids and solids. Actual
field samples of trapped solids from the Duke Power Company were submitted
through R. M. Statnick, OAP Project Officer. Descriptions of samples, the
analytical procedure and a discussion of results are presented in Appendix D.
It is noteworthy that the XRF analysis of several of the proposed speci-
mens detected some errors in the assumed composition. In the case of cal-
cium taken in solid samples identified as No. 1, No. 2 and No. 3, the cal-
cium present in the added fly ash was not taken into consideration. Correc-
tion for the 8.51% calcium in the ash as determined by XRF gave relative er-
ros of 1.5, 1.0 and 0.5% for calcium contents which ranged from 16% to 30%
w/w. The value of the XRF technique was further exemplified when a discrep-
ancy in the reported sulfur value taken versus that found was elucidated by
interpretation of the XRF data. In essence, the analytical data were suffi-
ciently precise and consistent to permit identification of the use of an an-
hydrous sulfate salt as opposed to a dihydrate salt as reported for sample
preparation. Based on this postulation, relative errors for total sulfur were
computed to be 6.6% (0.4% absolute), 1.6% and 0.
2.4.1.3 X-Ray Equipment Vendor Contacts - Literature on laboratory and pro-
cess X-ray analysis equipment was requested from the following list of ven-
dors:
ARL Philips
Siemens JEOL Inc.
G. E. Picker
Information was received from each, however, only ARL and G.E. offered
on-line process equipment. Consequently, within the constraints described
previously, only their units have been evaluated in detail.
In the course of acquisition of information on available instruments,
an alternative approach was identified which is capable of monitoring sulfur
content and other elements in the wet scrubber operation. This alternative
involves non-dispersive X-ray spectroscopy and utilizes radiation from radio-
active sources (Fe, Cd, Am) as the primary source of excitation.
-23-
-------
17551-6012-RO-OO
The following seven firms sell these instruments:
Canberra Industries, 45 Gracey Avenue, Meriden, Conn.
KEVEX Corporation, 898 Mahla Road, Burlingame, Calif.
Nuclear Equipment Co., 931 Terminal Way, San Carlos, Calif.
Nuclear Diodes, P. 0. Box 135, Praire View, 111.
ORTEC, 101 Midland Road, Oak Ridge, Tenn.
Panametrics, 221 Crescent St., Waltham, Mass.
Princeton-Gamma-Tech, Box 641, Princeton, N.J.
Of these, KEVEX and Nuclear Instruments were contacted for information on
performance of their respective instruments. Several major points should
be made:
• Sulfur can be detected with the isotope-activated non-
dispersive systems. KEVEX personnel have demonstrated
that 0.02% S in steel can be detected. Those experi-
ments were admittedly performed under special condi-
tions and it seems reasonable to expect useful analysis
at Q.2% sulfur level under less than optimal conditions.
Nuclear Equipment Company has used their standard system
to determine that certain solid samples from the New
York Air Pollution Control District contained 0.15T^
sulfur.
• A number of the units require liquid air refrigeration
and cannot be allowed to warm up to room temperature.
The samples need not be cooled.
t The units utilize 25-100 mC of radioactive material
(Fe55, Cd108, Am21tl) and require a license from the
Atomic Energy Commission. The radioactive material
does not mix with the slurry stream.
• The KEVEX Corporation instrument has an automated sam-
ple handling capability.
• Elements other than sulfur can be detected and in the
ideal case, the lowest atomic number element that can
be detected is #11 (sodium). For on-line slurry situ-
ations, sulfur may be the lowest atomic number detected.
• The energy selectivity of the non-dispersive system is
an order of magnitude poorer than for other X-ray fluor-
escence techniques and could be a source of trouble in
very complex samples that would offer no trouble to an
ARL-type unit.
t The approximate costs of the non-dispersive units are
from $14K to $28K depending on the type of data re-
trievable that is necessary.
-24-
-------
17551-6012-RO-OO
As a consequence of these findings, additional consideration was given
to the isotope activated non-dispersive X-ray spectrometer technique for
monitoring of wet scrubber slurries.
2.4.1.4 Experimental Evaluation at Vendor Application Laboratories - For
the purpose of evaluating instrumentation at the two selected vendor appli-
cation laboratories, i.e., ARL and KEVEX, samples were prepared using pul-
verized fresh and spent scrubber solids as matrices together with known
quantities of calcium sulfate and iron which were homogeneously cast into
plastic resins which are substantially transparent to X-rays. Thus the
particulate samples were frozen in solids at concentrations which are likely
to be found in 1) dynamic scrubber samples of high and low concentrations
or in 2) static samples which contain relatively low residual water content.
The sample specimens consisted of a 1.25-inch diameter cylinder having a
0.125-inch thickness. The simulated slurry samples had the composition
identified in Table VIII, which were prepared in either epoxy resin (in the
case of the low solid concentration) or polyvinyl alcohol (in the case of
the 99% solids loading).
At the Applied Research Laboratories at Sunland, California, the 14
listed samples were analyzed on an ARL X-ray Quantometer 7200. Although
this instrument is a laboratory unit and only accomodates batch samples, it
contains the same basic X-ray fluorescent spectrometer as the process unit
(Model PCXQ) and, therefore, it served the dual purpose of evaluating both
types of equipment.
A more complete description of the visit to the ARL Application Laboratory
with a discussion of limits of detection, repeatibility, matrix effects,
quantitative analysis, particle size effects and liquid cell signal attenu-
ation is presented in Section 3.2, Continuous On-line X-ray Fluorescence
Methods for Cations and Elements. The complete technical and economic evalu-
ation of this instrument compared to G.E. and KEVEX instruments is presented
in Appendix D. The ARL laboratory instrument was found to meet or exceed
technical requirements of:
• Sensitivity - >0.1% for Ca, Mg, S and other elements
of interest
% Precision - >Z% of measured value
-25-
-------
17551-6012-RO-OO
TABLE VIII
XRF STANDARD SAMPLES FOR INSTRUMENT EVALUATION
TRW
Sample
001
002
003
004
005
006
007
008
009
010
on
012
013
014
^^••^•i
Material
CaC03
CaC03 +7.9% CaS04
CaC03 + 14.6% CaS04
Limestone
Limestone +7.9% CaS04
Limestone + 14.6% CaS04
CaC03 +7.7% CaS04 + 0.9% Fe
Limestone/TVA-Flyash
Limestone/Zurn-Flyash
TVA-Flyash + 98% Epoxy
TVA-Flyash +91% Epoxy
Zurn-Flyash 98.5% Epoxy
Zurn-Flyash +91% Epoxy
CaCO, +7.2% CaSO, +7.2%
Na,SO.j *
Form
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Purpose or
Function
Standard for
background
Determine de-
tectabillty
limit for
sulfur
Determine de-
tect ability
limit for
sulfur
Standard for
background
Detectability
limit
Detectability
limit
Determine
interference
Determine
S/Ca ratio
Determine
S/Ca ratio
Effects of
dilution
Effects of
dilution
Effects of
dilution
Effects of
dilution
SO* - S04=
resolution
^^^^^^^•••M
Sulfur
Content (% w/w)
1.85 calculateda
3.44 calculated
0.05 nominal
wet chemical
1 .87 calculated
3.44 calculated
1.81 calculated
1 .765 wet
chemical
0.045 wet
chemical
-
-
-
-
3.43 calculated
•^••^••^^^^^^
aThe calculated sulfur percent values are based on the known stoichlometric
ratios in CaC03, CaS04, and Na2S03-
-26-
-------
17551-6012-RO-OO
Utilizing the XRF technique and an ARL lab unit it is estimated that approx-
imately 1000 elemental analyses can be performed in an eight-hour shift by
a single technician at a cost of ^$0.96/element. These specification time
and cost factors are far superior to any other technique with the possible
exception of the atomic absorption technique for liquids.
2.4.2 Atomic Absorption Spectrophotometric Methodology
Laboratory experiments were performed to evaluate and develop accepta-
ble atomic absorption procedures (AA) for the rapid determination of calcium
and magnesium primarily, but also other elements of concern such as sodium,
potassium, iron, manganese, titanium, etc., in the scrubber slurry. The in-
vestigation was limited in scope to studies of the applicability for analy-
sis of clarified or filtered slurry liquor exclusively. However, dissolution
of solids with appropriate dilution is standard AA technique that requires
more time in sample preparation but it can be equally applicable to the slurry
solids for quantitative analysis in the absence of the recommended XRF capa-
bility.
Analyses were conducted utilizing a Perkin-Elmer Model 290 Atomic
Absorption Spectrophotometer. It should be pointed out that significant
instrumentation advancements have been made recently to yield much higher
sensitivities, and more versatile units with multiple lamp turrets for more
rapid element change over and to capability for simultaneous atomic absorp-
tion/atomic emission.
Of primary concern were 1) pH effects, 2) optimum fuel and oxidizer,
and 3) effects of established methods of interference inhibition. The re-
sults of this study are delineated in Table IX. These findings show that
calcium, in a solution made from saturated filtered limestone solution which
has then been treated with a small amount of H^SO., and finally adjusted with
H SO* to pH 4 to 7, is stable in the burner flame.
Best results were obtained by using a lanthanum inhibitor and acetylene/
N?0 flame. At pH 2, absorption decreases to an unreliable reading even when
the inhibitor is used. The literature indicates that this effect is produced
not by the increased acidity as such, but by a change in the viscosity and
surface tension of the solution. This, in turn, produces a decrease in
atomizer efficiency resulting in inaccurate response.
-27-
-------
17551-6012-RO-OO
TABLE IX
EFFECT OF FLAME OXIDIZER, pH AND INHIBITOR ON ATOMIC ABSORPTION
DETERMINATION OF CALCIUM IN LIMESTONE SOLUTIONS3
Solution
PH
9.5
7
6
4
2
Acetylene/Air Flame
With Inhibitor
Calcium Found
% w/w
0.01
0.02
0.021
0.02
—
Acetylene/N20 Flame
with Inhibitor Without Inhibitor
Calcium Found Calcium Found
% w/w % w/w
0.02
0.026
0.026
0.026
0.011
0.01
...
0.026
___
0.0006
Analytical calcium concentration taken = 0.026 % w/w
Typical suspect constituents present in filtered slurry samples obtained
from the operational limestone wet scrubber units described above, were ana-
lyzed by atomic absorption spectrometry. The results of these tests are re-
ported in Table X.
TABLE X
ATOMIC ABSORPTION ANALYSIS OF SELECTED SAMPLES OBTAINED
FROM OPERATIONAL LIMESTONE WET SCRUBBER UNITS
Element
Magnesium
Calcium
Sodium
Potassium
Copper
Mercury
Iron
Nickel
Chromium
Manganese
Aluminum
Titanium
Zurn
Feed
Slurry
Liquor
ppm
4000
435
10800
420
0.35
<4.0
<0.1
<0.2
<0.05
0.02*
<1.6
<5.0
Zurn
Spent
Slurry
Liquor
ppm
2700
1088
6800
420
0.10
<4.0
<0.1
<0.2
<0.05
0.16
<1.6
<5.0
KPL
Spent
Slurry
to
Clarlfier
ppm
<0.01
1360
1500
20
<0.05
<4.0
<0.1
<0.2
<0.05
<0.02*
<1.6
<5.0
Make-up
Liquor
from
Clarifier
ppm
0.4
340
1200
8.2
<0.05
<4.0
<0.1
<0.2
<0.05
<0.02
<1.6
<5.0
MMB^M^^^_i
1 div ^ limit of detection of instrument utilized
-28-
-------
17551-6012-RO-OO
For the purposes of further evaluating the AA technique (as well as the
other candidate methods presented in subsequent sections of this report), a
simulated, static liquid slurry was prepared from reagent chemicals and de-
ionized water. The added quantities of anions and cations are delineated
below together with found concentrations in solutions.
Constituent
Ca2+
Mg2+
Fe3+
Cu
Mn2+
A13+
Fe2+
Added
(ppm)
1000
1040
25.0
1.1
1.2
5.2
25.2
Found
480 C\2mM]
1040 (43mM)
£0.1
<0.05
1.2
<1.6
<0. 1
Constituent
K+
Na+
NT2+
so42-
N03~
Cl"
co32-
Added
450
200
1.0
6400
570
60
500
Found
450 (11 .5mW)
200 (8.7mW)
0.1
5828 (61mM)
570 (9.2mW)
60 (1.7mA/)
315 (5.2mW)
All found cation values were determined utilizing the P-E Model 290 AA unit.
Excellent correlation was found for the more soluble species Mg, Mn, K and
Na, adding further credance to the arguments in favor of the AA technique
over other proposed techniques, especially the manual EDTA titration methods.
The anion determinations were performed by the standard methods described in
Sections 2.6 - 2.8. The apparent mass balance discrepancies can be ex-
plained by partial precipitation of some of the species by carbonate, sul-
fate and hydroxide species. Although no single AA instrument can be recom-
mended for this application, the following candidate companies are suggested
as a minimum list for consideration.
Bausch & Lomb
Beckman
Corning
Instrument Laboratories
Varian
Jarre! Ash
Perkin-Elmer
Spectra Metrics
Technicon
2.5 SPECTROPHOTOMETRIC ANALYSIS OF SULFITE AND BISULFITE
Sulfite ion or the S(IV) species in solution was identified early in the
program as one of the major species to be determined in 1) assessing process
efficiency and optimization and 2) the performance of a mass balance about
-29-
-------
17551-6012-RO-OO
sulfur. In several wet scrubbing processes the ratio of sulfite to sulfate
is critical to the economics of the process in terms of reagent regeneration
or ultimate product synthesis for scrubbing process credit.
During the review of candidate laboratory analytical methods for the
determination of dissolved sulfur dioxide as HS03" and S03~, it was deter-
mined that no satisfactory instrumental methods existed for 1) determining
the anticipated concentrations (Table I) in the limestone slurry mixtures,
or 2) met the general criteria of speed, accuracy, automatability, adapta-
bility to routine operation by semi-skilled operators and use of relatively
low cost equipment. One other important aspect in considering candidate
methods was the likelihood of successful development and implementation of
the method by January 1972.
Experimental or theoretical evaluation studies of several spectro-
photometric methods including methods based on direct sulfite absorbance,
pararosaniline, fast blue salt, fuchsin-aldehyde, 5-aminofluoroescein and
furfural. As a result of this evaluation the furfural method was deemed to
offer the best potential for meeting the above criteria and indeed, a new
method based on bisulfite bleaching of the furfural UV absorption was de-
veloped. This method is based on the chemical reactions in Equations 1 - 4
and depends on the bleaching of the 276 nM absorption of furfural by reac-
tion with bisulfite.
HS03" * C4H3OCHOHS03~ (1)
C4H3OCHOHS03H t C4H3OCHOHS03" + H+ (2)
3
H+ + HS0" (3)
HS03" j H+ + S03= (4)
The absorbance, A, at 276 nM is directly related only to the amount of fur-
fural in solution when the pH of the media is maintained around 4.0 in ac-
cordance with the Lamber-Beer-Bouguer Law.
A = abcF (5)
where a = molar absorptivity of furfural, liters/mol-cm
b = optical path length, cm
Cr = concentration of uncombined furfural, M
-30-
-------
17551 -601 2-RO-OO
The equation of the bleaching reaction (Equation 1) is governed by the
formation constant, K
CF[HS03']
(6)
where c. = concentration of furfural -sulfite adduct, M
= co -
[HSCL~] = concentration of uncombined bisulfite
M
c = analytical concentration of bisulfite taken, M
Combining Equations 5 and 6 results in a relationship of absorbance and bi-
sulfite ion as shown in Equation 7.
\ - "
Consequently, the function of I/A is linearly proportional to the free bi-
sulfite concentration; also, the formation constant can be calculated from
the ratio of the slope to intercept of the straight line relationship. It
is interesting to note that this method was first developed for the deter-
mination of furfural and prior to this study has not been used for the de-
termination of bisulfite (Reference 7). The reason for this is because in
most situations colorimetric procedures are used for determining low concen-
trations of chemical species but in the limestone scrubber case the concen-
tration of bisulfite (1-150 mM) is too large for trace analysis methods
(without massive dilution) and not readily adaptable to common macro titri-
metric procedures (without using large volumes and dilute titrants).
Detailed studies of the effect of pH, diverse ions, temperature and
time to constant color development are described below. The effort has
culminated in a rapid, simple instrumental method which can utilize a single
reagent condition consisting of furfural, phosphate buffer and sulfamic
acid (to remove trace concentrations of nitrite interference). The repro-
ducibility of the method has been determined to be better than 2% relative
or 0.2 mM absolute whichever is higher. The formalized procedure in a for-
mat geared to chemical technician implementation is presented in Appendix E.
-31-
-------
17551-6012-RO-OO
2.5.1 Effect of pH
As the furfural bleaching was reported to be a direct function of com-
plexatlon by the bisulfite 1on (Equation 1 above), 1t was necessary to max-
imize the bisulfite concentration through pH control and determine the effect
of varying pH on the reaction equilibria. The relative concentrations of
sulfur (IV) species, i.e., undissolved sulfurous acid, bisulfite ion and
sulfite ion, as a function of pH are shown in Figure 2. Based on the bi-
sulfite curve, it was originally postulated that pH control in the region of
approximately pH 3.5 to 5.0 would be useful for analytical purposes. The
actual optimum pH range was determined experimentally and is described below.
pH
Figure 2. Sulfur IV Species as a Function of pH
Two principal buffer systems were evaluated, i.e., acetate-acetic acid
and phosphate-phosphoric acid. Prior to determining the effect of pH on
sulfite determination, the effect of pH on the absorbance of neat furfural
(baseline) was investigated. The results of these tests are shown in Table XI
-32-
-------
17551-6012-RO-OO
TABLE XI
pH DEPENDENCE OF FURFURAL ABSORPTION SPECTRUM AT 276
pH
7.88
6.94
6.52
6.40
4.24
4.24
3.70
3.51
2.80
Absorbance
0.777
0.775
0.772
0.768
0.760
0.760
0.758
0.767
0.772
Buffer System
-
-
-
-
Phosphate-Acetate
Phosphate
Phosphate
Phosphate
Phosphate
Concentration of furfural = 50.7 uM
It is interesting to note that an apparent minima occurs in the absorbance
of furfural in the 3.7 - 4.3 pH range. Although the differences ar'e small,
the monotonic nature of the curve lends credence to the belief that the
effect is real. As a consequence, it appeared desirable from both the sta-
bility of the furfural absorption and the major effect on ensuring the
maximum bisulfite S(IV) form to definitize further the optimum pH range
for the furfural bleaching method.
Utilizing reagent grade sodium bisulfite in an unbuffered system,
absorbance versus concentration of S(IV) gave the data in Table XII. The
least squares plot for these data gave
I/A = 2.359C + 1.349 (8)
Where C = analytical concentration of bisulfite millimolar
The values listed for bisulfite found were obtained by Equation 8.
Linearity deteriorated appreciably for bisulfite concentrations greater
than 2.25 mM. The sulfite concentration range of interest (Table I) was
identified from equilibrium calculations to be 1 to 150 mM. Consequently,
it will be necessary to dilute the scrubber liquor to a point in which the
bisulfite content does not exceed 2.25 mM to obtain reliable results.
-33-
-------
17551-6012-RO-OO
TABLE XII
DETERMINATION OF BISULFITE BY FURFURAL BLEACHING IN UNBUFFERED MEDIA
Bisulfite"
Taken
mM
0.000
0.106
0.211
0.317
0.422
0.528
0.845
1.06
1.12
2.11
2.25
•M^HMMM
Final
pH
6.94
5.25
4.70
4.40
4.30
4.12
3.96
3.93
3.80
3.60
3.3
IMBH^HH
Absorbance (A)
0.776
0.643
0.541
0.478
0.427
0.383
0.295
0.257
0.245
0.163
0.148
I/A
1.228
1.555
1.848
2.09
2.34
2.61
3.39
3.89
4.08
6.13
6.75
Bisulfite
Found
mM
-0.023
0.090
0.214
0.316
0.422
0.536
0.866
1.08
1.16
2.03
2.29
Deviation
mM
-0.023
-0.010
+0.003
-0.001
0.000
+0.008
+0.021
+0.020
+0.040
-0.080
+0.040
Solutions prepared by dilution of stock sodium bisulfite solutions
(acidified to give pH =3.1)
The analytical concentration of furfural in the final solution was
50.7 MM
The results of analysis of standard sodium bisulfite solution using
the sodium acetate/acetic acid buffer system are tabulated in Table XIII.
As can be seen, excellent precision was obtained in the 3.6 to 3.9 pH range
The sodium dihydrogen orthophosphate/phosphorlc acid buffer permitted ex-
tension of the furfural-bisulfite method Into the pH 4 region, thus, to-
gether with the acetate system, allowing pH control through the theoreti-
cally useful range of the technique.
The determination of bisulfite by furfural bleaching in the phos-
phate buffer system was performed at two pH regions, 3.6 and 4.4 with the
combined results listed in Table XIV. The furfural concentration in these
experiments was 50.1 vM and the bisulfite content was varied between 0.2
and 2.3 mM. The relationship between absorbance and bisulfite content for
this data is given by the following expression which is an average of the
two sets of constant pH data:
I/A = 2.522C + 1.329 /9)
The significant differences between Equations 8 and 9 clearly point out the
need for measurement in a buffered (constant pH) media.
-34-
-------
17551-6012-RO-OO
TABLE XIII
DETERMINATION OF SULFUR (IV) BY FURFURAL BLEACHING
Sulfur (IV)
Taken3
mM
0.000
0.101
0.202
0.306
0.407
0.509
0.814
1.020
P«b
3.9
3.9
3.8
3.7
3.7
3.6
3.6
3.6
Absorbance (A)c
at 276 nM
0.760
0.643
0.559
0.483
0.437
0.388
0.303
0.264
I/A
1.315
1.555
1.789
2.07
2.29
2.58
3.30
3.79
Sulfur (IV)
Found
mM
0.001
0.100
0.196
0.311
0.401
0.520
0.815
1.016
Deviation
mM
+ .001
-.001
-.006
+ .005
-.006
+ .011
+ .001
-.004
aAdded from stock sodium sulfite solutions, concentrations listed are
those of diluted analyzed solutions.
bA sodium acetate - acetic acid buffer was utilized for pH control.
cThe analytical concentration of furfural in final solution analyzed
was 50.7 \iM
where C is the concentration of bisulfite (mM), the intercept is 1.329
(similar to that obtained for the acetate buffered system) and the slope
is 2.522. The slope for the acetate buffered system was 2.439. This equa-
tion was used to calculate the values for bisulfite found in the table. The
effect of pH <3.6 and pH ^4.4 can be evaluated by treating the data as two
separate sets of data as shown in Table XIV. The least squares equations
and the errors about the equations are listed. At pH ^3.6,
I/A = 2.527C + 1.346 (10)
while at pH ^4.4,
I/A = 2.513C + 1.305 (11)
The significance of the difference between these equations and the equation
(9) derived from the composite data has been evaluated by statistical treat-
ment of the errors. The significance of the differences was tested by com-
parison of the variance ratio (or F test) of the pooled variance of the two
equations with the variance of the composite equation.
-35-
-------
Bisulfite Taken
. [mtf]
0.000
0.227
0.511
0.511
1.520
1.520
0.000
0.194
0.437
1.500
2.230
TABLE XIV
EVALUATION OF PH EFFECT IN DETERMINATION OF S(IV) IN PHOSPHATE BUFFER
JDH
3.5
3.5
3.5
3.7
3.5
3.7
4.2
4.3
4.3
4.5
4.6
Absorbance(A)
at 276nM
0.763
0.529
0.379
0.377
0.190
0.190
I/A = 2.
I/A
1.311
1.890
2.634
2.653
5.263
5.263
527C + 1.346(+0
Variation for
0.760
0.557
0.421
0.196
0.145
I/A = 2.
for
51 3C +
Variation for
Standard
for
intercept
slope
1.316
1.795
2.375
5.102
6.897
1.305(+0
intercept
slope
deviation
Bisulfite
Found FmA/1
0.014
0.215
0.512
0.517
1.550
1.550
Deviation
[mAfl
+0.014
-0.012
+0.001
+0.
+0.
+0.
006
030
030
Relative
Error(%)
1.4
5.3
0.?
1.
2.
2.
2
0
0
.00476)
= 0
= 0
.00166
.00112
0.004
0.195
0.425
1.511
2.225
+0.
+0.
-0.
+0.
-0.
004
001
012
on
005
0.
0.
2.
0.
0.
4
5
7
7
2
.000604)
- 0.
= 0.
= 0.
000245
000165
0246
^•^•^H
^m^mmm
^^^
(10)
(11)
^^^^•^•B
01
01
ro
i
§
o
o
-------
17551-6012-RO-OO
The variances for the individual and pooled data are as follows:
at pH 3.6: S2 = .00476 9 5 d.f.
xy
at pH 4.4: S2x_y = .000604 @ 3 d.f.
S2 • <5 x -00476) t (3 x .000604) , 8
P o
For all data, the variance is:
S2 = 0.00345 @ 10 d.f.
Thus, the ratio gives:
.00345
h8,10 ~ .0032 = k08
which is not significant at the 99% confidence level.
Using the two buffer systems (acetate and phosphate) and combina-
tions, where necessary, an experimental curve of bisulfite concentration
versus pH was obtained. For each point on the curves, the solutions con-
taining l.OmM and Q.SmM bisulfite and 5 x 10" M furfural were adjusted to
the desired pH with the appropriate buffer and the furfural absorption
(A) measured at 276 my. Because the term I/A is directly proportional to
bisulfite concentration, the shape of the I/A vs. pH curve (Figure 3) pro-
vides a measure of the useful pH range (at maximum I/A), as well as the
effect of small pH variations on bisulfite determination in this range.
The experimental curve for the 1 mM concentration of HS03" com-
pares favorably with a theoretical best fit curve generated by computer
for the idealized quadratic equation:
2
y = a + bx - cx (12)
Examination of the curve and experimental and computed I/A values
together with the relative error, listed in Table XV, indicates excellent
agreement with the theoretical curve (Figure 3). Assuming a maximum ex-
perimental relative error of +3% in the pH region of maximum bisulfite con-
centration, the usable analytical pH range extends from 3.5 to 4.5. One
can infer from Figure 3 that the method may be applicable to the determina-
tion of HS03 in HS03" - S03~ mixtures, however, the accuracy of the mea-
surement at low HS03" concentration plus the apparent pH dependency of molar
absorptivity preclude the use of the method for this determination.
-37-
-------
CO
Ul
Ul
I
cr>
o
i
§
o
o
Figure 3. Bisulfite-Furfural Absorption as a Function of pH
-------
17551-6012-RO-OO
TABLE XV
EXPERIMENTAL AND COMPUTED pH VERSUS 1/A(HS(K~)
1
2
3
3
3
3
4
4
4
4
4
5
6
EH
.66
.52
.35
.53
.73
.80
.00
.25
.34
.40
.40
.50
.85
1
2
2
2
2
2
2
2
2
2
2
2
1
I/A
.0740
.3536
.9378
.9483
.9768
.9088
.9429
.9151
.8701
.9018
.9880
.3704
.0604
I/A,, Deviation in I/A
1
2
2
2
2
2
2
2
2
2
2
2
0
v
.2189
.2171
.7893
.8627
.9231
.9389
.9692
.9757
.9694
.9628
.9628
.4853
.9771
+0
-0
-0
-0
-0
+0
+0
+0
+0
+0
-0
+0
-0
.1449
.1365
.1485
.0856
.0537
.0301
.0263
.0606
.0993
.0610
.0252
.1149
.0833
Relative Error, %
13.5
5.8
5.1
2.9
1.8
1.0J
0.91
2.1
3.4
2.1
0.8
4.8
7.8
I
=
f «/>
fD
C
^ •"
-$
0>
3
rt>
y = a + bx - cx
y = -1.874 + 2.326 x -0.2788 x2
2.5.2 Effect of Diverse Ions and Total Ionic Strength
The effect of diverse ions expected to be present in the wet scrubber
filtered liquors on the bisulfite-furfural bleaching method was investi-
gated. The results of these screening studies are shown in Table XVI at
relatively high additive concentrations.
Threshold interference level of that concentration of species causing
an absorbance deviation greater than 3% relative from an identical sample
without the interferent was used to define a positive interference.
Calcium, magnesium, sodium, nickel, copper and manganese ions were deter-
mined to be non-interfering at concentrations expected 1n the wet scrubber
media. The following threshold values were determined experimentally for
Fe"1"1"*", Fe++ and N02":
-39-
-------
17551-6012-RO-OO
Interference Level.
0.001 - 0.01
>0.01
>0.01
Concentration in the final diluted analytical sample
where the S(IV) concentration is optimally on the
order of 1 mM, the furfural concentration is 0.1 mM
and pH is adjusted to 4.0 with phosphate buffer.
TABLE XVI
INTERFERENCE SCREENING TESTS FOR BISULFITE-FURFURAL BLEACHING3
Bisulfite Additive
Concentration
mM Compound
0.00
0.00 CaS04
0.00 MgS04
0.00 NaCl
0.00 Fe(N03)3
0.00 NaN02
0.528 |
0.528 CaS04
0.528 MgS04
0.528 i NaCl
i
0.528 j Fe(NO,),
1 -00
0.528 j NaN02
Concentration
mM
_
Absorbance
0.755
1
0.0367
10.20
9.00
9.75
10.50
-
0.0367
10.20
9.00
9.75
10.50
0.769
0.765
0.767
>2.0
0.850
0.408
0.398
0.392
0.392
>2.0
0.831
"Unbuffered
b
Analytical concentration of furfural = 50.7 pA/; absorbance at 276 my
-40-
-------
17551-6012-RO-OO
Carbonate was added as sodium bicarbonate, and nitrate was added as
potassium nitrate to evaluate potential interference from those components.
Neither species at an unrealistically high 1M concentration caused a
measurable deviation. Ionic strength* was investigated through the range
0.217M to 1.519M by varying the phosphate buffer concentration. As with
the carbonate and nitrate, no interference could be detected and therefore
these slurry components were deleted from further consideration. Nitrite,
iron II and iron III, on the other hand, caused sufficiently severe perturb-
ation to the analytical accuracy that further study and corrective action
was necessary.
2.5.2.1 Elimination of Nitrite as an Interferent - The nitrite inter-
ference level cited above is considered to be a realistic concentration
that can be encountered from absorption of NO from the flue gas. The NO
»» X
consists of mixtures of nitric oxide and nitrogen dioxide. There appears
to be little affinity for NO absorption in the limestone scrubber, however,
N02 will react quantitatively with water by disproportionate to give
equimolar quantities of nitrite and nitrate. At elevated temperatures
nitrite can undergo further disproportionation to give nitrate and nitric
oxide. The extent of these reactions that occur concurrently during the
sulfur dioxide abatement process will depend on actual temperatures,
concentrations, and residence times. The need for control of the nitrite
interference becomes quite evident from examination of the following data:
N02" (NaN02) % Deviation from Control
Cone., m Absorbance
0.01 4
0.1 12
1.0 93
The compounds sodium azide and sulfamic acid were selected as
candidate additive reagents for elimination of the interference in the
furfural-bisulfite complex absorbance exhibited by nitrite ions. The
reactions of these compounds with nitrite are as follows:
* y = 1/2 E (mi x z^)
-41-
-------
17551-6012-RO-OO
NaN + HN0
2
HN0
NaOH
(13)
2 - 2 24 2 (14)
The azide reagent, however, was found to be ineffective as a supressant.
Actually it caused a greater error in the method than the nitrite ion. Sul-
famic acid, on the other hand, was shown to be effective as the nitrite in-
terferant suppressant from the data in Table XVII.
TABLE XVII
EFFECT OF SULFAMIC ACID ON NITRATE INTERFERENCE
Furfural , yAf
50.24
50.24
50.24
50.24
50.24
50.24
50.24
50.24
50.24
50.24
50.24
50.24
HS03 , mtf
-
1.0
1.0
-
-
-
1.0
1.0
1.0
1.0
1.0
1.0
N02", mM Sulfamic Acid, mM
-
-
0.10
1.2
0.6
0.1
1.2
0.6
0.1
0.10 1.2
0.10 0.6
0.10 0.1
A
0.777
0.281
0.352
0.778
0.775
0.779
0.289
0.278
0.278
0.286
0.293
0.275
No apparent change occurred in the furfural absorption or the furfural-
bisulfite absorption upon addition of sulfamic acid at concentrations from
10"4 to 10 M. From the table it can be seen that the increase in fur-
fural absorbance (HS03~ present) upon addition of 10" M nitrite is -\-25%
and upon addition of sulfamic acid in the concentration range from 10
o
to 10 M that this interference is essentially eliminated. The addition
of sulfamic acid has been incorporated into the procedure as a standard
precautionary measure.
2.5.2.2 Iron as a Potential Interferent - Iron is a significant consti-
tuent of coal burning power plant flue gas fly ash and can be found as a
minor constituent in limestone. As the pH of the scrubber liquor varies
-42-
-------
17551.6012-RO-OO
so will the iron solubility and the potential for interference. Assuming
some degree of solubility, the 276 nM absorbance of solutions containing
phosphate buffer (pH ^4.2), 1.0 mM bisulfite and a furfural concentration
of 0.10 mM> was measured as a function of ferric ion concentration, added
as ferric nitrate. Table XVIII shows the results of the ferric study.
TABLE XVIII
INTERFERENCE OF Fe^3 IN THE FURFURAL-BISULFITE METHOD3
Bisulfite cone. ,.mM
1.000
1.000
1.000
1.000
0
0
Fe+3 conc.,uM Fe(N03)3
-
0.10
1.0
10.0
10.0
0.0
Absorbance
0.464
0.460
0.472
0.490
1.412
1.378b
%Deviation
-
-0.9
+1.7
+5.6
+2.5
-
aPhosphate buffer, pH 4.2 Furfural cone. = 100.0 \M ^ A = 1.378
Deviations in the absorbance of the furfural-bisulfite complex begin to be-
come significant at concentrations greater than 10 MM ferric ion (+1.7% de-
viation at 0.001 mM). Ferric ion was found to interfere with the furfural
absorbance alone, without HS03" present, as well as with the complex. The
significance of this potential interference is discussed below.
The aqueous chemistry of ferric iron affords a plausablt explanation
of the above observed interference. Ferric ion in aqueous solution has a
strong tendency for hydrolysis and/or complex formation. The aquo ion that
is formed will hydrolyze significantly at pH's 1n the range of 2-3, as
indicated by the following equilibrium equations and constants:
[Fe(H20)6]3+ » [Fe(H20)5(OH)]2+ + H+; K = 8.9 x 10"4
[Fe(H20)5(OH)]2+ = [Fe(H20)4(OH)2]+ + H+; K » 5.5 x 10"4
(15)
(16)
At pH's greater than 2-3, condensed species are formed and eventually
colloidal gels appear. For alkaline pH's ferric ion will become insoluble
as the hydrous Fe203 is precipitated.
-43-
-------
17551-6012-RO-OO
The aqueous UV absorption spectrum, which is believed to arise from
the condensed forms of the hydrolyzed aquo complex, shows an intense
charge transfer band in the far ultraviolet which exhibits a tail exten-
ding into the visible region of the electromagnetic spectrum. This tailing
effect is present at 276nM: the analytical wavelength in the furfural
bleaching method and hence, gives rise to increased absorbance values.
The efficacy of several complexing agents for the elimination of
the iron (II and III) interference by measuring the molar absorptivities
of the complexes at 276nM was investigated. The results were as follows:
Component Mixture
Aqueous Fe +
Fe3+ - EDTA
Fe - citrate
Fe3"1" - tartrate
3+
Fe - Phosphate buffer
Aqueous Fe
Fe2+ - EDTA
Fe - citrate
2+
Fe - tartrate
2+
Fe - phosphate buffer
a (276 nM)'*
3,200
5,400
4,200
4,110
4,000
600
7,800
1,200
600
600
*Liter/mol-cm
It is readily apparent from the data that the interference of the
aquo ferric iron complex is significant and that a bathochromic shift
occurs for the charge transfer bands in the presence of the complexing
agents. The increase in molar absorptivity of the complexed species
precludes the use of complexing agents for the elimination of ferric
ion interference.
Ferrous ion exhibits a lesser interference except in the presence
of strong complexing agents such as EDTA or citrate ion. Thus, the addi-
tion of complexing agents to the scrubber solution would compound the
interference if iron is present in the Fe (II) state.
-44-
-------
17551-6012-RO-OO
The impact of the apparent iron problem becomes negligible, however,
if the scrubber is operated in the planned pH range of 6.5 to 7.5. In this
pH range the iron will be present in the condensed form as colloidal suspen-
sions and would be removed by solid liquid phase separation techniques. In
+3 +2
the simulated static slurry described in Section 2.4.2, Fe and Fe were
added at a 25 ppm level (0.45 mM), but after filtration (Whatman No. 41) no
iron could be detected in solution by atomic absorption spectrophotometry.
It appears that efficient removal of colloidal species will be essential to
the elimination of iron interference. In the event the scrubber is operated
at a more acidic pH (3-4.5), it will be necessary to adjust the pH of the
*
analytical liquor sample to neutrality and, after allowing sufficient time
for nucleation, refilter the sample. It may be more expedient to utilize
a cation exchange procedure for this purpose. This contingency is noted
in the standard method, Appendix E.
2.5.3 Time Dependence of the Furfural-Bisulfite Complex Formation
In order to achieve automation of the furfural technique for the analy-
sis of sulfur species, one of the basic criteria for selecting a method for
development, a series of experiments were conducted to determine the param-
eters affecting time to equilibrium for the formation of furfural-bisulfite
complex. The experiments were conducted on a Beckman DK-2A ratio recording
spectrophotometer utilizing the time drive mode of operation and a constant
temperature cell holder.
From examination of the results of the experiments, given 1n Table XIX
it is apparent that the time dependence for complex formation is effected by
temperature, concentration of bisulfite, concentration of furfural and pH.
It is surprising, however, that the strongest dependence is on the pH of the
solution. An attempt to fit the data to pseudo first order (large excess
of HS03") and second order kinetic plots were unsuccessful which is Indica-
tive of complex solution kinetics.
At first glance the kinetics would be expected to follow second order
equations for the reaction of one furfural molecule with a molecule of bi-
sulfite. This, however, does not reflect the strong dependence of the
kinetics on the solution pH. For the system under consideration it would
-45-
-------
17551-6012-RO-OO
TABLE XIX
TIME DEPENDENCE ON THE FURFURAL-BISULFITE EQUILIBRIUM
Temperature HSO,
(°C) (mM)
22 0.1
0.5
1.0
1.5
22 0.1
0.5
1.0
1.5
22 0.1
0.5
1.0
1.5
22 0.1
0.5
1.0
1.5
39 0.1
0.5
1.0
Furfural
(mM)
0.06
0.06
0.06
0.12
0.06
pH Time to Equilibrium
(Min)
3.7 10.5
9.5
7.2
5.8
3.9 8.0
6.5
4.5
3.5
4.1 5.5
5.0
4.0
2.8
3.9 7.0
5.8
4.6
3.6
3.9 6.0
5.5
4.5
appear that the sulfurous acid-bisulfite and/or the bisulfite-sulfite acid
base equilibria also participate in the reaction kinetics. This participa-
tion could yield additional terms in the rate expression resulting in a
strong pH dependence.
The dependence of the reaction rate on pH will be utilized in the auto-
mated procedure to shorten the reaction times and yield minimum instrument
hold times. Independent operating parameters include furfural concentra-
tion, bisulfite concentration, pH and temperature. For the automated pro-
cedure, furfural concentrations will be limited by available detection
techniques and because bisulfate concentration is the variable to be deter-
mined, increased temperature, though not as effective in enhancing rates as
the pH, will be used for enhancement. It is believed that by judicious se-
lection of temperature and constancy in the rate of color development curve
would permit analyses after 5-6 minutes at a nominal pH of 4.0.
-46-
-------
17551-6012-RO-OO
2.5.4 The Effect of Temperature on the Furfura1-B1su1f1te Analysis
This section describes the study undertaken to determine the tempera-
ture dependence of the bisulfite calibration and the extent of control re-
quired to meet the specified accuracy. The studies employed a Beckman
DK-2A spectrophotometer fitted with a constant temperature cell holder. To
investigate the temperature effect, bisulfite calibration curves were run
at temperatures of 21.0°, 35.O6 and 62.2°C. Temperatures were recorded
utilizing a chromel-alumel thermocouple immersed in the spectrophotometric
cell liquid immediately proceeding measurement of each data point.
As previous reported the linear portion of the bisulfite calibration
curve follows Equation 7, and the equilibrium constant K for the complex
format ion may be determined by:
K _ Slope of the calibration curve /*j7\
Intercept of the calibration curve l
Changes in K with temperature are indicative of the temperature sensitivity
of the analytical procedure. Additionally, because K represents an equili-
brium constant for the formation of the complex a logarithmic plot of K
versus 1/T would be expected to be linear and allow evaluation of tempera-
ture effects at any point within the range of temperatures Investigated.
Calibration curves were obtained at three temperatures and are pre-
sented in Table XX. Six point calibrations were obtained for the
experiments at 21.0° and 35.0°C, while only four calibrations were used
at 62.2°C. A least squares evaluation of the data gave the following values
for the equilibrium constants:
Temperature. °C K (liter/mol)
21.0 1,800
35.0 740
62.2 141
A graphical presentation of In K versus 1/T 1s shown in Figure 4. The re-
sultant graph yields a straight line of slope 1.65 x 10"4 and Intercept 2.16
x 10"3 (1/T, Y axis and In K, X axis).
Because the calculated concentration of bisulfite for a given absorb-
ance is inversely proportional to the equilibrium constant, the temperature
-47-
-------
17551-6012-RO-OO
TABLE XX
THERMAL DEPENDENCE OF THE BISULFITE CALIBRATION CURVE
Temperature
21.0
35.0
62.2
HSOq Cone.
_
0.185
0.370
0.555
0.740
0.925
_
0.185
0.370
0.555
0.740
0.925
_
0.186
0.371
0.558
I/A
1.385
1.818
2.222
2.632
3.145
3.584
1.445
1.653
1.852
2.049
2.252
2.449
1.540
1.560
1.600
1.640
effect resulting in a 5% error may be evaluated. Calculations show that
the 5% error band is 1.2°C (or 2.2°F). The bisulfite calibration is sensi-
tive to temperature fluctuations and thermal control is required.
2.6 INSTRUMENTAL ANALYSIS OF SULFATE
As in the case above for accurate, rapid and frequent determination of
S(IV) species, sulfate ion is similarly important for sulfur mass balance
and process optimization as well as being a major participant in calcium
sulfate scale formation. The sulfate ion may be present in filtered slurry
liquor or it may be present in the separated solids. The sulfate concen-
tration range for the liquor was, by definition, set at 1-500 mM (See
Table I). In the search for a suitable instrumental method, it was con-
cluded that the automated laboratory techniques that were available at that
time were not immediately acceptable without a critical review of the state
of the art. Consequently, several candidate sulfate methods were evaluated
including:
-48-
-------
I
4*
1O
2.6 2.8 3.0 3.4
1A (°KX103)
Figure 4. Change in Furfural Bisulfite Complex Equilibrium Constant with Temperature
01
01
__J
I
at
o
INJ
•ya
o
i
o
o
-------
17551-6012-RO-OO
• Titrimetric analysis
• Spectrophotometric analysis
Turbidimetric
Colon'metric
Atomic absorption
Infrared spectroscopy
Atomic emission
t Specific ion electrode
The results of our assessment of the technique, useful range,
interferences, and relative ease of automation are presented below.
This evaluation was used as the basis for selecting methods for limited
feasibility testing in the laboratory.
2.6.1 Theoretical Evaluation of Sulfate Methods
The methods described below are for the most part indirect methods
based on precipitation, usually with barium.
2.6.1.1 Titrimetric Methods - The titrimetric method involves the
precipitation of sulfate with a barium salt with thorin, Sulfanazo, or
other colorimetric or potentiometric end points. To avoid co-precipita-
tion of carbonate and sulfite with barium, the analysis is conducted in
acid solution. The range of the technique is 5-1000 mg sulfate/liter
(0.05 to 10 mM) with a precision of 1-5%. Interference by nitrate, phos-
phate, chloride and cations is concluded to be either negligible or cor-
rectable. Automation of the titration technique can be accomplished in
a batch analysis process using a potentiometric or colorimetric end point,
however, the technique for acidifying and titration will require con-
siderable development.
2.6.1.2 Turbidimetric Method - The turbidimetric method utilizing a
photometer or turbidimeter is generally used in routine analysis of large
volumes of samples where a precision of 5 to 10% is satisfactory. The
sulfate concentration range applicable for this method is 0.1-1.0 mM with
a minimum detection level of around 0.01 mM sulfate. Usually, barium
chloride is used to precipitate the sulfate and interferences can be
remedied. The time period between precipitation and turbidity reading is
critical (up to 4 minutes with a maximum usually reached in 2 minutes).
-50-
-------
17551-6012-RO-OO
In spite of the time lag for precipitation, and large dilution requirement,
the fact that at least one automated Instrument was marketed justified
further consideration of the technique.
2.6.1.3 Barium Chloranilate Colorimetric Method - The indirect colorl-
metric determination of sulfate can be accomplished by using barium chlora-
nilate to precipitate barium sulfate and release and equivalent amount of
colored acid-chloranilate ion. Cations must be removed because of their
interference, pH 1s critical (use pH 4.0), filtering and shaking are
necessary and even at low concentrations the standard solutions may not
obey Beer's Law. The long reaction times and required filtration times
are also disadvantageous. The range of the method is from .02 - 4 mM
sulfate. Calcium, ferric and aluminum ions completely precipitate the
acid chloranilate ion and therefore must be removed by ion echange.
Recent findings under Contract 68-02-0008 (Reference 1) have shown that
chloride and ni.trate concentrations present in the filtered slurry solution
cause low results by more than 2%. Nonetheless, the method has been auto-
mated and deserves further consideration.
2.6.1.4 Benzidine Colorimetric Methods - Sulfate can be precipitated with
benzidlne hydrochloride and the excess can be measured by a variety of
Colorimetric methods:
• from generation of iodine
t reaction with furfural
t diazotlzation followed by coupling with phenol
Then in each case, the color complex is measured at the appropriate wave
length.
2.6.1.5 Infrared Spectrophotometric - Quantitative IR analysis of solids
for sulfate in calcined S02 reacted limestones has been reported by
E. F. Rissmann and R. L. Larkin (Reference 8). The extrapolation of this
technique to an automated method for dissolved sulfate 1on in wet scrubber
mother liquor appeared feasible.
2.6.1.6 Atomic Absorption Method (AA Method) - This technique is also
based on precipitation of sulfate with a barium salt (chloride in this
case), but measures the excess barium by atomic absorption spectroscopy.
The useful range of this method is 0-1 mM sulfate with results comparable
-51-
-------
17551-6012-RO-OO
to volumetric and gravimetric measurements being reported. The AA method
involves less operator handling than the volumetric method, but has a
longer "start-to-finish" time because of long reaction time for equili-
brium precipitation of barium sulfate under the established conditions.
Modifications of the method reported by Dunk, et al (Reference 9) to
reduce the solubility of barium sulfate at the working pH could result in
decreased reaction time. The technique might be acceptable for laboratory
application, however, it would not be a method of choice for on-line
analysis.
2.6.1.7 - Specific Ion Electrode - Quantitative precipitation of sulfate
is accomplished in a 50% dioxane 50% water solution using an excess of
standard lead perchlorate. The excess lead ion may be quantitatively
measured using a lead selective electrode. Activity of lead (measured
by the electrode) must then be related to the concentration of lead by
calculation using a knowledge of the total ionic strength. This tech-
nique has a dynamic range from fractions of a ppm to thousands of ppm
lead. Excess lead is related directly to a sulfate concentration. The
major drawbacks associated with the technique are high and variable ionic
strength, the limited service life of cells, the large potential for
interference and as with most of these methods the excessive elapsed
time for complete precipitation with the resultant constraint on a
continuous mode of operation.
2.6.1.8 - Sulfate By Difference from Total Sulfur - Recognizing the
time limitation inherent in the precipitation methods and the time plus
interference drawbacks of the chloranilate method, an alternative total
sulfur technique was sought. It was believed that a combination of the
furfural spectrophotometric method for S(IV) species coupled with a total
sulfur technique could yield rapid and sufficiently accurate sulfate data
to meet the process development criteria. In reviewing available direct
instrumented methods it was found that a gap exists, *>Q.3 to 10 mM
sulfur wherein sulfur cannot be determined.
As mentioned earlier, X-ray fluorescence provides a lower limit of
approximately 10 mM, whereas a standard flame emission instrument, such
as the Melpar FPD unit has a practical upper limit of approximately 0.3 m
(10 ppm). From theoretical considerations and work performed by other
-52-
-------
17551-6012-RO-OO
experimenters it appears possible to construct an instrument of the
atomic emission (flame photometer) type for the analysis of total sulfur
in a liquid stream that would fill the concentration range void.
In the region between 210 to 820nM, atomic sulfur has only one emis-
sion frequency, a sharp line at 216.9nW, respectively which would result
from the combining of carbon and oxygen. High resolution spectrometers
(^O.lnW) which are available will reduce or eliminate the interference.
A total dissolved sulfur analyzer for the wet scrubber process stream
should meet the following requirements:
• Capable of atomizing (vaporizing) a representative portion of
the process stream on a continuous basis
• A hot zone, with optical windows, capable of disrupting all
molecules
0 Have sufficient spectroscopic resolution to reduce or
eliminate all interferences
• For atomic absorption analysis have a light source that
is capable of producing an emission at 216.9 my
The feasibility of this approach is based upon utilizing the 216.9 my
atomic sulfur line which is from the excitation of non-ionized sulfur
atoms. Recently, it has been demonstrated that sulfur can be measured
quantitatively with both A.A. and A.E. techniques by utilizing microwave
excitation (Reference 10). Spectrometrics Inc., Burlington, Massachusetts
produce a unit that is similar to the type required for this determination,
however, the instrument generates a plasma which would tend to ionize many
or most of the sulfur atoms.
In the work of Syty and Dean (Reference 11) utilizing a fuel rich, air-
hydrogen flame, they found a lower limit of detection of 15 yg/ml sulfur
(^Q.SmM). They report a linear signal function with the square of concen-
tration and a noiseless flame background. This technique is recommended
for future development to augment RF and extend the detection limits to
lower concentrations.
2.6.2 Experimental Screening of Candidate Sulfate Methods
Based on the findings above, a turbidimetric, an infrared and a titri-
metric sulfate method were selected for experimental screening. To facili-
tate testing, the static scrubber slurry simulant described in Section 2.4.2
-53-
-------
17551 -601 2-RO-OO
was utilized as the standard. The sulfate value of 5828 ppm was obtained
gravimetrically, the referee method utilized as the standard for all
comparisons.
2.6.2.1 Turbidimetric Method - Although the problems associated with
turbidimetric sulfate analysis have previously been documented, the lack
of more promising sulfate methodology dictated a brief investigation of
this method. The method utilized is given in Reference 12. A calibration
curve was constructed and the synthetic scrubber solution analyzed with
the following results:
2_
Method SO, Concentration, ppm % Deviation
Gravimetric 5828 (61
Turbidimetric 5460 (57 mM) 6.3
Not all possible interferants are present in the synthetic blend, however,
the results indicate that a turbidimetric analysis may be the only alter-
native to an automated analysis considering the paucity of more promising
methods. Experimentally the lower limit of detection (0.01 mM) is that
previously reported while the upper limit can readily be adjusted through
automatic dilution. The previously stated objections to this analysis are
still valid but may possibly be circumvented and the method optimized to
provide acceptable accuracy. In the single analysis above, the error of
6.3% is not acceptable in accordance with the allowable error of 5%,
stipulated in Table I.
2.6.2.2 Infrared Spectrophotometric Method - The method of Rissman and
Larkin (Reference 8) for the quantitative determination of sulfate in
calcined SOp reacted limestone was investigated for its applicability to
wet scrubber mother liquor. During this brief examination it was not
possible to manufacture the 0.003 mm cells by vacuum deposition of metal
which is a time consuming and costly procedure. Two experiments utilizing
0.015 mm path length cells with Irtran windows were conducted. In the
first experiment, a sample of distilled water was run followed by the
synthetic scrubber solution and then a 960 ppm sulfate reference standard.
The results are as follows:
-54-
-------
17551-6012-RO-OO
Solution Absorbance SO^2" Concentration
Reference 0.043 960 ppm (10
Synthetic Scrubber 0.153 3420 ppm (36
Solution
A second experiment was conducted utilizing matched 0.015 mm cells
blanking out the water absorbance in the reference beam. The results
of the experiments were:
Solution Absorbance SO.2" Concentration
Reference 0.046 960 ppm (10 mw)
Synthetic Scrubber 0.163 3400 ppm (35 mw)a
Solution
Calculated
For these analytical conditions the deviations from the gravimetric values
(5828 ppm) are greater than 40% error and hence the method is not recom-
mended for further consideration.
2.6.2.3 High Frequency Titration Method - A high frequency tltratlon for
sulfate content of scrubber liquors was conducted using a Sargent Model V
Oscillometer with BaCl2 titrant. A sample of standard Na^SC. and the
synthetic scrubber solution were analyzed. The results of the titration
were:
Experimental Known
Solution Results Concentration Deviation
Reference 957 ppm (10. mW) 960 ppm (10 mtf) -1.0%
Synthetic
Scrubber 6140 ppm (64 mW) 5828 ppm (61 mtf) +5.0%
Solution
Synthetic
Scrubber 5557 ppm (58 mW) 5828 ppm (61 mtf) -5.0%
Solution
The second scrubber sample was analyzed 1n 40% methanolic solution as
opposed to aqueous solutions for the previous two runs. This method also
demonstrates an apparent acceptable accuracy, however, the analysis time
is considered excessive requiring on the order of 20 minutes for a single
titration. This 1s because of equ1Hbrtt1on time required between
successive titrant additions. The use of mixed methanol solvent reduces
-55-
-------
17551-6012-RO-OO
the solubility of BaS04 and significantly shortens the equilibration times.
It may also heighten interferences by other precipitating salts because
the barium salts will become generally less soluble in the methanolic
solution. The method has potential but will require development time and
effort.
2.7 INSTRUMENTAL ANALYSIS OF CARBONATE
The required analysis range specified in Table I is 1-20 mM with an
allowable error of 15%. Examination of the carbonate species distribution
curves as a function of pH, presented in Figure 5, reveals that in a wet
scrubber system operated in the range of pH 6.5 to 7.5, the bicarbonate
species will predominate with moderate concentrations of carbonic acid.
Historically three analytical techniques have been utilized for the
analysis of carbonate in aqueous solutions. The reactions involved are
a) the acid-base equilibria between carbonate, bicarbonate and carbonic
acid, b) the extremely low solubility of Ba, Ca and Sr carbonates and fin-
ally, c) the thermal removal of C02 from carbonic acid at low pH's. The
relative merits of each of these with their appropriate end point detection
devices are discussed in the following paragraphs.
2.7.1 Acid-Base Determinations
Titration of carbonates with either acid or base yields two end points
corresponding to the following equilibria:
H2C03 + H20 = HC03" + H30+ K = 4.30 x 10"7 (is)
HC03" + H20 = C032" + H30+ K = 5.61 x 10"11 (19)
Standard methods of end point detection would include conductometric, high
frequency, thermometric, colorimetric (indicator) and potentiometric. All
of these analysis end points would, however, be interfered with by HN02,
H2S03> HS03" and HS04", metal ions and consequently, the methods are impractical
2.7.2 Precipitation Reactions
The precipitation of Ca, Ba or Sr carbonates has been utilized for the
gravimetric and turbidimetric analysis of carbonates. Representative solu-
bilities are:
-56-
-------
100
Ol
en
•«4
I
o»
o
PO
s
figure 5. Carbonate Species as a Function of pH
-------
17551-6012-RO-OO
CaC03 = 0.0015 g/100 ml HgO 9 25°C
- BaC03 = 0.002 g/100 ml H20 @ 18°C
SrC03 = 0.0011 g/100 ml H20 9 18°C
In this case as in the preceding acid-base reactions, strongly interfering
species will be present as evident from the following solubility data for
scrubber sulfur species:
CaS03-2H20 = 0.0043 g/100 ml H20 @ 18°C
BaS03 = 0.02 g/100 ml H20 9 18°C
BaS04 = 0.0002 g/100 ml H20 9 18°C
SrS03 = 0.0033 g/100 ml H20 9 18°C
SrS04 = 0.0113 g/100 ml H20 9 18°C
Again these impurities appear to make these methods impractical for carbon-
ate analysis.
2.7.3 Thermal or Acidimetric Removal of CO,,
The equilibrium between gas phase C02 and aqueous solution has been
utilized extensively for carbonate analysis. The equilibrium may be re-
presented by:
H20 + C02 - H2C03 (2Q)
Many methods of estimating the evolved C02 include acidimetry and gas analy-
sis. Gas analysis may be accomplished by gas chromatography, gravimetry
(ascarite absorption) or manometric methods. For the scrubber solutions,
potential interference from dissolved S02 is encountered.
S02 + H20 = H2S03 (21)
However, the large difference in the first proton dissociation constants
may allow C02 removal to be conducted specifically, while maintaining the
sulfite species as bisulfite. The most promising methods which would elim-
inate the requirement for selective (XL removal however, are those utilizing
specific detection systems such as non-dispersive infrared (NDIR) detection
or, as a lesser candidate, gas chromatography.
-58-
-------
17551-6012-RO-OO
The NDIR approach to simultaneous measurement of thermally liberated
C02 and S02 is applied routinely to metals 1n a Laboratory Equipment Corp.
thermal analyzer. The Instrument output 1s displayed directly as % of C
and % S in 60 seconds elapsed analysis time for both.
Given the ranging capability of NDIR, this detector should be readily
adaptable to the ASTM acidimetric method for carbonate in lime (Method C25)
which in turn can be used for scrubbing slurry.
Solids can be handled directly on the Leco Carbon-Sulfur Combustion
Analyzer after separation and drying. This approach was demonstrated
utilizing TRW's instrument (see Section 3.3 for discussion). Further eval-
uation and development of the thermal/acldimetrlc methods for COg (and SOp)
are certainly warranted on the basis of the foregoing results.
2.8 SURVEY OF NITRITE/NITRATE INSTRUMENTAL ANALYSIS METHODOLOGY
Although methodology for nitrite/nitrate measurements in the wet
scrubber liquor was not among the list of primary constituents under study,
a preliminary review of methods was performed. The following brief listing
of candidate methods and references 1s provided to assist current Investi-
gators and for consideration for future experimental evaluation and develop-
ment. A comprehensive review of methodology for nitrogen-oxygen compounds
is incorporated 1n the recently published monograph on analytical chemistry,
(Reference 13).
2.8.1 Brucine Colorlmetric Method
Several versions of the brudne colorimetrlc method for nitrite/nitrate
determination are found in the literature. For total nitrite plus nitrate
the EPA Federal Water Quality Office Method (Reference 14) is recommended for
evaluation. The procedure 1s applicable for 0.1 to 2 ppm nitrate nitrogen
(0.007 to 0.14 mM concentrations) and extreme care must be taken to control
reaction conditions 1n order to obtain repeatable, accurate results. The
degree of interference from scrubber constituents would have to be deter-
mined. The method reported by Fisher, et al, (Reference 15) allows for dif-
ferentiation between nitrite and nitrate by varying add concentration.
FWQO recommends the brudne method as a manual technique, but it is po-
tentially automatable.
-59-
-------
17551-6012-RO-OO
2.8.2 Dlazotlzation After Reduction
FWQO (Reference 14) recommends two automated nitrite/nitrate methods
based on reduction with cadmium copper catalyst or hydrazine sulfate, reac-
tion of NOp with sulfanilamide and coupling to an azo dye. Precisions for
an interlaboratory evaluation are reported (+5.75, +18.10, +4.47 and
-2.69% Bias). In testing saline water, standards containing 10 salt water
constituents must be utilized.
2.8.3 Ion Specific Electrodes
Instrumental measurement of nitrate ion utilizing nitrate ion specific
electrodes has been reported. In a flowing gas system designed by Dr.
Martini (Reference IS), nitrogen oxides were reacted with ozone and the re-
action products absorbed in a sodium nitrate aqueous solution. Measurement
of nitrate concentration was achieved in a flow through liquid cell equipped
with an Orion specific ion electrode. In a study funded by EPA under con-
tract CPA 22-69-95 Driscoll, et al (Reference 17) found good correlation be-
tween results by the time consuming ASTM POS method and the nitrate ion
specific electrode. Reduction of nitrate to ammonia and use of an ammonium
ion specific electrode has been suggested as an improved approach to elim-
inate certain interferences.
-60-
-------
17551-6012-RO-OO
3.0 TASK II - DEVELOPMENT OF PROCESS INSTRUMENTATION
The primary objective of this program was focused toward this task
and was aimed at the identification of instrumental analytical methods
suitable for on-line analysis of selected chemical species in the EPA Shawnee
Power Plant wet limestone scrubbing process. The development of on-stream
analysis methods will permit the rapid acquisition of data for the effect-
ive characterization of the process and timely elucidation of process
parameter variations. High analytical accuracy (0.1% relative) is not a
requisite of the needed methodology but rather the instrumental techniques
must be reliable, reproducible, cost effective and the equipment easily
maintained, and possess 2-5% relative accuracy.
On review of the scrubbing process variables several sampling require-
ments were identified relating to characterization of the scrubber mixture
These sampling requirements are shown in Table XXI. During a literature
TABLE XXI
LIMESTONE SLURRY SAMPLING REQUIREMENTS
Slurry Solids Content - 0 to 15% w/w
Slurry Sample Quantity - 0.5y particles in liquid
• Lag Time - <30 seconds
Sampling Rate - 30 samples/hr, minimum
Analysis Time - 2 min, max.
Easily Maintained
review phase, sampling, separation and quenching of reactants were identi-
fied as major problem areas that had to be resolved prior to application
of any analytical techniques to on-stream analysis. A system capable of
handling a sampling rate of 30 samples per hour necessitated the use of a
rapid separation of slurry and isolation of solid and liquid phases and
was a key milestone prior to developing analytical techniques. The sampling
rate was established assuming specific combinations of scrubber designs
-61-
-------
17551-6012-RO-OO
and analysis location and sample frequency. An example which fulfills
this requirement is three different scrubber design processes sampled
every thirty minutes at five different locations. Variations of sampling
locations up to eight and sampling frequency of up to fifteen minutes cover
a wide range of samples to be analyzed. For the purpose of establishing
the ability of an instrument to meet the continuous on-stream analysis re-
quirements a total of 30 samples per hour was taken as a nominal value.
The effort to identify and develop suitable sampling and rapid slurry
separation methodology involved vendor contacts, laboratory evaluation of
prototype and standard equipment and testing of candidate equipment at
vendor application laboratories. In order to assess the adequacy of a
laboratory developed technique of analysis, it was necessary to produce a
dynamic flue gas/scrubbing slurry sample. To accomplish this, a small
scale laboratory scrubber test loop incorporating both a flooded bed and a
Venturi type scrubber and a gas blending feed system was designed and con-
structed. Thus samples streams were generated which were simulations of
the pseudo-equilibriated scrubber system, i.e., removal of a sample from
the stream or change in operation would immediately be reflected in sample
instability or dynamic change.
The following sections describe the activities associated with:
• The development of slurry sampling and separation
techniques
• Extrapolation of X-ray methodology to continuous,
on-line applications
• Design and plan for automation of the TRW furfural-
bisulfite spectrophotometric method
t Operation and process variation capability of the
laboratory bench scale wet scrubber process loop
3.1 SLURRY SAMPLING, SEPARATION AND QUENCHING
During the course of this task several vendors were contacted to de-
termine whether they had equipment available which could separate a lime-
stone/dolomite slurry meeting the following operating parameters:
-62-
-------
17551-6012-RO-OO
• Flow rate —to 300 Ib/m1n (a portion of this flow
could be diverted prior to the separator)
§ Solids, % 0.5-15
t Particle size, micron 5-300
• Density of solids (unpulverized) g/ml 2.7-2.9
• Density of liquid, g/ml 1.005-1.080
• Temperature, °F to 150°F
0 System to exclude air during and after separation —
both phases
t Time to effective separation 15 seconds
The directory of 48 separation equipment vendors and manufacturers
compiled for this purpose is included as Appendix F. Of these ten com-
panies replied positively that they had equipment which might fit these
operating parameters. The separation principles identified Included con-
tinuous discharge centrifuges, in-line filter cartridges, belt filters,
and a continuous cyclone cone centrifuge.
Laboratory evaluation of these principles was undertaken using spent
slurry obtained from the Key West Electric Company and equipment sold by
deLaval, Sharpies and Demco. A summary of the findings are shown in
Table XXII. It was found that neither the cone centrifuge nor a combination
of the solid bowl centrifuge-centrifugal cone provided clear-cut separation
as indicated by slight cloudiness 1n the discharge fluids. An optically
clear fluid would demonstrate excellent solids rejection and is needed for
any subsequent colorimetric characterization of the liquid phase. However,
inclusion of a polishing filter, such as an Acroflow in-line convoluted
cartridge filter downstream resulted 1n a high capacity unit providing con-
tinuous transparent liquid for periods as long as several hours depending
on the initial solid loading. A dual, parallel filter system from AFM-Caro
was identified as a strong candidate for the downstream polishing filter
component.
Several cartridge filter types Including a wound cellulose fiber
(Micro-Wynd), convoluted fiber screen (Cuno-Cal) and sintered metal screen
(Micro-screen) with a variety of flow and particle retention properties are
-63-
-------
U551-6012-RO-00
TABLE XXII
SUMMARY OF LABORATORY EVALUATION OF SEPARATION METHODS
Continuous centrifugation - deLaval Laboratory Gyro-tester
Performance - 30 sec. operation at 0.5 gpm feed - 3% Zurn slurry
Results - very nearly clogged
Cone centrifuge (cyclone) - Demco 18mm cone
Performance - continuous - pretreatment device
Results - very promising
Solid bowl centrifuge/cone-Sharpies solid bowl/Demco
Performance - minimum one hour continuous operation
Results - slight turbidity
Polishing filter - Acroflow in-line convoluted cartridge
Performance - high capacity - quick interchange
Results - optically clear output
available. The former two cartridges are low-cost and disposable while the
metal screen is cleanable for reuse. A switching valve permits flow diver-
sion from one filter housing to the second while a quick disconnect bolt
allows rapid filter-cartridge removal and replacement.
3.1.1 Continuous Cyclone Separation/Filtration
The cyclone cone separator was fabricated by Demco as a prototype to
meet TRW's design requirements and is shown schematically in Figure 6. The
device consists of an 18-mm cone fabricated from 316 stainless steel and
possesses an adjustable orifice control. The unit operates with a 35 psi
minimum pressure differential with an inlet feed velocity of 46 ft/second
and a minimum volume demand of 1 gpm. Throttling the underflow to cause an
overflow to underflow ratio of 45, resulted in an overflow to underflow
solids content ratio of 0.0204 with a 3% limestone slurry. Consequently,
operation of the Demco in this mode permitted rejection of approximately
98% of the original solids content. The solution containing 2% of the
original solids was readily handled through a continuous discharge centri-
fuge and polishing filters to provide optically clear liquid. It must be
emphasized that the cyclone was operated at the minimum design pressure
-64-
-------
INLET
OVERFLOW
17551-6012-RO-OO
and flow because of slurry sample supply and pumping constraints within the
laboratory. Much higher solids loading (to the 15% maximum) should be
readily accommodated with equal or better solids separation at more
optimum run conditions. That is, the
efficiency of the cone can be Improved
at the Shawnee plant where up to ^-
10 gal/min of slurry can be diverted
and pumped at higher pressure through
the cyclone. The life time of the
filters are at least one hour and use
of the parallel bank system such as
described above permits back flushing
or cleaning to reactivate a spent
filter when it is isolated from the
flow loop. Low cost cartridges can
•SEAL be removed and discarded.
The sampling and phase separa-
tion system shown schematically in
Figure 7 is the recommended approach
to permit continuous sampling for on-
stream analysis. Two alternative
systems (I and II) downstream of the
cyclone are presented for considera-
tion, however, the simpler and less
' costly of the two (Option I), 1s re-
commended for initial implementation
and checkout.
Utilizing Option I, a clarifica-
tion test was conducted on a 3% slurry
made from the TVA high fly ash con-
tent limestone. The slurry was run
through this laboratory scale clari-
fication apparatus (cyclone, centri-
fuge and filter) for over 60 minutes, producing a clear, water white liquid
on exit from the filter. All aspects of the test appeared normal and
examination of the equipment post test revealed no anomalies. Option II
UNDERFLOW
Figure 6,
DEMCO Centrifugal
Separator
-65-
-------
BANKED
PARALLEL
FILTERS
SWITCHING
VALVE
AP CONTROL
PUMP
PISTON
VALVE
HIGH LOADING CAPACITY-
QUICK DISCONNECT
OPTION II
en
i
o
ro
73
o
i
o
o
CYCLONE
FEED
PUMP CYCLONE
SEPARATOR
ALTERNATIVE
SYSTEMS
CLARIFIED
SCRUBBER -
LIQUOR
ANALYZER REQUIRING
^OPTICALLY CLEAR OR
PART ICULATE FREE
LIQUID SAMPLE
\
GENERALIZED
SCRUBBER
UNIT
SOLIDS ENRICHED
UNDERFLOW
SOLIDS ANALYZER
POLISHING
FILTER
CONTINUOUS
LIQUID DISCHARGE
CENTRIFUGE
(CONTINUOUS
OPERATION > 1 -2 HRS)
OPTION I
Figure 7. Schematic of Continuous Slurry Separation Approaches
-------
17551-6012-RO-OO
could not be fully evaluated for operation time because the dual filter
system was not available, but nonetheless it is expected to be very nearly
as cost effective on an operation basis by optimizing the cyclone operation
and utilizing re-usable filters.
Because of the concern that mass transfer between solid and liquid
phases in a non-equilibrium slurry might cause a problem (when separating
by the Demco cone-polishing filter technique because of filter cake build-
up), a worst case plan was implemented to determine the effect of filter
cake build-up using a batch filtration method.
The experiment permitted comparison of the chemical composition of
filtrates taken from consecutive filtrations through an increasing filter
cake size and fresh filtrations. The experiment consisted of taking 200-ml
aliquots from a highly agitated slurry (3% w/v solids) and filtering through
a millipore filter in a deaerated environment. After a period of time the
slurry was sampled again and an additional 200-ml aliquot was taken and
placed in the filter already containing the filter cake from the previous
filtration. A comparison aliquot was taken at the same time and was fil-
tered through a clean filter. The time of filtration was governed by the
time required to pass the filter containing the consecutive runs; the rate
of filtration through the fresh filter was regulated by the downstream
pressure. This operation was repeated two additional periods to provide a
total of four sequential sets of aliquots. The filtrates were retained in
deaerated environments and analyzed for bisulfite and calcium contents and
for pH measurement. Details of the experimental procedure are presented
below and the results are presented in Table XXIII.
To 4360 ml of boiled, argon sparged deionized water were
added 90 g limestone, 5.17 g NaHS03 and 45 g of fly ash (metals
added, which had been neutralized with HC1* and then washed in
*Because of insufficient TVA fly ash to conduct these experiments, fly ash
obtained from the Nevada Power Plant, Moapa, Nevada was used. The chemical
composition of the Moapa fly ash is different than that of the TVA fly ash
and consequently, metals were added to provide comparable metallic consti-
tuents to that expected from the TVA fly ash. In addition, the fly ash was
considerably more basic, consequently, the fly ash was neutralized prior to
use
-67-
-------
17551-6012-RO-OO
TABLE XXIII
EFFECT OF FILTER CAKE BUILDUP ON SIMULATED
WET SCRUBBER FILTRATE COMPOSITION
Filtering Type
Consecutive Filtration
Fi
Fi
Time, min
Itering Variables F1Uer Cake>
g/in2
PH
Itrate Composition Bisulfite, rnM
Calcium, mM
Fresh Filtration
Fi
Fi
Time, min
Itering Variables Filter Cake,
g/in2
PH
Itrate Composition Bisulfite, mM
Calcium, mM
Slurry
5
15
4
7.2
5.65
2.43
15
4
7.2
5.65
2.30
Mixing
45
21
8
7.2
3.46
1.05
18
4
7.6
3.41
1.12
Period,
75
60
12
8.0
2.99
0.70
60
4
7.8
3.21
0.80
Min.
140
50
16
7
2
0
50
4
8
2
0
.8
.60
.55
.0
.71
.62
deionized water). This yielded a synthetic scrubber solution
with 3% solids and 12 mM bisulfite. This slurry was placed in
a magnetic stirrer in a GN2 purged dry bag. Also present in the
dry bag were two glass millipore filtering sets with 10y Teflon
filters and 1.5-in2 of filter area.
The results show good comparison between the consecutive single batch
filtration filtrate composition indicating little contribution of liquid-
solid mass transfer to the key liquid phase chemical constitutents. It 1s
interesting to note that the bisulfite and calcium ion concentrations de-
crease as a function of elapsed time of slurry mixing which, as will be
discussed later in Section 3.4, is attributed to the kinetics of calcium
sulfite precipitation. Because the elapsed time of separation anticipated
when the Demco cone-polishing filter are used will be considerably less than
-68-
-------
17551-6012-RO-OO
that experienced during these experiments, it is believed the findings here
will be valid until comparable filter cake buildup (approximately 0.5-inch)
is observed. It is recommended however, that similar experiments be con-
ducted using the full scale Demco (1 gal/min) flow on the TVA process units
to confirm the results of our findings prior to using the apparatus in the
field.
3.1.2 Solids Discharging Methods
A continuous staged separation concept was devised which is capable
of achieving "instantaneous quenching of reaction" within an arbitrary al-
lotted time of 15 seconds in such a manner as to present a "dry" stream
of slurry solids for continuous analysis. In one conceptual design shown
in Figure 8, the slurry feed may be taken from the slurry stream and fed
to a solids discharging centrifuge. In this system the process stream at
the point under scrutiny is split and the sample stream enters a liquid/
liquid/solid separator. A second heavy liquid phase such as a Freon, tri-
chlorethylene or other heavy inert solvent would be added to the slurry as
it entered the separator. Because of the toxic nature of many of the can-
didate solvents, appropriate containment of the vapors during handling and
evolution is required via forced air hoods or vents. As shown in the
schematic drawing, the light, clear aqueous phase is separated from an an-
nular zone near the center, the denser non-aqueous phase is ejected from
an intermediate zone while the solids, essentially free from aqueous liquid
contamination are continuously discharged from the outermost zone and trans-
ferred to the quartz filter carrier belt.
Whereas the liquid/liquid/solid type centrifuge with its inert solvent
wash potential capability was not demonstrated, a Sharpies Super-D Canter
P-600 was tested at the Pennwalt Corporation application laboratory. A
20-gallon aliquot of Zurn spent slurry containing 3.86% solids was pro-
cessed at ambient temperature, various feed rates and two bowl speeds. The
following experimental data were obtained and are presented graphically in
Figure 9.
-69-
-------
en
en
SLURRY
ANALYZER
"^
1-
SCRU
ALTERNATIVE 1 LIQl
i
ALTERNATIVE 11 FREON -]
CLARIFIED
AQUEOUS
SOLUTION - i
FREON - — J
RECOVERY If
SOLIDS DISCHARGING
CENTRIFUGE
c~
'LJ
>BER
JOR
,
>
1
1
|
OVERf
t
1
1
O
(o
) V
x^ \
o
ro
73
1
o
o
-LO'/v
CYCLONE
SEPARATOR
fuNDERFLOVv
•h GAS EX IT
( -*-
HOT GN2 ..,. ||
•^ DETECTOR 1
jOLIDj CARR ER <-n, mc —
AKIAIYZPB /~AC iULIL):)
(2) ANALYZER
^^^ 1^^ (3)
FILTER AND , , ||! ill 1
BLOCK f \ PYROIYSFR OVEN ' f V^
/ L ±/ ^-^
f 1 \ / / f \
(°) QUARTZ BELT / / ( ° ]
c«>s£v ALUMINO-SILICATE GLASS CONSTRICTIONS \^S
FEED SPOOL COLLECTING
SPOOL
Figure 8. Schematic Design of Continuous Solid Separation and Analysis Apparatus Concept
-------
17551-6012-RO-OO
200 400 600 800 1000
PPM INSOL SOLIDS IN CLARIFIED LIQUID
Figure 9. Flow Rate vs Clarity for Sharpies Super-D Canter
-71-
-------
17551-6012-RO-OO
Test No.
Clarified Liquid
rate, gph
insol . solids, ppm
Mechanical Conditions
bowl speed, rpm
conv. diff., rpm
oond
1
50
500
5000
50
3
2
16
290
3
28
410
4 5
63 126
520 870
6 7
25 63
320 530
finnn
DUUU
fin
3
8 9 10
32 95 126
360 600 680
The feed slurry contained 3.86 % w/w (38,600 ppm) insoluble solids. The
composite solids sample collected during the test program contained 20.0 %
w/w moisture.
Thus it can be seen that the P-600 provides a clarified liquid that can
be fed to a polishing filter and solids discharge with ^20% moisture amen-
able to the solid analysis system in Figure 9.
Alternatively, the centrifuge unit (which is a major cost item) may be
omitted and the underflow from the DEMCO cyclone used as the feed to the
moving belt filter. In addition to concentrating the solids in the under-
flow, classification of size distribution also occurs. For the 3% w/w
solids loaded Zurn spent slurry, the size distribution in a microscopic
examination of under and overflow gave the following:
Overflow Underflow
Size, y Cum % No. Size, y Cum % No.
<10 87 <10 53
<20 99 <20 88
<30 100 <30 96
<40 98
<50 99.5
<60 100
Tnis obvious classification is a variable dependent on cyclone effi-
ciency parameters. Figure 10 shows photomicrographs of the particles iso-
lated from the overflow and underflow. Qualitative XRF analysis of the
solids from the separation indicated no significant classification by
species or elements. With a 98% w/w concentration of solids in the under-
flow, some differentiation by compounds will be tolerable.
-72-
-------
OJ
I
OPERATING SPECIFICATIONS:
INLET FEED, FPS 46
VOLUME DEMAND, GPM 1
PRESSURE, PSID 35
SEPARATION CAPABILITY 97% OF
PARTICLES LARGER THAN 4.3 MICRONS
WITH SPECIFIC GRAVITY OF 2.3
(300
3 TO 10 MICRONS
TEST RESULTS:
PARTICLE SIZE <10^
FRACTION OF SOLIDS
FRACTION OF VOLUME FLOW
XW CHEMICAL ANALYSIS
(SIGNIFICANT DIFFERENCES)
UNDERFLOW
53%
0.524
0.022
Ti, Fe, Br, Cl, K, S, Sr
OVERFLOW
87%
0.476
0.978
Ti ABSENT, Fe, Br, Cl,
K INCREASED S, Sr SAME
i
-------
17551-6012-RO-OO
Relative and approximati ons for the three slurry separation subsys-
tems exclusive of the moving belt filter for solids transport but inclusive
of the dual polishing filters and pumps described above are as follows:
DEMCO Cyclone ^$600
Solid Bowl (Liquid Discharging Only) ^$3,000
Liquid/Liquid/Solid Discharging ^$15,000
Super-D-Canter ^$10,000
In the filter/drying housing which is the second stage residual inert
solvent or remaining moisture is volatilized in heated high pressure dry
nitrogen stream before the solids pass into the solid analyzer sections,
3.2 CONTINUOUS ON-LINE X-RAY FLUORESCENCE (XRF) METHODOLOGY
As described in Section 2.4.1 the XRF technique was judged the best
candidate instrumental technique for laboratory and continuous, on-line
analysis for elements and cations in the composite slurry, the separated
solids and for dissolved species in the liquor (sensitivity permitting).
Limiting the evaluation and development effort to those candidate instru-
ments that could conceivably provide the on-line technology, three units
are evaluated and ranked in Appendix D, i.e., 1) ARL, 2) G.E. and 3) Kevex.
Timely experimental evaluation utilizing the standard specimens delineated
in Table VIII was possible only at the ARL and Kevex laboratories. Details
of the experimental activity related to the ARL unit that were pertinent
to the conclusions presented in Appendix D are discussed below while the
Kevex data is highlighted in the appendix.
A trip was made to Applied Research Laboratories (Sunland, California)
for the purpose of evaluating their process control X-ray quantometer
(PCXQ) for on-line analysis of selected chemical species in the Shawnee
plant limestone wet scrubbing process. The 14 specimens listed in
Table VIII were run on ARL's laboratory X-ray Quantometer 72000. Although
the Model 72000 accommodates only dry specimens, ARL personnel assured us
that it is the same basic X-ray fluorescent spectrometer instrument as the
Model PCXQ minus slurry presenter modifications.
Several PCXQ units with slurry presenter modifications in beginning
stages of fabrication and assembly were examined. Up to 15 slurry streams
-74-
-------
17551-6012-RO-OO
can be analyzed sequentially in an automated mode for elements from magne-
sium upwards in the periodic table with the PCXQ. In addition, an override
can be placed on the automated mode and batch solid samples may be analyzed
by shutting off the appropriate slurry feed streams. Nine spectroscopic
channels of information are available and nine elements in each of the 15
slurry streams can be simultaneously detected and analyzed. A pulp density
monitor is incorporated into the system along with a fixed external standard.
Flow rates through the slurry sample cell can be 5, 15 or 50 liters/minute.
The data generated on the ARL-PCXQ are in the form of voltage ratios
and are denoted
V'ES
where: I = voltage output from sample
IES = voltage output from a fixed external ARL standard
The signal data are presented in Table XXIV. An analysis of these data
and curves generated from them lead to the observations and conclusions
presented below.
3.2.1 Limit of Detection
The limit of detection for sulfur is 0.03% absolute* in CaCOg and lime-
stone-base specimens on the ARL-72000 unit. An on-line slurry unit PCXQ
utilizing a helium X-ray path and a Kapton cell window will have a poor
limit of detection. The limit of detection is significantly less than the
0.25% absolute value which has been discussed as the lowest reasonable value
which is likely to be encountered for the slurry but higher than the low-
est 1 mM for the liquid phase (0.03% is equivalent to 9.4 mM}. For the
other elements of concern, e.g., Ca, Mg, Fe, Cl, etc., the limit of de-
tection is much lower than the 0.1% tentative requirement (Appendix D).
3.2.2 Repeatability
The repeatability of the 72000 unit was demonstrated to be more than
adequate for projected use. The precision of the slurry model should be
close to the same because the precision is largely controlled by both the
X-ray spectrometer design and signal processing technique both of which are
Conditions: vacuum, 0.25-mil thick polyethylene window, 2-minute count
-75-
-------
17551-6012-RO-OO
TABLE XXIV
SIGNAL DATA FROM ARL QUANTOMETER 72000
SAMPLE
001
CaC03 (reagent grade)
Rerun 2 days later
002
CaC03 +7.9% CaS04
(1.85% Sulfur)
Rerun 2 days later
003
CaC03 + 14.6% CaS04
(3.44% Sulfur)
Rerun 2 days later
004
TVA Limestone
(0.05% Sulfur)
005
Limestone + 7.9% CaSO.
(1.87% Sulfur) *
005 (Repeat)
Limestone + 7.9% CaSO,
(1.87% Sulfur)
005
With 0.25 mil Kapton
004 (Repeat)
004
With 0.25 mil Kapton
006
Limestone + 14.6% CaSO,
(3.44% Sulfur) H
Mg
0.220
0.210
0.220
0.212
0.206
0.198
2.881
2.655
2.630
0.306
2.879
0.285
2.330
Al
0.010
0.010
0.018
0.019
0.007
0.008
0.167
0.140
0.140
0.036
0.168
0.041
0.134
S
0.054
0.054
1.294
1.332
2.697
2.666
0.074
2.031
2.052
1.000
0.076
0.059
4.544
Fe
0.344
0.360
0.416
0.422
0.360
0.390
1.276
1.166
1.169
1.182
1.279
1.294
1.127
Ca(a)
10.448
0.524
9.653
7.873
9.068
7.412
9.933
9.352
9.354
8.512
9.928
9.037
8.799
^'Sensitivity changes were made in calcium determinations so the numbers
obtained two days apart for calcium will be different.
-76-
-------
17551-6012-RO-OO
TABLE XXIV (CONTINUED)
SAMPLE
007
CaCO. + 7.9% CaSO, +1% Fe
(1.81% Sulfur) H
008
Limestone-TVA FA
009
Limestone-Zurn FA
010
1.5% TVA FA in 98.5%
Epoxy
on
10% TVA FA in 90% Epoxy
012
1.5% Zurn FA in 98.5%
Epoxy
013
10% Zurn FA in 90% Epoxy
014
CaC03 + 7.2% CaS04 +
7.2% Na2S03 (-40 Mesh)
1(3.43% Sulfur)
New Run
New Sample
Na2S03 (rea9ent grade)
Na2S04 (reagent grade)
NaCl (reagent grade)
NaCl with 0.25 mil Kapton
ARL (Reagent Grade)
CaC03
003
With 0.25 mil Kapton
Mg
0.211
4.805
2.026
0.307
0.786
0.299
0.344
0.192
0.194
0.197
0.088
0.100
0.249
0.125
0.272
0.266
0.147
Al
0.008
6.275
0.082
0.313
1.976
0.060
0.062
0.008
0.008
0.008
0.005
0.005
0.915
0.200
0.033
0.032
0.007
S
1.408
1.023
0.149
0.096
0.298
0.068
0.072
1.310
1.385
1.439
6.649
5.832
0.077
0.054
0.105
0.092
1.351
mm^mmm
Fe
6.082
4.000
1.877
8.710
4.000
2.160
2.180
0.361
0.426
0.380
0.585
0.700
0.493
0.528
0.490
0.505
0.424
mmmmmmi^mfm
Ca
9.592
2.576
9.850
0.133
0.866
0.362
2.376
9.612
7.786
7.662
0.031
0.031
0.030 1
0.029 1
8.382 1
8.3981
6.670 1
-77-
-------
17551-6012-RO-OO
the same in the PCXQ and 72000 systems. The repeatability of a single
measurement was checked by determining the sulfur content in sample 002
eleven times. The sample contains 1.85% S and the mean voltage ratio was
1.397 ^0.0054 where the uncertainty is the standard deviation. The per-
cent standard deviation is 0.4%, far less than the 3% which has been viewed
as a requirement.
System repeatability after a period of time, is good as shown from an
experiment in which the same samples were analyzed two days later. The
findings presented in Table XXV shows that the two sets of data always
TABLE XXV
ARL-72000 VACUUM QUANTOMETER REPEATABILITY DATA
Specimen and
Element
Sample 001
Mg
S
Fe
Sample 002
Mg
S
Fe
Sample 003
Mg
S
Fe
Voltage Ratios
First Data
Set
0.220
0.054
0.344
0.212
1.294
0.416
0.206
2.697
0.360
Second Data Set
2 Days Later
0.210
0.054
0.360
0.220
1.332
0.422
0.198
2.666
0.390
•— —«•__!
Rel ati ve
Reliability
5
0
3
4
3
2
4
3
8
•M«^
agree to within 10% of their nominal value and more frequently agree to
within 4%. Again, the conclusion is that a 10% fluctuation around a nominal
1% sulfur value will be readily detected.
3.2.3 Matrix Effects
Figures 11 and 12 demonstrate that both elemental sulfur and calcium
working curves are influenced by the matrix material in which the sulfur
-78-
-------
10
I
GO
LU
VES
Figure 11. Working Curves for Sulfur Analysis
en
CJl
CT>
o
ro
i
§
§
-------
17551-6012-RO-OO
10
in
t—t A
-^ 9
o
A
0
•F—
4->
ra
(Y
M
0
>
8
7
r—
H
'
'
-
• H
-
:
—
-
--
_
-
—
—
-i
'
— l
>-i
...
H
-
-
\
:.
-
' '
—
-
—
—
^
i
-
:
-
_
.-
•
r™
~
—
7
'-
_
-
:;
...
r
•
-
-
•
-
h-
'I
-
—
-
'
;
_
~
-
-
-
-
_.
'
-f
~
-.
...
1
—
•
-
-
x
-
-
-
—
-
--•-
,4-
•f-
2
r
-j -
k -i-
...p
r
1
1
-•
4"
xx
T
•
t
-
•
S
-
-
t
•
-
-
x
-
>
1
.-
_^
f-
-
x
»
-
-
-
.1
V
S
1
-
-
3
-
:
v
p
&
~
-
-
i
-
,
'e
i
-
N
/ r
';
-
""
-
—
•-
-
<>
>v
^
?
.
1
-
-
-
--
X1
-
$
-
x
-
S
K
--
-
-
-
*
-
(c
V
X
-
-
-
._
#
X1
:
-
f
^
<•
'
-
-
-
X1
-
X
\
-
-
••
X
-
-
{
^
—
-
-••
-
-
— '
*
--
-
t
Y
-
:
-
...
:
—
^
-
-
r1
.
-
-
_.
X
#
_
^
x
-
~
c
x
2
^
?
-
-
--
*
x
-
:
/
-
-
-
?
-
-
-
*
-
-
-
~
Ti
x
-
-
x
(I
ff
—
-
-
(
x
-
-
_
/[
1
-
-
-
^
>
-
•-
-
'
I
-.
-
-
£
-
-
-
-
-
-
a
?
x^
-
-
r-
-
-
-
...
b.
x-
•
^
—
-
-
X
"•
-
-
-
-
-
-
...
-
X
-
-
—
-
::
-
-i
x
-
...
_
-
-
-
-
-
X
-
-
-
-
-
.
-
-
-
-
-
-
-
c
x
:
-
-
-
-
>
-
-
-
-
j
»
—
-
-
-
-
-•
-
-
"
/
-
-
-
-
-
-
-
i
-
-
-
-
s
—
—
^
:.
-
-
1
-
—
__
-T
i_
Ti
-
- -
_ -
~f
. . .
;
:.;-j
I
Weight % Calcium
Figure 12. Calcium Working Curves
-80-
-------
17551-6012-RO-OO
is contained. In Figure 11, two curves were generated for specimens with
the same sulfur contents; the only difference in the specimens is that one
set contained limestone (natural CaC03) as a diluent for CaSO^ and one set
contained reagent grade CaCO~. Particle sizes were about the same. It is
clear from the figures that the output ratio from the Quantometer is in-
fluenced by the matrix, and that separate working curves are needed for
CaCO, and limestone matrices. 'The explanation for this effect may rest in
0
the presence of trace elements in limestone and it may very well be possi-
ble to take this effect into account by monitoring the magnesium content
on a separate spectrometer.
3.2.4 Quantitative Interpretation
Using the limestone + CaSO, working curves in Figures 11 and 12, the
following information was determined for the limestone/TVA flyash and lime-
stone/Zurn flyash specimens:
t For quantitative analysis it is essential to prepare refer-
ence specimens having all the same ingredients including
limestone and flyash. In the studies reported here poor
agreement with wet analysis was obtained using working
curves which did not include flyash. However, in previous
work conducted at TRW, excellent agreement was obtained on
the TVA specimens when flyash was added to the matrix.
0 The TVA specimen appears to be considerably different than
the Zurn flyash specimen with regards to other elements
also. The TVA flyash specimen contains approximately
2-1/2 times the quantity of Mg, 75 times the quantity of
aluminum, twice as much Fe, and 1/4 as much Ca as the Zurn
flyash specimen. Specimen 007 contained 0.9 % w/w Fe to
check interference between iron and sulfur in a CaC03-base
sample. The curve in Figure 11 (points #2 and #7) indicates
that the 0.9 % w/w iron had virtually no effect on the 1.8 %
w/w sulfur determination.
3.2.5 Particle Size Effects
The output signal is influenced by particle size in the specimen.
This effect was dramatically illustrated by samples 006 and 014. Both
specimens contain 3.44 % w/w sulfur except 014 contains about half the
sulfur as the sulfite ion. At first, it was found that.the signal from
-81-
-------
17551-6012-RO-OO
014 was about half of what was expected and it was thought that a dis-
tinction might be made between the S04~ and S03~ ions. Auxiliary deter-
minations on reagent grade Na2$04 and Na2$03 proved that such a distinc-
tion was not possible, and instead it was observed that a large difference
existed between the particle sizes of CaS04 and Na2S03> An additional
sample was prepared in which all the constituents were passed through a
400 mesh screen (37 micron opening). The data for this sample showed that
as long as the particle size remains small (<37 micron), spurious results
due to particle size will be eliminated. The particle size effect should
be common to all X-ray fluorescence units and is not unique to one vendor.
No problem is anticipated in actual wet scrubber operation where over 80%
of all particles are smaller than 30 microns (This observation was made from
classification of spent limestone slurry solids from Zurn, KPL and Shawnee
wet limestone facilities).
3.2.6 Dilution Effects
Dilution effects (or, stated another way, the effects of solids load-
ing) are handled in a straightforward manner by use of a concentration-
loading-signal map. Figure 13 contains experimental data in a hypothetical
map. The main point is that the signal strength is proportional to the
solid loading fraction for the TVA flyash. Secondly, the broken lines in-
dicate the manner in which such a map might be used in a real operation.
A signal strength value for sulfur is first determined. A percent solids
loading is next determined via a pulp density gauge which is built into
the ARL system. The point where the two lines intersect gives the weight
percent sulfur in the solid portion of the slurry sample.
3.2.7 Analysis of Liquid Samples
Throughout the evaluation session it was realized that the ARL-72000
was a highly optimized laboratory unit and that a lower level of perform-
ance would be expected in the PCXQ slurry system. As a result, the follow-
ing series of experiments was conducted to determine, in an appropriate
manner, the behavior which might be expected on a PCXQ system.
The experiments consisted simply of covering the dry samples with a
0.25-mil sheet of Kapton*. The voltage ratios were comoared with those ob-
tained with Kapton absent. The data are presented in Table XXVI. For
*Kapton is the slurry cell window material in the PCXQ system.
-82-
-------
17551-6012-RO-OO
100
0.1
SULFUR SIGNAL IS/IES
Figure 13. Conceptual Concentration-Loading-Signal Map for Sulfur
every element except iron**, the X-ray intensity (voltage ratio) decreased
when Kapton was introduced. The reduction was most severe for magnesium
and this was expected because the magnesium radiation has the longest wave-
length (hence, softest X-radiation). A comparison for sulfur was made be-
tween sample 004 (Limestone) and 005 (Limestone-7.9^ w/w CaSO.) under the
**The slight increase in the iron signal may be due to the 30-100 ppm iron
impurity which is frequently found in all grades of Kapton except the
electronic grade.
-83-
-------
17551-6012-RO-OO
TABLE XXVI
EFFECT OF KAPTON WINDOW ON X-RAY SPECTROMETER
PERFORMANCE: 0.25-MIL SHEET
Sample
002 Mg
AT
S
Fe
Ca
003 Mg
Al
S
Fe
Ca
004 "Mg
(Lime-. Al
stone)
Fe
Ca
005 Mg
Al
S
Fe
Ca
Voltage Ratio Values
No Kapton
0.212
0.019
1.332
0.422
7.873
0.198
0.008
2.666
0.390
7.412
2.879
0.168
0.076
1.279
9.928
2.630
0.140
2.052
1.169
9.354
0.25-Mil Kapton Sheet
0.148
0.010
0.753
0.463
7.098
0.147
0.007
1.351
0.424
6.670
0.285
0.041
0.059
1.294
9.037
0.306
0.036
1.000
1.182
8.512
two situations and the results were extrapolated to the case where a 1-mil
Kapton window would be used. The results demonstrated that although the
Kapton does degrade the performance of the unit, the detection limit with
a 1-mil window is still below the 0.25% level (e.g., 0.20 % w/w).
-84-
-------
17551-6012-RO-OO
3.2.8 Advantages of the ARL System
ARL claims at least two patented features of their Quantometer, an
external standard to optimize system stability and a slurry density gauge.
The external standard consists of a brass or titanium disc positioned to
intercept a portion of the primary X-ray beam adjacent to the sample. The
resulting non-dispersed X-ray signal is detected by a sealed detector.
The detector output is integrated across a capacitor simultaneously with
all other detector signals being measured, and is subsequently divided into
the intensity values obtained in the elemental capacitors, hence, the IS/EES
outputs. This ratio technique is automated by terminating the overall in-
tegration period when the external standard capacitor reaches a predeter-
mined fixed voltage (e.g., 4 volts). The slurry density gauge utilizes
backscattered X-rays from the white spectrum to monitor changes in solids
concentration. A curve of voltage from the scattered radiation channel vs
wet specific gravity (measured as slurry density) is produced. The curve
is good only for a slurry of a specific type. A major matrix change from
lime concentrate to clay slime, for example, would require the use of dif-
ferent pulp density correction curves. Slurry density (or pulp density)
correction curves must be generated for each element of interest.
The ARL PCXQ slurry unit is the recommended system for wet-scrubber
on-line process monitoring of Ca, Mg, total sulfur and the other elements
of concern. Its strong point is its proven ability to present and analyze
slurry specimens. The weakness of X-ray fluorescence in general seems to
be its sensitivity to specimen matrix, particle size, and slurry density.
Corrections for these effects could be readily made with the aid of digital
computer programs which are available from ARL. A discussion of multiple
utilization of this computer for data acquisition and reduction of the out-
put of other analytical instruments at the test site is presented in
Section 4.
Solids Analyzer 2 and Solids Analyzer 3 shown in the design concept
for continuous solids separation and analysis were originally considered
as one XRF and one XRD unit. No commercially available on-line XRD unit
has been found to date however, the concept is feasible with a dry powder
presenter. As presently conceived, Analyses 1, 2 and 3 are three feed
-85-
-------
17551-6012-RO-OO
streams to the same XRF instrument. The advantage of this multi-stream
approach is that total sulfur and elements are determined in the composite
slurry; after filtration (with or without solvent washing) and mild drying,
major differences between dissolved and solid constituents may be estimated;
and, finally, after pyrolysis for thermal removal of C02 from "available"
carbonate and S02 from test labile sulfur species, the concentrations of
Ca, Mg and other low level constituents are maximized in the solids to im-
prove detectability and eliminate dilution and matrix effects. Automatic
dry powder presenters for the last XRF channel are available from several
sources. G.E. presents a slightly compressed, smoothed sample contained
in a moving cup (XEG System) while ARL offers a briquetting unit for pel-
letizing the samples at 40,000 psi.
3.3 CONTINUOUS ON-STREAM CARBONATE ANALYSIS
It should be noted that in addition to demonstrating the utility of
X-ray fluorescence in the continuous solids analysis system, Figure 10,
the feasibility of the continuous thermal liberation of C02 from still
"active" slurry solids with quantitative C02 measurement was amply demon-
strated. Filtered, dried solids from the Demco cyclone efficiency tests
(Section 3.1) (utilizing the spent Zurn slurry) were obtained from the
overflow (to polishing filter for analyses requiring optical clarity) and
underflow (to solids analyzer system). It will be recalled that although
size classification occurs in the cyclone, the XRF analysis indicated no
discernible compositional differences in the overflow and underflow solids.
The carbonate values shown below tend to corroborate that conclusion:
Carbonate Content, % w/w
Demco Overflow 57.0
58.5
Average 57.8
Demco Underflow 57.8
57.0
Average 57.4
These analyses, exclusive of separation and drying required approxi-
mately five minutes per test. The Leco laboratory unit provides simultan-
eous C02 and S02 elapsed analysis time of 60 seconds. Automation of Leco
-86-
-------
17551-6012-RO-OO
sample injection operations would provide an interesting alternative to
XRF determinations which should take less time.
It is believed that continuous measurement or certainly consecutive
batch sampling and determination of the available carbonate in the slurry
scrubbing process will provide key process control information on the
capacity of the spent stream and recycle stream that will permit maximum
utilization of the slurry and thereby significantly improve process cost
effectiveness.
3.4 DEVELOPMENT PLAN FOR WET SCRUBBER BISULFITE ANALYZER (WSBA) PROTOTYPE
FABRICATION AND EVALUATION
With the completion of the laboratory development phase of the UV fur-
fural bleaching method for sulfite, the next phase, i.e., to provide a pre-
liminary design for automation of the procedure, was implemented. In order
to support most effectively the 1972 start-up date for the EPA Limestone-Wet
Scrubber Test facility at Paducah, Kentucky, the plan for provision of a
sulfite method has been first to provide a detailed instrumental method,
which has been accomplished (see Appendix E), and second, to design a batch
automated type analyzer for laboratory utilization. The objective is to
provide the most economic procedure in terms of 1) speed and simplicity of
performance, 2) maximum versatility, and 3) ease of implementation.
A flow diagram design of the batch automated analyzer to meet these
objectives is presented in Figure 14. This design is based on the follow-
ing considerations to achieve complete automation with attainable required
analytical accuracy:
• Sampling system: a dual sample system that has a 9-minute
time cycle for each sample position is required. The dual
system is to have a 4.5-minute overlap between samples
yielding an initial throughput capacity of approximately
100 samples/shift. This can easily be doubled or tripled
at a later time.
• Sample dilution ratios; the optimum ratios for a 2 dilution
system are approximately 17:1 for the reagent add stage of
the low concentration sample, and 170:1 total dilution for
the high concentration sample.
• Bleach development time: a minimum development time ap-
pears to be around 5 or 6 minutes.
-87-
-------
,IXING COIL;
MIXING COIL
HEAT EXCHANGER
LAUDA MODEL KS481
DELAY COIL
WASTE
BLUE
BLUE
O
O
MI/MIN
0.80 NITROGEN
1.60 DISTILLED WATER
0.16 SAMPLE
0.16 SAMPLE
2.90 REAGENT
1.20 NITROGEN
2.00 WASTE
PROPORTIONING PUMP TECHNICON MODEL !
CIRCULATING SYSTEM
CONSTANT TEMPERATURE BATH-PRECISION
UV MONITOR MODEL 1280
LABORATORY DATA CONTROL
SARGENT RECORDER MODEL SR
at
CD
ro
i
§
o
O
CHEMINERT
CHROMATRONIX
3 WAY VALVE
VALVE, SKINNER
4-WAY SOLENOID
TIMER, CRAMER
NITROGEN GAS
Figure 14. WSBA Flow Diagram
-------
17551-6012-RO-OO
• Temperature: 5% error band is induced by a 1.2°C change in
temperature.
• Analyzer cell; the cells must be temperature controlled
and be capable of having prior samples pumped out.
• Spectrometer: a dual beam grating or prism instrument with
a reproducibility in transmission units of approximately
0.2% at 276 m is required (a line source at 276 «M could
also be used).
Recommended future studies will concentrate initially on factors to
evaluate the selection of the parastaltic pump. These studies will deter-
mine the dilution and reagent addition steps, and flexible tube stability.
The first experiments will investigate the dead space and mixing effects of
the total system under test. Subsequently, a study of temperature effects
will be conducted with the pump and reagents at 22°C and 40°C using various
dilutions with primary emphasis on the 17:1 dilution ratio.
In addition, it is recommended to determine the extent of detector
error through repetitive calibration runs (e.g., five samples, five times
for five different days). Premixed standards will be used for all tests.
In addition, other experimental parameters will be assessed individually,
such as flow - stop flow in the UV cell, temperature controller, dilution
ratio, mix chamber, bubble rate, furfural adding, and sample adding effects,
When the extent of individual factors have been determined, the key output-
dependent characteristics will be varied in the total system to determine
whether the continued effects are additive. The magnitude of these error
effects will be quantitized through statistical treatment of the data.
3.5 BENCH SCALE WET SCRUBBING PROCESS SIMULATOR
A modular bench scale test loop wet scrubber was designed and fabri-
cated to permit evaluation of the recommended methods under simulated use
conditions. A loop system was selected because of the necessity of: 1)
closely approximating the full scale operating unit, 2) accurate control,
and 3) producing stable (equilibrium) and unstable (non-equilibrium) con-
ditions for evaluating candidate instrumental methodology under known,
controllable conditions with realistic compositions. A schematic diagram
is presented in Figure 15 while a photograph of the fully assembled system
is shown in Figure 16.
-89-
-------
17551-6012-RO-OO
GASEOUS NITROGEN
MAKEUP CONNECTION
VdGAS RELIEF VALVE ,
GAS BLOCK VALVE 1
1 t>d
r~| f| r"|ROTOMETERS
| | | | |_J 0 - .5 CFM
SO2 CO2 \°2 {
((
^ C~T~^K"'°'
10 CFM BLOWER,
W/ DRIVER
t.»
1/20 HP ELECTRIC STIRRER
-------
GAS CIRCULATING BLOWER
FLOW METER FOR
PACKED BED
PACKED BED SCRUBBER
1
VENTURI SCRUBS 1 0 A'1ALVZER > GAS CHROMATOGRAPH
MULTI-GAS BLENDER
THERM'O-METER
DEMCO CONE SLURRY FLOW
& FILTER METER-TOTAL
CONDUCTIVITY
A;IALYZER
SLURRY CIRCULATING
o
o
Figure 16. Photograph of Instrumented Bench Scale Process Scrubber Simulator
-------
17551-6012-RO-OO
The system consists of a bench scale Venturi scrubber with a second
stage packed bed, fitted with a recirculating gas stream. The pressure
drop associated with the packed bed is about 0.5-inch of water, the pressure
drop due to the Venturi is about 1 inch of water and the pressure drop as-
sociated with the ducting is 0.4-inch of water. The packed bed is 9-1nch
deep and has a diameter of 4-inch. The ducting is 2-inch I.D. throughout.
The Venturi has a throat size of 1 inch.
The recirculating gas stream is moved via blower K-101. The composi-
tion of the recirculating gas stream is controlled by the Minor Gas
Addition facility. This facility allows the addition of small amounts of
gases via rotometers and bottled gas. Gases such as SOg. C02 and 02 are
controlled in this manner. Nitrogen is bled into the system to make up
that amount which has been absorbed by the circulating slurry and maintain
a positive inert gas blanket. The level of slurry in the liquid separator
V-103 is controlled in this manner. The composition of the gas stream is
monitored by gas analyzer AR-104, which gives compositions of S02> C02 and
Op in the circulating stream.
The liquid slurry exits the Venturi scrubber via the liquid separator
V-103. The temperature in the downcomer is measured and recorded by TR-105.
Analyses of the slurry can be provided in the downcomer by tapping a port
for analyzer AR-105. The liquid stream from the liquid separator dumps
into a 15-gallon delay tank (T101), where it is agitated with a 1/20 horse-
power electric laboratory stirrer (M-101), and the temperature is adjusted
and controlled by a tank heater (E-101). The temperature is measured and
recorded by TR-101. The residence time in this delay tank is about one
hour with a design slurry flow of 0.2 gpm.
The liquid slurry travels to the process feed tank along one of two
routes. It can travel along the straight transfer section, or it can be
diverted through a filter. The purpose of the filter is to take out solids
from the circulating slurry. The composition of the slurry exiting the
delay tank is monitored and recorded by AR-101.
The solids content of the slurry in Process Feed Tank (T-102), a 15-
gallon, 304 stainless steel tank equipped with a 1/2 horsepower laboratory
-92-
-------
17551-6012-RO-OO
stirrer, is adjusted by adding limestone from the solids hopper. The com-
position of the tank is monitored and recorded by the process analyzer
AR-102. The temperature in this tank is maintained by tank heater E-102
and is measured and recorded by TR-102.
The adjusted slurry from T-102 is transported along the transfer line
with a positive displacement pump, P-101. This pump has the capacity of
0.1 to 0.3 gpm. This range is required so that the liquid to gas ratio
present in the packed bed Venturi scrubber is capable of being changed.
Accurate flow of the pump output can be adjusted via the recycle stream to
T-102. The temperature in this section of line is recorded on TR-103. The
flow then splits, part going through the counter-current flow section of
the packed bed. The other part of the flow goes to the thread of the
Venturi. The flow which goes to the packed bed section is measured on flow
indicator FI-104, while the flow to the Venturi scrubber is determined by
difference utilizing the flow monitored on flow indicator FI-103.
All liquid lines present in the bench loop simulator are of 1/4-inch
polypropylene, with an 0.028-inch wall. Utilizing this type of tubing,
the flow velocity is about 1.8-feet per second.
This unit has been used to test the applicability of the recommended
methods under controlled conditions, as described below. For this series
of experiments the process feed tank (T-102) was by-passed from the delay
tank (T-101) directly to the variable speed slurry pump (P-101). Contin-
uous pH measurement of scrubber effluent from the Venturi downcomer was
accomplished in the 15-gallon delay tank (T-101) by means of a combina-
tion electrode connected to a Corning Model 12 research pH meter. Readout
was monitored continuously on a strip chart recorder. This may be con-
sidered in-line analyzer AR-105. For a brief description of available
process pH monitoring equipment see Section 3.6 and Appendix G.
Initial check-out and operation with gaseous S02 feed and limestone
slurry (3.6% w/w solids) revealed three mechanical problems that were recti-
fied prior to proceeding to demonstrating proposed method applicability
and varying processing parameters. The first difficulty was associated
with the 10 cfm gas circulation blower (K-101) which utilized air intake
-93-
-------
17551-6012-RO-OO
to cool the motor. It was necessary to seal the blower and provide exter-
nal cooling. In the slurry circulation check-out run, the phenolic im-
peller of the Flowtek, Inc., circulation pump P-101 split in half, abort-
ing the run after one hour of operation. After replacement and continuous
operation of the replacement impeller, it was examined, showing considera-
ble abrasive wear. Because no rubber or stainless steel replacement im-
pellers were available, this pump was replaced in favor of a Jabsco pump
with rubber impeller. The last modification to the scrubber was required
in the packed bed packing material which initially was Raschig rings. This
bed packing tended to collect and accumulate slurry solids to the point of
flow stoppage. This problem was remedied by replacing the Raschig rings
with large rings (3/4-inch O.D. x 1-inch long) made from PVC pipe.
The following paragraphs describe the experiments that were performed
to determine 1) the inherent oxidation effect of the scrubber system in the
absence of oxygen and flyash, and 2) demonstrate the utility of the recom-
mended procedures. Additionally, several scrubber experiments were per-
formed through in-house support to examine the effect of temperature, fly-
ash and oxygen.
3.5.1 Bench Scale Scrubber Tendency for Slurry Oxidation
To establish the system baseline tendency for oxidation of sulfite to
sulfate, prior to flue gas, flyash, limestone or oxygen addition, the bench
scrubber was charged with 10 liters of deionized, deaerated water prepared
by nitrogen sparging and 6.244 g NaHS03 reagent (^6 mM). A zero time
sample (blank) was taken from the catch tank prior to slurry circulation.
Samples (100 ml) were taken every 15 minutes for the two-hour duration of
the experiment. A positive pressure or blanket of inert N2 gas was main-
tained throughout the test. The pH of the catch tank was determined by
monitor AR-105. Table XXVII presents the results of the analyses for bisul-
fite ion as well as the pH of the catch tank solution at the time each
sample was taken. The data correlates very well between pH and the apparent
sulfate formed from oxidation. Elapsed time for the sulfite analysis by
the laboratory instrumental method (Appendix E) was approximately five
minutes for the time of sampling whereas the furfural/buffer addition was
performed immediately upon sampling. More important than gaining experi-
ence and confidence in the methods was establishing the baseline behavior
-94-
-------
17551-6012-RO-OO
TABLE XXVII
BISULFITE ANALYSIS FOR BISULFITE OXIDATION
BENCH SCRUBBER EXPERIMENT 104
•^^•M
No.
0
1
2
3
4
5
6
7
8
Time, Min.
0
15
30
45
60
75
90
105
120
HS03" Cone., mM
5.85
5.50
5.10
4.69
4.32
4.14
3.93
3.54
3.29
S(IV) Removed
-
6
13
20
26
29
33
40
44
PH
4.95
4.00
3.85
3.75
3.70
3.60
3.60
3.50
3.40
of the scrubber system to help elucidate the results of the forthcoming
experiments. At the present time it is only possible to point out several
possible causes for the observed oxidation phenomenon. They may include
surface oxide coating, residual dissolved oxygen and last, perhaps most
significant phenomenon, the spontaneous decomposition of bisulfite reported
by Chertkov (Reference 18) as follows:
4HS03" •*• S3062" + S042" + 2H20 (22)
The formed trithionate then hydrolyzes to yield more sulfate and thiosul-
fate which then reacts with an intermediate HS03 radical to yield more tri-
thionate:
S3°62" + H2° * S2°32" + S042" * 2H+ (23)
HS03 + OH + HS-HS03 * (HSO^-S + HgO (24)
The possibility that added sulfur (as SO^) was being lost to the sys-
tem walls in a passivation process rather than the more probable oxidation
or disproportionate reactions described above was investigated further.
A second sodium bisulfite run was conducted with an approximate charge of
-95-
-------
17551-6012-RO-OO
6 mM HSO^". Both bisulfite and total sulfur as sulfate* were determined
every 15 minutes over a two-hour period of recirculation. Oxygen was added
after the first hour of operation. After the first hour the sulfite de-
creased from 6 mM to ^3 mM (^50% decrease) and at the conclusion of the
second hour the sulfite level was 1.5 mM (^75% decrease). Total sulfur in
solution remained essentially unchanged throughout.
3.5.2 Evaluation of Recommended Methods for Characterization of the
Limestone Scrubber Process
The two experiments described in this section, utilizing the bench
scale wet scrubber loop with fresh TVA limestone and gaseous S02, were di-
rected specifically toward demonstration of the utility of the developed
WSBA method and applicability of atomic absorption spectroscopy and pH
monitoring.
3.5.2.1 Standard Operating Procedure for Bench Scale Scrubber - In each of
the experiments described in this section as well as those in the following
Section 3.5.3, the procedure for charging the scrubber, feeding the simu-
lated flue gas and operating the process were very nearly identical. They
were as follows:
Three pounds of screened (less than 30 mesh) TVA feed
limestone was added to 10 gallons of nitrogen-sparged, de-
ionized water in the process run tank and mixed. The feed
pump was then started and the flow adjusted to run condi-
tions, 0.1 gpm through the Venturi scrubber and 0.1 9Pm
through the packed bed scrubber. The blower was then
started and sulfur dioxide was introduced at a flow rate
of 0.015 cfm. The sulfur dioxide addition was continued
for a period of 45 minutes and then terminated. The cir-
culation of the gas system at approximately 12 cfm was con-
tinued for 75 minutes more, with the addition of 0.0177 CFM
nitrogen. One hundred milliliter samples of the slurry mix-
ture were taken from position AR 101 at the start of the ex-
periment (before S02 flow) and every 15 minutes thereafter.
The samples were immediately filtered through two Whatman 41
filter pads while under a nitrogen blanket maintained by a
flow of dry GN2- The clear filtrates were collected and one
aliquot analyzed immediately for bisulfite ion [total S(IV)]
*Total dissolved sulfur values were obtained by oxidation of the aliquot
with 3% hydrogen peroxide, precipitation with barium chloride and gravi-
metric determination of the barium sulfate.
-96-
-------
17551-6012-RO-OO
using the furfural bleaching method. Atomic absorption analy-
sis for cations was performed on a separate aliquot of the
filtered sample. The pH was continuously monitored using
AR 105, previously described.
3.5.2.2 Results of Characterization of Scrubber Slurries - The two scrubber
experiments were designated Runs No. 105 and 106 and were identical with
the exception that No. 106 incorporated an oxygen addition at the 180 min
elapsed run time and involved more comprehensive characterization for com-
parison to process parameter variation runs detailed in Section 3.5.3. The
data obtained for these runs are shown in Table XXVIII.
TABLE XXVIII
BISULFITE ANALYSIS FOR THE LIMESTONE SLURRY
BENCH SCALE SCRUBBER EXPERIMENTS
Time
No. Min.
0 0
1 15
2 30
3 45
4 6C
5 75
6 90
7 105
8 120
9 180
pH
1105 1106
9.40 9.50
7.30 7.45
6.85 6.90
6.65 6.60
6.65 6.60
6.65 6.65
6.70 6.70
6.70 6.70
6.80 6.75
6.95
Oxygen flow started
10 315
11 390
7.50
7.70
Bisulfite
(mM)
1105 #106
0 0
1.25 1.19
2.67 2.19
3.31 2.81
1.53 2.22
1.74 1.68
1.45 1.51
1.39 1.39
1.10 1.24
0.69
after Sample 9
0.03
0.10
Total Sulfur
in Solids
(9) 106
-
-
...
2.6
-
-
-
-
-
-
taken
8.0
8.9
Total Sulfur
in Filtrate
(mg), 106
-
.
-
12.8
-
-
-
.
_
6.6
7.2
Fe in
Filtrate
ppm, 106
.
_
<0.1
<0.1
<0.1
-
-
_
_
<0.1
<0.1
-
Ca in
Filtrate
mM, 106
5.16
4.81
5.30
4.71
The results available for comparison between Runs 105 and 106, i.e.,
through a run time of 120 min for bisulfite and pH, show very good agree-
ment. The significance of this finding is that the run parameters are
stable and controllable with the TRW bench scale design and, in addition,
that 1) the WSBA laboratory instrumental method is an improvement over
currently available methodology, and 2) automation will provide further
economic advantage.
-97-
-------
17551-6012-RO-OO
It should be noted that the continual loss of sulfite through this
point in the scrubber operation remained perplexing. The process para-
metric study discussed in the next section provided invaluable information
on at least one important phenomenon contributed to this loss and that may
be an important factor in the TVA operation.
3.5.3 Study of Wet Scrubber Process Variables Using the TRW Bench Scale
System
As described above, the bench scale limestone scrubber unit was suc-
cessfully used to demonstrate the utility of the recommended analytical
procedures for rapid characterization of filtered limestone slurry mixture
composition. The additional studies described below were conducted at TRW
through partial in-house support, and although these studies are beyond the
scope of the current program, they are reported because they 1) are relevant
to the chemistry of limestone scrubbing processes, and 2) clearly identify
future areas of technical endeavor required for detailed elucidation of the
significant processes needed for systematic process design improvements.
3.5.3.1 Experimental - The bench scale limestone scrubbing process simu-
lator was used exclusively in these experiments. Figure 16 shows the unit
with associated heat-traced lines and analysis instrumentation located
nearby to facilitate rapid analyses. In these experiments, sampling was
accomplished by periodic removal of 100-milliliter samples of the slurry
mixture from the sampling point AR 101, rather than using the recommended
continuous Demco/polishing filter approach (because of the high liquid
flow requirements of the latter unit).
In these experiments, operating temperature, fly ash and oxygen were
variables evaluated for their influence on the limestone scrubber chemistry
In these studies the total solids loading was maintained at 3.1 to 3.6% w/v
When limestone/fly ash mixtures were employed the limestone content was
2.5% w/v and the fly ash content was 0.6% w/v. Because insufficient TVA f|v
ash was on hand for the entire series of experiments, it was decided to use
the available fly ash from Nevada Power Plant, Moapa, Nevada. However, its
chemical composition was considerably different than that previously obtain
from the Shawnee Power Plant, Paducah, Kentucky. Consequently, it was modi-
fied by addition of oxides of iron, cobalt and nickel to bring its metal
-98-
-------
17551-6012-RO-OO
composition to values directly comparable to that of the Shawnee Power Plant
fly ash. In addition, the Moapa fly ash was considerably more basic than
that from the Shawnee Power Plant, hence, the charge was neutralized (in a
slurry form) to pH 7 with hydrochloric acid prior to system addition. In
this way, the pH of the system during sulfur dioxide addition was maintained
between 6.3 and 7.8, which is representative of operational units of the
Kansas Power and Light Company and the Key West Electric Company.
The general operating procedure consisted of charging the reservoir
with 12 gallons (45.5 liters) of deionized water, adjusting the temperature
to the desired level and deaerating the water by sparging it with argon gas
and stirring it for two hours. Limestone and neutralized fly ash slurry
were added as desired for the experiment. Gas phase preparations consisted
of heating the system to the desired temperature and purging it with nitro-
gen gas maintained at a flow rate of 0.018 cfm. At the beginning of the run,
slurry pumping through both the flooded bed and Venturi scrubber was initi-
ated and then sulfur dioxide was introduced into the system at a flow rate
of 0.015 cfm (at 22°C). After 45 minutes, the sulfur dioxide feed was
terminated and the nitrogen gas flow was allowed to continue throughout the
remainder of the experiment to preclude any oxygen pick up from the surround-
ing environment. Subsequent mass spectroscopic analysis of the reagent
grade sulfur dioxide gas revealed an assay of only 86% v/v with the remainder
consisting of air. Therefore, the 45-minute addition of sulfur dioxide re-
sulted in introduction of a total analytical concentration equivalent to
15.0 mM S(IV) present in the 45.5 liters of slurry. The corresponding oxygen
content introduced during sulfur dioxide addition was 0.48 mM Og. These
values are true as well for Runs No. 105 and No. 106 above, but the quan-
tity of oxygen added with the S02 does not account for the bisulfite loss.
In Run 112, when oxygen was added deliberately to the system, the nitrogen
feed was reduced to 0.014 cfm and oxygen was added at a rate of 0.004 cfm
(resulting in a total oxygen content equivalent to 5.46 mM, if soluble in
the 45.5 liters of slurry). On termination of the sulfur dioxide and 02
feed after 45 min in Run 112, the nitrogen feed rate was increased to
0.018 cfm.
-99-
-------
17551-6012-RO-OO
As mentioned previously, 100-ml aliquots of the slurry were taken
from the delay tank prior to sulfur dioxide introduction and every 15 min-
utes after experiment initiation. The samples were filtered immediately
through a Whatman 42 filter paper under a nitrogen-purged plastic dry bag.
The clear filtrates were collected in 100-ml volumetric flasks and then
analyzed within two minutes of sampling for bisulfite ion [total S(IV)]
using the furfural bleaching method. The pH was monitored continuously at
AR 105 using the Corning Model 12 pH meter and a combination electrode
coupled to a strip chart recorder. Selected filtrate samples were analyzed
for total sulfur content (peroxide oxidation followed by barium ion pre-
cipitation) and total calcium ion content (atomic absorption).
3.5.3.2 Results - The findings of these experiments were somewhat surpris-
ing. Figure 17 shows typical smoothed curves for liquor bisulfite ion con-
centration as a function of time. At room temperature (Run 106), it is
seen that approximately 20% of the available sulfite is retained in the
liquid phase and that after sulfur dioxide addition, there was a monotonic
decrease in bisulfite ion as a function of time. Initially, it was believed
this decrease originated principally from air oxidation of the bisulfite;
however, inspection of the data reported in Table XXIX for analysis of the
sulfate [S(IV)] content of these solutions shows that the sulfate content
did not change appreciably during the experiment. This finding suggests
that the slow, continual reduction in bisulfite originates from precipita-
tion of calcium CaSO-* 1/2^0 (which was observed by X-ray diffraction to be
the only sulfur-containing species in the solid phase).
Another interesting finding was that the sulfate sulfur increased sig-
nificantly (2-3 fold) and associated sulfite content decreased when fly ash
was added. This observation suggests enhanced oxidation of sulfite through
catalysis by trace fly ash metal ingredients. Because the extent of oxida-
tion is considerably more than that capable through direct combination with
the available dissolved oxygen (0.48 mw), it appears that some reduction of
the metal oxides or salts may also take place or that the bisulfite dis-
proportionation is significantly enhanced through catalysis.
-100-
-------
I
o
I
BISULFITE
CONCENTRATION,
mM
72eF NO FLY ASH
. 125°F NO FLY ASH
72°F FLY ASH ADDED
THEORETICAL 100%, HSO3 - CURVE
20
40
60
80 100
TIME, MIN
Figure 17.
Effect of Fly Ash and Temperature on Soluble S(IV)
in Bench Scale Wet Scrubber
en
en
73
O
O
O
-------
o
ro
t
TABLE XXIX
LIMESTONE SLURRY LIQUOR COMPOSITION AS A FUNCTION OF OPERATING CONDITIONS AND TIME
Run
106
no
113
m
112C
Operating Conditions3
Temperature
°F
72
72
125
125
125
Limestone/
Fly Ash
% w/v
3.6/0
2.5/0.6
3.6/0
2.5/0.6
2.5/0.6
Time of
Sampling
Min.
15
45
180
15
45
120
15
45
105
15
45
120
15
45
105
Filtered Liquor Composition
PH
7.45
6.60
6.95
6.70
6.60
6.60
6.70
6.35
6.70
7.65
7.15
7.45
6.90
6.75
6.85
S(IV),
mM
1.19
2.81
0.69
1.49
1.22
0.12
1.15
1.15
0.50
0.86
0.74
0.09
0.96
1.01
0.66
S(VI),
mM
b
1.19
1.38
3.06
2.95
3.25
1.16
1.18
1.06
2.07
2.19
2.53
2.60
2.42
2.34
Total
Calcium
mM
b
5.16
4.71
8.73
9.35
7.98
3.24
4.11
2.87
7.73
7.73
7.48
8.23
8.98
8.73
Continuous closed loop scrubbing of influent S02 gas; S02 feed terminated after 45 minutes of addition
[total charge equivalent to 15 mM S(IV)]; system continued recycling for remainder of experiment with
N2 gas feed to ensure no oxygen pick-up.
Not determined.
C0xygen added together with SO- at 0,,/SO,, molar ratio of 0.267 and terminated concurrently
after 45-minute addition. c *
on
en
cr>
O
i
•yo
O
O
o
-------
17551-6012-RO-OO
Comparison of Run 113 with 106 and Run 110 with 111 permits evaluation
of effect of temperature. The observation that increased temperature re-
duces the sulfite content is explained through enhanced CaSOo'l/ZHLO preci-
pitation kinetics. It is believed that the lower sulfate content at ele-
vated temperature was caused by lower solubility of oxygen, which, of
course, controls the kinetics of oxidation.
In the experiment when a significant quantity of oxygen was added
(Run 112) its affect appears to be of second order magnitude. A significant
higher quantity of sulfate was observed compared to its corresponding Run 111
however, the sulfate content was observed to decrease as a function of time
suggesting either 1) calcium sulfate coprecipitation within the calcium sul-
fite hemihydrate, or 2) a poorly mixed sample.
Another expected finding was that the fly ash introduced a considerably
higher soluble calcium into the slurry than the limestone. This stems from
the fact the fly ash calcium form is the more soluble calcium oxide (hydrox-
ide) and in our case that was partially titrated with hydrochloric acid.
It should be noted that the products of the total calcium ion and the
total bisulfite ion concentrations readily exceed by two orders of magni-
tude the solubility product of calcium sulfite hemihydrate (8.4 x 10~8)
whereas the products of the total calcium and total sulfate ion concentra-
tions are of the same magnitude as the solubility product of the calcium
sulfate dihydrate (2.4 x 10" ) inferring that under the conditions of the
experiment calcium sulfite hemihydrate precipitates and calcium sulfate
dihydrate does not. It is recognized that this simple calculation did not
take into account considerations of activities of the ions and the distri-
bution of the analytical concentrations in other species, such as CaSO-°,
CaHC03+, CaS04°, and CaC03°.
An attempt was undertaken to provide a detailed compositional analysis
using the data of the Radian Corporation (Reference 2 ), however, on exten-
sive evaluation of that work it became obvious that the many assumptions
used in generating the thermodynamic disassociation constants and varia-
tions as a function of temperature, together with total disregard of
known thionate chemistry equilibria in this media, would make the results
-103-
-------
17551-6012-RO-OO
of such calculations open to considerable question. In addition, the con-
centration of some of the input materials in our experiment (e.g., acid
needed to neutralize the fly ash) were known only semi quantitatively, hence,
calculation of the ionic strength employed was not well-established. Con-
sequently, the theoretical analysis was abandoned at this time, however, it
is clear that additional effort is warranted in this area, particularly
for generation of fundamental or "effective" constants under use conditions
of elevated temperature, solids loading, and non-equilibrium conditions.
3.5.3.3 Recommendations - As is seen from the previous discussion, labora-
tory bench scale experiments can be very useful in elucidating the chemistry
involved in the wet limestone scrubbing sulfur dioxide abatement process.
The bench scale studies permit studying a wide range of well-controlled
operating conditions in a relatively short time and at considerable lower
cost than full scale process unit. Consequently, it is recommended that
bench scale studies be conducted to augment and complement full scale tests
by identifying the more significant parameters for detailed testing thereby
reducing the need and, hence, the cost of extensive full scale evaluation.
The significant finding observed in the tests reported above, namely,
the rate dependent step in holding tanks having controlled environments is
that of calcium sulfite precipitation, should not be under emphasized. The
rate of precipitation could control the effectiveness of sulfur dioxide ab-
sorption in full scale scrubbers operated at pH 6-7. It is clear that near t
term bench scale studies of this type are warranted in which detailed mater-
ial balances are made, together with use of operating conditions which closely
represent full scale operational ranges. Findings from these studies will
permit 1) early fixing of operational ranges, 2) identifying conditions
leading to problem areas such as scaling, erosion, etc.), and 3) improved
scrubber design and operational modes.
3.6 PROCESS MONITORING FOR pH
On stream pH monitoring devices applicable for wet limestone scrubber
systems have been reviewed. Of those available on the market, three can-
didate systems, manufactured by Beckman Instruments, Universal Interloc,
Inc., and Leeds and Northrup, have been selected for recommendation.
Appendix G lists the specifications, available materials of construction,
-104-
-------
17551-6012-RO-OO
special features and costs. Both Beckman and Universal Interloc produce
second generation amplification systems for pH readout; the main improve-
ment results from the differential amplification of the glass and reference
signals eliminating ground loop interferences. Mechanical strength is a
very important factor in the final selection of a system. The availability
of electrodes constructed of non-scaling or non-fouling material should be
considered. In scaling studies performed at TRW, Teflon was found to be
superior to other polymers and metals and should be used whenever possible.
-105-
-------
17551-6012-RO-OO
4.0 TASK III - DATA ACQUISITION AND PROCESSING
The purpose of this task was to identify data collection and reduction
methodology that will be suitable for computer input. Data format or in-
strument signal must be capable of interfacing with the eventual engineer-
ing computational system to be operational at the TVA wet scrubber plant
site.
Depending upon 1) the mode of operation of the analytical instrumenta-
tion, i.e., laboratory or on-line, 2) the allowable elapsed time from
sampling-to analysis-to data hard copy, i.e., real-time data output versus
delayed batch data handling, and 3) the funds available and appropriate
cost trade-offs that must be made from a position of total project overview,
several degrees of sophistication of data acquisition systems are possible.
Three alternative systems are described below. The computer data acquisi-
tion system associated with the recommended XRF on-line process instrument
is the primary candidate system because of its proven application for
similar analyses and availability to meet the demonstration start schedule
at TVA.
4.1 COMPUTER DATA SYSTEM FOR XRF ON-LINE PROCESS INSTRUMENTATION
The recommended on-line continuous slurry analyzer for calcium, mag-
nesium, total sulfur and other selected elements, as described in Section
3.2, is the ARL PCXQ 4400 XRF unit. Likewise ARL's laboratory XRF instru-
ments were considered excellent candidates for laboratory analysis. These
systems are equipped with a Hewlett-Packard 2114C computer as a standard
option. Discussions were held with ARL technical sales personnel to deter-
mine the basic computer configuration, its utilization by the XRF unit and
its potential for monitoring other laboratory or on-stream analysis in-
strumentation.
Table XXX describes the computer system recommended for the X-ray
analysis function. Note particularly that the computer processor is avail-
able approximately 80% of the time the instrument is operating and is, of
course, available totally when X-ray functions are not being performed.
The use of this computer for non-X-ray monitoring is limited primarily
by the amount of computer memory (core) available, interfacing devices, and
-107-
-------
17551-6012-RO-OO
TABLE XXX
RECOMMENDED COMPUTER SYSTEM FOR XRF DATA ACQUISITION AND REDUCTION
Item
Hardware
Configuration
Utilization by Quantometer
Computer
Hewlett-Packard 2114C
central processor
4K memory
16 1/0 channels
A/D Converter
Sampling Program
Quantitative Analysis
Sample Selection
Curve Fitting
SOFTWARE
supplied by ARL
(core resident)
supplied by ARL
(core resident)
supplied by ARL
(core resident)
supplied by ARL on
tape. Must be read
over other programs
to use , then others
reloaded.
<20% of CPU time
50% to 100% depending on
functions and instrument
configuration
2 or 3 channels (13 unused)
Utilized only once M sec)
per sample to convert
capacitor voltages (inte-
grated counts) to digital
values.
Collects and stores
capacitor voltages (1 sec
computer time/sample)
Uses sample and standard
values to compute weight
percents of the various
elements (5 sec computer
time/sample).
Determine next sample for
analysis (3-5 sec. computer
time/sample)
Calculate new calibration
curves by least squares
the availability of software to perform the required additional functions.
If these are real limitations in the existing unit, they may be overcome
by the purchase of additional equipment and programming services through
ARL or Hewlett-Packard. Implementation of a system which performs non-
X-ray functions, in addition to the X-ray analysis is a straightforward
computer program development task.
-108-
-------
17551-6012-RO-OO
4.2 ALTERNATIVE COMPUTER DATA PROCESSOR FOR GENERAL ANALYTICAL
INSTRUMENTATION
General laboratory instrumentation data acquisition and processing
systems may be designed around such highly sophisticated and costly com-
puters as the IBM 1800 or Sigma 3. Assuming that such a system is not
available, however, at Paducah, it would be much more cost effective to
establish such a capability around a suitable minicomputer. As part of
this task, the feasibility of the latter approach was investigated.
Recently, Digital Equipment Corporation (DEC), has introduced a general
laboratory data processing system based on one of their minicomputers and
compatible with various high data rate laboratory instruments. It is be-
lieved that this system, costing approximately $50K, unquestionably meets
the needs of the laboratory and possesses the potential for process control
as well.
The DEC POP 12/LDP is designed as a simple-to-operate tool for a wide
variety of real-time data-handling and research applications. Performance
characteristics of the POP 12/LDP have been optimized around a complete
hardware/software system containing two processor modes identical to those
on the widely used POP 8/L and PDP-8/I. The Analytical Instrument Package
includes all hardware for meaningful instrument/processor A/D interfacing.
The Floating Point Processor allows high speed real time processing of data
from high data rate instruments. The basic unit can accept data from four
instruments simultaneously (expandable to sixteen instrument inputs at
$1K/4 channels).
The POP 12/LDP is a fully bundled system capable of sharing software
programs (at no cost) from all POP 12, POP 8, LINC, and LINC-8 users through
DEC's users group library (currently over 500 programs). This very important
feature means that the system while big enough to handle the real-time pro-
cessing requirements of high data rate instruments, such as mass spectrometry
(MS) and fast Fourier transform infrared (IR) or nuclear magnetic resonance
(NMR) spectroscopy, can also be used for process instruments and process
control loops. Simplified programming languages are also available for
special purpose programming by other than professional programmers. In ad-
dition, DEC has just opened an analytical instruments applications laboratory
-109-
-------
17551-6012-RO-OO
where they will continue to expand the available software for chemical
applications.
A specific recommendation between the two choices of minicomputer
systems available, i.e., DEC and HP, cannot be made at this time. Such a
decision, as indicated previously, must be made based on overall project
cognizance. However, if such a decision is pending, it is suggested that
competitive bids be solicited from both vendors.
4.3 NON-COMPUTER DATA ACQUISITION
Non-computer data acquisition methodology is the last of the alterna-
tives to be considered and, as the simplest, requires very little discus-
sion. The principle recommended instrumental methods of analysis are X-ray
fluorescence, atomic absorption, UV and NDIR spectrophotometry. Each in-
strumental output may be readily and inexpensively obtained in a variety
of forms including digital, hard copy form directly in ppm or percent of
species. For example, the ARL XRF unit is available with chart readout,
typewriter, tape punch or tape printer as standard options. Multi-lamp AA
units and spectrophotometers may be equipped with digital (BCD) output and
linked through an appropriate logic board to the ARL data system for punch
taping and/or printing. In this way data acquisition for quick look and
manual computation is combined with computer interface capability.
-110-
-------
17551-6012-RO-OO
5.0 CONCLUSIONS AND RECOMMENDATIONS
During the performance of this program methodology for laboratory and
on-stream measurement of calcium, magnesium, sulfite, sulfate and H con-
centration were reviewed, evaluated and developed. In a subordinated effort,
methodology for nitrite/nitrate and carbonate were examined to permit an
evaluation of direction and scoping of future work. For the instrumental
techniques recommended, the alternative data acquisition and reduction sys-
tems to permit the widest range of choice in automation versus cost have
been identified. The conclusions and recommendations regarding methodology
provided in the following paragraphs are presumptive of the given concentra-
tion ranges and accuracy requirements as delineated by Bechtel and Radian
Corporations.
For on-line slurry analysis for total calcium, magnesium, sulfur and
other elements that may be of importance to the process operation and as
iron, manganese, cobalt, chloride, etc., X-ray fluorescence (XRF) has been
judged the best process analyzer. Extensive evaluation of available tech-
nology has led to the conclusion that, as of this writing, Applied Research
Laboratories (ARL) offers the best process XRF system. Because sample pre-
treatment and phase separation is not required with the ARL slurry presenter,
continuous, real-time readout of up to 15 separate elements can be obtained.
A scanning channel is recommended for detection of trace or other composi-
tional variations that may impact process operation.
The choice of a laboratory XRF unit is not as clear-cut as several
vendors offer competitive equipment. Nonetheless, the ARL laboratory units
are definitely among the best available in terms of sensitivity, versatility,
and automatability. In addition, both the process and laboratory units
come equipped with several optional, but most importantly, field proven
data acquisition and reduction systems ranging from digital recorder output
of signal to printed hardcopy of percent composition distribution utilizing
sophisticated computer methodology for correction of matrix effects and
automated calibration for substrate variations.
Because all other analytical techniques require rapid phase separation
at the point of process sampling, on-line separation techniques were de-
veloped in the laboratory and are recommended for evaluation and adoption at
-111-
-------
17551-6012-RO-OO
TVA. The system recommended is comprised of a cyclone centrifugal separator,
such as the Demco unit tested in this study. This unit significantly de-
creases the solid content of the analytical effluent but does not yield an
optically clear stream. Therefore, a second stage separator is necessary.
With the given information is is not possible to determine if a polishing
filter will suffice or if an intermediate centrifuge stage will be required.
For initial field testing, the two-stage cyclone + filter separator system
based on the design in Figure 7 is recommended. A high capacity, parallel
dual cartridge filter unit such as the AFM-Cuno unit warrants testing for
this application.
Whereas XRF can provide elemental analysis of the heterogeneous slurry
as well as separated liquid and solid, the very low solubility of some
species and trace concentrations of others of concern may require a more
sensitive liquid phase analyzer. Atomic absorption (AA) spectrophotometry
has been shown to meet this requirement in the experimentation reported in
this document. At the present time, however, there are no on-line or batch
flow automated instruments that can be considered process units. On the
other hand, multiple lamp (element) units for very rapid and simple labor-
atory AA analyses are commercially available. Several units are available
as dual AA/atomic emission (AE) spectrophotometers while even more versatile
instruments have interchangeable flame!ess attachments. In selecting a
laboratory AA, the trade-offs of simplicity and speed, cost and versatility
must be made by the user, TVA. A list of AA instrument manufacturers re-
commended for consideration or for contact for requesting competitive bids
was provided in Section 2.4.2.
Atomic or molecular emission techniques have been identified as ex-
ceptionally promising for the analysis of total dissolved sulfur in the
aqueous phase of wet scrubber slurries. Current methodology is based on
the tedious and time-consuming total oxidation to convert all sulfur species
to sulfate, followed by precipitation. It is suggested that a small feasi-
bility study be funded to determine if the aspirator-burner assembly design
parameters can be optimized and fixed to eliminate variations in excitation
energy. The potential range of applicability appears to fill a void in cur-
rent technology for sulfur measurement and is in the range anticipated in
the wet scrubbers.
-112-
-------
17551-6012-RO-OO
Of the sulfur species in the slurry, the sulfite (S03=) and bisulfite
(HS03~) S(IV) species were considered extremely important for process op-
timization considerations including efficiency, "scaling" and reagent re-
generation. No acceptable method for the rapid, automated or on-line
measurement of S(IV) species was available and consequently the UV spectro-
photometric method utilizing the bleaching of the furfural UV absorption at
276 nM was developed. The recommended instrumental laboratory method is
presented as Appendix E. A plan for complete automation of the method was
developed (Section 3.4) which will permit automatic analysis of approxi-
mately 100 samples per eight-hour shift. Immediate implementation of this
plan is urged to support start-up and operation schedules at TVA.
Several potentially automatable instrumental laboratory methods for
sulfate species in solution were evaluated theoretically and experimentally
(see Section 2.6). Based on the data generated and information from vendors
and users, two methods specific for sulfate are recommended for further
evaluation and development, i.e., 1) sulfate precipitation as barium sulfate
with either a measurement of turbidity optically or by measurement of excess
barium by AA, and 2) the ion exchange reaction with barium chloranilate and
measurement of the free chloroanilic acid colorimetrically. Because these
methods are either time-consuming or suffer from species interference, the
technical approach required would parallel that utilized to develop the sul-
fite/bisulfite method to determine and eliminate interferences present in
the wet scrubber environment and optimization of parameters. The sulfate
could be determined by difference very rapidly and conveniently if the total
sulfur analyzer is selected for development.
The feasibility of the batch automated or continuous carbonate deter-
mination by pyrolysis/acidimetric C02 release followed by quantitative NDIR
measurement has been demonstrated experimentally. The method should be
rapid, specific accurate and simple to instrument for specific candidate
scrubber processes. The nitrite/nitrate methodology literature was reviewed
and three candidate methods were identified and evaluated. It is not pos-
sible to select one for recommendation at this time. Further evaluation is
necessary.
-113-
-------
17551-6012-RO-OO
Finally, the bench scale wet scrubber process loop which was designed
and constructed for this study was found to be exceptionally valuable in
providing dynamic simulation of the process stream for methods development
and evaluation. Furthermore, the system was shown to be extremely useful
for understanding the wet scrubbing reaction mechanisms. For example, it
was determined that for bench scale scrubbing experiments, the rate limit-
ing step in holding tanks is that of calcium sulfite precipitation which
could control the effectiveness of sulfur dioxide absorptions by wet lime-
stone scrubbing. The unit will find utility in the development of other
process analysis techniques and is sufficiently versatile to simulate most
other candidate scrubbing processes.
-114-
-------
17551-6012-RO-OO
6.0 NEW TECHNOLOGY
This section provides documentation of new methodology and technology
contributions specifically beneficial to the analysis and characterization
of wet scrubber processes that were developed or conceived under EPA
Contract 68-02-0007.
Two of the new technology innovations were of sufficient novelty and
potential use to warrant submitting invention disclosures to the TRW Patent
Office. The subject matter of these disclosures are listed below:
Docket No. Title
72-096 Wet Scrubber Bisulfite Analyzer
72-100 Total Sulfur Analyzer in Process Streams
in addition to the invention disclosures three other new innovations were
identified during the program which offer unique approaches to character-
ization of wet limestone scrubber process streams. In addition to the
invention disclosures the following new technology items are described below:
• Application of X-ray fluorescence methodology for numerous
slurry elements
• Automatable solid and dissolved carbonate method
• Combined use of a small scale cyclone and dual filter
separator system for continuous phase separation
5.1 WET SCRUBBER BISULFITE ANALYZER
An instrument based on bisulfite bleaching of the 276 nM UV absorption
of furfural was conceived. This approach has been developed to be free from
et scrubber species interference, and is currently laboratory operational
M/ith elapsed analysis time for each sample of five minutes. A plan for
automation to allow approximately 100 analyses per eight-hour shift with
little operator involvement was proposed and is discussed in Section 3.4.
6<2 TOTAL SULFUR ANALYZER FOR PROCESS STREAMS
A new conceptual technique for determining sulfur in solution or dis-
solved from the solid phase based upon atomic emission or molecular emis-
sion utilizing flame, microwave and other suitable excitation was generated
-115-
-------
17551-6012-RO-OO
during this program. The sulfur analysis range, believed to fill the cur-
rent void in the state of the art from the low ppm level to 10 mM, is di-
rectly applicable to the wet scrubber slurry and is adaptable to dissolution
of slurry solids. The method possesses the potential simplicity accuracy
and cost effectiveness of the widely accepted AA technique.
6.3 X-RAY FLUORESCENCE ANALYSIS OF ELEMENTS
The utility of the X-ray fluorescence technique was demonstrated for
application to all wet scrubber samples. The state of the art of this
analytical technique for automated laboratory and continuous on-line analy-
sis permits analysis of calcium, magnesium and most other elements possibly
of importance to scrubber operation. Total sulfur can be determined quan-
titatively above approximately 0.03% absolute. The decided advantage to
the wet scrubbing program is accuracy, simplicity and cost effectiveness
on this single versatile analytical tool. Utilizing the recommended ARL
equipment, up to nine slurry streams can be analyzed for up to 15 elements
of a simultaneous and continuous basis with no sample pretreatment. The
technique is effective for the two-phase slurry or separated liquor and/or
solids.
6.4 PYROLYSIS/ACIDIMETRIC CARBONATE METHOD
Experimentation with actual and simulated scrubber solids and liquor
was performed in this study, utilizing pyrolysis for solids and acidifica-
tion for liquids to liberate carbon dioxide. In the preliminary tests,
quantitative detection was accomplished by ebulliometry and gravimetry,
however, an automated instrumental method has been proposed with non-dis-
persive infrared as the primary detection candidate and gas chromatography
as a reasonable back-up detector. The range of applicability is extemely
wide.
6.5 CONTINUOUS SLURRY PHASE SEPARATOR
A continuous staged separation system has been designed which is capa-
ble of achieving "instantaneous quenching" of liquid/solid reaction in less
than 15-30 seconds. The system which provides an optically clear liquor
stream for continuous analysis or grab sampling, is comprised of a mini-
ature cyclone centrifugal separator from DEMCO and a downstream dual-parallel
-116-
-------
17551-6012-RO-OO
polishing filter unit, The current design calls for minimum slurry feed
of one gallon/min and is easily incorporated in a self-contained portable
cart.
-117-
-------
17551-6012-RO-OO
APPENDIX A
SURVEY OF STANDARD ANALYTICAL METHODS FOR SLURRY COMPONENTS
This appendix delineates wet and some standard instrumental methods
for 16 of the key chemical species potentially present in limestone slurries.
Tabulated are the components to be measured, the principles of the method,
potential interferences and related comments. In general, utilization of
these methods requires separation of the phase prior to analysis.
-119-
-------
TABLE A.I
SURVEY OF STANDARD ANALYTICAL METHODS FOR SLURRY CONSTITUENTS
Components
so2
S03
N0x
N03"
N02-
"3"
SO,' v
HS03"
Reference
Techniques
West-Gaeke
T1tr1metr1c(H202)
Turbidimetrlc
lodo metric
lodometrlc
Sa1tzman(N02)
Bruclne
Color! metric
Color! metric
lodometHc
Tiirtaldlmetrlc
Tltrlmetric
Quarternary
«M>n1iM salt
A.C. polarographlc
Colvrimetrfc and
U.V.
spectropnotometHc
Method
Colorimetric(560mu) automated
ColoHmetric (SSOmu automated
for NO and NO, (KMnO. + NO ^
N02 ^ 4
Color1metr1c(470mp)
2,4 phenoldlsulfonlc acid
(480 np)
(520mp) dlazotlze-sulfanlllc
add 1-naphthylanine
I2-titr. excess with Na2S203
(or arsenlte)
As BaS04
Add Bad 2 and Indicator
Tltr. with perchloric add
Interferences
Ozone + N02 should be 8 mg
References
18, 19, 20, 21
18, 19, 20, 21
21
22
22
18. 23
18, 24
18. 25. 26
tn
tn
«j
t
CT>
O
ro
i
73
O
I
O
O
I
O
-------
TABLE A.I (CONTINUED)
I
ro
i
xxnponents
co3"
HC03
"a"!
HC03"|
Cl"
Ca"
Hg"
K+
Na+
Reference
Techniques
Gravimetric
UV spectroraetric
Gravimetric
Titrlmetric
Turbldimetric
Titrlmetric
Gravimetric
Volumetric
EDTA
Flame emission
Atonic
absorption
EDTA- Vol Metric
• Gravimetric
AtMlC
absorption
Gravimetric
Plane Emission
Gravimetric
Method
CO^ on a scar He
C02 on a sea rite
Sr+* titr. each ion separately
AgN03-AgCl(560mu)
AgN03 + dye for end pt.
(NH4) Mo04 weigh as Ca Mo04
Mssolve CaOx in dil. H2SOd
•Ml titr. with KMn04
Eri chrome Black-T indicator
(pH 10)
ppt. with (HH4)2 HP04
ppt. as KC104 or ICgPtClg
(767 nip)
Interferences
tony sepn. nee.
Mg interferences
Ca interferences
Separate Ca as CaHo04
"Use "radiation" buffers
Comments
Also manometric
Also manometric
Nephelometric even more sensi-
tive. Very sensitive 5xl(HNin
Cl and above
Good separation from Mg.
Also a colorimetrlc technique.
Lengthy
References
28, 29
28
4
28
28, 30
26, 30
31
31
en
01
en
o
i
20
o
o
o
-------
TABLE A.I (CONTINUED)
en
en
ro
ro
I
Components
Fe~
Fe+~
Mn"
Co"
Reference
Techniques
Flame Emission
Titri metric
Titri metric
Color! metric
EDTA
Atomic
absorption
Colorimetric
Tltrinetric
Volhard
Tltrlnetrtc
Atomic
absorption
VolMKtrlc
Color! wtrlc
Method
(589mu)
Redox- Ce"" in acid soln.
or KMn04 or K2Cr207
Redox- in acid + NH.SCN titr.
with std. TiCl3 or *T12(S04)3
1,10-phenanthroline + NH4OH-
HC1 (or hydroquinone)
(540my) oxld. with KIO. In
H^ to Mn04- *
Redox-NaB103 titr. with KMn04
or(HH4)2S203 + AgN03 titr. with
As02~~
Voluwtr1c-titr. with KNn04
K104(ox1d. to Mn04). Hg(N03)2
to ppt. - add FeSO. and back
titr. with KMn04 4
T1tr. with K« or EDTA
K1tro-R salt-control pH
Interferences
Use "radiation" buffers
Many metals
Many metals
B1, Cl", Sn, Br, 0,. I"
N02~. S03. and Fe(fl)
Ce. Cl~, Co, Cr, F~, HNO?,
Ni and U
Interferences by ppt.
Sepn. of interferrlng Ions
Comments
J_L A
Or oxid. to Fe and use color-
imetric, EDTA, etc.
Sepn. nee.
Sepn. nee.
26, 32
26, 30, 32
32, 33
26, 28, 30
ro
i
73
O
O
-------
TABLE A. I (CONTINUED)
ro
CO
i /
••^••MMH
•H-M-+-
u* ' ' '
Kg**
••••••••Ml
Reference
Techniques
Gravimetric
Atonic
-absorption
Potentlometric
Colorl Metric
Colorlnetric
Atomic
absorption
Voluwtric
Atonic
absorption
Colorl metric
Volumetric
GraviMetric
Atonic
absorption
Method
Ppt. with a-nitroso-B-
naphthol. Ignite and weigh as
co3o4
Oxdn. V > V(V) and titr. with
FeSO. using Pt and colonial
electrodes.
Phosphotungstate method
(409 my)
8- hydroxyqu incline and ext.
with CHCl3(55flBu) pH 3.5-4.5
Red. to V(IV) and titr. with
KMn04
Dlthizone Method (500 my)
[Hg(II)d1th1zonate or (610 my)
excess dlthizone]
Interferences
Bi ,r,K,Mo,NH*SCN~Sb.Sn,
Ti, and Zr. Colored ions such
as Co, Cr04> Cu, etc.
As, Cr, Fe, Pt, H2S, and Sb
Many metals and an Ions
Many metals and anions
Comments
Remove interferences by extraction
of CHC13 layer with alk. (pH 9.4)
aq. soln. (V to aqueous phase)
then reextract V with CHC13.
References
26, 28, 3C,
26. 28, 30
28, 30
••MMBMMM
tn
tn
CTi
O
ro
i
§
i
o
o
-------
TABLE A.I (CONTINUED)
—i
en
en
ro
i
Components
Al***
Pb"
As*"
Reference
Techniques
Gravimetric
Volumetric
Colorlmetric
Atomic
absorption
Colorlmetric
Atomic
absorption
TltHmetrlc
Colorlmetric
Atotric
absorption
Method
Ppt. with 8-hydroxyqulnoline
Ppt. with 8-hydroxyquinoline add
excess acid and std. KBrO.-KBr
soln. then excess Kl and litr.
liberated I, with Na,S,0,, starch
indicator. z z z 3
Reac. with NH. auHntricarooxylate
(525 nn.) 4
Oithizone (520 mp) or Di-(3-naphthyl-
thlocarbazone.
Using KMn04 as titrant
AMKmlun molybdate (840 nm)
Cu + Ni
Other metals
Many separations necessary
Other colorimetHc methods.
28, 33
26, 35
30. 33
ro
i
o
i
o
o
-------
17551-6012-RO-OO
APPENDIX B
SURVEY OF ELECTROCHEMICAL METHODS FOR ANALYSIS
OF DISSOLVED OXYGEN AND SULFUR DIOXIDE
This appendix summarizes electrochemical methods for the determination
of dissolved oxygen and sulfur dioxide contents. Criteria used for evalu-
ation of these methods included:
• The present state of development (laboratory technique,
availability of laboratory or process stream instruments)
• Sampling mode
• Specificity and interferences
• Pretreatment of sample for analysis (concentration, filtra-
tion, adjustment of pH)
• Useful concentration range, sensitivity, accuracy, pre-
cision
• Temperature requirements
t Analysis time
• Data reduction capability
t Requirements for further development
• Maintenance requirements
• Cost
• Life time and cycle life
0 Commercial instruments
Table B.I identifies electrochemical methods for the determination of oxygen
and Table B.I I electrochemical methods for the determination of sulfur
dioxide.
-125-
-------
ro
CT>
i
TABLE B.I
ELECTROCHEMICAL METHODS FOR THE DETERMINATION OF OXYGEN
B.I.I Direct Polarographic Determinations
The Present State of Development:
Sampling Mode:
Specificity and Interferences:
Pretreatment of Sample for Analysis:
Use Concentration Range, Sensitivity
Accuracy, Precision:
Temperature Requirements:
Analysis Time:
Data Reduction Capability:
Requirements for Further Development:
Maintenance Requirements:
Cost:
Life Time and Cycle Life:
The technique is based on the measurement of diffusion current for the reduction of dis-
solved oxygen gas on mercury, gold or platinum electrode. The reduction on mercury is
well studied. Reduction waves on solid electrodes are complicated by reduction of ad-
sorbed oxide film when studied by scan voltammetry. Most of the work is accomplished
using solid electrodes with a membrane. The membrane provides selective sampling for
oxygen gas. The fine particulate suspensions in the slurry liquid may poison the solid
electrodes if used directly. Because of the relatively short life of amalgam electrodes,
the direct reduction method is advisable for dropping mercury electrode only. The drop-
ping mercury electrode is used for the polarographic determination of dissolved oxygen.
The technique is suitable for laboratory bench.
Batch sampling only.
Specific for oxygen in the slurry composition.
(vs. see).
Current measured between -0.3 to 0.6 V
The sample can be used directly without the adjustment of pH.
Can determine down to 0.5 pom. Higher concentration limit depends upon the slurry ex-
tract composition. Can go up to 100 ppm. The technique is capable of providing better
than 5% precision at the lowest limit.
Can be operated at any temperature. Should be calibrated accordingly.
As the gas is reduced directly from the solution, response is high. Results obtained
immediately.
Current readings should be converted to concentration by analog procedure. The current
should always be measured at the same time, e.g., at the end of drop life.
Very little.
Mercury electrode should be carefully maintained.
Inexpensive.
Mercury reservoir should be continuously replenished.
en
en
ro
i
•?o
O
I
o
o
-------
TABLE B.I (CONTINUED)
B.I.2 Polarographic Determination Through Membrane
The Present State of Development:
I
ro
Sampling Mode:
Specificity and Interferences:
Pretreatment of Sample for Analysis:
Use Concentration Range, Sensitivity
Accuracy, Precision:
Temperature Requirements:
Analysis Time:
Data Reduction Capability:
Requirements for Further Development:
Maintenance Requirements:
Cost:
Life Time and Cycle Life:
Commercial Instruments:
The basic principle is the same as in A.I.I. The electrode is a solid, planar electrode
and is separated from the solution by a membrane through which oxygen diffuses. The solid
electrode is polarized at the diffusion region for the oxygen gas. The electrode could be
polarized with an external source or by a galvanic couple like lead-silver system. Lead
electrode is the anode in the basic electrolyte system contained inside the assembly. The
limiting diffusion current for oxygen reduction is due to the membrane permeability to
oxygen. It is assumed that steady state is attained for dissolved oxygen on the sample
side of the membrane. Many process instruments based on the above principle are available
in the market. The technique is fairly well developed.
Either continuous or batch.
Membrane provides selectivity for oxygen.
To prolong the life of the membrane it is better to minimize slurry particles in the sample.
0-40 ppro dissolved oxygen. Quite sensitive.
Operates from 32°F - 110°F.
Fast response, usually 30 seconds.
Can be read directly as ppm by weight of dissolved oxygen.
Both the galvanic and potentiostatic methods are well developed and commercial instruments
are available.
The electrolyte inside the system should be replenished periodically. Membrane life is
usually long, unless damaged by violent physical shock. Membrane can be replaced without
difficulty.
Commercial oxygen analyzers are produced in large scale. Price is competitive.
As the system is well-sealed and there is no moving part, the detector could last from six
months to two years without servicing.
Potentjostatic: 1) Instrumentation Laboratory, Inc., 2) Honeywell, Model S 914-21 Do-meter,
3) Yellow Springs Instruments Co., YSI Oxygen Meter, 4) Weston and Stack, Model 400,
5) Delta Scientific, Series 3210 and 6) Sectarian Instruments, Inc., Model 735.
Galvanic: 1) Bio Marine Industries, Model DOA 555, 2) International Sales Associates Model
QM 10, and 3) New Brunswick Scientific Co., Inc.
en
en
en
o
ro
i
70
O
O
o
-------
TABLE B.I (CONTINUED)
B.I.3 Conductivity Measurements
en
01
ro
CO
i
The Present State of Development:
Sampling Mode:
Specificity and Interferences:
Pretreatment of Samples for Analysis:
Use Concentration Range, Sensitivity
Accuracy, Precision:
Temperature Requirements:
Analysis Time:
Data Reduction Capability:
Requirements for Further Development:
Maintenance Requirements:
Cost:
Life Time and Cycle Life:
Commercial Instruments:
The technique is based on the principle that oxygen reacts with thallium metal in aqueous
medium to produce soluble conducting thallium hydroxide. The conductivity increased by
35 micro mhos/cm for every ppm of oxygen. Instruments based on this principle are avail-
able for process stream.
Sampling can be continuous.
Although the principle is specific for oxygen, as the technique is based on conductivity
measurement, the solution should be demineralized for analysis to provide useful
sensitivity.
Solution should be free of ions.
The technique is extremely sensitive to oxygen. Conductivity is measured before entering
thallium tube and again after it leaves the thallium tube - 0-1000 ppb (parts for billion
by weight) can be measured.
Temperature should be constant. It is preferred to keep it near 25°C.
Fast.
Conductivity difference is directly read as parts per billion.
This method is useful for relatively pure systems, like demoralizing plants. For slurry
studies, further development needed.
Ion-exchange column and thallium column should be attended to periodically.
Relatively expensive.
The ion exchange column and the thallium column should be charged periodically or the ion-
exchange column should be recycled.
Beckman Instruments, Inc., OA-5K DO Analyzer.
ro
i
73
O
I
O
O
-------
B.I.4 Coulometric Determination
TABLE B.I (CONTINUED)
The Present State of Development:
ro
<£>
i
Sampling Mode:
Specificity and Interferences:
Pretreatment of Sample for Analysis:
Use Concentration Range, Sensitivity,
Accuracy, Precision:
Temperature Requirements:
Analysis Time:
Data Reduction Capability:
Requirements for Further Development:
Maintenance Requirements:
Cost:
Life Time and Cycle Life:
The method is based on the principle that oxygen is quanti-
tatively reduced by electrogenerated free radicals like
methyl viologen cation radical. Though the method is more
suited for gaseous stream, it can be used for dissolved
oxygen by batch process. A known amount of sample is added
to a solution containing totally reduced methyl viologen.
The reoxidized methyl viologen, is reduced by passing cur-
rent, and the charge passed is equivalent to oxygen added to
the sample.
Batch sampling.
Non-specific for oxygen; any oxidant will give the same
result.
Slurry solution does not need any special treatment. pH
should be near 7.
Not known yet. Should be in sub ppm level. Very high con-
centrations can be determined.
Any temperature.
Time of electrolysis.
current.
Would be 5-10 minutes depending on
Directly convertible to concentration.
Needs further study.
Maintenance is simple.
Inexpensive.
Methyl viologen solution may have to be replaced after every
20 analyses due to dilution and increased volume.
CJl
Ul
I
en
o
no
i
;0
O
I
CD
O
-------
co
o
B.I.5 Corrosion Probe Method
The Present State of Development:
Sampling Mode:
Specificity and Interferences:
Pretreatment of Sample for Analysis:
Use Concentration Range, Sensitivity,
Accuracy and Precision:
Temperature Requirements:
Analysis Time:
Data Reduction Capability:
Requirements for Further Development:
Maintenance Requirements:
Cost:
Life Time and Cycle Life:
TABLE B.I (CONTINUED)
This method is based on the measurement of corrosion po-
tential of aluminum electrode in the solution. The poten-
tial is measured with respect to a gold electrode, which
behaves like a "quasi" reference electrode. The aluminum
electrode is in contact with a wetted frit, which separates
the gold electrode, immersed in distilled water. The oxygen
gas is sampled from the test solution, by means of a carrier
gas. The sudden variation of the potential of the couple,
as a transient, is related to the concentration of oxygen.
This technique is still a laboratory curiosity.
Suitable for batch sampling.
Any gas affecting aluminum corrosion will give wrong reading.
S02 may affect readings.
No special treatment is needed, as the carrier gas carries
the dissolved gas.
Sensitivity is 0.1 ppm.
Ambient.
4-6 minutes analysis time.
Analog output.
Needs a lot to be done, as the method is based on empirical
approach.
Maintenance is delicate, as it involves measurement of cor-
rosion potential.
Inexpensive.
No information is available.
tn
tn
en
O
ro
i
20
o
i
o
o
-------
I
CO
B.II.l
TABLE B.II
ELECTROCHEMICAL METHODS FOR THE DETERMINATION OF SULFUR DIOXIDE
Coulometric Determination
The Present State of Development:
Sampling Mode:
Specificity and Interferences:
Pretreatment of Sample for Analysis:
Use Concentration Range, Sensitivity
Accuracy, Precision:
Temperature Requirements:
Analysis Time:
Data Reduction Capability:
This method is based on the principle that sulfur dioxide solu-
tion is oxidized by bromine or iodine, which could be electro-
generated. Exact amount of bromine or iodine is produced by
anodic oxidation of bromide or iodide solution and the amount of
bromine or iodine produced is coulometrically found from the
total current passed. Any excess bromine or iodine generated
is detected by an indicator electrode, which is used for stopping
electrolysis. This technique is well developed for process stream,
but only for S02 in gases. Using these developed instruments for
continuous analysis of solutions is not practical. However, these
commercial instruments can be used for sulfite analysis in aqueous
solutions by batch sampling.
Batch sampling.
Other reducing agents in solution, may interact with iodine or bro-
mine. However, the slurry does not contain any other reducing agent.
The sample should be clear, free of particles. The solution should
be acidic.
This technique detects sulfur dioxide from 0 to 1000 ppm range, in
gases. Sensitivity in ppb.
Ambient conditions.
There is a lag time, as this technique needs mixing and electrol-
ysis. Maximum time requirement is about 5 minutes.
The readout is in coulombs, which is directly convertible to ppm.
Requirements for Further Development:
Maintenance Requirements:
Cost:
Commercial Instruments:
Very little further development.
For solution analysis, the reagents get continually diluted.
Solutions may have to be replaced after a set of analyses.
Price on these instruments is competitive.
For gas samples: 1) process analyzers, Titrilog II, 2) Barton ITT,
Model 286. 3) Phillips PW-9700, 4) Beckman Model 906. and 5)
Atlas Electric. Modes 1 200 and 210.
en
cn
i
en
o
ro
k
o
o
-------
TABLE B.II (CONTINUED)
B.II.2 Polarographic Determination Through Membrane
en
The Present State of Development:
CO
ro
i
Sampling Mode:
Specificity and Interferences:
Pretreatment of Sample for Analysis:
Use Concentration Range, Sensitivity
Accuracy, Precision:
Temperature Requirements:
Analysis Time:
Data Reduction Capability:
Requirements for Further Development:
Maintenance Requirements:
Cost:
Life Time and Cycle Life:
Commercial Instrument:
The method is based on the principle that sulfur dioxide can be electrooxidized on a sens-
ng electrode, at a given applied potential between the sensing electrode (Au) and a counter
electrode (Pb62). A thin membrane, selective for S02, separates the sample and the sensing
electrode. The limiting current is due to diffusion of S02 across the membrane, and is pro-
portional to SO, concentration in the sample. The technique is S1milar to oxygen detection,
except the membrane is selective to S02 and instead of reduction current, oxidation current
is measured. One such commercial instrument is available. The device can be used for both
gases and solutions. However, laboratory studies have shown that the current at the sensing
electrode is dependent upon the previous history of the electrode.
Applicable to both continuous and batch sampling.
The specificity depends on the membrane selectivity for S02 alone, and the oxidation poten-
tial. In the slurry the interfering ions would be barred by the membrane.
The membrane isolates the sample and the electrolyte. The sample should be free of fine
suspension.
Concentration range 0-5000 ppm.
Ambient 40°F - 110°F.
Less than 2 minutes.
Analog readout, so directly convertible to ppm S02
The stability and reproducibility of the sensor electrode is not fully studied.
Not known. Expected to be stable for at least three months.
Not expensive.
Life time of the sensor is expected to be long. This could be easily replaced.
Dynasciences Corporation.
ro
i
73
O
I
O
O
-------
TABLE B.II (CONTINUED)
B.II.3 Direct Polarographic Determination
The Present State of Development:
CO
CO
I
Sampling Mode:
Specificity and Interferences:
Pretreatment of Sample for Analysis:
Use Concentration Range, Sensitivity
Accuracy, Precision:
Temperature Requirements:
Analysis Time:
Data Reduction Capability:
Requirements for Further Development:
Mai ntenance Requi rements:
Cost:
Life Time and Cycle Life:
The principle is the same as in A.II.2 except that there is no
membrane separating the sample from the indicator electrode.
This is possible only in the case of the liquid samples. The
limiting current in both the cases is proportional to the S02
concentration. The direct reduction can be accomplished
either on mercury electrode or Pt electrode. However, the
direct reduction method can be used only as a batch process.
Direct oxidation of $03" on platinum electrode can be for
analysis.
Batch sampling.
Can be made specific for S02 by potentiostating.
The pH of the solution should be near 0.
Not known, but should be in the same range as in A.II.2.
Ambient.
Sample has to be transferred to the electrochemical cell. The
cell should be ready with supporting electrolyte. Time of
analysis will be about 10 minutes.
Analog readout. Can be read as ppm from calibration chart.
The system should be studied before employing it as a routine
method.
Simple.
Inexpensive.
Lifetime, practically infinite.
C71
ro
i
3
o
O
-------
17551-6012-RO-OO
APPENDIX C
INITIAL SHAWNEE PROCESS DEMONSTRATION OPERATIONAL MODES
This appendix provides details of the initial experiments for assess-
ment of configuration, flow rates and solid loadings planned for testing by
Bechtel Corporation of the wet limestone scrubbing process for abatement of
sulfur dioxide emissions from the Shawnee Power Plant, Paducah, Kentucky.
These details were used to identify sample locations and sampling rate in
the studies conducted during this program.
-135-
-------
VtUTUB.1
Co
CT>
I
HOLD
(z
••
at
T»
EFFI
Utt
® "WTR1*
ffi "®
SLURRY
SfttAM
MS
1
^
3
4.
5
C.
7
8
_!.*•
i
to
FLOW[GPM] « % SOL IOC
MAI.
GPM
leOO
511
512
375
«S
650
IVJO
Z3
JOO
1
«*•
V?
14
15
IS
C.Z.
30
O7
o
o
14
to
SO
T-«»lC.»l.
&l"M]«iS
400 , 2
11 ' tS
JOO (.5
75 j a-i
n :lo
5«. i c3T
SOO O
15 !'o"
100 | j
S !<>o
n ,3o
!
i
i
1
M.U.
«'!!;
:io
20
0
• 0
n
4JO
4-<7.
-
O
50
4-
iiT
%s
o /
4.f
O<
O
(0
<3 2.
O
"«
O./
_..
60
"iS
L
PttiiU2L[r(l6)
nu.
30
UMH.
1C
— ;
JO
(9
\
niM.
5
S
'——.-
L
j
©
-
Linesi
rONE. POMO
I
OPE.RATIMG
101
MODE!
: i
• NOT TO BE USED FOK DESIGN.
QUANTITIES SHOWN AM fOK
TEST MOGKAM EVALUATION.
6lM
1 **
BECHTEL
<>*•• r«*i, ct;to
N. X. P C A
ALKALI scauBfiNi T£6f
SLURB.Y now SHEET
M- tot - Mt-
— I
Ul
en
i
cr>
O
ro
i
o
o
-------
SLOO.RV n.ow
STKAM
1
2.
3
4-
_5_
"7
8
q
10
IA
II
IT
FLOVl[lfM] t V. SOLIDS
MAX.
GPn
boo
Sit
S7t
575
•»0
*5
t.
KS*
sso
20
390
1
45
**,
.4
15
IS
O.I
o
30
7o
o
0.7
o
14
(ft
30
1YMC.AI,
£-?•£ ,
12.
-— •
3OO L S
75
11
• n
500
J Z
0
SO
10
o
0.7
IJ o
too
S
to
n iso
j
;
j
HIM.
ZiO
£S
«./
Jo | 04
o i O.4-
3
II
n
2
427
450
O
JO
4 .
0
0
•e
/a
£0
j
01
0
0.1
bo
\5
•
I
P^«iU«C[M«l
MAX »io«H.
30 j <0
1
!
HIM.
J
30 | 10
i
J
i
5
OPERATlNS.MoOE
102
NOT TO IE USED FO« DESIGN.
QUANTITIES SHOWN ME FO*
TEST F-ROGKAM EVALUATION.
~1-
BECMTEL
N. A.. P. C. A
ALKALI SCBUB*IN&
SLUR8.Y FLOW SUCCT
&1SS
M-IOt- Mi-J.
01
tn
en
o
ro
o
o
n i ii f mi
-------
OJ
CO
I
SLURRY
STREAM
1
2
3
4
5
&
._!_
«?
10
II
12.
IA
13
14
FLOW [fcrrt] t
fA«.
Gm
too
511
i'?!.
57 I.
S75
3S
4.
Si.;
850
20
Z.GC.
a
45
%i
14
IS
15
IS
15
0.7.
o
30
•70
o
-------
ictut
HOIO
IEB.
T«k
SfTL
KIK.
co
SLURRY r».ov«/
STREAM
N!
1
"z
5
4
5_
6
7
.„..
~"io"
n
12,
3A
13
'I*-
PLOW I'.Ftil t
r-«AX.
GPn
(.00
toe
too
511
572.
35
4.5
;o
eSo
gjo
20
Zoo
45
10
14.
1*
1*
IS
15
0.2.
0.(
jo
7o
0
0.7
0
14
3o
to
V; SOLIDS
1YC10AL | MlW.
i-DO
4CO
5
5
5
^^!c>.s
3OO 6.'J
rT.
n
S
soo
530
/Oo
0
o;
JO
0
0.7
0
"s
30
60
'
S'M 51S
2?0 !<5./
/oo Jo./
/oo
ZO
o
0
II
2
So
n
i
O.I
0.1
O 1
O.I
10
to
0
0.2.
0
O.I
IS
(TO
Pfttssu«t[r*i
.« '„•.•;"
BECHTEL
r, » •. r • » N c i •, c o
N. *. I*. C . A
scRuseiNfr T£«T
VE.NTU*!
SLURRY FIOW SHEET
M- 108 - M)-4
Ol
cn
ro
i
§
-------
I
CD
•vj
Ul
01
I
<7>
O
STREAM
i
z
3
4-
5
6
7
8
3A
10
II
Ptow[<,»M]< ?< SOL IDC
MAK.
(.00
4,00
Coo
572.
joo
£?C
. /z
20
200
ZfcO
e
10
%i
3 7
0.7
0 7
11
2-1
o
2.1
O
0.7
O
fro
(0
TYPICAL
>\7'J
4co
<:o
31E
O
3&i
5io
(4
/OO
IZZ
8
0
"/. S
o
o
a
1.5
AS
O
o
e>
to
to
MlN.
&PM_;
4OO
4oo
400
ML
o
3*U
0
Jo
4-
4-
JSS
o
o
•J
O.J
O.J
c>
01
O
C
o
*0
4.0
p«t$iuat[ni6)
MA,.
. .'
30
30
Uo*M.
10
10
HIM.
5
S
j
TOO
300
* NOT TO BE USED FOR DESIGN.
QUANTITIES SHOWN ARE FOR
TEST PROGRAM EVALUATION.
BECHTEL
s « * r» «« c i •, c o
N. A. P. C A
ALKALI SeauaiMNfr TECT
PLOW SMCET
M • IO» - Mb- S
-------
.n..
VtMTUB.1
SCE.UB5EK ffT,
HOLD
TA.KIK.
®
WATtft
HOLD TANK.
DOLOMITE. POND
STRUM
NS
1
Z
3
A
5
6
1
&
IK
<\
10
II
PLOW(*,PM] < */„ SOLI DC
MAX.
6PM
COO
511
-SI*
577-
>00
Sic
Sil
20
Zou
2£o
8
10
% s
O7
/ S
I.S
/.s
/.s
0
/.£
0
0.7
o
60
60
TYPlOL,
C.PM
400
HZ.
•55.
0
I.S
i'6
SLURRY FLOW
M- lot
CJI
cn
i
o*
o
IN5
§
O
o
-------
SLUR.RY
STREAM
MI
1
2.
3
d-
-f-
7
a
q
10
it
Ti~
13
FtOw['.fti] ( % SOLIDS
MAX.
GPI-I
LOO
£.00
'.CO
S"»2-
S71
«SC
350
25
/« «
7-OO
to
14-
14-
14
IS
IS
0 1
30
O
0.7
O
>4
lo
to
TYPICAL.
4-so
40C
-------
SLURRY
s
I) AT A
VtUTOttJ
SC.KU&M.K.
.£»
oo
STREAM
N»
1
._-„.
3
4-
S
&
7
2
<^
10
II
36.
IZ
FLOW^.PM] $ % SOL IDS
MAX.
too
to.-)
(.CO
sit
5'TL
SIS
15"
tib
*S
850
8
Zoo
4?
ss_
U
14.
1*
IS
IS
0.7.
O
0
30
AS
60
/4
TYPICAL
6PM
4*0
4-30
??-
ICO
% s
5
—
5
S
t.s
«..S
"
SLURR.Y FLOW 5 MEET
01
I
o^
O
PO
73
O
O
O
-------
|ftbsvu««.y
TO.IWIK.
Puow
1
•2.
3
4
5
(o
_7
q
1(7
(1
I A,
17.
FLOW[',PM] 4 % SOLI DC
MAX.
iOiD
512.
S12.
ST2.
57S
1?
iJS
4S
SSC
£
200
45
»ii
14-
IS
IS
15
IS
e.t
0
o
30
60
14-
zo
TYPICAL] HIM.
400
5
3-!?.Ui
i^r
°i
t i
jcolts
7?
zcr
Soo
n
j^:
s
;oo
n
0.?.
o
o
if-
O.7
bo
5
10
!
G'M i^JS
ZJO io.l
<7C 0.4
.•£_
-°-J
e
o
04
0.4
0.4-
0-1
O
n j_'o
450JO.Z
4- ! 60
ffo
n
O.I
IS
PR«tURt[«tG)
MM.
JO
70
**H.
/O
to
MiN.
5
5
L '
d ,
/
®
; I ' DouoMifE TV>MP
MODE
107 B
• NOT TO BE USED FOR DESIGN.
1 QUANTITIES SHOWN ARE FOX
•' TEST PROGRAM EVALUATION.
'—rt'A-,—iKil —
BECHTEL
N. A. P. C.A
ALKALI scaueeiN&
SLURRY FLOW SHEET
1?.E
Wt- loft -MS- 9
cn
01
l
o
i
§
o
o
-------
tn
I
TAJJK.
WTR.
TA.UK.
1 i f
SLURRY Twow DATA
STOAM
N»
I
E
3
4
5
6
7
8
«?
\a
.'.A,
11
12.
Ptow[tni] ^ ^i SOLIDS
MAX.
6PH
foOO
too
too
SO
S7S
41
.'5
850
850
,30
jc»:
850
850
Xi
!T
14
14
IS
IS
o
40
o
z
o
(4
•L
O
TYPICAL
GPM^
40c'
4cc
%,s
10
10
Ase i 10
'jt
34-^
t-
£
1OO
503
/fc
ICO
v?s
Z'Jo
II
II
C
(.0
C
1
J
10
1
I.S
O
HIM.
6>M XS
100
100
iOO
ts
7.S
10
5
152.
4 SO
8
So
Z1S"
Z10
3
J
t
4
4
a
4c
c
O-S
0
3
t.4-
O
pftesiURt[rtt&3
MM.
Me**.
... f
30
30
10
10
MiN.
J-
5
.
to
l*o
MOD£
108
• NOT TO IE USED FOR DESIGN.
QUANTITIES SHOWN ARE FOR
TEST PROGRAM EVALUATION.
ItlL-fO l-Ca
I
lei"
BECHTEL
N. A . P. C .
FACILITY
SLURB.Y FLOW SHEET
W- 106 - MS- II
Ol
en
a*
o
ro
XI
o
i
o
o
-------
SLURRY Tuow
STREAM
N!
1 -
2
3
4-
5
C,
7
8
Cf
10
IA.
II
n..
FLOW^PM] i % SOLID:
WAX.
&PCI
L.OO
59J-
C'l
SO
ETS
tr-L
IS
a so
1C
JfPO
250
e«>
14-
14
I'r
'.S
IS
0
do
a
1
o
14-
i
0
TTPICAL
100 | (O
'j^J I0
i*O
34-L
4-L
8
Zoo
503
/b
IOO
Z<1S
t^O
o
- ....
n
il
a
C
1
o
IO
-^
MIN.
t 3O
!00
IC3
15
^5
/o
S
4-50
s
3
5
}
4.
4
0
4o
OS
cs
o
*> J3
W3 |o,f
110 JO
I
j
I
PdtStURt[MI6l
MAX
30
)0
MOKM.
10
10
MIN.
s
S
*
—I
tn
01
cr>
O
«^
ro
i
73
O
O
O
to
OPERATING MODE
108 B
•NOT TO BE USED FOR DESIGN.
. QUANTITIES SHOWN AR£ FOR
t TEST PROGRAM EVALUATION.
Hi*
~T
BECHTEL
N. A.. P. C . A
ALKALI scnueeiN* T£«T
> SrsjeH
SLURRY FLOW Sneer '
kA- IO&-
-------
'SLURRY Tuow DATA
STREAM
N»
1
Z
3
4.
5
6
7
S
3A
1
10
-i - —
Ftow[orri] 4 J4 SOL ICE
MAX.
GPtt
c,oo
OVi
600
S^L
zoc
J-7C
S1Z
20
AOO
lo
eso
M
07
0.7
22
T -t-
C
t f
o
0.7
60
0
TYPICAL
<;>PM :%S
4co; a
«0 i '->
4 3- a
31Z
c
3 at
530
14
foo
8
Uo
o
/ s
'.5
O
I.5
C
C
do
0
HIM.
GPM
4c'
•tCC
O
ro
70
O
O
O
-------
VEUTOB.I
SC.K.U6BCK.
3:
I
oo
•^ 1(3)
KEC.-E
WXTfcR.
SLURRY
STREAM
i
2
3
4-
5
6
7
a
IA.
<\
10
*
Ftow [<.m] < % SOLIDS
MAX.
6PM
too
S<>2.
112-
571
ZOO
s?o
s-r*.
io
AOO
10
SSo
%i
J-7
'•5
1.5
1.5
.' S
0
'5
0
0.7
£>D
O
TYPICAl-
6PI' '/.SI
4oo o
i-?z'
J?L
AS
/.5
S^ZJ/.S
O j/. 5
3S6.
530
If
/oo
8
0
0
(7
60
120 i o
___~
MlH
*30
.'?:-
its
o
0.}
si rz. lo.J
572
ii
4SO
0
50
0-5
03
0
0 J
a
0
4 |6o
ST4-|0
.
PREJlURE.[rUs I t««CUCO
M A. P. C. A
ALKALI scBoe»iN6 TE«r FACILITY
icRU3BI>i<>
SUURH.Y FLOW SHEET
M- IO6 -Mi-ll -
-------
STREAM
NI
\_
2
3
4
?
6~
7
a
<\
10
3A
II
1
TLOW[OPM] ( »4 SOLlDC
MAX.
GfriJ* S
tr>---
t,Cx-j
i.00
50
STJ
4^
15
CLO
Si'o
20
300
ts
H
14
!•«
15
IS
O
i.0
O
Z
O
14-
za
1V?ICAL
6PM
4c.o
4os
-v-,o
£o
i42
1T1
.«
Coo
533
/&
too
M
10
10
IO
ii
2L
0
X.
O
1
0
IO
ZS 1 18
HIM.
. C. A
3cRUBBi"6
PLOW Sneer
M- IQ6 -
171
01
ro
i
§
o
o
-------
vtuiuu
SCROOfttB.
M«(_OTA,KIK.
,^
L tm
MK.
1
ttCIR
Ul
C3
I
PltOCt^S VMTtft
WOLO TA.UK.
\F»\.Ttlt/
SLURRY n_ow
N!
1
2.
3
4-
5
-------
tn
•
\
STR«=*M
NS.
1
2
B
4
5
U
7
3
')
10
1 1
il
FiowfoPr^i % Souos
h-lfcX.
6PM '% 5
1.1 or
II!'.-
U'.:r
ui^;
«.:
•?v
6
£V.
iSO
?o
*c.
S
-•'•C,-^'
I -A
i z •.'•;•
U
5
3
1 23 '.
n
T
5.
G42.
5eo
I*.
n
n
OPERATING MODE
20 1
.' ' NOT TO BE USED FOR DESIGI'
* QUANTITIES SHOWN ARE FO
. TEST PROGRAM EVAlUATIOh
1
x- •' ~- ^- •: -.
??
jr>
C
to
d.l
C
0
^o
2o
f
Ml*
Pr-i
Joo
4'.
(J
re.
~
K,
4
450
412.
T
n
n
"I
1
J. '
K .
'. ;
f:
/i
£•
i'-*5
0
. >•",
-
-
IS
C.
J.C
w
c
o
lo
IS
WlSSUBE[«kn
r>x.
r;
A.M.
5"
Mill.
O.i
Vl?
*^
C».' "''f
.•^•;t5 fn VPvc.M.
-i^ir^—— r~
ioHPUWT UK)
ftflMfUHT 300
-1-4-1
f-.>''1
E--j£.-*?~
BECHTEL
;»» < XNCISCO
tJ A. PC. A
ALI.cn-iTx(<
Tr.A SC.Pu.miiU':, SY1«'»sM
SUUP.PX FLOW SwEGT
(@?
IM<.
i.^5
t**» k« ' Ml
M- 109 - MS- 1 0
tn
tn
i
m
o
i
§
o
o
-------
en
ro
1
?_
i
4
- 5
(o
1
&
9
10
1 1
a
Fiow(i>pr'i] £ % SOLIDS
r-iA*.
6PM '% ',
I.-.CC-
ib'::
!•:<
:57<
^1
il
fi
JS^
ISO
Zc,
fe
45
IS
li
••..
0
JO
c
u-
of
0
0
lo
3o
TYPICAI.
6i-M
;/:.-
.'•?'-•
;:;?.
(2C
:7
8
53o
Soo
/*
•j
n
VLL
IT
5
'J
*>
'/
£.0
O t
0
O
7o
2o
M.rt
f?M
^-•-.
•)•->
o
i ft
n
>.
4
450
42.T
7
7.
n
T-s
O
0-0
-
15
c
t*
c. i
j
C
to
IS
9ftEiSu«C«Kfl
:r-
••laiM.
5
Mill.
^•5
in
en
i
01
CD
I
70
O
O
O
100
3oo
OPERATING MODE
2.0 -L
>OE
'NOT TO IE USED FOR DESIGN.
I QUANTITIES SHOWN ARE FOR j
TEST PROGRAM EVALUATION. .
Gi •
BECHTEL
Skll IKKCMCO
fJ A. PC. A
Au*Au SC«u&l»lfJ&
TtA
SHEET
- 2
-------
C.OOL61.
up
CD ®
~ J '
"1 \**s
i
P8OCS85 WATER.
HOLE* TA.MK.
FV.OW DATA
Ni
to
94 sonos
V.5
IO
10
10
to
•SCO
530
to
Ml hi
f-P^**' ij
150
60
Mid.
OF-ERATif4G MODE
''}
^ • NOT TO BE USED FOR DESIGN, i
• QUANTITIES SHOWN ARE FO* '•
J TEST PROGRAM EVALUATION.
.'J%~*~» ^* ^ ^^ ^* \^\»» »\\ v^Vx^^ VW^~\~\\ . % \iN* v
*\\\*% \»\ \\\» \% »\» x\»*%» * » V \-.\v „ v\%
BEGHTEL
SAN fMANCISCO
iJ A PC A
TC.A
SLURRY Ft-ov« sweet
CIS 5
- icq -
en
en
i
o
ro
§
O
o
-------
en
-P.
SIURR.Y n.ow DATA
STMA.M
1
5
4-
S
la
~l
8
9
10
It
12.
13
Fl&w[GPKU JiSouOS
r-i*x.
<">Pf1 !
z<-
zn
l-l
LV-.
M-.
:r;
17
/Z.3
7
£
51o
172
/6
31Z
2
^8&
°AS
S
c
c
i'V
S
ft
'«.
o.t.
0
o
I.S
1,0
0
Ml>4
f-?v.
j.O
•»5
O
/7
.'a
6
*
4^0
?o
T
^Z
4-
ISfo
*. ^
•»
C.I
//,
-
n
fri
OJ
.<•
0
0.3
bO
o
paESSoetp
KV.K.
i
**M.
3
s«l
Mid,
o Z
E
OPERATING MODE
4 • NOT TO SE USED FOR DESIGN, i
i QUANTITIES SHOWN ARE FOR '
1 TEST PROGRAM EVALUATION. '
rwifUJT loo
\»Q
loo
BECHTEL
SAN INAMCIiCO
Nl A PC A
ALKAU scRLB'i'Nii Ttsr O.CH.ITY
TC.A
SHEET
109
-vl
VI
en
i
O
•ya
o
i
CD
O
-------
©
HVQR.AT6 MAKE UP
^
®
^©~
1
i
~
WA.T4B
,©
pBocsaa wATtK.
NOLP T4.KJH
PCMO
DATA >
N*
1
^
^
4
5
(o
1
6
9
10
il
\1
13
FIOW[GPM]£ yf sotios
I-1A.K.
IGPM i%s
/.;.-c
'£-.'•'
•JST
.3-;:.
*5
v'i
a
«i-..
ii'c,
/o
-SG
•1C.
10
!'-•
.5
/S
o
30
r;
60
•-1 ."i
•-
o
»"J
3o
to
TXflCAU
6PM :^s
I-W-,
:-o
/.-:;:
/Z3
/ 7
1
•
533
SCO
£
II
n
g
5
J
5-
-•
JC
0
to
<:.f.
-
c
£•
^o
to
.r.1:
Joo
*-S
G
,/;
'7
«
-»
-5SD
5
'0
n
4-
'}-'>
<.
Cf
-
-
I'J
(.
6.0
'i
r,
c.
o
ID
60
patsiutztc^Kn
r>K.
rJc^M.
M-.
(OM fUW 100
OPtRATlNG MODE
205
3
i
i ' NOT TO BE USED FOR DESIGN.
; QUANTITIES SHOWN AXE FOR
i TEST PROGRAM EVALUATION.
\
1C,' APCL/A*!
BECHTEL
s • •, rxkciSCO
tl A. P.O. A
ALKAU scftuBftiMO TtsT rxctuiTv
TCX
SLOR.RX
>Y'jl CM
SHEET
~nr
^ 1 M- 109 • wo-S | o
ro
i
73
O
I
o
o
-------
I
en
1 ®.
MAKu OP
J_?.l 1©
w
.(«.
P8OCS5S W»,Tt«.
MOLO TA.KIK,
^^ j»-
)
(M)
do)
OouOMlTB
1
I
1
4-
r
(o
7
8
<5
10
1 1
12.
13
14
FiowfoPri] 4 ?i tonos
CM AX.
.'.PM '
r i .
/,-.-.:
•)•;;•
ii .".
•n.
75
'7
1
&
rA2-
500
if.
n
n
P ^
5
i.
c-
(r
f
I.
JO
c-
6J
•3/
0
O
io
lo
M.H.
C H '*".
.-,-V
Z-'::
-!'J
*'.
'-
0
17
J
4
^D
4-IZ.
7
17
n
^•i
C
.-, /
0-1
c...r
-
-
IS
o
i.0
r>;
0
o
IO
IS
VTWSSORECPSKn
rVK.
A*M.
MicJ.
OPttlATlNG MODE
iO Co I
. * NOT TO K USED FOR DESIGN.
« QUANTITIES SHOWN ARE FOR
' TEST PROGRAM E VALUATION.
BECHTEL
1A P, I • A K C I '. C 0
M A. PC. A
AI.KAU SCRUB & INI& TtsT OS.CILITY
TCA.
Si.uR.fcr FLOW SHEET
M- IO9 - Mi- (>
tn
CJ1
ro
i
70
O
I
O
O
-------
coouc
I
en
SCRUbBCB.
)
iCRU
u»
SOE
VB '
«. tf r.
rtwK
U-J
1®
P80CC3S WATtH.
"V FXew
Nt
4-
10
.70
IS
Go
/J
S
1,0
ion
1-10
I5Z.
0 ',
40
Mlfl
OPERATING MODE
50E
i • NOT TO BE USED FOR DESIGN.
I QUANTITIES SHOWN ARE FOR ;
I TEST PROGRAM EVALUATION.
I
BECHTEL
S » '. f • A n C I 5 C 0
A PC A
TCA. c,CHl EiiI•.;<", IViTtM
SutET
01
tn
i
en
o
ro
O
O
-------
01
O3
I
StURR-V
STMAH
NS.
\
Z
3
4-
£
t.
7
8
9
Fiow(on4]4 >i SOL 10 s
MAX.
/
ze
\ *,
c
o
to
TYPICAL
«.?M
/•;.-.-.
•i^o
i-
25
20
^3A
Soo
35-
1
•is
/4
/•*
/,'
o
2.8
<3->
O
C
bo
rti»»
f-pt^
''. -i :,
?.So
W
/>
4
4CO
4-42.
0
4-
%^
1
7
7
O
Ifi
a.t
-j
c
&o
PftessoatCPSKiT
**.
*t*.
Mitl.
OPERATING MODE
Zoft
I * NOT TO BE USED FOR DESIGN.
QUANTITIES SHOWN AXE FOR
TEST PROGRAM EVALUATION.
&.U HI*
BECHTEL
hi. A. PC. A,
TE.ST
TCA ^CW)-r .I'.HJG i
SLURR.Y FLOVM SHE.ST
M - '03 - i/.s -
-------
SUURR.Y O.ow DATA
STMAM
7
10
II
4'J
(.0
TYPICAL
76
53i
Boo
35
60
f-PM
+50
i-n
to
OPERATING MODE
7. e. O
j ' NOT TO Bt USED FOR DESIGN.
.1 QUANTITIES SHOWN ARE FOR
{ TEST fROORAM EVALUATION. :
Mid.
100
r.;> f:. .'.
BECHTEL
-, • i, r • « H c I s c o
ALKAU
A PC. A
ttsr
SHF,CT
M- 109 -
tn
en
ro
§
O
o
-------
I
CTt
I
cJ
w
K1 ^ v
1;
\
PLOODEO BCD
Sca.uuae.ti.
CD
SLA.KIWC, TK.
MAKfi UO
V*.TSR
u
,
@
© @*
PeOCtSS WKTtK.
HOLD TA.MK.
aeiLoiavt »
TAKJK.
>'!
v»
FT-OW DATA
SmMi
Ni
I
2
}
4
5
t,
7
&
o
9
10
II
12-
FIOW[OPK] £ >i sotios
MAX.
M
li
i'j
!';•
tU
Cl
Jo
0
t
IS
"Jo
to
TVPltAL
(irK\
-'t<~ '-
,-••
3 . '.'.
ICC
:: •'<:
?i
n
Goo
Z?
0*2.
n
S
'(•"*
c
t
<
/^
{,
2
30
O
0
01
3o
60
MlW
r-?r,
i---
a;
/ Vi.
IS
r,
o
n
4.12.
i
450
n
4-
.
'!•;•;'
^
C.L
r.. '_
3.'j
-
0
IS
c
0
o y
10
&0
PfmSSUBSfl'SKil
r/.K.
•;j
/
tJo*M.
— '
1
Mid.
10
1
<-IM6«TO>J» POUO
OPERATING MODE
J • NOT TO BE USED FOR DESIGN.
J QUANTITIES SHOWN ARE FO«
I TEST PROGRAM EVALUATION.
I
o fei Ari«i»vi.\.
BECHTEL
•nu r ««nclsc0
hJ. A. P C. A
LOOO6D P.i.o
SuuR.o.r
W-HO - Mi-
tn
tn
o
I
o
o
-------
PfiOCSSJ WATtR.
HOLO TA.MH
SLURRY fXow DATA
STMMl
Ni
1
T.
3
4-
S
(o
1
8
«
10
II
IT.
Pl6w[o?M]4 y. SOLiOS
MAX.
r,pri iy.s
40;
i r.-
^,-". '•-
v=c
iS-i
i!.U
«
810
?-SO
25
b
9
M
i'j
•;
"
irj
0.1
2C
O
3 i
0
^B
fco
TXPJCAU
Gi'^^
.•,/.--
IGO
•ii..
iOr.
? "t'
?i
n
5oo
SJo
Z5
5
8
•t«,
^
t,
L
b
4,
o
JC-
o
o.t
o
10
t>0
Mirf
f-FM
_.-)0
V-
1-1 i.
/ ;
0
o
/ 7
4«17
4Co
7
7.
4-
r ^
c
c.c
c.f>
: r.
-
o
>~
•> i
o
60
fee.
PttlSSUfcEtPSKn
r>x.
70
/
t««M.
70
1
MliJ.
/o
1
OPERATING MODE
)
NOI TO K USED FOK DESIGN.
QUANTITIES SHOWN ARE FOR
TEST PROCKAM EVALUATION.
BECHTEL
S A :. f • • « c I i C 0
. A PC A
ISO
SMCET
t**«i*a •• I «t
M- nO • MS-2. I O
01
01
ro
73
O
I
O
O
-------
ro
i
FUow
STIUAH
Nt
10
11
n-
Sotios
IMA*. TYPICAL
ID
i.0
Sod
14
n
>s
1C
10
411
n
(.c
IS
/a
/s
MM.
ftoeut i.'
j« NOT TO SE USED FOR DESIGN. ,
QUANTITIES SHOWN ARE FOR i
TEST PROGRAM EVALUATION. •
BECHTEL
S*'l IXNCItCO
»J A. PC. A
At-KAu SC«o6S(M6 TtST fAClUTY
SLURRY
Sneer
C.*155
M- no -
Ol
en
ro
•yo
O
i
o
o
-------
.,-®-
FuOOC-tC> BtO
stau^BtR.
\o
1 1
15
14
IS
SOLIDS
GPM %S
•t'j
Zo
t'Ju
tc.
Of,
Z.2
11,
5JO
IZ2.
3 £6
'.5
to
Mlti-
to
n
450
e.3
1,0
it
HILJZ.OV
loo
ro
OPE.RATING MOCE
304- I
* NOT TO »E USED FOR DESIGN.
QUANTITIES SHOWN ARE FOR ,
TEST PROGRAM EVALUATION.
A
!/u
CLH m.v
I
::.;>«! r-l-
BECHTEL
SAX f«««CISCO
»J A. PC. A
ACKAU SC«l.'B9iM6 T6.ST
FLOW SHEET
M- HO - M4-4
01
en
ro
i
73
O
I
O
-------
10
SLURRY
13
1*
15
Pf1
4CC
Jo
45
01
7.5
lo
60
n
c-t.
to
fill*-
n
415
10
n
02
03
10
60
ryx.
It
OPERATING MODE
NOT TO SE USED FOR DESIGN.
QUANTITIES SHOWN ARE FOR
TEST PROGRAM EVALUATION.
A
/O
-J
in
in
ro
I
•yo
o
O
o
KlUfcOY
BECHTEL
A. PC. A
TE.ST
rLOoDto Btti otuimaiiici SYV»«M
SLURJW FLOVAI SHEET
CIS 5
M- HO -
1
-------
SLURRY rv.ow DATA,
Oto
cr»
en
STftfcAM
Ni
1
1
1
-•1
4
5
U
7
8
9
10
1 1
\1
13
14
Flow [&PH] 4 yf SOL.DS
MAX.
uI-M _,
t
\r>
'•S
IS
IS
C-.l
JO
o
4-0
Q4
*
6
45
O
TYPICAL
Gi'M
.1. .-,
/CO
J "
I?'
•i-tr
Kl
n
o
f
53.0
S2t
O
14
i
"rs
5
b
«,
<.
6
C
in
c
00
o.^
r/
0
S
o
Ml»4
f-FM
J.50
/".o
:-;z
zo
c
r
n
a
4
4EO
446
S
/o
1
>is
0.3
s ;:
0.7
.:• i;
-
0
/s
o
6C.
e.j
c
c
ff
0
patvsoRt(P5»6l
r>x.
tl
i
^M.
15
1
Mill.
IO
1
lof-eHATiNG MODE
I 5 o fa I
I
{ • NOT TO BE USED FOR DESIGN.
< QUANTITIES SHOWN ARE FOR
4 TeST PROGRAM EVALUATION. :
»-. » »»\ %» *^\ V^%\xx\^^^\vNvX \^" _*
x«»\\\» v* .>»xx»..x»».o »%^» ,;o*
A
Ci*. HV
tiiUCO F:- APPIO/*1.
"Vf.a~
BECHTEL
I«»HCISCO
tJ A. PC. A
AuKAu SCA>.'69IM6 T6.ST
eo ato
SUUR.RT
SMEET
M- HO - t/.S- &
tn
tn
i
cr>
o
ro
o
0
-------
en
i
H-ow
y*t*n
N*
1
'L
3
4
5
t,
1
8
9
to
\\
FLOW (am] £ Ji sotios
r-irfm.
GPM i% i
:.co
i«o
(.!> '
3 ;."
;'•'-'
'fl'
P:<
* £'-
Z?
45
a
•4
;.;
,b
/^,
2 /
Jo
e-3
c
a
^u
6r
TYPICAL
<.i-M
/io
<-r
,TO
i«r
e:
77
^2
5co
2-5
n
E
>s
S
r
t,
6
c
js
c> j
&
o
lo
60
M(»<
C-PM
/»<:
JOB
tC
c
0
r;
4 fro
4IZ
7
n
4
^i
0
c>
O *j
-
-
;S
£.--
e
o
10
to
PfUiSufcEtfSKT]
r>x.
7.
19
tJoRM.
2.
IS
Mill.
Z
/o
OPE.RATING MODE
30-; I
1 * NOT TO K USED FOK DESIGN.
QUANTITIES SHOWN ARE FO»
HST ntOCRAM EVALUATION.
BECHTEL
rJ A. PC. A
AcKAU SCRU6SIK16 T6.ST
F~UOOC>6O tt-O
SuuR.R.r FLOW
HO
Ul
en
ro
i
73
O
I
O
O
-------
J
FLOODED
StlU&OtB.
MMIMBMMM^K
®
1_
^^
in
S4At
V
kic,
\
\
x
r*.
0
/
Mfc
y.
1 THICKtUfeH.
^-^^^
^•^^•••M
•••••
1
At
•^
^^
n
r*
1
^•i
s*
-H
.U
KJ
r
>—
^
{
T
««
K.
-I
J
7
(T
V
•— -1 '
^_J
®
... 1 ^i
i
M**KI
• •• " /^s"i i ,- v
Cf.'l 4-")
process WATS*.
HOt. ft TAMK.
© '
I .UP
(fi
"3
•»M ' LiMCSToMfc PQMO
)
:J-
j '
SLURK.V R.OW DATA ->^-
Ni
l
Z
3
A
5
L
7
6
9
10
i i
\1
\ ':
K
FLOW [OPM] i y» SOLIOS
MAX.
Zco
•',cc
iCC
If.
i'~-
*5
*2
(0
3:c
810
20
,
5?i
51o
M
• r-
'5
:ii
O.I
20
o ;
iw
Ofi
0
0
0
7.2
O
TYPICAL
f.c
.Jco
,":r.
;:•>.•
33
/7
7
3
52»o
I30
/e.
o. 1
212
W
5
5
fe
<.
<•
io
O
to
*.c
o
Ci
f\
1.5
0
'.?M v. <;
IOC
ioo
zo
o
o
17
8
.;.
teO
K4
7
O
212
3l'J
-------
J,
C
<§)
SCURRY FLow O*TA> -f-
^ ^
Tl«l-i
"^,
Y
a
\
-f
>\
D
. /"
1*
rl
/
Tfc*..
». >- U.f*.
^-"
/
_vhr
1
•_t«.&L<-< m •
©
7
-, ®
J
C')|
7
(10)
" 7
" v.
t.r- >
I',
l-l'.K.
CATAUVST
ls.'/kTB<»
peaces WATEK.
MOLf> TA(0>1
fj
1
••'
•r
C
I T
STRtA-M I
Nt j
•^••••l^rvaWH
1 ?
b
4
5
t,
V
s
^
10
11
IZ
1?
rio«[G?r-iU y- s°>-
MAs.
c: f •''. ;
?•-:
i ;•
.; •
•','
-jr.
/y
-.',,
ff
JZ
jr
10
:'.
'•
.-i ,•
C.
L H
'••Z
T
•">
O
to
7YPi0
«
i
R
i
t-
.-.
L
-S-
c.
40
it.
o
o
o
^n
.OS
Mtfl
f-rri
.; .. i .
C-
'••
. :
; ;
--
j-
f
4EO
44 Z
7
(
4-
1
OPf-RATlNG MODE
4OT TO SE USED FOR OK\Ct
3UANTITIES SHOWN ARE FO
EST PROGRAM EVALUATION
1
* •
'.
L
~7
^ r
11 3*
0 I
-
C
IS
:
i-
'..-.
o
o
C
£o
VtKisoat[P-..il
^x.
V:
IrtlM.
ft'
T***_
«...
••-
• "•*• -
(•«t' », ^
11' .IEO Ft'.'. If", r. M '
\-
». _
••^M
1
4-un
trH
TiTi^i !
I...-;
BECHTEL
i • •! r « « x ci -. CO
rJ A. P
AtKAU 5C«ob^»-
1-'
LOOULO at u s
(Sff.
mw
^155
C A
»& Tt$r rxcu-iTv
CCi.xfc\i-:Ci '.Y'.Ti M
• tow 5waeT
M>«M «. M.
M- I 13 - I/. 5. -9 •-
en
01
ro
i
§
i
o
o
-------
17551-6012-RO-OO
APPENDIX D
DETAILED ASSESSMENT OF X-RAY ANALYTICAL METHODS
D.I EVALUATION AND RECOMMENDATION OF X-RAY INSTRUMENTATION
As a result of the studies performed on this program, the Applied
Research Laboratories (ARL) units were recommended for both laboratory and
on-line use. The details of this review are presented below under the four
following categories:
• Cost review of suitable X-ray instrumentation,
• Evaluation of technical capability of instrumentation,
t Manpower requirements for operation, and
t Preliminary specifications for X-ray instrumentation.
D.I.I Cost Review of Suitable X-ray Instrumentation
Wavelength Dispersive Type Spectrometer •
• Applied Research Laboratories (A.R.L.)
Model PCXQ 44000 (for on-line analysis of slurries)
Instrument Cost: a) Single stream without a computer,
$70K
b) Fifteen streams without a computer,
$100-110K
c) Fifteen streams with a computer,
$150-160K
Computer Programming: $6K for 6 man weeks plus $0.2K per
day as required
The PCXQ 44000 is the only instrument seen to date that could
serve as an on-line analyzer for liquid slurries. The prices
above are for units with a scanning spectrometer, six fixed
spectrometers, and an external standard. The instrument oper-
ates with a helium x-ray path since elements lighter than
manganese will be included in the analytical program.
Model VXQ 25000 with 46000 Console (LAB Unit)
Instrument Cost: a) ^$60K with computer capability
b) -490K includes computer (HP 2114C)
-169-
-------
17551-6012-RO-OO
The VXQ 25000 laboratory unit priced above would include a
scanning spectrometer, six fixed spectrometers, and an
external standard. This instrument can operate in the
manual or automatic mode in helium or vacuum.
Model VXQ 25000 with 245000 Console (LAB Unit)
Instrument Cost: a) Single scanning spectrometer, $41K
b) Up to 4 fixed spectrometers, add
$1.5K each
c) Six fixed spectrometers installed,
add $2.2K each
The VXQ 25000 with the 245000 console includes a scanning
spectrometer, a variable number of fixed spectrometers, and
an external standard. It is a unit that can be operated
manually or automatically. Readout is in direct %
concentrations on a single-sheet hard copy.
Purchase of any of the A.R.L. instruments entitles one
person to go to a tuition-free X-ray school for one week
at A.R.L. Also, training is provided during the plant
checkout and installation of the equipment.
• General Electric (G.E.)
Model XRD-710 (Automatic Lab Unit)
Instrument Cost: ^$90K including computer
The XRD-710 instrument is a fully automated, solid state
unit capable of sequentially analyzing 10 specimens for
49 elements from F upwards. Operation is in vacuum or air.
Model XRD 700VS (Manual Lab Unit)
Instrument Cost: a) ^$32K with a four specimen sample
holder
b) ^$41K with a ten specimen sample
holder
The XRD 700VS is strictly a manual unit operable in air or
vacuum.
-17C-
-------
17551-6012-RO-OO
Energy Dispersive Type Spectrometer
t Kevex Corporation
Model - Vacuum System (Lab Unit)
Instrument Cost: a) -\422K without a computer
b) ^$35K with a computer
The Kevex instrument has limited detectability and probably
the least versatile instrument of those evaluated.
D.I.2 Evaluation of Technical Capability of Instrumentation - Instru-
mentation provided by the ARL, GE and Kevex vendors has been evaluated for
technical capability to satisfy process monitoring requirements of the wet
limestone injection sulfur dioxide abatement process. Key points in this
evaluation which distinguish the instrumentation from each other are
presented below.
General Electric Instrumentation - The evaluation of the G.E.
instrumentation was accomplished through discussions with applications
personnel and assessment of the technical ability of the instrument for
the job. TRW laboratory specimens were not used because of a long lead
time necessary to acquire experimental data. Four G.E. wavelength
dispersive X-ray spectrometer systems were evaluated, 1) XEG - X-ray
Emission Gage I for computerized process control, 2) 700VS manual laboratory
unit, 3) XRD-410 vacuum tube automated laboratory unit, and 4) XRD-710
solid state automated laboratory unit. Distinguishing features of these
instruments are:
• The XEG unit was designed for dry powders only. Two
slurry units were sold by G.E. and they are a constant
problem.
• A slurry presenter is not feasible with any of the G.E.
units.
• G.E. has no automated equipment to convert slurries to
briquettes. In fact, they stated that an on-line slurry
dryer would not result in an efficient operation.
• All G.E. units discussed utilize the more inefficient flat
diffracting crystals as opposed to the curved diffracting
crystals used in Applied Research Laboratory Equipment.
-171-
-------
17551-6012-RO-OO
In summary, no G.E. unit can be utilized for on-line process control
of wet scrubbing process slurries. Their instrumentation should, however,
be considered for use in a laboratory environment.
Kevex Instrumentation - These comments are the results of
discussions with Mr. Richard S. Frankel, President, and Dr. Rolf Woldseth,
Applications Lab Director on 7, 19, 20 and 24 May 1971 and on 9 June 1971.
The data presented (see Table VIII - Section 2.4) were obtained by Dr.
Woldseth on the same TRW prepared specimens which were analyzed previously
by Applied Research Laboratories. The following points were developed from
the contacts made with Kevex Corporation:
• There is a basic problem with regard to the detection of low
levels of S in a matrix containing a high level of Ca with
energy dispersive X-ray spectrometer systems such as those
sold by Kevex. That basic problem is that the CaKe escape
peak of 2.27 KeV interferes directly with the SKcq peak at
2.27 Kev. This escape peak phenomenon occurs in the Si(Li)
X-ray detector when 4.01 KeV CakBi quanta ionize the Si and
cause 1.74 KeV SiKai X-rays to escape from the detector and
leave 2.27 KeV quanta to be counted. The effect of the CaKp
escape peak was demonstrated at Kevex with TRW specimen 009,
limestone/Zurn FA, containing 0.045 + 0.015% w/w S by wet
chemical analysis at TRW. The sulfur determination data on
specimen 009 were as follows for a 100 second count:
Gross count 655
Background count 510
Net count 145
Ca Escape Peak count 130
The 145 net counts and 130 calcium escape peak counts are well
within the counting statistical error of approximately 30
counts. The conclusion was that in this case the Kevex unit
used was not capable of detecting sulfur down to the 0.03 - 0.06%
w/w level. The vacuum path Kevex unit could probably detect
S in a limestone matrix to the 0.3 - 0.5 ±0.1% w/w level, i.e.,
a factor of 10 to 20 times worse than the ARL 72000 Quantometer.
The sensitivity of the Kevex instrument could be improved by
a factor of > 5 by counting for a longer time than 100
seconds, using a 50pCi source rather than the 25yCi Fe used,
and using an 80 mm detector instead of the 30 mm detector used.
• As in the case of wavelength dispersive X-ray spectrometers
(G.E. and A.R.L. instruments) the output signal is influenced
by specimen particle size. Particles sized below ^250 microns
can usually be analyzed accurately by X-ray instrumentation.
-172-
-------
17551-6012-RO-OO
• Kevex Corporation has no slurry presenter equipment.
t Kevex Corporation has no computer software available. All
computer interfacing must be with their supplier, Nuclear
Data, Inc.
• The A.R.L. unit detected Mg known to be present in more than
half of the 14 TRW specimens submitted; the Kevex unit
detected no Mg.
In summary, no Kevex unit, like G.E., can be used for on-line control
of wet scrubbing process slurries. In addition, their instrumentation is
somewhat limited even in a laboratory environment and does not compete
well with either the A.R.L. or G.E. units considered.
Applied Research Laboratory X-ray Instrumentation - The
evaluation of the A.R.L. X-ray instrumentation was based on results of
TRW prepared specimens. The A.R.L. wavelength dispersive X-ray spectro-
meter instrumentation has several advantages over G.E. and Kevex equipment.
These advantages are:
• Curved diffracting crystals are used which are more efficient
than flat crystals used by G.E., i.e., their use results in
higher count rates, thereby allowing shorter count times than
with flat crystals.
• Up to nine elements can be determined simultaneously since
A.R.L. has capability for nine fixed spectrometers.
• An external standard in A.R.L. instruments automatically
corrects for changes in X-ray tube voltage.
• A.R.L. has the only available and proven slurry presenter
apparatus.
t A.R.L. has a patented slurry density gauge as an integral
part of their on-line equipment.
-173-
-------
17551-6012-RO-OO
X-ray Instrument Availability
A.R.L. 90-120 days without computer
120 - 150 days with computer
G.E. 90 - 120 days
Kevex 60 days
D.I.3 Manpower Requirements for Operation - The manpower requirements
listed here are approximations based on experience and discussions with
the three vendors. Initially, perhaps for the first three months of
operation, any X-ray instrument will require one full time engineer or
scientist and one technician per shift. After the initial period, engineer
time should decrease to approximately 50% over perhaps the next nine months
of operation; technician time should remain at one man per shift. In
fact, 50% of an engineer's time for one shift per day should be adequate
after a year's operation; the other shifts could be run with one technician.
With computerized data acquisition and reduction equipment, it is
estimated that the following maximum number of analyses and elements per
analysis could be obtained for the four X-ray equipment operational modes:
Estimated Maximum Number of Total
Number of Analyses Elements Elemental
X-Ray Unit per 8-hour Shift Per Analysis Analyses
A.R.L. (on line) 240a 8 ^1800
A.R.L. (lab) 160a 7 MOOO
G.E. (lab) 160a 1 M60
Kevex (lab) 45b -^50 (16) ^640
aTwo-minute residence period in spectrometer
Ten-minute residence period in spectrometer
Unfortunately, of the 50 elements possible with the Kevex unit, only
about 16 are of primary interest to the limestone injection wet scrubbing
sulfur dioxide abatement process.
-174-
-------
Using the above estimates, the number of elemental analyses per 8-hour
shift, taking equipment maintenance requirements (5% down time) into account,
are estimated as follows:
A.R.L. (on-line) ^1800 elemental analyses
A.R.L. (lab) ^1000 elemental analyses
G.E. (lab) ^160 elemental analyses
Kevex (lab) ^640 elemental analyses
It is emphasized that these numbers are maxima and were calculated
under the assumption that the operation was quite routine and well-trained
personnel had been working the problem for about one year.
In addition, the calculations assume that there is a plentiful supply
of separated or non-separated slurry for analysis to ensure that the
maximum number of analyses per 8-hour shift is maintained. This assumption
can lead to high labor costs for acquisition of samples, separation, and
transfer to the laboratory while maintaining documented inventory and
clean-up prior to the next set of samples.
For the purpose of providing a cost of operation estimation, it is
assumed that five sampling points on each of the three scrubber units are
to be sampled every 30 minutes (total 30 samples/hour). Similar calcula-
tions can be made for other combinations resulting in 30 samples/hour rate
(e.g., 6 sampling points on one scrubber every 12 minutes). A summary of
the mechanical operations to be performed by technician labor are as
fol1ows:
Estimated Time Required,
Operation Minutes
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Install grab sampler unit on scrubber
Take sample*
Remove grab sample*
Perform separation
Take sample to laboratory
Log sample with identification number
Load sample into X-ray carousel sample
compartment
Remove sample
Clean grab sampler
Take sampler to scrubber
3
2
2
2
6
1
2
2
2
6
28
*-!hen Demco/fliter combination is used, these numbers are reduced but
not eliminated because it is necessary to purge the system with the
stream.
-175-
-------
17551-6012-RO-OO
Although the total time is 28 minutes/analysis, several operations
(such as Steps 5, 9, and 10) can be combined with other samples and, hence,
it is estimated that 8 to 9 technicians can provide an adequate number of
samples providing the operations can be programmed so that the technicians
do not run into each other. With the 10-minute residence time requirement
of the Kevex, fewer samples are required (6 per hour) and, hence, only two
technicians are required for sampling. A summary of estimated costs for
operation and acquisition of X-ray instrumentation is shown in Table D-I.
Costs required for computer operation are not included in these numbers,
however, it is readily seen that the costs associated with the two A.R.L.
units provide more cost effective operation.
D.I.4 Preliminary Specification for X-ray Instrumentation - The
instrument sensitivity standardization computer interfacing and minimum
required computer software for X-ray instrumentation suitable for both
on-line and laboratory use are comparable. Specific requirements for these
two operation modes have unique requirements in number of spectrometers and
analytical speed. The general and specific specification requirements are
presented below.
General Specification
Sensitivity and Precision
The instrument will be able to detect all elements above atomic
number 11 in concentrations down to 0.1 % w/w. Water slurries
with 1, 5 and 10% w/w solids (limestone and dolomite matrices)
will be analyzed to demonstrate specific sensitivity to Mg,
Al, Si, P, S, Cl, K, Ca, Ti, V, Mn, and Fe in both the liquid
and solid phases. The precision must be at least 2% of the
measured value.
Computer Interfacing
The instrument will interface with a computer supplied as part
of the system. Output will be typewritten hard copy.
Minimum Required Software
Computer software is required to 1) control the sequence and
functioning of all devices in the system, 2) convert elemental
-176-
-------
TABLE D-I
SUMMARY OF ESTIMATED COSTS ASSOCIATED WITH OPERATIONS AND ACQUISITION X-RAY INSTRUMENTATIONS
Operational Cost/8-Hour Shift
X-ray Unit
A.R.L. (on-line)
A.R.L. (lab)
L G.E. (lab)
1 Kevex (lab)
Capital
Cost
$K
160
90
90
35
X-ray
MRS
Hrs.
8
8
8
8
Unit
TS
Hrs.
8
8
8
8
Cost
$a
240
240
240
240
Sampling
TS
Hrs.
8
72
72
16
Cost
$a
80
720
720
160
Total
Cost
$
320
960
960
400
Total
Analyses
240
160
160
45
Total
Elements
1800
1000
160
640
Cost/
Analysis
$
1.33
6.00
6.00
8.88
Cost/
Element
$
0.18
0.96
6.00
0.62
Calculated assuming a burdened labor cost for professional employees (MRS) of $20/hour and
technicians (TS) of $10/hour.
tn
i
-------
17551-6012-RO-OO
count or voltage ratio output to percent concentration,
3) determine up to six interelement correction constants,
4) apply interelement correction constants to compensate for
interelement effects, and 5) determine working curves from
standard samples with graphic print out.
Specific Specifications
On-Line System Description
The on-line X-ray instrumentation will provide elemental
composition analyses of flowing slurries in a minimum of
three streams. Elemental readout will be typewritten in
percent concentration. Elements with atomic numbers greater
than 11 will be able to be analyzed automatically. There
will also be provisions to analyze dry and wet batch
specimens.
Analytical Speed
Approximately 1800 elemental analyses with elemental
percent concentration output on hard copy must be
feasible during an 8-hour shift. (Manpower: 1 laboratory
technician).
Number of Spectrometers
There will be one scanning spectrometer and six fixed
spectrometers for Mg, S, Cl, K, Ca, and Fe.
Instrument Standardization
Means will be provided to correct for minor variations
in the X-ray tube power output during analysis.
Slurry Flowrate
The slurry presenter system will maintain a constant flow
through the analyzer cell of 5 liters/minute.
Slurry Density
A means will be provided to determine the percent solids
in the flowing slurry.
Laboratory System Description
The laboratory X-ray instrumentation will provide elemental
composition batch analyses of slurries and dry solids. Elemental
readout will be on hard copy in percent concentration. Elements
with atomic numbers greater than 11 will be able to be analyzed
using a vacuum or helium X-ray path.
-178-
-------
17551-6012-RO-OO
Analytical Speed
Approximately 1000 elemental analyses with elemental
percent concentration output on hard copy must be
feasible during an 8-hour shift. (Manpower: 1 laboratory
technician).
Number of Spectrometers
There will be one scanning spectrometer.
instrument Standardization
Means will be provided to standardize the instrument.
Summary - In conclusion, it must be reiterated that the A.R.L.
on-line X-ray unit is the most cost effective and pays for its capital
cost with respect to the second most cost effective unit (A.R.L. lab) after
only 105 8-hour shifts (using the assumptions employed to generate Table VI).
Consequently, this on-line A.R.L. X-ray unit is tentatively recommended for
process characterization of the limestone wet scrubbing sulfur dioxide
abatement process. Should it be decided that a laboratory unit be employed,
it is recommended A.R.L. lab instrument be purchased.
D.2 X-RAY FLUORESCENCE CHARACTERIZATION OF ACTUAL AilD SIMULATED SOLIDS AND
LIQUIDS FROM POWER PLANTS
D.2.1 TVA Samples
The five liquid samples and six solid samples submitted by J. Barkley
of TVA were analyzed quantitatively by x-ray fluorescence techniques
employing TRW's G.E. Model XRD-5 laboratory x-ray spectrometer. It is
understood that, 1) these samples were intended to simulate a wide range
of sample compositions, both liquid and solid phases, that could be encoun-
tered under varying wet scrubber operating conditions and, 2) these
analytical results would provide additional information to aid TVA in their
efforts to evaluate candidate analysis techniques.
In order to eliminate the effect of disparate particle size between
standards and the unknown material, both were pulverized further and passed
through a 325 mesh sieve (>44 micron particle size). In previous studies
it was found that fluorescence intensity increased asymptotically with
decreasing particle size reaching a plateau in the range of 70-100 microns.
-179-
-------
17551-6012-RO-OO
The unknown samples were blended in a Wig-L-Bug and pellet!zed under
25,000 psi into a 1.5 in. diameter disc in a polyvinyl acetate (^0.1% wt)
binder. Calibration curves for Ca, K, Cl, S, Fe were prepared using reagent
grade anhydrous CaC03, K2C03, Nad, CaS04 and iron powder in a simulated
fly ash/SOp reacted limestone background composed of SiO^ and AlgOg-
The solution standards were prepared from reagent grade Ca(N03)2.4H20,
NaoSO., KC1 and NaCl. The special liquid cell with Mylar window was
employed for liquid analysis. Instrument operating settings were the same
for both liquids and solids, as follows:
Tube - Cr operated at 50 KVP, 30 Ma
Crystal - PET
Purge gas - He
Detector - flow proportional
Counting time - 10 or 100 sec. depending upon elemental
concentrations
The taken and measured concentrations for Ca, K, Cl, S and Fe together
with a comparison of TVA's reported pH values to TRW's measured pH values
are listed in Table D.II. The taken element values for the solid samples
were derived from the compound blends defined by TVA which are given in
Table D.III. For the liquid samples, the agreement between taken and found
for Ca and K is excellent while the chloride values are somewhat more
divergent. XRF sensitivity to sulfur, as discussed in several previous
reports, is below the concentration made up. It is quite apparent that
solutions 2, 3 and 4 have changed on standing as evidenced by the marked
differences in pH.
The utility of XRF for this specific application 1s exemplified by the
findings of the solids analysis. Upon superficial examination, the agree-
ment for Ca and S values appears rather poor. Closer study of the calcium
data reveals that the measured concentrations are consistently higher than
the reported taken values. The differences between the individual pairs
are, however, proportional to the ash content, i.e., (text continued on
Page 183)
-180-
-------
TABLE D.I I
i
oo
QUANTITATIVE X-RAY ANALYSIS OF LIQUID AND SOLID SCRUBBER SAMPLES FROM TVA
TVA Reported Values
TRW Measured Concentrations
Liquid
Samples
#1
#2
#3
#4
#5
Solid
Samples
#1
#2
#3
Hold Tank
Outlet
Clarifier
Bottom ppt.
Clarifier
Inlet
Ca
5.76
3.78
5.84
3.93
3.39
Caa
10.5
29.5
24.0
-
-
-
K Cl
2.26 8.07
2.44 11.6
2.71 15.1
2.53 12.24
1.63 11.14
K Cl Sa
4.9
10.0
8.6
- -
_
_
Molality(X10
S (PH)
5.83 (6.8)
5.48 (5)
7.86 (6)
5.12 (7)
4.84 (7)
% By Weight
£§. (Ash)
(64.95)
(3.99)
(30.1)
-
-
-
)
Ca
5.6
4.0
5.5
4.0
3.4
Ca.
16.1
30.1
26.1
7.2
15.0
30.0
K_ C_l_
2.1 10.4
2.4 15
2.5 14
2.7 14
1.6 9.6
K Cl_ S^
1.2 <.002 6.5
<.2 <.002 12.7
0.7 <.002 10.9
1.3 .008 4.7
0.9 <.002 8.6
<.2 .009 12.2
S^
<9.4
<9.4
<9.4
<9.4
<9.4
Fe
16.5
.9
7.1
12.4
13.2
1.7
aSee Table D.III for prepared composition by compound
(PH)
(6.7)
(3.2)
(3.3)
(5.7)
(6.9)
in
in
o
rv>
o
o
-------
en
en
00
TO
CaS03'l/2H20
CaCO,
Ash
CaO
MgC03
Total:
TVA
Sample No.
Wt % Ca
25.00 5.81
0.91 0.28
1.59 0.64
64.95
5.27 3.76
1.00
98.72 10.49
1ADL.E, U , 1 1 1
SOLID SAMPLE COMPOSITION BY COMPOUND
1
S Wt %
4.65 51.04
0.22 2.05
36.27
3.99
3.78
2.98
4.87 100.11
Sample No. 2
Ca S
11.87 9.50
0.64 0.51
14.50
-
2.70
-
29.71 10.01
Sample No. 3
Wt % Ca S
40.07 9.32 7.46
4.84 1.50 1.20
12.00 4.80
30.08
11.74 8.39
0.89
99.62 24.01 8.66
i
CTi
O
ro
TO
0
o
o
-------
17551-6012-RO-OO
Sample
No. 2
No. 3
No. 1
Taken
29.5
24.0
10.5
Calcium
Found
30.1
26.1
16.1
Absolute
A
0.6
2.1
5.6
Ash
3.99
30.1
64.95
Relative Error
Corrected for
Ash Ca
Contribution
1.0%
1.5%
0.62
Indeed, calculation of the least squares straight line equation using:
Y = fraction of total Ca
X = fraction of ash
gives; intercept, a = 0.317847 o = 0.000263
slope, b = -0.232775 a = 0.016217
and
X • 1
0.0399 0.30856
0.301 0.24778
0.650 0.16666
1.000 0.08507
Even considering the inhomogeneity of the added fly ash, the precision is
good yielding a calculated calcium content 1n the fly ash of 8.51%.
The sulfur disparities were more of an enigma, as they did not lend
themselves to easy explanation. Again, the concentrations found were con-
sistently higher than the taken levels but the differences were inversely
proportional to the ash content. A masking or antagonistic interference
was considered but discarded because of the lack of consistency or correla-
tion with the constituent levels. Assuming the analyses to be correct,
it was next postulated that, rather than an experimental weighing error
made in preparing the samples, another sulfate salt was used. Recalculation
of the sulfur taken on the basis of anhydrous calcium sulfate gave the
following correlation.
CaS04(%S) + CaS03-l/2 HpO(%S) = New Taken %S % Found
No. 1 (5.86M0.22) = 6.08 6.5
No. 2 (11.97M0.51) = 12.48 12.7
No. 3 (9.40M1.20) = 10.60 10.6
-183-
-------
17551-6012-RO-OO
These data strongly suggest that the anhydrous sulfate salt was used as op-
posed to the reported dihydrate. Most importantly, the XRF tool is suffi-
ciently accurate to allow elucidation of such anomalies.
D.3 DUKE POWER COMPANY SAMPLES
Several filter and impinger samples were submitted through the Project
Officer for characterization by XRF and other techniques. The sample de-
signations, elements determined by XRF and concentration values measured
are listed below. Whereas the emphasis in this program has been directed
toward Ca, Mg and S, note the utility for other pollutant species that are
of concern in terms of human health, crop and material damage notably
silicon phosphorous and chloride.
Element. % by Weight
Sample Cl_ S_ SJ_ £
"B" Filter Paper <0.1 3.7 16.5 0.65
Run 1-B
Probe-Cyclone Acetone 2.2 1.8 12.7 0.42
Filter #12376 <0.1 1.0 13.5 0.35
Number 44 0.64 2.3 14.7 0.54
-184-
-------
17551-6012-RO-OO
APPENDIX E
TENTATIVE METHOD FOR ANALYSIS FOR SULFITE AND BISULFITE
ION BY FURFURAL BLEACHING
(A Laboratory Instrumental Method)
SCOPE
This method is applicable for the analysis of sulfite and bisulfite
ion in aqueous media. While the method was developed especially for sul-
fur dioxide scrubbing processes, it has general applicability within the
given ranges.
SUMMARY
An amount of sulfite and bisulfite ion adjusted to be in the concen-
tration range of 1-5 mM is added to a measure excess of furfural. A none
UV-absorbing complex is formed and from a measurement of the decrease in
absorbance of the free furfural species at 276 nM and pH 4, the bisulfite
concentration can be calculated utilizing standard spectrophotometric
calibration techniques.
REAGENTS
1) Furfural Solution, Aqueous, approximately 2.5 x 10~4 M
Dissolve 24 mg +_ 2 mg in deionized water which has been
boiled to remove oxygen and cooled. Dilute to one liter in
a volumetric flask. This reagent should be made fresh daily
and used for preparation of a standard curve using standard
known bisulfite solutions.
2) NaH2P04 Buffer Solution
Dissolve 120.0 grams of reagent grade NaH PO in 1 liter
of boiled, cooled, deionized water. Adjust pH to 4.0 by drop-
wise additions of concentrated H P04. This solution is stable
and does not have to be freshly made.
3. Aqueous Sulfamic Acid
Dissolve 0.486 g of reagent grade NH2S03H in 200 ml of
boiled, cooled, deionized water.
NOTE: Minimize air exposure of all reagents.
-185-
-------
17551 -601 2-RO-OO
REQUIRED EQUIPMENT
1. Beckman DK-2A ultraviolet-visible spectrophotometer or
equivalent.
2. Matched silica sample cells with an optical path length of
10 mm.
PREPARATION OF STANDARD CALIBRATION CURVE
The separation of the standard bisulfite solutions and absorbance
curve should be carried out each day that samples are analyzed. Weigh a
quantity of reagent grade sodium bisulfite to be determined by the formula:
Grams NaHS03 =
where %S02 = %S02 in the NaHS03 (SOp content is the normally
reported assay value for NaHS03).
Quantitatively transfer the NaHS03 to a 500-ml volumetric flask, add
cooled, boiled deionized water to dissolve the salt and fill to the mark.
This yields a 5 mM solution (5 x 10" M) . Pipet 25 ml of this solution
into a 50 ml volumetric flask and 25 ml into a 100 ml volumetric flask.
Dilute each to the mark to yield solutions of 2.5 mM and 1.25 mM bisul-
fite, respectively. Pipet 5 ml of each of the three calibration solutions
into 25 ml volumetric flasks. Using a pipet, add 5 ml of the NaHpPO.
buffer solution, 5 ml of the furfural solution, and 1 ml of sulfamic acid
solution to each flask and also to an empty 25 ml volumetric flask (cali-
bration standard with zero bisulfite content). Dilute to volume with water
and mix well. With the instrument in the transmittance mode, adjust the
zero and 100% T on the UV spectrophotometer with a solution of 5 ml of
buffer and 1 ml sulfamic acid solution diluted to 25 ml with water (a
blank) placed in both sample and reference cells. Change the instrument
to the absorbance mode. Allow the calibration standards to stand at least
5 minutes but less than 20 minutes after preparation and then scan each of
the four calibration solutions from 340nM to 260nM while retaining the
blank solution in the reference cell.
Measure the absorbance (A) of the peak at 276 nM and calculate the
reciprocal of absorbance (I/A). Plot I/A (ordinate) vs the bisulfite con-
centration in millimoles per liter (abscissa) to yield a straight line
-186-
-------
17551-6012-RO-OO
function of the type y = a + b. Determine ordinate intercept (b) and the
slope (a) of the function by a least squares calculation.
ANALYSIS OF SAMPLES
The samples must be free of particulate suspension and contact with
the atmosphere should be minimized. The samples should have a bisulfite
concentration of 1 - 5 mtt. If upon analysis the level is too high, dilute
the sample appropriately using quantitative techniques. If filtering or
dilution is necessary, it should be done in an inert atmosphere dry bag or
glove box. The samples should have an initial pH between 5 and 9.
Pi pet 5 ml of the sample into a 25 ml volumetric flask. Add 5 ml of
the NaH2P04 buffer solution, 5 ml of the furfural solution, and 1 ml of the
sulfamic acid solution. Add water to the mark and mix well. Analyze all
the solutions using the method in the calibration section.
CALCULATIONS
Calculate the concentration of total sulfite plus bisulfite (SI7) in
each solution according to the formula:
R 1
Siv concentration, mM = 5 £• (TT)
where:
B = ordinate intercept of calibration straight line curve (A )
Y = slope of calibration straight line curve (A"1 m M°les -11
-187-
-------
17551-6012-RO-OO
APPENDIX F
DIRECTORY OF SEPARATOR MANUFACTURERS AND VENDORS
Ace Scientific Supply Co., Inc.
1420 E. Linden Ave.
Linden, N.J. 07036
Air Products & Chemical, Inc.
P. 0. Box 538
Allentown, Pa. 18105
Allied Engineering & Production Corp.
2421 Blanding Ave.
Alameda, Calif. 94501
American Air Filter Co., Inc.
215 Central Ave.
Louisville, Ky. 40208
AMF Cuno Div.
American Machine & Foundry Co.
400 Research Parkway
Meriden, Conn. 06450
Ami con Corp.
Scientific Systems Div.
21 Hartwell Ave.
Lexington, Mass. 02173
Barnstead Co.
225 Rivermoor St.
Boston, Mass. 02132
Beaver Filter Corp.
P. 0. Box 848
Port Ewen, N.Y. 12466
Belleville Wire Cloth Co. Inc.
135 Little St.
Belleville, N.J. 07109
The Bittner Corp.
181 Hudson St.
New York, N.Y. 10013
The Carborundum Co.
Graphite Product Div.
P. 0. Box 577
Niagara Falls, N.Y. 14302
Chemical Equipment Corp.
7454 E. 46th St.
Tulsa, Okla. 74145
The Chemical Rubber Co.
18901 Cranwood Parkway
Cleveland, Ohio 44128
Clay Adams
Div. of Becton, Dickinson & Co.
299 Webro Rd.
Parsippany, N.J. 07054
Columbia Filter Co. Inc.
199 - 7th Ave.
Hawthorne, N.J. 07507
The DeLaval Separator Co.
350 Dutchess Turnpike
Poughkeepsie, N.Y. 12602
The Dow Chemical Co.
P. 0. Box 1656
Indianapolis, Ind. 46206
The Duriron Co, Inc.
452 N. Findlay St.
Dayton, Ohio 45401
Eagle-Pitcher Industries, Inc.
American Bldg.
Cincinnati, Ohio 45202
Ertel Engineering Co.
62 Front St.
Kingston, N.Y. 12401
Filtros Plant-Ferro Corp.
601 W. Commercial St.
E. Rochester, N.Y. 14445
Fisher Scientific Co.
711 Forbes Ave.
Pittsburgh, Pa. 15219
General Nuclear Corp.
550 Fifth Ave.
New York, N.Y. 10036
Graham Manufacturing Co., Inc.
26 Harvester Ave.
Batavia, N.Y. 14020
-189-
-------
17551-6012-RO-OO
APPENDIX F (CONTINUED)
Heico, Inc.
Delaware Water Gap, Pa. 18327
F. R. Hermann & Co., Inc.
P. 0. Box 229
Mill dale, Conn. 06467
International Equipment Co.
300 Second Ave.
Needham Heights, Mass. 02194
International Sales Associates
116 N. Bellevue Ave.
Langhorne, Pa. 19047
Johns-Manville Products Corp.
22 E. 40th St.
New fcrk, N.Y. 10016
Kern Chemical Corp.
854 S. Robertson Blvd.
Los Angeles, Calif. 90035
Komline-Sanderson Engineering Corp.
100 Holland Ave.
Peapack, N.J. 07977
Millipore Corp.
Ashby Rd.
Bedford, Mass. 01730
MSE, Inc.
811 Sharon Dr.
Westlake, Ohio 44145
Norton Co.
Plastics & Synthetics Div.
P. 0. Box 350
Akron, Ohio 44309
Owens-Illinois, Inc.
Consumer & Tech. Products Div.
P. 0. Box 1035
Toledo, Ohio 43601
Pall Trincor Corp.
459 Chestnut St.
Union, N.J. 07083
Pall Trinity Micro Corp.
Route 281
Cortland, N.Y. 13045
Pennwalt Corp.
3 Penn Center
Philadelphia, Pa. 19102
Perry Products Co.
1421 N. 6th St.
Philadelphia, Pa. 19122
Planchefs Lab Products
P. 0. Box 1802
Ann Arbor, Mich. 48106
Pulverizing Machinery Div.
10 Chatham Rd.
Summit, N.J. 07901
Service Filter Corp.
7433 N. Harlem Ave.
Chicago, 111. 60648
Straightline Filters, Inc.
Box 1911
Wilmington, Del. 19899
Technican Corp.
511 Benedict Ave.
Tarrytown, N.Y. 10591
Testing Machines, Inc.
400 Bayview Ave.
Amityvllle, N.Y. 11701
Tri-R Instruments, Inc.
48 Merrick Rd.
Rockville Center, N.Y. 11570
Vanton Pump & Equipment Corp,
201 Sweetland Ave.
Hillside, N.J. 07205
Zena Co.
723 - 22nd St.
Union City, N.J. 07087
-190-
-------
17551-6012-RO-OO
APPENDIX G
PROCESS pH MONITORING SYSTEMS
BECKMAN
UNITS IN SYSTEM
Sensors
Glass Electrode
Reference Electrode
Thermocompensator
Electrode Assembly
Flow Chamber
Flow Chamber (Low
pressure epoxy)
Standard glass pH electrode with short
wide electrode bodies for strength and
rapid replacement features - $32.00
Conventional reference with electrolyte
reservoir or new Lazaran plastic un-
breakable, chemically resistant elec-
trode. Needs no electrolyte replenish-
ment. - $32.00 - $125.00
Temperature sensitive resistance ele-
ment for automatic correction of ana-
lyzer for temperature variation to read
true pH - $47.00
Available in stainless steel or poly-
vinyl dichloride useful at high pres-
sure and over a wide range of tempera-
ture. Other materials are special
Order. Easy snam out electrode assem-
bly - $150.00 - $210.00
Dimensions:
Height - 14-1/4" (plus 6-5/16"
for reservoir where necessary)
Diameter - 3-1/4"
Inlet - 1/2" NPT
Outlet - 1/2" NPT
Conduit - 1/2" NPT
Useful for pH measurement in streams
below 15 psig. Requires a unique 1"
diameter reference electrode. 1/2"
pipe fittings - $76.00
-191-
-------
17551-6012-RO-OO
Submersion Chamber
Separate Insertion Glands
pH Analyzer
pH Indicator and
Amp!i fi er/Transmi tter
Similar to standard flow chamber des-
scribed above but designed for submer-
sion in tank, etc. - $115.00 - $195.00
Used for inserting electrodes separately
into pipe, etc., threaded - $150.00 -
$140.00
May be purchased as separate units or
as a single compact unit. Two models
available. Differential amplification.
Direct output from analyzers to grounded
or ungrounded recorder or other readout
device. Low drift and noise. Solid
state circuitry. Scales for any 2, 5
or 10 pH units. Voltage or current out-
put. Automatic or manual temperature
compensation. Optional alarm contacts -
$610.00
UNIVERSAL INTERLOCK, INC.
UNITS IN SYSTEM
Sensors
Glass Electrode
Reference Electrode
Thermocompensator
Electrode Assembly
Flow Chamber
Standard glass pH electrodes with rugged
construction manufactured by Micro
Sensors, Inc. Combination electrode
also available.
Available with either a hardwood plug
liquid junction or a ceramic liquid
junction. Sealed industrial reference
available.
Automatic temperature compensator avail-
able.
Available in PVC, Penton, and Teflon.
Stainless steel special order. Can be
purchased with built in preamplifier
and features twist-open flow chamber
for easy cleaning. $365.00 - $540.00
includes preamp, electrodes, tempera-
ture compensator and cell holder.
-192-
-------
17551-6012-RO-OO
Insertion Assembly
Dimensions:
Mounted on a 10' x 10" PVC plate
Inlet - 1/2" NPT or 1" NPT
Outlet - 1/2" NPT
Unit contains electrodes, preamplifier,
automatic temperature compensator and
cell holders. Designed for pipe or
tank installation - $470.00 - $595.00
Dimensions:
pH Analyzer
Transmitter
Length - 10-1/8"
Diameter - 3"
Probe portion - 5"
Housing with 2-1/2" NPT
All the signals preamplified at source.
System terminates ground loop interfer-
ence and is compatible with grounded or
ungrounded instrumentation. Low drift
and noise. Automatic or manual temp-
erature compensation. Alarm contacts -
$470.00 - $685.00
LEEDS AND NORTHRUP
UNITS IN SYSTEM
Sensors
Glass Electrode
Reference Electrode
Thermocompensator
Electrode Assembly
Flow Chamber
Submersion Chamber
Standard heavy duty type - $30.00
Low diffusion calomel reference requires
less frequent filling. Non-fouling
characteristics - $30.00
Available as part of system - $44.00
Many available for various applications
(flow, pressures, temperature, etc.).
Non-fouling plastic (polypropylene) as-
sembly eliminates clogging and electrode
fouling in some applications. Available
also in stainless steel - $109.00
Many types available for various appli-
cations.
-193-
-------
17551-6012-RO-OO
pH Analyzer
Amplifier Preamplifier and amplifier are contained
in one unit. Measurement of pH reading
is performed by first generation system
directly amplifying the differential mv
signal - $608.00
Recorder Hi-impedance unit capable of direct am-
plification and recording of mv elec-
trode output - -v$l,000.00
-194-
-------
17551-6012-RO-OO
REFERENCES
1. Radian Corporation, Contract 68-02-0008, "Laboratory Analyses
For Wet Limestone Scrubbing Processes".
2. Radian Corporation, Contract CPA-22-69-138, "A Theoretical Description
of the Limestone Injection-Wet Scrubbing Process", June 9, 1970.
3. "Alkali Scrubbing Test Facility, Phase 2: Design Engineering", Bechtel,
August 1970.
4. L. Szekeres and F. Bakacs-Polgar, "Determination of Alkali Hydrogen
Carbonates in the Presence of Alkali Carbonates", Act. Chim. Hung.
Tomus 26, 1961.
5. H. T. Dyer, Advances in X-Ray Analysis 9_, 447 (1966).
6. B. P. Fabbi and W. S. Moore, Applied Spectroscopy, 24, 426 (1970).
7. J. F. Harris and L. L. Zoch, Anal. Chem. 34, 201 (1962)
8. E. F. Rissman and R. L. Larkin, Anal. Chem. 42, 1628 (1970)
9. R. Dunk, R. A. Mostyn, H. C. Hoave, Atomic Absorption Newsletter
8, 79 (1969).
10. H. E. Taylor, 0. H. Gibson and R. K. Skogerbee, Anal. Chem. 42_,
1569 (1970).
11. A. Syty and J. A. Dean, Applied Optics 1, 1331 (1968).
12. "Standard Methods/Water and Waste Water", 13th Ed. 1971, APHA,
AWWA, WPCF, Method 156C, p 334.
13. E. A. Burns, "Nitrogen Oxygen Compounds", Chapter in Analytical
Chemistry of Nitrogen and Its Compounds" Part 1, C. A. Streuli
and P. R. Averell, Ed., Wiley-Interscience, New York, 1970, p 85.
14. Environmental Protection Agency Water Quality Office, Methods for
Chemical Analysis of Water and Wastes, 1971.
15. F. L. Fisher, E. R. Ibest and H. F. Beckman, Anal. Chem. Vol. 30.
1972 (1958).
16. R. Di Martini, Anal. Chem., Vol. 42, 1102 (1970).
-195-
-------
17551-6012-RO-OO
REFERENCES (CONTINUED)
17. J. N. Driscoll, et. al., "Determination of Oxides of Nitrogen in
Combustion Effluents with a Nitrate Ion Selective Electrode," Presented
at 64th Annual APCA Meeting, June 1971.
18. B. A. Chertkov, J. Applied Chem. 3£, 2687 (1959).
19. L. C. Schroeter, "Sulfur Dioxide," Pergamon Press, New York, New York
(1966).
20. Am. Coc. Mech. Eng., "Flue and Exhaust Gas Analysis, Power Test Code
(1936)
21. P. W. West and G. C. Gaeke, Anal. Chem. 28_, 1916 (1956).
22. "Air Pollution," Ed. A. C. Stern, Vol. II Analysis, Monitoring and
Surveying, 2nd Ed., Acad. Press, New York (1968).
23. "Std. Methods for Exam, of Water and Waste Water," 12th Ed. Am. Pub.
Health (1965).
24. I. M. Kolthoff and V. A. Stenger, "Volumetric Analysis," Vol. Ill,
Titration Methods, 2nd Ed., Interscience Publ., Inc., New York (1947).
25. ASTM D516-68
26. "Handbook of Anal. Chem." Ed. L. Meites, 1st Ed., McGraw Hill, (1963).
27. I. M. Kolthoff and C. S. Miller, J. A. C. S. 63, 2818 (1944).
28. "Scott's Standard Methods of Chem. Anal." 5th Ed., D. Van Mostrand Co.,
New Jersey (1956).
29. R. G. White, "Handbook of Ultraviolet Methods," Plenum, New York (1965).
30. F. J. Weicher, "The Analytical Uses of EDTA," D. Van Nostrand, New
Jersey (1958)
31. I. M. Kolthoff and E. B. Sandell, "Textbook of Quantitative Inorganic
Analysis," 3rd Ed., Macmillan Co., New York (1952).
32. Hillebrand, Lundell, Bright and Hoffman, "Applied Inorganic Analysis,"
2nd Ed., Wiley, New York (1953).
33. Snell and Snell, "Colorimetric Methods of Analysis," Vol. II, 3rd Ed.,
D. Van Nostrand, New Jersey (1963).
-196-
------- |