::!S::i:::ij:^
Si:!:-::::::::::::::::^
U.S. E ^ -—^ N AGENCY
-------
-------
PROCEEDINGS
SECOND INTERNATIONAL
LIME/LIMESTONE
WET-SCRUBBING SYMPOSIUM
VOLUME II
November 8-12, 1971
Sheraton-Charles Hotel
New Orleans, Louisiana
ENVIRONMENTAL PROTECTION AGENCY
Office of Administration
Research Triangle Park, North Carolina
June 1972
-------
The APTD (Air Pollution Technical Data) series of reports is
issued by the Environmental Protection Agency to report tech-
nical data of interest to a limited number of readers. Copies
of APTD reports are available free of charge to Federal employ-
ees, current contractors and grantees, and nonprofit organiza-
tions - as supplies permit - from the Air Pollution Technical
Information Center, Environmental Protection Agency, Research
Triangle Park, North Carolina 27711 or from the National Tech-
nical Information Service, 5285 Port Royal Road, Springfield,
Virginia 22151.
EPA REVIEW NOTICE
These proceedings have been reviewed by the Environmental
Protection Agency and approved for publication. The contents
of this report are reproduced herein as received from the
authors. Approval does not signify that the contents neces-
sarily reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or commer-
cial products constitute endorsement or recommendation for
use.
Publication Number APTD-1161
11
-------
PREFACE
The Second International Lime/Limestone Wet Scrubbing
Symposium was held November 8-12, 1971, in the Claiborne
Room of the Sheraton-Charles Hotel, New Orleans, Louisiana,
sponsored by the Environmental Protection Agency, Office
of Air Programs, Control Systems Division.
The Symposium, under the chairmanship and vice-chair-
manship of Messrs. E.L. Plyler and D.R. Mayfield, began
Monday morning with the official welcome and opening remarks
by OAP's Sheldon Meyers, followed by an introduction by
Mr. Plyler.
The Symposium consisted of nine sessions, divided into
five different areas: fundamental research, pilot scale
research and development, prototype and full scale tests,
panel discussion on scaling, sampling and analytical methods.
Sessions 1 and 2, chaired by Philip S. Lowell, were
concerned with fundamental research. Frank T. Princiotta
of OAP was the chairman for Sessions 3 and 4 and the first
three presentations of Session 5, dealing with pilot scale
research and development.
The last two presentations of Session 5 and Sessions 6
and 7 were chaired by H.W. Elder. The 17 papers presented in
these sessions were concerned with prototype and full scale tests
The panel discussion on problems related to scaling in
lime/limestone wet scrubbing, chaired by A.V. Slack, consisted
of the five papers of Session 8. OAP's J.A. Dorsey chaired
Session 9, sampling and analytical methods, which concluded the
symposium early Friday afternoon.
AH papers presented during the symposium are included
in these proceedings except those which were given by notes
and for which there exists no written text.
111
-------
-------
CONTENTS
Title
VOLUME I
Preface
Philip S. Lowell
SUMMARY: FUNDAMENTAL RESEARCH - PARTS I AND II ........... 1
FUNDAMENTAL RESEARCH - PART I
C.Y. Wen and S. Uchida
Simulation of SC>2 Absorption in a Venturi Scrubber
by Alkaline Solutions ................................. 3
M. Epstein, C.C. Leivo, and C.H. Rowland
Mathematical Models for Pressure Drop, Particulate
Removal and SC>2 Removal in Venturi, TCA, and
Hydro-filter Scrubbers ................................ 45
Delbert M. Ottmers, Jr.
A Model for the Limestone Injection-Wet Scrubbing
Process for Sulfur Dioxide Removal from Power Plant
Flue Gas .............................................. 115
James L. Phillips
Precipitation Kinetics of CaSC>4. 2H20 ..................... 1 51
D.C. Drehmel
Limestone Types for Flue Gas Scrubbing ................... 1 67
FUNDAMENTAL RESEARCH - PART II
J.M. Potts, A.V. Slack, and J.D. Hatfield
Removal of Sulfur Dioxide from Stack Gases by
Scrubbing with Limestone Slurry: Small-Scale
Studies at TVA 195
L.H. Garcia
Absorption Studies of Equimolar Concentrations of NO
and N02 in Alkaline Solutions 233
J.D. Hatfield and J.M. Potts
Removal of Sulfur Dioxide from Stack Gases by Scrubbing
with Limestone Slurry: Use of Organic Acids 263
-------
CONTENTS (continued)
JL Page
J.S. Morris
Potential Water Quality Problems Associated with
Limestone Wet Scrubbing for S02 Removal from
Stack Gases 285
Linda Z. Condry, Richard B. Muter and William F. Lawrence
Potential Utilization of Solid Waste from Lime/Limestone
Wet Scrubbing of Flue Gases 301
Frank T. Princiotta
SUMMARY: PILOT SCALE RESEARCH AND DEVELOPMENT - PARTS I,
II, AND III 315
PILOT SCALE RESEARCH AND DEVELOPMENT - PART I
B.N. Murthy, D.B. Harris and J.L. Phillips
Sulfur Dioxide Absorption Studies with EPA In-House
Pilot Scale Venturi Scrubber 325
I.S. Shah
SO2 Removal Using Calcium Based Alkalies Pilot Plant
Experience 345
R.A. Person, C.R. Allenbach, I.S. Shah, and S.J. Sawyer
A Pilot Plant Test Program for Sulfur Dioxide Removal
from Boiler Flue Gases Using Limestone and Hydrated
Lime 373
R.J. Gleason
Limestone Scrubbing Efficiency of Sulfur Dioxide in
a Wetted Film Packed Tower in Series with a Venturi
Scrubber 391
T.M. Kelso, P.C. Williamson, and J.J. Schultz
Removal of Sulfur Dioxide from Stack Gases by Scrubbing
with Limestone Slurry: TVA Pilot Plant Tests.
Part I - Scrubber-Type Comparison 437
N.D. Moore
Part II — Experimental Design and Data Analysis for
Spray and Mobile-Bed Scrubbers 462
VI
-------
CONTENTS (continued)
Title Pag
PILOT SCALE RESEARCH AND DEVELOPMENT - PART II
A. Saleem, D. Harrison, and N. Sekhar
Sulfur Dioxide Removal by Limestone Slurry in a Spray
Tower 48
J.L. Shapiro and W.L. Kuo
The Mohave/Navajo Pilot Facility for Sulfur Dioxide
Removal 50]
J.H. McCarthy and J.J. Roosen
Detroit Edison Pilot Plant and Full-Scale Development
Program for Alkali Scrubbing Systems—A Progress
Report 52!
A.L. Plumley and M.R. Gogineni
Research and Development in Wet Scrubber Systems 54'
D.E. Reedy
Lime Scrubbing of Simulated Roaster Off-Gas 56'
VOLUME II
PILOT SCALE. RESEARCH AND DEVELOPMENT - PART III
John M. Craig, Burke Bell, and J.M. Payadh
Mobile Pilot Plant Study of the Wet Limestone Process
for SO2 Control 57!
Robert J. Phillips
Sulfur Dioxide Emission Control for Industrial Power
Plants 60:
Ivor E. Campbell and James E. Foard
Sulfur Oxide Control at the Copper Smelter 63?
H.W. Elder
SUMMARY: PROTOTYPE AND FULL SCALE TESTS - PARTS I, II,
AND III 65 •
PROTOTYPE AND FULL SCALE TESTS - PART I
James Jonakin and James Martin
Applications of the C-E Air Pollution Control System 657
vii
-------
CONTENTS (continued)
'itle Page
E.G. McKinney and A.F. Little
Removal of Sulfur Dioxide from Stack Gases by
Scrubbing with Limestone Slurry: Design Considerations
for Demonstration Full-Scale System at TVA 673
PROTOTYPE AND FULL SCALE TESTS - PART II
M. Epstein, F. Princiotta, R.M. Sherwin, L. Szeibert,
and I.A. Raben
Test Program for the EPA Alkali Scrubbing Test Facility
at the Shawnee Power Plant 693
R.M. Sherwin, I .A. Raben, and P.P. Anas
Economics of Limestone Wet Scrubbing Systems 745
J.D. McKenna and R.S. Atkins
The RC/Bahco System for Removal of Sulfur Oxides and
Fly Ash from Flue Gases 765
Gerhard Hausberg
The Bischoff-Process—Initial Results from a Full-Size
Experimental Plant 785
J.J. O'Donnell and A.G. Sliger
Availability of Limestones and Dolomites 799
PROTOTYPE AND FULL SCALE TESTS - PART III
Tsukumo Uno, Masumi Atsukawa, and Kenzo Muramatsu
The Pilot Scale R&D and Prototype Plant of MHI Lime-
Gypsum Process 833
Lyman K. Mundth
Wet Scrubber Installations at Arizona Public Service
Company Power Plants 851
J.H. McCarthy and J.J. Roosen
Detroit Edison Full-Scale Development Program for
Alkali Scrubbing Systems (The material in this
paper was included in Paper No. 4c.) 863
Vlll
-------
CONTENTS (continued)
Title
Robert R. Padron and Kenneth C. O'Brien
A Full-Scale Limestone Wet Scrubbing System for the
Utility Board of the City of Key West, Florida 865
J.A. Noer and A.E. Swanson
Air Pollution Control at the Northern States Power
Company Sherburne County Generating Plant 877
J.W. James
Ontario Hydro's Prototype Limestone Scrubber for SO2
Removal from Clean Flue Gas 899
D.T. McPhee
La Cygne Station Air Quality System 907
J.F. McLaughlin, Jr.
Sulfur Dioxide Scrubber Service Record Union Electric
Company—Meramec Unit 2 915
B.C. Gifford
Will County Unit 1 Limestone Wet Scrubber 917
H.P. Willett and I.S. Shah
A Summary Report—Chemico's Commercial Systems
Installations at Electric Power Generating Stations ... 931
A.V. Slack
SUMMARY: PROBLEMS RELATED TO SCALING IN LIME/LIMESTONE WET
SCRUBBING 943
PROBLEMS RELATED TO SCALING IN LIME/LIMESTONE WET SCRUBBING
A.V. Slack and J.D. Hatfield
Removal of Sulfur Dioxide from Stack Gases by Scrubbing
with Limestone Slurry: Operational Aspects of the
Scaling Problem 947
Bela M. Fabuss
Calcium Sulfate Scaling 965
Joan B. Berkowitz
Review of Scaling Problems in Limestone Based Wet
Scrubbing Processes 975
IX
-------
CONTENTS (continued)
Title Page
J.R. Martin
Deposition Problems and Solutions in the Combustion
Engineering Lime/Limestone Wet Scrubbing Systems 983
Philip S. Lowell
Use of Chemical Analysis and Solution Equilibria in
Predicting Calcium Sulfate/Sulfite Scaling Potential .. 1001
J.A. Dorsey
SUMMARY: SAMPLING AND ANALYTICAL METHODS 1013
SAMPLING AND ANALYTICAL METHODS
Klaus Schwitzgebel
Development and Field Verification of Sampling and
Analytical Methods for Shawnee 1017
E.A. Burns and A. Grunt
On-Stream Characterization of the Limestone/Dolomite
Wet Scrubber Process 1053
Terry Smith and Ronald Draftz
Particulate Emissions from Two Limestone Wet Scrubbers ... 1073
Terry Smith and Hsing-Chi Chang
Design Criteria for a Size-Selective Sampler for Lime/
Limestone Wet Scrubbers 1083
R.M. Statnick and J.A. Dorsey
Instrumental Methods for Flue Gas Analysis 1097
Gene W. Smith
EPA Recommended Source Test Methods for New Source
Performance Standards Testing 1109
-------
MOBILE PILOT PLANT STUDY
OF THE
WET LIMESTONE PROCESS
FOR
S02 CONTROL
BY
JOHN M. CRAIG, Ph. D.1
BURKE BELL^ ?
J. M. FAYADH, Ph. D.
Prepared for Presentation at the
SECOND INTERNATIONAL LIME/LIMESTONE
WET SCRUBBING SYMPOSIUM
NEW ORLEANS, LOUISIANA
NOVEMBER 8-12, 1971
1 Present Affiliation - Southern Services Inc., Birmingham, Ala.
2 Zurn Environmental Engineers, Washington, D. C.
-------
INTRODUCTION
One of the earliest processes utilized for the removal of
sulfur dioxide from combustion erases involved scrubbing with a
slurry of lime or limestone in water. This process was inves-
tigated in England in the 1930's and with some significant
variation has been studied since then by interested parties
throughout the world. The federal government^ utilities, and
hardware manufacturers have recognized the potential of lime-
stone scrubbing and are actively engaged in research and develop-
ment programs to develop the process and demonstrate its
commercial application.
A critical step in the commercial utilization of this
process is, of course, experimentation at the pilot plant scale
to develop firm data on design, operation, and economics and
to re-examine laboratory data in a more realistic environment.
The pilot plant stage allows a more critical review of factors
such as process chemistry, mass transfer mechanisms, scale
formation and control, effect of limestone characteristics,
optimum scrubber design, and identification of water pollution
problems.
Zurn Industries, Incorporated along with the Office of
Air Programs, Environmental Protection Agency conducted a joint
research project which involved a pilot plant study of wet
limestone scrubbing: utilizing a Zurn Air Systems Division
576
-------
"Dustraxtor" scrubber. The pilot plant was mobile and was
installed at two electric generating stations, an oil-fired
station in Key West, Florida and a coal-fired station in Paducah,
Kentucky. The study was applied in nature and investigated
variables important to the success of the limestone scrubbing
process. Data were developed which will ultimately be utilized
in a demonstration project to be discussed later during this
symposium. Since Zurn Industries has an interest in the
commercial development of this type of system, we contributed
significantly to the cost of the project through the design
and construction of the pilot plant.
CHEMICAL CONTACTING IN THE DUSTRAXTOR
The objective of flue gas scrubbing for mass transfer is
to remove as much material as economically possible out of the
gas and into the scrubbing liquid. There are physical, chemical
equilibrium, and rate relationships that limit the solubility
of material in liquid that control the mass transfer relations.
However, the amount of mass transfer not only depends upon the
equilibrium relationships, but also on the contacting scheme.
The "Dustraxtor" employs many contacting schemes in its operation.
The "Dustraxtor" scrubbing unit employed in the pilot
plant study is a type of turbulent contact scrubber and is
shown in Figures 1 and 2. The unit consists of a flooded,
collecting tube through which the flue gas must pass. The
collecting tube is installed vertically in the inlet plenum
chamber directly above the recycle hopper so that the bottom
of the tube is a short distance above the liquid level in the
hopper.
577
-------
Flue pases enter the inlet plenum chamber where they sweep
over the surface of the scrubbing; liquid and are directed up
the collecting tube. The velocity of the gas passing beneath
the collecting tube and above the collecting bonnet draws the
scrubbing slurry from the bonnet surface and upward into the
collecting tube. The shearing action of the gas atomizes the
scrubbing slurry into a dense spray as the gas-slurry mixture
continues up through the collecting tube. The result of this
action is a highly turbulent mixing zone which provides the
intimate contacting and time necessary for the chemical reaction
to occur. As the gases are discharged from the collecting tube,
they are directed into a curved liquid deflector which acts as
a separator by forcing the slurry downward onto the tube sheet.
This shower effect provides an additional mixing zone for
absorption with chemical reaction to occur. The cleaned gas is
discharged through the exhaust stack and the scrubbing solution
is returned by gravity to the recycle hopper. When employed
for limestone scrubbing, this built-in recycle of the scrubbing
slurry provides sufficient time for the limestone to go into
solution and react with the sulfur oxides.
The stages of contact in this type of unit are:
1. The initial shearing action as the flue gas passes
through the slot between the collecting bonnet and tube,
2. The highly turbulent mixing zone within the collecting
tube where the gas is in intimate contact with liquid
droplets.
3. Impingement of the gases and liquid upon tne surface
of the deflector.
578
-------
J4-. The passage of the gas through a highly turbulent curtain
of liquid being discharged from the liquid deflector.
The contacting: schemes found in this type of unit include
counter-current,co-current, and cross-current. Since there are
many contacting schemes taking place at the same tinie in this
unit, a theoretical analysis of the total mass transfer mechanism
would be very complicated to perform and may prove to be only of
academic interest.
The unit normally operates at a pressure drop of from 6
to 8 inches WG. This pressure drop is regulated by either
raisinpr or lowering the adjustable level control weir. One
advantage of this type of unit in limestone scrubbing applications
is that there are no moving parts, spray nozzles, or pumps
necessary for operation. By variation in weir height, bonnet
spacing or tube selection, a wide range of operating conditions
may be met at maximum efficiency.
DESCRIPTION OF THE MOBILE PILOT PLANT
Zurn Industries cooperated with the Office of Air Programs,
Environmental Protection Agency on a project in which a mobile
limestone scrubbing pilot plant was designed, manufactured, and
operated. The major objectives of the pilot plant study were
to evaluate the potential of this type of turbulent contact
scrubber in the wet limestone process, investigate process
chemistry and kinetics, and generate design data for scale-up
purposes. The first phase of the pilot plant study was conducted
in Key West, Florida, at the municipal power plant which burned
approximately 1-2 percent sulfur content fuel oil. After this
579
-------
phase of the project was completed, the mobile pilot plant was
moved to the TVA Shawnee coal-fired power plant for the second
phase of the study.
Figure 3 is a schematic flow diagram of the pilot plant.
The scrubber had a nominal capacity of 1500 cfm. The dry
collector permitted operation with and without fly ash when
used during the second phase of the study. Other salient
features of the pilot plant were capabilities to handle a wide
range of gas and liquid flow rates, construction materials to
withstand a wide range of pH (pH 3 to 9), and provisions for
ready access to the scrubber to check on scale buildup and for
scale removal, if necessary.
Six types of limestone reactants were evaluated (coral marl,
aragonite, Predonia Valley limestone, dolomite, precipitated
calcium carbonate, and lime). A background series of tests were
conducted with salt water. The stoichiometry of the reactant
and the effect of slurry concentration were studied for selected
limestones. Stoichiometric requirements were based upon the
molar requirements for the reactant to form CaSO~. For selected
reactants, the influence of particle size on sulfur dioxide
removal efficiency was studied. The effect of an iron catalyst
on sulfur dioxide removal efficiency and on the composition of
the spent liquid slurry was also evaluated.
The scrubber variables evaluated with respect to the
efficiency of sulfur dioxide removal for the most effective
reactant combinations include gas flow rate, system pressure
drop, stoichiometry, particle size, and slurry concentration.
Other factors analyzed during the test program included but
580
-------
were not limited to:
Corrosion of the various metals utilized throughout
the system.
Scale formation.
Operating; and maintenance problems associated with
operation of the system.
In order to evaluate these variables, the following data
were collected during each experimental run:
Boiler operating conditions.
Inlet and outlet gas samples including sulfur oxides,
nitrogen oxides, particulates, flow rate, temperatures
(wet bulb and dry bulb).
Compositions of the slurry feed to scrubber and scrubber
slurry discharge including solids concentration, calcium
and magnesium, sulfate and sulfite, nitrogen compounds,
chloride, pH. temperature, and flow rate.
The equipment utilized for this study included a modified
Dustraxtor containing one tube. The scrubber was modified by
the addition of an external weir to enable measurement of the
quantity of liquid drawn up the tube. The scrubber was designed
such that various tube diameters could be tested in the system.
(12 inch I.D. and 8 inch I.D. tubes were utilized during this
study). A continuous record of the inlet and outlet sulfur
dioxide concentration was taken during each experimental run
utilizing a Dynasciences Model SS-130 analyzer connected to a
strip chart recorder. The analyzer was calibrated twice daily
utilizing a known standard of certified calibration p;as.
Intermittent nitrogen oxide and particulate samples were taken
581
-------
utilizing standard stack sampling techniques and analyzed wiun
standard procedures. Gas flow rate was determined with a
calibrated "Annubar" flow meter.
The mobile pilot plant was installed on the Number 3 unit
of the Key West Municipal Power Plant during the first phase
of this study. This unit has a generating capacity of 20 MW
and was burning 1 to 2 percent sulfur content fuel oil during
the experimental period. During the second phase of this study,
the mobile pilot plant was installed on the Number 9 and 10
units of TVA's Shawnee generating station. Each unit has a
generating capacity of 150 MW and was burning 2| to ^ percent
sulfur content coal during the experimental period. The Number
10 unit at the Shawnee plant was also in the dry limestone
injection configuration and this gas was scrubbed in the mobile
pilot plant.
RESULTS OP THE TEST PROGRAM
This section of the report will discuss the results from
the three major phases of the test program as well as operating
and maintenance problems associated with the pilot plant.
Generally speaking, the pilot plant performed satisfactorily
with only the normal problems that might be expected in this
type of operation.
A detailed final report on this project is currently being
prepared and will be available soon. This presentation will
only summarize some of the tests conducted on an oil-fired boiler,
a coal-fired boiler, and a limestone-injection boiler. The
results were favorable and indicate that this type of scrubber
582
-------
should be considered for use in the wet limestone S02 control
process.
Key West Tests
Initial tests were conducted at the Key West site to
determine the quantity of liquid that was drawn up into the
tube. Figure ^ summarizes the data accumulated during this
phase of the test program. As can be seen from Figure ^, the
quantity of liquid drawn up into the tube is a function of the
gas flow rate and pressure drop across the scrubber. Liquid
to gas ratios of 100 to greater than 500 gallons per MCFH were
attainable with this scrubber configuration.
Figure 3 illustrates the scrubber configuration and sample
locations utilized during the limestone scrubbing phase of the
Key West test program. A fractional factorial design was
employed to determine the major parameters which effect SOp
removal in this type of scrubber. Tables 1 and 2 summarize
the field test data associated with the experimental design for
each primary reactant, coral and Fredonia Valley limestone.
Based upon the experimental design, stoichiometry, gas flow
rate and pressure drop are highly significant factors in the
design of a S02 -limestone scrubbing system using this type
of scrubber. Figure 5 summarizes the results obtained with the
two primary reactants studied. Generally, at a stoichiometric
ratio of 1:1, a pressure drop of 6.5 inches H20, and 2000 ACFM
gas flow rate, efficiences of 50-75 percent were attained
depending: upon the firrind of the reactant studied. An increase
in scrubber pressure drop to 12 inches H20 and a decrease in
flow rate to 1000 ACFM improved removal efficiency to a
583
-------
level of 75 to 90 percent. However, it should be pointed out
that the scrubber was in a once-through configuration and was
only operated for a period of 3 to ^ hours per test.
Figure 6 summarizes the effect of pH upon SOp removal
efficiency during this phase of the test program. Due to the
location of the fresh slurry feed line, in relation to the
spent slurry discharge, improper mixing occurred in the scrubber,
and the discharge pH was not as reliable an indicator of this
effect as the weir pH.
Another test series was conducted to determine the effect
of gas flow rate upon S02 removal efficiency. The test conditions
were fixed, and only the gas flow rate through the scrubber
was varied. Figure 7 summarizes the results obtained during
this series of tests. The results were as expected, when the
L/G ratio for a given tube diameter increased, the S02 removal
efficiency increased. The liquid flow rate utilized in the
L/G calculation is actually the liquid drawn up into the tube
based upon Figure k and not the actual liquid flow to the
scrubber.
Tests were conducted using other reactar.ts, dolomite, lime,
and precipitated calcium carbonate. The results of this phase
of the test program are summarized in Table 3. All reactants
ran satisfactorily except for precipitated calcium carbonate
which caused heavy foaming at the weir and inside the scrubber.
Shawnee Unit No. 9 Tests
Table ^ summarizes the Fredonia Valley limestone tests
conducted while the mobile pilot plant was connected to the
No. 9 unit at the Shawnee station. Due to pressure drop
584
-------
consideration across the entire system, it was necessary to
conduct all tests with an 8 inch diameter tube. A fractional
factorial design similar to that used during the first phase
of the test program was utilized for the primary reactant.
Commercially available grinds were used to determine the effect
of particle size on SG>2 removal efficiency. Each experiment was
conducted with the primary dust collector in operation as
depicted in Figure 3. The experimental technique was identical
to that utilized in Key West, wherein the liquor fed to the scrubber
was based upon the test stoichiometry, gas flow rate and SOp
concentration in the entering gas. Figure 8 summarizes the
results from this test series. The highly significant parameters
for this system were found to be pressure drop, gas flow rate,
and stoichiometry. Particle size was not a significant factor
due to the large amount of 100 mesh material passing the 325
mesh screen. This situation occurred in previous studies and
made it very difficult to determine the effect of a coarser
grind upon S02 removal efficiency.
During testing on No. 9 unit, little or no mechanical or
operating difficulties were encountered. Scale formation was not
a problem due to the installation of a device to continually
wash down the tube. Based upon the limited number of tests
and operating hours on a coal-fired unit, removal efficiencies
of 35 to 80 percent may be expected in a once-through
585
-------
configuration depending upon gas flow rate, stoichiometry,
particle size and pressure drop. However, additional testing
would be required prior to any large scale-up of this scrubber
for control of S02 from a coal-fired station.
An abbreviated test series was conducted inorder to
evaluate the effect of ionic strength upon S02 removal. A
increased ionic strength slurry solution was tested at a
similar pressure drop, gas flow rate and stoichiometry as was
used during the first phase of this study. The results are
shown below:
Location SOg Removal Efficiency %
No. 9 Unit 69.6
No. 9 Unit, Simulated Ionic Strength 75.5
(12,000mg/l Cl~)
Key West (24,000 mg/1 Cl~) 84.2
Shawnee Unit No. 10 Tests
Figure 9 illustrates the configuration of the mobile
pilot plant when installed on the No. 10 unit at Shawnee. All
gas to the scrubber, laden with calcined limestone from the
limestone injected boiler, by-passed the primary mechanical
collector and entered the scrubber. Table 5 summarizes the
results from this test series. As can be seen, S02 removal
586
-------
efficiencies of 60 to 9^ percent were attainable depending upon
stoichiometry, gas flow rate, and pressure drop. However, due
to the design of the scrubber, the entering gas and limestone
impinged upon the tube and built-up until the cake fell off due
to its weight. A re-design of the entrance to the scrubber would
be necessary before this unit could be applied on a limestone
in.lection boiler. The low inlet S02 concentrations found
during this test program can be attributed to leakage of
ambient air into the system.
Operating Problems
In addition to the operating problems discussed previously,
two other problems were associated with the mobile pilot plant:
corrosion and scale build-up on the blower. Since the scrubber
was initially tested utilizing salt water, severe corrosion
problems occurred in the scrubber and discharge piping. However,
once the salt water tests were concluded and fresh water was
utilized in the system, no additional failures due to corrosion
were encountered.
Scale build-up on the blower blades caused serious operating
problems more than once during the life of this project. The
scale build-up would occur on the blower blades, causing an
imbalance in the system, and in some cases, ultimate failure
of the blower. This problem could have been avoided by using
a "hot-side" blower instead of locating one down-stream of the
scrubber system.
Scale Formation in the Scrubber
Scale formation occurred at two points in the scrubber
system. Initially, scale was discovered at the tip of the tube
587
-------
in an area of high turbulence and then at a later date, some
slight crystal formation was noticed on the scrubber housing
at the water line. Heavy scale formation on the tube lip
could develop into an operating problem if not corrected, since
the crystal growth occurred directly in the gas path; therefore,
precautions were taken to eliminate scale formation at this
location. Table 6 indicates the major compounds found in a
sample of the scale formed on the tube.
TABLE 6
ANALYSIS OF TUBE SCALE
Compound Quantity (wt %)
CaSO^ • 2H20 30.5
CaCCU 68.0
MgC03 1.5
Rate of scale formation appeared to be a direct function of
the slurry concentration; however, no effort was made to
quantify this conclusion. An appreciable difference was not
noticed in the rate of scale formation caused by the primary
reactants tested, coral or Predonia Valley limestone.
588
-------
BALANCE VENT LINE
DEFLECTOR
COLLECTING TUBE
COLLECTING BONNET
WIER
WATER IN
LEVEL CONTROL
DUST AND WATER OUT
FIGURE 1. TYPICAL CROSS SECTION OF DUSTRAXTOR
589
-------
CLEAN AIR OUT
OUTLET to
ROUND BOOT CAS UK
WATH DDLBCKM
DIRTT GAS IM MOD WATH
HOPPER CLEANED MANUALLY OB
BY ADDITION Of MECHANICAL
mm®
COLLECTED MATERIAL
TO BOTTOM OF HOPPER
FIGURE 2. DIAGRAMMATIC VIEW OF SCRUBBING
ACTION WITHIN DHSTRAXTOR
590
-------
-------
FIGURE
f *
I 5
^ i
-------
8°--
60--
So--
£> LlHSSTofJ*
v Line
B $AiT l^fTf^.
A /xoi^vr^
4-0 • •
Jo--
/o --
FIGURE 5
593
7
-------
1
I
•a:
«
o
>"
-oil
§ v,
< s
C
1 s?
Hj K
* § 5
vi 8 >
«S
f N
W
M
^
>N
'«!
in
I N
I
4- tv.
fi
594
-------
fe
1
I
^3
-3
O
M
PH
I
A
595
-------
/oo 4-
fc
i
I
* \-
= 6.C
4
-6 ^
FIGURE 8
596
-------
O\
w
K
£>
c5
M
fr.
597
-------
I
<
S3
o
M
EH
O
<
W
CO
598
-------
TABLE 2
SUMMARY OF KEY WEST FRACTIONAL FACTORIAL DATA
FenoM WLity L'^e-sro^f- FIELD TEST DATA ^/?"^ -^-j^-)
,cp
7
IL,
1
U-.'
\ — '! —
L '
f Jl'f
Vl
H
_j
• j
_i...
if
Tji
. o!
\ i;
i o
tl ',
', r~
•i'
-9
, 1 (
: ^
'V
:
00 in -
vA • ••
+' o' «o,
; *j £ ? _
! *^ O. (Tl
r S?I - w
i
or' -i .« ; ui
•^•j -S . —
- I
1 t- 1 -
°i °. -
CT rO «o' \n
r ~"l ~'
o
J in frt
A^9. -1 *
or ifl KV
i
J) 0
c~ — i l!i ~
, • i
t w <":
r sil i! *
1 ;
to o' ~
^ ~i •• in
± -, ~
Vo o —
£"••• —
fQ
£— -*9 *"~
, i
°° ^. £ _
i i •
P •'•'* —
' ol rO
! rO ' *! VO - .
1 I ^ vi — *"
i j
' U.
! i
1
1 !
•1 ••
l -
i
1 to
r!
o
•H
f« jj
! 0) -M
' B ; o
i .P j->
w w
Ig-tl
& 0! rf, ,
J' ci •• *°
| ! '
"o', —
CM
x o ' r.
•^ .-HO
H • -P -H
«,H n rt ' 4^
o -H K ' c)
CO *— i,
•*— ' O 4-5
a -H c:
0) O (.. 01
4J ^. 4J Cl
-
Iti t/l o t(
to -H f,
01 o TJ
« »-, 4J rH
O Pi C/l 10
1 -. o S ° Ul
o ,9-2i ^Sy^r ° cot/1 oo^o
- w - w ' --<,| ^ * fi; , ^ 1/5 In
1 !
o; *0 tv-
°; tC —
cxi ' \n f |O
•tO , ' N, -
, -: i ly
I v/i ; o ! o^
r^ i — c**
10. . , CM -
Oj i <0 K>
Q ' ' Vl |O
— j I 1 M ~
1
, ift| J' u>
01 o W
Oj . o fJ
i ; ^
1 :
m ' • i o" o
N : • — to
"O . CM —
! •
0 j j 0 , C-1
• i
or ! o:vp
o i : &• cy
,
0^ i , ° *
"i ! .cv'
| i ;
! ' '"^
OS ' cP! tJ
°i ' fO, |rt
"i ! «° ~
101! ! rj' f"
M.1 ' - 01
-u) O 4J E-i E-i
^: O £
o (=-.
Ci, fl
- c
0 0
O o
Vn l/i
— . i " J I 1 y 0
•Til loS^^'i ivj|a3-; r^iiriviln^
i • < i : l
1 i ! i
VP *si 9
IT) ' ^' '•^
^L...^:.1^":..
1 ^
, v. — — —
1 L 31
• - "I — , ~r '
0 J «3 0
!£ I i~ ^.
j J . ,_
•*' l 00 f>
"* : o r); i f~ t-j o-
O i co — • ' c~ in ^*
i » ' i
r . j i M 0) .
"to i " 0!~«:~o"
! ift o £ Ji ^, «:
.-. T -u°°1 .7i_ ^i i W
' ! ! ' -J
j
£ SI -: -,
in ' ' in
Ul ' Qi Q t/1
1 "~^ «^\' 1
C-j | -; -(
~ ' J 00 jj
y> ' j] «-; 1^
£ l i «.S N l
°0 — , cN|
: i !
•^i t-J oo. to i
^j- ' i 10 • «j' in
5 -_ 00 o 0
•* CP s
6 °° • — c^ ^ «f
. : , !
- : J f* j - ."
r>i r-: Sr ^ in ^
r -s ° J *t *
ui *"! — ' r1 vr> IP
i :
. . . ^ .... .
^ .. w' ET .5
r; ; C~ ; 0 *
—I : — «* ' (r f tr
o< ! . ,-! o. >«:
I ^- ' C E •—
r r, t-i i, fj r; o P ' ^ ; i
0)01 ^^Clfll-HtlOlf^-W 4J
OO 4->4->004->— f^Ofl 'C
dC cd o c c -H ^s^^oi UP.
OO ^^.OO'dOlJJ 3 pi 01
oohrcJa>oo^4Jrt'di-t T)I-IP-
C o, a, o ci f, 01 tn «i <,< P.
c« « -H r K CM x' o (> o w ^ ti •. o f-
01 t, rH Cl K
^ f< fr. EH P.
3
r-l
:io
o'
1 • •*.!
•*5 o
CO
1 -
«~J lo :
In lo i1
' i.
01
01
Cl f~
3 <0
599
-------
TABLE 3
RESULTS OF SECONDARY REACTANT TEST PROGRAM
Reactant Gas Flow Rate Stoichiometric Pressure S02 Removal
(SCFM) Ratio Drop (in H?0) Efficiency (%)
86.4
94.9
93.4
97.6
32.6
51.2
51.7
51.8
67.4
85.5
81.9
a
Lime
1445
790
782
1455
Dolomite
1430
1430
772
772
Precipitated Calcium
1431
778
773
1435
1:1
1:1
1:3
1:3
1:1
1:3
1:3
1:1
Carbonate
1:1
1:1
1:3
1:3
6.5
12.0
6.5
12.0
6.5
12.0
6.5
12.0
6.5
12.0
6.5
12.0
a Sample contaminated due to heavy foaming and carry-over.
NOTE:
1. Slurry concentration - I"o by weight
2. Particle size - no control due to use of commercially
available materials.
600
-------
°. r>
cr -
03
—
,J 0 0
in -
oO j
oo i;
5
if
Ir
ti
(5
.
In
0, 0
a o i
— r—
1,
1 o. , o. 0 ^ ~ | |
.1 , {Q C>- V>
0
•*
, CO
K
O
1 T4
t, -p
CO CO
, 0) tt)
1 E-< EH
O
CM
X
> 0
t, --I
I-. 4J
3 t,
r-l CD
CO P^
CO
o
*H
4^>
•H
•C)
1 O
' O
CO
„ (IS
'' C
hf
R
«H
tn
4^
(tf CS
t, ^
(1)
'*- V,
O1
J
p
r-d
(3
t
ID
P
[3
tt)
f-i
ature (Twb)
Mcantratlon (ppm)
t. O
0) LI
p,
1= CM
a) o
EH CO
ncentration (ppm) ,
O i
o
X1
§1
r
f
^
i
tn •—
§ a
,H hr
-p —
^H
•d a)
C -P
O 0)
0 (-
>! O
t. rH
r< It,
3
<— 1
IO
r
a
Q)
EH
• ft?
0
? s
>— fl)
3
Tj H
0) VH
CU <«
0)
fc W
0)
P It)
p. P
0 3
3J H
601
-------
m
2
(K
W
EH
O
H
EH
O
W
•-3
W
o
EH
W
W
fc
O
IS
a
w
o
e
u
£
i
o
a.
o
I
1
U-
I
i—
i
j;
§
• i
h
J
Ql
H
r.!.
,
1 *"
ift
i ct-
I1"
U
iS
fr
h
,T
H
i ^
I
• '
K
4)
u
0)
H
"
j
1
i
1
1
—
•«»•
t
f\
cr
0
5
^
.£•
tsi
-
•
"-"
o
6
0
I'l
~~
VI
• 0
o
^4
a
W
chlometrlc
^4
o
4J
l/J
1
1
1
1
~1
1
,
1
I
1
|
I
M.
J
jc
NuO
C
0
tH
01
t.
ry Concent
i.
3
i-l
t/}
i#-i .
i
1
-
O
0
o
0
o
o
*^'
_3
—
*— s
£l
t*
*+-*
empersture
H
»>-^i»
r~
* •
M
o
SM
o
•a-
C\l
»o
N
N
M
.
C
O
O
o
(N
*
t>
M
(XT
O
M
c-
i
r
0
^
e
P,
N^
C
0
^H
tl
(,
02 Concent
0)
r
i
,
'
—
i
i
•
i
•
i
^
I
)
R
P.
^
fi
+j
01
^
Ox Concent
^
•-
i
1
—
•A
U
d
^
--
f
r
O
£
~
0
—
— '
0
/^
-
vo
(0
s
?
K)
£1
0
"11
E
r?
•
Jo
—
«•»«.
/^
^?
..^
enperature
H
r-
r-
r
f
—
—
«
Q
JS.J
-
le
"r^
o
—
^•n
^Q
F?
P,
a
01
H
• -
~
Vi
rt^
•O
0
0
—
o
o
in
ui
"VJ
«
s
Vi
^
^>
CO
R
^^
f.
•H
n
Oo Concent
Vi
-•
1
i
i
i
i
\
i
-
i
i
i
(
\
a
P,
P,
c
o
.p
01
Or Concent
U-
^.'
ii
^:
Hopper
—
1
I
I
^
1
I
t
\
t
I
I
.
\
4>
(1
O
«
W
V
3
H
—
;
k
I
..
\
i
i
Tj
'
— ;
!
1'
\
1
','
1 ;
) j
j
* i
!l
\t
;,
||
[J
j>
L
D
I]
J
t J
'i
i|
602
-------
SULFUR DIOXIDE EMISSION CONTROL
FOR
INDUSTRIAL POWER PLANTS
Robert J. Phillips
Project Engineer
GM Manufacturing Development
GM Technical Center
Warren, Michigan
NOVEMBER, 1971
603
-------
SULFUR DIOXIDE EMISSION CONTROL
BY
WET SCRUBBING
Recent limitations on the emission of sulfur dioxide
have forced industrial users of coal to convert to
natural gas in order to meet the new codes. This has
been necessary because economic means of controlling
the emission of sulfur dioxide has not been forthcoming.
Many approaches have been offered for the control of S0»
from large utility boilers, but these approaches are,
in almost all cases, based on the recovery of sulfur in
some salable form. The small industrial coal user cannot
usually justify the expense of such recovery equipment for
the small quantity of sulfur involved.
604
-------
General Motors is a heavy user of coal for industrial
powerhouses consuming approximately 2.3 million tons/yr.
Because of the long term availability of coal and with
the higher cost and deteriorating supply of natural gas,
General Motors has made it a policy to continue to use
coal wherever economically possible and do it within
existing pollution control codes.
605
-------
Most of the boilers in the corporation range from
steaming capacities of 50,000 to 150,000 Ib/hr.
with an average plant steaming capacity of 250,000
Ib/hr.
Any SO- control process that might be selected had
to first meet two corporate criteria.
First - The process must provide an economic
incentive over conversion to an alternate
sulfur-free fuel.
Second- The prime function of the powerhouse is to
make steam. The SO2 control process cannot
be so complex as to compete with the steam
generating function of the plant. In other
words , the S09 control process must be
£*
nearly control and maintenance free.
606
-------
We made an evaluation of possible approaches to the
control of S0~. Based on some promising results from
tests conducted by the coal industry and subsequent
tests run by TVA, it was first decided to try a dry
additive injection system to reduce SO0 emission from
^*
our boilers. The dry additive injection system utilizes
limestone, dolomite or red mud, either fed in with the
coal or injected directly into the boiler. The pulverized
stone is calcined in the boiler and reacts with the S02
to form a solid calcium sulfate. This sulfate is then
collected by the dust collectors.
607
-------
40
E 30
h—
o
20
CSI
O
CO
o
C£
UJ
Q_
10
0
50
100
150
200
25C
PERCENT STOICHIOMETRY
This system presented a very economical approach.
Tests were conducted at the Chevrolet-St. Louis
powerhouse. These tests showed only a 30% reduction
in S09 but, at the same time, particulate emissions
were increased beyond code limits. Because of the low
efficiencies and increased dust loading, it was
concluded that the dry additive injection system is
inadequate.
608
-------
Flue
Gas -
Feed
H20 -
Co(OH)2
or
CaCO,
Scrubbed
Gas
li
Scrubber Feed
t
ber
r
i .
i
i
i
Scrubber 1
i
i
1
Thickener
*j
ill
Slur
Tank
1
I
"2°
'H
y
J
>.
Effluent
i
1-
Mixing
Tank
»•
1
H
T
Holding
LIME OR LIMESTONE SLURRY SCRUBBING
At the same time as our studies were taking place,
other studies involving wet scrubbing of boiler
flue gases with lime were also being carried out.
The obvious advantage of scrubbing with lime is that
CaSOo is an ideal form for the disposal of the sulfur
since it forms an inert solid suitable for land fill.
By far the overwhelming problem encountered with the
lime scrubber is excessive plugging of the system with
insoluble CaSCU. For this reason, we felt that because
of the high maintenance required, lime scubbing is not
the answer to SO^ control for industrial size boilers.
609
-------
SODIUM CARBONATE SCRUBBING
i
SCRUBBER
v v
MIXING
TANK
A more practical means of controlling SO- can be
utilized by substituting soda ash for lime in the
scrubber. The obvious advantage of this system is
the formation of a soluble product thus eliminating
plugging problems. Of course, the formation of a
soluble product is an unsuitable form for the sulfur
because a water pollution problem has now been created.
610
-------
Flue
Gas ~
FeeJ
H20 -
Ca(OH)2
Scrubbed
Gas
Scrubber Feed
t
ber
_-
Soda Ash
Make-up
i
Thickener
r
\ „
-v
411
Ca
Prec
Tank
2°
.W.
"] •
r
Mixing
Tank
i
»•
Cau
Wash
Water
CaCOo
SODIUM HYDROXIDE SCRUBBING WITH LIME REGENERATION
We still felt that wet scrubbing would offer an economic
means of controlling SO,, emissions in industrial boilers..
The optimum scrubber system should involve the best parts
of both the lime and soda ash system. For this reason,
we selected a system using caustic soda for scrubbing with
regeneration using lime. In this system a caustic solution
is sent through the scrubber. The sodium sulfite formed
is then sent to a regeneration tank where lime is added.
The sulfur is precipitated from the system in the form of
calcium sulfite and sulfate. In the process of precipitating
the sulfur, the lime also serves to regenerate the sodium
scrubbing solution. In this lime regeneration system, the
only product leaving the system is an inert cake.
611
-------
After some favorable laboratory scale tests, it was
decided to run a pilot plant of the system on an
actual operating boiler. A critical piece of
equipment in the system is the scrubber itself. All
of our boilers are equipped with dust collecting
equipment so that the scrubber would not have to act
as a primary dust collector. Therefore, it was decided
that a low energy scrubber would be preferred for the
system. Nevertheless, the scrubber would still be
exposed to an environment of fly ash and also it must
stand up to possible boiler upsets and soot blowing which
would increase the fly ash loading a considerable amount
for short periods of time. We finally decided that two
pilot systems should be run, one utilizing a scrubber that
could be classified as a gas absorber for maximum S0~
removal. The second scrubber was selected for its ability
to handle particulates and was considered to be marginal
in absorption characteristics.
612
-------
The first pilot system was installed on a powerhouse
boiler at Chevrolet-Cleveland. Arrangements were
made to use a 2800 cfm packed bed cross flow scrubber
for the tests. This scrubber utilizes a plastic
packing noted for its good absorption properties. The
scrubber contains a packing depth of four feet. 2800
cfm of flue gas (approximately 10% of the total flow)
is drawn out of the exhaust stack into a transition piece.
613
-------
The transition piece was positioned so that it points
directly into the gas stream and its diameter was
selected so that the gas velocity into the transition
duct approximates the stack velocity. This insures
that the scrubber is seeing a fairly representative
dust loading. The flue gas first enters a quench
chamber where the 580° F flue gas is cooled to 130° F
by spraying city water thru a series of baffle plates.
614
-------
The saturated gas then passes horizontally through
the four feet of packing. Scrubbing solution is
simultaneously being sprayed vertically through the packing.
The scrubber is plastic and controls have been installed to
insure that the scrubber never sees a high temperature. The
caustic scrubbing liquor is sprayed into a series of low
pressure full cone nozzles at a rate from 25 to 45 gpm.
615
-------
The spent liquor from the scrubber is sent to the
regeneration tank which contains both a mixing and
a settling section. Lime is added to the mixing
section using a lime slurry feeder. The sulfite
precipitate is filtered and dried using a vacuum
filter. The regenerated liquor is sent back to the
scrubber.
616
-------
The boiler being used for the study is a 80,000 Ib/hr.
spreader stoker equipped with a multiclone dust
collector. A coal with an approximate sulfur content
of 2% is used at the plant. The boiler operates with
very high excess air (over 100%) resulting in S02
scrubber inlet concentrations of about 1000 ppm. The
boiler is not equipped with an economizer or air preheater
resulting in a flue gas temperature of 580° F.
617
-------
S00 concentrations were monitored using an infra-red
£
analyzer. Test ports were located at both the inlet
and outlet of the scrubber. This monitoring capability
gave us the chance of chap'j, /,g variables and measuring
their effect immediately, thus greatly reduced the time
necessary to obtain the needed scrubber design data.
The first tests on the scrvbber were run to determine
scrubber performance under different operating conditions.
The variables with the most affect on scrubber performance
were the liquid and gas rates to the scrubber.
618
-------
REMOVAL OF S02 VS. GAS RATE
o
40
• 39
38
o
37
ct:
CM
O
oo
36
L- 2000
1400
1600
1800
2000
2200
SCRUBBER GAS RATE (LB./HR.FT )
In general, the scrubber performance increases as the
gas rate is reduced in a scrubber. Our tests have
shown that at very high gas rates 2150 Ib/hr. ft.2,
this characteristic no longer holds. It was found that
the scrubber efficiency actually increased from under
37% per foot of packing to 38.5% as the gas rate was
increased from 1880 Ib/hr. ft.2 to 2150 Ib/hr. ft.2.
It is theorized that at a relatively high gas flow, the
increased turbulence that results more than compensates
for the reduced contact time in the scrubber.
The significance of this result is that high turndown
rates on the gas side can be accomplished without loss
in scrubber efficiency. This is of considerable
advantage for fluctuating boiler loads.
619
-------
PERCENT S02 REMOVAL
PER FOOT OF PACKING
LENGTH (FT. )012345678910
S02 CONC. 1000 600 360 216 130 78 47 28 16 10 6
% REMOVAL 0 40 64 78 87 92 95 97 98 99 99.4
SINCE THE FORSEEABLE REGULATIONS WILL REQUIRE LESS THAN 90% REMOVAL
THE PRELIMINARY ESTIMATE OF LENGTH IS FIVE FEET.
Our conclusions are that the packing efficiency will
be approximately 40% / ft. of packing. In the light
of forseeable emission regulations, a 90% S02 reduction
efficiency will be adequate. This will require that a
5 ft. bed of packing be used. It is interesting to
observe that if 99% efficiency is required, the length
of the scrubber packing must be doubled to obtain this
9% increase.
620
-------
EFFECT OF CAUSTIC CONCENTRATION
ON S02 REMOVAL EFFICIENCY
o
SULFITE-BISULFITE REGION
40 h
<
>
o
CSJ
O
20
10,
_L
_L
_L
4567
Ph OF SCRUBBER OUTLET
SULF1TE REGION
10
Another factor influencing the scrubber operation is
control of caustic in the system. Chemical conversion
of the caustic in the scrubber takes place in the
following manner:
1st; NaOH + SO,
4.
pH 11.5
Na2S03 + H20
pH 9.0
2nd:
SC>
pH 9.0
NaHSC>3
pH 4.0
The reaction can be controlled at any point merely by
supplying caustic to maintain a selected scrubber water
outlet pH.
621
-------
As can be seen, there is no loss of scrubbing efficiency
until the scrubber outlet pH falls below pH = 6.0,
which is greater than 50% into the bisulfite region.
There is considerable advantage in scrubbing into the
bisulfite region because of the limitations in producing
a high caustic concentration in the regeneration system.
If the system operates 50% into the bisulfite region,
then the caustic requirements become 1.5 moles NaOH/mole S
instead of 2.0 moles NaOH/mole S on a sodium sulfite cycle.
This reduction in caustic requirement to the scrubber can
be accomplished without loss in scrubber efficiency. The
advantage of such operation will become more apparent
after a discussion of regeneration chemistry.
622
-------
Regeneration
Complete regeneration of the spent scrubbing liquor
requires that the captured sulfur be precipitated
from solution as calcium sulfite and calcium sulfate,
These reactions are shown in the following equations:
Ca(OH)2 -»• 2NaOH + CaSC>3 ,
2) Na2 S04 + Ca(OH)2 -»• 2NaOH + CaS04 ,
It has been found that regeneration of sulfites
-(equation 1) presents no serious problems. Sulfites
are formed by the absorption of S0~. The relative ease
of sulfite regeneration results from the fact that the
solubility product (Ksp) of calcium sulfite is very low
(6.25 x I0~8(gm moles) 2 /£2 @ 1150 F)t when the
product of the dissolved calcium and sulfite ion concentration
exceeds the solubility product, calcium sulfite is
precipitated.
3) (ca ++) x (S03 =) = 6.25 x 1CT8
If only sulfites needed to be removed from solution,
regeneration would be relatively simple. A portion of
the sulfur absorbed in the scrubber appears as sodium
sulfate. The percentage of sulfate formation may vary
from as low as 10% of the sulfur absorbed to as much as
30% depending on the amount of oxidation that takes place.
623
-------
Sulfates do not regenerate as readily as sulfites.
The reason for this difficulty is that the
solubility product of calcium sulfate is 3.7 x 10
times as great as calcium sulfite. This high calcium
sulfate solubility shifts the reaction in favor of
sulfate over caustic formation. In order to establish
an equilibrium favoring caustic formation, an excess
of sulfate ion must be present in solution. This can
be done by keeping an excess of sodium sulfate in
solution. This excess Na2SC>4 in solution is necessary
for the precipitation of CaSC>4 and formation of a
dilute caustic solution.
624
-------
Design of Regeneration System
To effectively operate the regeneration system, there
is a minimum sodium requirement needed to precipitate
sulfates from the system and form a suitable concentration
of caustic. The minimum sodium required to remove a
gm mole of sulfur from the system can be summarized as
follows:
1) Free sodium (NaOH, Na2 003) required to
react with absorbed sulfur (SCU / SOo)
in scrubber. This includes 2.0 moles
Na/mole S03 plus,
a) 2.0 moles Na/mole SO*, (sulfite cycle).
b) 1.0 moles Na/mole SO- (bisulfite cycle).
2) Sodium sulfate required to reach a favorable
solubility product of CaSO* in solution.
625
-------
EQUILIBRIUM CAUSTIC FORMATION
IN Ca(OH>2 - Na2$04 SOLUTIONS @ 120°F
o
t—
<
I—
z
LU
O
z
o
.2
.15
.1
^£ .05
*—
to
3
<
0
I
I
I
I
I
1
.4 .6 .8 1.0 1.2 1.4
SODIUM CONCENTRATION (gm. moles/liter)
1.6
1.8
Tests have been conducted to determine the equilibrium
caustic formation in sodium sulfate - calcium hydroxide
systems. The figure above shows the maximum amount of
caustic that can be formed at 120° F in regeneration
solutions as a function of sodium concentration. As can
be seen, increased Na+ (indicating increased Na~S04)
concentrations decreases the Ca++ concentration allowing
more OH~ to exist at equilibrium. Beyond 1.0 molar
sodium the additional amount of OH~ that can be formed
does not increase appreciably.
This equilibrium data shows that no more than a 0.14 molar
caustic solution can be produced in 1.0 molar sodium
solutions.
The figure does show the limits of caustic that can be
formed in the regeneration cycle. It does not indicate
the rate of caustic formation which is needed for practical
design of the regeneration system.
626
-------
EFFECT OF LIME AVAILABILITY ON THE RATE OF CAUSTIC FORMATION
IN 0.5 MOLAR Na2$04 SOLUTIONS @ 120°F
.13
s
= .11
«
o
t .09
ce
I
.07
u
2 .05
t/>
ID
<
O
0.25
Hydroxide
Available in
Lime gm.moles
OH'/liter
0.125
0.07
6 8 10 12
MIXING TIME (MINUTES)
14
16
An additional series of tests were run to determine the
rate of caustic formation in 0.5 molar Na^SO, (1.0 M Na+ )
solutions.
These tests were run by adding a given amount of
calcium hydroxode to 0.5 M Na-SO, solution (a) 120° F
and drawing samples at 2 , 5 and 10 minute mixing intervals,
The samples were immediately filtered and titrated with
0.08 N HpSO.. In all, three different initial lime
concentrations were evaluated.
The figure above shows that the rate of lime conversion
can be maximized if lime is fed at a concentration near the
equilibrium hydroxide concentration (0.14 M). In the
0.125 M OH~ lime feed, 80% conversion of the lime was
obtained in 5.0 minutes and 0.1 M OH- was achieved. When
0.25 M OH - lime was added, 44% lime utilization was achieved
in 5.0 minutes and only 0.11 M OH - was formed.
627
-------
These tests concluded that:
1) Ca(OH)2 should be fed at or near the
equilibrium caustic concentration of
0.14 gnu moles OH~/liter.
2) Lime mixing efficiency can be 80% in
5.0 minutes. Recycling of the sludge
will have to be evaluated to determine
if increased utilization can be obtained.
3) With 5.0 minute mixing, it is felt that
0.1 molar NaOH can be produced for use
in scrubbing.
628
-------
A main advantage of this process is that S02 can be
removed in the scrubber as a soluble salt. This allows
us to use a low energy scrubber for the process. Any
calcium carry over from the regeneration cycle will
result in deposition in the scrubber. This carry over
must be avoided if the process is to retain its
advantages over lime slurry scrubbing.
Calcium plugging can occur from two sources -
1st - Carry over of suspended solids.
2nd - Precipitation of dissolved calcium.
Suspended solids can be handled by a gravimetric
settling system. A more serious kind of plugging can
result from carry over of dissolved calcium into the
scrubber. The dissolved calcium is at its solubility
product with respect to sulfite and sulfate. Exposure
629
-------
to S02 an^ SO- in the scrubber will cause a small
amount of precipitation of CaSO_ and CaSO, in the
scrubber if free hydroxide is allowed to enter the
scrubber.
We have evaluated one means of eliminating
precipitation in the scrubber by using Na2COo make up
as a water softener prior to entry into the scrubber.
We were successful in reducing the calcium concentration
from 400 ppm to 250 ppm. This, however, will not
completely avoid precipitation of sulfites in the scrubber.
Another means of eliminating precipitation in the scrubber
would be to operate the scrubber on a low pH loop. The
scrubber recycle tank could then be used to precipitate
dissolved calcium leaving the recausticizing section. To
operate under these conditions, a scrubber with excellent
gas absorption characteristics and high liquid to gas flow
rates would have to be used. We are now evaluating the
possibility of using such a mode of operation.
Pilot Study of Regeneration System
It was found in the pilot plant that production of
0.1 molar NaOH is easily obtainable in 1.0 molar
sodium scrubbing solutions at 120° F. During 130 hours
of pilot operation no difficulty was encountered in
recausticizing the spent scrubber solution. It did
become apparent that three variables had a significant
influence on the concentration of caustic produced.
1) Mixing time
2) Degree of agitation
3) Rate of lime addition
630
-------
The mixing section of the solids contact reactor
utilized a slow 16 rpm paddle for mixing. It
was found that 30 minutes of mixing was required
to obtain the desired caustic concentration. It
was also found that lime usage was only 70%.
Simulated regeneration tests using near stoichiometric
lime addition and high speed flash mixing showed that
the lime usage was increased beyond 80%. Also, the mixing
time required to produce the same quality caustic was
reduced from 30 minutes to 5 minutes.
The amount of lime present in the reaction zone also was
shown to have an effect on the concentration of caustic
produced as was demonstrated in the laboratory tests.
Although higher amounts of lime in the reaction zone
produced higher caustic concentration, lower lime
utilization was realized.
To further discourage the use of high lime concentrations
in the reaction zone, it was found that the spent sludge
was chemically inactive even though it contained as much
as 30% unreacted lime. Therefore, high lime concentration
in the reaction zone would only reduce lime usage because
sludge recycle does not appear to improve lime usage.
The final test conclusion was that near stoichiometric
lime addition, together with high speed chemical mixing,
must be used to optimize lime usage.
631
-------
Cake Analysis
Na2S04 2.6 %
Na2CO3 1.13%
NaHCO3 0.73%
A12(S04)3 1.84%
Ca(OH)2 0.16%
CaS04.2H20 39.30%
CaS03.2H20 2.93%
CaC03 4.85%
H0 46.5 %
Sludge Settling & Filtering
The CaS03 - CaS04 precipitate settled very rapidly.
At a clarifier rise rate of 1 gpm/ft. no turbidity
was noticed in the clarified liquid. The thickened
sludge contained 20% solids which was taken to 50%
solids in the rotary drum vacuum filter.
An analysis of the filter cake was made. This analysis
gives only an indication of the content of the cake and
cannot be considered representative.
A considerable period of time elapsed before the cake
was analyzed. The degree of sulfite oxidation and
carbonation cannot be determined. The high amount of
gypsum does indicate that high oxidation takes place in
the scrubber. There is no explanation for the high
Al?(S04)3 content in the filter cake. Silicon was
632
-------
also identified in significant quantities in the
sample indicating that aluminum may be present in the
form of aluminum silicate rather than sulfate.
This cake analysis also indicates better than 80%
lime usage where system material balances indicated
only 60 to 70% usage.
An evaluation of the settling of the precipitated
CaCCs in the softener was not made in the pilot plant.
Instead, an in-line filter was used to remove the
precipitated CaC03.
Livonia Scrubber
Concurrent with these tests being conducted, another
scrubber is being tested at the Chevrolet-Livonia
powerhouse. This unit is a wet cyclone scrubber. It
is primarily known for its ability to handle particulate
matter. The unit has the same capacity as the packed
scrubber unit and operates with a multiple stage impingement
wash of the flue gas. The pilot scrubber we are using
contains four stages with a capacity for two additional
stages.
633
-------
The scrubber is located on the exhaust of a 80,000 Ib/hr.
spreader stoker equipped with a multiclone collector. The
boiler has an economizer so the scrubber inlet temperature
is about 350° F.
The gas is cooled and saturated in the first scrubbing
stage rather than using a pre-guench as in the packed
scrubber. There is also no regeneration cycle in this
system since it was decided that all necessary information
for regeneration could be gained with the packed scrubber.
We mainly are interested in seeing whether this scrubber,
with marginal gas absorption characteristics, is capable of
removing sufficient amounts of SO-.
Sulfur dioxide removal tests were run on the 4 stage
scrubber. Results showed an 87% SO- removal efficiency.
This efficiency was higher than expected. The scrubber
pressure drop was 5.0" H-0 during the tests. An additional
two scrubbing stages is now being installed and the scrubber
will be re-tested. It is hoped that the additional two
stages will bring the overall SCU removal efficiency above 90%.
634
-------
BENEFITS
1. Greater than 90% sulfur dioixde
removal efficiency.
2. Low energy scrubber (3" - 7" H^
3. Soluble scrubber (no plugging)
4. Inert by-product
5. Economically feasible
It is our conclusion that the caustic-lime regeneration
system presents a practical means for control of SO2-
Efficiencies in excess of 90% are easily obtained with
standard low energy scrubbing equipment. The scrubber
is not in danger of plugging as in a lime system because
the lime is restricted to the regeneration tank. The
sulfite cake formed can be readily disposed as sanitary
land fill.
635
-------
PROCESS ECONOMICS
Installed Capital
Costs $ 3.20/lb. steam/hr.
Operating Costs $ 3-4/ton of coal
An economic evaluation of SO- control methods was
t£t
made. The results indicate that the caustic-lime
regeneration shows favorably over other possible
alternatives. Fixed capital costs for the system
are estimated at about $3.20/lb. steam. Operating
costs will be approximately $3.00 - $4.00/ton of coal,
636
-------
Another benefit of the system is that particulate
emissions now below code in the 0.2 - 0.3 grains/scf
range will be reduced by 70% +.
In our tests we have found no appreciable formation of
nitrite or nitrate ion in water analysis. We have
concluded that oxides of nitrogen are not being absorbed
in the system.
To sum up — the conclusions from our work are that we do
have a workable and economic process for the control of
sulfur dioxide emissions from coal burned in industrial
sized power plants. This should be of considerable benefit
to many industries in the very near future, and it certainly
will be of great benefit to our environment.
637
-------
o
CO
E
11
O
E
+
-------
SULFUR OXIDE CONTROL AT THE COPPER SMELTER
by
Ivor E. Campbell
Technical Director
Smelter Control Research Association
and
James E. Foard
Senior Smelter Engineer
Metal Mining Divisions
SECOND INTERNATIONAL LIME/LIMESTONE WET SCRUBBING SYMPOSIUM
New Orleans, November 8-12, 1971
639
-------
SULFUR OXIDE CONTROL AT THE COPPER SMELTER
Although the basic chemistry of t-he principal reactions in various
sulfur dioxide control processes is the same regardless of the source
of sulfur dioxide, conditions prevailing at the copper smelter relevant
to sulfur dioxide removal processes differ in several significant respect
from those prevailing at other industrial sources of sulfur dioxide.
This paper reviews the state of the art of sulfur dioxide control
as applicable to the copper smelter and the Smelter Control Research
Association1s-pilot-plant program. Some of you may not be familiar
with the Smelter Control Research Association. The Association was
founded and incorporated as a nonprofit association by the eight primary
copper producers of this country for the specific purpose of developing
processes for removal of sulfur oxides and particulates from smelter
stack gases and more especially for conducting pilot-scale investigation
of the removal of sulfur dioxides from reverberatory furnace off-gases.
The reasons for emphasizing the reverberatory gases will be discussed
later.
The eight founding companies of the Smelter Control Research
Association are: American Smelting and Refining Company, The Anaconda
Company, Cities Service Company, Copper Range Company, Inspiration
Consolidated Copper Company, Kennecott Copper Corporation, Newmont
Mining Corporation, and Phelps Dodge Corporation. These companies
account for all of the primary copper smelter production in the United
States.
Results of the Association's experimental programs will be published
and will be available to the public. It should be emphasized that the
640
-------
programs of the Association supplement extensive in-house development
work being conducted by the individual copper companies.
As background for our discussion of sulfur dioxide control as
applied to copper smelters, it will be helpful to first describe, in
general terms, a typical copper smelter operation and to note the opera-
tions in which sulfur dioxide is generated and the conditions under which
it is formed.
To simplify the description reference will be made principally to
the flow of the copper-bearing materials and of the sulfur.
Most copper ores today are much too low in copper content to be
processed directly. Therefore, it is necessary to separate the copper-
bearing minerals from other materials in the ore in which they occur.
After crushing and grinding of the ore, the copper-bearing minerals are
concentrated by a flotation process. The composition of the concentrate
varies from smelter to smelter, but the copper content is usually in the
range of 10 to 30 % and the concentrate contains, on an average, a pound
of sulfur or more for each pound of copper since it is produced from
sulfide ores. The concentrate also contains a number of trace consti-
tuents, some of which are known to affect certain of the sulfur dioxide
control processes, particularly those involving the use of catalysts.
Smelters throughout the world rely on oxidation processes to remove
the sulfur and this oxidation generates sulfur oxides. A typical smelter
produces approximately two pounds of sulfur dioxide for each pound of
copper, since a pound of sulfur forms two pounds of sulfur dioxide.
641
-------
Foreign smelters employ flash, electric, and blast furnaces, as
well as reverberatory furnaces; but our domestic practice, which has
been keyed to the available raw materials, is based on the use of
reverberatories. At some smelters, the concentrate goes through a
roasting operation before being fed to the reverberatory furnace.
The roasters, typically fluid bed reactors, generate a relatively
uniform, concentrated stream of sulfur dioxide, i.e., 10-15 % sulfur
dioxide. In most cases, the concentrate is fed directly to the
reverberatory furnace where it is fused in an oxidizing atmosphere,
forming a slag and a molten mixture of copper and iron sulfides known
as matte. Approximately one-third of the sulfur in the concentrate
is eliminated at the reverberatory furnace as gaseous sulfur dioxide
which is discharged to the stacks. The matte is discharged from the
reverberatory furnace into ladles in which it is transported to con-
verters o In the converters, the matte undergoes further oxidation,
additional slag is termed, and the balance of sulfur is discharged
as sulfur dioxide. The converter product is approximately 97 to 98 %
pure, and is known as blister copper. The blister copper is further
refined to produce the high purity product required for industrial
applications.
The reverberatory furnaces are large, rectangular, refractory
chambers, usually in excess of 100 feet long and 30 feet wide. They
are fired with gas, oil or coal. The reverberatory furnace off-gases
contain, on an average, around 1 % sulfur dioxide, along with a variety
of trace constituents as noted earlier. The gases, on leaving the
642
-------
furnace, are usually in excess of 2000°F, but a feed stream from a
typical fluesystem would probably be on the order of 400-500°F.
The converters are cylindrical, refractory-lined vessels, usually
30 feet long and 13 feet in diameter, although a few larger units are
now in use. The operation of the converter is cyclic, and the con-
centration of the sulfur dioxide in the exhaust gases varies widely
during the operation. The average sulfur dioxide content of the ex-
haust gases during the blowing operation is usually in the range of
4 to 6 %. This variability in sulfur dioxide concentration is a
definite handicap in the sulfur dioxide control process. Smelters
operating sulfuric acid plants try to program their converter opera-
tion, coordinating the cycles of converters in order to provide a more
uniform flow of sulfur dioxide to the acid plant. This objective is
difficult to attain and is a significant constraint on the smelter
operations and does not, in any event, produce a highly uniform flow
to the acid plant.
It will be seen, then, that two distinct levels of sulfur dioxide
concentration occur at the copper smelter. The converters and roasters
produce off-gases relatively concentrated in sulfur dioxide, averaging,
in the case of the converters, around 4 - 6 % sulfur dioxide. The
reverberatory furnace produces a much more dilute gas, ranging from a
fraction of 1 % to somewhat in excess of 1 % sulfur dioxide.
Technology is commercially available for converting the nulfur
dioxide in the more concentrated streams, i.e., from the converters
and roasters, to sulfuric acid. Seven acid plants are in operation
in the copper industry today, and additional plants are either under
643
-------
construction or in the planning stage. There is, however, no commer-
cially proved technology available for the removal of sulfur dioxide
from the more dilute gases produced in the reverberatory furnace.
The Smelter Control Research Association program is, accordingly,
directed as noted earlier, specifically to the development of processes
for removal of sulfur dioxide from reverberatory furnace off-gases.
It should be noted that although reverberatory gases are more
dilute than converter gases, they are on an average at least 3 to 5
times as concentrated in sulfur dioxide as typical utility stack gases.
In the wet-lime and wet-limestone scrubbing processes, this higher
concentration represents a significant difference, particularly with
respect to control of scale formation in the scrubbing system.
It is fortunate that there are strong similarities in the problems
involved in removal of sulfur dioxide at the three concentration levels
we have referred to, i.e., the 0.1 to 0.3 % and lower level prevailing
generally at utility and industrial boilers, the 1 % level at the
reverberatory -furnaces, and the 4 to 6 % and higher levels found at
the converter and the roaster. However, sulfur dioxide removal systems
for each application present distinct and separate problems. Verifica-
tion of any process under consideration must, therefore, be carried out
under conditions valid for the proposed application. Therefore, pilot-
scale tests on actual smelter gas are considered essential before the
very substantial commitment that would inevitably be involved can be
made to a given control process. For this reason the Association's
pilot-plant tests, as we noted earlier, are to be conducted on actual
smelter gas.
644
-------
In establishing priority to be assigned to investigation of various
possible alternatives for removal of sulfur dioxide from low-strength
gases at the smelter, the stringent time schedule being imposed on the
industry for sulfur oxide control dictated that priority be assigned to
alternatives deemed closest to commercial availability, even though
other alternatives might provide a more satisfactory long-range solution.
The time schedule and the economic facts of life dictated that the
control system be compatible with existing facilities, i.e., that the
system be what is commonly referred to as an "add on" system. After
reviewing more than 100 processes or variations of processes, the Smelter
Control Research Association selected wet-limestone scrubbing for its
initial pilot-plant study on the basis that this process, which has
been investigated extensively by the Environmental Protection Agency
and by the utility industry, was closest to commercial availability.
A pilot plant has been constructed and is in operation at the McGill,
Nevada smelter of the Kennecott Copper Corporation. The pilot plant
is rated at 4000 SCFM and is of sufficient size to permit extrapola-
tion to full commercial scale. (The extrapolation would be approximately
20 to 1.)
The pilot plant has been designed with a view to optimum versatility
and will be used in testing wet-lime and sodium sulfite scrubbing as
well as limestone scrubbing. In addition, lime regeneration of the
scrubbing lig_uid in the sodium sulfite system is to be evaluated. The
tests on the sodium sulfite system will, hopefully, with information
available from other operations on sulfur dioxide desorption permit
evaluation of the sodium sulfite absorption-desorption process vis-a-vis
645
-------
the throw-away process.
The plant employs a Turbulent Contact Absorber in series with a
venturi. Limestone-water slurry flows countercurrent to the gas
stream from the Turbulent-Contact Absorber to the venturi. A bleed
stream from the venturi is sent to a centrifuge where the reaction
products are separated and discharged, the liquid being returned to
make-up tanks for recycle. Operation with lime will be essentially
the same as with limestone.
When sodium sulfite is used as the absorption medium, the bleed
stream from the venturi will be sent to causticizing tanks where
calcium sulfite will be precipitated on addition of lime. The product
will then be sent to the centrifuge for separation of the calcium
sulfite which will be discharged as a slurry, as in the limestone
scrubbing system. The liquid from the centrifuge will be returned to a
holding tank where make-up caustic will be added to provide a fluid
suitable for recycling.
The pilot plant has been in operation with limestone approximately
two months, on a 24-hour per day, 5-day per week basis. In initial
operation, the scrubbing liquid was not recycled. Once-through
operation was employed to minimize plugging of equipment with reaction
products. This was done to facilitate operations while the operators
were familiarizing themselves with the system. The plant is now a
closed-loop operation.
646
-------
It is too early to present any final conclusions with respect to
;he effectiveness of the system in smelter gas control. Results to
3ate may be summarized as follows:
The average sulfur dioxide recoveries have been approximately
'0 per cent from 0.7 per cent sulfur dioxide streams and approximately
>5 per cent from 1.0 per cent streams.
A high calcium limestone has been used in all tests. Bench-scale
:ests of a number of high calcium limestones indicated that the lime-
tones tested were equally reactive. Bench-scale tests of additional
.imestones are planned.
Limestone usage has been varied from 1.2 times stoichiometric
.o 2.0 times stoichiometric.
Both 90% minus 200 mesh stone and 95% minus 325 mesh stone have been
sed in the pilot plant. Any difference in the reaction rates with the
wo grinds is apparently masked by other process variables. Although
ne might anticipate increased reactivity with the finer grind, no
ignificant improvement in sulfur dioxide recoveries has been noted to
ate with the use of the finer grind.
The sulfur dioxide recoveries do appear to decrease as the
oncentration of the sulfur dioxide as the feed stream increases.
In this regard, it should be noted that the concentrations of
ilfur dioxide in the feed stream varies and extended runs are required
i order to obtain meaningful data. Over shorter periods variations in
he gas concentration may mask the effect of other variables.
647
-------
Our current efforts are directed to establishing the factors
controlling sulfur dioxide removal efficiencies in order to optimize
sulfur dioxide recovery. Development of techniques for effective
control of scale formation is also a major objective.
We would like to emphasize again that there are significant
differences in the problems involved in sulfur dioxide control at the
smelters and at other industrial sources of sulfur dioxide.
The gas flows are generally lower than at the utilities. A
typical commercial unit on a reverberatory furnace would handle
approximately 80,000 SCFM.
The sulfur dioxide concentration of reverberatory gas is much
higher, averaging around 10,000 ppm, i.e., 1%, and is much less uniform
than that of industrial boilers.
The total dust loading at the smelter is lower than that at many
utility boilers because of the lower gas volumes, but the concentration
is somewhat higher than at typical coal-fired boilers. Trace constituents
of the smelter gas stream will not only be different than at the
utilities but will vary from smelter to smelter. These constituents
may have little or no effect on some control processes but are known to
affect others, and their possible effect must be determined in actual
pilot-plant tests.
The percentage of o::ygen in reverberatory gas streams is highly
variable, and is normally higher than that in exhaust gases from
utility boilers.
648
-------
The gas flow from the converter is cyclic. VThile the gas flow
from the reverberatory is much more uniform, it, too, is subject to
variation.
From the standpoint of disposal of waste products, the smelters
have, in most cases, for the moment at least, available sites to establish
sludge ponds, and this is not as serious a problem as it would be in
more heavily populated areas. However, the problems involved in handling
the very substantial quantities of raw materials consumed in the control
process operation and the tonnage of waste products that must be
disposed of do present a substantial materials handling problem.
YJhile the situations at the smelter, with respect to disposal of
the throw-away product, may be more favorable than that at the utilities,
the situation with respect to disposal of potentially marketable
products such as sulfuric acid and sulfur is much the reverse, since
the smelters, in many cases, are remote from potential markets.
As noted earlier, the Smelter Control Research Association
supplements extensive in-house developments by the individual Member
Companies. A few of these may be briefly mentioned:
1. A 1,000 SCFM ammonia scrubbing plant was operated for about
eight months at a copper smelter to determine whether the
process, which had been in successful operation for a number
of years at a lead-zinc smelter was, in fact, adaptable to
copper reverberatory gas.
The sulfur dioxide recovery achieved in the pilot plant was
in accord with recoveries from available theoretical data as
well as from prior practice. The problems encountered were:
(1) aggravated corrosion; (2) high sensitivity to irregular
draft conditions; (3) a stack discharge of highly visible
particulate matter which required additional collection
649
-------
devices; and (4) requirements for high volumes of low
temperature water, an extremely important consideration in
many smelter locations. Perhaps the most compelling factor,
however, is that the operation of a full-scale plant would
produce large quantities of soluble ammonium compounds which
would create disposal and pollution problems of their own
unless they had utility as usable or salable products.
2. A 400 SCFM pilot plant utilizing sodium citrate as the
absorbent for sulfur dioxide was operated for about six
months at another copper smelter. The process calls for the
sulfur dioxide loaded absorbent solution to be treated with
hydrogen sulfide to form elemental sulfur, part of which
would be reacted with methane to make up the hydrogen sulfide
requirement. The pilot operation, which included only the
absorption and regeneration steps, was beset with mechanical
problems, and solution losses were incurred which prevented
reliable evaluation of citrate consumption, an obviously
important parameter. Recovery of sulfur dioxide from the
reverberatory gas stream ranged between 90 and 99 per cent.
The .hydrogen sulfide generation step in the process was not
investigated.
3. Another copper smelter is currently installing an industrially
sized operating module to explore the commercial feasibility
of absorbing the sulfur dioxide in reverberatory gas with
dimethylanaline (DMA), and springing the sulfur dioxide in
highly concentrated form to feed an acid plant, or perhaps a
sulfur plant. DMA absorption has been successfully demonstrated
650
-------
in small lead smelter applications treating gases above
3.5 per cent sulfur dioxide. The unit under construction
will be treating a much leaner gas and is expected to be
on-line by mid-1972.
4. A $1.7 million sulfur reduction pilot plant designed to
produce 8 to 20 tons of elemental sulfur per day has been
recently placed in operation as a joint venture of two copper
smelting companies. Its main objective will be to perfect a
suitable process for reacting sulfur dioxide with reformed
natural gas to form elemental sulfur. The pilot plant will
operate on gases ranging from 12 to 100 per cent sulfur
dioxide.
5. A number of smelters have conducted bench-scale tests or
even operated small pilot plants to e;:plore the potential
of absorbing sulfur dioxide in naturally occurring or
industrially produced materials readily available to them.
Such waste products as fly ash concentrate tailing pulps,
furnace slags, and others have been tried. In general, the
sulfur dioxide loading capacity of these sorbents is not
high', and their usage would require prohibitatively large
tonnages unless they were effectively regenerated, and this
step has not been successfully demonstrated.
In addition to their work in the five areas mentioned, individual
smelters have been exploring, in pilot plants already completed or
being planned, changes in the smelting process itself by which pollution
abatement may be furthered. This test work relates particularly to the
development of methods of continuous smelting, which would greatly
.enhance the efficiency of appurtenant sulfur recovery processes.
651
-------
In addition, the potential of electric smelting is being explored
by several companies and a 7500 KVA electric furnace is presently being
installed at a large chemical plant with a small copper smelting annex.
It is evident that the copper smelting companies are on the move
in their efforts to control sulfur o::ide eraissiona.
652
-------
PROTOTYPE AND FULL SCALE TESTS
H.W. Elder, Chairman
Participants:
James Jonakin and James Martin
E.G. McKinney and A.F. Little
M. Epstein, F. Princiotta, R.M. Sherwin, L. Szeibert, I.A. Rabei
R.M. Sherwin, I.A. Raben, and P.P. Anas
J.D. McKenna and R.S. Atkins
Gerhard Hausberg
J.J. O'Donnell and A.G. Sliger
Tsukumo Uno, Masumi Atsukawa, and Kenzo Muramatsu
Lyman K. Mundth
J.H. McCarthy and J.J. Roosen
Robert R. Padron and Kenneth C. O'Brien
J.A. Noer and A.E. Swanson
J.W. James
D.T. McPhee
J.F. McLaughlin, Jr.
D.C. Gifford
H.P. Willett and I.S. Shah
653
-------
PROTOTYPE AND PULL SCALED-TESTS
(PARTS I, II, AND III)
Second International Lime/Limestone Wet Scrubbing Symposium
New Orleans, Louisiana
November 8-12, 1971
H.W. Elder, Chairman
SUMMARY
The increasing interest in preserving a quality environment
in this country and abroad has caused an accelerating movement toward
enactment of legislation to limit discharge of undesirable materials to
the atmosphere. The existing and pending laws have started a somewhat
frantic search for methods to control emission of materials classified
as pollutants. The participants and attendees at the Second International
Lime/Limestone Wet Scrubbing Symposium assembled for a critical analysis
of one of the more promising methods for removal of S02 from stack gases.
The session of the symposium on prototype and full-scale appli-
cation of the process covered topics from plans for full-scale installa-
tions to availability of raw materials. The main theme that evolved as
the session progressed was the concern over reliability of the scrubber
system as compared with the power plant. The problem of waste disposal
was a popular subject and, as might be expected, economics was a con-
troversial issue.
Participants in the prototype and full-scale session are
listed below.
Name Company
Masumi Atsukawa Mitsubishi Heavy Industries, Ltd.
A. 0. Blatter Union Electric Company
D. C. Gifford Commonwealth Edison Company
J. W. James Ontario Hydro
Jim Jonakin Combustion Engineering, Inc.
Ulrich Kleeberg Gottfried Bischoff KG
L. K. Mundth Arizona Public Service
J. H. McCarthy The Detroit Edison Company
J. D. McKenna Cottrell Environmental Systems
B. G. McKinney Tennessee Valley Authority
D. T. McPhee Kansas City Power and Light
J. A. Noer Northern States Power Company
J. J. O'Donnell M. W. Kellogg Company
R. R. Padron Key West Utility
F. T. Princiotta Environmental Protection Agency
R. M. Sherwin Bechtel Corporation
H. P. Willett Chemical Construction Corporation
654
-------
Use of lime or limestone scrubbing for control of S02 emission
in stack gas is of interest because it is relatively simple compared with
processes that produce salable products, both fly ash and S02 may be
removed, limestone is available in most parts of the country, earlier
work has shown the process is feasible, and the complexities of marketing
sulfur products are avoided. A better understanding is needed of factors
that influence chemical scaling, solids deposits, slurry settling rates
or mechanical dewatering, erosion, and mist carryover.
Limestone can be added to the system in two ways—either addition
in the boiler where the carbonate is decomposed or addition directly into
the scrubber circuit. The calcined material is more reactive and should
result in a lower raw material requirement. However, there is evidence
that scaling is more likely to occur during closed-loop operation with lime
than with limestone. Also, occurrence of calcium-containing deposits in the
convection pass was reported for operation during limestone injection. It
is likely that relatively wide tube spacing would be required to prevent
plugging. The risk of tube fouling could be avoided by addition of cal-
cined lime directly in the scrubber, but the cost of lime is high and the
scaling tendencies are a deterrent. However, the process apparently has
been developed successfully in Japan and a 20-mw prototype unit is being
installed on an oil-fired boiler.
The major interest appears to be in use of limestone in the
scrubber. Plans for equipping over 3000 mw with tail end processes were
discussed. The plants range in size from 37-580 mw and scheduled startups
vary from January 1972 to May 1977. It was interesting to note that all
stressed the importance of reliability and many have included scrubber
bypass systems, redundant installed equipment, and modular construction.
Methods of solids disposal varied from use of only settling ponds to barge
transport of slurry to a remote location followed by mechanical dewatering
and return of the clarified liquor. The concensus was that waste disposal
is likely to be a major cost item. Stabilization of sludge by the addi-
tion of fly ash was reported. Materials of construction varied but in all
cases the scrubbers are lined with polyester glass or rubber.
One presentation dealt mainly with a fairly detailed cost
analysis for a tail end system. The capital requirement for a l80-mw
retrofit unit amounted to $^9/kw and the operating cost was estimated at
about $^/ton of coal. Others reported estimated capital costs ranging
from about $20-$6o/kw.
In spite of the increasing commitment to full-scale application
of the lime/limestone scrubbing processes many questions remain unanswered.
The EPA-funded, prototype-scale demonstration of the process at TVA's
Shawnee Steam Plant is designed to fully characterize the process with
limestone added in the scrubber circuit, calcined lime added in the scrub-
ber, and limestone injected into the boiler followed by wet scrubbing.
655
-------
The system was designed by Bechtel Corporation. Three parallel scrubbers
each designed to handle 30,000 acfm of gas will be operated simultaneously
with closed-loop operation. Each train is equipped with a thickener and
the underflow from a single unit can be further dewatered in a centrifuge
or filter. A separate settling pond is provided for the test facility.
The test program will begin with a break-in phase to establish
system integrity followed by statistically designed screening experiments
to identify the effects of variables and will be concluded by long-term
operation at optimum conditions to verify reliability. The system is
extremely flexible and well instrumented to obtain maximum information.
It was reported that limestone is available in most parts of
the country in quantities that are sufficient for the installed generating
capacity. Moreover, because of the abundant reserves, the price is likely
to be relatively stable for the foreseeable future.
656
-------
APPLICATIONS OF THE C-E
AIR POLLUTION CONTROL SYSTEMS
JAMES JONAKIN
JAMES MARTIN
C-E Combustion Division
Windsor, Connecticut
Presented at
SECOND INTERNATIONAL LIME/LIMESTONE WET SCRUBBING SYMPOSIUM
November 8-12, 1971
Sheraton-Charles Hotel, New Orleans, Louisiana
657
-------
APPLICATIONS OF THE C-E
AIR POLLUTION CONTROL SYSTEM
James Jonakin
James Martin
Much has been said in the technical press on the subject of air
pollution from power plant boilers and the C-E systems for meeting
this problem. A brief review will serve to introduce the full-
scale commercial activities C-E has been engaged in.
Requirements for electric power in the U. S. will increase from
1-5 trillion kwh in 1970 to over 3 trillion kwh by 1980. To meet
this demand for power, generation capacity must increase from the
present 3^0 million kilowatts to about 710 million kilowatts and
all available fuel sources - nuclear, coal, oil, and others -
must be used. By 1980, nuclear energy is expected to supply about
21$ of the total installed capacity while a substantial portion of
the remaining capacity will be supplied by coal. In the face of
this continuing demand for electric energy from fossil fuels, utili-
ties and equipment manufacturers must work toward decreasing pollution
while at the same time keeping the cost of the end product, electricity,
at a reasonable level.
POLLUTION EMISSIONS
As coal is burned in the furnace of a steam generating unit, a number
of combustion products are formed, of which 98.7$ are non-pollutants.
The primary concern of the power industry is particulate matter, sulfur
oxides, and nitrogen oxides. Particulate matter accounts for 80$ of
the polluting effluent, while sulfur oxides and nitrogen oxides make
up the remaining 20$. These three pollutants represent only 1.3$ of
the total stack effluent, a rather small portion but a very signifi-
cant quantity.
When all combustion processes in the U. S. are taken into account
and the figures are "based on yearly emissions of pollutants into
the atmosphere, then the total is of concern, as shown in table 1.
These quantities will increase in proportion to the amounts of
fossil fuel consumed unless additional control methods are developed
and applied. Pollution from carbon monoxide and hydrocarbons are
almost two-thirds the total; however, their source is primarily from
automobiles, trucks, airplanes, etc. and will not be discussed in
this presentation. Note that the tons of particulate matter emitted
is the smallest of the totals listed. Although a "heavy" potential
polluter accounting for 80$ of the polluting effluent from a steam
generator, present technology is generally adequate to reduce
emission of particulate matter to the level shown. Electrostatic
precipitators are available with a removal efficiency of better than
95$-
658
-------
CONTROL OF SULFUR OXIDES
The control of sulfur oxides emissions, however, presents some
difficult problems. Trace quantities of sulfur oxides have to
be removed from millions of cubic feet of gas on a continuing
basis. As an example, over 2 million ft of gas is discharged
from the stack of a 700,000 kw steam generating unit each minute.
The gas resulting from the combustion of a coal containing 3%
sulfur contains only 0.2/J sulfur oxide.
Sulfur oxides can be controlled by two general methods:
1. by burning low sulfur fuels which occur naturally, or
are produced by removing the sulfur from the fuels
before combustion, or
2. by removing sulfur oxides from flue gases after com-
bustion.
Natural gas contains no sulfur, but the supply of this fuel in the
U. S. is limited and the proven resources are shrinking rapidly.
The use of this premium fuel by utilities is decreasing and cannot
be considered a general solution to the pollution problem.
Fuel oil produced from crude oils originating in North Africa and
the Far East contains less than 1% sulfur and many utilities on our
east coast are now burning this fuel. However, drilling and tanker
capacities are somewhat limited and its cost has increased as the
demand has increased. In addition, the rapidly changing political
situation in these countries jeopardizes this source of oil. The
high bid fur Central America and Middle East oils are more economical
and the supply is more stable.
The coal used by electric utilities in the USA contains an average
of 2.2% sulfur but ranges from <~L% to >5$- In the eastern and mid-
western states where about 80% of the total power is produced, most
coals contain >3% S- There are large supplies of low sulfur coal
in the Western U. S., but for the most part, it ic located far
from major population centers or market centers, therefore, it has
not been mined commercially on a very large scale.
STACK GAS CLEANUP
Of the two methods of controlling SOx emissions, the present emphasis
of industry is on stack gas cleanup rather than removing sulfur from
the fuel before conbustion. There are numerous schemes for removal
of SO and particulate matter from stack gases. They have been des-
cribed in some detail in various publications.
Wet scrubbing has received the most attention because of its low
cost and simplicity. Combustion Engineering is offering two wet
bcrubbing systems: "furnace injection" and "tail-end". These
systems can be termed "throw away" systems since the process is not
dependent on the recovery of a salable chemical by-product.
659
-------
SULFUR RECOVERY AND WASTE DISPOSAL
The descriptive term "throw away" implies indiscriminate rejection
of by-products, and carries the stigma of conversion of one form
of pollution to another. C-E is not blind to this situation and
continues to investigate all aspects of the problem.
A cost comparison of the C-E wet scrubbing process vs sulfur
recovery processes is shown in table 2, In the U. S., there-
fore, the present systems for stack gas cleanup must be used at
least as an interim measure. At the same time C-E has established
research programs for the investigation of by-product utilization
and waste disposal. This subject is covered in another C-E paper
at the conference, entitled, "Research and Development in Wet
Scrubber Systems".
FURNACE INJECTION SYSTEM
The furnace injection process involves injecting an additive which
contains a large percentage of calcium or magnesium, such as lime-
stone or dolomite, into the furnace of a steam generating unit.
The schematic arrangement has been shown in many papers. Here,
the additive is calcined, producing a more reactive compound
which then chemically reacts in the dry state with some of the
SO and SO in the combustion gases to form compounds of calcium
ana magnesium. About 20 to 30 percent of the sulfur oxides,
including all of the SO , is removed in the dry state.
Next, the flue gas containing unreacted SO and calcined additive
flows to the wet scrubber, a large tank-like structure containing
water sprays and a bed containing glass spheres. In the scrubber,
the remaining calcined additive chemically reacts with the water
and the remaining SO . At the same time, the fly ash is scrubbed
from the gas. The solution containing the coirroounds formed by
these reactions and fly ash drains out the bottom of the scrubber
to a clarifier or pond where the solids settle. Clarified water
is then available for recirculation. The cleansed flue gas
passes through a mist eliminator for removal of moisture and
then is reheated for fan protection and reduction of stack plume.
The development of C-E's air pollution control system started in
196^ with the construction of a small pilot facility in our labora-
tories. This was followed by a second pilot application on a unit
at the Detroit Edison Company in 1966 and 196? •
APCS CONTRACTS - FURNACE INJECTION SYSTEM
1. Union Electric - Meramec No. 2 - lUO MW
2. Kansas Power and Light - Lawrence No. h - 125 MW
660
-------
Installation of the APCS systems for these tvo existing
units was completed in late 1968. Operation of these
demonstration units has revealed problems , and modifica-
tions have been made to the scrubbers, water control
piping, and the additive injection procedure. Our
Research and Product Development Department has been
actively involved in the solutions to these problems,
which are described in detail in their paper on wet
scrubber systems. Data obtained to date show that
SO emission when burning coal with 3-5% sulfur has
been reduced to the equivalent of burning a 1% sulfur
coal, and that at least 99% of the particulate matter
is removed.
These units are pictured in Figures 1 and 2 res-
pectively.
3. Kansas Power and Light - -Lawrence No. 5 - ^20 I'M
The contract for this unit, obtained in 1968, is for a
new C-E boiler. This unit is just starting up and has
fired natural gas during startup. Coal will be fired
beginning about mid-November 1971 and the APCS will be
tested.
This unit is pictured in Figure 3.
k. Kansas City Power and Light - Hawthorne No. 3 - 100 MW
5. Kansas City Power and Light - Hawthorne No. h - 100 MW
The contracts for these two units were obtained in 1970.
They are presently being installed and are scheduled for
operation next year.
TAIL-END SYSTEM
The other C-E air pollution control system, designated "tail-end", is
similar to the furnace injection method except a slurry of pulverized
limestone or slaked lime is used to scrub the flue gases. No furnace
injection of additive is required. Results obtained in prototype
units show that oO% or more of the SO can be removed with a particulate
matter removal efficiency of 99% or better.
The schematic arrangement for the contracts with this system is
approximately similar to that for the KDL prototype unit, shown
in the paper on wet scrubber systems. The system is designed for
a variety of situations in which furnace injection can be eliminated.
Some of these situations are:
a. Providing flexibility in the selection of the most
suitable additives.
661
-------
b. Supplementing existing electrostatic precipitators on
older units.
c. Replacing precipitators for the removal of particulates
as well as SO , as in the case of one contract described
below.
APCS CONTRACTS - TAIL END SYSTEM
6. Louisville Gas and Electric - Paddy's Run No. 6 - 70 MW
The contract for this unit, an existing boiler built by a
competitor, was obtained in 1970. Of interest is the fact
that this contract will use carbide sludge, a waste material,
as the additive for reducing SO emissions. The utility
will obtain the carbide sludge from a local industrial firm
which has large quantities of this waste product produced
in the manufacture of acetylene. In other words, a "throw
away" product from another industry is recycled to improve
the environment.
To ensure that the carbide sludge would be an adequate
additive, tests were conducted in our new laboratory
prototype unit. This 12,500 cfm unit, the largest
laboratory test unit of its kind, has been running
almost constantly since its startup in early 1970. A
tremendous amount of research data has been collected
which has helped us in adapting systems to the specific
requirements of customers. Most important, however, it
has helped us to accelerate progress in the development
of a commercially operable APCS.
7- Union Electric - Meramec No. 1 - 125 MW
This contract was obtained in 1971- The boiler is a
duplicate of Meramec No. 2, for which a furnace injec-
tion system was supplied, in item 1 above.
8. Northern States Power - Sherburne No. 1 - 700 MW
This contract, also obtained in 1971> is significant
for two reasons:
a. It is the largest contract so far, for either
of the two C-E systems. The boiler is a 1970
contract, a C-E controlled circulation unit
burning a sub-bituminous coal.
b. The coal has a sulfur content of less than 1%
and experience has shown that this fact is detri-
mental to the efficiency of electrostatic preci-
pitators . It is for this reason that the wet
scrubber system was selected and the primary
object is particulate removal, although about
half the sulfur in the fuel will be removed by
the scrubbers. This system and the reason for
its selection will be described in a paper to be
given by Mr. Noer.
662
-------
APCS CONTRACTS - FIELD EXPERIENCE
The initial operation of the air pollution control system has
previously "been discussed. Subsequent operation has revealed
additional problems and also has allowed C-E to further refine
the modifications made to the original APCS design. The fol-
lowing is a summary of the modifications made to system com-
ponents .
Sootblowers
The installation of half-track sootblowers on the scrubber inlet in
1969 has eliminated the problem of massive inlet deposits. These
sootblowers have been in service at Union Electric Meramec #2 and
Kansas Power and Light Company Lawrence ffh for over 7000 hours of
coal operation. Only one serious buildup of deposits has occurred
during that time. The oscillating motor on one of the inlet blowers
failed and was not discovered immediately. In approximately 2U
hours a large deposit developed, which had to be removed manually.
The sootblowers have been operating at normal blowing pressures and
have not required any abnormal maintenance. All future C-E - APCS
are being supplied with inlet sootblowers similar to those on the
existing unity,
Ladder Vanes
The installation of ladder vanes on the units at KP&L Co. , Lawrence #U
and U.E. Co., Meramec #2 helped solve the problem of gas distribution
in the scrubber but in the final analysis could not successfully be
kept clean. The ladder vanes were removed in 1970.
Further work has been done at our Kreisinger Development Laboratory
to determine if a less complicated system of ladder vanes can be
used to improve the scrubber's gas distribution but not provide such
a large surface for deposition. In addition, more sophisticated
wash systems are being designed.
Spray Piping
Deposition on the underbed spray piping has been reduced dramatically
by the use of strategically placed wash nozzles and the installation
of some quenching nozzles at the inlets of the scrubbers.
The use of some synthetic materials to prevent corrosion has led C-E
to the conclusion that the material best suited to the environment
of the scrubber is fiberglass. Piping, either made from fiberglass
or coated with fiberglass hasbeen installed in the existing field
units. In addition, the use of fiberglass coatings on other scrubber
surfaces has been tried successfully.
663
-------
Nozzles
Underbed spray nozzle pluggage is another problem we have encountered.
Deposit formation in the bed area is directly related to proper
operation of spray nozzles and distribution of spray water. Nozzle
plugging with bits of scale and miscellaneous debris which happen
to enter the spray water system as well as a need for proper mixing
of spray water and recycle slurry water led to the development
of a special nonclogging nozzle. There have been no serious plug-
gage problems in the field units since installation of the nozzle.
Both overbed and underbed recycle of slurries have been accomplished
with these nozzles.
Bed Plugging
Bed plugging of two types has been experienced at both existing
demonstration units. They are defined as follows:
1. Deposits formed by mechanical means. Deposits formed because
of low gas velocities in sections of the scrubber bed fall in
this category. These deposits may be cementitious in nature
because of the mixture of solids found in the scrubber, but
usually can be washed away with a fire hose.
2. Deposits formed as a result of precipitation of a slightly
soluble calcium salt such as calcium sulfate, calcium sulfite, or
calcium carbonate. The scales formed by these deposits are
always crystalline in nature and are usually very difficult
to remove.
The control of these two types of bed pluggage has been accomplished
by the use of higher liquid to gas ratios, to prevent mechanical
deposits, the addition of extra make-up water to the system to dilute
the concentrations of the slightly soluble calcium sulfate, and the
implementation of closer controls on the system's chemistry to prevent
calcium sulfite and carbonate scale.
Work in our research laboratory is just being completed which is
expected to eliminate the use of the extra make-up water for cal-
cium sulfate scale control.
Mist Eliminators
The mist eliminators above the marble bed became coated with deposits
during early runs. The throughput gas velocity was raised, and
consequently, the dewatering capability was greatly reduced. The
demisters were cleaned, and a wash system was installed above the
demister section to be operated off-line, as required. Deposition
on the demister vanes has been reduced during operation under steady
conditions at moderate loads.
Laboratory studies were undertaken in 1971 to develop an improved mist
eliminator which could sustain higher gas velocities than present
designs and also lend itself to washing better than the existing demister.
664
-------
Water table studies suggested that a two stage L-shaped demister
would "be more satisfactory than the present Z-shaped demister.
Air-water models were constructed in our laboratory at Windsor,
Connecticut. The evaluation showed that the double L-shaped
demister would have considerably more throughput capacity than
the existing design and that it could be washed much more easily.
This two stage L-shaped demister is presently being installed at
Kansas Power and Light Company on both units and has been incor-
porated on all future APCS.
Reheaters
The reheaters have been "plastered" with deposits a number of times
during previous operation. The improvement in scrubber bed gas dis-
tribution and the installation of half-track sootblowers have made
it possible to maintain the gas side pressure drop across the reheater.
Operating experience has'led to the conclusion that, in the damp atmo-
sphere above the demisters, a steel-on-steel construction was necessary
to provide fin strength to withstand blowing pressures required for
Clean operation. The initial reheaters were replaced with an arrange-
ment of the heavy duty finned tubes to allow for efficient cleaning by
scheduled sootblower operation.
Our experience during the last twelve months since the redesigned
reheaters were installed has revealed that the present reheater
design coupled with half-track sootblowers are adequate if the
demister is performing as designed.
On two or three occasions during the last year, improper demister
operation has caused excessive reheater deposition at both Kansas
Power and Light Company, Lawrence #U and Union Electric Company,
Meramec #2. The new demister design previously described should
eliminate this problem.
Drain Line Scale
During June, 1969, operation of the APCS at Union Electric, a calcium
sulfite scale was found in scrubber drain line to the clarifier. It
was theorized that some of the calcium oxide entering the scrubber was
falling out in the scrubber bottom. This CaO would then begin to hydrate
as it was conveyed down the scrubber drain line to the clarifier. This
hydration would cause a rise in the pH of the scrubber effluent. Since
the scrubber effluent contains a high concentration of calcium bisulfite,
an upward shift in pH would cause a conversion of calcium bisulfite to
sulfite and a shift in solubility and probably produce the observed cal-
cium sulfite scale.
To eliminate this problem, the bed overflow pots were tied off and
piped to the clarifier separately thereby insuring no shift in solu-
bility. The calcium oxide that fell out in the scrubber bottom was
conveyed to a well-mixed recycle tank and then pumped up above the
scrubber bed in order to utilize this additive.
This system has undergone extensive testing at both Union Electric
and Kansas Power and Light Company. It has been determined that
665
-------
the calcium oxide cannot be injected in the "bed without calcium
sulfite scale occurring. A modification to the recycle system vas
made at KP&L Company in the spring of 1971. A portion of the pot
effluent slurry was pumped to the recycle tank and thoroughly mixed
with the calcium hydroxide slurry. It was determined that if the
proper amount of pot effluent slurry was used (i.e., utilized to control
the recycle tank pH essentially) no scale was produced when the
material was recycled and yet significant improvement in sulfur
dioxide absorption could "be achieved. This improvement ranged
from 200-^00 ppm additional sulfur dioxide removal depending on the
stoichiometric amount of limestone being fed.
Furnace Operation
Kansas Power and. Light Company has been able to operate the APCS
for periods of four weeks or longer at a time without serious
plugging developing in the back-pass of the boiler. Additional
sootblowers have been ordered but have not been installed to
date.
Union Electric Company Meramec #2 has experienced back-pass plugging
which has required high pressure cleaning. The economizer of this
unit is extremely close spaced and is normally difficult to clean.
In addition, the tubular air heater does not lend itself to the
furnace injection APCS as well as the regenerative air heater. The
application of the furnace injection system to existing units requires
a thorough evaluation of the back-pass of the boiler to prevent prob-
lems similar to those experienced at Meramec #2.
Other Field Work
During 1970 the application of wet-scrubbing for particulate removal
from boilers fired with low sulfur and low ash coal was evaluated at
the Naughton Station of Utah Power and Light. A scrubber having one
marble bed and two marble beds in series, and a rod type scrubber
were tested. Results indicate a particulate reduction to 0.02-0.035
grains per Standard Cubic Foot (SCF) of dry gas for the marble bed and
rod type scrubbers operated at a 6-10 inch pressure drop. Increasing
the pressure drop across the rod type scrubber resulted in a reduction
of the particulate matter below 0.01 grains per SCF of dry gas.
Field testing was continued at Meramec Station #2 of Union Electric
Company and Lawrence Station ffh of Kansas Power and Light Company
to evaluate a new nozzle design, variations in recycle methods and
limestone addition on the removal of SO and deposit formation on
the scrubber components. S0? removal up to 70 percent was achieved.
Interrupted operation due to change to gas firing to meet load demands
did not permit completion of scheduled testing program at Lawrence.
(This unit is capable of operation at higher loads on gas than coal.)
Following equipment revisions to incorporate new knowledge and experience
test programs were undertaken at both field units in 1971 in the follow-
ing general categories:
1. Operation without recycle
666
-------
2. Operation with overbed recycle
3. Operation with underbed recycle
k. Simultaneous operation of both underbed and overbed recycle
Results were:
Kansas Power and Light Lawrence #U
The test program on recycle optimization was completed. Tests were run
with above and below bed recycle, with and without dilution (bed over-
flow water directed to the recycle tank) at QO MW.
In general,results show:
1. The APCS can be run with above bed diluted recycle without scaling
at or below stoichiometric additive feed rates.
2. Modifications are required (additional punrp capacity) to fully test
underbed recycle.
3. Dust efficiency is between 98.5 and 99-2 percent.
k. Overall SO removal efficiency ranged from 50 to 75 percent depending
on mode of operation.
Approximately 1500 hours of operating time was accumulated in 1971 up
to mid-June 1971 despite a three week annual outage.
The Environmental Protection Agency ran a weeklong series of tests at
our installation in Kansas in order to set national standards in terms
of particulate and gaseous pollutant emissions.
Kansas Power and Light Lawrence #5
The unit has been in service since March, 1971, on gas operation.
The APCS under went preliminary check-out in September, 1971- The
unit was operated on a partial coal firing basis for five days prior
to a scheduled one-month unit outage. It is presently planned to
return the unit to service the week of November 15, 1971, at which
time full coal operation will commence.
During the preliminary operation of the APCS at the end of September,
two problems were observed.
Initial operation revealed poor gas distribution in the scrubbers.
High velocities were found in the rear of the beds and conversely
the front of the beds (side above the inlet) were not active. Tests
on gas flow models at our research lab indicate that the installation
of three ladder vanes will give the necessary improvement in gas dis-
tribution. These systems are currently being installed with a wash
system to minimize deposition. This ladder vane assembly is a sim-
plified version of those previously tried at Union Electric and KP&L
Lawrence #k.
667
-------
A second problem was encountered during this brief period of operation
in September. The distribution of furnace injected limestone to the
six scrubbers was found to be very poor. A modification to the lime-
stone injection system is currently being made to improve the distri-
bution of the additive in the furnace.
The correction of the imbalance in limestone addition at the point of
intoduction should allow us to cope with small variations, the scrubbers
by employing the recycle concept previously mentioned.
CONCLUSIONS
In summary, C-E is one of the pioneers in the development of methods
to minimize emissions from stacks of power plants. For example, initial
work on the C-E pollution control system for controlling SO and dust
emissions began some seven years ago at our Kreisinger Development
Laboratory to meet the needs of our electric utility customers. The
project was undertaken without government funding. Today, after the
sale of a number of systems, our program is still going strong. Our
efforts have been expanded and new areas are now being explored.
668
-------
TABLE 1
AIR POLLUTION MISSIONS
IN THE U.S. FOR 1969
MILLIONS
POLLUTANT TON/YEAR %_
Sulfur Oxides 26 18.3
Nitrogen Oxides 13 9-2
Particulate Matter 12 8.U
Carbon Monoxide 72 50.7
Hydrocarbons 19 13.k
TABLE 2
COST COMPARISON OF CE WET SCRUBBING
PROCESS VS SULFUR RECOVERY PROCESSES
CAPITAL COST OPERATING COST
PROCESS $/KW MILLS/KWH
CE 13 to 25 0.2
Sulfur Recovery 30 to 60 0.7 to 1.5
669
-------
Fig.
-1
670
CO
o
o
O
o
-------
^!
CE-APCS
KANSAS POWER & LIGHT CO.
LAWRENCE, KANSAS
LAWRENCE it
Fig. -2
671
-------
CE-APCS
KANSAS POWER & LIGHT CO. - LAWRENCE 5
LAWRENCE, KANSAS
Fig. -3
672
-------
REMOVAL OF SULFUR DIOXIDE FROM STACK GASES
BY SCRUBBING WITH LIMESTONE SLURRY;
DESIGN CONSIDERATIONS FOR DEMONSTRATION FULL-SCALE SYSTEM AT TVA
By
B. G. McKinney
Division of Power Resource Planning
Tennessee Valley Authority
Chattanooga, Tennessee
A. F. Little
Division of Chemical Development
Tennessee Valley Authority
Muscle Shoals, Alabama
Prepared for Presentation at
Second International Lime/Limestone Wet Scrubbing Symposium
Sponsored by the Environmental Protection Agency
New Orleans, Louisiana
November 8-12, 1971
673
-------
REMOVAL OF SULFUR DIOXIDE FROM STACK GASES
BY SCRUBBING WITH LIMESTONE SLURRY;
DESIGN CONSIDERATIONS FOR DEMONSTRATION FULL-SCALE SYSTEM AT TVA
By
B. G. McKinney
Division of Power Resource Planning
Tennessee Valley Authority
Chattanooga, Tennessee
A. F. Little
Division of Chemical Development
Tennessee Valley Authority
Muscle Shoals, Alabama
ABSTRACT
Design considerations are presented for the limestone slurry
scrubbing system planned for TVA1s Widows Creek unit 8 (550-mw; plant
located in Northeast Alabama). The scrubber type will be selected on
the basis of results from pilot plant tests. Basic design premises for
the overall installation are given and the following major design factors
are discussed.
'Limestone handling and grinding.
'Scrubber design (e.g., gas velocity, entrainment separation,
turndown, liquor flow rate, limestone amount and particle size,
solids content of slurry).
*Particulate removal.
'Equipment arrangement.
*Flue gas handling and conditioning.
^Instrumentation and control.
*Materials of construction (for resistance to both corrosion and
erosion).
*Method of solids disposal.
-------
REMOVAL OF SULFUR DIOXIDE FROM STACK GASES
BY SCRUBBING WITH LIMESTONE SLURRY;
DESIGN CONSIDERATIONS FOR DEMONSTRATION FULL-SCALE SYSTEM AT TVA
By
B. G. McKinney
Division of Power Resource Planning
Tennessee Valley Authority
Chattanooga, Tennessee
A. F. Little
Division of Chemical Development
Tennessee Valley Authority
Muscle Shoals, Alabama
The Tennessee Valley Authority (TVA) has a major research and
development program under way, mainly in cooperation with EPA, on methods
for removing S02 from power plant stack gas. Small-scale and pilot plant
work, plus the EPA-TVA large-scale test program at the TVA Shawnee Steam
Plant, have been described in other papers in this symposium. In addition,
TVA decided in late 1970 to install a full-scale S02 removal system on
generating unit 8 at Widows Creek Steam Plant (in Northeast Alabama, near
Chattanooga, Tennessee). The primary objective is to work out design and
operating problems that affect both S02 removal efficiency and process
reliability, with emphasis on the latter. Hopefully, a removal system can
be designed and demonstrated that will serve as a model for any future
installations required.
Among the reasons for selecting Widows Creek No. 8 as the demon-
stration unit was the fact that the electrostatic precipitators there are
quite inefficient. Since additional dust removal capacity was needed anyway,
it was decided to install a wet scrubbing process and thus remove both the
residual dust and the S02. Thus the dust problem was a major factor in the
decision to remove S02, which otherwise might have been delayed until design
data were available from the EPA-TVA Shawnee project.
In selecting a process, the first choice to be made was between
recovery and throwaway operation. Since no proven recovery methods were
available, and marketing problems were considered quite difficult, throwaway
was selected even though it involves a solid waste disposal problem.
The next choice was between lime and limestone as the absorbent.
Lime is more active but can be produced economically only by injection of
limestone into the boiler and catching the resulting lime in the scrubbers.
Several major problems made this course unattractive:
675
-------
1. Experience by TVA and others has shown that injection of
limestone into the boiler can cause fouling and plugging
of boiler surfaces and tube assemblies.
2. The use of lime, particularly when introduced with the gas,
aggravates the scaling problem. Scaling can be reduced or
eliminated by operating the scrubber at low pH or by high
blowdown and replacement with fresh water. Both are unde-
sirable, the latter particularly because it entails outflow
of scrubber liquor to watercourses. It was decided that the
TVA system must operate as a "closed loop," that is, with
return of all liquor to the scrubber.
J. Since it was desired to retain the precipitators, the presence
of lime in the gas from the boiler would have complicated
operation.
4. Dry grinding of limestone creates a dust problem.
5. If lime is generated in the boiler, there is the problem of
getting the same CaO:S02 ratio to all of the scrubbers serving
the boiler.
It was decided, therefore, to introduce the limestone, after wet
grinding, directly into the scrubber circuit. Experience by others has
indicated that this reduces the scaling problem by a major degree. It also,
of course, avoids the other problems listed for boiler injection.
An attempt is made in this paper to define and evaluate the design
considerations associated with limestone scrubbing. Design areas considered
are limestone handling and grinding, scrubber design, particulate removal,
equipment arrangement, stack gas handling and conditioning, instrumentation
and control, materials of construction, and method of solid disposal.
Figure 1 shows the general area at the Widows Creek Steam Plant,
with the expected location of equipment. The new waste solids pond is also
shown.
Basic Design Premises
In preliminary planning for the scrubbing facility, it was first
necessary to establish major design premises. The more important of these
are as follows:
676
-------
\
w
677
-------
1. Coal analysis (as fired basis)
a. Ash content, 25$
b. Sulfur content, 4.3$
c. Moisture, 5-0$
d. Heating value, 10,000 Btu/lb
2. Capacity
a. Maximum power generation rate for unit No. 8, 550 mw
b. Stack gas rate at capacity,1 1,600,000 acfm at 280°F
(5,325,000 Ib/hr)
3. Sulfur dioxide removal
a. Percent removal, 80
b. Inlet concentration, 3^-0 PPm (wet basis); 37^-0 PPm (dry basis)
c. Outlet concentration, 650 ppra (wet basis); 750 ppm (dry basis)
k. Particulate removal
a. Inlet particulate loading,2 5.6 gr/scf (dry); 5.! gr/scf (wet);
3.6 gr/acf (280°F)
b. Particulate level at scrubber exit,2 0.020 gr/acf (l25°F saturated)
0.022 gr/scf (wet); 0.026 gr/scf (dry)
5. Stack gas reheat temperature, 175°F (50°F rise)
These premises have been used for preliminary engineering purposes but will
be reviewed and updated as required prior to detailed engineering.
1 Based on a total of 33$ excess air including air heater leakage.
2 Based on the conventional ASME sampling technique.
678
-------
Limestone Handling and Grinding
A wet grinding system for limestone was chosen over dry grinding
because it is less expensive and does not produce a dust problem. Even
if dry grinding were used, the ground limestone probably would be slurried
before feeding to the scrubber.
A schematic drawing of the limestone handling and grinding system
is shown in Figure 2. The facility is designed for receiving limestone by
both rail and truck from the quarry. The limestone is conveyed from an
unloading hopper to either the live storage silo or the dead storage area.
Material will be reclaimed from the dead storage area as required to maintain
an adequate level in the live storage silo. Capacity of the latter is
sufficient for 3 days' operation.
Limestone is conveyed from the live storage silo to a wet ball
mill where it is ground from the purchased size (probably 1-1/2 by 0 in.)
to the desired size. The resulting slurry is pumped from the ball mill
through a classifier where the oversized particles are separated and re-
cycled to the ball mill. The product slurry (50-60$ solids) from the
classifier goes to a hold tank from which it is pumped to the scrubbing
system.
Final selection of the particle size has not been made, but it
will not likely be coarser than 70$ minus 200 mesh or finer than 90-95$
minus 325 mesh. Selection of the size will be made after completion of
additional pilot plant tests. Hopefully, the additional tests will produce
sufficient data to permit determining whether the advantages of a finer
grind justify the increased grinding cost. The potential advantages are
primarily (l) less erosion of piping and equipment and (2) less limestone
requirement for a given S02 removal rate, other conditions being equal.
Final selection of limestone feed stoichiometry has also not yet
been made but probably will be between 1.2 and 1.5. The final selection
will depend upon results of additional pilot testing and selection of a
particle size.
The limestone source probably will be deposits in the vicinity of
the Widows Creek plant. Estimates have indicated that the cost of shipping
limestone from greater distances would not be economical unless some major
advantage could be demonstrated, which has not yet been done. The local
deposits are fortunately quite low in magnesium, which could cause a ground
water pollution problem if present in large amounts. The composition of the
limestone is given in another paper in this symposium ("Removal of Sulfur
Dioxide from Stack Gases by Scrubbing with Limestone Slurry: TVA Pilot Plant
Tests—Part I, Scrubber-Type Comparison" by T. M. Kelso, P. C. Williamson,
and J. J. Schultz).
679
-------
o
CVJ
3
en
ff
•r-l
T3
C
•rl
M
O
a)
t>0
•1-1
680
-------
The flowsheet shows recycled pond water as the liquid portion
of the mill feed. Use of recycled water at this point reduces the amount
available for washing the mist eliminators in the scrubber, for which a
substantial flow of clear liquid is needed. Further attention will be
given to the problem of portioning out the available clear liquor for the
various needs.
Scrubber Design
Several scrubber designs and configurations for limestone
scrubbing have been considered in studies conducted at TVA. Three types
remain under consideration for the Widows Creek facility: (l) a venturi -
spray tower system, (2) a venturi - mobile-bed combination, and (3) the
mobile-bed type alone. Flow diagrams for the three are given in Figures 3>
ij-,- and 5- In each of the three schemes, four scrubber trains will be used--
one for each of the four stack gas ducts from the boiler.
In the venturi - spray tower system (Fig 3)» tne Sas exiting
the electrostatic precipitator is cooled and presaturated prior to entering
the absorption tower. The latter is an open spray tower equipped with a
venturi in the bottom of the tower. The venturi, through which a portion
of the scrubber circulating slurry is injected, serves to remove fly ash
and to atomize the scrubbing liquor for S02 absorption. The remaining
circulating slurry is injected into the top portion of the spray tower
through spray headers equipped with nozzles.
Under the current plan, limestone slurry from the mill, recycled
pond water, and makeup river water are introduced into the absorber circu-
lation tank (however, as mentioned earlier, it may be necessary to use all
available water and clear liquor to wash the mist eliminators). The resulting
slurry is circulated from the tank through the presaturator, venturi, and the
two spray tower headers, with liquid to gas ratios (L/G) of about J, 20, 20,
and 20 gal/Mcf, respectively, to the four points.
The pressure drop across the venturi is 10-15 in. H20. Recent
pilot tests indicate that a lower pressure drop than this is adequate for
dust removal. However, the tests also indicate that the higher pressure
drop is necessary for adequate S02 removal in the venturi, presumably because
the pressure drop gives a more effective spray pattern in the lower part of
the scrubber.
The superficial design velocity for the spray tower is 6-8 ft/sec,
and the solids content of the circulating slurry is 10-15$ by weight.
681
-------
QQQ
oo oo
OO
ru
OOO OOO
ru
Dri
s
•H
O
(1)
4-1
CO
I
to
tJ
a,
4-1
c
O O
O O
682
-------
OO 00
OO
OOO OOO
g
M
toO
W
•H
Q
s
o
CO
>>
CO
T)
(1)
A
5
c
o o
683
-------
oo oo
oo
5
ooo ooo
«J
•i-l
P
§
•d
-------
In the venturi - mobile-bed system (Fig k), stack gas from the
electrostatic precipitators enters the venturi where it is cooled,
saturated, and the fly ash removed. The cooled gas then flows to the
mobile-bed scrubber where the S02 is absorbed. Limestone slurry and
recycled pond water (see earlier discussion) are added to the absorber
circulation tank. Slurry is circulated from this tank to the upper of
the two beds through a spray nozzle header (low pressure) and flows down
through the scrubber. Slurry overflows from the absorber circulation
tank to the venturi circulation tank from which it is circulated through
the venturi.
The absorber is operated with a liquid:gas ratio of about
kO gal/Mcf and a pressure drop of about 6-8 in. H20. Corresponding values
for the venturi are 20 L/G and if.-5 in. H20. The superficial design
velocity in the mobile-bed scrubber is 12-13 ft/sec. The solids content
in the circulating slurries has been set tentatively at 5 and 8$, re-
spectively, for the mobile bed and the venturi. The 5$ was selected because
general practice is to avoid higher contents in order to limit wear of the
plastic balls. However, the pilot plant has been operated at 10-15$ solids
to reduce scaling, improve S02 removal, and reduce pumping load for trans-
porting the waste solids to the pond and the separated liquor back again.
Finer limestone particle size may allow use of the higher solids content
without excessive ball wear; this is being tested in the pilot plant.
In the scheme involving use of mobile beds alone (Fig 5); tne
flue gas is cooled and saturated in a spray chamber presaturator before
entering the mobile-bed scrubber, which removes both fly ash and S02.
Limestone slurry and recycle pond water are added to the absorber circu-
lation tank from which the slurry is circulated to the upper of the three
beds through a spray header and flows down through the other beds.
The absorber is operated at an L/G of if-O-^O gal/Mcf and 9-12 in.
H20 pressure drop. An L/G of about 5 is used for the presaturator. The
absorber superficial gas design velocity is 12-lJ ft/sec.
The three beds are specified because this was the arrangement
tested in the pilot plant. There is some evidence from the pilot plant work,
however, that the upper bed contributes very little to S02 removal. Further
tests will be made in an effort to find the best combination.
In each of the three scrubber arrangements, the absorber is equipped
with an entrainment separator. Provisions will be made for continuous or
intermittent flushing with recycled pond water. The selection of the entrain-
ment separator is further discussed later under the problem of particulate
removal.
685
-------
Some of the advantages and disadvantages of each of the three
schemes are as follows:
1. The venturi - spray tower has a lower superficial gas
velocity, which results in a larger scrubber cross-
sectional area than the other two schemes.
2. Erosion of the plastic spheres in the two schemes using
mobile beds is a major potential problem.
3. The turndown capacity of the mobile-bed scrubber is limited,
which may lead to pluggage and other problems at reduced
loads. Also, fly ash removal may be a problem at reduced
load in use of the mobile-bed scheme in which the scrubber
must remove both fly ash and S02.
4. For the large cross-sectional areas involved, there may be
the problem of poor sphere distribution in use of the mobile
beds. It may be possible to keep the spheres in place by use
of vertical partition screens but this has not been proven
in practice.
5. The higher gas velocity in the mobile bed results in a higher
loading on the entrainment separator.
6. The mobile-bed scrubber is a better contacting device than
the spray tower, which gives better S02 removal efficiency
or possibly a lower limestone consumption rate.
7- Spray nozzle erosion is more severe in the venturi- spray
tower combination because spray nozzles with higher pressure
drop are required.
8. A higher pressure head on the pumps is required for the venturi -
spray tower system.
9. The venturi - mobile-bed system is the most flexible of the
three because dust and S02 removal units are separated.
Alterations to the configurations of the above schemes are under
consideration that may have potential advantages.
1. Packed section in spray scrubber. A short, open packed section
in the upper part of the spray tower should increase S02 re-
moval and perhaps be free of scaling if open enough.
686
-------
2. The three-bed TCA scrubber performed fairly well in the
pilot plant with the balls removed. Thus a scrubber with
several horizontal, open, thin grids (similar to the re-
taining grids in the TCA scrubber) spaced from the scrubber
top to bottom, might give adequate S02 removal without the
problems associated either with the usual type of fixed
packing or with the mobile-bed type of scrubber.
The choice between the various scrubber schemes will be made
after completion of additional pilot plant tests. The investment and
operating costs of the three will be compared before a decision is made;
preliminary estimates, however, indicate little difference among them.
It is likely that the primary factor influencing the final selection will
be the experience gained regarding operating reliability.
Particulate Removal
Pilot tests conducted at Widows Creek on unit 8 stack gas indi-
cate that the fly ash can be removed satisfactorily with a pressure drop
of ij--5 in. H20 and relatively low liquid:gas ratios, about 10-15 gal/Mcf.
The main problem, as discussed earlier, is entrainment of slurry solids
into the gas stream. Thus development of efficient mist elimination
equipment is essential for the Widows Creek project.
In most of the pilot plant tests to date, a vane-type entrain-
ment separator has been used—selected because of low pressure drop and
large open area which minimizes pluggage and is easy to flush. Plugging
and poor efficiency have been major problems. Further studies will be
conducted to determine the best type of entrainment separator for the
scrubbers at Widows Creek. One course being explored is use of two mist
eliminators in series, the first irrigated with clear liquor to replace
slurry mist with clear liquor mist and the second to remove the clear liquor
mist. Use of the upper TCA bed, which apparently does not contribute much
to S02 removal, as a mist eliminator may also be helpful since the movement
of the balls should prevent plugging. Also, if a fixed-bed section in the
top of the scrubber should prove feasible, mist carryover probably would be
reduced.
Equipment Arrangement
A typical equipment arrangement study drawing is shown in Figure 6.
The general arrangement will be similar for each of the three scrubber
schemes under consideration. Detailed equipment arrangement drawings will
be made after a specific scrubber scheme is chosen.
687
-------
f
utvimoit
SOUTH UtVMION
FIGURE 6
General Arrangement Study
688
-------
The general plot plan shown in Figure 1 shows the relative
location of the three major areas (scrubbing system, limestone handling-
grinding, and solids disposal) with respect to each other and the
unit 8 powerhouse.
Stack Gas Handling and Conditioning
As noted earlier, the gas will be humidified and cooled prior
to entering the S02 absorber. Cooling the gas prior to the absorber is
desirable primarily for two reasons: (l) better S02 absorption and (2)
elimination of temperature problems in materials selection for the absorber.
The gas entering the absorber will be cooled to a temperature of 150°F or
less.
The gas leaving the scrubber will be reheated to desaturate and
to provide buoyancy. The reheater will be designed to provide up to 50°F
temperature rise (125-175°^)- Both direct reheating by burning oil and
indirect heating with low pressure steam are being considered. Primary
consideration is being given to indirect steam but no decision can be made
until the economics are further evaluated.
Soot blowers or liquor sprays will be installed where there is
a likelihood of deposits forming, particularly in the presaturator and
the reheater wet-dry interface areas.
The gas will be moved through the system by induced-draft fans
located downsteam of the reheaters to permit operation of the fans with
a "clean" unsaturated gas. Consideration is being given to designing the
fans with sufficient static pressure to permit converting the boilers from
the present pressure type to balanced draft. A plenum will be provided
connecting the four ducts to the scrubbers to permit operation of the boiler
up to 75$> load with one scrubbing train out of service. Also, bypasses
will be provided around the scrubbers to the stack to prevent an undue
amount of boiler downtime because of scrubber malfunction, particularly
during the initial shakedown operation of the system.
Provision for scrubber turndown is a potential problem. If the
individual scrubbers cannot be turned down without operating difficulty,
then provision must be made to cut one or more scrubbers out of the system
at low boiler load so as to maintain adequate gas velocity through the
active scrubbers. This could require frequent operator attention to cut
scrubbers in and out of the system and would put emphasis on finding
dampers that open and close easily, are resistant to sticking, and close
tightly.
689
-------
Instrumentation and Control
In general, the scrubbing system will be well instrumented to
permit flexibility in control and monitoring of the system since it will
be a demonstration unit. Also, adequate controls are required to mini-
mize operating personnel requirements.
Since there are four similar trains, one of them will have
more instrumentation than the others to permit flexibility in determining
optimum operating conditions. In addition to being useful in optimizing
the Widows Creek facility, this should give data useful in design and
operation of any subsequent installations.
Operating variables to be controlled include limestone feed
rate to the scrubber and solids concentration in the circulating scrubber
slurry. The limestone rate will be regulated to maintain the desired
stoichiometry, either directly or indirectly. The solids content of the
slurry will be maintained at the desired level by regulating the slurry
purge rate to the pond. Additional considerations on instrumentation
and control of the system are as follows:
1. The induced-draft fans will be instrumented and controlled
similarly to induced-draft fans for balanced draft furnaces.
2.
Analyzers will be provided for monitoring the scrubbing
system inlet and exit S02 concentrations.
One of the four trains will be equipped for regulating the
slurry circulation rate to the scrubber. Amperage recorders
will be provided for the scrubber circulation pumps on all
four trains for rough monitoring of liquid circulation rates.
Sampling points, temperature recorders, pressure and differ-
ential pressure recorders, etc., will be provided as deemed
necessary for monitoring and controlling the system.
Materials of Construction
Selection of the most economical materials for adequate resistance
to corrosion-erosion is a major problem. Attempts have been made in the
pilot plant to evaluate various materials of construction including metals,
coatings, and linings. However, it is very difficult, to obtain conclusive
corrosion-erosion data in the pilot plant because of the short running time
of the tests. Evaluation of coatings and linings is particularly difficult.
690
-------
The equipment items most susceptible to erosion-corrosion are
pumps, piping, nozzles, and areas where there is slurry impingement against
surfaces. Corrosion is generally not thought to be very severe in the
system except for its enhancement by erosion. The material that appears
to be most promising for the high erosive areas is carbon steel lined with
a soft elastomer such as neoprene with a 55-60 Shore A hardness. However,
in addition to being expensive, neoprene-lined carbon steel has some distinct
disadvantages; field modifications and repairs are difficult, expensive,
and time consuming. Such linings also generally have a maximum allowable
continuous operating temperature of only about 150°F.
For tanks where erosion is not as severe, carbon steel lined or
coated with a less expensive material than neoprene should be satisfactory.
It may also be feasible to use unlined steel, as was done in the early ICT
work in England--particularly where the pH is above 6.0-
Solids Disposal and Water Recycle
The solids purge stream will be pumped to the pond without any
concentration of the solids. The supernatant clear liquor will be recycled
from the pond to the scrubbing system for reuse. River water makeup will be
used as necessary to maintain the water balance in the system. Water is lost
from the system by evaporation in the scrubber, hydration of reaction products,
pond water evaporation, pond seepage, and entrainment with settled solids as
interstitial water. Water enters the system as rainfall into the pond and as
river water makeup to the scrubber. An overall water balance is given in
another paper in this symposium ("Removal of Sulfur Dioxide from Stack Gases
by Scrubbing with Limestone Slurry: Operational Aspects of the Scaling Problem"
by A. V. Slack and J. D. Hatfield).
Thickeners could be used to concentrate the purge slurry and reduce
the pumping rates to and from the pond. However, the savings in pumping costs
do not justify the additional capital investment required for thickeners.
Also there is some doubt as to how effective thickeners would be in concen-
trating the purge slurry because of the very poor settling characteristics of
the product solids.
One of the undesirable aspects of the solids disposal system is the
possible low bulk density of the settled solids in the pond. Data from the
TVA pilot plant indicate that the settled solids contain about $0-60% by
weight water. The corresponding bulk density is 89-82 lb/ft3 of slurry, which
results in a large pond volume requirement for storing the accumulated solids.
Further work is planned aimed at reducing the settled solids volume.
691
-------
-------
TEST PROGRAM
FOR THE
EPA ALKALI SCRUBBING TEST FACILITY
AT THE
SHAWNEE POWER PLANT
By
M. Epstein
F. Princiotta
R. M. Sherwin
L. Szeibert
I. A. Raben
Bechtel Corporation
Environmental Protection Agency
Bechtel Corporation
Bechtel Corporation
Bechtel Corporation
Presented at the
Second International Lime/Limestone
Wet Scrubbing Symposium
New Orleans, Louisiana
November 8-1 2, 1971
693
-------
INTRODUCTION
In June, 1968, a three phase program was initiated whose aim was the
testing of a large, versatile prototype system to fully characterize
wet limestone scrubbing for removal of sulfur dioxide and particu-
lates from boiler flue gas. The Office of Air Programs (OAP) of the
Environmental Protection Agency (EPA) is sponsoring this program,
with Bechtel Corporation of San Francisco as the major contractor.
Phase I of the test program, which has been completed, consisted of
preliminary engineering, equipment evaluation and site selection.
Phase II, which is close to completion, involves the detailed design
and construction of the facility and the development of the test-plan
and mathematical models for predicting system performance. Phase
III -will comprise, primarily, the testing portion of the program.
The test facility consists of three parallel scrubber systems, each
capable of treating approximately 30, 000 acfm of flue gas, which are
integrated into the flue gas ductwork of an existing coal-fired boiler
at the Tennessee Valley Authority (TVA) Shawnee Power Station,
Paducah, Kentucky. The facility was designed for maximum flexibility
and has a high degree of instrumentation for control and recording of
data over a wide range of operating conditions. Construction is approxi-
mately 80% complete, with Phase III start-up presently scheduled for
March 1972.
694
-------
EPA has utilized the capabilities of several organizations to maxi-
mize the efficacy of the total program. For example, Bechtel
Corporation, as the major contractor, has prepared the detailed
design of the test facility, is developing the test program and mathe-
matical models and will direct the test efforts. TVA is constructing
the facility and will operate the unit during the test program. Other
contributors include, Radian Corporation, who has provided impor-
tant support in the analytical determination areas, and McCrone
Associates, who participated in the development of the particulate
mass loading and size distribution measurement procedures.
695
-------
TEST PROGRAM OBJECTIVES
The overall objective of this program is to evaluate the feasibility and
economics of closed-loop limestone wet-scrubbing processes. The
following are the major goals of the program:
Investigate and solve operating and design problems,
such as scaling, plugging, corrosion and erosion.
Generate test data to characterize scrubber and system
performance as a function of the important process
variables.
Study various solid disposal methods.
Develop mathematical models to allow economic scale-
up of attractive operating configurations to full-size
scrubber facilities. A parallel economic study will be
performed to enable estimation of capital and operating
costs for the scaled-up system designs.
Determine optimum operating conditions for maximum
SO and particulate removal, consistent with operating
cost considerations.
Perform long-term reliability testing.
696
-------
TEST FACILITY
The following discussion provides a brief description of the test
facility. Reference 1 provides a recent, more complete description.
The test facility consists of three parallel scrubber systems, each
•with a different scrubber design and its own slurry handling system.
Scrubbers will be of prototype size, each capable of treating approxi-
mately 30, 000 acfm of flue gas from the TVA Shawnee coal boiler #10.
Therefore, each circuit is handling the equivalent of approximately
10 Mw of power plant generation capacity. The equipment selected
was sized for minimum cost consistent with the ability to extrapolate
results to commercial units. The 30, 000 acfm scrubber train was
judged to meet these requirements.
SCRUBBER SELECTION
The major criterion for scrubber selection was the capability of remov-
ing both sulfur dioxide and particulates with high efficiency (sulfur
dioxide removal greater than 85% and particulate removal greater
than 99%). Other factors considered in the selection of the scrubbers
were:
Ability to handle slurries without plugging or excessive
s caling.
697
-------
• Cost and maintenance requirements.
• Ease of control.
• Pressure drop.
Based on the limited information available in the literature, the follow-
ing scrubbers were selected:
(1) Venturi followed by an after-scrubbing absorption section
(spray or pall-ring packed-bed).
(2) Turbulent contact absorber (TCA).
(3) Marble-bed absorber (Hydro-Filter)
The venturi (manufactured by CHEMICO) contains an adjustable plug,
which permits control of pressure drop under a wide range of flow
conditions. The TCA (manufactured by UOP and described in Ref. 2)
utilizes a packing of low density plastic spheres which are free to move
between retaining grids. The Hydro-Filter (manufactured by National
Dust Collector Corporation and described in Ref. 3) utilizes a packing
of glass spheres (marbles) which are normally in slight vibratory
motion. A "turbulent layer" of liquid and gas above the glass spheres
increases mass transfer and particulate removal.
Models describing pressure drop, particulate and sulfur dioxide removal
within the three scrubbers have been presented in Ref. 4.
SYSTEM FLEXIBILITY
The test facility has been designed to achieve great flexibility:
698
-------
The three scrubbers can be tested in parallel.
Scrubber internals and piping configurations can be changed.
For example, the TCA can be operated as a one, two or three
stage unit, with a variety of liquor flow piping arrangements.
The facility can be operated under various alkali addition
modes. These include: limestone in the scrubber circuit,
hydrate in the scrubber circuit and limestone (or dolomite)
injection in the boiler. For the scrubber circuit addition
modes, alkali can be added at two locations, the Scrubber
Effluent Hold Tank or the Process Water Hold Tank.
Various solid disposal configurations can be evaluated. They
include: Clarifier/Pond
Clarifier / Centrifuge /Pond
Clarifier / Filter /Pond
Each scrubber has been furnished with a quench spray to
permit humidification and cooling of the flue gas prior to
contact with the scrubbing slurry.
A heat exchanger has been provided to permit cooling of inlet
scrubber liquor (only one scrubber at a time).
Each scrubber has its own oil-fired reheater (and stack) to
increase the temperature of the exit gas over wet-bulb values.
Each scrubber will accommodate either a chevron or centrifu-
gal mist eliminator (demister).
SYSTEM RELIABILITY
Difficulties in achieving reliable operation of wet limestone scrubbing
systems attributable to scaling, plugging, erosion and corrosion have
been reported by many investigators.
The following are among the design and test features included to maxi-
mize system reliability:
699
-------
• The scrubbers are either of stainless steel or Neoprene-lined
carbon steel construction.
• All major piping, pumps and tanks are lined with rubber or
fiberglass reinforced polyester.
• Variable speed pumps have been selected, where practicable,
for flow control to avoid the solids build-up which can be
produced by throttling flow controllers.
• Sootblowers have been installed at each scrubber inlet to mini-
mize solids deposition.
• The flue gas has been heated downstream of the induced draft
fan to reduce solids deposition within the fan.
• The ability to operate over a wide range of independent
variables has been designed into the facility, to increase the
likelihood of achieving operating conditions which are free of
scaling and other operational problems.
• Scrubbers have been selected which are relatively immune to
plugging and are designed to minimize solids build-up at
liquid contact zones.
SYSTEM DESCRIPTION
Typical schematic flow diagrams for the Venturi, TCA and Hydro-Filter
systems are shown in Figures 1, 2 and 3, respectively. Not shown are
flue gas saturation sprays, heat exchanges for cooling of liquor, flow
controls, and process details. These figures are intended to illustrate
the range of operating conditions possible at the test facility.
In Figure 1, a venturi system is shown with a packed-tower (Pall-ring)
after-scr ubber, with hydrate (Ca(OH) ) addition to the Scrubber Effluen
L,
Hold Tank and a Clarifier/Pond combination for waste disposal.
700
-------
In Figure 2, a TCA system is shown with the TCA in a three-stage
configuration, with limestone addition to the Scrubber Effluent Hold
Tank and a Clarifier/Filter/Pond combination for waste disposal. In
Figure 3, a Hydro-Filter system is shown with limestone addition to
the boiler and a Clarifier/Centrifuge/Pond combination for waste dis-
posal. Any of the scrubber systems can be operated with any of the
solids disposal systems and with several piping flow configurations.
For all configurations, gas will be withdrawn from the boiler ahead of
the power plant particulate removal equipment so that the entrained
dust, including lime during injection-scrubbing tests, can be introduced
into the scrubber. Gas flow rate to each scrubber will be measured
by venturi flow tubes and controlled by dampers on the induced-draft
fans. Concentration of sulfur dioxide in the inlet and outlet gas will
be determined continuously. The efficiency of the process for removal
of nitrogen oxides also will be determined by periodic checks of inlet
and outlet concentration.
Control of the scrubbing systems will be carried out from a central
graphic panelboard. An electronic data acquisition system will be
utilized to record the operating data. The system is hard wired for
data output in engineering units directly on magnetic tape. Onsite
display of selected information will be available. Also, important
process control variables will be continuously recorded and trend
recorders will be provided for periodic monitoring of selected data
sources.
701
-------
.2> s £
II
r>
°£i_
LU t
°ogq*
§3oi
QliZlxS
"uj
CO
-®
"
gll
A
A
16
u
lAl
O O- 01
00®
702
-------
ro S j«
« 2 £
^ M. C»
.S° « ^
;H u |_
a.
• —
(/I O
si
O _i
I
I
II
•*
fe
oo®
703
-------
o
5
CO
1 =
i,
E-UP WATER^>— ^
(/l
E£g*
£«°i
%* = *-
(•)
j
/*
^
bo »J
frt o
^ gf
I
o
A
A
ra 4-*
£ ^
II
1
081
II
'
|i|
O OL (/l
O®@
704
-------
Batch samples of coal, limestone, slurry and gas will be taken periodi-
cally during each test run for chemical analysis, particulate size
sampling and limestone reactivity tests. The locations of these sample
points are indicated shown on Figures 1, 2 and 3.
The Shawnee pilot facility contains five major areas: (1) the scrubber
area (including tanks and pumps), (2) the operations building (including
laboratory area, electrical gear, centrifuge and filter), (3) the
thickener area (including pumps and tanks), (4) the utility area
(including air compressors, air dryer, limestone storage silos, mix
tanks, gravimetric-feeder, and pumps), and (5) the pond area. Figure
4 is a perspective sketch of the pilot facility. The pond area is shown
at the extreme right; the three Clarifiers are near the center, adjacent
to the operations building; and the scrubber area is in the foreground
with the boiler facility to the left.
705
-------
/
-------
TEST PROGRAM
DESIGN CONCEPTS
Data from the test program (Phase III) will be utilized to:
Develop mathematical models for process and equipment
scale-up to commercial (multi-hundred megawatt plant)
scrubber systems.
Evaluate the feasibility and long-term reliability of closed-
loop limestone wet-scrubbing systems.
The following test periods have been defined:
(1) Air-Water Tests.
(2) Sodium Carbonate Scrubbing of Air-SO? Gas Mixtures.
(3) Break-In Testing for Wet-Scrubbing of Boiler Flue Gas
with Limestone and Hydrate in the Scrubber Circuit and
with Limestone and Dolomite in the Boiler.
(4) Screening Testing for Wet-Scrubbing of Boiler Flue Gas
with Limestone and Hydrate in the Scrubber Circuit and
with Limestone and Dolomite in the Boiler.
(5) Primary Testing for Wet-Scrubbing of Boiler Flue Gas
with Limestone and Hydrate in the Scrubber Circuit and
with Limestone and Dolomite in the Boiler.
707
-------
Air-Water Tests. The experiments with air and water are designed
to:
(D Determine pressure drop model coefficients for all three
scrubber systems without solids.
(2) Observe the fluid hydrodynamics within the scrubbers in
a clean system, e. g. attempt to determine the droplet
sizes in the venturi throat region.
(3) Determine the quantity of entrained liquid in gas leaving
scrubber demister.
Sodium Carbonate Tests. In these tests, water solutions of sodium
carbonate will be used to scrub SO^ from mixtures of air and SC^.
The purpose of these tests is to determine uncertain model coeffi-
cients for gas-side controlling mass transfer for all three scrubber
systems. These data may also be used to evcJuate and design other
systems which rely on sodium carbonate aqueous scrubbing.
Break-In Tests. The break-in test periods will be used to identify
and resolve possible operating problems prior to initiation of screen-
ing testing. The primary objectives of the break-in tests are to:
(1) Determine scaling and plugging tendencies and evaluate
various methods for scale removal. The limits of the
levels of the most important scale-related variables which
must be set during subsequent testing in order to prevent
scaling will be identified, e. g. suspended solids concen-
tration in scrubber inlet slurry, effluent hold tank resi-
dence time. An attempt will be made to develop reliable
methods for detecting potentially deleterious scaling
before complete pluggage or major scaling has occurred,
708
-------
e. g. pressure drop information, liquor conductivity
measurements, visual observation. Mechanical methods
(chiseling, scraping) as well as chemical methods (muri-
atic acid, oxalic acid, soda ash) will be tested as means
for removing scale.
(2) Determine solids separating and handling sapabilities.
Various solids separating configurations will be evaluated
for their effect upon settling characteristics, scaling ten-
dencies, dewatering capability, reliability and the ability
to approach steady state within the time constraints
imposed during screening testing.
(3) Establish criteria for achievement of steady state condi-
tions and ascertain the time required to achieve steady
state for various operating changes.
(4) Determine the limitations of the system with respect to
control of the independent variables over their required
levels. Also, determine the replicability of dependent
variable response to selected settings of independent
variables.
(5) Evaluate the efficacy of the systems-data measurement
capability.
(6) Supply information regarding the selection of levels of
some of the less important independent variables which
are to be held fixed during the screening testing. Also,
evaluate selected combinations of scrubber/liquor piping
flow configurations.
(7) Determine system reliability for subsequent screening
and primary test requirements.
Screening Tests. The screening tests are designed to:
[1) Characterize, as completely as practicable, the effect of
important independent variables on particulate removal
and SC>2 removal for the three scrubber systems.
709
-------
(2) Compare model predictions for pressure drop, particu-
late removal, and SO,, removal with the data and deter-
mine the best-fit values for uncertain constants and
coefficients within the models.
The screening tests are divided into two general categories:
(1) Factorial tests where the most important independent
variables and selected "secondary" variables are treated
in a fractional factorial test matrix.
(2) Sensitivity tests where less important variables are
tested in an abbreviated test design.
A certain number of undefined tests will be allotted for the screening
tests. This is to allow flexibility by permitting selection of further
test runs influenced by results of prior factorial and sensitivity test-
ing. Also, some of these runs can be used to replicate a previously
tested set of conditions for a relatively long term test (1-2 weeks).
This would allow preliminary verification on a longer term basis of
potentially attractive operating modes, prior to the primary testing
effort which is performed as the last portion of the test program.
Primary Tests. The objectives of the primary tests are to:
(1) Perform selected short-term testing to ascertain close-
to-optimum operating conditions for each scrubber type
and alkali injection mode.
(2) Perform long-term testing (two to four months) on the
most attractive operating modes for each scrubber
system to determine reliability of operation and to
develop data for process economics and for scale-up
to larger systems.
710
-------
TEST SCHEDULE
The Break-In and screening testing will be divided into three blocks
which are to be tested sequentially. They are:
(1) Block #1 - Limestone in Scrubber Circuit
(2) Block #2 - Hydrate in Scrubber Circuit
(3) Block #3 - Limestone in Boiler
The preliminary schedule for the test effort is presented in Figure 5.
The estimated date for initiation of System Check-Out is March 1,
1972. The presented schedule should be considered approximate only,
since it may not be feasible to adhere to a rigid schedule in a research
and development program of this type.
Table 1 presents a description of the reports which are presently
scheduled for general distribution.
TEST PROGRAM VARIABLES AND LEVELS
The levels of the independent variables for the air-water, sodium
carbonate and screening experiments for all three scrubber systems
are given in Tables 2 through 12' . The following conventions have
been adopted in these tables (see Figures 1, 2, and 3):
These represent the levels of the statistically designed runs.
A small number of runs in the test sequence have not been
statistically designed.
711
-------
a
UJ
I
O
O <
iZ X
O_ 111
8
£?
s
I
m
!
I
1
a
s
o
£>•
CO
IX
•O
IO
ROGRAM FUNCTION
i
i
i
i
i
I
I
ae m
M 2
« w
at:
=
ot=
o
at
si
si
fc;
o
o
z
o
.. 6
G
fe ^
888
Si 5* 3J
712
-------
0)
I—*
.a
rt
H
2
H
u
w
w
Q
O
(X
w
Q
4)
4->
nl
P
a
•o .2
4> •*->
-u rt)
nl U
a 3
*•§
m /if
W &
i-H
at
rH
4)
G
4)
O
fH
4)
^
nt
co
rC
^J
O
a
o
r>
£
H
*
to
C
O
•rH
4J
3
i-H
O
CO
/*%
CJ
O
• |H
-l_>
Informa
TJ
0
nt
CO
C
C
4)
i— i
rQ
0
rH
ft
1— 1
n)
C
o
of operati
lx
rH
h
^
3
to
F—1
=fc
M
CJ
0
r-l
«
m
0
d
o
• rH
4-»
4)
r-l
n.
If
0
O
fH
4>
-M
?
fH
•iH
«4
CO
«rH
O
CO
H->
r-l
f*
S3
CO
4)
fH
*
CO
c«
bo
•H
CO
CO
Tl
H->
CO
4)
-M
1— 1
nl
3
H->
CJ
n)
T)
C
nl
T3
0)
a
C
trl
(U
rH
ft
CO
H-*
to
4)
H
•rH
1
M
fH
«
r-H
4)
T3
O
g
H
fH
o
£
*rt
(V
4->
nt
TJ
««
O
C
o
ilizati
4->
3
i
fH
4>
H->
• H
rC
H^
• H
fe
r
bo
•rt
H->
CO
4)
H->
C
•H
ri
nt
4)
rH
rO
M-l
O
CO
be -^
•S 3
-M 4)
: *
-^ .
rT«
O 4)
u a
j^k
rt o
£
T3
C
nt
r-l
4)
4)
T)
i
rt
T3
<+H
O
C
O
•rH
-U
n)
4>
fH
ft
o
o
o a
4) 4) bO
4) 7; C
fH ft "H
£ C 0)
O nj
s«w
£ 4! -H
H n! *
bO
CJ
• rH
S " CO
O n co
1 « g
w
w
ft
i
a
o
*H <« bo
§ °.s
g c a
C o 4)
o 33 £
i^ « n
r <-* °
EH ft t/)
fi r*
0 nt
•H 4)
3 ^
r5 rO
S «H
m O
4)
fH CO
nt
H->
nt
^4
S
°
rt -
nt
a
O
0
5 °.S
c c d
C O 4>
o 33 £
^ « M
r •—' °
H ft OT
fH
T3
C
nt
CO
a
4)
r-l
rO
O
fH
ft
bo
C
• rH
H->
rt
fH
4)
ft
O
WH
O
>^
fH
p
a
P
w
^
CO
4)
fH
•s
CO
CJ
be
•H
CO
4)
T)
4->
CO
4)
-u
i— i
nt
3
-u
O
ni
TJ
C
nt
TJ
4)
CJ
CJ
nj
r— 1
ft
*M
O
CO
B
nl
H->
CO
•k
nt
-u
nl
T3
«*H
O
s
o
interpretati
CO
4->
CO
4)
-4-1
H->
1 developmen
4>
T3
O
a
TJ
C
n)
CO
a
4)
i-H
rCJ
O
fH
ft
r- 1
nt
a
o
• rH
H->
re)
rH
4)
ft
O
MH
O
rT
nt
H
a
$
3
CO
4>
rH
CO"
c;
bo
•H
CO
4>
T)
HJ
CO
4>
4->
ed and actual
JH
nt
t— i
ft
bo
• rH
4->
CO
4)
H->
screening
<+H
O
CO
4->
r-H
3
CO
0)
fH
*
bo
CJ
•|H
•*J
CO
4)
•»->
H-«
4)
fH
ft
rH
4)
-»->
CJ
• rH
•elopment,
r>
4)
Tl
I-H
4)
T)
0
a
CO
m
4)
U
0
fH
ft
above
ro
=tfc
4->
fH
O
ft
4>
fH
CO
nt
4)
a
$
o
^§
4) ™
HJ fH
*+* bo
n! o
S o
a«
C O
4)
i-H
-I-)
• rH
H
-t-> r;
r-l ™
3 °
1 "
c!
O
CO
Te
- L:
o c
H->
• H
3
u
£- -
•S * u
s -s ^
2l|
»?1
«: o «
•H (M
C! =tfc
41 .>
2
.
o
CO
4)
bO
fl
•s
4>
4)
fH
0
CO
-o
CJ
m
CJ
Break-i
-*'
CO
=fe
ty
rX
O
O
rH
r4
«2
CO
H->
1— 1
3
CO
4)
PH
4-J
CO
4)
H
•
fH
4)
t— i
• rH
O
W
CJ
•iH
Limestone
HH
>-J
fH
O
ft
4)
PH
i-H
n)
C
• H
fc
*
in
713
-------
•s
I
h
CU
a
x
W
0)
•l>
cd
>
^
8.
ID
•a
s
•4->
CO
CJ
4-»
•— 1
[r!
O
TJ
$
g
tn
^,
w
^
u
H
§
»
03
3
C
9)
Level!
01
1
|H
£
1
CU
1
h
cd
>
CO
V
J
01
3
2
M
rt
1
0
O O O
O O 0
o o o
o o o
^H CM CO
^H CM CO
tl
4-1
a)
?
O
S
*"
60
Clt
en
^H CO
—1 CM
CO
>
~t CM
c
o
15
a
c
o
u
Q)
^2
|
U
1
o —• N ^
^-i ro m TJ<
i
1>
"Ji *
M 2
o a
(4-1 .
|?
J
1
O
o o o
o o o
o o o
o" o" o
-H CM CO
*-< CM CO
tt)
1
Ci
h
^
1
U
o o o
o o o
o o o
o o o
^ CM CO
- CM CO
CU
K
O
.r^
S§§
O I-H (M
- CM -0
,
4-*
cd
O >•
^ 2
o •
2-3
J
M O O O
O O fM
^ Mm ^
11
*{
CO
0 %
•-H CQ
^ 1-1
O m
If
^
1
60 o o o
O i— i ro ^O
~H ro ro -
rt
t-i
O
Q
2*
c
00
in ^H
-H CM
0)
**
g
TJ
CU
(U
a
en
en
CO
ct)
O
S ^
rt "
^ CM
CU
4J
at
T3
»
O
a
A
a.
o o
^ CM CO
1
« t.
*J "
^ !o
"^-t CO
n ^t
3
-------
c
D
I
|H
8.
w
0
XI
ai
U
id
H
01
—4
XI
nj
6
en
£
0
i
a
s
0»
m
l/>
U
H
i
4J
01
ji
c
0)
en
a;
0)
3
2
^
CD
CO
r-4
01
01
J
in
XI
nl
u
cd
to
r-4
0)
J
0)
XI
a
rd
>
**-<
U
0 0 O
o o o
o o o
00*0
.-« OJ PO
- CM m
0>
nl
o
E
^,
^
V
5
.-I (M
-1 CM
CO
(U
(D
CO
"3
h
XI
s
§,
V
XI
1)
0> fc
C U n
_o to 3
CD 00 &
•- .3 £
Is 2
4) fll pl
t> ft to
— t OJ CO
(3
0
'•s
a
a
o
(J
0>
3
w
s
a
00
o o o
o un o
-4 (NJ Tf
- N fO
U
"rt m
r> fn
1 s-
<*H C
0 °
3 Ji
a- o
J
1
U
o o o
O O 0
o o o
o in o
r-t — 1 OJ
~* OJ CO
0)
4-1
rt
g
o
<«
^
s
"-H
o o o
0 O O
o o o
o un o
--( I-H CM
<-* OJ CO
"rt
f-l
o
.tt
^
1
O —< OJ
^H OJ CO
rt
o ^*
JH ft
o »
^ 0
J
a
o o o
o o o
oJ m oo
i-t OJ CO
CO
4->
2
o ^
^ t-t
o «*
SH O«
"~
1
00
o o o
o to o
-* rj ^
-I (M fl
£
Id
s
o
o
s*
J
00
in ^H
-. N
0)
4J
01
0)
0)
xi
a
CO
CO
fl!
a
fi ^
"*
""
-H (M
01
0)
•0
00
1
U
p,
o o o
Tf v£) 00
^H CM CO
c
0
•y
to
o
a.
00
3
ft
c
CM fo in
-H CM fO
00
c
3
U
n)
a
o
S
00
S
c
'"* O ^f
^ rf -H
-H CM c<1
00
C
U
a
MH
o
no
0)
u
rt
w
c
O CO sO
i-H OJ CO
4J
bp
QJ
XJ
•- .s
£ a
s *^
o
B1
0)
S
P.
P.
nt
T3
XI
O
C
01
U
C
0
U
(M
O
in
715
-------
w>
-S3
n
c
4)
g
t
'?>
.£
n
C
4)
W
1
n
V
V
J
n
C
0)
£
"C
I
s
• H
t)
n
4>
4)
J
g
£
0
O
^
r
f-
D
^ 00
0 ^C
t. JH
I-M
=Ji ^
Venturi 1
Venturi 1
i-i CM
00
c
*?
(0
OH
Venturi &
t<
v
^
o ^
-?•£
h t-
OJ V
'(H h
JS jS
u u
•— CM
TD
c
0
p.
Clarifier
fl
o
a
4J
n)
5
T)
C
ft
Clarifier/
C
o
rt
h
p
S1
Solids handling con
D
£ £
O O
O O
OC 00
2 3
0 0
f-i JH
0 O
^ N
m
0)
£
O
cO
00
P
0
0
a-
4)
£
O
O
00
p
o
£
V
N
'm
Limestone particle
h
W
h
W
H
Rl
^H fS]
H
ffi
w
*^
d
'o
ft
C
Limestone injectio
l-H CO
- CM
CO
CO
CO
Excess air
< ffl
~H rsj
<
«J
O
U
sajq-ei-iBA. A-r-epuoDag
0
6
6
o
1 Saturation sprays
|
o
o
o
o
t-,
o
p
cr
"c
Vn
O
bo
fl
O
0
U
o
0
o
o
I Coagulant
P3XT^t
PT3H
716
-------
If
h U
en
C
V
£
s
Q.
X
W
>*•
_^
n
C
OJ
CO
n
!
•iments 1
a
W
5
0
u
o
a>
a>
J
H
a
0
U
G
,
nl
U
£
*-i
u>
D
flj
^
a
g
0
c
7
m
3
O
en
D
O
o
on
"S>
?0% throu
Limestone particle size
H
a
w
H
X S
w ft
-1 ISi
H
a
w
H
H X
a ^'
w cu
#
Limestone injection point
Is
— fN]
CO
sS
00
CO
Excess air
< £0
-M
<
V
a
Ir
•a
0
U
00
..
O
c
O
c
Q
M
ft
h
a
CO
c
0
rt
1
a> o
— N
o
Z
0
Z
o
0
1
u
'o
a
"o
0
U
™—
c
0
CL
flj
Clarifi.
T)
C
0
O-
Clanfiery
T3
C
O
D.
Ll
Clarifie
C
-I1
QJ
rt
§
n)
3
00
g
u
W)
1
^
n
T2
o
o
2
o
0
Z
o
Coagulant
c
c
c
rj
G
Packing diameter
c
0
c
o
G
O
c
0
M
C
,5
u
S.
op
cu
u
S
w
paxij;
717
-------
I s?
HII
M
•D C
TT 1-*
m
C
&)
E
4)
fi1
w
•*j
'>
.*H
(0
a
4)
4)
,4
m
a
1
S
s*
M
"3
'E
0
-4J
U
rt
Ji
CO
4)
i— I
S1
o
O
~"~"
c
i— i
nj
4)
^
CQ
FH
CQ
4)
5
_
A.
0
O
(5
""I*
^
a
^
ffl
4>
4)
|4
[0
O
•4-1
U
Controlla
o
in
m m
co r--
-~
o
in
in o m o
«M m r- o
I-H rj co •*
.2
"rt
Stoichiometric
o
00 r-<
-M
£2
" N
•^ CO 1-4 «-H
i— i f\] CO Tj<
•o
0
(1)
u
'o
4)
Percent solids
g
u
o
o
o
o*
PO
p
£
u
o o
o o
o o
in o
-co
g
u
o
o
o
o"
rO
g
is
u
o
o
o
in
4)
m
o
C
to
o
1
ac
0
o
B
p.
bo
o o
0 0
•-H CJ
PH
o
o
E
a
be
o o
o o
CO Tf
•-H f\J
CO
rt
PH
I/I
g
0
"o
4)
rt
O
S
er
c
1
O
c
E
^H ro
-S
1
o
c
E
U") O
•-j (M
4}
O
c
4)
O
(H
H
W
C
c
Tf CO
-< ro
^
c
c
u
p-
D
>
0
rt
be
Overflow weir '.
tsj
CO
ro co
CO rO
-ca
S
ro co
rO co
1-1 ro
*
C
O
Flow configura
4)
bO
^
m *^
C C
0 4)
P. O
Clarifier,
Clarifier,
r-n ro
c
0
'-^
Clarifiei
4)
CJ
IM
TJ -M S
O 4) ^
a u c!
0) 4) 0)
rt rt «j
u G G
«-• ro ro
o
a
OJ
^H
rt
5
*
C
c
'•£
CO
(H
ao
c
u
Solids handling
- ro
<
<
Limestone type
^
CQ
0)
E
O
O
CO
A
M
0
1
o
tn to
E E
O O
O O
oo rj
& -C
2 2
0 0
O O
i-- (M
to
4)
O
o
CO
_
be
0
o
-C
4>
g
O
0
CO
_
be
o
tn
0
to
4)
U
Limestone part
•-H CO
^ ^
CO
CO ^D
-H CM
CO
*
Excess air
< CQ
- ro
<
<
0>
P-
C
U
<§§
I-H ro
o
o
0
w
Satura tion spra
j
o
Z
o
0
in
4) O
^ Z
I-H ro
o
cr
C
C
be
C
o
0
U
c
0
QJ O
>• z
" N
c
*
Coagulant
U
Z
o.
00
o o
0 0
rH CNJ
^H fvj
O
z
g
a
00
o
o
^
m
a
P.
0
IH
0
C
3
cr
3
— —
H
K
W
H
EC
W
H
EC
W
f-H
EC
W
_!_,
C
'3
a
c
0
u
Limestone inje
CO
c
CO
c
CO
in
j
CO
IT
lameter
ID
01
OJ
p.
w
y>
a
0
c
cO
c*
CO
C
CO
C
CO
C
U
ra
P-
bC
0)
K
PI^H
718
-------
« 6
c £
a
H
•
rt i
h U
T3 I.
C u
0) W5
a
D B
"O '
C tl
01 >.
riments
(X
X
U
• Factoria
s,
u>
0>
OJ
J
,3
i
rt
ffl
o
[0
QJ
J
M
o
4->
O
(U
Controllab!
tower
£
D M
1 '
Venturi a
Venturi £
— < ro
0)
o
4-1
nt
t-.
CO
Venturi &
o
"rt
t-t
Scrubber configu
O CO
o *-*
^H ro
ON ^ CO i«
- M CO ^
O
fl)
Stoichiometric r
^ (NJ
CO ^
-H N
6^ {NJ ^
-H N m *
"U
•£>
£
O
0)
Percent solids r
1
u
o o
0 0
0 0
iD O
^H CO
-H (N]
a
u
0 0
o o
o o
-H rvj
0)
rt
o
CO
nJ
O
tx
GC
O o
0 o
^ N
a
O
o
4-i
G
>
1)
"rt
fn
O
'^-^
O
a
GO
0 0
O O
CO O
- N
ex
GO
o o
o o
-
-------
W u
K H
0)
H
CO
C
g
fn
OJ
W
15
'£
O
u
o
to
0)
(U
5
^
rt
0)
M
CQ
*n
5
to
!
Controllable Factors
CO
00
rt
to
to
00
rt
Cfl
Number of stages
o m
O — i
. . . .
^H (M r<1 -^
Stoichiometric ratio
^ tM
00 -H
— « (M
^ (M xO
•— « CM cO ^
TJ
1)
"rt
Percent solids recircul
..„,„„
g
u
o o
o o
0 0
m o
1
u
0 0
0 0
0 0
ITi O
~N
+H
0
00
c
'o
o
u
CO
CO
Excess air
PT
sajqT
u
O.
-H
rt
o
U
3H
o
Z
o
Coagulant
L
C
CM
C
CM
~*
Packing diameter
on
c
X
c
H
720
-------
SI
B «
s|
£ 2
° n
B
"
Height of packing
PT^H
an
B
.c
B
H
*
721
-------
is
C -M
^
CO
cu
§
cu
a
W
£
r Sensitivi
0
en
4)
cu
J
n
•M
c
cu
£
i-t
rx
w
• Factor la
0
CA
O
TJ
jj
1>
0
O
f~l
bC
0
O
in tn
£ £
O 0
O O
OO fM
QD W)
0 0
!H J-(
0 O
-i 00
VI
cu
g
o
o
X
00
o
(-1
4-1
o
cu
CO
CU
o
rd
a
CU
TJ
CO O
—i m
(-1 fM
CO
CO
Excess air
< CQ
-i fM
g.
4J
•3
O
U
O 0
-> rg
O
c
o
Saturation sprays
0
Z
o
Z
o
z
0
u-
cu
"fj
o
oc
G
"o
0
u
o
o
o
-t-1
c
oJ
o
u
^,
£
0
a)
a
cn
Venturi &
a
o
.
nj
(-1
(X
Venturi &
V
o
>>
(Tj
tx
M
"G
Scrubber configuration
paxij
PT3H
722
-------
en
C
g
JH
CO
^"
W
>x
• Sensitivil
o
{0
CO
CO
CO
"c
CO
SH
(X
Factorial E
IH
O
CO
CO
4)
r
re
£
0
OJ
V
Factors
to
_Q
Controlla
07
00
0)
(U
00
re
(M rO
w
re
-u
en
(M
CO
CO
re
Number of :
o
m
un m
r\] r-
-H rq
m o m o
rj m r- o
^ ^ ^ ^
o
re
u
'£
Stoichiomet
o
00 "-<
^f CO -H -H
-H (NJ fO Tf
TJ
"rt
3
u
o
tu
T3
Precent sol
1
u
o
o
o
o
u
0 O
o o
0 O
un O
-H t\]
g
u
0 O
0 0
0 O
LO O
i— i C.J
0)
re
o
n)
O
a
00
0
o
a
0 O
O IN]
0 -H
— i (M
£
a
0 O
O r\]
""
(T)
a.
a.
0
"re
^
0
o
cr
d
B
0
g
in o
— < sO
-H CNJ
C
g
m o
~-, -.£,
-.
OJ
£
OJ
u
CU
"H
H
W
< CQ
<
cu
rt
"a
0
-u
0)
O
sajq-ei IBA A.-i'eujT j^
rj
CO
N r^
tN] CN]
CN oo
— ' CM
*
O
"S
^
GC
c
o
o
TJ «
O "^
a >*-
Clarifier ;
Clarifier;
^H (N]
•a
c
0
a
tn
(M
„.
OJ
a
U
sa^q-eij-eyv Aj-epuooag
O
c
O
0
rt
a.
Saturation s
o
Z
o
Z
o
S-.
o
3
CT
CU
*^i
Cooling of i
o
2
o
Z
0
Z
Coagulant
C
rj
C
(M
C
fN]
-------
M h
S 2
•*H M
- 1' h
M a>
w rf -^
3 > o
•O •"
g S
£"!"
"3 ^3
CO
V
1
V
OH
)4
w
>.
|S
•"*
-« (M
- CM m Tf
'O
0)
^5
Percent solids recircu
1
u
o
0
0
o*
1
u
o o
o o
o o
m o
-* CO
f-4 f\]
g
O
o o
o o
O 0
in* o"
- (M
0}
O
<*-<
CO
O
83
g
U)
o
o
a
two
O O
o o
(M •*
1
DO
O O
O O
*" N
m
ro
ft
CO
£
0
41
ro
H
O
g-
j
iq«!"
c
1
O
sO
C
e
If) O
-H (M
q
e
in o
- CM
ii)
E.H.T. residence tirrn
A11W
C
sO
.3
C
DO
u
rt
ft
cu
O
Overflow weir height a
ud
<
< n
Boiler injection plane
tM
CO
(M m
-i CM
CM m
— i CM
*
Flow configuration
0
Z
i
HO
§§
^ CM
r-l CM
1
O
O
2
CL
in
a
(U
re
In
&
O
tr
4)
§£
Limesi
Dolomi
^ CM
4)
Limes ton
ri
^imestoi
Additive type
CO
a>
£
O
o
CO
00
3
0
O
10 CO
S S
o o
o o
<*"> fM
bO bC
2 2
A j3
^ ^
o o
<-> CM
CO
V
S
o
rO
^
M)
y
0
JH
O
a>
N
CO
U
nJ
a
(U
.s
8TqB"T
-H CO
1-1 CM
00
CO
Excess air
A
< ffl
^ rg
0)
a
o
U
"""
0 0
-H rxi
O
fi
o
Saturation sprays
in
0) 0
o
Z
0
Cooling of inlet liquor
C
0
ft
^
Clarifi*
•8
O
ft
Clarifier/
13
a
ft
>larifier ,
w
c
O
(_,
Solids handling configu
o
0
o
Z
Coagulant
F
sa^
C
CO
g
CO
in
C
00
Glass sphere diameter
pax^
PH
qi2T it
C
S
.c
00
c
u
rt
a,
Q
tUD
OJ
[
-
724
-------
(1) Limestone type A and B refers to limestone from two
different quarry sites.
(2) Coal type A and B refers to coal of two different sulfur
concentrations, i.e. high and low sulfur coal.
[3) Boiler injection plane A and B refers to two different
temperature regimes, for additive injection, within the
boiler.
[41 E. H. T. and P. W. H. T. refer to the Effluent Hold Tank
and the Process Water Hold Tank, respectively.
5) N. C. refers to a variable which is "not controlled. "
The flow configuration numbers refer to specific flow configurations
for each scrubber system. As an example, the TCA flow configura-
tion numbers 22 and 23 are shown in Figure 6 and 7, respectively.
Selection of Independent Variables
Generally, the independent variables chosen for a specific set of runs
were those variables considered to have a major effect on the cri-
terion (dependent) variables in question. These judgments were based
on the limited existing pilot plant data and the results from mathema-
tical models which predict criterion variables as a function of the
independent variables (see Ref. 4). Also, the number of variables
vhich can be tested within a set of runs is limited by the schedule time
limitations and the desired statistical "resolutions" (see next section).
For the break-in experiments, the dependent variables are SO?
removal, particulate removal, plugging and scaling of critical pieces
of equipment, solids separating capabilities (settling, etc. 1, length
of time to reach "steady state" and system reliability.
725
-------
©r
J
O
Hz
«i O
«s
w 2
B S
^i2
H In
Z Z
W O
W U
U
S w
si
is
x;
<
h
is
Is
Z B!
S w
a H
« <
9*
J UJ
o y5
w w
^
ww
a
j
<
Q
W
OS
O
g
3
W H
-------
7;
H
a D
II
J
2
2 Z
w o
w U
w
BJ W
a
Q
z OS w
s 3s * 5
2 || g a
5 §» 2 5
S 08 J „
m u ., (., * a
fiW'<;>i(Ui
BS X S3 ^ U H <
p g=> JH .5
" o, g a w 3 w
-s
53
-1 < H
Q OS O 5
32<«<
O g p W M
OT H W ^ 'j
KM W
b< OS
W 2 0.
SOU
H U OS
o c M ti <
p p°.3^(
h ^ Z ffl w ,
££££3
2- Z ^ U S (
wo?wg
PS H
o z
u. w
° i s
° s
£ §
« 3 S
S -
= -
S 3
S 2 5
I I
- i 1
S •" c
K g S 5 g
>
f
727
-------
For break-/in experiments with limestone in the scrubber circuit
(Tables 4, 5 and 6), the designations of Group I and Group II refer
to two different experimental sequences (designs^, based upon dif-
ferent groupings of the variables to be investigated, e. g. in the ven-
turi system, the limestone injection point is studied (varied) in Group
II and not in Group I. All of the probable variables which may affect
the criterion variables have been included in these designs (Group I
and II). For example:
• The E. H. T. residence time (studied in Group I for TCA
and Hydro-Filter) may affect the degree of supersaturation
in the slurry system and, hence, influence scaling.
• Addition of coagulant (see Group II, Hydro-Filter) can
affect clarifier settling characteristics and solids build-
up within the system.
• Solids handling configurations (see Group II, Hydro-
Filter) can affect solids separating (dewatering) and
handling characteristics. Also, different configura-
tions can affect reliability and the duration of the approach
to steady-state after system upsets.
• Stoichiometric ratio and the percent solids recirculated
can directly affect system plugging tendencies, clari-
fier settling characteristics, SO^ removal and system
overall chemistry.
• Excess air (studied in Group II, Hydro-Filter can affect
the degree of oxidation of sulfite to sulfate and, hence,
the system settling characteristics, scaling tendencies
and SO2 removal.
• Plug position (Group I, venturi) can affect SG^ removal,
particulate removal and scaling within the venturi.
728
-------
For break-in tests with hydrate in the scrubber circuit and lime-
stone/dolomite in the boiler, (Tables 7 through 12) the variables to be
studied are fewer, since it is assumed that many of the operating
problems will have been solved during testing with limestone in the
scrubber circuit.
For the screening experiments, the criterion variables are SO~
removal and particulate removal. The variables considered to have
significant effects upon these dependent variables are shown in Tables
4 through 12. There are two types of variables considered in the
screening experiments: (1) primary and (2) secondary variables.
Primary variables are those considered to have a major effect upon
the criterion variables while secondary variables are considered to
have lesser, though significant effects. The screening experiments,
as mentioned previously, are of two types:
(1) Factorial tests where the primary variables and "selec-
ted" secondary variables are treated in a fractional
factorial test matrix.
(2) Sensitivity tests where the remainder of the secondary
variables are tested in an abbreviated test design.
Selection of Levels for Independent Variables
The levels of the independent variables have been chosen using engi-
neering judgment, pilot plant test results, the restraints of the system
and results from mathematical models which relate the criterion and
independent variables.
729
-------
In many of the experimental sequences it was preferable to vary cer-
tain controllable factors over their maximum design ranges. For gas
flow rate, the stable operating range is from about 10, 000 to 30, 000
cfm in each scrubber system. For liquor flow rates to the venturi
and after-scrubbers, the controllable operating ranges are from
about 100 to 600 gpm. For the Hydro-Filter bottom and top sprays,
the allowable limits of liquor flow rates are from 100 to 600 gpm and
100 to 400 gpm, respectively. For liquor flow rates to the TCA, the
controllable operating range is from about 200 to 1200 gpm.
As mentioned previously, two of the stated objectives for the break-
in periods are: (1) to provide information regarding the limitations
of the system with respect to control of the independent variables
over their "desired" levels and (2) to define the limits of the levels
of the important scale (or operating-problem) related variables.
Therefore, severe (or unfavorable) operating ranges have been selec-
ted for a number of the independent variables, e. g. stoichiometric
ratio and percent solids recirculated, during the break-in testing.
For the screening experiments, the variations in stoichiometric ratio
from 1. 25 to 1. 75 and from 0. 9 to 1. 3 for limestone addition and
hydrate addition, respectively, were based upon limited pilot plant
data, which indicated reasonable SO,, removals at these levels. The
selected variation of percent solids recirculated from 8 to 12% was
also based on limited data, which indicated that severe scaling or
plugging would likely occur outside of these limits. Maximum ranges
of gas rate variation (10, 000 to 30, 000 cfm) and liquid-to-gas ratio
730
-------
variation (about 10 to 60) were chosen in order to determine scale-up
effects as accurately as possible. Obviously, the choices of the values
of many of the variables listed in Table 4 through 12 may have to be
modified once break-in testing has been completed.
STATISTICAL DESIGNS
The test sequences (presented in Ref. 5) are, generally, partial fac-
torial designs based upon the chosen variables and levels and the
restraints of time as outlined in Figure 5. For the screening (fac-
torial) experiments, the statistical designs are partially replicated
fractional-factorial sequences of Resolution IV which have been selec-
ted with the understanding that an additional few experiments will be
?'; ;'=:
run, if necessary, in order to achieve Resolution V designs for the
significant variables. It is assumed, therefore, that the number of
"significant" variables will actually be less than the number of pri-
mary variables listed in Tables 2 through 12.
TEST SCHEDULE
The total number of scheduled test runs are given in Table 13 and the
estimated duration (line-out and "steady-state") of the runs in Table
14. The complete test schedule is presented in Ref. 6.
A Resolution IV design is a plan in which no main effect is con-
founded with any other main effect or two-factor interaction, but
where two-factor interactions are confounded with one another.
A Resolution V design is a plan in which no main effect or two-
factor interaction is confounded with any other main effect or
two-factor interaction but two factor interactions are confounded
with three factor interactions.
731
-------
Table 13
NUMBER OF SCHEDULED RUNS IN
EPA TEST PROGRAM
Test Program Function
Venturi
Air-Water 20
Sodium Carbonate 24
Limestone in Scrubber (Block #1)
Break-In 32
Factorial 40
Sensitivity 8
Undefined 8
Hydrate in Scrubber (Block #2)
Break-In 16
Factorial 32
Undefined 8
Limestone in Boiler (Block #3)
Break-In 19
Factorial 40
Sensitivity 10
Undefined 8
Number of Runs
TCA Hydro -Filter
20 20
20 20
30 32
32 32
10 10
8 8
16 16
32 32
8 8
19 19
40 40
10 10
8 8
Primary Testing To be defined
732
-------
P
W
J
D
O
u
o
C t3
S <«
H 5
oiracooitncnwtntntntnoiw
>^>,>,>,>x>N>,>,!>,>,>,;>,>,
—i
oj O
O O
O H
m m
H y
< H
a; w
w 3
* a,
tn tn
H
O -H —<
Q
W
H
en
W
0
s
M _ _ t!
T
e
fin
733
-------
In Table 15, examples of TEST SCHEDULE SHEETS for TCA system
factorial experiments with limestone in the scrubber circuit (Block
#1) are presented. The second column refers to the "statistical
design" table and run numbers, which have not been presented.
Generally, the test design for all three scrubber systems must be
"meshed, " so that the scrubber systems can run simultaneously.
Variables to be meshed (i. e. , variables which are common to all
scrubber systems) include limestone size, limestone type, and coal
type. In addition, stoichiometric ratio must be meshed for runs
where limestone is injected into the boiler (Block #3).
The s-ame conventions used in Tables 2 through 12 have been used in
Table 15.
ANALYTICAL SCHEDULE
Batch samples of coal, limestone slurry and gas •will be taken peri-
odically during each test run for chemical analyses, particulate size
sampling and limestone reactivity tests. Locations of sample points
have been shown on Figures 1, 2, and 3. In addition, clarifier settling
tests, filter leaf tests, and filter and centrifuge "operational tests"
will also be conducted periodically, during the break-in periods. A
typical analytical schedule for break-in experiments with limestone
in the scrubber circuit is shown in Table 16. Normally, there will be
a separate analytical schedule sheet for each group or sub-group of
tests.
734
-------
W
O ~
H 2
w W
W £~*
B! <
W o u
§r
i " ."
IHH
g«8
OT § «
H OJ u
HX w
WS
H
U
6*
3 IS
.3
U
jfi
£ P-
,
§ a
cr rt
"1 Pi
•H qj
HH
HH
H H
H PH
K ffi ffi ffi
W W W W
W W
co oo ro ro
no ro CM
~-< ro to
ro ro r\) ro
i i i i
co a* o ~H
ro ro ro ro ro
.2 °
% ^
u 2 ;
a
>
£
a
>
U
•D
>
U J IX tn
735
-------
p ft
c
o
^1
0 .
•O -H
« s
ll
nj rt
O rf
o «
HHHH
ffi K
W U
£ S ^ K K K ffi ^
p,' pi CU A W W H M pi
& X X
A W W
•0 a
•3 o
a X
tn W
.2 ™'
3 g
DO .5
EH
ro
a ^ .c
W J. CO
^_ oj aj ^ *
3 .§ I S «
U J J (X m
736
-------
c s as
S B XS
M
• ii
!M
ill
5
§88
Ssl
IS§6.
a, > O Sj <
ss 5 5 J
JZ25* I
H
- 5
i 5
w
2 S
Is
o
ti •
£-2
'It
= E "8 3
2 2 P
!l<
u i. X *J
8.8.^-fi
C
_«> .
II
o n
^ "
Jj
*° p^
_£*•-!(
£ "8-
JP
jll
«£ 2
. IU UJ
3 2 s
U DC D
|o§
C x
I I
1 <^
i.'
I "31SM
if
COAL
LIMESTONE
HVDRATE
PARTICIPATES aCUEB t.^E
s
™
POND RETURN ' IQUtC
NOWWOD
5*1
n
! PARTICULATES.FLJE GAS FEE
^1 i irt ^
PA«T,CULAI£S,SCRJB«ED GA
PARTICULATES,SCRU»MD GA
i
S
1^
o
z
s
^
iTii^
Q
?
i
SSiS,
3 > >•
"" at1 or
PROCESS WATER HOLD TANK
AFm-SCRUMER OUKET iLU
AFTER-iCRueMR OUTLET SLU
CLARIFIER UNDERFLOW"
' a 1^1
CIARIFIER OVERFLOW
FILTRATE RETURN LIQiX>R
CENTRATE RETUBN LIQJJOR
FILTER CAKE
1
>HniN3A
H
o
S
PAHTICULATES.SCBUMED GA
•5
ix-
PARTICULATES,SCIUMED GA
RECIRCULATING TANK INLET
, 3
!
> 3
ri
>- i-
RECIRCUIATING TANK OUTL
5
o
1
1
jjj
u
^
W
j
8
<
J
i
TCAQUILtl SilJRftY 14}
TCA OUTLET SIU«RV (5)
CLARIFIER UNO£RKtj»'
CLARIFttR OVERFLOW
FILTRATE RETURN LiCuOR
CENTRATE RETURN LIGUOP
FILTER CAkE
<
J
1
n
m
CN Csi
9
1
O
•?
PARTtCULATES.FLUE GAS FEE
PARTICIPATES, SCRUBBED GA
CO
9
01 '
fA«T(CULATES,SC«U«SfO GA
OVERFLOW WEIR OUTLET SLL*
s
1 11,
i- u O ' ^
SLAKING TANK O'HlET SLUR
PROCHSWATER HOtDTANK
PROCESS WATER HOLoTANK
a 311
HYOHO-FUJER OUTLfT StUflR
CLARIFIER UNDERFLOW
"CIARIFIER OVERFlOW
FILTRATE RETURf- .lQUG«
CINTRATI RfTJ*' -I1,|0«
FIL'TEB CAf-E
CENT»lfUG£ rAM
-O«OAH
CO
•*
r
737
-------
In order to meet the formidable slurry, coal, and alkali analytical
requirements of the facility at reasonable costs, equipment has been
selected that minimizes manpower. For example, an x-ray fluores-
cence unit has been purchased, which should allow for comprehensive
slurry analyses with reasonable manpower requirements. A summary
of batch analytical methods for determining important species in
slurry, coal and alkali is presented in Table 17.
In the particulate sampling area, equipment will be available to mea-
sure mass loading in the scrubber inlets and outlets. If the results
of an EPA sponsored program with McCrone Associates are success-
ful, this data will be supplemented with particulate size distribution
data obtained with a multi-cyclone device.
Equipment has been carefully selected for gas analyses. For
analysis, six "Du Pont" photometric analyzers will be utilized for con
tinuous operation at all three scrubber inlets and outlets. For HO,
LJ
> N and Q£ determinations, two "Process Analyzer" gas chro-
matographs have been purchased for intermittent operation at each
scrubber inlet and outlet as well as two "Environmetrics" NO
analyzers. pH will be monitored on a continuous basis using fifteen
"Universal Interlox" pH analyzers. Three "Universal Interlox"
electrolytic analyzers will be used to monitor electrical conductivity.
738
-------
Table 17
FIELD METHODS FOR BATCH CHEMICAL ANALYSIS
OF SLURRY, COAL AND ALKALI SAMPLES
Species Desired
Sodium
Potassium
Calcium
Magnesium
Choride
Total Sulfur
Field Method
Atomic Absorption
X-Ray Fluorescence
Total Sulfite and Bi-Sulfite
Dead Stop lodometric
Total Carbonate and Bi-
Carbonate
Nitrite
Nitrate
Infrared Analyser
Ultra-Violet Technique
739
-------
DATA ACQUISITION
Over 150 pieces of data (flow rates, temperatures, pH's etc. ) are to
be recorded automatically and continuously onto magnetic tape at the
pilot plant site. At Bechtel Corporation, in San Francisco, all field
process data (including results of analytical analyses and manually
recorded information) will be placed into a "random access" computer
file. The data will then be printed and plotted in suitable formats for
distribution and assessment by Bechtel, EPA, and TVA personnel. A
computer program will average data during the "Steady-state" run
sequences, identify possible erroneous data points, etc. Another
program will "adjust" data which are functions of temperature and/or
pressure and convert certain data to more useful units.
DATA ANALYSIS
Statistical computer programs will analyse the data to determine, for
the major dependent variable (SO? removal), the magnitude of the main
effects and the two-factor interactions for all the major independent
variables, as well as the magnitude of time trends, systematic
upsets, and random error.
Model predictions for the dependent variables of pressure drop, par-
ticulate removal, and SO_ removal (see Ref. 4) will be compared
•with the data and the "best-fit" values for uncertain constants and
coefficients within the models calculated. For pressure drop and
particulate removal, the form of the proposed models can also be
checked. All models will be modified, if necessary, for better fit
to the data.
740
-------
SUMMARY
The Office of Air Programs (OAP) of the Environmental Protection
Agency (EPA) has sponsored a program to fully characterize wet
limestone scrubbing for removal of sulfur dioxide and particulates
from boiler flue gas. Thes test facility consists of three parallel
scrubber systems, each capable of treating approximately 30,000
acfm of flue gas, which are integrated into the flue gas ductwork of
an existing coal-fired boiler at the TVA Shawnee Power Station,
Paducah, Kentucky. Construction at the TVA power station is approxi-
mately 80 percent complete, with start-up presently scheduled for
March, 1972.
Bechtel Corporation, as the major contractor, has prepared the
detailed design of the test facility, is developing the test program and
will direct the test efforts. TVA is constructing the facility and will
operate the unit during the test program, which is scheduled to last
30 months.
Three basic test periods have been defined:
(1) Break-In Tests
(2) Screening Tests
(3) Primary Tests
741
-------
"Break-in" testing will ascertain operational problems and variable
control limitations and establish some attractive operating configura-
tions.
"Screening" testing will define the influence of process independent
variables on the dependent (performance) variables. Results from
these tests will be used to perfect the mathematical models, which
will then be valid for optimization and commercial scale-up.
Primary testing will be used to ascertain conditions for maximum
performance of the scrubber systems and to demonstrate long-term
process economics and reliability.
742
-------
REFERENCES
Bechtel Corporation, Alkali Scrubbing Test Facility - Operating
Manual for the EPA Alkali Scrubbing Test Facility at the TVA
Shawnee Power Plant, October 1971
Universal Oil Products, Air Correction Division, Bulletin No.
608, "UOP" Wet Scrubbers", 1967
National Dust Collector Corporation, General Catalog,
December 23, 1968
M. Epstein, et al, Bechtel Corporation,"Mathematical Models
for Pressure Drop, Particulate Removal and Sulfur Dioxide
Removal in Venturi, TCA, and Hydro-Filter Scrubbers, pre-
sented at Lime /Limestone Symposium, New Orleans, Nov-
ember 8-12, 1971
Bechtel Corporation, Alkali Scrubbing Test Facility - Statistical
Design of Experiments for Venturi, TCA, and Hydro-Filter
Scrubbers, to be published.
Bechtel Corporation, Alkali Scrubbing Test Facility-Test Pro-
gram Manual for the EPA Alkali Scrubbing Test Facility at the
TVA Shawnee Power Plant, October 1971
743
-------
-------
ECONOMICS OF LIMESTONE WET SCRUBBING SYSTEM
R.M. Sherwin
I.A. Raben
P.P. Anas
Bechtel Corporation
Prepared for
Second International Lime/Limestone
Wet Scrubbing Symposium
New Orleans, Louisiana
November 8-12, 1971
745
-------
ECONOMICS OF LIMESTONE WET SCRUBBING SYSTEM
R. M. Sherwin
I. A. Raben
P. P. Anas
of Bechtel Corporation
ABSTRACT
This paper describes factors influencing the costs of full-scale
limestone scrubbing installations. Data on capital costs are
presented, together with process and mechanical design criteria
which most greatly affect these costs. Among the latter are
corrosivity, erosion, scaling, local site factors, environmental
impact, weatherizing, front and back end logistics, retrofit
limitations, operating control philosophy. A case history is
cited for an eastern utility planning retrofit scrubbers on a
multiple-boiler installation. Operating costs are broken down
into their components, and a range for each of these is given.
The importance of these cost considerations in planning programs
for compliance is examined.
Prepared for presentation at the International
Symposium for Wet Limestone Scrubbing,
Nov. 8-12, 1971, New Orleans, Louisiana
746
-------
INTRODUCTION AND BACKGROUND
The Environmental Protection Agency, acting in accordance with the
mandates of the 1970 Clean Air Act, has set in motion the regulatory
mechanism to achieve primary ambient standards of air quality within 3
years following June,1972. The states are busy preparing their programs
for implementing this time table, and these are to be submitted to EPA
by February, 1972. Many of the major urban areas have already enacted
portions of these plans into local regulations which must be complied
with by June, 1975.
Major operators of stationary combustion equipment are concerned with
the effects of this activity on their overall operations. Not only are major
capital outlays indicated, over and above those traditionally associated
with fossil fuel equipment, but extra operating costs as well. Those of us
who are following the progress of stack gas cleanup the closest would be
less than frank if we did not recognize that these costs are exceeding
estimates made during the early, conceptual stages of this program.
Nevertheless, we should keep one thing in mind: the same factors contri-
buting to spiraling costs here are to some degree at work in the other
avenues available for emission control. While of scant comfort to a major
utility, this should be somewhat reassuring to the engineer charged with
finding the lowest-cost means of compliance. The recognition of what
these cost components are, furthermore, provides a basis for challenge,
comparison, and evaluation of progress being made elsewhere in the whole
spectrum of pollution-control technology.
Early this year a major eastern utility retained Bechtel Corporation to
prepare a study and estimate of fuel-burning alternates for one of its major
existing facilities. Involved in the picture were four identical 180 Mw
pulverized coal boilers burning 3-1/2% sulfur coal. Because of scheduled
availability only one unit could be converted at a time, and it was necessary
to make the conversion over a three-year period, in time for EPA deadlines.
The company was confronted with a number of possibilities, including
(a) high-sulfur coal with both particulate and SO scrubbing, as well as,
(b) low-sulfur coal with particulate scrubbing only. The existing rquipment
was laid out in a typically close-coupled configuration, raising many questions
of ducting, fan placement, and tandem vs. booster staging of fans. The
capital costs for these two alternates are shown in Exhibits I and II (slide).
It will be noted that a system for particulate removal only involved about
60% of the outlay of a system for particulate and SO together, using rear-
end limestone addition.
747
-------
The following factors helped contribute in a major way to the capital cost
of this facility, and these should be particularly noted by anyone who leans
on generalized cost concepts, such as the convenient (but undependable)
"0. 6 factor" approach for SCX removal in power plants.
Raw Material Handling. Limestone was delivered by barge 9 months
out of the year. This required a totally new conveying system, dedication
of area for stockpiling, apart from the coal pile, and displacement of
existing secondary service facilities. Supply logistics dictated a relatively
high investment of this type to provide maximum long-term limestone
economy and reliability of supply. Conveyors were sized for high-speed
discharge rates to minimize conflicts in scheduling with coal barge
deliveries. High reclaiming rates with large 24-hour silos were indicated
to minimize the use of non-day shift
Under conditions of delivery by truck, this portion of the investment could
have been lower by $2/kw.
Limestone Milling. A complete, integrated milling operation was planned.
The delivery of pulverized limestone by pipeline or truck from a central
milling point would have lowered costs. However, the feasibility of such
an operation was felt to be conjectural without a more extensive study.
Should future study demonstrate the practicality of off- site milling, the
investment in both milling and raw stone handling could be transferred to
a second party.
Scrubbing System. The scrubbing system is the heart of the operation and
incurs the major cost burden, being 60-65% of the total investment. Some
of the design-cost factors which have confronted system designers are
listed below:
(1) Should a booster fan be added in series with existing fan,
or should a completely new fan be specified?
(2) Should "pusher" fans be used, at the risk of eroding the fan
blades, or should induced-draft fans be employed, with the
possibility of high dewpoint corrosion and imbalancing?
(3) What corrosion mechanism is most likely, and what material
is most suitable for fan blades following the scrubber?
Stainless steel is costly and may be susceptible to stress
corrosion. Peripheral speed considerations limit the use of
748
-------
rubber-lined rotors. Certain alloys of interest, such as
Corten, cannot be welded with safety. Deposits of slurry
may build up excessive unbalanced forces on the fan blades,
requiring periodic washing. Can washing be done on the
line?
(4) Are reheat temperatures dictated by plume-drop considerations,
(minimal reheat) or by dew-point corrosion factors (maximum
reheat)? Is the existing stack properly lined with impervious
insulating material so as to accept a flue gas with dewpoint
characteristics?
(5) What is the most economical and lowest maintenance method
of reheat —direct injection of combustion gas, or steam
coils ?
(6) What is the most suitable corrosion-resistant medium for
field fabrication and assembly? For example, a shop-applied
soft rubber lining appears to be the ideal protection for bi-
sulfite slurry environments. (Field-applied rubber linings
are generally more costly and less reliable than those applied
under shop conditions. ) For large vessels, however, there
is no choice but field fabrication (of the steel), which means
field application of the coating. Under such circumstances
the cost of rubber linings is hard to justify.
(7) How far can one go in achieving economy of scale? For
present scrubbers, the size limit for a single train is thought
to be about 400, 000 cfm. There are a number of reasons for
this, including non-uniform gas distribution problems (e. g. ,
if scrubbers are bigger, inlet ducts must be bigger) and the
limited availability of matching fan capacity.
The size of I. D. fans will be physically limited by the maximum
flow and pressure drop of the largest boiler units now being
designed.
For ball mills, there is almost no limit, although at through-
puts above 100 tons per hour (enough for 1000 Mw), there is
greater concern for reliability. With skilled maintenance,
bearing and liner repairs can normally be taken care of within
48 hours.
749
-------
(8) What should the philosophy be in terms of providing standby
pumps, mills, and scrubbers? For example, one major
installation includes seven large circulating pumps, although
the theoretical number needed is only four. Proposed EPA
standards for new sources do not permit loss of control (i. e. ,
as stated in the allowable Z-hour limit) more than once a year.
This is a serious and costly constraint in terms of capital
investment.
(9) Does damper design need to be positive shut-off, and if so,
is double-dampering a reliable answer? Double dampers with
intermediate air purging are preferred by some, whereas
others insist upon a heavy crane-supported gate.
(10) To what extent is winterizing necessary? Here is an installation
without winterizing (slide) and here is the same installation
with enclosure (slide). Closing in of the equipment reduces
exposure to freezing in small lines during down periods and
also provides greater comfort during unscheduled maintenance.
It may also encourage closer attention to the equipment, a
matter of paramount interest during the early stages of equip-
ment start-up.
(11) What type of instrumentation is most likely to satisfy conditions
of power plant reliability? What are the response characteristics
of fan-damper combinations under conditions requiring emergency
by-pass? During such a period the operator must isolate 30 inches
pressure drop out of a system total of 40 inches or more without
pressurizing the boiler. This problem relates directly to the
aforementioned question of maximum sizing for fans. If you are
using a booster fan in series with an existing I. D. fan on a
retrofit job, the problem may be compounded.
(12) What method of solids thickening is to be used? Does it stop at
a pumpable consistency, or must filtration be employed? Can
hydroclones or centrifuges be used without fear of slimes build-
up? Are clarifiers the only safe answer? So far we have en-
countered no proposals which stop short of complete ponding or
clarification.
All the foregoing considerations were realistically evaluated and taken into
account in the aforementioned study. The results were neither excessively
conservative nor overly optimistic. For instance, a successful working design
throughout was anticipated as a matter of fact, despite the total lack of
operational experience. On the other hand, reasonably generous sparing
was provided.
750
-------
Solids Handling. One of the more difficult aspects of limestone scrubbing is
how to handle the waste solids. As typical wet filter cake, these account
for approximately 40% of the coal tonnages. The rheological properties of
these materials have proven to be somewhat unpredictable. In a scrubbing
system where the alkali efficiency is low, virgin limestone comes through
unaltered, a relatively high degree of dewatering will occur, the density
and solids content will be high, but the cake may still slump. At higher
efficiencies dewatering is incomplete, the density and solids content drop;
on the other hand, the cake may retain its stiffness. How to stockpile, reload,
and transport this material is a major challenge. Liquidity and plasticity
of the filter cake are important criteria for design of such systems, and are
variant.
For the plant in question, ponding was conceived as the best method for
disposal —at a point four miles distant. (An alternate consideration involved
total dewatering, folio-wed by day-time reloading, and truck disposal.) Rubber-
lined pipe delivered underflow from a battery of liquid cyclone thickeners out
to the pond, and carbon steel pipe delivered the pond effluent back to the
process.
One of the major cost elements here -was a decision to provide all of the
diking required for the life of the pond in advance, rather than in small incre-
ments involving a greater earthmoving total over the long haul. The argument
here is manifested in a discount cash flow (DCF) trade-off of operating costs
vs. capital costs. The capital costs shown in Exhibit I could be substantially
reduced at the expense of increased operating costs in subsequent years.
Diking, incidentally, accounted for $3.40 per kilowatt, or about 35% of the
total solids disposal costs.
In summary, we have so far described a retrofit conversion job -which would
cost $43 per kw if done all at once. Since that is impossible, it is necessary
to add escalation factors. These bring the cost to approximately $50/kw.
A large-scale modern installation would hope to achieve economies of layout,
scale, and logistical planning such as to reduce all elements appreciably.
751
-------
OPERATING COSTS
It is important to note the overwhelming effect which capital investment
has upon operating costs. Exhibit III shows what may be expected for
operating costs of limestone scrubbing across a wide range of power plant
sizes (slide). (This data is typical only, and does not represent information
from operating records or projections of Bechtel clients. ) These are shown
in terms of cost per ton of coal, as these are the units in which operations
people think. Costs are divided into two categories —those which are a
constant expense burden, regardless of load factor, and those which vary
with the amount of fuel burned.
Capital charges account for about 95% of the non-varying costs. As percent
of investment they may range from 12 to 15%. We have assumed a figure of
14%. Financial and accounting people might object to calling this a non-
varying cost. It is really an average cost, made non-variant by a process
of amortization. A typical exercise showing how such a figure is developed
has been presented by H. W. Elder in a previous paper (1).
Prediction of maintenance costs is always a tricky business, since these are
traditionally developed from operating experience. It must be remembered
that such experience cannot yet be brought to bear on this type of installation.
Chemical plants typically allow for a percentage of investment ranging from
less than 2% to over 4%. A more rational approach is the "investment-year"
basis (2), a typical example of which is shown in Exhibit V (slide). If the
experience of SO_ scrubbers proves equal to coke plants, for example, one
might approach 2% maintenance after the fifth year. For the purposes shown
we have been cautiously optimistic in assuming a figure of this magnitude.
This compares with a rate of 1. 4 to 1. 7% for many power plants today.
The number of operators required, -while low, depends on the size and number
of units and upon considerations of retrofit vs. advance-planning.
We have not attempted to assess non-varying costs on a "per ton of coal" basis,
as these change too much with size and capital investment. However, for
each individual case one can approximate a unit cost, which will be somewhere
in the range shown.
Among the variable costs, the largest unit item, next to limestone, may be
waste disposal if trucking is considered. This is not included in Exhibits III
and IV because the station has access to ponding. As previously noted, the
cost of amortized pond preparation ends up instead in the capital charge
account.
752
-------
Power is a significant item for high-energy scrubbers —approximately 4%
of boiler capability. The cost of reheat may require 2-1/2% of steam
capability, unless the plant resorts to direct oil-or-gas fired injection. In
this case station efficiency is not affected, but fuels more costly than coal
are required. There are other potential maintenance and operating advantages
to this type of reheat —for example, freedom from tube scaling. The
ultimate trade-off still remains to be seen. We have yet to encounter a
regenerative reheat system, although the economics are claimed to be
favorable.
We have seen that the capital costs of scrubbing for particulate removal only
run about 40% less than scrubbing for both SO _and particulate removal.
Exhibit IV (slide) shows what the combined effect of altered technology and
reduced investment does to operating costs. This demonstrates that the
reduction in operating costs for particulate removal only is likewise about
40%.
753
-------
COMBINATION OF PRECIPITATORS FOR PARTICULATE REMOVAL WITH
SCRUBBERS FOR SC> REMOVAL
A frequently-asked question is: "Are there cost advantages to an electrostat'
using a precipitator for particulate removal, followed by a wet scrubber for
SO ? " Some of the disadvantages are:
L*
o an excessive space requirement, particularly in retrofit
installations;
o the need for two separate waste handling systems.
On the other hand, there may be some advantages, such as:
o a reduction in capital cost (compared with 2-stage, all-
wet scrubbing);
o an improvement in handling characteristics of waste solids,
obtainable by blending of the lime solids filter cake with the
precipitator fly ash.
From what we have just seen, if we were to leave off the second stage, i. e. ,
omit SO_ scrubbing, we would achieve about 40% reduction in costs. Looking
at it in reverse, omission of the first stage and retention of the second might
realize a cost reduction of $10-20 per kw which could be applied toward an
investment in precipitators. This does not automatically establish a pre-
ference for 1st-stage electrostatic precipitators. On the other hand, the
deliberate collection of dry fly ash for blending with filter cake may offer
important savings in material handling costs. Exhibit VI (slide) shows the
approximate relationship between limestone efficiency and the degree
of mechanical moisture typically encountered in a filter cake. The greater
the limestone efficiency, the higher the moisture content to be trucked away.
As previously noted, this material may be much more manageable in terms
of stiffness, low plasticity, etc. , than one of lower water (but higher
limestone) content. The handling problem can be eased somewhat by blending
dry solids with wet filter cake, but its feasibility depends entirely on local
circumstances.
754
-------
OPPORTUNITIES FOR COST REDUCTION
There is great interest in knowing how truly representative these costs will
be for other situations, particularly for new, large-scale fossil plants at
the 500 Mw level and higher. We are making no claims on this score, as
we believe each situation must be considered unto itself. It is, however,
only being realistic to limit the installation of first-generation scrubbers
to retrofit situations. These are at a distinct disadvantage in terms of:
economics of scale, cramped hind quarters, and protracted construction
schedules.
The following criteria deserve continuing investigation, as they would appear
to offer the biggest opportunities for cost reductions. Detailed study is
already underway in many quarters on these possibilities:
o Limestone mills - With good logistics and economics of
scale, large modern power plants should not require more
than one, or at most, two mills for their entire supply.
o Use of lime - If scaling problems can be solved, the potentially
greater efficiencies associated with lime will reduce the
tonnages of raw material and waste solids. This could open
the door to over-the-fence supply economics and do away
with milling requirements altogether.
o Elimination of clarifier - Some day, when process constraints
have been more fully appraised, it may be found possible to
by-pass the clarifier. This was proposed many years ago
by ICI-Howden, but the conditions for its success have not
been fully demonstrated.
o Plant layout - Progress in reduction of cost by appropriate
planning offers hope for appreciable savings. Structural
work and duct fabrication are major cost components so
affected.
o Pond management - Much remains to be done on the flow
classification properties of the various process slurries
obtainable under different conditions before ponding practices
are optimized. These findings will help dictate future operating
practice. There is precedent in the ore beneficiation field,
where the management of pond tailings has been developed into
a fine art.
755
-------
REFERENCES
(1) "Economics of Limestone - Wet Scrubbing"; Elder, H. W. ,
International Symposium on Lime/Lime stone Scrubbing
for SO Control, NAPCA, Pensacola, Florida, March
1970.
(2) "Predicting Maintenance Costs"; Chem. Eng. , July 13, 1959.
756
-------
ffi Q CO
H H Z
£ H 0
^2£
H W n
CO O Q
O pj Q
U On <
O O O O
O O -
•*
H
H
CO
o
U
vD
o
-*
ro
in
in
xD
00
oo
O
Q
CQ H
« O ^
Pi
x
w
H r. I
Z H H
O 2 H
H 1=) p{
co i II
WH^
a J
3 5
•H
n)
r—I
U
(L)
P5
bo
ni
h
O
-M
CO
n)
o
CD
C
O
^_>
CO
cu
0)
a
o
-M
CO
0)
CO
bo
C
O
CO
co
O
ft
CO
•H
Q
T3
fi
O
QJ
-4->
o!
o
•H
C
O
rt)
O
CD
H
in
a
o
CO
C
O
O
bO
•H
CO
cu
In
757
-------
o
u
rt
w
ft
ffi Q w
H W '
g H O
H rt £
co O Q
O rt Q
U ft <
O
U
CO
o o
m I-H
• •
•* in
CM
PO
O
00
00
00
B
EXHIBIT
*&»
A o ft
>BING ON
IT RETR'
E-YEAR
S 2 w
P D oJ
H
H
U
>—i
H
tf
bo
fi
•H
^
0)
i-H
CO
CO
bo
O
0)
CH
(U
fi
fi
V
o
a
o
O
CQ
T)
•H
I-H
O
CO
0
0)
CO
bO
"§
^
U
CO
n)
co
O
ft
CO
•H
Q
o
ft
«— t
T3
CO
4-1
n)
^
O
fn
ft
0)
X!
O
4->
CO
rH
fl
o
•H
_j._\
aJ
r— \
n)
o
CO
W
fl
o
• H
-4->
O
53
h
-»j
CO
ti
O
O
ao
a
•I*H
f-i
2
T3
4->
CO
0)
^
a>
1
in
CO
758
-------
o
a
o
H
f-t
v
OH
i S G N
°
w 1
T °
1-3 0
CO
H i
W ' .--
!> 0 r\
G COSTS FOR "V
PLANTS IN 20
- WITH PONI
§ g
H^
HH
< PQ
rf pq
W &
n rr*
I-M PM
O U
CO
4->
a
cu
Jj
CQ
CO
4j
rt
cu
Requirem
T3
CU
s
J3
CO
CO
<
V
>
fi
percent i
•^
^H
4->
C
0)
&
J *
CO
cu
>
rt
percent i
ro
CO
0 manhour
o
o
o
i— i
i
m
CO
0 manhour
o
o
o*
CO
i
in
i — i
n
MM/year
*J3
•
^
t
i— i
^
CM
•be-
CO
H
co
O
U
w
J
ffl
CO
N
XI fi *H
U cd O
C -*-1
al 2
±J c o
a --j JD
ni J5 ni
U S J
c
o
superv
fn
O
^3
Cti
4->
O
be
.s i
cu
ex
c
o
H
M g
cu °
cxU
•
CO O
o
U
X m
ga
co .T5
co U
•-i o
-M
C d
cd o
o in o
v£) O ^O
ro
m
in
S
o
H
cd xl
cu cu cu v
ex ex ex ex,
vO
moo
.-H O Tt
CO • • •
-eq- -te- -te- -to-
o in o
ro co o
• • • •
o
o
nJ
O
O
rt
o
H
co
H
CO
O
U
CO
W
J
CQ
3
rt
<
HI
^
F)
O
4J
CO
cu
a
i3
ater
ower
•£ 0,
cd
0)
(H
Xi
cu
^i
0
o
o
r— (
:eam
co
^i
n)
4->
O
J->
JD
cH
O
H
Q
tf
O
0) 0)
cx
m
in
n)
CO
O
CX
CO
CO
ni
O
a
759
-------
a
H "3
^O
_ CO CO
percent i
manhour
manhour
0 0
o o
o o
co r-* o*
' CM
ro i
O
i— i
II
f-t
ni
0)
•^
5
i
2
o
2
CO
cu
60
^
n)
41
O
,— 1
ni
S
AH
ni
U
iance
0)
4->
a
••-<
n)
s
t-
O
4->
n)
fH
o
42
n)
J
rvision
0)
£
CO
4!
t-t
O
43
ni
i— i
bo
•4-)
nJ
^
V
A,
O
tn
a
o
H
ts's
o
u
T3 CQ
c
d +s fe «j
.B n) o a>
n)
•4->
O
-4->
I
•§
CO
H
O
EH
0
£
w
O
760
-------
EXHIBIT V
TYPICAL MAINTENANCE
FOR COKE PLANT
4—Investment-oge, million dollors x yeors
80
70
60
50
40
30
20,
10
Cl I I I I
I 1 I I
Fig. 2
I l I
100 200 280
Annual maintenance, thousand dollors
M = 0. 004 I x t - 83, 100
where M = annual maintenance
I = investment
t = age, years
Source: Chem. Eng.
761
-------
EXHIBIT VI
EFFECT OF CHEMICAL REACTION ON THE TRANSFER AND^TORAGE OF FILTER CAKE
80-
CQ
z
<
fe
o
to
O
to
60-
40
(dense, fluid material)
(bulky, friable material)
20 40 60 80
% CONVERSION OF LIMESTONE TO REACTION PRODUCTS
100
762
-------
EXHIBIT VII
ESCALATION
% annual gain
15
Skilled trades
ENR-20-cities avg. rates for bricklayers.
carpenters, and structural ironworkers
1966
71 72 73
Source: ENR
763
-------
-------
THE RC/BAHCO SYSTEM FOR
REMOVAL OF SULFUR OXIDES
AND FLY ASH FROM FLUE GASES
By
J. D. McKenna
R. S. Atkins
COTTRELL ENVIRONMENTAL SYSTEMS, INC.
A Division of Research-Cottrell
Bound Brook, New Jersey
Prepared For:
Presentation at Second International
Lime/Limestone-Wet Scrubbing Symposium
New Orleans, Louisiana
November 8-12, 1971
765
-------
INTRODUCTION
The first widespread application of a flue gas desulfuri-
zation system appears to be underway. There are a number
of variations on the theme but basically it appears that
a throw-away calcium based scrubbing system offers the
most feasible alternative at this point in time. Research-
Cottrell in the belief that no one system can meet the
needs of the total market has endeavored to provide a
number of systems. One of these systems which utilizes
a packed tower has been previously described at this
symposium. A second and completely distinct system
now being offered by Research-Cottrell under license
from Banco will be reported on in this paper.
In 1964, AB Bahco of Sweden initiated investigation of
sulfur dioxide control and after preliminary screening
studies, decided to develop a calcium based scrubbing
system.
In 1966, AB Bahco installed a 1400 scfm pilot unit at
their central heating plant. This facility was operated
for three heating seasons, and studies were conducted on
766
-------
hydrated lime, burned lime and limestone. Results
obtained during this operation were previously reported
by Gustavsson.
The first commercial installation of the system went into
operation in November, 1969, at So'dersjukhuset, a large
hospital in Stockholm. With the success of the first
unit, two additional units were installed at that site.
Each unit accommodates an oil fired steam boiler with
a capacity of 33 metric tons of steam per hour. The
H
first slide shows the installation at Sodersjukhuset.
Gustavsson's paper presented at the 1st International
Lime/Limestone Symposium indicated that this system
was significantly beyond any other in its stage of
commercial development. That symposium fostered the
Research-Cottrell - Bahco relationship which culminated
in their recent license agreement. In order for Research-
Cottrell to arrive at a decision to proceed with the
licensing negotiation, a rigorous techno-economic and
market evaluation of the Bahco SC>2 removal system first
had to be executed. Portions of the techno-economic
evaluation will be reported on here.
767
-------
PROCESS DESCRIPTION
Figure 1 illustrates the major functions of the RC/BAHCO
two-stage scrubber. Flue gas is supplied by an I.D. fan
with the typical draft control mechanism. A secondary
damper, which admits make-up air and automatically main-
tains optimum gas flow in the scrubber during varying
boiler operations, is connected to the vacuum side of
the I.D. fan.
Flue gas is forced into the scrubber inlet (1) where it
is distributed and directed against the surface of a
hydrated lime slurry. There the flue gas reverses
direction and enters the first venturi stage (2). The
gas liquid contacting creates a vigorous cascade of
droplets.
Due to the large flue gas velocity in the column inlet,
the droplets are transported through the venturi scrubber
at a high concentration and turbulence. The droplets are
separated from the flue gas in a centrif'igal force drop
collector (3). Liquid is returned to the first stage
contact zone and the gas passes to the second stage
venturi (4) where the gas-liquid contact process is
repeated.
768
-------
In the second stage drop collector (5), the gas becomes
free of droplets and exits through the chimney (6). The
scrubbing liquid (13) is recirculated to the first stage
contact zone (14). The scrubbing liquid is milk of lime
prepared by dissolving slaked lime in water.
The slaked lime is fed from a bin (7) by a screw conveyor (8)
through a mixer (9) into a dissolver (10) . From the
dissolver, the milk of lime is pumped to the second stage
impingement zone (11) where the liquid height is controlled
by a level tank (12). The liquid overflow is recirculated
to the mixer.
The lower scrubber stage is provided with partially spent
slurry from the two drop collectors. The overflow returns
to the dissolver, which is also the level tank for the
first stage. A regulator keeps the level constant in
the dissolver by supplying fresh water.
A fraction of the return flow from the first drop collector (16)
is continuously separated using a concentration regulator (17)
into a thickener (18) which automatically feeds viscous
waste sludge into a large basin (19). Sludge then can be
pumped to a storage truck for transport to a refuse dump.
769
-------
Alternately, the sludge can be pumped to a settling pond
or disposed of directly; another option would be filtra-
tion of the clarifier bottoms to achieve a higher solids
content before trucking.
PERFORMANCE
In December, 1970, a team of Research-Cottrell's technical
personnel visited Sedersjukhuset in order to observe and
test the Bahco SC>2 removal system. At that time, only
Unit #1 was installed and operating. The main objectives
of this visit were verification of the Bahco S02 removal
claims and also observation of the system's operational
reliability and simplicity. Particulate removal testing
was not given priority at that time since Research-Cottrell
believed that based on its extensive venturi scrubber
experience it could reliably predict the particulate
removal efficiency.
On-site observations of the unit were made for a three
week period. During this time, the unit operated in a
trouble free and reliable manner requiring very little
attention. The scrubber operation was attended to by
the same operator who serviced the boiler. His main
task with respect to the scrubber was that of measuring
770
-------
the pH of the outlet slurry to the sedimentation tank.
This he did hourly and this task required 10 minutes
at the most.
Lime delivery was observed twice. The lime was pneumati-
cally delivered from the truck through a rubber hose to
the lime hopper. This task required about thirty minutes
and was wholly performed by the truck driver.
Waste removal was observed once. The sludge was pumped
to the removal truck and this task was performed again
solely by the truck driver.
An interview of the hospital's chief engineer indicated
that the greatest problem encountered in the first
season's operation of the Bahco scrubber was that the
dampers and flow measuring devices became inoperable
at the very low ambient temperatures prevalent (-25°C).
The sludge handling portion of the unit, therefore, had
to be insulated. At the end of the first season's
operation (i.e. seven months), the unit was still operable
even though there was about one inch of build up
on the wall. The end of the season cleaning was executed
manually by three men in thirty-two hours.
771
-------
Assessment of Performance
While at Sfldersjukhuset, CES personnel measured overall
SC>2 removal efficiencies of the Banco system. These
tests confirmed the efficiency claims of Bahco.
Due to the unseasonably mild weather enjoyed by Stockholm
residents during the first three weeks of December, the
boiler was only operating at 30% of full capacity. The
average steam production was 10 metric tons per hour at
^
a pressure of 24.0 Kg/cm- (341 psi) and temperature of
360° Ci Heavy #4 oil with a sulfur content of approximately
1".5% was consumed at an average rate of 800 liters/hour.
The scrubber was operated at an average pressure drop of
400 mm (15.7" water) and a pH of 5.5-6.0. Lime consumption
was about 70 Ibs/hour. Make-up air, introduced via a
calibrated mechanical damper, was mixed with the flue
gas before entering the scrubber to insure a constant
liquid to gas ratio.
The gas stream was sampled simultaneously at the inlet
(before make-up air is introduced) and the outlet of the
scrubber. The gas was absorbed in 150 ml of 3% hydrogen
peroxide for eleven minutes at a rate of 0.4 cfm. A
suitable sample was extracted and titrated with 0.1 N
sodium hydroxide to a methyl-orange endpoint. Three
772
-------
different tests verified the Bahco claims of high SC>2
removal efficiency. These tests are shown in Table 1.
A comparison of these results with Bahco data are shown
in Figure 2. It must be noted that the values obtained
for SOj concentration at the inlet are actual measurements
of SC>2 concentration coming from the boiler and need to
be adjusted for the dilution of fresh air.
ECONOMICS
A plot of RC/Bahco system capital costs versus gas
volume is shown in Figure 3. Represented here are the
estimated bugetary selling price for an installed
module; thus, all equipment shown in Figure 1 are
included. This, however, is not a turnkey cost; and,
as such, does not include any unique installation costs
such as interconnecting duct work, utility connections,
*
remote instrumentation, etc. Turnkey costs are often
significantly higher than installed costs for historical
air pollution control applications.
Operating costs for a single case have been provided in
Table 3. Here it is shown that for a 40 megawatt unit,
the annual operating cost would be about $316,000. The
* Total installed system cost.
773
-------
largest single item is the lime cost at about 57% of
the total. The second largest is the power at about
13% of the total. The corresponding material balance is
shown in Table 2.
Figure 4 illustrates the annualized cost of operating
a RC/Bahco SC>2 scrubbing system for various flue gas
SCU concentration levels and unit sizes. The scrubbing
annualized operating costs are also compared with an
assumed cost of a fuel switching option for meeting
pending SC>2 legislation requirements.
SUMMARY
In 1970, Bahco licensed their system for the Japanese
market. A number of installations are already in
operation in Japan. A list of the systems sold to
date are provided in Table 4. The last group of
slides show the installation at Yoshinaga. This
installation consists of three units each with a gas
handling capacity of 44,000 scfm. The units are
installed on oil fired boilers and employ sodium
hydroxide scrubbing. The reaction product is used in
the pulping process.
At this time, Research-Cottrell intends to apply the
774
-------
system to fossil fuel boilers with capacities in the
range of 60,000 to 500,000 pounds of steam. It is
anticipated that once Research-Cottrell is fully geared
up to respond to the market,the delivery time will be
nine months. Discussions are presently underway with
a number of companies for application of the system to
both utility and industrial boilers.
775
-------
REFERENCES
1. Gleason, R. J., "Limestone Scrubbing Efficiency of
Sulfur Dioxide In A Wetted Film Packed Tower In Series
With A Venturi Scrubber", Paper presented at the
Second International Lime/Limestone-Wet Scrubbing
Symposium, November 10, 1971, New Orleans, Louisiana.
2. Gustavsson, C. A., "Bahco SC>2-Scrubber CTB-CTK 1.5
Pilot Plant Connected to Bahco's Central Heating Plant",
Paper presented at 1st International Lime/Limestone
Symposium, March, 1970, Pensacola, Florida.
3. Gustavsson, C. A., "Bahco SO2-Scrubber, Commercial
Installation at Sfldersjukhuset, a Swedish Hospital
in Stockholm", Paper presented at 1st International
Lime/Limestone Symposium, March, 1970, Pensacola,
Florida.
776
-------
(M IM
O * IM
CO H
M
0) V4 r+t
£3ol
g
to a
H
4J
4) •
H a
8"
S** •
W Ml
D C 5o
OS H H
CO
K
a
o
M
a a
M W
w tj
H O -S
3 O ij
io
OIO
I 4
SH X-H
10 M
C -P
0 4)
HS
:Ti
S"o»
ft S4
(1)
•p
Q
at
i»«
4J
M
VO
0
o
o
*
M
in
o
M
VO
00
f)
O
H
H
00
•
in
o
H
O
M
91
O
3
o
0
H
O
C4
O
X
on
X
fM
H
M
O\
O
0
in
*
N
in
in
0
Ot
VO
00
o
o
tH
^
•
VO
O
o
o
(N
9l
O
fl
O
0
H
O
(N
O
X
a\
X
r>t
H
<•>
VO
e\
o
o
VO
H
ft
N
•
in
in
VO
<*
o
o
H
O
•
VO
o
H
O
OM
*""
O
en
-------
o
o o
o ir
«• N"\
•sr t—
o
oo
oo
LO
O
UJ T
<-> I
CSI
-------
Utilities
Power
Make-up Water
TABLE 3
Estimated Annual Operating Costs
For
40MW RC/Bahco S02 System
Quantity
Required
600 KW
40 gpm
Cost
$/Unit
0.008 $/KW hr
0.3 $/l,000 gal
Annual Cost
$/yr
42,400
6,300
(1)
Chemicals
Lime (91% CaO)
0.935 tons/hr 22.0 $/ton
(2)
180,000
Other Operating Expenses
Operators 0.5 man/shift 4.0 $/hr
Supervision (25% labor)
Maintenance (3% of capital cost)
Direct Overhead (75% of Labor & Maintenance)
Taxes and Insurance (2% of capital cost)
17,500
4,400
20,200
31,600
13,500
Estimated Operating Costs $315,900
(1) Assumes 8,760 hours/year; does not include waste disposal
cost or gas reheat cost.
(2) Includes delivery; i.e., lime trucked approximately 50 miles.
779
-------
CD
2TH;
P3U-
oa.1-
oo:
CO
rn
o
in zc
o CD
CSJ
^^>
rc
o
OCD
«
O
<
CO
i
<_>UL
C_-
I— c/:
CD CD O CD CD CD CD=> CD
CD CD CD CD CD CD oCD CD
I— rv. CNI CD CD hn
^ S X X S K
iv, cr cr CD CD oo
ce:
c§°
CO
o
^
«a:
CQ
uu
O
CD
CJ
CD
O
CD
CD oo
CO 00 >- O
oz 0 CO h-
OO •—i >—• •—.
CO
or
LU
,§
CO
oo
CO 00
CD
rr:
CD
01
00
<_) 1-H
rn
-
-------
781
-------
FIGURE 2
-scrubber
782
-------
FIGURE 3
0&
i5
PLANT SIZE, SCFM
783
-------
CM
I ^v.
I CD
\i
a:
I ^
I O
I <
i *^
00
784
-------
THE BISCHOFF PROCESS—INITIAL RESULTS FROM
A FULL-SIZE EXPERIMENTAL PLANT
Dr.-Ing. Gerhard Hausberg
Gottfried Bischoff
Essen, Germany
Presented by:
Ulrich Kleeberg
Prepared for
Second International Lime/Limestone
Wet Scrubbing Symposium
New Orleans, Louisiana
November 8-12, 1971
785
-------
The Bischoff Process - Initial results
from a full-size experimental plant
by Dr.-Ing. Gerhard Hausberg
For reasons well known it is advisable
to use experimental plants of the greatest
possible size to gain knowledge on the
reactions and transitional processes taking
place in heterogeneous mass flows.
The flow rates through such experimental
units should be approximately comparable
to the partial streams passing through
each of the several units that are connec-
ted in parallel in high-capacity commercial
plants. Taking, for example, six scrubbers
connected in parallel for an 000-MW boiler,
than each group will have to handle a flue
gas rate of approximately 500 000 m5NTP/h.
Scrubbers of the same design for flow rates
of this magnitude and even of up to 10 m /
NTP/h have been in use for several years in
the iron and steel industries.
The tests we carried out in 1968 and 1969
in a pilot plant designed for a maximum
throughput rate of 5 000 m NTP/h were from
the outset regarded as preliminary tests
only. There was agreement that whatever
results would be obtained from this pilot
plant they certainly could not be used
without further experience as basis for
engineering full-size commercial plants.
786
-------
An initial report on the resuls obtained
with the pilot plant was given in October
1969.
Although the informative value and use-
fulness of the data derived from small
scale plants are often higher than what
one would expect in view of the equipolent
influence of several laws of similarity,
the only thing it was hoped could be
gained from the pilot plant was an ade-
quate justification for planning and
building a much larger experimental unit.
Since the results obtained from the small
version were indeed encouraging, the
decision was made, early in 1970, to go
ahead with a full-size experimental plant
for a flue gas throughput of 140 000 m^NTP/h.
This plant, which was installed in the
Steag power station of Kellermann at Lunen,
West Germany, went into operation at the
beginning of February 1971.
The large scrubber unit is, basically, of
the same design as that of the small version,
which had been dismantled in the meantime
and which is now being used in a steel
plant. Owing to the fact that the full-size
unit is located immediately after the new
350-MW boiler of the Kellermann power
generating unit, minor changes in the
dimensional proportions had to be made.
The throughput ratio between the two
units is approximately 30:1.
Pig. 1 gives a size comparison between
the two experimental units.
787
-------
The scrubber of the large unit is again
a two-stage design with centrally dispo-
sed spiral jet nozzles, some of which are
of the double-sided type. The raw gas
pipe has initially been connected to the
boiler at a point upstream of the electro-
static precipitator and enters the primary
scrubber stage at the top. The flue gases
flow at first only through the primary
scrubbing and cooling zone which has no
internals other than the spiral jet
nozzles.
The second stage is accommodated in the
same circular cylindrical housing. It
contains an adjustable annular gap
washer. The two stages are separated
from one another by a sloping plate.
The scrubbing fluid for the second
stage is injected through a spiral jet
nozzle also arranged in the centre of
the housing.
Downstream of the annular gap washer,
the gas flows through a water separator
of the axial-flow rotor type whereby
the residual water present in the form
of droplets is removed by centrifugal
effect.
The induced draft fan arranged at the
downstream end of the system is provided
with variable-pitch guide vanes. In
combination with the adjustable annular
gap washer it permits the system to be
operated at a constant pressure differen-
tial while the flue gas rate varies. The
clean gas leaving the scrubber passes
through the clean gas flues to the power
station stack. V/ith the system operating
788
-------
at full capacity, approximately 10 per
cent of the flue gas volume discharged
from the 350-MW "boiler passes through
the scrubber.
With the present system layout, all of
the water required for the annular gap
stage is taken from the overflow of the
settling tank. This amount is withdrawn
from the collecting tray of the secon-
dary washer and passed to the two bottom
nozzles of the primary scrubber.
Of the remaining amount of water needed
for the primary stage, only that quantity
required for the annular gap stage is
pumped to the settling tank. The larger
portion is taken from the collecting
tray of the first stage and directly
injected again into the primary scrubber.
In the tests still to be run, the layout
of the' water cycles will be varied in
several ways, although at the present
stage no details can be given. An
elevated circular cylindrical vessel
with central inlet and conical bottom
section is used for clarifying and
settling.
The next illustration (2) shows the
general layout of the system with its
water circuits.
The sludge accumulating in the conical
discharge section of the settling tank
is pumped to the dump area by a slurry
pump via a plastic pipe over a distance
789
-------
of abt. 700 m.
Before the tests were started, three
wells were drilled in the area where
the sludge was to be dumped and the
ground water was analyzed in its ori-
ginal condition. After the tests were
begun and sludge dumped in this area,
the ground water was checked at regular
intervals.
Like in the small-scale plant, dry
pulverized white lime is still being
used as alkaline additive in the full-
size unit. The material is introduced
through a vertical pipe arranged cen-
trally in the raw gas inlet. A propor-
tional belt weigher located at the lime
bin serves to adjust the amount of lime
fed into the system.
After the full-size plant was put into
operation, the lime-to-SOp ratio was
at first adjusted as far as possible
to the values which experience gained
with the pilot plant had shown to result
in an SOp removal efficiency of around
80 per cent. The flue gas rate was
varied in the range from 90 000 to
133 00 m%TP/h and the loss of head
accross the annular gap washer held
between 200 and 250 mm WG. So far,
the system could not be operated at
the maximum flue gas flow of 140 000 m^
NTP/h as this would have led to a loss
of head exceeding the maximum allowed
value of 250 mm WG. Pull-load tests will
be run, however, after making a minor
790
-------
modification on the annular gap washer.
If we enter the values obtained from
the full size plant in the SOp removal
diagram plotted for the pilot plant,
then, as we see from the next slide (3)»
the values already known are largely
confirmed by the full-size plant.
In this connection, it should be noted
however, that the raw gas from the
350-MW boiler contained between 3,0
and 5,0 g/m^NTP of S0p, that is, about
•2 ^
1,5 g/m NTP or 0,0525 per cent by volume
more than the flue gas of the small
boiler with liquid ash discharge.
It is also apparent from Pig. 3 that the
flue gas volume, and hence, the period
of detention do not exercise any distinct
influence within the range of the through-
put rates used. This may be due to the
fact that the effects of several process
parameters partly cancel each other. Por
example, it is not possible to determine
what influence the S02 content of the
raw gas has on the removal efficiency,
as the concentration of SOp in the raw gas
from the 350-MW boiler was constantly
fluctuating within the above-mentioned
range.
Measurements made to check the dust
removal efficiency of the full-scale
plant showed that dust concentration
in the clean gas was 50 mg/nrNTI when
the system throughput was 133 000 m^NTP/h
and the dust load of the raw gas was
15 g/nrNTP. The removal efficiency,
791
-------
therefore, on average was 99,7 per cent.
In the small-scale unit, dust concentration
in the clean gas ranged between 10 and
•Z
15 mg/m NTP at a mean entry concentration
of abt. 10 g/m5NTP. The light coloured,
yelloish grey fly ash of the 350-MW boiler
appears to be much finer than the much
darker ash from the small boiler with
liquid ash discharge. So far, no dust
analyses have been made.
The final dust concentration of 50 mg/nrNTP
obtained v/ith the full-size plant is far
below the maximum allowable value of
150 mg/m5NTP.
With the knowledge now available on S0?
and dust removal and its dependence on
volume flow, it would be possible to
build units for throughput rates of,
. C -2
say, 10 m NTP/h and over, based on the
tested design and layout.
While the tests so far have proved in
the main that the results obtained with
the pilot plant v/ith respect to separa-
tion efficiency also apply to the full-
size version, the purpose of further
tests v/ill be to try out a series of
designs and process variants some of
which are aimed at an optimization of
the dimensions of the primary scrubbing
stage. Operational reliability will be
the most important aspect.
Reproducibility of guaranteed values in
continuous operation at a minimum of
792
-------
maintenance costs is the standard by
which operational reliability must be
gauged. Therefore, one thing for which
the full-size plant was constantly
being observed and checked was the
formation of deposits and incrustations.
It was particularly in the conical section
and in the upper third of the cylindrical
shell of the scrubber that substantial
incrustations and asymmetrical deposits,
due to the given inflow conditions,
occurred.
After the design was changed to give
better flow conditions and additional
spray nozzles for intensive shell
washing, as shown in Fig. (4), were
installed, a considerable improvement
was obtained.
By varying the rate of wash water flow
at constant gas throughput, it was
proved that the rate of growth of the
deposits on the shell surface was
essentially dependent on the thickness
and velocity of the film running down
the shell surface.
A temporary disturbance in the operation
of the full-size unit occurred as a result
of insufficient clarification of the
return water. In designing the settling
tank, the dimensions had intentionally
been held rather small because plans were
for this part of the system to be extended
only at a later stage. It thus could happen
793
-------
that the water recycled to the scrubber
was for a while entering the annular gap
washer as a slurry containing 200 g/1 of
solids. While this appeared to have no
influence on SOp and dust removal, and
the nozzles were not clogged, heavy
erosion must be expected to occur in
the long run on such items as pumps,
flow restriotors and nozzles.
After the inflow conditions and the
pump-out conditions at the discharge
cone of the elevated vessel were im-
proved, the solids concentration in
the overflow dropped to between 10
and 20 g/1. It is intended to test
still other settling tanks of modified
design and with larger dimensions. In
this connection, it is noted that the
mean settling velocity at concentrations
up to 50 g/1 was about 4 m/h and that
it slowed to 1 m/h at a solids concen-
tration of 100 g/1.
The slurry forming in the conical thicke-
ning section of the settling tank con-
tains an average concentration of solids
of abt. 300 g/1. Tests carried out by
Westfalia Separator of Oelde, \7est
Germany, have shown that this sludge
water can be largely clarified with
the aid of decanters. Residual water
in the solids discharge is as low as
abt. 30 per cent by weight. In spite
of this low water content, the mixture
still flows and can be pumped, and it
may greatly reduce handling and dumping
costs.
794
-------
But since it was not our intention
to content ourselves with dumping the
sludge, we requested three institutions
and specialist firms to cooperate in
analyzing the sludge and studying the
possibilities of processing it for
commercial use.
Initial results of these studies
indicate that it will be possible, for
example, to use the sludge consisting
of fly ash and reaction products as an
additive in the production of lime sand
bricks. Test bodies made by Rheinisch-
Y/estfalische Kalkwerke, Dornap, using
a 20-per cent addition of sludge were
found to have a maximum compressive
p
strength of 300 kgf/cm as compared
P
with 240 kgf/cm for the standard lime
sand brick.
In this connection it should be noted
that gypsum in the form of dihydrate
was found at the time in the sludge of
the pilot plant in addition to the
hemihydrate. In contrast to this ,
calcium sulphate in the form of
anhydrite was in some instances found
to prevail in the full-size plant.
One reason for this may be temperature
zone differences between the two systems.
For instance, one point where dihydrate
is converted into anhydrite is around
40 deg C. Moreover, as is known, the
conversion points may shift owing to
the presence of attendant materials,
795
-------
such as certain constituents of the
fly ash. Part of the heat generated
during the slaking of CaO will surely
be dissipated more quickly in the small
unit than in the larger plant.
Another use of the sludge seems possible
if the gypsum produced in the process
is again obtained in the form of dihy-
drate. If a unit for dry separation of
fly ash is arranged to precede the
scrubber, then it will be possible to
produce high-quality gypsum board by
the well known Giulini process.
After completion of the first set of
test runs, the full-size plant will,
therefore, be connected to follow the
electrostatic precipitator so that
measurements can be made for separate
removal of dust and SOp.
The fact should be emphasized here that
it is not possible at the present stage
to make a final statement on the sludge
produced, because only a few samples
were tested so far.
After completion of the tests in which
pulverized white lime is injected into
the raw gas inlet, the plant will be
operated on milk of lime. It can be
expected that the use of this additive
in the form of slaked lime or calcium
hydroxide will result in a further
improvement of S0? removal. This was
seen to be the case with the small
pilot plant in which pulverized calcium
hydroxide and milk of lime were used
796
-------
at times. In those tests a removal effi-
ciency of 88,8 per cent was obtained
at a concentration in the raw gas of
only 2,25 g/m NTP or approximately
0,08 per cent by volume.
The use of milk of lime will offer
still another advantage. So far, all
of the additive was introduced, as
already mentioned, upstream of the
primary scrubbing stage. The annular
gap washer, operating in a zone of
lesser concentration, received no
additional lime. A check on the pH
values of the wash, fluid passing
through the annular gap indicated
11,0 for the inflow and 5,5 for the
outflow of that stage.
This means that the annular gap
washer was working partly in the
acid range.
When we change to milk of lime, we
shall feed the additive to both
stages in varying proportions.
The first and most important result
of the tests run so far with the
full-size plant is the fact that the
information obtained with respect to
SOp and dust removal is also applicable
to the design and layout of large
commercial units.
The tests were again prepared and
carried out in close cooperation
with Steinkohlen-Elektrizitatsgesell-
schaft-AG, Essen, which has proved
so successful in the past. / „
797 ..
-------
Isvirvvoi ef$0i east Cj'i frc.Ti flue 'c^'-.s
ELpeiuiH-nKi! ur,tti tar Hi COia'i/tryii art £8X)i,i',:fn/li
.1
I I I j I I 1 -11,-" I '"I*
f
Schematic Diagram of Pfant for ffts
Removal of Dust and S02 from Flat Qasas
70/IB
. i
TO
SO
id
30
30-
0
t
si
X *«
1.0
y
./
:Lpfc.
iff with
»r wilfi
i
2L
tfr/ otfi
quid ash
*
m *
'iseharg
titictiarg
Of
•
t
o
—
~ uiL
tor
l,fj
* so.c
-T~'~
crtrfcn
-
e< .>;o-«
Cm'HrP/f, 't
-—
Wbctlw
"&
if 2 M^
G£ ; £3 1 SO/- removal from flue go$e$
\LJ? I
—
with
71/11
SOi and dust removal from flue gases
full-size ttpefimentol plant-raw gas Intel
I •
.3
798
-------
AVAILABILITY OF LIMESTONES AND DOLOMITES
J.J. 0'Donne11
A.G. Sliger
Research and Engineering Development
The M.W. Kellogg Company
Piscataway, New Jersey
Presented at the Second International
Lime/Limestone Wet Scrubbing Symposium,
New Orleans, Louisiana, November 8-12, 1971
799
-------
INTRODUCTION
Power in the United States is produced largely from
plants which burn coal or oil as the primary fuel. The
use of limestone or dolomite to control sulfur emissions
from these plants is contingent upon several factors.
Among these, proximity of adequate carbonate rock deposits
to fossil fuel-fired power plants, and the relationship of
carbonate rock production to possible demand are important
considerations in establishing the relative merits of any
carbonate rock-based process. This study, performed under
contract with the Office of Air Programs, had as its objec-
tive the determination of the availability and costs of
limestone and similar materials throughout the contiguous
United States, thus providing a basis for determining the
feasibility and economics of limestone-based SOK removal
processes for any particular power plant site.
800
-------
POTENTIAL LIMESTONE DEMAND BY POWER PLANTS
Figure 1 shows the location of the major (^200 MW)
power plants in the United States which burn either coal
or oil as the primary fuel. Included are a few plants
which, although not yet constructed, are being designed
exclusively for either coal or oil and are scheduled to
be on stream by 1975. Of the 275 plants shown, 84% are
coal-fired. The oil-fired plants are all located along
the eastern coast, with most of these in the northeast.
About 90% of all power plants shown are located east of
the Mississippi River, with locally high concentrations
in the northeastern quarter of the country, particularly
in many of the major metropolitan areas.
Based on 1969 fuel consumption statistics, an estim-
ated 20 million tons of sulfur oxides were emitted by
power plants in the United States. More than 40 million
tons of limestone would have been required to remove these
oxides from the stack gases. This potential limestone
demand has been broken down by region, with the regions
801
-------
being defined as follows
Region
States Included
New England
Middle Atlantic
East North Central
West North Central
South Atlantic
East South Central
West South Central
Mountain
Pacific
Connecticut, Maine, Massa-
chusetts, New Hampshire,
Rhode Island, Vermont
New Jersey, New York,
Pennsylvania
Illinois, Indiana, Michigan,
Ohio, Wisconsin
Iowa, Kansas, Minnesota,
Missouri, Nebraska,
North Dakota, South Dakota
Delaware, Florida, Georgia,
Maryland (incl. Washington,
D.C.)f North Carolina,
South Carolina, Virginia,
West Virginia
Alabama, Kentucky, Mississippi,
Tennessee
Arkansas, Louisiana, Oklahoma,
Texas
Arizona, Colorado, Idaho,
Montana, Nevada, New Mexico,
Utah, Wyoming
California, Oregon, Washington
Table 1 itemizes potential demand by region. The East
North Central region far outranks any other region in poten-
tial limestone requirements with over one-third of the total.
With the exception of New England, all regions in the eastern
half of the country had potentially large demands for lime-
stone.
802
-------
CARBONATE ROCK RESOURCES
Deposits of carbonate rocks, including limestone,
dolomite, shell, marble, and marl, occur in some form in
every state. Total reserves have never been estimated,
but are known to be enormous.
Figure 2 shows the distribution of surface carbonate
rocks in the United States and includes limestone, dolo-
mite, and marble. The map shows that surface deposits of
carbonate rocks occur throughout the nation, but are par-
ticularly in evidence in the eastern half of the country.
A band of deposits beginning in Vermont extends southward
along the Appalachian Mountains into central Alabama.
Extensive deposits are found in the states surrounding the
Great Lakes, reaching southward into northern Alabama.
Large areas of Minnesota, Iowa, and Missouri are covered
with carbonate rocks and broad outcrops occur in Kansas,
Oklahoma, Arkansas, and Texas.
Particularly in the central lowlands, carbonate rock
deposits frequently occur as thick, horizontal formations
covering large areas. In general, the deposits found in
western states are different. They commonly occur as
steeply dipping or vertical beds of small areal extent.
However, notable exceptions to this are found, particularly
in Colorado, Arizona, and New Mexico where large outcrops
occur.
803
-------
Limestone occurrences, including chalk but excluding
dolomite, are shown in Figure 3. The map is similar to
Figure 2 and shows that although limestone is found through-
out the country, the more numerous and extensive deposits
occur in the eastern half of the nation. In the western
states, the deposits tend to be discontinuous and relatively
small in areal extent.
The formations shown in Figure 3 include limestones of
different degrees of purity. Estimates have been made that
only about 2% of the known reserves of commercially usable
limestone is chemical grade (>95% carbonate content), and
that the bulk of these reserves will be exhausted in 40-50
years. Much of this high purity limestone occurs in the
area extending from the Great Lakes southward to Alabama.
Chalk deposits are shown in several of the central
states and in a curving belt through Alabama and Mississippi.
In general they are not high purity limestones, but locally
they may contain over 95% calcium carbonate.
Figure 4 shows the location of high grade (>25% magnes-
ium carbonate) dolomite quarries. Although originally drawn
in 1941, it is a useful guide to the important occurrences
of dolomite. As with limestone, the largest deposits are
located in the eastern half of the country. Two major areas
are noted. First, a belt of dolomite extends from Vermont
to central Alabama, along the Appalachian Mountains. This
coincides quite well with a similar band of limestone pre-
viously noted. Second, large formations of dolomite occur
in the states encircling the Great Lakes. These deposits
coincide with or adjoin large limestone deposits in the
region.
804
-------
The most significant deposits of marble are found
along virtually the entire length of the Appalachian Moun-
tains in the east, and as scattered occurrences in the Rocky
Mountains in the west. Although eastern marbles are pre-
dominantly calcitic (high calcium) , dolornitic types also
occur. Both types are found in the west.
Shell limestone occurs primarily in Gulf Coastal
waters, but it also is found in bay waters along both the
east and west coasts. It is usually a very pure type of
calcium carbonate.
Marl deposits exist in several areas, notably around
the Great Lakes, and along the southeastern coastal plain.
This soft, relatively impure form of calcium carbonate
varies considerably in character, that of the Great Lakes
area being a precipitated calcium carbonate, while that of
the coastal plain is an impure shell deposit. Limited
occurrences in other regions are generally impure chalks or
soft limestones.
PRODUCTION
Crushed carbonate rock production in the United States
in 1969 was distributed, by type, as indicated in Table 2.
As shown, the total national production amounted to 652
million tons, of which over 85% was limestone. Dolomite
ranked second with more than 10%. Marl and marble production
were very low, amounting to less than 5 million tons combined,
or 0.7% of the total.
805
-------
No limestone production was reported in four states,
viz., Louisiana, Delaware, New Hampshire, and North Dakota.
In fact, the latter three states did not produce any type
of carbonate rock. Dolomite was produced in twenty-four
states, chiefly in the northeastern quarter of the country.
Production of limestone and dolomite, by region, is
shown in Table 3. Within most regions, production rates
of individual states varied from near zero to tens of
millions of tons. The New England and Mountain regions,
however, had fairly uniform (but low) outputs. The East
North Central region was also exceptional, in that all
states reported large quantities of limestone and dolomite,
ranging from 16 million to 55 million tons. Nationwide,
Pennsylvania and Illinois were the leading producers of
limestone and dolomite, respectively.
Shell was dredged from bay waters along all three
coasts. However, 83% came from Texas and Louisiana, with
the latter being the leading producer. Small quantities
of marl were produced in Indiana, Michigan, Minnesota,
Mississippi, Nevada, South Carolina, Texas, and Virginia.
Eighteen states, principally in eastern and western moun-
tainous regions quarried marble, with Alabama recording the
highest production at 632,000 tons.
Over 4,700 quarries, producing 861 million tons of
crushed stone of all types, were in operation in the United
States in 1969. Since production of crushed carbonate
rocks totalled 652 million tons, it can be assumed that,
806
-------
roughly, over 3,500 of these quarries produced limestone,
dolomite, and related stones. More than one-third of all
quarries had annual production rates of less than 25,000
tons. The large operations (over 900,000 tons/year) pro-
duced one-third of the total crushed stone, although they
represented less than 4% of the total number of quarries.
END USE
Carbonate rocks are unique among the different types
used in this country. Not only do they find use in appli-
cations where their physical properties are important, but
also in markets which utilize them for their chemical prop-
erties and composition. Almost two-thirds of the total
production of carbonate rocks were used for various con-
struction purposes in 1969. Most of this stone was used
as an aggregate material in road construction. The second
largest use was in cement manufacture, which consumed more
than 105 million tons. About 70 million tons of high grade
limestone and dolomite were used in lime manufacture and
other applications requiring a high purity material.
SELLING PRICE
The average unit value, or net selling price at the
quarry, for all crushed carbonate stones produced in the
United States in 1969 was $1.49/ton and varied by type of
stone as shown in Table 4. The average values ranged from
a low of $1.0I/ton for marl to a high of $9.69/ton for
807
-------
marble. The high unit value of marble reflects its prim-
ary use as a decorative material.
Unit value also varies with end use. Stone used for
construction purposes averaged $1.44/ton while stone used
in applications requiring a high purity material had an
average value of $1.69/ton. Average prices ranged from
$0.69/ton for fill to $6.00/ton for exposed aggregate
(decorative stone). Stone for most applications, however,
averaged under $2.00/ton. The variation in price depends
not only on supply and demand, but also on the chemical
and/or physical properties required for the particular
application.
Average unit values of limestone and dolomite, by
state, are shown in Table 5. Within most states, average
prices were $1.00-2.00/ton, although spot prices ranged
from- $0.12-25.00/ton. New Jersey and several states in
both the New England and Mountain regions reported average
values greater than $2.00/ton. Rhode Island showed the
highest average value: $7.57/ton for limestone. Prices
shown for the Pacific region are unusual in that those
for limestone are comparatively low while dolomite prices
are quite high. In all cases where average unit values
were high, production of the particular stone was limited.
With the exception of Virginia, which reported $3.92/
ton, average prices of shell were $1.00-2.00/ton. Unit
values for marl were all below $1.15/ton. Marble varied
widely in price.
808
-------
TRANSPORTATION
Trucks dominated in the transportation of carbonate
rocks from quarry to consumer, accounting for almost three-
fourths of all stone. Rail and waterway hauls, amounting
to one-fifth of the stone shipments, were about equally
divided.
Trucks generally are used for shorter hauls of under
50-100 miles, while rail is employed for longer distances.
Where conditions permit, shipment of stone by barge or boat
is preferred, since this is usually the cheapest method of
transportation.
So many factors influence transportation rates and
costs that it becomes very difficult to establish average
rates, even within a single area. Most freight rates for
crushed stone in the United States, however, fall within
the ranges shown in Figures 5-7. The lower curves corres-
pond to rates which could be obtained under favorable con-
ditions; i.e., high volume movement, periodic shipments in
lot-sized quantities, intrastate shipment or commodity
rates, etc. The upper curves correspond to unfavorable
conditions.
DELIVERED PRICE OF LIMESTONE
The delivered price of limestone to a power plant is
the sum of the price of the stone at the quarry plus the
transportation charges. To see what this actually would
809
-------
be, the delivered price of limestone to 37 selected power
plants was determined based on the assumption that a high
calcium limestone would be required. The results of this
investigation are shown in Table 6.
Most of the plants are located in the eastern half of
the United States where the major coal- and oil-fired
power capacity is found. Prices range from $1.95-13.20/ton.
Half of the plants could be supplied at under $4.00/ton,
while all but 3 could obtain limestone at under $6.00/ton.
The latter 3 plants are located in the west in areas where
base prices are higher or limestone deposits are remote.
For several eastern seaboard plants, particularly in New
England, the availability of low-cost high calcium lime-
stone is contingent upon the acceptability of an imported
stone. Domestic sources are either inadequate or too dis-
tant to provide a low-cost material.
PROJECTED COSTS
Carbonate rocks historically have been stable, low-
priced commodities. Based on average unit values for the
years 1960-1969, projected average base prices for 1975 are
as follows:
1969 1975
Limestone and Dolomite $1.45/ton $1.67-1.82/ton
Marl $1.01/ton $1.28-1.48/ton
The average value of shell has dropped considerably since
1960. It is unlikely that it will continue to decrease
810
-------
through 1975. More probably, it will parallel limestone
and dolomite but not exceed them in value. Average unit
values for crushed marble are highly variable, reflecting
the sensitivity of price to market conditions.
Transportation rates during the next few years should
rise about 6%/year, on the average. Estimates by type of
transportation are as follows:
Truck 4- 6%/year
Rail 6- 8%/year
Water 5-10%/year
In general, rate increases should follow wage increases
granted to labor. The estimates, offered by stone produ-
cers, are predicated on the present state of the economy
and a continuance of the present inflation rate. Any
changes will, of course, affect these estimates.
SUPPLY/DEMAND RELATIONSHIP OF CARBONATE ROCKS
FOR POLLUTION CONTROL
Proximity of Carbonate Rock Deposits to Power Plants
Comparison of Figures 1-4 indicates that the major
deposits of carbonate rocks largely coincide in location
with the power plants. This is particularly true in the
East North Central region where huge reserves of stone
occur. Both high calcium limestone and high grade dolo-
mite abound, and many deposits are found near the major
power generation centers throughout the region.
811
-------
The New England region is not as fortunate. While
the power plants are located primarily in coastal areas,
the rock deposits occur in the mountainous western sec-
tions. Most of the deposits are highly crystalline stone
or marble, and many are dolomitic. The suitability of
these materials would have to be determined before in-
cluding them as a possible source.
Availability of stone should not be a problem for
power plants in the Middle Atlantic region. All types
of stone occur and nearby deposits can be found. Plants
in western and eastern Pennsylvania, particularly, are
fortunate in that large reserves of high grade stones
are present.
In the South Atlantic region, most plants are located
near an adequate source of stone, particularly if the cry-
stalline limestones and dolomites of the mountainous areas
<«
prove suitable. Several coastal plants could use shell
or coral limestone, or marl. However, for some inland
power plants in, for example, North Carolina, no nearby
deposits exist.
Large quantites of high calcium limestone occur
throughout the East South Central region and power plants
should experience no difficulty in obtaining adequate
supplies. Many plants located in Alabama and Tennessee
also could obtain dolomite quite easily.
The less abundant and more widely scattered carbon-
ate resources of the western states are of minor impor-
tance, since there are few coal-fired power plants in
812
-------
this area. With the exception of the two power plants
located in North Dakota, which are far removed from any
commercially important carbonate deposits, the few plants
that do exist are fairly near reserves of high calcium
stone.
Potential Demand Relative to Production
As previously mentioned, the potential demand for
limestone by coal- and oil-fired power plants in 1969 ex-
ceeded 40 million tons. This represents only 7.3% of the
national limestone production of 559 million tons. Dolo-
mite production, on the other hand, was roughly 50%
greater than the potential demand. Shell production was
only one-half of the demand. The quantities of calcar-
eous marl and crushed marble quarried were relatively in-
significant. Obviously, limestone is the only stone pro-
duced in sufficient quantities to warrant nationwide con-
sideration as an agent for SOX removal. Dolomite and
shell, however, are quarried in large enough amounts to
make them important materials in some regions. Marl and
crushed marble may be useful in certain localities where
the other rocks do not exist, but their limited occurr-
ence and production do not permit wide-scale use of these
materials.
Table 7 shows the relationship of limestone and dolo-
mite production to potential demand, by region and state.
The third column of the table is the ratio of total lime-
stone and dolomite production to potential demand, for 1969
813
-------
With the exception of New England, the eastern re-
gions of the country all have large relative supplies of
limestone and dolomite. There are, though, some excep-
tions to this among the individual states. New Jersey
and Delaware have a potential need which exceeds their
production. Adjacent states, however, are large producers
of stone and could provide the necessary tonnages.
Mississippi and both of the Carolinas have low relative
supplies of stone. If they could not otherwise be supplied,
marl deposits, which occur extensively in all three
states, could be used. Georgia and West Virginia, with
comparatively low relative supplies, could easily obtain
needed stone from surrounding states. It is interesting
to note that the region with the highest limestone demand,
i.e., the East North Central region, also has the highest
limestone and dolomite production.
The New England region faces a shortage of limestone
and dolomite, with two states producing less stone than
potentially required. The marble resources of the region
could improve the situation somewhat, but power plants
would have to rely on shipments of stone from nearby states
such as New York, or, perhaps, on imports.
Most states in other regions of the country have ample
production. North Dakota, with no production, is an out-
standing exception to this, however. A few other states
have low relative supplies of stone, but either the demand
is quite small, or the production could easily be expanded
to meet the need.
814
-------
CONCLUSIONS
€> Enormous deposits of carbonate rocks occur in the
United States, and reserves are more than adequate for the
foreseeable future. Availability of high purity stone may
become a problem several decades hence, but, with the prob-
ability that the required quality will depend on other pro-
cess and economic factors, no shortage of suitable stone
is foreseen.
o The major deposits of carbonate rocks occur in the
eastern half of the United States, where the vast majority
of fossil fuel-fired power plants are located. Large re-
serves in these eastern areas provide a nearby source of
stone for most power plants.
© Relative to the potential demand for carbonate rocks
by power plants, production of these materials is quite
large in most states. However, current production is in-
adequate to supply the potential needs of power plants in
several Atlantic coastal regions, notably New England.
• Limestone is the only type of carbonate rock which is
produced in large enough quantities to merit consideration
for widespread application in the removal of SOX from
stack gases. In many areas, however, ample amounts of
other carbonate rocks are produced, particularly dolomite.
0 Most of the power plants in the eastern half of the
United States could be supplied with high calcium lime-
stone at less than $6.00/ton. Many could obtain stone at
815
-------
less than $4.00/ton. Costs for power plants located in
western states generally would be higher, owing to the
lack of suitable, nearby deposits.
• Based on projections of material cost and transporta-
tion charges to 1975, the delivered price of limestone to
most power plants should not increase by more than $1.00-
2.00/ton.
816
-------
TABLE 1
POTENTIAL LIMESTONE DKMAND BY PO'./EK PLANTS IN THK UNITED STATES
Region
New England
Kiddle Atlantic
East North Central
West North Central
South Atlantic
East South Central
rfest South Central
x-
Kountair
Pacific
Total
Limestone (H Tons)
1,369
6,583
14,307
2,738
8,471
5.6A5
5
1,351
266
40,735
% of Total
3.4
16.2
35.1
6.7
20.8
13.9
-
3.3
0.6
100.0
817
-------
TABLE 3
PRODUCTION OF CRUSHED AND BROKEN LIMESTONE AND DOLOMITE
IN THE UNITED STATES IN 1969, BY REGION
Region Production (MM Tons)
New England 2.4
Middle Atlantic 90.9
East North Central 185.6
West North Central 92.5
South Atlantic 87.5
East South Central 81.3
West South Central 58.3
Mountain 10.7
Pacific 18.8
Total 628.0
818
-------
TABLE 2
PRODUCTION OF CKJ-HJi!) AND BROKEN C.-JIBOKATE STONES
IN THE U::iTKi) 3TATr;M 1H 3V69, BY TYPE
Prpductipn (M Tons)
Liiusstone 558»793
Do lorn ic 63,330
Shell 19,731
Calcareous 1'arl 2,490
819
Total, All Types 651,665
-------
TABLE 4
UNIT VALUE OF CRUSHED AND BROKEN CARBONATE STONES
IN THE UNITED STATES IN 1969, BY TYPE
Ty_£e Unit Value ($/Ton)
Limestone 1.45
Dolomite 1.55
Shell 1.42
Calcareous Marl 1.01
Marble 9.69
Average, All Types 1.49
820
-------
TABLE 5
UNIT VALUE OF CRUSHED AND BROKEN LIMESTONE AND DOLOMITE
IN THE UNITED STATES IN 1969, BY REGION AND STATE
Average Unit Value ($/Ton)
Region and State
New England
Connecticut
Maine
Massachusetts
New Hampshire
Rhode Island
Vermont
Middle Atlantic
New Jersey
New York
Pennsylvania
East North Central
Illinois
Indiana
Michigan
Ohio
Wisconsin
West North Central
Iowa
Kansas
Minnesota
Missouri
Nebraska
North Dakota
South Dakota
South Atlantic
Delaware
Florida
Georgia
Maryland
North Carolina
South Carolina
Virginia
West Virginia
Limestone
NR(I)
1.32
4.14
7.57
1.46
2.49
1.56
1.46
1.44
1.32
1.02
1.54
1.17
1.49
1.40
1.32
1.39
1.87
-
1.22
1.31
1.50
1.57
1.62
1.51
1.52
1.62
Dolomite
4.20
- (2)
5.24
"
1.53
1.97
1.73
1.45
1.28
1.45
1.48
1.21
1.72
-
1.38
1.13
-
-
-
-
-
-
1.37
1.62
(1)"NR" indicates that value was not reported
(2) "-" indicates no production
821
-------
TABLE__5 (Cont'd)
UNIT VALUE OF CRUSHED AND BROKEN LIMESTONE AND DOLOMITE
IN THE UNITED STATES IN 1969, BY REGION AND STATE
Region and Stat_e_
Easj: Sgu,th Contra. I
Alabama
Kentucky
Mississippi
Tennessee
West
___
Arkansas
Louisiana
Oklahoma
Texas
Average Unit Value ($/Toii)
Limestone Dolomite
1.17
46
00
33
1.36
1.30
1.35
1.60
NR
1.17
Arizona
Colorado
Idaho
Montana
Nevada
New Mexico
Utah
Wyoming
Pacifi c_
California
Oregon
Washington
64
04
10
24
64
1.51
2.25
2.11
1.07
1.00
1.28
2.83
1.77
NR
2.66
2.91
6.00
Total Un.i ted States
1.45
1.55
(1) "NR" indicahes that value v/as not reported
(2) "-•" indicates no production
822
-------
TABLE 6
DELIVKKED PRICE OF HIGH-CALCIUM LTMESTC"i; TO SKLKCT-MJ PO'-.TiR PLaiTS
Power Plant
Benning
Gorgas
Cherokee
J. McDonough
Devon
Fisk
R.S. Wallace
V.'ill County
D.H. Mitchell
Wabash
Fiivcrside
Lav/rence
Cane Run
Elmer Smith
Riverside
Edgar
L Street
Delray
High Bridge
Hawthorn
Sioux v
Essex
Port Jefferson
Waterside
Allen
Lcland Olds
Tidcl
Mi and. Fort
Acme
Horseshoe Lake
Elrania
Schuylkill
Wateree
Bull Rur.
Cabin Creek
Karr^ner
LakeHide
Location
Delive re d P ri ce
Washington, D.C.
Gorgas, Alabama
Denver, Colorado
Cobb County,' Georgia
Killford, Connecticut
Chicago, Illinois
East Peorio, Illinois
Lockport, Illinois
Gary, Indiana
Terre Haute, Indiana
Iov;ana, Iowa
Lav/rence, Kansas
Louisville, Kentucky
Owensboro> Kentucky
Baltimore, Maryland
N. Weyr.outh, Massachusetts
Boston, Massachusetts
Detroit, Michigan
St. Paul, Minnesota
Kansas City, Missouri
'vest Alton, Missouri
Newark, New Jersey
Port Jefferson, New York
New York, K^w York
Belmont, North Carolina
Stanton, Korth Dakota
Brilliant, Ohio
North Bend, Ohio
Toledo, Ohio
Horseshoe lake, Oklahoma
hlrama, Pennsylvania
Philadelphia> Pennsylvania
Rockland City, South Carolina
Oak Ridge, Tennessee
Cabin Creek, t.'est Virginia
Capti.na, ,.'e.-,t Virrinia
St. Francis, Uiscoiicin
3.23
6.36
4.50
A.50*
2.AO
3.30
3.30
2.65
2.25 (75-9A2
1.95
3.66
3.00
3.72
3.B5
A. 50*
A. 50*
2.AO
3. CO
4.60
3.10
4.50*
4.50*
A. 50*
5.39
13.20
3.80
2.45
2.45
8.00
5.55 (92; CaC03)
4.50*
3.90 (se,; cacoo)
4.24
6.00
4. CO (H0;; CaCOv)
2,60
* -- Source of stone is outside of U.S. (i3ahar:~jr,)
823
-------
TABLE 7
AVAILABILITY OF LIMESTONE AND DOLOMITE FOR POLLUTION CONTROL (1969)
Region and State
New England
Connecticut
Maine
Massachusetts
New Hampshire
Rhode Island
Vermont
Totals
Middle Atlantic
New Jersey
New York
Pennsylvania
Totals
South Atlantic
Delaware
Florida
Georgia
Maryland
North Carolina
South Carolina
Virginia
West Virginia
Totals
East South Central
Alabama
Kentucky
Mississippi
Tennessee
Totals
East North Central
Illinois
Indiana
Michigan
Ohio
Wisconsin
Totals
West North Central
Iowa
Kansas
Minnesota
Missouri
Nebraska
North Dakota
South Dakota
Totals
Production
(M Tons)
275
800
750
800
33,457
56,667
90,900
40,729
4,334
9,804
4,500
1,900
17,829
8,405
87,500
17,752
30,158
300
33,109
81,300
54,844
25,157
39,066
50,595
15,937
185,600
26,200
15,334
4,127
41,200
4,663
989
92,500
Potential Demand
(M Tons)
Ratio
9.19
16.9
4.23
17.8
14.4
57.8
365
6.37
37.7
42.4
0
28.3
33.8
824
-------
V 3JSAT
(6961) J05ITMOD
HP'S 3TIMOJOQ QUA. 3MOT83MIJ 10 YTIJIHAJIAVA
oi^BH (anoT M)
61M fr
0 I>
SpttfiJ I >
ep^BJ I >
SplBJ V >
8.3* 02
I3.fr 85£
00
e.ei frv
8. SI 8V
VV.S 5fr£
8.5£ V3
IV. I eV£
Se.V I5£,I
ir • C O O O £
OV< 83£>
(anoT M)
3V3,5
00£,85
053a
05S
000,1
see
OOV,OI
OOfr,VI
02£
020,1
008,81
iBiztnsD rl^uo8 ^asW
BnBiaiuoJ
3BX9T
obBioloO
OflBbI
£ nfi -it noM
SDfiVSW
ooixsM W9M
oiiioB'I
Bin'xolilBD
nop9ttO
noipnxriaBW
\ 5S8
-------
UJ
Ql
UJ
LUl
I
UJ
It
UJ
I
Q.
tr.
UJ
X
o
<
cc
o
§
or
m
-------
-------
CO
o
CL
UJ
Q
UJ
CO
UJ
o:
CO
UJ
fe
CO
O
UJ
H-
UJ
X
co
H
(O
o
a.
UJ
o
_i
a
X
O
o
h-
oa
CO
o
to
ro
O
UJ
ID
m
CO
UJ
UJ
££
m
CO
o
cc
u_
Q
UJ
CC.
Q.
UJ
or
828
-------
+
UI
oc.
CD
U_
CO
UJ
1-
2
to
o
UJ
h;
Z
13
UI
X
h-
z
to
UI
or
or
o
UI
o
0
o
UJ
o
or
CD
i
X
CD
X
U.
O
Z
0
I
^
o
o
^^
CVJ
O)
r~-
d
•z.
cc
IS
o
cc
Lu
Q
UJ
l-
E
Q_
UJ
cc
829
-------
CO
h-
2
UJ
X
CO
cc
o
LL_
CO
UJ
l/> \—
LJ cc
§ *
o o
c ?
cc
o
0.
CO
z
<
(T
O
o:
CO
UJ
h-
h-
co
Q
UJ
H
UJ
UJ
z
o
CO
o
UJ
X
CO
01
o
o
o
o
o
00
tn
U
O
in
(O
o
o
ro
O
00
- o
31IW-N01/SJ.N33 '31VH
830
-------
LJ
X
en
Co w
h-
LU <
cc K
1 §
Q:
o
Q.
CO
a:
l-
cc
UJ
h-
h-
Q
t;J
UJ
X
U-j h-
o!
LU
b
en
Q
UJ
x
en
Q:
o
m
CM
31HM-NOJL/S1N3D '31VM
831
o
m
CM
in
CM
CM
O
O
CvJ
O
m
CM
CO
LU
LL)
CO
Q
O
O
in
O
in
m
- O
O
-------
N
UJ
0
H-
LJ
CL
1C
CO
tlL
O
u.
CO
LU
or
2:
o
<
1-
c:
o
CL.
CO
z
o:
i
i—
>_
i
ce
LJ
l-
co
LJ
1-
H
CO
Q
LJ
H-
•z.
LJ
!_,
^
LJ
O
1-
co
Q
LJ
X
CO
CC
0
UJ
in
6
o
m
o
in
O
O
o
m
ro
O
o
ro
o
CJ
CO
UJ
_1
5
Uj"
h-
co
Q
o
o
CM
O
m
O
o
o
' m
3HIW-N01/S1N30 '31VH
832
-------
THE PILOT SCALE R&D AND PROTOTYPE PLANT
OF MHI LIME-GYPSUM PROCESS
Tsukumo Uno, Masurai Atsukawa, Kenzo Muramatsu
Mitsubishi Heavy Industries, Ltd.
Tokyo, Japan
Presented at
Second International Lime/Limestone Wet Scrubbing Symposium
Sponsored by
Environmental Protection Agency, U.S.A.
Office of Air Program, Division of Control System
Sheraton-Charles Hotel, New Orleans, Louisiana
8-12 November, 1971
833
-------
THE PILOT SCALE R&D AND PROTOTYPE PLANT
OF MHI LIME-GYPSUM PROCESS
ABSTRACT
Tsukumo Uno, Dr. *
Masumi Atsukawa **
Kenzo Muramatsu **
The lime-gypsum process is a method in which flue gas is scrubbed with
lime slurry to remove SO2 and gypsum is recovered as byproduct through air
oxidation of calcium sulfite formed in the absorber. This process has been
attracting particular attention as an advantageous desulfurization process in
Japan because of its remarkable features - abundant and cheap absorbent,
large demand for by-product gypsum, no waste to be thrown away, simple de-
sulfurization system, etc.
The technical feasibility is already proved by the six years operation
record of 62,500 Nm /hr plant which was constructed for a sulfuric acid plant
by our company under the license of Japan Engineering and Consulting Co.
We have conducted recently a pilot scale research and development on this
process to make the process fit to the treatment of flue gas from an oil fir-
ing boiler. On the basis of the test results, a prototype plant, treating
100,000 Nnr/hr flue gas is under construction at a power station. Here de-
scription is made shortly on the results of the pilot test and outline of the
plant.
* Manager, Hiroshima Technical Institute
** Manager, Environment Equipment Research Laboratory,
Hiroshima Technical Institute
** Process Engineer, No.2 Project Engineering Department,
Chemical Engineering Center
834
-------
INTRODUCTION
¥e have been conducting research and development of a lime/limestone
scrubbing process since about ten years ago. In 1962 - 1963, we carried
out a pilot test at Amagasaki No.3 P/S, Kansai Electric Power Co. which treats
5,400 Nm3/hr flue gas from an oil firing boiler. Through this pilot test, we
have obtained technical data on gas cooling and carbon dust removal by means
of a spray tower and on the SO2 removal by means of various types of scrubbers.
In 1964 was constructed the first commercial lime-gypsum plant treating waste
gas from sulfuric acid plant at Koyasu works, Nippon Kokan Co. This plant has
been working smoothly for about six years producing high quality byproduct
gypsum for sales on the market.
Though the technical feasibility of the lime-gypsum process has proved that
it can treat waste gas from sulfuric acid plant as aforementioned, applica-
tion of the process to the power plant includes some problems to be solved.
That is, we have to overcome difficulties due to the differences of the
characteristics of these plants as shown in Table 1. This paper describes
shortly the results of pilot tests which have been carried out to solve these
problems as mentioned hereunder and an outline of the prototype plant being
under construction at Amahigashi P/S, Kansai Electric Co.
(l) How to design an economical and efficient scrubber suitable for treat-
ing a large volume of flue gas with lime slurry.
(2) How to prevent the plugging due to deposition of sludge and occurrence
of hard gypsum scale.
(3) How to obtain high quality gypsum without the influence of impurities
from flue gas and in raw materials.
(4) How to satisfy both high S02 removal and high Ca reactivity in case thin
S02 is to be treated.
835
-------
OUTLINE OF MHI LIME-GYPSUM PROCESS
Though the process differs somewhat depending on conditions of waste
gases to be treated hereby, it consists of the following sections basically
as shown in Pig. 1.
(l) Gas cooling and dust removal section
This is a section to clean, moistened and cool flue gas with water be-
fore the gas is sent to absorbers.
(2) Absorption section
This is a section in which oxides of sulfur are absorbed by lime slurry.
Ca(OH)2 + S02 + Aq = CaS03'|H20 + Aq (l)
Ca(OH)2 + S03 + Aq = GaS04-2H20 + Aq (2)
CaC02 + S02 + Aq = CaS03-|H20 + C02 + Aq (?)
CaC03 + SOj + Aq = CaS04-2H20 + C02 + Aq (4)
Ca(OH)2 + C02 + Aq = CaC03 + Aq (5)
CaS03-|H20 + y02 + Aq = CaS04'2H20 + Aq (6)
The main reaction of S02 absorption produces calcium sulfite as see;n
in equations (l) and (3) and some amount of calcium sulfate is also pro-
duced depending upon 803 and 02 contents of the flue gas and operating
conditions of absorber as seen in equations (2), (4) and (6).
(3) pH value adjusting section
The pH value of the spent liquor from the absorption section is adjusted
here to 4 - 4-5 by adding a small amount of sulfuric acid.
(4) Oxidating section
This is the section of converting calcium sulfite to gypsum by air oxida-
tion at 4 - 5 kg/cm^G.
(5) Gypsum filtering section
Gypsum is filtered out of the slurry delivered from the oxidation tower.
The filtrate is fed back to the cooling tower.
836
-------
(6) Impurities separating section
The impurities brought with flue gas or raw materials (lime) are sepa-
rated out of the system by neutralizing the waste water of the cooling
tower with slaked lime and filtering this liquor.
The filtrate is fed back to the process.
PILOT SCALE RESEARCH AND DEVELOPMENT
The photographs of the pilot plant at which the tests on the MHI lime-
gypsum process have been conducted are shown in Fig. 2 and 3. This plant was
designed to be able to produce 0.7 T/D of gypsum by treating 2,000 - 3,000
Nm3/hrof flue gas and installed at Hiroshima Technical Institute in October
1969. The test results are as follows.
1. Selection of the scrubber type - economical, efficient and free from
scaling
As they well know, the prevention of scaling trouble is the most
important problem at the designing of lime/limestone scrubbers.
From this standpoint, we adopted a spray tower as a scrubber of simple
structure for the treatment of waste gas from sulfuric acid. In this
case, the application of this process is rather easy because the amount
of gas is rather small and also it has higher S02 content, negligible
amount of C02 and no particulates.
To apply this process to other sorts of plant as shown in Table 1,
we have carried out a series of pilot test and found that the plastic
grid packed tower is suitable to the process «t± it can be operated without
*=>
837
-------
scaling trouble at a higher superficial gas velocity and lower draft
pressure loss. For the flue gas containing 0.1 vol % of SC>2, 90 to 98 %
of SO 2 removal is attainable at a superficial gas velocity of about 1.2
x 104 mVm2-hr and a total draft pressure loss about 90 mmH20. The flow
rate of scrubbing liquor is 70 m^/m^ hr, in this case.
2. Reaction rate in the scrubber
The absorption reaction occurred in the scrubber is mainly control-
led by the dissolving rate of lime as shown in Fig. 5 and the dimension
of the scrubber can be determined by the equation (l) and (2).
(i)
a
Kd = A-C0a-(L/G) (2)
From these equations and Fig. 4 and 6, it is clear that a higher pH
value is requested together with a higher liquid-gas ratio for attaining
a higher absorption efficiency. However, a lower pH value is necessary
for attaining a high conversion ratio of lime to calcium sulfite or sul-
fate (this is designated here as Ca reactivity). To accommodate these
conflicting factors, we have adopted two scrubbers combined system in
which one is operated at a lower pH and the other at a high pH.
The two scrubbers are installed in series in relation to gas stream - the
former is countercurrent and the latter is cocurrent. The scrubbing
liquor is at first fed to the latter and the effluent is then sent to the
former. The volume of each scrubber necessary for 90 % 302 removal" and
98 % Ca reactivity is related with pH value of the latter scrubber, as
shown in Fig. 7. This figure shows that the most economical point exists
at the pH value of 9, but practically we have to use a lower pH at which
the formation of calcium carbonate is not predominant in the scrubber
838
-------
when carbon dioxide content is high in the gas to be treated.
3. Prevention of plugging in the scrubber
Troubles often encountered in the lime scrubbing process are the ac-
cumulation of soft deposit and the formation of hard scale in the scrubber.
The former is usually caused by the deposition of calcium carbonate or
sulfite and it is preventable by operating the scrubber at a suitable pH
and avoiding the stagnant portion in the scrubber. The latter is caused
by the growth of gypsum crystals and then the desupersaturation of scrub-
bing liquor is the most effective countermeasure for the scaling. For
this purpose, addition of the seed crystals and installation of the delay
pipe are well known. And also to avoid stagnant flow or portions dried inter-
mittently in the scrubber is very important because the crystalline scale
of gypsum often begins to grow by concentration of the liquid at these
portions. We have realized all of these countermeasures at the pilot
test and found no sign of scaling after a 500 hours continuous operation.
4. Oxidation of calcium sulfite
The oxidation of calcium sulfite is conducted in a tower equipped
with the rotating cylinder type of air atomizer at 4 - 5 kg/cm2 and 50 -
80 C. The principle and features of this specially devised atomizer were
explained at the first symposium. However, we like to again emphasize
that this may be the most suitable one for the lime-gypsum process as it is
operated at a very high oxidation efficiency without fear of plugging.
The rate of oxidation depends on pH value of the slurry as shown in
Fig. g and we adopt pH 4 - 4.5 at practical operation. This figure
839
-------
suggests us that HS03 ion in the slurry may take an important role in
the oxidation reaction. The seed crystal added in the scrubbing liquor
serves again for the growth of gypsum crystals.
OUTLINE OP A PROTOTYPE PLANT
A prototype plant of MHI lime-gypsum process treating 100,000 Nm3/hrof
flue gas from an oil firing boiler is under construction at Amahigashi P'/S,
Kansai Electric Power Company. This plant is designed to remove 90 % of S02
from flue gas containing O.llvol % S02 and produce daily 19 tons of gypsum
for sales in which quick lime is used as absorbent. It is scheduled to be
completed in March 1972 and operated for about one year after the guarantee
run to confirm such problems as concerned with materials of construction,
quality of byproduct, performance characteristics and economics through the
long run. The production/demand of gypsum and distribution of limestone in
Japan are shown in Pig. 9 and 10 respectively.
SUMMARY
The lime-gypsum has been attracting particular attention in Japan as an
advantageous desulfurization process because it produces gypsum which has a
great possible share in the market by the simple process using lime as ab-
sorbent which is cheap and abundant in Japan. Here described are the test
results on the process at the pilot plant treating 2,000 NmVhr of flue gas
from an oil firing boiler, especially on the rate of absorption reaction and
the techniques for prevention of scaling. The technical feasibility and eco-
nomics of this process will be made clear through the operation of the proto-
840
-------
type plant which will be started to operate in March 1972.
NOTATION
Yj_n, Iout S02 concentration at the inlet and outlet ppm
of the scrubber, respectively
PO Pressure of gas in the scrubber atm
C0 Effective concentration of Ca in slurry kg-mol/m3
# Ca reactivity
Z Effective height of scrubber m
G" Molar superficial gas velocity in the scrubber kg-mol/m2.hr
Kd. Mass transfer coefficient based on dissolving 1/hr.atm
of Ca
841
-------
h-
U.
O
O
CO
U
s
o ^
^5
CO f-
ll-
u.
o
CO
o
8
1
CE D
o°.§
cE z ^
UJ ^ -J
CO 2 {/)
i
e?
z
tr .
Cfe
O Q-
PETROLEUM
REFINERY PLANT
FURNACE
o
^
o
a:
2Z
^<
W Q.
O
^
8
in
§
n
L
*•
8
in
o
o
(VJ
^
8
o
o
'
JANTITY
l03xNrr^H)
"""^ »
Q5 v"1^'^
s
8
n
^
o
r~
in
^*
i
o
n
8
n
*
%
— ^'"""*
is
(T
TEMPE
8
^
in
CJ
o
<
3
3
o
11
i
Q
N
c> ^ !2 —
^ ^ ) j
(VI /-N _«.
^5 ^^ ^^ O^
o ~
n
CJ «-' ^f ~~
. "* ™~
l * *
o n ~ o
0 ~
£? ^ oo n
• • • ^ •
O °O
-------
&
oo
oo
1 1 1
U
0
C£
Q_
-
(D
\
LU
•<-
<
1
n:
, ,
I
en
U_
00
o
1—
Q_
CXL
0
/ s\
00 }
o 5
00 <
/
^
^
LU
S
1 L
v
— i—
— \
^
^ o
O
1— ^
< — >
^^
o
X
O
t c
tx. [
O
< U
(D
t
00
<
ID
x-X
LU
Z)
1
u_
\^
U
1
^
^
<
— p>c
— ^(.
»
(
^— -
>^
3
/)
i_
V
>—
f-\
J)
o
JN
C
(\l
i
k
(3
00
oj
U
/
\
00
a:
LU
h-
-------
Fig. 2 (left)
Photograph of the scrubbers
at which pilot tests have
conducted on the lime-
gypsum process.
Fig. 3 (under)
Photograph of the pilot
plant besides the scrubbers
at which the tests have
conducted on the lime-
gypsum process.
This picture shows inside
of the house which is
seen in the Fig. 2.
-
i§
844
-------
SP
ON
100
so
5 60
o
51
ol
O
CO
20
-j
^/
TS
LOW •*•
PH
HJG-H
REMOVAL/OPERATING COMDITION
500-WOO
2-7
Remarks
O
•
Grrld pitch.
2-5 Turn.
3 0 mm
4-0 w*
0-1 0.2 04 0-£ 0.8 1.0 Z.
EFECTIV/E Ca COWC,-Co (
F..F- RATE OF ABSORPTION REACTION
845
-------
X
X
o
-J
0 o
0 20 4-0 60 80 100
REACTIVITY (%)
Fig. 6 EFFECT OF PH ON Cai REACTIVITY
846
-------
UJ
QO
ca
a
a:
o
to
u.
o
LU
r
8.Qa
Ca Reactivity
TOTAL
No-2
SCRUBBER
4 5 6 7 8 9 10
PH IN Nlo-2 SCRUBBER
Fig. 7 RELATION- SCRUBBER yOLUME/pH IN No.2
847
-------
NO
v 0.8
**:
0.6
04
0-2
0.10
s
o
oc.
Of.
0.06
0.02
CO
I
456
pH IN OXIPATION TOWER
EFFECT op PH ON OXIDATION R/ATE
o 3
— Demand
|| Production
for
OTHERS
for
CEMENT
for
GrYPSUM
8 '69 '70 '71 '72 '73 F. YEAR
PRODUCTION/DEMAND OF
IN T/\PAN
848
-------
UME STONE, PRODUCTION CHART(1959)
CF.9.IO)
TOTAL 91.3QaOOQTon
WIT 1QQQTon
849
-------
-------
WET SCRUBBER INSTALLATIONS
AT
ARIZONA PUBLIC SERVICE COMPANY
POWER PLANTS
PRESENTED BY:
LYMAN K. MUNDTH
VICE PRESIDENT, POWER PRODUCTION
ARIZONA PUBLIC SERVICE COMPANY
PHOENIX, ARIZONA
PRESENTED AT:
INTERNATIONAL
LIME/LIMESTONE WET SCRUBBING SYMPOSIUM
SPONSORED BY DIVISION OF CONTROL SYSTEMS
U.S. ENVIRONMENTAL PROTECTION AGENCY
NEW ORLEANS, LOUISIANA
NOVEMBER 8 - 12, 1971
851
-------
WET SCRUBBER INSTALLATIONS
AT ARIZONA PUBLIC SERVICE COMPANY
POWER PLANTS
INTRODUCTION
Arizona Public Service Company is pioneering the installation of wet
flue gas scrubbing equipment on coal fired power plants with work
currently progressing at two of its operating stations, one in New
Mexico and the other in Arizona.
At the Four Corners Plant, near Farmington, New Mexico, scrubbers
are being installed on three units with a total net generating capacity
of 575 MW (nominal) .
The other installation is at the Cholla Power Plant at Joseph City,
Arizona, a single unit, with a rating of 115 MW (nominal).
The Four Corners Plant is fueled with sub-bituminus coal from the
Navajo Mine which is adjacent to the plant. This coal has a high
ash content; ranging from 6 to 25%, and averages about 22%. The
coal has a low sulfur content; with a range of 0.4 to 1.9% and aver-
ages about 0.7%.
The Cholla Plant coal is from the McKinley Mine at Gallup, New
Mexico. The ash content varies from 5 to 15%, with an average of
about 8%. The sulfur content is from 0.4 to 1.0% with an average
of about 0.5%.
852
-------
I. EQUIPMENT CONCEPT
At the time this report was prepared both the state of New Mexico and
Arizona had particulate emission regulations, but neither state had
adopted SO emission limits for power plants. However, both states
are currently in the process of adopting SOo emission regulations.
The wet scrubber concept of cleaning flue gases was selected because
of concern for the effectiveness of electrostatic precipitators with low
sulfur coal and for the SOo removal capabilities of the scrubbers.
A. Four Corners Plant The scrubber installation at the Four
Corners Plant is being performed by the Chemical Construction
Company (CHEMICO) and will use Chemico's variable throat,
high energy, wet approach venturi for particulate removal. A
pilot plant test has been conducted to obtain design data and
operating parameters for the scrubber system.
B. Cholla Plant The Cholla installation has been awarded to
Research-Cottrell, Inc. for design and construction and will use
a high energy, flooded-disc venturi for particulate removal and
an absorber tower following the venturi for SO removal. A
h
5000 CFM pilot plant is presently being installed to optimize recycle
rates, verify materials of construction, and to conduct various
chemical control tests.
853
-------
II. ADDITIVE ABSORBER
Neither plant has a completely engineered SG>2 removal system at the
present time, although the basic concepts of SC>2 removal have been
studied and provisions are being incorporated.
A. Four Corners Pilot plant tests conducted at the Four
Corners Plant by Chemico demonstrated that approximately
10 to 20% SO7 removal occurs simultaneously with particulate
£1
removal in the venturi. The capability exists for further SC>2
removal by injecting supplementary sprays, after the venturi,
in the lower section of the scrubber vessel and appropriate
connections are being provided at this time for these sprays.
B. Cholla Plant In the Cholla scrubber, SO2 removal will
take place in a second stage, utilizing a wetted film, fixed
packing absorber tower with an alkaline solution as the absorp-
tion medium.
III. PROCESS CONFIGURATION
Both the Four Corners and the Cholla scrubber systems winl employ a
continuous loop scrubbing cycle with part of the cycle liquor bled-off
to continuously agitated thickeners. The clarified water will be returned
to the scrubbing cycle for make-up.
854
-------
Both installations provide for reheat of the flue gas before entering
the stack to minimize negative bouyancy effects of the plume and to
prevent condensation of the flue gas within the stack.
Both scrubber installations are retrofits to existing plants with the
major part of the construction taking place with the units in service.
A. Four Corners Plant At Four Corners, space is limited.
The venturi scrubber vessels (two venturi scrubber vessels
on each unit) have been mounted above the existing mechan-
ical dust collectors and induced draft fans. The particular
arrangement was dictated because of limited construction
space and priority of continuous plant operation during the
scrubber installation. With two scrubbers on each unit, and
appropriate ducting and baffles, the unit will be able to
operate at 50% load if one of the scrubbers is out of ser-
vice for repairs or maintenance .
New induced draft fans will be placed downstream from the
scrubbers, and designed to handle the wet gas. The fans will
have stainless steel rotors and rubber lined casings. The
fans will discharge into mist eliminators which will help
reduce the flue gas reheat requirements by drying the flue
gas before it enters the reheater.
855
-------
B. Cholla Plant At the Cholla Plant, the existing mechanical
dust collectors and induced draft fans will be retained. The
new scrubber system will be installed in series with , and
downstream from the existing mechanical collectors and I. D.
fans. New booster fans are required to operate in series
with each of the existing induced draft fans to satisfy the addi-
tional pressure drop requirements of the scrubber system. More
construction space is available at this site, and the scrubbers
will be installed at ground level. Again, two scrubbers will
be installed to provide the capability of half load operation
when one scrubber is out of service.
IV. WASTE DISPOSAL
At both plants it is planned that the concentrated solids (fly ash and
sulfates and sulfites of calcium) in the thickener under-flow will be
transported to the existing plant ash disposal area in a separate trans-
port system. Presently, wet sluice systems transport the fly ash to
retention ponds.
V. PRESENT STATUS OF CONSTRUCTION
Construction and installation of the Four Corners scrubbers is nearly
complete. The scrubbers are scheduled to be in operation by December,
31 of this year (1971). These scrubbers will be "tied-in" during
856
-------
scheduled maintenance outages continuing through the latter part of
this month (November) and into December.
The Cholla scrubber is scheduled to be in operation by December 31,
1972. The design work and materials procurement is currently in
progress. The pilot plant is being installed at the plant at the present
time, and scheduled to be operating after the first of the year.
VI. DESIGN CONSIDERATIONS
The nature of an electric utility requires a very high degree of reliability
in its generating facilities. The installation of largely untried and com-
mercially unproven pollution control equipment will most certainly de-
grade the reliability of the generating units on which it is installed.
Consequently, an extreme amount of care must be taken in the design
and selection of wet scrubbers, or any other SO2 removal equipment
to maintain this high degree of reliability. Specific items to be con-
sidered are:
1. Redundancy of Critical Equipment Because emission regu-
lations will, in most cases, prohibit the operation of generating
units without the pollution control equipment in service, the same
redundancy is required in the design of this equipment as is used
857
-------
throughout the rest of the plant. This means that for any
critical piece of equipment, a spare should be installed
with provisions for on-line isolation and repair.
2. Controls The Controls, and the control system philos-
ophy of the removal system must be compatible with, and
integrated into, the control system for the rest of the plant.
3. Plant Personnel Plant personnel must be trained in the
operation and maintenance of the removal equipment.
4. Materials Careful specification of materials of con-
struction is required throughout and each component of the
removal system must be individually studied to determine
the severity of duty to which it will be subjected.
5. Equipment Layout Thus far, the experience of industry with
wet type removal systems has demonstrated many problems and
poor reliability. Consequently, careful attention must be
given to the layout and arrangement of the system. Ready
access must be provided for maintenance, repair, and re-
placement of equipment.
858
-------
6. Guarantees On the subject of guarantees, a manufacturer
may be able to design equipment to meet given proformance
specifications and meet a certain particular guarantee. However,
I do not believe you will find a manufacturer who will guarantee
the performance of his equipment over the period of its useful
life. This is, in effect, what the regulatory agencies require
of the utility operator.
859
-------
^N
S*
•^ c\
O
t—t
H
O p re
O re O
H W £5
O § Q
1—I f-^.
> re ^
w O O
CO CO i-J
,. I PH
U H ^
O ^
Is
re
O
860
-------
*
o
I—I
H
§
< EH
^ >5
w PQ n
O CQ O
CQ
CL. I—1
K Cu
< w 5;
O
ffi
o
861
-------
-------
DETROIT EDISON FULL-SCALE DEVELOPMENT PROGRAM
FOR ALKALI SCRUBBING SYSTEMS
J.H. McCarthy
J.J. Roosen
Detroit Edison Company
Detroit, Michigan
Prepared for
Second International Lime/Limestone
Wet Scrubbing Symposium
New Orleans, Louisiana
November 8-12, 1971
863
-------
The text material for this paper was included in Paper
No. 4c, "Detroit Edison Pilot Plant and Full-Scale Development
Program for Alkali Scrubbing Systems—A Progress Report." There-
fore the paper is not presented here. J.H. McCarthy and J.J.
Roosen of Detroit Edison Company were the co-authors of both
presentations.
864
-------
A FULL SCALE LIMESTONE
WET SCRUBBING SYSTEM FOR THE
UTILITY BOARD OF THE CITY OF KEY WEST, FLORIDA
ROBERT R. PADRON
Superintendent of Engineering
Utility Board of the City of Key West, Florida
KENNETH C. O'BRIEN
Supervising Engineer
R. W. Beck and Associates
Denver, Colorado
In early 1972, the Utility Board of the City of Key West will put
into operation a full scale limestone wet scrubbing system on the initial 37 MW
unit of its new power plant, now under construction.
This presentation outlines the decisions which led to the installation
of the system and includes a description of the system, economics, proposed
operation, and possible operating problems.
865
-------
INTRODUCTION
Key West, an island city, is approximately five square miles in area
and located 160 miles southwest of Miami, Florida. Key West, as well as the
entire State of Florida, is widely known for its tourism, fishing and water
recreation, and is boastful of its clean and clear water.
The Utility Board of the City of Key West provides electrical energy
for Key West, the U. S. Navy, and adjacent communities in the lower keys.
The generating facilities for the system are currently comprised of approxi-
mately 60 MW of steam power plant generating equipment and 17 MW of package
diesel generators.
In early 1968, design was started on a 37 MW power plant to be located
on Stock Island, adjacent to the City of Key West. The plant was required to
maintain firm capability to meet the growing system load.
In recognition of the increasing emphasis on the protection of the
environment and pollution control by both private citizens and governmental
agencies at all levels, the Utility Board established a policy of considering
all practical methods of controlling pollution at the new power plant.
At the time the plant design was being completed, the air pollution
control requirements for the State of Florida were under revision. Consul-
tation with the Florida State Director of Air Pollution and the Southeastern
Regional Director of Air Pollution for the Federal Government revealed that
the revised criteria would be relatively strict with regard to S02 and '
particulate flue gas emissions.
The fuel for the new plant was residual oil, selected on the basis
of availability and economics. The residual oil available at the time contained
approximately 2.75% sulfur. The fuel oil suppliers advised that residual oil
would be available at some time in the future with a sulfur content of about
0.8%. However, even with the lower sulfur content, preliminary calculations
indicated that the future air pollution control requirements could not be
met without some form of pollution control equipment.
The Board's policy to control pollution, coupled with the certain
adoption of new air pollution restrictions and public opinion, resulted in the
inclusion of an SOo and particulate scrubbing system as a part of the steam
generator specifications. The scrubbing system was included as an alternate
bid item to allow the scrubbing systems to be evaluated separately.
A number of scrubbing processes were at various stages of develop-
ment at the time, but there was some doubt as to their commercial success. However
there appeared to be sufficient prospective economies in using a scrubbing
system to warrant careful investigation.
DESIGN PARAMETERS
Once it was established that some form of flue gas scrubbing should
be considered, a survey of candidate processes was made.
Some of the factors considered were:
1. Estimated capital costs.
2. Estimated operating costs.
3. State of development.
4. Simplicity of Process.
5. Availability of raw materials.
6. Operating flexibility.
Based on the evaluation of these factors, the limestone wet
scrubbing system was selected for primary consideration.
866
-------
The scrubbing systems normally employ scrubbing devices with
proprietary designs. For this reason the specifications were developed
on a functional or performance basis, with the primary purpose to obtain
proposals on equipment which would be suitable for continuous operation with
a base loaded power plant.
Some of the functional requirements included were:
1. A guarantee of SC>2 and particulate removal as follows:
a. 85% of S02 produced when burning fuel oil with
a 2.75% sulfur content.
b. 95% of all particulate matter generated during
the combustion process.
2. The scrubbing liquid was to be a slurry consisting of
sea water and native coral marl. The marl (limestone)
is the primary constituent of all land areas in the Keys,
and is consequently readily available. Sea water rather
than fresh water was specified due to the extremely high
cost of fresh water in the area (in excess of $2.00/1000
gallons.)
3. A complete system was required, including tanks, conveyors,
pulverizers, storage bins, pumps, piping and controls.
The raw marl delivered to the system was specified to be
a maximum size of 2 inches in diameter.
4. Materials of construction were required that would withstand
the corrosive effects of a sea water and marl slurry in con-
tack with flue gas. It was further required that the scrubber
be designed to safely withstand a gas temperature of 650° F
in the event of an air preheater failure.
5. It was required that the spent slurry be of such quality that
the sea water could be discharged into the bay and meet all
applicable pollution restrictions once the solids had been
precipitated out in a settling pond.
6. The manufacturers were required to submit operating costs
for the system to be used in evaluating the proposals.
The specifics of design necessary to meet these requirements were
established as the responsibility of the manufacturer submitting the proposal.
EVALUATION OF PROPOSALS
Four proposals were received for steam generating equipment, three
of which included proposals for limestone wet scrubbing equipment.
Values listed in the proposals for the scrubbing equipment ranged
from $345,000 to $738,000. Although it appeared that there was some economic
justification in installing a wet scrubbing system, a bid analysis was
completed which indicated that the lowest bid for the steam generator and wet
scrubbing system was substantially in excess of the funds available. For this
.reason, the steam generator contract was awarded without the wet scrubbing
system.
At this time it was learned that the National Air Pollution Control
Administration was in a position to make certain funds available for a full
scale limestone wet scrubbing system.
867
-------
A request was made by the Utility Board of NAPCA that Key West
be considered for the installation of a limestone wet scrubbing system, to
be jointly funded by the Utility Board and NAPCA. A series of negotiations
took place which resulted in a jointly funded full scale limestone wet
scrubbing system for the power plant. The system selected for the project
was designed by Zurn Air Systems. The Zurn system was selected on the basis
of the lowest total evaluated bid which complied in principle with the
specifications. The scrubbing device is described by the manufacturer as
a modified impingement type scrubber.
Concurrently with the negotiations between the Board and NAPCA,
a series of pilot plant tests were conducted at the existing Key West Power
Plant using the Zurn Air System's scrubbing device, which demonstrated the
feasibility of the system.
A flow diagram for the complete scrubbing system and a schematic
diagram of the scrubbing device are shown on Exhibit I.
DESIGN AND CONSTRUCTION
A change order to the steam generator contract was executed in
October, 1970, to include the wet scrubbing system. Design and construction
progressed at a rapid rate and at the present time the installation is
approximately 70% complete.
A considerable amount of review during the design phase was
exercised by the Utility Board and their Consulting Engineers as well as by
the Office of Air Programs for the Environmental Protection Agency (previously
designated as NAPCA) to provide as much constructive input into the design
as possible.
Design areas of particular interest were as follows:
1. Material selection -
The ability of the materials used to withstand the corrosive
conditions was considered critical to the successful opera-
tion of the system. Although it was known that it might be
necessary to replace certain parts of the system with more
"exotic" materials during the initial operating period,
materials were selected which appeared to be the most
practical from economic as well as performance aspects.
In general, stainless steel combined with non-metallic
coatings was used whenever corrosive conditions were
anticipated. Some material selections were based on
experience gained during the pilot plant tests.
2. Safety -
The performance of the wet scrubbing device is directly
related to the slurry level in the scrubbing device.
Close control of the slurry level is required. Interlocks,
automatic overflow devices, and an automatic flue gas
bypass were included in the design to avoid the possibility
of the slurry level rising above the tubes and consequently
blocking off the flue gas flow. Additional interlocks and
safety devices were included to protect materials not
designed to withstand the high inlet gas temperatures in
the event of sea water supply failure.
868
-------
3. Operation -
The control system was designed to allow normal continuous
operation of the scrubbing system under a relatively constant
load. Automatic operation and simplicity were considered
to be key design criteria for the controls. Provisions for
instrumentation required during the Office of Air Programs'
test were included.
Additional design criteria were established by the manufacturer
as a result of the pilot plant tests being conducted in Key West.
The general arrangement of equipment for the system is shown on
Exhibits 2 and 3.
The entire power plant, including the scrubbing system, is
scheduled for start-up in the early part of 1972.
PROPOSED OPERATION
No operating experience other than that gained during the operation
of the pilot plants is available since the wet scrubbing system for Key West
is still under construction. Therefore, the procedures outlined in this
section are tentative, and will be modified as required when the system is
put into operation.
The various components of the wet scrubbing system and the material
flows are shown on the flow diagram, Exhibit I.
In general, the wet scrubbing system is divided into two sections:
The first section is the marl preparation plant. Here coral marl, 2 inches
in diameter or less, is received in a pick-up hopper, and after initial
crushing to 3/4 inch size is placed in the storage bin. The bin has
sufficient capacity to hold a 2-day supply of marl under full load conditions.
The material handling equipment up to this point is operated from local manual
control stations, on an intermittent basis.
The marl is then fed through a roller mill which reduces the particle
size to 90% - 325 mesh, and then on to the surge bin, ready for mixing with
sea water. The roller mill feed rate is not easily adjustable and for this
reason must be set for the average expected load condition. It must be started
nanually and will run continuously.
The second section of the system is the actual scrubbing section.
Here marl is fed in correct proportion to sea water from the surge bin into
the mixing tank to form the desired slurry concentration.
The slurry is pumped to the two scrubbers, each handling half the
:lue gas. The partially spent slurry is pumped out of the scrubber, some
jeing recirculated to the scrubber and the remainder being pumped to the
settling ponds. The slurry flow to the scrubbers and the slurry level in
;he scrubbers are automatically controlled, based on manual set points
established in the main power plant control room. The set points are
jetermined by the unit load and readings from the S02 monitor located
n the stack.
The flue gas passes from the induced draft fan through a spray
;ection where the gas is cooled from 350° F to 160° F. It then passes
;hrough the scrubbers where the S02 and particulates are removed, and
;hen to the stack. The flue gas pressure drop across the scrubber is
bout 12 inches F^O and is a function of the slurry level in the scrubber.
869
-------
The scrubbing section can be run on a relatively automatic basis
from the main control room once the equipment has been started. A number
of automatic controls are included to allow safe operation of the unit. A
bypass around the scrubbers is provided for maintenance purposes.
The Office of Air Programs of the Environmental Protection Agency
will be performing tests on the system during the first yeac. of operation.
It is anticipated that these tests will aid in establishing the optimum flow
rates and equipment operation.
Two settling ponds will be provided. One will be emptied while
the other is in use. It is anticipated that a front-end loader will be
required to handle this material.
POSSIBLE OPERATING PROBLEMS
Anticipated operating problems, particularly in the areas of corrosion,
scaling and disposal of spent slurry, are major factors that must be taken into
consideration in formulating operating procedures.
The problems of corrosion and scaling were observed in the initial
operation of the pilot plants and resulted in material and design changes. It
is anticipated that it will be necessary for the contractor to make a number
of additional changes once the system is put into operation.
To combat corrosion, the stainless steel scrubbers were epoxy lined
and the tube material was changed to fiberglas. It is possible that the change
of tube material will also solve the scaling problem, but if not, it was dis-
covered that by installing sprays and maintaining the tubes in a constant wet
condition the scaling problem would be eliminated.
Disposal of 50 tons per day of spent slurry is an obvious operating
problem and has not been completely resolved to date. The initial thinking
was that of utilizing the product to fill approximately twenty acres of City
owned baybottom land adjacent to the steam plant site. Although this matter
is still being pursued, the State of Florida's recent stand of opposing
the filling of any submerged land productive to marine life has made it a
less plausible solution.
Initially the slurry will be placed on land above sea level to avoid
this problem. Other methods of disposal are currently being investigated.
Wind tunnel tests were conducted during the design of the system
tg determine if a reheat system would be required for the flue gas. The
results were inconclusive. It will therefore be determined during the initial
operation of the plant if this will be a problem.
SYSTEM ECONOMICS
The cost information shown on Exhibit IV summarizes the estimated
capital investment requirements and operating costs for the scrubbing
system. A cost evaluation, also shown on this plate, relates the system
costs to the plant capability and output, and also compares them to the
cost of burning low sulfur fuel.
870
-------
CONCLUSIONS
The mandate has been issued to limit S02 and participate emissions
from power plants.
The limestone wet scrubbing process does not appear to be the
universal answer in reducing these emissions, but is considered to be the
best economical selection for the Stock Island Plant. It is likely thac
the selection of limestone wet scrubbing equipment will become economically
more attractive as the demand for low sulfur fuel increases.
A certain number of operating problems must be expected during
the initial phases of operation, but it is anticipated that the solutions to
these problems are available within the limits of the current technology.
871
-------
H 'A J. OZ HOJ.VA313
i3«na
HdO.01 5
>-.'
0
n
Ul
z.
t-
y
\
\
872
-------
0.1
I
z
LU
5
LU
O
C£
<£
LU
LU|
O
CC
o
(0
Ul
/--• -'
xf x
x, ;-•
^x
v.
tur
873
-------
874
-------
o olo
o olo
o olo
o
o;
NNUAL OPE]
U
COSTS
"
ne stone)
^.
t,
1
hi
V
°S
££
.5 D
'x ""*
3 .
*^ ei
t
00
C
•O
C
"rt
bi
0}
rt
*
j-
fM
g
2
U
00
C
1
u
*V h.
JD
in
'rt
CU
£
(4
O
C
V
.S
2
1
"
t/) P
J 'o
0 |
«$•
t)
sf conduit wiring
nstructlon including wet
control center
Mechanical co
t/i
3
S
a
00
C
o.
'S.
1
S.
o
hi
Ling equipment for marl and
scrubbing con!
control tubing
Material hand:
spent slurry
•+Q
TOTAL
Engineering
r-
(4
0)
O
o
o
0
n)
2
S
o
r-
00
X
in
00
X
^
O
o
0
r-
X
J<
(D
^3
1
r-
vO
o"
l(
«
V
w "1
x: S
> >-.
v c
U («
*• 5
* ^
gl
§ 5
u ^
c °
0 <=
11
f4 ^3
•° ^
* u
sS;
^ Z
rt
S
a.
'5
4)
00
1
3
hi
u
T3
r4
1
c
'o
^*
in
*>
i ma. ted
t.
c
n
S
.2
.3
0.
(4
U
"o
hi
rd
O
>*
hi
S.
o
*»
(4
-o
5
S K _
•3 S
'o
§s
*> pq
.S
'a
W
to
O
U
O
g
H
U 2
°* t
° 3
Q 2
U] O
2 ^
yj Z
< **•
-
w
O
u
rt
•d
0
*j
C
V
rt
h
C*
C
S
u
3
C
00
1)
T)
hi
m
V
d
|
§
'x
g
"c
conte
)H
3
4-1
M
0
u
I
0
u
p.
OO
1
hi
3
*3
U)
£
0
O
in
O *•
U w
o
rt u
^ h
1_l
VOTUZ1
V
•S
M
C
j5
*• "§
V) hi
0 0
U (0
K ^
U °
? >.
0 h
P*. ti
>.s
.
, S £ .
-
>< 3 «
2 i! v
^
^
rt
a
U
"c
fS
S
^_,
o
U
d
rt
U
O
o
o
o"
* «
•C 0
o U
i 1
^ * I
o ** rt
Tp O ^
r- [-^ ^
0
h
«.
OB
0
pe rating c
o
*rt
3
G
G
<
<4
M (Nl
O
O
in
r-
^
OB
O
nt annual <
u
rt
^
3
CT
Ul
*•
«
>*
o
o
00
„**
^2-
^
1^
« H
IB
4
5.
n)
" in
o W
H
O
H
O
O
U,
^ S
jj hi
1 1
a +?
f 3
c S
9 'M
« U
f-a- 6
? V 0
^ O -H
-C U *
*j ui O
*. fl 0.
S^ 2
•2 « a
c ^ c
03 o
•O j T3
« TJ «
fl C ^
W (4 CQ
— —,
C —
* S
"^
11
« 8
1 ••
h ^
31
6^
o «
j5
c «
M O
« **
C T3
0 0
T3 G
V 3
-------
-------
AIR POLLUTION CONTROL AT THE
NORTHERN STATES POWER COMPANY
SHERBURNE COUNTY GENERATING PLANT
By
J. A. NOER
Mechanical Engineer
Plant Engineering and Construction Department
Northern States Power Company
Minneapolis, Minnesota
A. E. SWANSON
Director - Nuclear Activities
Black & Veatch, Consulting Engineers
Kansas City, Missouri
Presented Before
The Second International
LIME/LIMESTONE WET SCRUBBING SYMPOSIUM
New Orleans, Louisiana
1971
877
-------
s—
z
-------
ACKNOWLEDGEMENTS
The authors express their appreciation for the suggestions by the
many people from Northern States Power Company and Black &
Veatch who reviewed the paper.
879
-------
ABSTRACT
The paper describes the Air Pollution Control System and related consider-
ations of the two-unit Sherburne County Generating Plant. The plant will be
located near the Mississippi River about 40 miles northwest of Minneapolis
and will have a net capacity of 1360 MW. Fuel will be low sulfur western
coal delivered by unit train. Commercial operation of the two units is
scheduled for 1976 and 1977 respectively.
Northern States Power Company (NSP) encouraged the formation of the
Citizens Advisory Task Force comprising concerned citizens and represen-
tatives of public agencies and conservation organizations. This group's contri-
bution in the plant site selection and in the development of environmental
protection criteria is discussed.
The authors present the background and considerations which led NSP to the
decision to use a limestone wet-scrubber for paniculate collection and SC>2
removal. The scrubber selected is the Combustion Engineering Tail-End Lime-
stone Additive System using a 10 inch layer of glass marbles as the con-
tactor. The paper describes the components and physical arrangement of
equipment as well as simplified flow diagrams and operating features. The
status of the scrubber system development including effluent control, cost
estimates, and schedule requirements is given.
880
-------
AIR POLLUTION CONTROL AT THE
NORTHERN STATES POWER COMPANY
SHERBURNE COUNTY GENERATING PLANT
INTRODUCTION
Northern States Power Company (NSP) will construct a two-unit, 1360 MW
electric generating plant near Becker, Minnesota for commercial operation in
1976 and 1977. The fossil fuel plant, named Sherburne County Generating
Plant, will burn low sulfur western coal.
The two previous NSP base-load generating plants are nuclear plants, and it
was NSP's desire to build a nuclear plant for commercial operation in 1976
and 1977. Due to the regulatory uncertainties at both the State and Federal
level, the growing public resistance to siting of nuclear plants, and finally the
amount of nuclear generating capacity in the Company's system, a decision
was made to proceed with a fossil fuel plant. Through concerted efforts to
involve the public in environmental planning, the use of low sulfur coal, and
the use of advanced technology for pollution control, it appears that a
coal-fired plant will provide acceptable environmental protection and avoid
delays associate^, with nuclear plants.
881
-------
PLANT DESCRIPTION
The Sherburne County Generating Plant is located in Becker Township,
Sherburne County, Minnesota, approximately 40 miles north of Minneapolis.
(See FIGURES 1 and 2.) The site area is about 1,300 acres.
The site is now cultivated land with an average elevation of 965 feet above
mean sea level. The land is nearly level, with a maximum elevation differ-
ential of about 10 feet. The main plant facilities are set back more than half
a mile from the Mississippi River, so that an undisturbed zone is maintained
along the waterfront. (See FIGURE 3.) The normal river water elevation is
920 feet mean sea level.
The plant will comprise two electric generating units (Units No. 1 and No. 2)
each rated at 680,000 kilowatts net electric generating capacity. The steam
generators, turbine generators, wet scrubbers, and associated auxiliary equip-
ment will be enclosed in the main building structure. Unit No. 1 is scheduled
for commercial operation on May 1, 1976, with Unit No. 2 scheduled for
operation one year later.
Black & Veatch (B&V), Kansas City, Missouri, is the Architect-Engineer and
has overall design responsibility for the plant. Management of design, con-
struction, and quality assurance is the responsibility of NSP's Plant Engineer-
ing and Construction Department.
Capital expenditures are estimated to be over $360,000,000 for the plant.
This includes more than $25,500,000 for the air pollution control system.
The steam generators will be supplied by Combustion Engineering, Inc., and
will be of the balanced-draft type designed for burning pulverized coal. Each
steam generator will be rated at 4,985,000 pounds of steam per hour and
will require a gross heat input of 6,757 million Btu per hour.
Each turbine generator will have a gross capacity of 720,000 kilowatts and
will be supplied by General Electric Company. The turbine inlet steam
pressure will be a nominal 2,400 pounds per square inch (psig). The inlet
882
-------
steam temperature will be 1,000 degrees Fahrenheit (F) and the reheat
temperature will be 1,000 F. Each generator will have a rating of 800,000
kVA at 0.90 power factor.
The furnace will be equipped with tilting tangential burners and will be
designed for low combustion temperatures which are expected to reduce to
the practical minimum the formation of nitrogen oxides.
Each steam generator will be equipped with a wet flue-gas scrubber, capable
of removing 99 per cent of the paniculate matter and 50 per cent of the
sulfur dioxide in the combustion gases. The scrubbers are described in greater
detail later in the paper.
Combustion gases leaving the scrubbers of the two units will be emitted to
the atmosphere from a single chimney at least 550 feet tall.
Each electric generator will be connected through a power transformer to a
transmission substation. The substation will feed four 345 kV transmission
lines supplying power to the existing NSP electrical transmission system.
The turbine condensers will be cooled with water from cooling towers
operated as part of a closed-cycle circulating water system. The towers will
be of the wet, mechanical-draft, cross-flow type and will be oversized to
minimize visible plumes. The cooling tower for each unit will consist of 20
cells, each equipped with a 28-foot diameter fan capable of moving about
1,500,000 cubic feet of air per minute. The air-vapor mixture will be
discharged from individual fan stacks about 60 feet above grade. The circu-
lating water flow rate for each unit will be approximately 240,000 gallons
per minute (gpm). The cooling tower for Unit No. 1 will be located 3,000
feet from a similar tower for Unit No. 2 to reduce the combination of plumes
from the two towers.
Fuel for the plant will be low sulfur (0.8 per cent or less) subbituminous
coal from the Colstrip area of Montana. Coal shipment is to be made by
unit trains. Coal handling facilities at the plant site will include automatic
sampling facilities, rotary car dumper, stacker-reclaimer, crushing facilities,
883
-------
and a system of conveyors for transfer of coal between elements of the
system. A generally circular railroad track, which will handle the unit trains,
will enclose the coal storage area at the plant site. Storage at the plant site
will provide a 90 day supply of coal. This will require approximately
1,600,000 tons of coal, stored to a depth of about 40 feet. The maximum
design burning rate is 814 tons per hour for the two units.
Ash from coal combustion will be collected in the bottom ash hoppers and
scrubbers associated with each unit. Ash from these collection points will be
sluiced to a water-filled ash storage basin formed by an earthen dike. The
bottom of the storage basin will be sealed to minimize seepage into ground
water.
Water sources for operation of the plant will be the Mississippi River and
wells. After maximum reuse, all process return water from the various plant
systems will be directed to a common water holding basin with a minimum
retention time of 24 hours. Water for the ash handling systems will be reused
process water. All process return water which cannot be recycled will be
treated to meet applicable water quality standards prior to release to the
river at a single point of discharge.
BACKGROUND OF ENVIRONMENTAL CONTROL
In Minnesota, as in most other states, there has been no effective method of
resolving conflicting viewpoints concerning the environmental aspects of
power plant siting and development. In an effort to find a new approach to
resolution of conflict, NSP discovered that many environmental conflicts
involved in its past construction programs arose from a lack of early public
participation in its environmental planning activities. Northern States Power
Company also realized that the public hearing process is an inadequate
method for communicating with the public or involving the public in plan-
ning and decision making processes. As a result of evaluating past environ-
mental planning programs, NSP created the Citizens Advisory Task Force to
884
-------
provide a forum for in-depth discussions on environmental problem solving
and power plant siting.
Participating in the new open-planning group were representatives from the
Minnesota Conservation Federation, the Scientific Study Areas Committee of
the Minnesota Academy of Science, Minnesota-Wisconsin Boundary Area
Commission, Minnesota Chapter of the Sierra Club, Minnesota Committee for
Environmental Information, Minnesota Environmental Control Citizens
Association, Minnesota Chapter of the Isaac Walton League, Minnesota Envi-
ronmental Defense Council, Save Lake Superior Association, Zero Population
Growth, St. Croix River Association, Clear Air — Clear Water Unlimited, and
the League of Women Voters.
Northern States Power Company believes that it is important to seek out
those individuals who can effectively represent the concerned public to assist
in major decisions relating to power plant siting and pollution control. There
are two reasons for this approach: (1) the installation will more closely
represent the public interest; that is, it will be consistent with what the
public appears to be willing to pay for, and (2) delays in securing permits
and licenses may be avoided by using the advice of those who represent the
public.
After about six weekly meetings, the Citizens Advisory Task Force generally
lost its hostility toward NSP and developed a deep concern in accomplishing
the objective of plant site selection. The group took under advisement four
alternate sites to determine which site should be selected for a 680 MW fossil
fired power plant to be in service in 1976. The Task Force recommended
the Sherburne County site and NSP concurred, even though another location
was the Company's first choice.
Northern States Power Company conducted a thorough investigation of alter-
nate methods of air pollution control for Sherburne County, in line with the
following Task Force guideline contained in the group's report:
Emission of all air-pollutants should be reduced to the minimum
technologically feasible. The cleanup should include not only
885
-------
paniculate matter but also gaseous pollutants such as sulfur
oxides. The Task Force noted the Company's stated intention to
use low sulfur western coal in its next plant, but cautions the
Company against regarding this use as removing the need for
sulfur oxide cleanup as it becomes technologically feasible for
other coals. Adequate space should be provided for future addi-
tions of air pollution control equipment to accommodate possible
future technological developments and revised air emission
standards.
The collection of fly ash from low sulfur coal is difficult with a "cold"
electrostatic precipitator (with a flue gas temperature at the precipitator inlet
of about 300 F) due to the high resistivity of the ash at normal exit gas
temperatures. There are also problems of energizing a "cold" precipitator at
low loads and startup. For these reasons, a "cold" precipitator was not
considered practical. The alternatives considered were a "hot" precipitator
(with a flue gas temperature at the precipitator inlet of about 700 F) or a
wet flue gas scrubber. Even though economics were somewhat in favor of a
"hot" precipitator, NSP selected the wet scrubber because there is a high
degree of confidence in collecting particulates and there would be some
reduction in the sulfur dioxide emissions.
SCRUBBER DESCRIPTION
The air pollution control system includes a tail-end limestone wet scrubber
designed to remove both particulate and 809 from the boiler flue gases. The
manufacturer guarantees that the system will remove 99 per cent of the
particulates and 50 per cent of the SO 2 entering the scrubber if either high
calcium limestone or dolomitic limestone is used as an additive.
The flue gas enters each module below a bed of glass marbles at a design
temperature of 290 F. (See FIGURE 4.) As the gas turns upward, large particles
are separated immediately. Sprays provide a constant supply of water to the
underside of the bed of marbles. Water entrained in the gas stream floods the
spaces between the marbles. Violent mixing and physical contact occurs which
886
-------
results in the capture of substantially all of the particles larger than 5
microns and a high proportion of the particles smaller than 5 microns. The
cleaned and cooled moisture laden gas goes up thraugh demisters which
remove entrained water. The gases are then reheated from 120 F to about
163 F to (1) provide dry gas for protection of the ID fans, (2) reduce visual
pluming at the stack exit and (3) enhance buoyancy which increases the
effective stack height. The water containing the particulates overflows the
turbulent bed and is drained to a clarifier for particulate concentration.
The SC>2 is removed from the flue gas by chemical reaction with the
limestone additive in the marble bed contactor. The overall reaction can be
shown as:
CaCO? + H0O + SO9 *• CaSO, + CO9 + H9O
J £ £ j £• £
and continuing
2CaSO3 + O2-*-2CaSO4
The calcium sulfate and calcium sulfite are only slightly soluble in water and
are therefore mostly precipitated. The precipitated solids are carried with the
scrubber bed drain water along with the fly ash to the clarifier for concen-
tration prior to disposal in the ash storage area. The recirculating tank
is the source of the scrubber spray water and is made up of the fol-
lowing:
* Clarifier overflow
• Additive slurry
* Makeup water to replace evaporation and blowdown
A soot blower is located in the gas inlet duct to each module to clean the
scrubber at the dry gas — moist gas interface. Soot blowers also clean the
heat transfer surface of the reheater. Water washers are provided to clean the
demister section. Experience gained from operation of the scrubbers at Unit 4
of the Lawrence Station on The Kansas Power and Light Company system
indicates that this cleaning equipment will be adequate.
887
-------
Velocity of the gases through the bed of marbles must be maintained at
above 75 per cent of rated flow to create the required turbulence for
paniculate removal. This is accomplished by automatically removing scrubber
modules from service in groups of three as the load is reduced.
The modules are arranged in four groups of three as shown on FIGURE 5,
Plan View. Flue gas crossover ducts at the air heater outlet and ID fan inlet
permit flexibility of ID fan and module operation. The arrangement will
enhance the reliability of operation of the flue gas scrubber system.
The additive will be wet ground and sluiced to the recirculation tank. The
feed quantity will be controlled as a function of the amount of SO? in the
inlet to the scrubber. It is expected that 100 per cent of the stoichiometric
ratio will be required for 50 per cent SO^ removal. Dolomitic limestone (40
per cent MgCO-j — 50 per cent CaCO-j) is being considered as an additive
because it is available near the plant. High calcium limestone would have to
be shipped from areas remote from the plant. The MgCOj will react with
SC>2 to form MgSO? and MgSO^. The solubility of MgSO^ is very high,
which may complicate the ability of the scrubber to form a solid precipitate
of the 804 ion. This could result in a less desirable blowdown from the
scrubber for eventual discharge to the environment.
RESEARCH AND DEVELOPMENT
The scrubber components are well along in the development stage. The
scrubber system development, howe/er, is still in progress, with emphasis on
the following items:
• Reducing quantity of the scrubber system effluent
• Determining quality of the effluent and the effect on river
water quality
• Selecting type and determining amount of the additive
• Confirming system and component reliability
• Confirming scrubber performance with respect to the guar-
antee
-------
A task force comprising personnel from NSP, B&V, and Radian Corporation
is working closely with Combustion Engineering to solve these problems.
The highest priority is placed on minimizing the impact of the effluent with
respect to both quantity and quality. Tests are being conducted on the
Combustion Engineering pilot system in Windsor, Connecticut, to minimize
the blowdown from the scrubber system by using a seeded crystallizer tank
to precipitate calcium sulfate and calcium sulfite. The tests are being run
initially with a high calcium limestone additive. Further tests are planned
with dolomitic limestone and western coal fly ash. Radian Corporation is
supplying input with respect to reviewing bench studies and analyzing pilot
system scrubber chemistry. The design of the scrubber components and
scrubber system is under careful review to maximize reliability and perfor-
mance.
TABLE I and TABLE II summarize the important river water quality para-
meters that can be affected by the scrubber effluent. The total dissolved
solids would be slightly increased immediately downstream of the plant, but
this effect would be hardly noticeable at the Minneapolis water intake
approximately 35 miles downstream.
TABLE I
RIVER WATER ANALYSIS
AT SHERBURNE COUNTY GENERATING PLANT
Prior to Plant After Plant Goes
Operation Into Operation
River Flow, cfs 1,223* 3,322**
Dissolved Solids, mg/1
Ca 40.1 43.6 41.4
Mg 15.6 18.1 16.5
Na 4.8 4.8 4.8
M(HC03) 181.0 178.4 180.1
S04 10.0 30.2 17.4
Cl 3.6 3.6 3.6
NO^ 1.9 1.9 1.9
SiO^ 5.8 5.8 5.8
Total 262.8 286.4 271.5
Carbonate Hardness 148 146 148
Noncarbonate Hardness 16 37 24
* Flow exceeded 90 per cent of the time
* * Flow exceeded SO per cent of the time
889
-------
TABLE II
RIVER WATER ANALYSIS
AT RIVER INTAKES FOR MINNEAPOLIS AND ST. PAUL
Prior to Plant After Plant Goes
Operation Into Operation
River Flow cfs 1,729* 4,781**
Dissolved Solids, mg/1
Ca 46.9 49.4 47.8
Mg 15.4 17.1 16.0
Na 2.1 2.1 2.1
M(HC03) 207.4 205.6 206.7
S04 10.8 25.1 16.0
Cl 4.0 4.0 4.0
N03 0.7 0.7 0.7
Si02 8.6 8.6 8.6
Total 295.9 312.6 301.9
Carbonate Hardness 170 168 169
Noncarbonate Hardness 10 25 16
* Flow exceeded 90 per cent of the time
** Flow exceeded SO per cent of the time
Northern States Power Company is actively involved in air pollution control
planning. There are a number of areas that NSP believes require more
research and development in the near future. Much effort has been made to
develop systems that will provide SC>2 control and particulate removal. These
areas will require more attention. NOX emission is becoming a major concern,
and NO reduction systems will need to be developed.
The water effluent from scrubbing systems must be minimized or eliminated.
if there is an effluent, it must be of acceptable quality. It is not prudent to
exchange one problem for another. Much effort has been made with respect
to particulate removal and SO2 removal, but very little concern has been
given to the blowdown effluent from scrubbing systems.
The design and development of scrubbing systems must consider reliability of
operation since the scrubbing system is an integral part of the power plant.
Unless the scrubbing system is reliable, the plant itself cannot be reliable.
€90
-------
REFERENCES
California Institute of Technology, Environmental Quality Laboratory, People,
Power and Pollution: Environmental and Public Interest Aspects of Electric
Power Plant Siting, September 1, 1971. Report No. 1.
Black & Veatch, Northern States Power Company Environmental Report,
Sherburne County Generating Plant, May 24, 1971.
891
-------
LIST OF FIGURES
FIGURE 1 PLANT SITE LOCATION MAP
FIGURE 2 PLANT SITE VICINITY MAP
FIGURE 3 PLANT SITE ARRANGEMENT
FIGURE 4 FUNCTIONAL SCHEMATIC
FLUE GAS SCRUBBER MODULE
FIGURE 5 PLAN OF
FLUE GAS SCRUBBER
MODULES
892
-------
SHE\BUR|NE NATIONAL
T. -CtlOtm- - —V -(WILDLIFE REFUGf
PLANT SITE
-St+ERBttRNE COUNTY.,
GENERATING PLANT
PLANT SITE LOCATION MAP
5 0 5 10 15
SCALE MILES
INTERSTATE HIGHWAY
U.S. HIGHWAY
893
FIGURE I
-------
;-- - -|30 ^J
B ~-
- SHERBURNE COUNTY ^
GENERATING PLANT „,
EXISTING NUCLEAR
PLANT
PLANT SITE VICINITY MAP
SCALE - MILES
CONTOUR INTERVAL - 20 FEET
DATUM IS MEAN SEA LEVEL
894
FIGURE 2
-------
EFFtJpT
TREftTMT
FACILITY
PLANT SITE ARRANGEMENT
895
FIGURE 3
-------
O OD
CO CD
•< O
Z CO
O
- CO
I— •<
O CD
896
-------
ID FAN
AIR HEATER GAS OUTLETS
INLET DAMPER
(TYP)
STEAM GENERATOR
PLAN OF
FLUE GAS SCRUBBER
MODULES
897
FIGURE 5
-------
-------
ONTARIO HYDRO'S PROTOTYPE LIMESTONE SCRUBBER FOR
S02 REMOVAL FROM CLEAN FLUE GAS
J.W. James
Ontario Hydro
Toronto, Canada
Prepared for
Second International Lime/Limestone
Wet Scrubbing Symposium
New Orleans, Louisiana
November 8-12, 1971
899
-------
ONTARIO HYDRO'S PROTOTYPE LIMESTONE SCRUBBER FOR
S02 REMOVAL FROM CLEAN FLUE GAS - by J.W. JAMES
ABSTRACT
Ontario Hydro is currently designing a 30 MW demonstration
limestone scrubber and expects to install it at one of its
stations near Toronto.
The system, which is planned for initial operation in
1973, will use limestone slurry in a spray tower contactor to
scrub SO- from clean flue gas. The gas will be cleaned by an
existing 99% efficient electrostatic precipitator. This prototype
will be sufficiently large to allow confident upscaling to a full
size 150 MW module, and should be particularly applicable to
existing units fitted with fly ash collectors.
The design of the prototype is based on a system devised
by Ontario Hydro's Research Division and tested in a 4000 cfm
pilot plant with flue gas from a coal-fired boiler.
The reasoning involved in selection of this system is
presented, along with the considerations to be investigated
during the tests.
Presented at the Second International Lime/Limestone Wet Scrubbing
Symposium, New Orleans, November 11, 1971
900
-------
ONTARIO HYDRO'S DEMONSTRATION LIMESTONE
/AL OF S02 Ff
J. U. JAMES
SCRUBBER FOR REMOVAL OF S02 FROM CLEAN FLUE GAS
1. Most of us are now aware that the route to the development of
a successful S0? removal process leads through a beautiful garden
having many inviting and expensive pathways. Choice of the route
to follow is influenced by faith, hope, and much scientific
endeavour; and is fraught with concern of being led down the wrong
garden path. - Having had some exposure to this process, I would
like to speak briefly today on the recent choice at Ontario Hydro
in favour of a 30 MW scrubber to demonstrate the removal of S02
from one of its coal-fired boilers.
2. During the past few years, our top management have been
anxious to develop an S02 removal process, through work on a large
demonstration unit installed in one of our own plants. Many systems
have been analyzed with this objective in mind. For a variety of
reasons, all were rejected. Some were already being developed in
demonstration units, while others were either not suited to our
needs or appeared to hold little promise. During this period of
search, we entered into two multi-sponsor agreements to support
well known major Research and Development projects. However, it was
not until the completion of our Research Division's most encouraging
work on a 4000 cfm pilot plant, using limestone slurry in a spray
tower scrubber, that we had a viable candidate for a large demon-
stration system for development on one of our generating units.
3. The pilot plant and its performance is described in a
paper entitled "Sulphur Dioxide Removal by Limestone Slurry in
901
-------
a Spray Tower" by A\ Saleem, D. Harrison and N. Sekhar. This
paper was presented during Tuesday's session. The 4000 cfm pilot
plant achieved about 75% removal of S0? from clean flue gas which was
drawn off the ductwork following the electrostatic precipitator of
a 300 MW coal-fired unit. The spray tower operated with a high
degree of reliability; and inspection, following a 1000 hour
continuous run, indicated no hard scale formation, although some
surfaces did have a soft deposit.
4. In general, the large scrubber will be similar to the
pilot, and its design will be based on the optimized process
variables from the pilot work. Provision will be made to operate
at off-design conditions as indicated by the demonstration tests.
- Aerodynamic model tests may be necessary to establish the large
scrubber vessel shape. These would optimize vessel configuration
to accommodate structural and space considerations, while ensuring
a satisfactory gas velocity profile in the scrubber and demister
over a 25 to 100% gas flow range.
5. The choice of limestone slurry in a spray tower contactor
to scrub clean flue gas from 30 MW of generation was based on the
following reasoning:
a) the limestone slurry process appeared to have a better
chance of early development than other SCL removal processes.
b) the choice of a spray tower scrubber was based on the pilot
plant experience. Some of its advantages are:
- it holds promise of fairly high reliability. This
results from the simplicity of the tower, which has
902
-------
a minimum surface on which deposit can form or
collect. Also, deposit formations on the demister
and scrubber surfaces are expected to be sufficiently
soft to allow removal by slurry or water sprays.
- it is much less sensitive to gas flow turndown than
either fixed or moveable bed scrubbers. This is
particularly important on our generating system
because of the cycling demand on fossil units.
- it has a very low flue gas pressure-drop, and this may
allow it to be retrofitted to existing units with
minimum modification to the ID fans.
- its SCL removal efficiency is expected to be at least 70%
and limestone consumption is about 1.3 stochiometric.
This is considered to be a satisfactory performance if
high reliability is achieved.
c) It was decided to demonstrate SCL removal while scrubbing clean
flue gas since it reduced the chemical problems introduced by
fly ash. All of Ontario Hydro's existing fossil plants are
fitted with high efficiency electrostatic precipitators, thus
a clean gas system can be directly applied to its existing
plants. For new power plants, the environmental pressures for
dry disposal of fly ash may result in continued specification
of precipitators. Alternatively, substitution of a high efficiency
venturi scrubber for particulate removal is not expected to
adversely affect the spray tower operation.
d) The 30 MW size was selected to provide the fastest and most
economical means of developing a full-scale scrubber. It was
903
-------
assumed that full-scale scrubber modules are unlikely
to exceed 150 MW in size during the next few years.
Further, it was thought that a successfully operating
30 MW scrubber could be scaled-up to full size with
confidence, and applied to a number of operating
boilers. Alterations to the 30 MW unit during develop-
ment will require considerably less time and cost than
for a larger scrubber.
6. Areas to be investigated in the demonstration unit are those
common to most limestone slurry systems. Some of these are noted
below:
a) Scrubber
- The optimized gas velocities and L/G ratios derived in the
pilot module are to be verified.
- a major consideration is gas velocity distribution in the
scrubber and demister. Attempts will be made to measure and
optimize this at critical locations.
b) Demister
Test data are needed in at least three areas:
- The effects of velocity maldistribution and demister cleaning on
the amount of solid and liquid particles escaping the demister.
- The optimization of moisture removal versus pressure drop.
- The methods of on-loaid cleaning.
c) Reheater
Initially the unit will have a direct fired reheater. Upon
successful demonstration of the scrubber/demister, other methods
of reheating will be tried.
904
-------
d) Waste Slurry
This will be discharged to a divided settling pond where
the slurry is expected to decant to 80% solids. The liquor
will be recycled to the process and the solids disposed
off-site.
Items requiring investigation include:
- The control of dissolved constituents in the recycled liquor.
- The most economical means of dewataring the slurry on site.
Alternatives to the settling pond include a clarifyer, or a
vacuum filter.
- The environmental problems of the solids disposal.
e) Materials
Corrosion rates will be investigated for various materials in
contact with the liquid phase.
f) Assessment of process availability and reliability.
7. Regarding design and construction status, preliminary design of
the 30 MW unit is in progress and the commitment to final design and
construction is scheduled for next March. The unit is expected to be
commissioned and ready for testing by September 1973.
8. Ladies and Gentlemen, I am well aware that many plans for the removal
of S02 have come and gone. However, the plan presented here today is
based on considerable recent R&D, and is one in which we have great
confidence. I trust that at a future Symposium, we can present
demonstrated results of the successful operation of this 30 MW spray
tower scrubber.
905
-------
-------
LA CYGNE STATION AIR QUALITY SYSTEM
by
D. T. McPHEE
KANSAS CITY POWER & LIGHT COMPANY
KANSAS CITY, MISSOURI
presented at
SECOND INTERNATIONAL LIME/LIMESTONE WET SCRUBBER SYMPOSIUM
SHERATON-CHARLES HOTEL
NEW ORLEANS, LOUISIANA
November 8-12, 1971
907
-------
LA CYGNE STATION AIR QUALITY SYSTEM
La Cygne steam electric generating station is a 820 raw, 3500 psi
cyclone-fired fossil-fuel unit utilizing a local low grade coal and is
scheduled for commercial service in 1973. This is a joint venture be-
tween Kansas City Power & Light Company of Kansas City, Missouri, and
Kansas Gas and Electric Company of Wichita, Kansas. It is estimated to
cost between $180 and $190 million and each of the companies will have
a 50% interest in the plant. Kansas City Power & Light Company is the
operating agent and will operate the plant. The fuel will consist of
some 2 million tons per year of a low grade bituminous coal obtained
from strip mines within a few miles of the plant. Some of the charac-
teristics of this coal are 22% ash, 5-1/4% sulfur, 10,000 Btu per pound.
In 1968 when the two companies decided to construct this plant, the
decision was made that an adequate air quality system would be installed
to fulfill our environmental responsibilities.
This decision has been a bit difficult to carry out in that we have
been faced with a moving target in trying to identify our environmental
responsibility and as originally anticipated, have had to carry out a
crash program to develop a technology for removal of S02 as well as
particulate matter. I do not want to belabor the moving target part
of the problem, but bear in mind that it certainly has been and still
is substantial. With a vast amount of emotionalism, and a fair degree
of bureaucratic and political nonsense that invariably creeps into
important new broad base concerns of man, it has been a bit difficult
to accurately and realistically identify the environmental problem.
908
-------
However, regardless of this and regardless of the fact that a new tech-
nology must be hurriedly ushered onto the electrical generating scene,
Kansas City Power & Light Company and Kansas Gas and Electric Company
keenly feel that the new dimensions of our environmental responsibility
requires that we forge ahead with an adequate air quality system for
La Cygne.
A substantial development program was carried out preparatory to
the decision on the type of equipment. Some $300,000, mostly with
Ebasco Services, Inc. and Chemical Construction Company, went into
this effort over an 18-month period. We also took a look at the sulfur
and sulfuric acid market for the metropolitan area of Kansas City. The
results of this investigation were not favorable for a recovery type of
system. Our review of the status of technology also indicated that some
type of lime or limestone non-recovery type of system should be used.
Discussions were carried out with Chemical Construction Company, Babcock
& Wilcox and Combustion Engineering for providing a system that would
remove particulate matter and sulfur dioxide to insure ambient air
quality levels required by the Environmental Protection Agency's primary
and secondary standards.
Our final decision was to use Babcock & Wilcox equipment consisting
of a venturi scrubber for removal of particulate matter and an absorber
for sulfur dioxide removal. The schematic arrangement of this system is
as per attached Exhibits C and E. The first stage of the system is a
variable throat venturi scrubber for particulate removal. The second
stage is a packed type of absorber using hollow plastic spheres for the
packing material. Particulate matter is removed in the first stage by
909
-------
means of a water spray. Sulfur dioxide is removed in the second stage
of the absorber section by precipitating calcium sulfate and calcium
sulfite. The gas then passes through de-misters and steam reheat coils
where it is reheated 25 degrees. The gas stream from the seven identical
parallel modules is combined in a plenum from which it is discharged by
six induced draft fans, 7000 hp each, through a 700 ft chimney.
Some 500,000 tons of limestone is expected to be used per year.
This limestone, of approximately 927, calcium content, will come from
local quarries and will be delivered by the supplier to a limestone
hopper adjacent to the coal receiving hopper. It will be handled by
the plant coal handling system for the initial part of its route to
the limestone facility. Two full capacity 110-ton per hour wet ball
mills will be available to grind the limestone to a fineness permitting
907. passage through a 325 mesh screen.
Since the S02 removal system will produce low pH water and since
the limestone and the fly ash that accumulate in the slurry may have
abrasion characteristics, the material throughout the system has to
have special consideration. The venturi section will be made of #316
stainless steel and lined with 2" refractory material. The sump tanks
under the venturi and absorbers will be made of carbon steel with a #316
stainless sheath bonded to the inner surface. These tanks will also have
some refractory material lining in areas where abrasion may be a problem.
The absorbers will be made of 316 stainless including the wire mesh
baskets that contain the hollow plastic balls. The de-mister in the top
of the absorber will be made of fiberglass. The steam reheat coils will
be made of 5/8" diameter stainless steel tubing. The breeching from the
910
-------
reheater to the fans and to the stack will be carbon steel. It is expected
that no corrosion will take place in this area since the reheat of the
gas should prevent the condensation of moisture on the walls of the
breeching. The pumps and the piping in the recirculation system will
be rubber lined for both corrosion and abrasion protection.
The total cost of the air quality system is estimated at $32.5 million.
It is expected that 99% efficiency will be achieved for particulate matter
removal and 807» for removal of sulfur dioxide. The design of the system
is basically complete. Construction is underway and testing is scheduled
to begin in August and September of 1972. We are certain that there are
challenging times ahead for making this an effective system. We are
convinced that this will be accomplished; however, recognize that the
operating and maintenance costs will be substantial which, of course,
will have to be added to the electric rate payers' bill.
911
-------
GAS
OUTLET
FROM LIMESTONE
SLURRY STORAGE
VENTURI
RECIRCULATION
PUMP
ABSORBER
RECIRCULATION
PUMP
TO SETTLING POND
RECYCLE AND
MAKE-UP WATER
VENTURI-ABSORBER MODULE
912
Exhibit C
-------
X
UJ
s
UJ
fc
a
CO
GO
a
UJ
I
913
-------
-------
SULFUR DIOXIDE SCRUBBER SERVICE RECORD
UNION ELECTRIC COMPANY—ME RAMEC UNIT 2
J.P. McLaughlin, Jr.
Union Electric Company
St. Louis, Missouri
Prepared for
Second International Lime/Limestone
Wet Scrubbing Symposium
New Orleans, Louisiana
November 8-12, 1971
915
-------
SULFUR DIOXIDE SCRUBBER SERVICE RECORD
UNION ELECTRIC COMPANY—MERAMEC UNIT 2
The following summary shows total days during which
the precipitator was blanked off to direct the boiler flue
gas through the sulfur dioxide scrubber. The type operation
during each test period is indicated by showing the number
of days the unit was either out of service, firing gas, or
firing coal.
No Gas Firing Coal Firing Total
From To Load 50-125 MW 50-55MW 100-110 MW Days
9-9-68 10-5-68
11-11-68 12-5-68
2-15-69 3-2-69
6-16-69 6-21-69
10-3-69 10-10-69 —
11-24-69 12-22-69
2-16-70 3-25-70* 10 1/2
8-31-70 9-6-70
5-2-71 6-4-71
Total
*Unit was shut down temporarily from March 6 to March 17 but
not converted to precipitator operation.
9/27/71
J.F. McLaughlin, Jr,
916
4 17
5 1/2
8
1/2
— —
2
10 1/2
— —
15 1/4
30 32 3/4
—
11
6
1
3
6
6
2
3
41
1/2
1/2
1/2
1/4
1/2
3/4
4
9
1
3
3
20
20
4
14
79
1/2
1/2
1/4
1/4
1/2
25
25
16
5
7
28
37
6
33
183
1/2
3/4
1/4
-------
WILL COUNTY UNIT 1
LIMESTONE WET SCRUBBER
by
D. C. QIFPORD
COMMONWEALTH EDISON COMPANY
CHICAGO, ILLINOIS
presented at
SECOND INTERNATIONAL LIME/LIMESTONE WET SCRUBBER SYMPOSIUM
SHERATON-CHARLES HOTEL
NEW ORLEANS, LOUISIANA
NOVEMBER 8-12, 1971
917
-------
Commonwealth Edison in order to gain technical under-
standing and to determine the economics and feasibility of sulfur
dioxide removal, embarked on the installation of two separate and
different facilities for the removal of sulfur dioxide from boiler
flue gas.
One system is a pilot plant at our State Line Station
that will produce elemental sulfur. This is a joint research
project with Universal Oil Products of their sulfoxel process.
The second system is a full size limestone wet scrubber
that will remove particulate and sulfur dioxide at our Will County
Station, Unit 1. This is the system I will discuss. I will cover
a progress status report as well as the technical aspects of this
system.
In January of 1970, the existing electrostatic precipita-
tor was found inadequate to meet the existing particulate emission
standards.
Accordingly, in the spring of 1970, we contracted with
Bechtel Corporation to investigate the sulfur removal systems
available and to recommend a system that had the greatest chance
of success.
Bechtel recommended a wet scrubber system using limestone
or lime. A specification was then prepared by Bechtel and released
for bid. Of the nine bidders that were solicited, only seven
proposals were received. After detail study and bid evaluation
with consideration of the project schedule, Babcock and Wilcox
was given authorization to begin the detail engineering in Septem-
ber, 1970. A formal purchase order was issued in November, 1970,
with a project completion deadline of December 31, 1971- This
completion date was established by the Illinois Commerce Commission
as part of a recent rate case.
The Babcock and Wilcox designed process is guaranteed to
remove 98$ of the fly ash and 7-6$ of the sulfur dioxide, but is
anticipated to remove 99$ and 83$ respectively. These efficiencies
are based on a dus't inlet loading of 1.355 grains per standard
cubic foot at 70 degrees P. and burning lj.$ Illinois sulfur coal.
In considering scrubbers, the pressure drop across a scrubber
with the same dust removal capability differs greatly between a
cyclone boiler with its smaller dust sizing and a p ilverized fuel
boiler with its larger dust sizing.
The Will County wet scrubber is being backfitted on a
163 net megawatt Babcock and Wilcox radiant cyclone boiler that
was put in service in 1955*
The wet scrubber is indicated on this property plat
of Will County Station.
918
-------
The wet scrubber, like gaul, is divided into three
parts; a limestone milling system, the wet scrubber, and the
sludge disposal area.
The milling system as shown on slide 2 consists of a
limestone conveyor, two 260 ton capacity limestone storage silos,
two full sized Allis Chalmers wet ball mills, and a slurry storage
tank. Each silo when full can supply the wet scrubber for 2l|.
hours of operation. The limestone required should be high in
calcium carbonate, above 97^« It should be noted that the reactivity
of the limestone is not necessarily related to the chemical analysis
of the limestone.
Each wet, ball type mill Isdesigned to pulverize 12
tons of limestone per hour so that 95>^ will pass through a 325
mesh screen. The mill output product is in the form of a water
slurry with 20fo solids. The slurry is piped to the 1± hour capacity,
62,^00 gallon slurry storage tank where it is pumped to the wet
scrubber system.
The wet scrubber system Is made up of two identical sys-
tems each taking half the boiler flue gas. Each system consists
of two recirculation tanks, slurry recirculation pumps, a Venturi
fly ash scrubber, a sump, a sulfur dioxide absorber, flue gas
reheater, and ID booster fan.
For clarity I have broken the wet scrubber system into
two subsystems, a gas system and a slurry system.
Slide 3 shows the flue gas path. Flue gas passes from
the boiler after the precipitator and goes to the Venturi. Here
the gas is forced through a pressure spray of water coming from
nozzles on each side of the venturi.
The gas pressure drop through the venturi xa 9 inches
of water. The removal of fly ash is effected by the collision of
the particles with small water droplets (the ability to collect fly
ash is a function of water droplet size).
From the venturi the gas turns through the s unp and
then upwards into the absorber. Here the sulfur dioxide is removed
as the gas at greatly reduced velocity is forced through two
separate stages of plastic spheres. These spheres, coated with
limestone slurry provide a wetting surface for the chemical reaction.
They also act as cleaners to prevent buildup of solids. The
abosrber outlet has a chevron type demister. The gas pressure
drop through the absorber is 6 inches of water. Space for a third
stage of plastic spheres is available if found necessary.
From the abosrber the flue gas is reheated from 128
degrees F to 200 degrees F to give the gas buoyancy and to limit
condensation in the fans, ducts and existing steel brick lined
stack. The bare tube reheater is divided into three sections, the
919
-------
first Is made of 30lf stainless steel, the other two sections are
corten steel. Each reheater has four sootblowers to maintain
tube cleanliness.
To compensate for the draft loss across the wet scrubber,
all ID booster fan Is used that discharges to the existing boiler
ID fan. It is intended that the new booster fan maintain a zero
differential across the bypass damper and that the existing ID
fan will continue to control the furnace pressure.
Slide If shows the slurry recirculation system. There
are three venturi recirculation pumps and four absorber recircula-
tion pumps connected to their own header. Normal operation will
be with each venturi and absorber system isolated from each other,
but the flexibility is there to enable online pump maintenance.
The sump is designed so that there is little or no mixing of the
venturi and absorber recirculation flows. The fresh limestone
slurry at 20$ solids is added to the absorber recirculation tank
where water dilutes the slurry to 8$ solids. The spent or waste
slurry is taken off the venturi pump discharge line.
Tank level differences are compensated by an inter-
connection between the two tanks. Each tank holds lj.2,000 gallons
and provides a reaction hold up time or four minutes in the
absorber recirculation tanks and six minutes in the venturi
recirculation tanks. Space Is available to add two more tanks
to increase the hold up time to six minutes for each absorber
system if the system performance characteristics dictate.
The flow of slurry to each venturi is 5800 gallons per
minute and to the absorber is P-750 gallons per minute. This
gives a liquid to gas ratio of l8.1j. to 1 and 28 to 1 respectively.
The variable throat venturi and the three sections in
the absorber allows a load range from 30 to 100$ boiler load.
The waste slurry Is pumped to a settling pond and all
the water runoff is recycled to the we.t scrubber and milling
system. With our pond arrangement there could be a limited
blowdown but Illinois law allows a dissolved solids limit on
water discharges to the canal of only 750 parts per million. With.
total recycle the dissolved solids concentration Is expected to be
2500 parts per million. To compound the difficulty of any blow-
down, the canal water used for makeup has a concentration of 625
parts per million.
Slide 5 is a sketch of the entire system showing the
milling system, Venturis, absorbers, recirculation tanks, pumps,
re heaters, fans, and duct arrangement.
The materials used for the construction of the system-are:
Flues from the boiler to the venturi, carbon steel
The venturi, cart-on steel with plasite 7122 and two
inch kaocrete
920
-------
The sump, corten steel with flake line 103 'and two
Inch kaocrete
The sump bottom, lined with firebrick
The absorber, rubber lined corten steel
Flues from absorber to re heater, corten steel with
flakeline 103
Flues from reheater to the ID booster fan, corten steel
ID booster fan, corten steel housing, carbon steel wheel
The pumps, recirculatlon tanks, valves, and all piping
in contact with slurry above six Inches in diameter is
rubber lined carbon steel
Any piping less than six inches in diameter is 316 L
stainless steel
The power requirement for the entire wet scrubber system
is nine megawatts or 5.1$ of the unit gross capacity of 177
negawatts. This is nearly equivalent to the auxiliary power
consumed by the rest of the unit. The eleven largest power con-
sumers are:
Two ID booster fans, 22^0 HP each
Two limestone mills, lj.00 HP each
Four absorber recirculation pumps, 200 HP each
Three venturi recirculation pumps, 350 HP each
The controls for the wet scrubber system are all located
on a ten foot long central control board that has approximately
as many Instruments as the present boiler board. This new control
board will allow complete start up, operation, and shut down of
the mill and wet scrubber system remotely.
Slide 6 shows that the estimated cost for the system is
in excess of $8 million, with equipment about $1|,750,000, erection
about $2,600,000, and professional engineering about $7^0,000.
This amounts to %9 per net kilowatt hour.
The limestone cost delivered is about $5.00 per ton.
Full load operation requires 1$ tons per hour or 130,000 tons
per year.
921
-------
With sludge production anticipated at 19 tons per hour
at full load, it will be necessary to dispose of 166,000 tons per
year. The cost to get rid of the sludge will approach $5»00 per
ton. This cost per ton would take the sludge from the pond and
convert it from toothpaste consistancy to a solid, stable, non-
reverting material.
The above costs related in cents per million BTU's are
as follows:
Carrying charges on $8.1 million for 15 years, 11.5
Limestone, 5«0
Sludge disposal, 6.5
Manpower (one shift position), 1.0
Auxiliary power, 2.0
for a total of 26 cents per million BTUs. This total does not
include maintenance or property tax.
So far I have just talked about the equipment and design.
Construction presents a great many probelms both physically and
schedule wise. Slide 7 shows how it was necessary to sandwich
the scrubber between the boiler house and service building with a
substantial canteliver. Also shown on the slide is the complexity
of the duct arrangement.
Due to foundation problems the equipment erection did not
start until mid May, 1971. Judicious use of overtime will allow the
erection of equipment to be completed by the end of this December.
To conclude my presentation, here are some slides that
I took at various stages of construction.
922
-------
c c
s s
ooo
,1
0)3
•i
923
-------
Figure 2
.MILLING SYSTEM
eclaim
bpper
Slurry
Storage
Tank
I
Recycle
Tank &
Pumps
To Wet
Scrubbe
924
-------
Figure '•
FLUE GAS PATH
Boiler
Stack
Existing
ID Fan
Byp iss
Dam >er
(TF-J
\
\
Electro).
Precip
ID
Booster
\
Rdheater
^
Deznister
Absorber
QQOQQOQOQQ
Sump
925
-------
Pi gun
SLURRY RECIBCTJLATIOK SYSTEM
To Sludge
Waste Pond
Absorber
Sump
Venturi
ecirculatlon
Tank
I
Absorber
Rfecirculation
Tank
Venturi Pumps
From ^
Systt
Absorber Pumps
926
-------
E
o .t:
c
o '
M C
=5 o T
UJ ~ U
oo.
££ <
o •£
O ^
927
-------
Estimated Costs
Wet Scrubber System
Investment
Equipment, buildings and foundations
Erection
Professional Engineering
Figure 6
$4,750,000
2,600,000
750,000
$8,100,000
Operating
Carrying charges on $8,100,000 for 15 yrs.
Limestone at $5.00/ton (130,000 tons)
Sludge disposal at $5.00/ton (166,000 tons)
One shift position
Auxiliary power
11.50/MBTU
5.0^/MBTU
6.50/MBTU
1.00/MBTU
2.00/MBTU
*26.0
*Note: This does not include maintenance or property tax.
928
-------
929
-------
-------
Chemical Construction Corporation
Pollution Control Division
A SUMMARY REPORT - CHEMICO'S COMMERCIAL SYSTEMS INSTALLATIONS
AT ELECTRIC POWER GENERATING STATIONS
H. P. WILLETT - Vice President
I. S. SHAH - Chief, Process Engineering and Development
Pollution Control Division
Chemical Construction Corporation
320 Park Ave. , New York, N. Y. 10022
Presented At
2nd International Lime-Limestone
Wet Scrubbing Symposium
New Orleans, Louisiana
November 8 - 12, 1971
931
-------
unemicai construction Corporation
Pollution Control Division
A Summary Report of Chemico's Commercial Sysfrejtru3*4nstallations
At Electric Power Generating Stations
H. P. Willett, Vice President Pollution Control Division
I. S. Shah, Chief, Process Engineering And Development
Chemico's extensive research and development work. Bench scale, pilot plant
cale and prototype scale, is aimed at developing capabilities to offer suitable
olutions - both technically and economically feasible - for pollution problems of the
tility industry. The major pollution problems of the utility industry are emissions of
.y ash, sulfur dioxide (SC^) and nitrogen oxides (NOX). Chemico has developed and
3 continuing to develop technology for:
(a) Fly Ash Removal
(b) Simultaneous removal of Fly ash and SC>2 using lime -
•limestone throw away processes
(c) SC>2 recovery using Magnesium base SO2 recovery process, to
produce saleable products.
i this presentation, we would like to summarize the various projects, one in oper-
•J.on andthe others unaer construction, in the utility industry, and briefly describe the
nportant features for each installation. The various projects are summarized in
able I.
ly Ash Removal - Holtwood Station of Pennsylvania Power and Light Company
oiler No. 17 is a pulverized coal fired balanced draft boiler having generating capacity
' 72 MW. The coal used is Anthracite, dredged from the river basin, having an ash co
17 - 20% and a sulfur content of 0. 5-0. 6%. The boiler was originally equipped with
932
-------
Chemical Construction Corporation
Pollution Control Division
Fly Ash Removal - Holtwood Station of Pennsylvania Power and Light Company (Cont'd
mechanical collectors and electrostatic precipitator for dust collection. The scrub-
ber system is designed to handle 358,000 ACFM of flue gas leaving the air heater at
360°F and -8" WG. This represents 80% of the total flue gas. The balance, 20%
of the gas, flows through the precipitator and partially blanked mechanical collector
section. The hot gases leaving the precipitator, mixes with saturated gas leaving the
scrubber, thus providing a reheat of approximately 35-40 F. The mixed reheated
gas is exhausted to the atmosphere through existing I.D. Fans and stack.
The liquor system consists of a thickener, recycle pump tank, and neutralization
tank (to neutralize thickener underflow with lime). The underflow after neutraliza-
tion is then sent to an ash pond through existing fly ash disposal pumps.
The scrubber system is designed to reduce the outlet dust loading to 0. 04 grains/SCF
dry when the inlet dust loading is 4.5 grains/SCF day or less. In case the inlet
dust loading is higher than 4.5 grains/SCF day, the scrubber system will provide
99% efficiency.
"'he scrubber system is successfully meeting the guaranteed
performance, even though the ash content of coal has increased from the design value o
17% (4. 5 gr/SCFD) to 38% (10 gr/SCFD).
Fly Ash Removal-Four Corners Station - Arizona Public Service Company
Jnits No. 1, 2 and 3 are pulverized coal fired balanced draft boilers having generating
933
-------
Chemical Construction Corporation
Pollution Control Division
Fly Ash Removal - Four Corner Station - Arizona Public Service Company(Conttd)
capacity of 175 MW, 175 MW and 225 MW respectively. The coal used contains 28. «°,
ash and 0. 6% sulfur. The boilers were originally equipped with mechanical collecto
Each of the three boilers will be equipped with two scrubbers, I. D. Fans, (wet)
mist eliminators, and reheaters. Each scrubber handles a flue gas volume of
407, 000 ACFM at 340°F and -10" W. G. , in the case of units 1 and 2, and a flue gas
volume of 515, 000 ACFM at 340°F and -10" W. G., in the case of unit 3. The flue ga
from units 1 and 2 are discharged to the atmosphere through one common existing st
whereas unit 3 has its own existing stack. Adjustable throat mechanism are provide
to maintain constant dust removal efficiency at varying loads.
A common liquor system, for all three boilers, consists of thickeners, pump tanks
and existing ash pond. The thickener overflow, and liquor from ash pond are re-
turned to scrubber system.
The scrubber system is designed to provide an outlet dust loading of 0. 4 grains/SCF
with an inlet dust loading of 12 grains/SCFdry. 'The flue gas is reheated by
20°Ftoavoid condensation of water vapor in t h e high velocity stack.
The system will be in operation before the end of this year.
Fly Ash Removal - Dave Johnson Station, Unit 4, Pacific Power and Light Company
Unit no . 4 is a new pulverized coal fired balanced draft boiler having generating
capacity of 360 MW. The coal fired has an ash content of 16% and sulfur cor
934
-------
lical Construction Corporation
Pollution Control Division
r Ash Removal - Dave Johnson Station, Unit 4, Pacific, Power and Light Company(Ccn)
%. The total flue gas of 1, 487,100 ACFM at 270°F and -12" W. G, and having
t loading of 12 grains /SCF dry, leaving the air heater is handled by three
•ubbers, three wet I. D. Fans, and one common low velocity wet stack. The
urated flue gas is not reheated. The bleed liquor from the scrubber system
Dumped to an ash plant, and cleared liquor from the plant is returned to the scrubbers.
'ustable throat mechanisms are provided in each scrubber, to maintain constant
:ssure drop to achieve constant dust removal efficiency at varying boiler loads.
3 scrubber system is designed to provide outlet dust loadings of 0. 04 grains/SCFdry
•vided the inlet dust loading is 12. 0 grains/SCFdry or less. The plant is under con-
uct ion and scheduled for start-up early in 1972.
2 Recovery - Mystic Station, No. 6 Unit, Boston Edison Company, and Essex
smical Company's Acid Facility at Rum ford, Rhode Island
'.t No. 6, rated at 155 MW generating capacity is equipped with air heaters, elec-
static precipitator (de-energized when burning oil fuel), two induced draft fans
a stack. At 155 MW rating, the flue gas volume leaving the I. D. Fans is
, 000 ACFM at SOOOp and +1" W. G. The SC-2 loading is 1410 ppm (by volume,
gas basis). The flue gas leaving the two I. D. Fans enter the two new F. D. Fans,
one Venturi Type SO2 absorber. SC>2 is absorbed by MgO slurry in the absorber,
ning a slurry of MgSOs, MgSO4 and unreacted MgO. The bleed from absorber is
t to a centrifuge, to produce a cake containing approximately 5% surface moisture.
935
-------
hemical Construction Corporation
Pollution Control Division
SC>2 Recovery - Mystic Station, No. 6 Unit, Boston Edison Company, and Essex
Chemical Company's Acid Facility at Rumford, Rhode Island (Cont'd)
The centrifuged cake is then dried in a drier by removing both crystalline and
surface water. The dry product is stored in an existing fly ash silo and then trucked
away to Essex Chemicals sulfuric acid facility at Rumford, Rhode Island. The dry
product containing MgSC^, MgSO4 and MgO is calcined in a calciner, to produce
SC>2 rich (12 - 16%) flue gas, and regenerate MgO. The flue gas after proper cleaning.
enters the sulfuric acid plant to produce 98% H2SO4 acid.
The regenerated MgO is returned by truck to Mystic Station of Boston Edison, for '
reuse in the absorber, Make up MgO will be added to the system at Boston
Edison.
The SO2 recovery process plant will reduce the inlet SO2 concentration of 1410 ppm
'n the boiler flue gas to less than 150 ppm which is equivalent to burning less than
). 3 percent sulfur content fuel oil.
Approximately 50 tons/day of crystal MgSO3, MgSO4 and MgO will be produced
it the power plant. The recovered sulfur dioxide from the power plant stack flue gas
vill be equivalent to the entire feed requirement of the 50 tons/day sulfuric acid plant
it Rumford, Rhode Island. The crystals will be shipped by truck only five days a
veek and only during the 8 hour day shift. At the acid plant, approximately 20 tons/day
)f MgO will be regenerated.
936
-------
cal Construction Corporation
ollution Control Division
Recovery - Mystic Station, No. 6 Unit, Boston Edison Company, and Essex
mical Company's Acid Facility at Rumford, Rhode Island (Cont'd)
plants at Mystic Station and Essex Chemical are under construction and
eduled for start-up before the end of 1971.
Ash and SC"2 Removal - Phillips Station of Duquesne .Light Company
.he Phillips station, there are 6 boilers having a total generating capacity of
MW. Each of the boiler^ is a pulverized coal fired balanced draft unit, and
resently equipped with mechanical collector-precipitator and separate stack.
coal burnt has 21% ash and 2. 3% sulfur.
• flue gas leaving the air heater from each boiler enters a manifold. The total
gas in the manifold is 2,190, 000 ACFM at 362OF and -22" W. G. , and the dust
ding and SO2 loading are 5. 9 Grains/SCFdry and 1370 ppm. This total volume
lue gas is handled by four scrubbing trains, each consisting of first stage scrub-
, wet I. D. Fan, and a mist eliminator.
ommon reheater and common stack are provided for all the scrubbing trains. One
ubbing train also includes a second stage absorber in place of the mist eliminator.
lis two stage scrubber-Absorber, simultaneous fly ash and SO. will
amoved using lime as absorbing agent. The other trains will remove only fly
using water as scrubbing liquor. The bleed from each scrubber is sent to an
pond, and if found necessary, additional lime will be used to neutralize the
937
-------
Chemical Construction Corporation
Pollution Control Division
Fly Ash and SO2 Removal - Phillips Station of Duquesne Power and- Light Company(O
total bleed. The clear liquor from the ash pond is returned to scrubber,
To maintain pressure drop across the scrubber throat, to attain constant
dust removal efficiency at varying load conditions, the venturi scrubbers are provide
with adjustable throats. As the load decreases, the throat area is reduced by clos-
ing the throat, and as the load increases the throat area is increased
The system is designed to provide an outlet dust loading of 0. 04 grains/SCFdry, whet
the inlet dust loading does not exceed 5. 9 grains/SCFdry. The outlet SC>2 concen-
tration from one scrubber train will be 274 ppm or less. Design and engineering
work is in progress, and the plant is scheduled for start up during February 1973.
Fly Ash Removal - Elrama Station of Duquesne Light Company
At the Elrama station, there are 4 boilers, having a total generating capacity of
494 MW. All the boilers are pulverized coal fired balanced draft units, and are
presently equipped with mehcanical collector-precipitator. The coal is similar
to that used at Phillips Station.
The flue gas leaving the air-heater from each boiler enters a manifold. Tne total
flue gas in the manifold is 2, 211, 000 ACFM at 304°F, and -22" W. G. The inlet
dust loading and SO2 concentration are 7. 32 grains/SCFdry and 1570 ppm. Five
938
-------
wuiiau ui/uuii
'Dilution Control Division
Ash Removal - Elrama Station of Duquesne Power and Light Company (Cont'd)
ubbing trains will handle the total gas flow, and each train consists of a
ubber, wet I. D. Fan, and mist elimi nator. A common reheater and common
• stack are provided for all the scrubbing trains. All scrubbers are provided
i adjustable throats to maintain efficiency at varying loads. At this plant,
ially only fly ash will be removed using water as scrubbing liquor. The bleed
TL scrubbing system is sent to the ash pond, and clear liquor from pond is re-
led to scrubber. The system is designed to provide an outlet duct loading of
74 grains /SCF dry, provided the inlet dust loading is 7. 32 grains /SCFdry or
3. The plant is scheduled for start up during February 1973.
kerson Station - Potomac Electric Company
s unit no. 3 of the Dickerson Station is a pulverized coal fired balanced draft
er rated at a generating capacity of 195 MW. The coal fired has an ash content
3% and a sulfur content of 3. 0%. The boiler is presently equipped with mechani-
collector-precipitator system for dust removal.
ty percent of the total gas flow will be treated in the prototype scrubber-absorber
tern, where the fly ash is removed in the first stage using water as scrubbing liquor
1 cleaned flue gas then enters the absorber where SO2 is removed using MgO
~ry as absorbing liquor. The slurry of MgSOs, MgSO4 and MgO is centrifuged,
cake is dried in the drier. The dry crystals are then trucked to an acid faci-
of Essex chemicals, where upon calcination, MgO is regenerated, and flue gas
939
-------
Pollution Control Division
Dickerson Station - Potomac Electric Company (Cont'd)
rich in SC>2 is produced to make 98% H2SO4 acid. Approximately 50 tons/day
of acid will be produced.
The ductwork is so arranged that the flue gas either before the precipitator or
after the precipitator can be withdrawn. The flue gas volume will be 295, 000
ACFM at 259°F and -11. 0"W. G. , and the dust loading and SO2 concentration are
5. 95 Gr/SCF dry and 1850 ppm respectively. The system will provide an outlet
dust loading of 0. 03 Gr/SCF dry, and an outlet SO concentration of 185 ppm or les
LJ
A n adjustable throat mechanism will be provided to maintain constant efficiency
at varying loads. Approximately 15°F reheat will be provided using off gases
from the dryer. At present, engineering and design is in progress and the plant is
scheduled for start up in 1973.
Fly Ash and SO2 Removal At a Power Generating Station In Japan
A 155 MW generating station burning pulverized coal having 20% ash and 3% sulfur,
in a balanced draft boiler is presently equipped with mechanical collectors-pre-
cipitator for dust removal. A flue gas volume of 451, 000 ACFM at 277OF and 0"
W. G. , .having a dust loading of 0. 25 Gr/SCF dry, and SO2 concentration of 2350.
ppm will be handled in one two stage scrubber train. The gas volume represents
75% of the total flue gas. A 5QOF reheat will be provided using fuel oil burners.
940
-------
cal Construction Corporation
'Dilution Control Division
Ash and SO2 Removal At a Power Generating Station In Japan (Cont'd)
"bide sludge will be used for simultaneous SC>2 and fly ash removal in a two
?e venturi scrubber system, including a delay tank. The bleed from the
ubber system will be pumped to the ash pond, and clear liquor from the ash
d returned to the scrubber system.
j scrubber system which handles flue gas leaving the existing pracipitator will
vide 90%-fSO2 removal and an outlet dust loading of 0. 025 grains/SCFD. The
it is under construction and is scheduled for start up during March 1972. This
it is designed for initial disposal consumed alkali, but for future conversion to
lufacture gypsum.
imary
>mico designed systems to date include four plants for fly ash removal using
er as scrubbing liquor, one plant each for simultaneous removal of fly ash
SC>2 using lime slurry and carbide sludge, and two SO recovery plants using MgO
~ry process, one each for coal fired and oil fired generating stations. Chemico
igned systems will handle a total flue gas volume of 10, 090, 100 ACFM resulting
n a total generation capacity of 2293 MW. Chemico systems designed to date
. remove 4320 Tons/day of fly ash and 400 Tons/ day of SO2, from flue gases,
thus help reduce air pollution and clean the air.
941
-------
O O O ^
- >J I—I
CO ft i r-
steam Adjustab'
in efficiency at
et I D Fans
locity wet stack
djustable throats
00 to 3 01 <
C 4-1 >
*4 C.
3 n w o w
6 T3 — i C
4J (0 10
(0 O O - U,
0) in 00 01 O
1-. 4-1 C .C
TO -. (U h-l
(n O >> 1-
CO 1- M *J
0 SI to 0 4i
Z. 3
o o
00 >*-i u"i C
C >w ^- O =0
BJ h O
11 X 00 ^
C ti E u
1* T3 (D b-
00 O
00 4J >
TJ C O ^
01 rt -^ tn Q
H 3 u x
r- (0 > U TO
- QJ e u
O CN U -C
am c X
>, — CO 01
O - in ^~. a>
£ §S,5.S
0
C
CO
*J
OJ
oil burners
c
(TJ
-------
PROBLEMS RELATED TO SCALING IN LIME/LIMESTONE WET SCRUBBING
A.V. Slack, Chairman
Participants:
A.V. Slack and J.D. Hatfield
Bela M. Pabuss
Joan B. Berkowitz
J.R. Martin
Philip S» Lowell
943
-------
SUMMARY
PROBLEMS REIATED TO SCALING IN LIME/LIMESTONE WET SCRUBBING
Second International Lime/Limestone Wet Scrubbing Symposium
New Orleans, Louisiana
November 8-12, 1971
Participants: A. V. Slack, Chairman
B. M. Fabuss
J. Berkowitz
A. L. Plumley
J. R. Martin
P. S. Lowell
SUMMARY
One of the major problems in removing S02 from waste gases by
lime/limestone slurry scrubbing is scaling in the scrubber and other slurry
handling equipment. Although study of the problem dates back to work in
England in the 1930's, much remains to be learned about the problem--par-
ticularly in regard to the effect of differing conditions among the units
producing the S02-laden waste gas.
The mechanisms involved in scrubber scaling are complex, much
more so than for scaling of boilers and desalination equipment. Among the
design and operating factors to be considered are (l) amount of S02 absorbed
per unit of slurry passed through the scrubber, (2) content of calcium sulfite
and calcium sulfate crystals in the recirculated slurry, (3) degree of de-
supersaturation accomplished outside the scrubber before return of the slurry,
(k) pH levels at various points in the circuit, (5) scrubber design as related
to tendency of solids to settle out of the slurry onto surfaces, (6) scrubber
design as related to the scouring effect of the slurry, (7) degree of oxi-
dation in the scrubber, and (8) point of lime introduction into the circuit.
Scaling can result from deposition of calcium sulfate, calcium
sulfite, or calcium carbonate--although carbonate scaling is rare. Calcium
sulfate is the usual scaling species but calcium sulfite is often encountered,
particularly when lime is the absorbent.
The consensus is that lime gives more scaling than limestone, but
no conclusive reasons for the difference have been advanced. The generally
lower pH in limestone systems seems to be a factor, at least in regard to
sulfite scaling. At the present level of development, lime systems must
be operated with blowdown (dilution with water) or at low pH (less than the
stoichiometric amount of lime) to reduce scaling to an acceptable level.
944
-------
Closed-loop operation (no blowdown to watercourses) is feasible with
limestone but careful attention to operating conditions is essential to
avoid scaling. For adequate S02 removal, the slurry circulation rate
and slurry solids content must be much higher for limestone scrubbing
than for lime; since this should also reduce scaling, it may be that the
particular level of operating variables required for good S02 removal
with limestone is the reason for limestone superiority in regard to
scaling.
The driving force for scaling is supersaturation. The formula
developed in the early English work for avoiding scaling was limiting the
degree of supersaturation developed in the scrubber to a level low enough
to avoid crystallization on scrubber surfaces; homogeneous crystallization
on sulfite and sulfate crystals was promoted, however, by carrying a large
surface area of such crystals in the slurry. The supersaturation developed
in the scrubber was then released in delay tanks before return of the
slurry to the scrubber.
The limiting upper value in the range of supersaturation that can
be tolerated is that at which bulk nucleation takes place in the solution
even in the presence of seed crystals. In recent work it has been deter-
mined that this level (for calcium sulfate) is about l.J for the nonstoichio-
metric Ca++/S04~ ratio in the scrubber solution. (Supersaturation is defined
here as aCa aS04~ / KspCaS04> where a is activity and Ksp is the solubility
product constant.) The critical level for calcium sulfite has not yet been
determined.
Data from recent successful pilot plant operation (nonscaling)
indicates a supersaturation level of 1.19 at the scrubber outlet, well below
the l.J critical level. The value decreased to 1.02, near the saturation
level, in the delay tanks before return to the scrubber. Sulfite super-
saturation was much higher--8.15 at the scrubber outlet and 6.^9 returning
to the scrubber--indicating that the critical level is much higher than
for sulfate.
The effects of scouring by the slurry or mechanical accumulation
of carbonate or sulfite crystals on surfaces followed by knitting together
or converting to sulfate have not yet been adequately evaluated. The
composition of the scrubber surface does not seem significant from tests
so far, and little is known regarding the effect of additives or of ionic
strength.
Further study is needed on all phases of the scaling problem.
In the meantime, the best course appears to be (l) use of limestone rather
than lime, (2) high recirculation rate, (j) high solids content in slurry,
(k) adequate delay time, and (5) use of scrubbers of the spray or mobile-
bed type. Since both of these scrubber types have major drawbacks--low
mass transfer rate and excessive packing wear, respectively—it may be better
to use a very open type of fixed packing. Recent work with a set of wire
screens as packing has given excellent absorption, relatively low wear, and
no apparent scaling.
945
-------
-------
REMOVAL OF SULFUR DIOXIDE FROM STACK GASES
BY SCRUBBING WITH LIMESTONE SLURRY:
OPERATIONAL ASPECTS OF THE SCALING PROBLEM
By
A. V. Slack and J. D. Hatfield
Division of Chemical Development
Tennessee Valley Authority
Muscle Shoals, Alabama
Prepared for Presentation at
Second International Lime/Limestone Wet Scrubbing Symposium
Sponsored by the Environmental Protection Agency
New Orleans, Louisiana
November 8-12, 1971
947
-------
REMOVAL OF SULFUR DIOXIDE FROM STACK GASES
BY SCRUBBING WITH LIMESTONE SLURRY:
OPERATIONAL ASPECTS OF THE SCALING PROBLEM
By
A. V. Slack and J. D. Hatfield
Division of Chemical Development
Tennessee Valley Authority
Muscle Shoals, Alabama
ABSTRACT
Scaling in lime-limestone scrubbing for S02 removal is a very
complicated process; much more is involved than the simple crystallization
of calcium sulfate from solution in scaling of boilers and desalination
equipment.
The possible effects of the following process variables on scaling
are discussed.
*Degree of desupersaturation in the surge tank.
*Amount of S02 absorbed per unit volume of liquor recirculated.
*Use of limestone rather than lime as the absorbent.
"Various factors related to scouring, such as slurry rate, slurry
velocity, solids content of slurry, particle size of solids,
scrubber type (e.g., mobile-bed vs fixed packing), and impingement
angle between slurry and surface.
*Factors affecting tendency of solids to silt onto surfaces, in-
cluding surface roughness and presence of transverse surfaces
that catch solids.
*Point of lime introduction into the scrubber circuit.
'Presence of fly ash in slurry.
*Degree of oxidation in scrubber.
*Nature of scrubber surfaces (e.g., plastic vs steel) in regard to
strength of bond developed between crystal nucleus and surface.
948
-------
Use of additives that weaken the bond between crystal and
surface.
At the present state of the art, th° most effective measures for
avoiding scaling appear to be (l) adequate delay time for desupersaturation
in surge tank, (2) use of limestone rather than lime, (3) high recirculation
rate (needed anyway with limestone for good S02 removal), (U) elimination
of surfaces that collect solid particles from the slurry, (5) high solids
content in slurry, and (6) use of scrubbers designed to maximize the scouring
action.
949
-------
REMOVAL OF SULFUR DIOXIDE FROM STACK GASES
BY SCRUBBING WITH LIMESTONE SLURRY:
OPERATIONAL ASPECTS OF THE SCALING PROBLEM
By
A. V. Slack and J. D. Hatfield
Division of Chemical Development
Tennessee Valley Authority
Muscle Shoals, Alabama
Because of marketing problems, the power and smelter industries
are generally turning to throwaway processes—production of a waste solid—
as a means of coping with the S02 emission problem. Lime or limestone is
the preferred absorbent since only a very low cost material can be con-
sidered when there is no return from sale of product.
Since the use of limestone as sorbent in a dry system has been
generally unpromising, most of the current effort is centered on absorption
of the S02 by a slurry of lime or limestone. Although this has shown con-
siderable promise, and is the method that has been selected by most of the
power and smelter companies that are planning to install full-scale S02
removal facilities, there are some major technical problems that remain
unsolved. The main one of these is scaling, that is, growth of an adherent
crystalline deposit on scrubber surfaces that eventually causes shutdown
because of interference with gas or liquid flow.
In this paper, the effect of operating factors, both chemical
and physical, on the scaling problem will be reviewed. The material
presented is based mainly on small-scale and pilot plant studies carried
out at TVA.
The Basic Problem
The reaction of S02 with CaO or CaC03 in a scrubbing operation
produces mainly crystalline CaS03-0.5H20. There is some dissolved sulfite,
however; the equilibrium concentrations of the various species produced
vary with pH, which depends on whether CaO or CaC03 is the absorbent and
on whether a countercurrent or backmixed scrubber is used. For CaO and
a backmixed scrubber, which gives the highest pH, the principal dissolved
sulfite species are HS03~, S03", and CaS03(aq) with CaS03(aq) preponderant.
With CaC03 and a countercurrent scrubber, which gives the lowest pH, the
HS03~ concentration is higher and the S03~ lower.
If oxygen is present in the gas, as it is in most situations, there
will be some oxidation of dissolved sulfite during passage of the solution
through the scrubber. Part of the resulting sulfate remains in solution, the
amount depending on several factors. The remainder crystallizes as CaS04-2H20,
which in most cases is the species that causes scaling—by crystallizing on
equipment surfaces.
950
-------
Solubility data for sulfite and sulfate are given in Table I.
Of more importance in regard to scaling, however, is the tendency of both
calcium sulfate and sulfite to supersaturate. Lessing2, in development
of the ICI-Howden process, found that CaS04-2H20 would supersaturate, in
a simulated solution, by about four times the saturation concentration.
In recent TVA pilot plant tests, the indicated degree of supersaturation
at the scrubber outlet has averaged about 1.8 to 1.9 times saturation
concentration. The degree of calcium sulfite supersaturation has appeared
to be even higher; these results will be checked in further tests.
TABLE I
Effect of pH on Solubility3 in the System
CaO-SOg-SOa-HgO at 50°C (l22°F)
Parts per million
pH Ca S02P
7-0
6.0
5.0
4-5
4.0
3-5
3.0
2-5
6?5
680
731
841
1,120
1,763
3,135
5,873
23
51
302
785
1,873
4,198
9,375
21,999
1,320
1,314
1,260
1,179
1,072
980
918
873
a Solution saturated with CaS03'0-5H20
and CaS04-2H20.
Sulfite.
c Sulfate.
The basic operational factors involved in scaling are illustrated
by Figure 1. The flowsheet shown is for a countercurrent scrubber and for
CaO or CaC03 introduction into the recirculation tank. Variations from this
include (l) backmixed or cocurrent scrubbing and (2) injection of limestone
into the boiler, in which case the resulting CaO enters the scrubber with
the gas.
1 Slack, A. V., Falkenberry, H. L., and Harrington, R. E. "Sulfur Oxide
Removal from Waste Gases: Lime-Limestone Scrubbing Technology." Paper
presented at 70th National Meeting, American Institute of Chemical Engineers,
2 Atlantic City, New Jersey, August 29-September 1, 1971.
Lessing, R. J. Soc. Chem. Ind. 5J_, 373-88 (Nov. 1938).
951
-------
I
TO STACK
GAS FROM
BOILER
SCRUBBER
MAKE UP H20
CoO OR
Cocoa
B
RECIRCULATION
TANK
THICKENER
OR POND
OVERFL'
TO WATf
COURS
-*. WET SOLIDS
FIGURE 1
Flow System Involved in Scaling
Operation of the system will be discussed in terms of the slurry
composition at points A, B, and C in the recirculation loop. At A the slurry
contains the solid species CaS03-1/2H20, CaS04-2H£0, and CaC03 (or CaO).
The liquid phase hopefully is unsaturated with sulfite-sulfate species because
of the water addition just before A and insufficient time for solid sulfite
and sulfate to resaturate the solution.
In the scrubber, S02 is absorbed and forms various dissolved sulfite
species. At the inlet pH involved in limestone scrubbing (about 6.0), the
main species present, HS03~, is in equilibrium with a much smaller amount
of S03=. The pH decreases as the solution flows down through the counter-
current scrubber, thereby bringing Ca++ into solution. This causes the
solubility product of Ca++ and S03- to be exceeded (under normal scrubbing
952
-------
conditions) with the result that a driving force for CaS03-0.5H20 cry-
stallization is developed. Crystallization does not necessarily occur,
however, because the CaSOo/0. 5HpO tends to supersaturate. Even a mild
tendency to supersaturation may have a major effect because the rapid
passage of the solution through the scrubber leaves little time for
nucleation and crystal growth.
Even if crystallization takes place it can occur on the surface
of calcium sulfite crystals already present in the slurry (homogeneous
crystallization) or on other solids (CaS04-2H20, CaC03, fly ash; hetero-
geneous crystallization). If the driving force for crystallization becomes
high enough, however, the capacity of these mechanisms to hold sulfite in
a harmless form will be exceeded and crystallization on scrubber surfaces
will occur. This is usually expressed as "critical degree of supersaturation,"
that which must not be exceeded if scaling is to be avoided.
The decrease in pH as the solution flows through a countercurrent
scrubber increases the amount of total sulfite species present at saturation.
Hence the solution can hold more sulfite in the lower part of the scrubber
without exceeding the critical degree of supersaturation.
Calcium sulfate also tends to supersaturate in the scrubber and
also has a critical degree of supersaturation. However, since its solubility
decreases with pH (when the solution is saturated with sulfite) there is no
advantage from the pH drop in the scrubber.
At point B the solution is supersaturated with both calcium sulfite
and sulfate. In the recirculation tank the pH rise from CaO or CaC03 addition
reduces sulfite solubility and promotes desupersaturation. The retention
time in the tank and in the thickener (or pond) also aids in desupersaturation
by allowing time for crystallization to take place. The objective is to
have the solution at or near saturation at point C. Although unsaturation
would obviously be desirable, such a goal seems impracticable except by water
addition close to the scrubber inlet. Unsaturation without water addition
could be achieved only by removing all the sulfite and sulfate crystals and
then crystallizing further by usual crystallization techniques; there are
several process and economic considerations that make this course undesirable.
Lessing proposed the following equation for rate of scaling.
K = (G! - C)/(C - C2)
pt
where C = concentration of CaS04'2H20 at time t
G! and C2 = initial and final CaS04-2H20 concentrations
p = amount of sulfate crystals
The value of K, about 1.5 in the ICI work, depends somewhat on the type of
sulfate crystals; small, thin ones have more surface area and therefore are
more effective in dissipating supersaturation than are blocky ones.
953
-------
Use of Diluting Water
Of the various process steps available for reducing scaling,
addition of water to the slurry entering the scrubber is one of the more
effective. For example, if the slurry is diluted by 20%, i.e., addition
of 20 gal water per 100 gal of slurry liquid phase, the capacity of the
resulting liquid to take up S02 in the scrubber is increased by about
on the basis of solubility alone (assuming that without the water addition
the saturated liquid phase can absorb only that amount resulting from a
drop in pH from 6-0 to 5,5 in the scrubber). In addition, the benefit of
supersaturation is increased in proportion to the volume of water added.
The main disadvantage of diluting water is that liquid must be
removed from the system (blowdown) to maintain the liquid volume in the
system at a constant level. Eventually this blowdown must be drained to a
watercourse, carrying with it dissolved sulfite and sulfate plus soluble
constituents introduced by the limestone and the boiler gas (Mg, Cl, Na, K) .
The water pollution aspects of this are discussed in another of the papers
in this symposium ("Potential Water Quality Problems Associated with Lime/
Limestone Wet Scrubbing for S02 Removal from Stack Gas" by James S. Morris).
Since there are some unavoidable losses of water from the system,
a certain amount of water can be added without need to blow down part of the
scrubbing solution. A water balance for typical scrubber conditions is shown
in Figure 2. The toal allowable makeup, 0.7^ ton water per ton of coal
burned, is quite small in comparison with the amount of liquid recirculated.
The makeup is equivalent to a blowdown of only about 1.0%.
Reduction in pH
There is an increasing body of evidence to the effect that reduction
of pH in the scrubber decreases scaling. No conclusive explanation for the
effect appears to have been advanced.
When CaO is used, the reduction in pH can be accomplished by cutting
back on the CaO:S02 ratio to less than stoichiometric. This reduces S02
absorption, of course; the question then is whether the reduction in S02
absorption required to avoid scaling will make it difficult to meet S02
emission regulations. No data appear to have been reported on the point.
Dilution with water can be combined with pH reduction, of course, to improve
absorption without incurring scaling.
The beneficial effect of low pH may be associated with the deposition
of solid CaC03 that can occur at high pH. In small-scale continuous tests at
TVA, use of Ca(OH)2 rather than CaC03 as the feed material (in countercurrent
scrubbing) resulted in rapid deposition of CaC03 in the upper part of the
scrubber near the slurry inlet. The deposit also contained CaS04'2H20; it
may be that CaC03 was converted to CaS04-2H20 in place. In the work reported
by Lessing, the accumulated scale contained as much as lQ% CaC03, which also
indicates the possibility of CaC05-CaS04 conversion on the scrubber surfaces.
954
-------
STACK GAS, 11.7
(0.57 H20)
STACK GAS, 12.1
(O.99 H20j 0.42 H20 PICKED
UP IN SCRUBBER)
SCRUBBER
r
MAKE UP H20, 0.76
ASSUMING 50% H20 IN
FINAL SETTLED OR
FILTERED SOLIDS AND
RECIRCULATION RATE
OF 50 GAL/MCF
CIRCULATION
TANK
1.86 H20
0.33 SOLIDS
FILTER
OR POND
H20, 69.6
SOLIDS, 12.3
.53 H20
0.33 WASTE SOLIDS;
0.34 H20 (0.33 AS H20,
0.0061 AS CoS03'0.5H20,
0.0043 AS CaSO4'2H20)
FIGURE 2
Water Balance for S0g Removal from
Quantities are tons per ton of coal burned
The use of CaC03 instead of CaO, which also gives a lower pH level
in the system, has generally reduced scaling. In the TVA tests mentioned
above, there was no deposition when CaC03 slurry was used instead of Ca(OH)2-
955
-------
The pronounced effect of pH on sulfite solubility may also be a
factor. A given reduction in pH in a low pH range (say, from pH 6 to k)
gives more increase in sulfite solubility than a similar reduction at a
higher level (say, from pH 9 to 7)-
Solids Content of Slurry
Much of the available data on scaling comes from the early work
in England on the ICI-Howden process1?2>3>4. One of the more important
variables in this work was the content of calcium sulfite and sulfate
crystals in the recirculating slurry; the concentration specified for the
full-scale unit constructed at the Fulham station was 3 to 5$ each of
calcium sulfite and calcium sulfate.
Although a large number of sulfite-sulfate crystals circulating
in the scrubber loop provides surface on which dissolved sulfite and sul- '
fate tend to crystallize preferentially, this alone does not seem to be
sufficient for preventing scaling. In the ICI work it was necessary to
adjust other factors also to get nonscaling operation. In TVA pilot plant
tests, scaling occurred even though CaC03 slurry was used and the solids
content of the slurry was about 15$.
Desupersaturation in Surge Tank
Another variable found important in the ICI work was retention
time of the slurry before return to the scrubber. It was considered necessary
to provide enough time to dissipate the supersaturation developed in the
scrubber, since it is apparent that any degree of supersaturation at point C
in Figure 1 reduces the capacity of the solution for absorbing S02 in the
scrubber without incurring scaling.
ICI specified a retention time of about 2.5 minutes in the recir-
culation tank. The effects of delay time and crystal concentration on
supersaturation, as reported by Lessing, are shown in Figure 3- Conclusive
* Lessing, R. J. Soc. Chem. Ind. 57, 373-88 (Nov. 1938).
Rees, R. L. J. Inst. Fuel XXV (lt8), 350-57 (March 1953)-
3 Hewson, G. W., Pearce, S. L., Pollitt, A., and Rees, R. L. Soc. Chem.
Ind. (London), Chem. Eng. Group, Proc. 15, 67-99 (l933)«
4 Pearson, J. L., Nonhebel, G., and Ulander, P. H. N. J. Inst. Fuel VIII
(39), 119-156 (February 1935)-
5 However, the species distribution was 5.0/0 CaS03'0.5H20 and 1.5/0 CaS04-2H20
(remainder ash), and therefore the amount of CaS04-2H20 may not have been
sufficient. Moreover, the excess of calcium sulfite may have been harmful;
Lessing pointed out that the presence of sulfite crystals can cut in half
the beneficial effect of sulfate crystals on sulfate desupersaturation.
956
-------
9000
6000
2
a 7000
•
z
2 6000
5
8 5°°°
0 4000
-------
however, that the solution returning to the scrubber is still supersaturated
in regard to both sulfate and sulfite; further data will be obtained in an
effort to resolve the question.
It should be noted that the rise in pH in the hold tank due to the
dissolution of CaO or CaC03 results in a much lower solubility of calcium
sulfite (Table l) and consequently a very high degree of supersaturation
can be developed even if the solution entering the hold tank is only saturated
with calcium sulfite. For example, if the pH rises from 5.0 to 6.0 in the
hold tank and the entering solution at pH 5.0 is saturated with calcium sul-
fite, the 302 ppm of S02 in the incoming stream is about six times the
saturation amount in the returning stream (at pH 6.0) to the top of the
scrubber. Unless sufficient delay time is permitted to precipitate the
excess sulfite, it is not surprising that the returning solution will like-
wise be supersaturated. The extent to which the solution entering the hold
tank is supersaturated will serve to increase further the supersaturation
in the solution returning to the top of the scrubber. The opposite effect
is obtained with calcium sulfate; a pH rise from 5 to 6 in the hold tank
increases the sulfate solubility (Table l) by about 5$. However, since the
resulting decrease in supersaturation is small, delay time is necessary to
dissipate the supersaturation before return of the solution to the scrubber.
Slurry Circulation Rate
A further process variable emphasized in the ICI work was amount
of slurry circulated per unit of S02 absorbed. After optimization of surge
tank retention time and solids content of the slurry, the critical degree
of supersaturation in the scrubber could be determined. From this the amount
of slurry circulation necessary to avoid exceeding the critical value was
calculated. For the inlet S02 concentration involved in the ICI tests (about
1000 ppm S02), the circulation rate required was about 130 gal per Mscf of gas.
This amount of circulation is almost intolerable for the situation
in the eastern part of the United States, where the high sulfur content of
the coal produces an inlet S02 concentration two to three times that in the
British work. For an inlet S02 content of 3000 ppm, which is not uncommon,
the required slurry rate would be about 400 gal/Mcf to give a liquor:S02
ratio similar to that used by ICI. Capital and operating costs for such a
pumping load would be extremely high.
This objection would not hold for the western part of the United
States, where the sulfur content of the coal is quite low, for example, 0.6$.
For O.S'/D sulfur, the slurry rate equivalent to the ICI practice would be
only about 50 gal/Mcf.
In use of CaO in this country, the general practice has been to
operate with a relatively low L/G (gal/Mcf) because good S02 absorption can
be obtained with CaO at low liquor rate (on the order of 10-30 gal/Mcf).
958
-------
Hence scaling has been promoted by the low liquor:S02 ratio. For CaC03,
which is much less reactive than CaO, it has been necessary to use higher
L/G (on the order of 40-6o) to get good absorption, which is favorable to
reduction of scaling. This is another factor, in addition to lower
system pH, that makes CaC03 a better absorbent in regard to scaling.
One possibility would be to use two or more scrubbing stages
with separate recirculation circuits. Enough delay time could be de-
signed into each circuit to desupersaturate the solution so that the
sulfite-sulfate make in each circuit would not be large enough to cause
precipitation and scaling. Such a system would be expensive but might
solve the problem.
Erosive Effect of Slurry
There is some evidence that the erosive or scouring effect of
the slurry may be a very important factor in scaling. In the TVA pilot
plant work, severe scaling was encountered when stack gas was scrubbed with
CaC03 slurry in a crossflow scrubber. There was no significant scaling,
however, in spray and mobile-bed scrubbers. The slurry circulation rate
and slurry solids content were somewhat higher in these tests, but the main
difference was the intense scouring effect of the high velocity sprays in
the spray scrubber and the bouncing spheres in the mobile bed--as compared
with the relatively slow flow of slurry through the crossflow. The thin
spines of CaS04-2H20 formed on the crossflow packing are shown in Figure k.
It would be expected that the erosive action in the other two scrubbers
would break off such spines as fast as they formed.
Even without solid particles in the liquor, it would be expected
that high velocity flow of liquid past surfaces would discourage adherence
of nuclei in the formative stage. This point was emphasized in the ICI work.
There are numerous factors that may affect the magnitude of the
scouring effect.
1. Slurry pumping rate obviously is important, but has generally
been fixed more by S02 removal requirement than by other
considerations.
2. Slurry velocity. In a spray scrubber a relatively high liquor
velocity from the spray nozzles is necessary for good spray
distribution. Velocity probably is lowest in the fixed packing
type. Data on effect of liquor velocity are not available.
3. Solids content of the slurry should be as high as practicable.
However, the 12 to l^% used in the TVA and ICI work for other
reasons (mainly to provide crystal surface) may be as high as
should be attempted. The slurry not only scours crystals away
from surfaces but also erodes the surfaces themselves. TVA
959
-------
FIGURED
_; a tg Scale^ on_Scrubber Packing
960
-------
has under way a test program aimed specifically at the
erosion problem. This work may indicate that a lower
solids content should be used to avoid excessive erosion.
Particle size of solids will also be varied in the TVA
tests, since small particles should not be as erosive as
large ones. They may also be less effective in removing
scale.
Silting
One of the more puzzling aspects of scaling is the effect of
silting, that is, the accumulation of solids on scrubber surfaces by
mechanical means — either by settling onto transverse surfaces or into
crevices or by fine particles being caught bodily on rough surfaces. If
calcium sulfate crystals are accumulated in this way, dissolved sulfate
should crystallize on them as it does on crystals in the bulk slurry. The
difference is that nucleation on surfaces is bypassed when crystals accumu-
late on surfaces by mechanical means; crystallization on the accumulated
crystals can cement them together and onto the surfaces, thus providing
an additional mechanism for scaling.
Calcium carbonate and calcium sulfite crystals can accumulate
on scrubber surfaces in the same way. Calcium carbonate is not stable
in the system except at high pH. In reacting with the solution, however,
it may become converted in place to calcium sulfite or sulfate and thus
cause scaling. Calcium sulfite can also form a relatively stable form of
scale although the scale found in TVA tests has been mainly sulfate.
No data appear to be available on the effect of silting. In the
TVA crossflow scrubber work, relatively loose deposits of CaC03 formed in
the packing, presumably by silting. Calcium sulfate also formed but it is
not clear whether or not the silting contributed to the sulfate scale
formation. However, the fact that the spray and mobile-bed scrubbers, in
which silting would not be likely, did not scale is a possible indication
that silting is a factor. Further work on the point appears desirable.
Degree of Oxidation in Scrubber
It seems logical that the "make" of sulfite and sulfate in the
scrubber, per volume of solution, should have a major effect on the degree
of scaling; as noted earlier, this was the basic consideration in the ICI
work. The limited data on the point from U.S. work are confusing. In the
recent TVA tests, only 10 to 20$> oxidation of sulfite occurred in the scrubber
(as indicated by the solid phase composition) yet the scale was mainly sulfate.
961
-------
Work by others has produced sulfite scaling, however, even at a degree
of overall oxidation on the order of 50$. It is not clear what makes the
difference and whether it is better to promote or inhibit oxidation as
far as scaling is concerned.
It is obvious that much more study is needed in this area, since
there are ways to change oxidation rate if such a change would be helpful.
Data are needed on the relationship between sulfite and sulfate in regard
to (l) supersaturation driving force needed under various conditions to
initiate nucleation, (2) activation energy of nuclei formation, (3) strength
of bond in adherence to surfaces, (k) rate of crystal growth after nucleation,
and the effect of increasing amounts of homogeneous surface for growth, and
(5) effect of other dissolved constituents and of ionic strength.
Nature of Surfaces
<
The type of construction material and condition of the surface
may be significant. ICI adopted wood packing and pointed out that corrosion
roughening of steel surfaces promoted scaling. Today wood is not favored
as a packing material and steel surfaces likely will be covered with rubber
or plastic to reduce corrosion and erosion. In the TVA pilot plant tests,
scale grew well on polypropylene packing and in small-scale work scale
formation occurred on glass.
It does not seem likely that type of construction material will
be a major factor in preventing scaling. However, it may be important enough
to be a trade-off alternative in determining the most economical solution to
the problem.
Use of Additives
In the desalination field various additives have been proposed
for reducing or avoiding scaling. Several mechanisms can be considered,
including nucleation inhibition, weakening the bond between crystal and
surface, formation of films that alter surface properties, and alteration
of crystal habit to change crystal growth pattern. A major drawback is that
such additives are likely to be expensive and that they will be lost from
the system with the liquor in the discarded wet solids. However, the cost
might be justified if a major benefit resulted. Experimental work is indi-
cated since experience in the desalination field will not likely be applicable
because of the wide differences between the two systems.
962
-------
Summary
Although a considerable amount of experience has accumulated on
scaling in lime-limestone systems, much remains to be learned regarding the
mechanisms involved and the best way to arrive at a reliable and economical
method for eliminating the problem. The iCI-Howden formula of adequate
retention time in the recirculation tank, high crystal content in the slurry,
and high recirculation rate per unit of S02 absorbed does not appear appli-
cable to high-sulfur coals because of the extremely high pumping load
required; however, it may be usable for low-sulfur coal. Since the ICI
method was developed from basic chemical considerations, there is no obvious
way to improve on it from the standpoint of process chemistry except for
using limestone instead of lime, which, for some reason as yet not clearly
identified, reduces scaling considerably.
There are, hox-jever, some mechanical factors that may be helpful.
With slurry composition and scrubber design aimed at producing an intense
scouring effect, it appears that scaling can be controlled at slurry re-
circulation rates no higher than those required for good S02 absorption.
The next step should be optimization of the system to give the
most economical combination of conditions. This will require a considerable
amount of research and development on the quantitative effect of the various
factors and on the basic mechanisms involved.
963
-------
-------
CALCIUM SULFATE SCALING
Bela M. Fabuss
Lowell Technological Institute
450 Aiken Street
Lowell, Massachusetts 01854
Prepared for
Second International Lime/Limestone
Wet Scrubbing Symposium
New Orleans, Louisiana
November 8-12, 1971
965
-------
CALCIUM SULFATE SCALING
by
Bela M. Fabuss
LOWELL TECHNOLOGICAL INSTITUTE RESEARCH FOUNDATION
450 Aiken Street
Lowell, Massachusetts 01854
Distillation processes are usually regarded as the most
effective means of producing potable water from saline or brackish
water. The evaporative desalination is often hampered by the
formation of calcium sulfate scale on the heat transfer surfaces.
This scale may be deposited at low temperatures in the form of
gypsum, at high temperatures as hemihydrate and may undergo trans-
formation to anhydrite.
Figure 1 shows a summary of the solubility curves for
these three modifications. It can be seen that both the anhydrite
and the hemihydrate show a strong inverse solubility. Figure 2
shows pilot and demonstration plant data of the Office of Saline
Water projects plotted on this diagram. It clearly indicates
that scale-free operation was frequently achieved well above the
arhydrite solubility curve and even in some instances above the
hemihydrate solubility curve.
Figure 3 shows several sea water heating runs at a
series of heating rates and on two different surfaces, stainless
steel and an epoxy resin. Figure 4 shows a series of experiments
when the supersaturation was achieved by evaporation at a constant
temperature. Finally, Figure 5 summarizes these data giving two
sets of precipitation curves superimposed on the calcium sulfate
966
-------
solubility diagram. In summary, this work clearly shows that the
precipitation limits and scale formation depend on the operating
conditions of the unit. Concentration of sea water by non-
boiling heat transfer permits operation at significantly higher
supersaturations than by boiling heat transfer. The effect of
other variables such as the heating surface materials, additives,
and heating and evaporation rates was slight. Scaling was con-
trolled primarily by kinetic factors, determined by the residence
time of the solution in the unit.
In applying these considerations to lime scrubbing of
stack gases, let us take a look at the equilibrium concentrations
of the ions and molecules in the system. We are dealing here only
with the dissolved ions. Figure 6 shows the calculated concentra-
tion of the ions in the solution at equilibrium.
If we consider that scaling occurs by calcium sulfate
precipitation and not by calcium sulfite or hydroxide deposition,
this can occur only at high sulfate ion concentrations since the
calcium ion concentration is controlled by the solubility equi-
librium of calcium sulfite. Thus, the scaling should depend on
the rate of oxidation and on the pH of the solution. The rate
of oxidation strongly depends on the pH of the solution. Above
pli. 7 and at high suspension concentrations, the suspensions were
practically stable and little or no oxidation occurred. Even at
pH 3, the oxidation of CaSO., suspensions proceeded slowly after
a significant induction time.
Based on the presented evidence, we would like to draw
the following tentative conclusions:
967
-------
(1) Lliniinating CaSO^ , Ca(OH),., and CaCO^ as potential
scale formers, scaling should occur only at high conversions at
low pH values.
(2) The scaling is most probably the result of a sequence
of processes: dissolution of calcium sulfite , oxidation in the
dissolved state, precipitation of calcium sulfate with potential
further conversion to anhydrite scale.
(3) The kinetics of each of these processes must be studied
to identify the rate controlling step.
(4) There is sufficient evidence from desalination practice
that even if the oxidation to sulfate cannot be prevented, scale-
free operation can be achieved by proper selection of operating
variables affecting the equilibria and kinetics of the process.
968
-------
o
o
c\j
UJ
o:
LU
OL
S
UJ
O
CD
O
OJ
O
o
ro
O
CD
OJ
CD
in
UJ
CC
Q
O
<\J
OJ
UJ
o:
(T
UJ
CL
S
LU
CD
M
O
O
ro
— o
O
CO
NOIlVdlN30NOO
969
-------
IO
ro CJ
NO I lVdiN3DNOD
O
970
-------
X X
o o
0- CL
LJ LJ
O LJ «
E
0
o
ro
CD
O
c c
E "E
o o
0 0
CO 00
CD CD
D •
1
O
lO
c
"E
o
o
O
O
*
.E "^ h
"E ^v*>.
o ^A/ly" "~ """•--
lO (
*
*
lilt
J LJ
h-
0
00
)
O
o
O O O O o^
-------
to o> o
10 ro si-
— \ _) ID — >
d .E .£
_,. CJ
51 CM ro
I
CO
O
O
o
ro
O
OJ
O
J-
w
CD
% 'QBlVlldlOBdd
972
-------
O
O
C\J
o
CO
o
o
o
LU
a: o
=> ^
<
a:
LU
a. o
2 o
Ld ""
O
00
O
CO
O
O
OJ
10
Q
I
I
I
ro <\j —
NOIlVdiN30NOO
O
CM
o
O
O
ro
U_
o
O -
CD LU
N EC
o
c\J
O
CO
O
si-
H
o:
LJ
LU
O
O
(D
to
W
«
D
CD
973
-------
K)
° i
CM
CJ
to
O
CVJ
ro
in
CDNODJOOT
974
-------
REVIEW OF SCALING PROBLEMS IN LIMESTONE BASED WET
SCRUBBING PROCESSES
By: Joan B. Berkowitz
Arthur D. Little, Inc.
Cambridge, Massachusetts
November 11, 1971
Arthur D Little, Inc
-------
REVIEW OF SCALING PROBLEMS IN LIMESTONE BASED WET
SCRUBBING PROCESSES
I. General Considerations
Scaling involves the precipitation of insoluble salts from aqueous
solutions onto process equipment surfaces. Any salt may precipitate if its
solubility limit is exceeded at some point in the processing stream. From
the point of view of thermodynamics or equilibrium, solubility is a
function of specific concentrations of the precipitating ions, total
ion concentration, local temperature, and pH. Kinetically precipitation
will not necessarily occur, even if the theoretical solubility limit is
exceeded for a given salt, since some degree of supersaturation is
typical of crystallization phenomena generally, and in many practical
cases a very high degree of supersaturation can be sustained. Precipita-
tion per se is not scaling. A precipitate becomes a scale when it
attaches itself to a solid surface, either by nucleation and growth
directly on the surface or by migration of particulates from the bulk
of the solution to the walls. A precipitate which forms within the body
of a solution can be carried in suspension within the processing stream
and will not result in scale formation unless it is carried to the walls
and tends to adhere there.
The salts most comonly found as components of scale are:
(anhydrite), CaS0^.2H20 (gypsum)? CaSO^.1/2 H20 (hemihydrate) ;
CaS03.2H20; and Mg(OH)~. In fact, any wet process in which calcium or
magnesium sulfites or sulfates must be handled is prone to scaling
problems. For example, saline water distillation processes, wet phosporic
acid manufacturing processes, as well as limestone/dolomite wet scrubbing
processes for removal of sulfur dioxide have all been very much troubled
by the deposition of scale on heat transfer and other surfaces. Accumula-
tion of scale in pipelines, orifices, and other flow passages results
in plugging of the equipment, often to the point where it becomes
inoperable. In the wet phosphoric acid process, calcium sulfate formed
by reaction between calcium phosphate ore and sulfuric acid, typically
crystallizes in lines to the extent that periodic shutdown is necessary.
976 Arthur D Little, Inc
-------
The so-called scale forming compounds listed above have two signifi-
cant characteristics in common. First, the salts exhibit inverse
solubility behavior; i.e., they become less soluble as solution tempera-
ture increases. Second, the salts tend to form relatively stable, super-
saturated solutions. The inverse solubility as a function of temperature
is probably the major factor responsible for ordinary boiler scale, and
for the scaling of heat transfer surfaces in saline water evaporation
plants. Scaling under relatively isothermal conditions may often be
ascribed to uncontrolled precipitation from highly supersaturated
solutions unto receptive surfaces of process equipment.
II. Wet Limestone Scrubbing Processes
A generalized wet limestone scrubbing process is schematically
depicted in Figure 1 and represents several alternative methods of
operating an SCL scrubbing process using calcium based reactants. As in
the wet limestone/dolomite injection process developed by Combustion
Engineering and Union Electric, the limestone can be calcined in the boiler
and hydrolized as it is removed from the gas stream in the scrubber.
Alternatively, as in the Howden-ICI process, lime or limestone can be
added outside the scrubber loop, thereby allov;ing greater flexibility of
scrubber pH and scaling control.
A. Scale Control in the Howden-ICI Process
The introduction of alkali outside of the scrubber loop is not in
itself sufficient to prevent scaling. It does, however, permit the
application of a number of simple scale control techniques. In one of
the early wet scrubbers, which was set up in Fulham around 1935, a scale
2-3 inches thick was found on scrubber surfaces within 72 hours after
start-up. The key to eliminating the problem lay in the understanding
and control of supersaturation behavior in calcium sulfate and sulfite
solutions, the primary products or "make" of the wet scrubbing process.
It was recognized that calcium sulfite and calcium sulfate solutions
exhibit an apparent stability under conditions of fairly high super-
saturation. If the scrubbing liquor is never permitted to become
anything more than slightly supersaturated, then sulfite and sulfate
precipitation in the scrubber loop will be very slow. The slight super-
saturation can then be destroyed by precipitation at preselected sites
Arthur D Little, Inc
-------
o
CM
I
1
fc
I
m
CO
D
cc
cc
cc
D
LU
O
V)
LU
5
_l
CC
O
LU
to
.O
.0
I
"8
JQ
I
/g I
3 «2 3
II
ht
c/5 UJ
O)
.£ .*
2 g
LU
CC
D
C3
8
u. (D CD
978
Arthur D Little,!
-------
so as not to interfere with normal scrubber operation.
The ICI group applied their understanding of supersaturation to
effectively overcome scaling in their wet scrubbing operations. A three-
pronged approach was used. First, the "make" of calcium sulfite and
calcium sulfate per pass through the scrubber loop was controlled, by
empirical adjustment of absolute flow rates and L/G ratios, so that the
solutions formed were only slightly supersaturated. Second, a delay
tank was introduced where supersaturation of the scrubbing liquor could
be dissipated prior to recirculation. Third, suspended crystallites of
calcium sulfite and sulfate, 3-5% of each, were carried in the circulating
liquor to provide sites for homogeneous nucleation.
The above three measures taken to control supersaturation in the
ICI wet scrubbing process went a long way towards alleviation of scaling
problems. Other design factors, however, had to be taken into account
before the problem could be completely eliminated. The solubility of
calcium sulfite is decreased dramatically as pH is increased, even in
the range 6 to 7. The pH must therefore be controlled so that the
solubility change does not occur in the scrubbing tower where calcium
sulfite might precipitate onto the packing. In the ICI process, the lime
or limestone slurry is added just before the delay tank where the
increase in pH assists in dissipating supersaturation. The rate of
addition of alkali is adjusted so that the pH at the bottom of the
scrubber is maintained at about 6.2. The total pH change through the
scrubber is therefore from about 6.8 at the top to 6.2 at the bottom.
It has been implicit in the discussions so far that solutions and
slurries are homogeneous in composition. It is naturally of prime
importance that such uniformity in composition be maintained, at least
to the extent that supersaturations are not exceeded in localized areas
within the scrubbing tower. In the ICI work deep plates inserted in the
lowest section of the scrubber tower provided for even gas distribution.
The plates were also in a region of high sulfur loadings and the main-
tenance of high liquor velocities as well contributed to elimination of
scaling in the system. It is interesting to note that in spite of the
substantial progress made by ICI in the 1930's towards prevention of
scaling by proper adjustment of design parameters, ICI was still troubled
by occasional scaling problems when wet scrubbing operations were resumed
979 Arthur D Little, Inc
-------
in the 1950' s. The problems were usually the result of incomplete
irrigation of the scrubbing tower grids due to accidental blockage of
flow elsewhere in the system.
Finally some materials and surface finishes are more resistant to
nucleation, growth, and adherence of scale than others. In the ICI work,
it was found that scaling could be minimized by constructing scrubber
grids of smooth, planed red deal wood. The critical factors are probably
corrosion resistance and surface smoothness.
B. Scaling in Limestone Injection Wet Scrubbing Processes
Although some of the ICI work involved the use of lime slurries,
the bulk of the effort by far was concentrated on limestone additions.
It is generally believed that lime is more efficient than limestone for
removal of SCL. However, the use of externally calcined lime adds
substantially to the cost of scrubber operations. By introduction of
limestone directly into the boiler, calcination of limestone may be
accomplished very inexpensively. Unfortunately simultaneous introduction
of lime and flue gas into the scrubber circuit has introduced scaling
problems which are yet to be brought under control .
The key mechanism responsible for S09 absorption in limestone
injection wet scrubbing process is believed to be the reaction of SCL(g)
in the flue gas with a circulating saturated slurry of calcium sulfite,
resulting in the formation of calcium bisulfite in solution:
CaS03 (sat. soln.) + S02(g) + HZ) (]_) -> Ca(HSC>3)2 (soln-)
A sudden increase in pH in local areas, where hot lime particles from the
boiler first contact the scrubber liquor, can force calcium sulfite out
of solution with resultant plugging problems. The lack of pH control at
the bottom of the scrubber may be a major factor contributing to scaling
in limestone injection systems. Solution to the problem is not easy and
might require major design changes.
In pilot plant experience with limestone injection wet scrubbing
processes, scaling has been more the rule than the exception. In one
installation which has been described in the literature, calcium sulfate
deposited on overflow drain screens in the scrubber and drastically
restricted water flow. Scaling and plugging were also encountered in
the marble bed scrubber and in the reheater. The most serious scale
980
Arthur D Little, Inc
-------
problem was encountered at the scrubber inlet where temperature is at
a maximum. It might be anticipated that this would be a critical position,
due to the inverse temperature dependence of calcium sulfate solubility.
Scaling at the scrubber inlet may be controllable through the use of
saturated sprays to pre-cool the flue gas and to avoid the formation of
a sharply defined wet/dry interface. This approach has been suggested
by Bechtel, and is expected to be tested in the pilot plant later this
year.
While the specific scale prevention methods devised by ICI for
limestone slurry scrubbing are not all directly applicable to boiler
calcined limestone injection scrubbing, the insights into the factors
responsible for scale formation can provide guidance to the development
of appropriate control techniques. The factors of primary importance
are pH, both local and global; gas and liquor distributions; liquor
velocities at scrubber surfaces; "make" of calcium sulfite and sulfate;
scrubbing liquor composition; and materials of construction. Under EPA
sponsorship, we are currently building a laboratory bench scale scrubber
to explore the effect of these factors on scaling behavior.
III. Scale Composition
The principal components of the scale formed in the ICI limestone
slurry process, before the scale control methods were optimized, were,
gypsum (CaS04.2H20), 60-90%; CaS03.l/2H20, 1-40%; and calcite (CaC03),
1-5%. The ratio of sulfate to sulfite in the scale seems to depend on
the degree of oxidation of sulfite in the scrubber circuit. This in
turn seems to be highly sensitive to catalysis by trace quantities of
transition metals.
Very little information seems to have been published about the
composition of scales formed in the boiler calcined limestone injection
processes. The absorption process is generally described in terms of
sulfite-bisulfite reaction, but in the presence of fly-ash, oxidation
of sulfite to sulfate is expected to be quite rapid. Since control of
scale depends to some extent on composition and supersaturation behavior,
the extent of sulfate formation could be an important and possible crucial
process parameter. When a limestone slurry is used as a reactant, it is
hardly surprizing that CaCO., might be a component of the scale. When
limestone is calcined in the boiler prior to introduction to the scrubber,
981 Arthur D Little, Inc
-------
the calcining process is complete and virtually no CaCO.,(s) is carried
in the flue gas mixture. Any CaCO_ found in the scale would thus have
to be due to the reaction of CO- with scrubber liquor components. The
role of C0~ in the injection scrubbing process is very much in need of
clarification. The critical step in the ICI process is supposed to be
the reaction of CaCO- with CO- in the flue gas to form the bicarbonate
which subsequently reacts rapidly with SO- to form sulfite. In the
injection process, the absorption mechanism seems to involve primarily
sulfite/bisulfite rather than carbonate/bicarbonate. If there is a real
difference in mechanism, a different approach to control may be required.
982 Arthur D Little, Inc
-------
DEPOSITION PROBLEMS AND SOLUTIONS IN THE COMBUSTION
ENGINEERING LIME/LIMESTONE WET SCRUBBING SYSTEMS
J.R. Martin
Combustion Engineering, Inc.
Prepared for
Second International Lime/Limestone
Wet Scrubbing Symposium
New Orleans, Louisiana
November 8-12, 1971
983
-------
Lime/Limestone Wet Scrubbing Symposium
Thursday, November 11, 1971
Session on Scaling Problems
J. R, Martin
Combustion Engineering, Inc.
It would appear that Combustion Engineering has been asked to
participate in this session on scaling problems in lime/limestone
vet scrubbing because of our vast experience in producing scale.
The deposition problems which we have experienced in the C-E - APCS
fall into two general categories. The first area includes all those
deposits which are mechanical in nature (i.e., deposition of solids
due to drop out at low gas velocities). The second area is limited
to scale formation as a result of chemical reaction. My initial
remarks will be related to the first type of deposition; mechanical.
Mechanical Deposition
The system schematic shown in Figure I of the C-E - APCS
(limestone-furnace injection) at Kansas Power and Light Corapary,
Unit #k will serve as the reference for this discussion. The C-E
APCS at Union Electric, Mersmec Station is similar except for
employing a clarifier rather than a pond; therefore, the statements
herein are considered to be applicable to both systems.
The first deposition probiera that was encountered was the
mechanical plugging of the scrubber inlet. The inlet became ^0-
50 percent plugged within 2-H hours of operation of the system.
Figures II and III show the scrubber inlet plugged and clean.
The clean inlet is as a resuK of insta] "1 ing a sootblower which
prevents excessive build-up of d<-porits. This de-position is
984
-------
caused bv "the wet-dry interface at the scrubber inlet. The inlet •
deposit is typically composed of flyash, calcium sulfate, and cal-
cium oxide bound together in a cemcntitious mixture. The calcium
sulfate which has been found in the inlet deposit does not exhibit
the crystalline type properties of calcium sulfate scale. It is
our conclusion that the calcium sulfate found in the inlet deposit
is a result of the removal of sulfur trioxide in the boiler by the
injected limestone; this is verified by the fact that the composi-
tion of the inlet deposits are quite similar to that of the dust
entering the APCS.
Another area of mechanical deposition is the area under the
marble bed. This area includes the underbed spray system, the
marble bed structural supports, the gas straightening vanes, and
the scrubber walls. Most of this deposition is also a result of
wet-dry interfaces. The flue gas entering the scrubber is 300-
350°F and is cooled down to its saturation temperature (llO-128°F)
in the area under the marble bed. During this cooling, some of the
hot dry flue gas impinges on partially wetted surfaces and deposi-
tion of the dust being carried by the flue gas can result.
The reheater and demister (shown in Figure l) are two components
of the APCS with which we have experienced deposition and scaling
problems. The deposition of mud (c?iemically uncombined solids) on
the demister occurs in normal operation of the system to a minor
extent. This build-up of solids IL cleaned by the utilization cf a
demister wash system, but vheu the marble bed is not operating
985
-------
correctly, excessive "build-up of solids can occur which the vash
system cannot cope vith. Additionally, calcium sulfate scale is
formed in the demister when the scrubber becomes supersaturated
vith calcium sulfate. The liquid vhich the demister is separating
has the highest concentration of calcium sulfate in the scrubber
and therefore, the greatest tendency to scale is at this point.
The problem of supersaturation of calcium sulfate and the resulting
scaling will be discussed in subsequent remarks. Also the APCS
reheater builds up calcium sulfate scale when the demister is not
operating properly. The excess liquid impinging on the reheater
is evaporated leaving anhydrous calcium sulfate. This deposit is
very hard even withstanding sand blasting.
Chemical Scale
Basically, we have formed three types of chemical scale in the
C-E - APCS limestone scrubbing process: calcium carbonate, calcium
sulfate, and calcium sulfite. These three types of scale have been
found in many locations throughout the APCS. Ky remarks on this
subject will be limited to where we have formed these different
scales, what we think the mechanism or reaction is that causes
them to form, and how we have eliminated or minimized their formation.
The marble bed is where we have experienced our most severe
scaling problems. Scaling of both calcium sulfate and calcium
sulfite has occurred on the overflow pots, recycle piping, and an
the marble bed.
986
-------
Calcium Sulfate
Figure IV is the marble "bed at Meramec Station, Union Electric
Company APCS, after about 2\ hours of operation in the fall of 1968.
The deposition on the overflov pots is a mixture of calcium sulfate
scale and flyash. The flyash appears to get trapped in the deposit
as the calcium sulfate is scaling. The deposit has a definite
crystalline shape and reflects light similar to chips of glass.
The calcium sulfate scale also was found on the scrubber walls.
This problem of calcium sulfate scale was not encountered in
our earlier pilot plant work and was therefore unexpected. Ini-
tially, we thought the scaling might be due to the retrograde
solubility that calcium sulfate exhibits (i.e., liquid tempera-
ture in the scrubber is higher than the liquid temperature in the
clarifier). This theory was weakened when it was determined that
the form of calcium sulfate we were scaling was gypsum and not
*
anhydrous. There is no change in gypsum's solubility in the
temperature range that the C-E - APCS operates.
Next, we switched to the use of a dolomitic limestone and the
higher magnesium seemed to depress the overall calcium in solution
to a point where it did not scale calcium sulfate. In the late
fall of 1969} the calcium sulfate scaling problem occurred again.
This occurrence was probably due to longer sustained periods of
operation of the APCS. It was at this tine that we added some
additional dilution water to the system in order to maintain the
987
-------
calcium sulfate concentration below the scaling level. The rate
of dilution water we determined that was required to run Meramec
was equivalent to about two gallons per thousand acfm. We are
not suggesting that this is the best way to prevent sulfate
scaling, but at that time it alleviated the problem and enabled
further operation of the system.
To date, we have never seen sulfate scaling in the scrubber
at Kansas Power and Light, Lawrence No. k. Tests this year in
our laboratory suggest that the. pond in Kansas (we go directly
from the scrubber to the pond) acts as a desaturating vessel, in
other words, the scrubber effluent which is supersaturated in
calcium sulfate solution going to the pond will "be reduced to
saturation given sufficient time. In fact, the calcium sulfate
concentration in the spray water from the pond has never gone
higher than 1200 ppm. Whether this is due to the action of the
pond or that we have not operated *for a long enough period of
time to completely saturate the pond is not known.
Calcium Sulfite
The sulfite scaling problem we encountered occurred when we
tried to recycle the solids from the bottom of t?.e scrubber to
above the bed. Figure V shows a recycle nozzle at Kansas Power
and Light; the spray pattern of this no7.£le is a hollow cone; it
looks like an inverted umbrella. All the deposit on the bottom
of the nozzle is pure calcium Fulfite. We have explained the
988
-------
formation of this type of scale as follows: there is a localized
area highly concentrated in calcium hydroxide where the recycle
slurry enters the bed. This highly alkaline slurry raises the pH
of the bed liquid causing a shift in the bisulfite-sulfite equi-
librium. Scaling of calcium sulfite results because of its rela-
tively low solubility. Later, when we installed a system where
part of the pot effluent water was pumped to the recycle tank and
controlled the recycle slurry pH, ve found that we could operate
the recycle system without sulfite scale.
Figure VI shows another view of the same marble bed. The
diagonal white line down the middle of the figure is where the
division plate is located under the marble bed; it is a loca-
lized area of low gas velocity which readily plugs during normal
operation. The other light areas are the type of deposit which
results when the recycle slurry pK is not controlled.
Figure VII shows the marble bed at K. P. & L., Lawrence Ho. h
after we revised our recycle system and rearranged the overflow
pots. ¥e have been able to run for a two-and-one-half week period
last June with this revised above-bed recycle system without any
sulfite scale formation in the scrubber system.
Calcium Carbonate
The other type of scaling problem we hr-ivc exr/ffiencc-d is
calcium carbonate scaling. Tn January of 1971 ">•'•:> incurred
989
-------
severe scaling of calcium carbonate at Kansas Power and Light
Company. Vie were trying to run some tests on the system to
determine the effect of recycle at the time. One particular
series of tests required running the system without recycle.
During this test very high limestone feedrates to the furnace
were required to obtain reasonable sulfur dioxide removal.
Since there was no recycle of the scrubber reject slurry, the
scrubber effluent slurry to the pond rose to a pH of 10.
Gradually, the pond pH started rising and when it rose to 10,
we developed a serious problem of calcium carbonate scaling
in the spray nozzles of both scrubbers.
The scrubber effluent comes in at the left of the pond
(Figure VIII is a picture of the pond) and then the liquid
travels around the pond to the spray water pumps. The pond
is simultaneously used for the plant's cooling tower blowdown,
their bottom ash blowdown, and their pyrites blowdown. At the
time that, the CaCOo scaling problem occurred, the cooling tower
blowdown was coming into the pond at the A^CS spray water pump
suction. The cooling tower blowdown had been pumped into this
area of the pond since the system was started up in 1968 with-
out any problems.
The cooling tower blcwdown water has about ^00 Dprn of
bicarbonate. Further, the auentit'r of cooling tower blowdown
water entering the por;d i p, roughly equivalent to 1500 gpm whereas:
the spray water to the APC3 is 3oOO gpi.i. It has been theorized
990
-------
that the pH rise which occurred in the pond last January caused a
shift in the bicarbonate-carbonate equilibrium and since calcium
carbonate is a very insoluble compound it became saturated and
subsequently scaled in the spray piping system as it left the
pond. Naturally, the first place where the problem would become
serious would be in the orifice of the spray nozzle.
After analyzing the problem, we moved the location where the
cooling tower blowdown enters the pond to the same place where
the scrubber effluent slurry enters the pond. In this way the pH
rise which may or may not occur depending on the APCS mode of
operation will take place on the inlet side of the pond, thereby
allowing the bicarbonate-carbonate shift and the resulting pre-
cipitation of the calcium carbonate to have sufficient time to
take place in the pond rather than the scrubber. Subsequent
operation of the system has not produced any calcium carbonate
scale.
991
-------
&a-9
-------
H
-------
\"
-------
H
0)
^1
3
d
•rH
995
-------
>
Q)
tr
•H
996
-------
H
>
Q)
tn
•r-l
997
-------
998
-------
H
H
H
tJ>
•H
999
-------
-------
adian Corporation
8500 SHOAL CREEK BLVD. • P. O. BOX 9948 • AUSTIN, TEXAS 78757 • TELEPHONE 512/454-9535
USE OF CHEMICAL ANALYSIS AND SOLUTION
EQUILIBRIA IN PREDICTING CALCIUM
SULFATE/SULFITE SCALING POTENTIAL
by:
Philip S. Lowell
Presented at:
SECOND INTERNATIONAL LIME/LIMESTONE
WET SCRUBBING SYMPOSIUM
8-12 November 1971
Sheraton-Charles Hotel
New Orleans, Louisiana
Sponsored by:
Environmental Protection Agency
Office of Air Programs
Division of Control Systems
1001
-------
\sOrpOr3ilOn ssoo SHOAL CREEK BLVD • p. o 60x9943 • AUSTIN. TEXAS 73757 • TELEPHONES^ -154 9535
1.0 INTRODUCTION
This paper presents experimental data from a TVA
pilot plant. These data are consistent with a proposed
quantitative measure of scaling potential. Both calcium
sulfate and calcium sulfite precipitation could be antici-
pated in the system studied. The possibility of scale
formation in the scrubber or on vessel walls existed for
both sulfate and sulfite, while sulfite scaling could also
take place on limestone feed crystals, a phenomenon known
as "blinding".
The principle object of understanding scaling phenomena
is to be able to design and operate a process that does not
scale. This requires that three types of information be known:
A description of the individual major
pieces of equipment used in the process
including kinetic data, equilibrium con-
ditions, and mass and heat balances.
A description of the entire system.
The ability to predict in quantitative,
measurable terms the scaling potential
in all the different parts of the system.
The equipment descriptions for particular types of
scrubbers have been discussed in Papers la and Ib and the first
two requirements have been treated in general by Dr. Ot;tmers in
Paper Ic.
1002
-------
Radian Corporation
SHOAL CRFLK BLVD • P O BOX 7918 • AUSTIN, TEXAS 78757 • TELEPHON C 512 454 "'533
The third requirement for design of a scale-free
process, a quantitative measure of scaling potential will be
discussed in more detail in Section 2.0. The proposed method,
which was first presented by Mr. J. L. Phillips in Paper Id
involves some function of the relative supersaturation, i.e.,
the quotient of activity product and solubility product
constant.
The experimental results presented here were obtained
at the Tennessee Valley Authority's Colbert Steam Plant. A
pilot plant is being operated there in support of a full-scale
plant to be designed for the Widow's Creek facility. The
objectives of the Colbert pilot operation are to obtain design
and operating data and to find out whether the scrubber and
ancillary vessels can be operated without scale formation.
The results during an extended period of operation at relatively
constant conditions showed that no scaling occurred. Radian
collected data in the middle of this period.
2.0 A QUANTITATIVE MEASURE OF SCALING POTENTIAL
In this section, the basis for the proposed scaling
potential function will be discussed and some basic laboratory
results will be referred to. Although laboratory studies have
been conducted only for calcium sulfate dihydrate precipitation,
it is expected that the same kind of behavior will be shown for
calcium sulfite in future work. While anhydrous calcium sulfate
is the thermodynamically stable form above 40°C, the formation
kinetics are much too slow for it to be of significance in any
system.
1003
-------
Pad \r. PS
Tf
-------
^oo SHOAL CPLCK BLVD • P o Box??-i8 • AUSTIN TEXAS 73757 • TELEPHONIST 1549535
FIGURE 2-1
RELATIONSHIP BETWEEN ACTIVITY COEFFICIENT AND IONIC STRENGTH
1005
-------
Radian Corporation
•ibOO SHOAL CREEK 81.VD • P O. BOX 9948 • AUSTIN, TEXAS 78757 • TELEPHON: 512 454-9535
Subsaturation: a ++acn= < K (2-4)
v_>a oU^ Spp cr\
(dissolution ten- <-asu4
dency)
Supersaturation: ^a^SOr > Kspc SQ ^^
(precipitation ten- 4
dency)
Since the magnitude of K varies widely for different
compounds, it is convenient to "normalize" the means of express-
ing the degree of saturation by using the quotient a1aa/K
sp
Then the relationships in 2-6, 2-7 and 2-8 are true for any com-
pound formed from cation 1 and anion 2.
Equilibrium: ^a- = 1 (2-6)
sp
Subsaturation: a, a,, , /o -7\
~v— < 1 (2-7)
(dissolution ten- sp
dency)
Supersaturation: 'a, a, , ,9 Q\
T, > 1 ^Z -O )
sp
The quotient a, a5/K has been termed the relative Supersaturation,
sp
r» It is proposed that some function of the relative supersatura-
tion is a valid and useful quantitative description of the tendency
towards scale formation.
The phenomena of scaling and nucleation are somewhat
related. Precipitation on an existing crystal requires only
that the relative Supersaturation be greater than one. Very
small crystals are more soluble than large crystals. There is,
therefore, some value of r greater than one that must be attained
in the solution before a nuclei can be formed.
1006
-------
Radian Corporation
CC^k, SL'.D • ~ O 3CX '743 • AUSTIN TIXAS 73757 • TELEPHONE 512 4549535
Scaling is somewhat similar to nucleation in that a new
species is being formed. If the electrostatic forces and lattice
spacing of the hoat surface is very close to that of the scaling
species a value 01 r close to one might be sufficient to cause
scale nuclei. If the host surface is very dissimilar the value
of r will probably be near that required for bulk nucleation.
While this is perhaps a rather folksey description of a complex
phenomenon, it does appear to have merit in correlating the
results. The important point being made is that value of r
required for nucleation should have some relationship to scaling.
Results obtained by J. ~L. Phillips in the Radian
laboratory (Paper Id) show clearly that for calcium sulfate
there is indeed a critical value of relative supersaturation
beyond which nucleation and a greatly increased rate of preci-
pitation takes place. The results are illustrated in Figure
2-2 which shows that beyond a value of ar -j_facn=/K of 1.3 to
U3 ^^4 Sp
1.4 the rate of calcium sulfate dihydrate precipitation markedly
increases. Photographs of the crystals formed at values of the
ratio greater than this critical value clearly indicate the
formation of small new crystals where nucleation has occurred.
Again, the data for calcium sulfite precipitation have not yet
been obtained. However, the same behavior is anticipated for
the sulfite.
3.0 EXPERIMENTAL CONFIRMATION
The equipment and flow arrangements for the pilot
plant scaling studies are given in Figure 3-1. Again it should
be pointed out that scale formation was not a problem in the
system shown. This can be explained by examining the values of
relative supersaturation for calcium sulfate and calcium sulfite
throughout the system. For similicity, let rl be the relative
1007
-------
I <•
I-'
( a
1 6
u
i Or—
.9
\
.8
.7
.6
5
.4
.3
.2
.1
1
Si
i
, — -f
45
Effi
-h-f-
ir£
r-
j -
_i
13
rrrr
.:;i
>
1 r i '
t-
i-
•t
Ij
1
t
t'
~G
*
4
I
S
4
'}
<
4
•r
1-
C
r"
i
i
•H
-f-r
1
t,
-1
U
J
^
4
1)
g
•H
i-l
•H
S
1
QJ
4J
1
4-J
4-1
O,
0
0)
PH
lit
t • :
%
0 ~
;!j!
J i ,1...
'RE
Jon
4'™
rtiT
"tTp"
-rH-r
-T — L-
-H
7-1 1:
rtt;
M -i
::: 1 ."
:.: 1
TT1
I
CII
stc
C
IF-
~T | .
— L
: — -~
::'
'IG
'IT
a
)ic
).5
TTTT
....
,,r
.mz-
— ~
> ,'~
••
URI
AT]
t 4
hie
% E
z^rv
:~
" r i
. \"-.
• t
: 2
:ON
1-5°
>me
>ee
T — :
-- t
. J
]
-'•Er
—
.
:.:v)
•'==1
=~
:_r
-~
_..
N
-2
01
C
tri
d S
x
• 1 1
. ,
' C
.c
>lu
-I--— -
-* ,-L-l
•-rr.|
...
:.^.|
-
--U-i
aSC
So]
rr>
TUT
_
.._
_
--—
r^
. ...
-•-
—
I;-.-'
• _:
I.:!.
:;::|f,|--:.
:r,
)4-
-Ut
r
TTT
.
— -T»
. ,-
;
"-I ^
:•_:!
2Ha
ior
Z^t"r-^-
,-, — .
p
— -'
— -
-
-O
r:-
^/
/-'
r *
0 g
i
i- •-
- 1- -
* — -- f-t-— —
*"- r "~
r
" " "*" r r ' " ' '
i
-. ~.rl _ HT-
1
71^71:-
-V--
^ . , i _,
:->_^
. : J — ,/
ZT r ' ' tl
• "\~S
--- - /-
1 /- —
/ - -
. ~~'.
;.::
_ -— «-
—
i
r~ 1
^^ ;
--t-
f i :-:
I -.
' —
•:''
Jr|
i
"~~"'
— *.
I -r, :
•::r
--
—
.....
:=.
r -
0
..„
p
z^n
.-::r|
~-
—
» i - -
—
1 :
f— - -
zrrr
—
t:- -
--— -
- ' t -
-itn.
!~* *
, —
r—
-
rzr~~r
_ _^.
Supersaturation - araagr) /Ks
, . . | . ,.. j . . . , f. , | . . . j, , , , j. . . . | . .,.(
...._„..., ,.,.., .—.,..-„ „..» ..., ^^-...^
J_«X X • ^- J. • ^J
1
t-,
.. —
— —
-- —
— rr*
•— -
-—
^—-
,=
. —
t---
k— -
-•-
•— -
.....
' r
Hi
| • ' • • • -•] i ' • r |
—
tTr-
t..i _
—
vn
:r-)
— ^
'
— t- -
1
"
—
~~
.-•1
!i:
. . 1 1
rxi.
_^_
~
t —
*
1
1
-
^
:
.— ,
, i . -ii •'.
::
:.--
~— Z
—
-r~
=
1 1
- —
~ -
_|
p
__.
--
~ .
• —
:" |
. ;. i
-r-
- — _.
' i •
ft!!
~~~
-:
t .— —
• •
.
::,
•~
-
- • —
- — .
— .
-— |
m;:
•
,...
—
h__.
~:_:
r_ -
._.
.::-
.. .
-
:=-
-
—
:.-.
:;•
'
_-;-:.
•— ;
- — -
_
i1'
i
." —
I--
t-"-
~.~7
1
.:::
-
."4 ^"~ "1.5 ""
1008
-------
to
5.
g
i
CO
w
p
O
1009
-------
Rrdian Corporation
3MOAL CPi LK BLVC • P il BOX 1715 • AUSTIN T'/AS 73757 • TE LE°HC NT 5' 2 4SJ 9535
supersaturation with respect to calcium sulfate and ra be the
relative supersaturation with respect to calcium sulfite, i.e.,
'SpCaSO
3
respectively. Using analytical chemistry methods described by
K. Schwitzgebel (Paper No. 9a) the molalities at appropriate
points in the system were measured. Using a programmed equili-
brium model, activities of the ionic species were calculated and
it was thus possible to describe r: and ra in any piece of
equipment in the system. These values of rx and rs are given
in Figure 3-1.
From the figure it can be seen that the slurry from
the scrubber is caught in the hold tank. In the hold tank rl
and Ty decrease (de-supersaturation occurs) . The hold tank
overflows into the delay tank. Here further de-supersaturation
occurs and rx becomes almost unity. The highest value for r1
occurs in the clarifier where there is a relatively long hold
time in contact with the air and the solids settle out. The
reason for this appears to be that the clarifier is a poor
liquid-solid mass transfer device. The total sulfur in the
clarifier (sulfite plus sulfate) stayed essentially the same
which indicates that little or no precipitation occurs. However,
since the solution is in contact with a large air volume for a
relatively long time period, oxidation from sulfite to sulfate
occurs. Therefore, rx increases and r2 decreases. The same
events apparently take place in the limestone feed tank where
the total sulfur remained essentially constant but r1 increased
and r2 decreased.
1010
-------
p^^:~.
B''X?948 • AUSTIN. TEXAS 73757 • IFLEPHONr 5l2 <5-f 9535
From Figure 3-1 it can be seen that rx for calcium
sulfate stays below the critical value of 1.3 to 1.4 in every
part of the system. On the basis of these values of the relative
supersaturation we would predict that calcium sulfate scale
formation would not take place. A possible exception would be
the clarifier outlet. The pilot plant observations are in
agreement xvith our predictions in that no sulfate scale forma-
tion occurred.
Since the laboratory investigations for the calcium
sulfite system have not yet been conducted there is no critical
vnluc of r_ with v.Mch to compare the pilot plant values. The
pilot plant data showed however that a value of r2 up to at least
8 can be tolerated without the formation of calcium sulfite scale,
In addition there is probably some value of Ty above which sul-
fite blinding occurs in the system.
4,0 SUMMARY
Laboratory investigations and pilot plant observations
indicated that sulfate scale-free operation will occur at values
of relative supersaturation of less than 1.3. Pilot plant
observations also indicated that sulfite scale-free and blinding-
free operation occurs at values of relative supersaturation with
respect to calcium sulfite less than 8.
These types of information are required for use in
the design procedure. After heat and material balances have
been made throughout a system, the scaling potential indicators
can be used to predict operability with regard to scald formation.
1011
-------
Radian Corporaticn
~;i • AUSTIN Ti/AS 78757 • TELEPHONIST 454-953$
ACKNOWLEDGEMENTS
This work was funded by the Tennessee Valley Authority,
Division of Power Research,
1012
-------
SAMPLING AND ANALYTICAL METHODS
J.A. Dorsey, Chairman
Participants:
Klaus Schwitzgebel
E.A. Burns and A. Grunt
Terry Smith and Ronald Draftz
Terry Smith and Hsing-Chi Chang
R.M. Statnick and J.A. Dorsey
Gene W. Smith
10.13
-------
SUMMARY
SAMPLING AND ANALYTICAL METHODS
J.A. Dorsey, Chairman
The concluding session of the symposium dealt with the measurements
programs developed primarily for the EPA prototype scrubber tests at the T
Shawnee Steam Generating Station in Paducah, Kentucky. The methods are
also applicable to other lime/limestone process development studies.
The EPA program is designed to acquire extensive data on the composition
of the process streams under varying operating conditions. The data will
be utilized to perform a complete evaluation of the process chemistry,
define operational problems, and model the process. This program results
in a requirement for extensive sampling and analysis with a high degree of
accuracy. While the basic chemistry of the process is rather straight-
forward, in actual operating practice the system presents a three-phase
system that does not achieve equilibrium in the scrubber pass. This failure
to achieve equilibrium (coupled with reactions producing undesirable
soluble species, side reactions producing shifts in the sulfite-
sulfate oxidation rates, and potential scale-producing species)
presents significant sampling and analytical difficulties.
Sampling problems were discussed in the papers by K. Schwitzebel and
by G. Burns as they relate to separation of the liquid and solid
phases in the unstable slurry from the scrubber downcomer. Several
techniques were devised, one based on centrifugal separation followed by
filtration and one employing only filtration. Both of these techniques
provide separations in less than 15 seconds. Sampling techniques for
gaseous and particulate species in the gas phase were discussed in
papers by R. Draftz and by R. Statnick.
Characterization of the slurry components requires analysis of nine
ionic species in both the liquid and solid phases. This produces a labo-
ratory load of over 450 analyses per day. K. Schwitzgebel discussed accu-
rate referee methods and manual field methods for the required analysis.
G. Burns presented evaluations of instrumental methods and development of
on-line monitors for the slurry solids and liquor samples.
The gas-phase analysis of sulfur and nitrogen oxides was discussed
1014
-------
by R. Statnick. An evaluation of instruments suitable for continuous
monitoring was presented. The development of a manual-size selective
particulate sampling system for suspended particulate in the gas phase
was described by R. Draftz.
Finally, a discussion of the proposed EPA methods for new source
performance standards was presented by G. Smith. These methods or an
equivalent will be required to define compliance with emission standards.
Hence, anyone developing control systems should incorporate similar tests
into his program,in addition to the tests being used for engineering
analysis. A copy of the December 23, 1971, Federal Register, containing
the final version of the methods, is included with the symposium
papers.
1015
-------
-------
• • ^^ M^
B ^^ • •
8500 SHOAL CREEK BLVD. • P. O. BOX 9948 • AUSTIN. TEXAS 78757 • TELEPHONE 512/454-9535
DEVELOPMENT AND FIELD VERIFICATION OF
SAMPLING AND ANALYTICAL METHODS
FOR SHAWNEE
By:
Klaus Schwitzgebel
Presented at:
SECOND INTERNATIONAL LIME/LIMESTONE
WET SCRUBBING SYMPOSIUM
8-12 November 1971
Sheraton-Charles Hotel
New Orleans, Louisiana
Sponsored by:
Environmental Protection Agency
Office of Air Programs
Division of Control Systems
1UI7
;HEMICAL RESEARCH • SYSTEMS ANALYSIS • COMPUTER SCIENCE • CHEMICAL ENGINEERING
-------
Radian Corporation
8500 SHOAL CREEK BLVD • P. O. BOX 9918 • AUSTIN, TEXAS 78757 • TELEPHONE 512 - 454.9535
1.0 INTRODUCTION
The work presented here describes the analytical and
sampling techniques for the forthcoming test facility at Shawnee.
One of the main objectives is the collection of engineering
design information for lime/limestone based SOE removal processes
Therefore, the demand for accuracy of the analytical chemistry
methods is more stringent than the demand for accuracy in control
processes.
The problem areas in analyzing unstable slurry streams
are sampling, sample handling and analysis. Two kinds of
analytical methods were selected, referee methods (used also as
back-up) and rapid field methods.
The ultimate use of the analytical results is for
chemical engineering purposes. The chemical engineer uses
the data to describe:
vapor-liquid mass transfer characteristics
in the scrubber
solid-liquid mass transfer rates throughout
the system
scaling potential
A mathematical description of the reaction kinetics
of the mass transfer steps as a function of the liquor composi-
tion is a prerequisite for the process engineering of limestone
based sulfur dioxide removal processes. The driving force term
in these rate equations is a function of the difference between
actual and equilibrium conditions. The driving force term for
1018
-------
Radian Corporation
8500 SHOAL CREEK BLVD. • P O. BOX 9948 • AUSTIN, TEXAS 78757 • TELEPHONE 512 - 454-9535
liquid-solid mass transfer is a function of the difference
between actual and equilibrium activities. The driving force
term for gas-liquid mass transfer is a function of the difference
between actual and equilibrium vapor pressures.
The solutions are non-ideal. Concentrations are not
suitable quantities for describing the driving forces. Thermo-
dynamic concentration, or activity, must be used. Only the
total amount of a species in solution is measured by chemical
analysis. The activities of the species of interest differ
markedly from chemical analysis values due to complexation and
the deviation of the solution from ideality. An example is
given showing how the activities of the important species can
be extracted from the results of the chemical analyses.
2,0 PROBLEM DEFINITION
The basic equipment arrangement for limestone
injection wet scrubbing (LIWS) processes is shown in Figure 1.
The three streams entering the system are flue gas, particulates
and make-up water. Three streams leaving the unit are cleaned
stack gas, solid waste products, and scrubbing liquor. The
composition of the incoming streams provides a means of predict-
ing the liquor composition on a qualitative basis. The important
species in the LIWS process are:
Group I Group II Group III
calcium sodium trace elements
sulfite potassium iron
sulfate magnesium cobalt
chloride nickel
nitrate copper
nitrite manganese
carbonate
1019
-------
GAS SPECIES
FG, SG
STACK GAS
SG
WATER
MAKEUP
WM
1. S02 '
2. C02
3. NOX
4. H20
D. U2
6 CO
' ,. SCRUBBER
7. N2 s
1 1
FLUE GAS
FG * A
1
SCRUBBE
BOTTOMS
SB
/
SCRUBBER FEED
SF
y v L
Limr-.'-mara t-rmwr
PROCESS
V/ATER
HOLD TANK
P
SLURRY RECYCLE SR
LIMESTONE
FLY ASH
SOLIDS
LA
r. CoO
2. MgO
3. CaS04
4. MgS04
5. CoS05
6. MgSOs
7. CoC03
8. MgC03
9. FLY ASH
10. SOLUBLE No
II. SOLUBLE Cl
SCRUBBER
EFFLUENT
HOLD TANK
E
CLARIFIER
LIQUID
CL
CLARIFIER
FEED .
CF
CLARIFIER
C
CLARIFIER
BOTTOMS
CB
FILTER
F
FILTER
LIQUID
FL
FILTER
BOTTOMS
FB
PROCESS SOLID SPECIES
(CF.SR.CB, FB, SF)
I. CoO
2. Co(OH)2
3. CoC03
4. CcS03 • xH20
5. CoS04 • xH20
6. MgO
7. Mg(OH)2
8. MgC03-xH20
9. MgS03 • xH20
10. FLY ASH
PROCE^
LIQUID
SPECIE
SB.CF.SR,
FB.CL.FL,
I.H*
2. OH~
3. HSOj
4. SOf
5. SOf
6. HCOj
7. COf
8. HS04
9. H2S03
10. H2C03
II. Co++
12. CoOH"
13. CoS03
14. CaC03
15. ColiCC
16. CoS04
17. CoNOj
18. N03
19. Mg + +
20. MgOH'
21. MgSO/
22. MgHCC
23. MgSO:
24. WgC03
25. No+
26. NoOH
27. NoC03
28. NoHCO
29. NoS04
30. NoN03
31. cr
FIGURE 1 WET SCRUBBING SCHEME
1020
-------
Radian Corporation
8500 SHOAL CREEK BLVD • P. O. BOX 9948 • AUSTIN, TEXAS 78757 • TELEPHONE 512 - 454-9535
The components listed in Group I are the most
important. They dominate the process by participating in the
gas-liquid and liquid-solid mass transfer steps. The species
listed under Group II contribute to the process performance
in three ways. First, they influence solubilities which are
dependent on the ionic strength of the solution. Second, they
form ion pairs with Group I compounds. Finally, they influence
the driving force for the mass transfer rates. The components
in this group form very soluble compounds with the exception of
magnesium hydroxide and calcium carbonate. In a closed loop
operation there is a buildup of the soluble compounds, since the
only stream in which they can leave the scrubbing unit is the
liquor adherent to the solids. This fact must be kept in mind
when selecting analytical methods. The procedures must give
accurate results in those cases where the soluble species build
up to a high level. The implications for the selection of methods
for sulfate and sulfite will be discussed later.
The third group is comprised of species leached
from the fly ash and impurities in the limestone. The concentra-
tion of these elements is never very high, since it is limited
by the solubility of the hydroxides in the alkaline parts of
the scrubbing unit. Their importance is based on the fact that
they are excellent catalysts for sulfite oxidation, even if
present in the parts per billion range.
The process simulations, discussed earlier by
D. M. Ottmers, Paper #lc, gave a valuable basis for estimating
anticipated concentration ranges. Estimation was necessary
since no data on a closed loop system operated over an extended
period of time were available at the time of analytical method
development.
1021
-------
Radian Corporation
8500 SHOAL CREEK BLVD. • P. O. BOX 9948 • AUSTIN, TEXAS 78757 • TELEPHONE 512 - -64-9535
As a general rule, the higher the accuracy demand
of an analysis, the higher are its costs. This fact raises the
question as to the ultimate use of the analytical results. The
accuracy requirements for routine, day-to-day operation are
less stringent than the requirements for process analysis. One
key objective of the forthcoming tests at Shawnee is the collec-
tion of engineering design information. From an engineering
point of view the following areas are of ultimate interest:
gas-liquid mass transfer rates in
the scrubber
dissolution and precipitation rates
as function of liquor composition
scaling potential.
The driving force term in the mass transfer equations
describing these rates is a function of the difference of the
actual process condition and the equilibrium condition of the
system„ In other words, the rates are a function of the
difference of two activity expressions. The closer the system
operates to equilibrium the more severely analytical errors will
influence extracted rate correlations. For LIWS processes the
analyses of the species listed in Group I are therefore the most
important. Error propagation calculations showed that the error
in these analyses should not be greater than about 2%. The con-
centration of the species influencing the ionic strength (Group
II) must be known within about 4%. The accuracy requirements for
the trace elements effective as catalysts are still less stringer
Twenty to fifty percent is considered to be sufficient.
1022
-------
Radian Corporation
8500 SHOAL CREEK BLVD. • P. O BOX 9948 • AUSTIN, TEXAS 78757 • TELEPHONE 512 --154 9535
The analytical results are influenced by three steps:
solid-liquid separation
sample handling
actual analysis
These problem areas will be discussed next.
3.0 SAMPLING
The scrubbing system can be divided into an acidic
and a basic part. The environment is acidic in the scrubber
itself and in the pipe between the scrubber and the effluent
hold tank. The solutions circulated in the rest of the system
are alkaline. For sampling purposes it should be noted that
the scrubbing slurry, especially in the acidic part of the system,
is not in thermodynamic equilibrium. The sorbent tends to dissolve
and sulfite and sulfate tend to precipitate. The technique often
used to sample this stream is collection of a slurry sample in a
beaker and filtration through a Buchner funnel. This technique
results in only semi-quantitative results for the follow-on
chemical analysis for three reasons:
loss of acidic gases (SO,, , C03)
especially if a vacuum is used
solid-liquid mass transfer during
the sampling procedure
sulfite oxidation by air oxygen.
Because of these sources of error most of the pilot plant data
presently available must be considered to be qualitative in nature
and not suitable for the extraction of engineering design informa-
tion. In-line, positive pressure filtration was the sampling
1023
-------
Radian Corporation
8500 SHOAL CREEK BLVD • P O BOX 97-18 • AUSTIN TEXAS 78757 • TELEPHONE b!2 4549535
method selected after field tests at several pilot units (see
Figure 2). The sampling apparatus consists of a positive pres-
sure pump, a membrane filter holder and lines and valves to
control sampling and purge rates. Flow rates used in the tests
were about 1300 ml/min. The residence time of the slurry is
about 2.3 seconds in the filter and approximately seven seconds
in the entire sampling equipment.
The degree of mass transfer in the filter cake, which
is by nature a good contacting device, was checked by taking
consecutive samples and plotting the chemical analysis results
as a function of the filtered volume. Extrapolation to zero
volume of filtrate represents the true aqueous phase composition.
With the exception of carbonate, the amount of solids dissolved
or precipitated in the filter cake was within the experimental
error of the chemical analyses.
Loss of acidic gases is avoided by the positive pressure
filtration, and air oxidation of sulfite is prevented by fixing
the sample immediately.
4.0 FIXING OF THE FILTERED LIQUID
After filtration care must be taken that the liquid
samples do not undergo further change. This is especially true
for the sulfite analysis. Sulfite losses can occur by :
evaporation from acidic samples
oxidation by air oxygen
interaction with nitrites
Nitrites can be formed by absorption of NO and N0a from the flue
gas. All three sulfite losses can be avoided by quenching the
sample in a solution of pH = 6 with knoxvn iodine content.
1024
-------
J-l
cu
O
a
0)
4J
-00-
B
CO
•H 0)
•H 3
tO CO
O to
P-i CU
J^
P-I
a
CU -rl
B
H
En
PI
•H
ctf
^
H
bO
C
•H
!-l
Cu
S
C(J
00
CO
ciJ
O
•H
CO
1025
-------
Radian Corporation
8500 SHOAL CRLTK BLVD • P. O BOX 9948 • AUSTIN TEXAS 78757 • TCLEPHONC 51? •!5'i-9535
Carbonate losses from acidic liquid can be avoided
by quenching the sample in a solution of pH = 10. EDTA must
be added to the buffer in order to avoid calcium carbonate
precipitation at this pH.
Sulfate in the presence of sulfite is determined
as the difference between the total sulfur and the sulfite
sulfur. In order to avoid sulfite losses and sulfate precipi-
tation the sample for total sulfur analysis is quenched in a
HP Op-water solution. Hydrogen peroxide oxidizes the sulfite.
Dilution with distilled water prevents sulfate precipitation in
the sample bottle.
5 • ° LTQUID PI-IASE ANALYSIS METHODS
The literature was surveyed through 1970 for analytical
methods which might be applicable to solutions of interest. The
sources consulted were:
Kolthoff and Elving, "Treatise on Analytical
Chemistry"
Biannual Reviews on Analytical Chemistry
1969 Book of ASTM Standards
FWPCA Methods for Chemical Analysis of
Water and Wastes
Chemical Abstracts
Pertinent Original Articles
1026
-------
Radian Corporation
8SOO SHOAL CRfcEK BLVD. • P. O. BOX 9748 • AUSTIN, TEXAS 78757 • TELEPHONE 512 - 154-9535
Promising methods for application to the analysis of
wet scrubbing liquors were checked in the laboratory and through
visits to manufacturers. Two types of analysis methods were
sought: referee (and back-up) methods and rapid routine procedures
5.1 Back-Up Methods
The literature review revealed that the choice of method
for sulfate, calcium and sulfite in the key or Group I species is
rather limited. The methods published for sulfate analysis can be
divided into five groups.
1. Gravimetric Procedure
2. Direct Titrimetric Procedures
3. Indirect Titrimetric Methods
4. Colorimetric Techniques
5. Turbidimetric Procedures
All these methods, with few exceptions, are based on
the formation of insoluble barium or lead sulfate. They all,
therefore, show the same potential interferences, namely, copre-
cipitation errors, occlusion of foreign salts, and errors due to
supersaturation ; The techniques most widely used are the gravi-
metric method (ASTM referee method), direct titration in a water
ethanol mixture using 133(010,,.);, or BaClP as titrant and thorin
as end point indicator, the barium chloranilate method (FWPCA
method), and the turbidimetric technique.
The gravimetric procedure was rejected for two reasons.
First, it is extremely time consuming, and second, there are
interferences expected in scrubbing solutions with high salt back-
ground. Alkali metals cause errors due to occlusion. Calcium
causes serious errors due to coprecipitation. Nitrate is reported
1027
-------
Corporation
flOO SHOAL CRICK BLVD • P O BOX 99-53 • AUSTIN. TEXAS 7P757 • TELEPHONE 512 - 454 9515
to cause errors by occlusion. The occlusion and coprecipitation
problems due Co cations can be avoided by use of ion exchange
resins.
The direct titration using thorin as end point detector
was rejected due to severe anion interferences. Figure 3 shows
the errors caused by several common anions.
FIGURE 3
Reference: Fritz, J. S.,
and S. S. Yamamura, Anal.
Chcm., 27_, 1461-1464~(T9"55)
il'B of tilfntio'i of ?,n!f,'i!o in presence of
coincorUralH-tis at cotiVJion among
Compensation for these anion interferences can be made by
standardizing the titrant solution with a Hr SO., standard con-
taining the foreign ions at a concentration corresponding to
that of the unknown solution. This technique may be useful in
routine analytical work, but is not acceptable for sulfate
determination in varying environments.
1028
-------
Radial! COTOrStiOn SMO SHOAL CRUKBLVD • PO.BOXWB • AUSTIN, TLXAS 7875? • TELEPHOW bi? - in-9535
The turbidimetric sulfate determination is recommended
by ASTM mainly as a control procedure where concentration and
type of impurities present in the xvater are relatively constant.
Kelly and Baldwin [Chora. & Ind_. , 1283-1285 (1969)] automated
this technique using an autoanalyzer. They report results more
consistent than with manual operation. Compared with gravimetric
techniques, however, they found deviations of ± 10%, which is
unacceptable.
Extensive laboratory effort was devoted to the barium
chloranilate method recommended by FWPCA. The laboratory results
revealed ni trate and chloride interference if these anions are
present at higher concentrations as well as a critical dependence
on pll. In addition, inconsistencies were found if different
batches of reagents were used. The time required for complete
reaction and precipitation to take place as well as the time
required to separate the very fine barium sulfate precipitate
from the acid chloranilate solution gave very little hope for com-
plete, fast automation of the method for very accurate determinations
These precipitation difficulties were one of the main reasons that
the barium chloranilnte method was also abandoned by Technicon.
The method ultimately adopted is an ion exchange
alkalimctric procedure. Sulfate in aqueous solutions is deter-
mined as sulfuric acid after passage of the sample through a
hydrogen form cation exchange resin. The aqueous acid mixture
obtained after the cation exchange is evaporated to a few milli-
liters on a steam bath or a hot plate. After this step all the
acids a.nd the water are driven off by evaporation at 75° C. Only
1L SO and other nonvolatile acids such as H5rO.: remain. If
]J:,P(\ is absent, the sulfuric acid can be titrated directly.
If II,PO. is present, !L S0: is driven off at 275°C and the H3PO.t
.is determined by alb :liircl;ri c titration. The method is free
(: r om c a I: i o n i n t e r f c r < • n c e .
1029
-------
Radian Corporation
8500 SHOAL CRCEK BLVD • P O BOX 9943 • AUSTIN, TEXAS 78757 • TELEPHONE 512 454 9535
Phosphoric acid presents the most severe anion
interference. This acid must be determined separately if it
is present in large amounts. Table 1 compares the results
obtained with the volumetric method to the results obtained
with the gravimetric procedure. The method was checked in
Radian's laboratory using high salt backgrounds. Laboratory
results are presented in Table 2. Results in analyzing field
samples agreed within 270 with the X-ray fluorescence technique
discussed later.
The methods for sulfite determination in the presence
of nitrite are also rather limited. The normal iodine thio-
sulfate procedure gives erroneous results due to nitrite-sulfite
interaction at the low pH values used in the procedure. The best
method found was to quench the filtered liquid in a buffer of pH
= 6 containing a known amount of iodine. The back titration must
be done at this pH value with arsenite instead of thiosulfate,
since thiosulfate is partly oxidized to sulfate in the presence
of nitrite at pH =6. A dead-stop technique for the end point
detection was chosen. At low sulfite concentrations this leads
to better results than the starch indicator normally used.
The most convenient method for the determination of
the third species in Group I, calcium, was found to be atomic
absorption. A 5% HC1, 1% LaCl3 solution used to dilute the
samples into the optimum range for A.A. measurement x\?as found
to suppress all the interferences.
Table 3 summarizes the referee methods found to be
most suitable for scrubbing liquor anlaysis.
1030
-------
Radian Corporation
P500 SHOAl CRf [K BLVD
O. BOX 99.48
AUSTIN, TEXAS 737S7
TLLLPMGNC 512
TABLE 1
GRAVIMETRIC AND VOLUMETRIC DETERMINATIONS OF SULFATE
o
Sample
1
2
3
4
5
6
7
8
9
10
11
12
Sea Water
(one sample)
Sexv/age0
(one sample)
IN VARIOUS SAMPLES
Sulfate Found, n
Gravimetric
427
421
391
188
167
161
104
96.5
87.5
60.2
39.4
27.0
2,650
2,640
2,640
125
124+
124
Volumetric
426
419
390
188
165
162
101
95.7
87.0
59.7
40.1
26.4
2,630
2,630
2',630
124
124
124
a. Natural waters (surface streams, wells, reservoirs, etc.)
b. Diluted 1 to 50 prior to ion exchange.
c. Filtered and broininated prior to ion exchange.
1031
-------
TABLE 2
DETERMINATION OF SULFATE IN LIMESTONE
INJECTION SIMULATION SOLUTIONS USING DIFFERENT RESINS
Exper.
t
1
4> U> N>
Amberlite
5
6 x
7 1
Q
8
9
10
X
Uu
f*j
13
SAMPLE
10 ml Simulation Soln.*
n n
" plus H202
11 plus Na2Si03
10 ml Simulation Soln.*
it n
" plus H202
11 plus Na2Si03
10 ml Simulation Soln.*
n ii
11 plus HS02
11 plus Na2S103
10 ml Pure K2S04 Soln.
ml 0.0502N
NaOH used
3.92
3.94
3.96
3.95
3.95
3.95
3.96
3.96
3.95
3.96
4.04
3.96
3.97
m mole
Theory
0.0990
n
it
tt
it
11
n
it
it
it
11
11
2 S04
Exper.
0.0984
0.0989
0.0994
0.0991
0.0991
0.0991
0.0994
0.0994
0.0991
0.0994
0.1014
0.0994
0.0996
Percen
Error
-0.7
-0.1
+0.4
+0.1
+0.1
+0.1
+0.4
+0.4
+0.1
+0.4
+2.3
+0.4
+0.6
Resin column dimensions: 1.2 cm I0D. x 18 cm high.
Resins used: Amberlite CG-120, Dowex SOW, and Rexyn 101.
All were 100-200 mesh size.
Simulation solution contained:
0.009904 M K2S04
0.150 M Ca(N03)2
0.100 M NaCl
0.05 M HC1
Samples 3, 7, and 11 contained 2.2 m moles H202 which was added as 5 dro
of 30% solution.
Samples 4, 8, and 12 contained O.lm rnole Na2SiOa which was added as 1 m
of 1 M solution.
1032
-------
•(
•1
D
y
3
E
P
E
•H
4-J
CX
O
h
o
cd
5-J
p
o
o
<
d
O E
•H a
4-J CX;
CO
5-1 CJ
•!-) K
£ C
QJ ftf
O 5-1
d
0
o
G
o
•H
4J
cd
4-J
a!
o;
Pi
P
r4
•I-J
C:
f ;
H
o
4J
CD
^
CD
CD
5-1
CD
CD
&
B-? B^ ^S
cn m CN
in
•
rH VO P-.
VI VI VI
1 1
0 0
4J 4J
O <3
rd co
CM ^ H
O W O
5-1 0- U
4~> n P
O C\; rH
CD t-i •.-(
CX P,
CO *
r-l CJ
• Q) T<
J-i
O
1
•H
rH
cd
*^
rH
4-J
0!
P
r/j
d
0
•H
CO
o
•H
£_l
4-J
CD
E
O
4-J
co cd
C M
E 4J
P -H
rH -!J
• O
O O
•H
0) 4-J
^ ') Cj
c. E
cj o
.a 4.1
0 p
;>•' (vj
Wv_x
0
O
rH
r4
QJ
4-J
d
M
O
•H
£
4J
Q)
E
P
5-i
4-J
co
d
O
O
1
0
o
cn
4-1
(I)
5-J
P
pq
CXM
rj
!-i
bO
0
d
c j
CJ
o
o
T3
O
r]
4J
Q)
)2^
d
o
•H
4J
P
rH
O
J>
w
rH
Cd
C
O
•H
4-1
Cj
d
d
o
•H
CO
•r-l
Q
0
5-1
PM
|
nj
4-J
•r-)
H
O
•H
5-1
4-1
CD
E
•H
P
O
5^
CD
f£
0
4-J
cd
5-i
4J
•H
4J
O
•H
4-.J
1 j
E
O
4-J
rj
cj
*-^
^_^
O
•H
5-t
-P
Q)
r"
P!
O
•H
4-J
d
CJ
4-J
O
CX
^
Ci
0
•H
4-J
in
CN
i
m
CO
P
4-J
cd
5_(
CO
CM
•<]
co
CO
«5
rH
O
T3
d
CO
m
ffi
21
O
4J
d
o
•H
4-J
O
r^
T3
CJ
&4
Cu
E
CO
d
•H
bO
^
d
o
•H
4J
rd
rH
rH
•H
4-J
CO
•H
Q
O
O
o
o
o
rH
QJ
4-J
QJ
E
O
4-J
O
cx
o
5-1
4J
CJ
QJ
CX
CO
0
•H
5-i
4J
QJ
E
•H
5-1
O
rH
O
CJ
M
0
•
^>
•
£3
0
•H
a
o
5-1
4-J
o
s
d
!_)
r\
O
rd
4-J
•H
5
B-S
CO
O
o
o
o
o
rH
0)
4-J
QJ
E
O
a
o
5-1
4-1
O
QJ
CX
CO
O O
CN r-l
,0
^n pt
CX CM
CX
O
OO
CN CN
I I
m o
co
cx cx
§§
r-lrJ
o o
•U 4J
CM CM
r4 5-1
O O
CO CO
O O
•H-H
S E
O O
-U 4-1
p
O
CO
CO
OJ
•H
CJ
C)
fl
CO
+ rcj bO
^ U >^
(V
O
CO
cn
O
CO
re
O
nj
4-J
O
H
l n
O
O
co d
ctf
4J «
CD QJ
QJ «•>
O P-H
5-J
H
1033
-------
ssoo SHOAL CKCCKBLVD • p. o. BOX 99<8 • AUSTIN, TEXAS 73757 • TELEPHONE 512- 454-9535
5 .2 Routine Field Methods
The selection of routine field methods was based on
the type and number of analyses per day required at peak load
operation „ This breakdown is shown in Table 4. Table 5 gives
the time requirements per day in the event that the back-up
methods are used. Duplicate analyses were assumed if not in-
dicated otherwise. The last column in Table 5 shows the total
man-rninute/day necessary for each type of analysis. The costs
show the following pattern: Total S > Ca > Mg > S02 > COS > Cl
= K, with total sulfur being the most expensive determination.
X-ray fluorescence appeared to be a technique which
cut the expenses drastically. However, no data were found in
the literature describing the use of this technique in analyzing
liquors of the composition encountered in lime/limestone based
scrubbing processes. Figure 4 shows the principle of this tech-
nique. The specimen is radiated by a primary X-ray source. The
elements present in the sample emit characteristic secondary
emission lines whose wavelengths and intensities are measured
using an analyzer crystal and a counter.
Quantitative X-ray fluorescence analysis is subject to
interferences as are most of the other analytical procedures.
The intensity of the emission line of an element can be reduced
or increased by the other elements present in the sample. An
increase is observed if secondary excitation occurs. A reduction
of intensity is caused by absorption effects. Correction factors
must therefore be determined and the measured intensities correctei
Tables 6 and 7 present preliminary results in analyzing
simulated scrubber solutions for sulfur and calcium. The RMS
errors for the two most important key species are quite acceptable
Preliminary results indicate that chlorine, potassium, and magne-
sium also can be determined by this technique.
1034
-------
Radian Corporation
8E.OO SHOAL CRtTK BLVD • P O BOX 99<3 • AUSTIN, T[XAS 78757 • TELEPHONE 512 - 451 9535
TABLE 4
Species
Ca-H-
Mg4^
K+
Na+
Total S
SOS
Cl"
CO,
Total N
N0~
NO;
NUMBER OF ANALYSES PER DAY
Number of Analyses
Sought at Steady State
53
48
9
9
53
36
9
53
9
9
9
AT PEAK LOAD OPERATION
Analyses During
Line Out
27
27
27
27
27
27
Total
80
75
36
9
80
63
36
53
9
9
9
459
1035
-------
m
w
rJ
PQ
co
Q
O
ffi
w
w
Pi
w
w
Pi
p
(H
C/J
CO
H
w
s
o-
w
Pi
w
M
H
Q
W
H
(H
fc
W
K*1
CO
Q
^~^
«
d
•H
^Ej
i
C
H
tH
4J
O
H
o o o o o
r^ oo o LTI o^
CM O •
cO
Q
CO
OJ
CO
>-
CO
c
^
c
•H
CO
•H
CO
fx.
tH
cd
d
QJ
E
•H
H
13
O
4J
QJ
S
4-J
bC
O
CO
CO
0)
•H
O
cu
0
CO
00 CN O O H 0)
CO T3 CO 13
V— / \~-S
O MD CM O> CTs
VD o r*^
rH rH
o m m o o
CO rH rH CO CM
13 /-N
0 0
& -H 2 d OJ
4-J O £-1 O3
0 O 0) -H 4J O -H cr1
bC-f-< S M OJ 4J 4J -H
d J-i 4J E co d
CTJ4J d QJO drH X
^
O i — 1 I> CJ CL< QJ •
H,
~-^ CO
Q
d ^^
S M
D
0 0
rH ffi
00 1
d
m cO
rH S
m
•
CO
CM
j_l
O
1036
-------
"nary X-ray Beam
)n!inuum + Anodo Characteristic)
Secondary Emission Lines
(Sample Characteristic)
ay Tube
velength Measured, X
Analyzing Crystal
Changer
Flow
Proportional
Counter
>29
-Scintillation
Counter
v A = 2d-sin0
FIGURE 4
1037
-------
Radian Corporation
Sample
Number
14
15
13A
16
17
18
19A
20
2-1
22
23
24
25
26
27
28
29
30
8409 RESEARCH BLVD. • P O. BOX W8 • AUSTIN, TEXAS 78758 • TELEPHONE 512 - 454-7S35
TABLE 6
Results of Sulfur Analyses
Nitrogen and
Corrected
Magnesium Interference
for
Only
Corrected Values
Sulfur Content
(mmoles/jO
25
25
25
25
25
25
50
50
50
50
50
50
50
50
50
50
50
50
Instru-
ment I
25.3
25.2
24.7
24.6
24.9
25.2
48.9
49.2
49.0
49.2
49.7
50.2
49.6
49.9
49.7
50.3
50.4
51.4
RMS =
% Error
1.2
0.8
-1.2
-1.6
-0.4
0.8
-2.2
-1.6
-2.0
-1.6
-0.6
0.4
-0.8
-0.2
-0.6
0.6
0.8
2.8
1.8
Instru-
ment II %
25.5
25.1
24.8
25.6
25.5
52.2
50.6
50.0
51.8
49.9
49 „ 4
50.8
50.0
50.9
49.9
53.0
RMS =
Error
2.0
0.4
-0.8
2.4
2.0
4.4
1.2
0.0
3.6
-0.2
-1.2
1.6
0.0
1.8
-0.2
6.0
2.4
1038
-------
Corporation wot RESEARCH BLVD. . P.O. BOX r>48 . AUSTIN, TEXAS 73753 • TELEPHONE $12 • 454-9535
TABLE 7
Results of Calcium Analyses Corrected
Sample
Number
86
77
78
79
80
81
82
90
91
92
Nitrogen
Chlorine,
Calcium Content
(mmoles/ A)
10
25
25
25
25
25
25
50
50
50
, Magnesium
and Sulfur
Instru-
ment I
10.0
25.0
25.0
25.0
25.0
25.0
25.0
50.3*
50.2*
50.0*
, Potassium,
Interferences
Corrected
% Error
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.6
0.4
0.0
for
Values
Instru-
ment II
9.7
25.0
25.0
24.9
25.0
24.8
24.6
50.2
49.8
49.2
70 Error
3.0
0.0
0.0
-0.4
0.0
-0.8
-1.6
0.4
-0.4
-1.6
RMS = 0.2 RMS =1.3
•*•
50 mmoles/4 Calcium values exceed linear range
1039
-------
Radian Corporation
SHOAL CRT! K BLVD
AUSTIN, TLXA5 78757
TEILPHONL '
Table 8 compares the results of the proposed referee
method and of X-ray fluorescence in analyzing total sulfur in
samples taken from different streams at TVA's Colbert Station
pilot plant. The agreement can be judged as being good.
A big advantage of X-ray fluorescence not yet mentioned
is the speed of analysis, especially if a minicomputer is used to
perform the matrix interference corrections. The estimated saving,
in time for the Shawnee tests are reflected in Table 9. The man-
minutes/day are cut by a factor of four as compared to the referee
methods.
6.0
SOLID ANALYSIS
The solid analysis comprises three steps:
phase identification
solids dissolution
analysis of the liquid phase.
The phase identification of crystalline compounds
collected on the filter previously described is most conveniently
done using X-ray diffraction. Important solid species (other
than fly ash) potentially present are:
Magnesium Compounds
Calcium Compounds
1.
2.
3.
4.
5.
6.
7.
8.
9.
CaO
Ca(OH)
CaC03
CaC03
CaS03 •
CaS04 •
CaS04 •
g-CaSC
v-CaSC
2
(aragonite)
(calcite)
%H3 °
2H20
%HaO
'4
1.
2.
3.
4.
5.
MgO
Mg(OH)s
MgCCv
Mgso3- :
MgS03-
1040
-------
col
w
Q
i-3
W
M a\
PM M
O H
CO
col
DO W[
y
.-)
pq
<£
H
Kl
t
H H
W
«
-i
i—i
O
o
H
53 HI
O
c
0)
M
(!)
m
m
•M
Q
E-£
cu
o
c
0!
CJ
W
0) ^
i , «.
M ^^-
O W
P CD
i—4 t— 1
Pn O
xi
!•— < f3
"3
f^
><
w
•r4
W
>i
•^ di
^*-^ r*
^^~ \-4
W <|
cu
r-4 r-4
O Ctf
6 o
C3 . 1
vS- P
f-i
CO 00 00 O> CM
vO CO v£> vO ^- ^~ 4J
•H ~^ \
rQ H J^ CO
C
Pi O O P4 cd Pi!
> H H 0 H 0
/*~N /^~S X~S /^^\ X*^S X*^>
r-l CM CO
-------
CTi
w
h-1
ft-
%
1 — 1
rt
G
rd
0)
•M
Cd
O
rH
3"
T3
j^
c;
>
0)
£
*
CO
0)
s
rd
H
• •
co
•H
CO
rd
s
Ctj
Q
^
G
•r-
1
d
rt
S
r-l
K(
t>
O
H
>*,
rl
O
~--~.
co
CD
CO
,">
r-l
$
d
<£.
co
•r-l
CO-
pi
r--!
rj
CJ
•^
•
G
«r-|
^£j
1T3
0
rC.
CD
r"* -i
T3
r-l
0
•r-
f^H
4-|
'oi
b
CO
CO
CD
•H
O
D
CO
O O O
00 CO VO
rH CM ~,
G I >,^i CO
O B r-l rt
3 rj w T3
T) kl G co G
CD <3 ^ (U 6
r: G Ki cj 'o o
O O PH CD f
rH G rH \-
W) H W) 0
CO) C 4-
••H 4J -H Q
CO 0) CO T
^-^TTJ "•— '
vO O> CTv O
O
rH
f
O O O
rH CO CM
f"
cu a
'U p ^
c G cr c
Cd O '>H 'f~
•rH C C
C C 4J ,G ,E
o o rt cj c
•H -H rH a; a
4-1 ID 4J rH H t-
2 0 CJ-rH
rH rd D 4-1 •
O 4-' "O CO > >
> JU CD -H • __
2
rH
rt
o) 4J 1 rel
O 0 O C
CJ H ^^
0} £>-i
Q cri
\ Q
• "^^^
G co
3 D
O
O ffi
CO I
CM G
st" 1
|J-V
,
O
N
* J— i
: o
^
4
j
j
j
3
N
)
J
r
1
4
)
)
\
•
•0
1042
-------
Radian Corporation
8500 SHOAL CREEK BLVD • P O BCX 97-13 • AUSTIN, TEXAS 78757 • TELEPHONE 512 454-9535
The characteristic diffraction patterns for these compounds
are listed in the "Inorganic Index to the Powder Diffraction
File".
The diffraction patterns of the collected solids can
be obtained using a film or a goniometer technique. The princip]e
of these methods is outlined in Figures 5 and 6.
The next stop in solids analysis is the chemical
analysis of the individual compounds. Carbonate is determined
on a weighed sample by the evolution technique. Sulfite is
determined by dissolving a x\7eighed sample in an iodine solution
of known iodine content. Calcium, magnesium and total sulfur
are determined in a sample dissolved in a weak mineral acid con-
taining HP0S for the oxidation of sulfite. The total calcium,
magnesium, and sulfate in solution can then be determined by X-ray
fluorescence or the referee methods described earlier. Typical
analysis results are presented in Section 7.
7•° FINDINGS ON PILOT PLANT STUDIES
The sampling, sample handling and referee methods
were developed and tested by analyzing data from several pilot
units. Samples were collected during
GAP in-house studies
pilot plant runs at the Tidd Plant in
Brilliant, Ohio
pilot plant runs at Key West
pilot plant studies at TVA's Colbert
Steam Plant
pilot plant studies at Shawnee
1043
-------
K3di3P COrpOr3*IOn *w> RESEARCH BLVD •
P.O. BOX ws • AUSTIN, TEXAS 78/ss • TELEPHONE 512 - 454-9535
X-ray Beam
Debye Cones
—Cylindrical Film
FIGURE 5 - Principle of the Debye-Scherrer Method Using
a Cylindrical Camera
X-Ray Tube
FIGURE 6 - Use of the Goniometer Technique to Obtain
X-Ray Spectra from Powders. The intensity
of the reflected radiation is monitored by a
counter and plotted with a strip chart recorder
as function of the angle 29. The specimen is
rotated with half the speed of the counter.
1044
-------
Radian Corporation
foOO St'OAL CPTCK BLVD • P O BOX 9^,3 • AUS1I N. TEXAS 78757 • TFLCPHONI 512 - 454-9535
Results will be presented for the pilot studies at TVA's Colbert
steam plant. The system arrangement at Colbert is shown in
Figure 7. Samples were taken and analyzed at the scrubber
effluent (sample point 2), scrubber spray (sample point 1),
affluent hold tank F-12 overflow (sample point 3), and the
process liquor tank F-13 (sample point 4). The results of the
liquid and solid phase analyses are shown in Tables 10 and 11.
The buildup of inerts is very small in this arrangement since
nost of the fly ash was removed by the raw water spray.
The accuracy of the methods is reflected in the
imbalance .
. . . • z.
/. i i/pos i.\ i i/neg
The pH measurements and analytical results shown in
Table 10 were used to calculate this imbalance. The imbalance
should ideally be zero for zero errors in the analytical deter-
ninations. Another source of ionic imbalance is the presence
of species for which no analysis was made.
The results of the solid phase analyses are presented
in Table 11. The concentrations of the solid species add up to
nearly 10070 with exception of the solids of the scrubber spray
which contains most of the fly ash. Compounds leached from
the fly ash for which no analysis was made may be responsible
for the low values found .
1045
-------
cfl
60
O
O
o
CO
(U
•H
•H
cfl
r-l
O
4-1
O
cy
,0
u
C/3
CJ
a
o
4J
en
cu
1-4
<:
£
i
t^-
O
M
1046
-------
g
H
-4
to
c
rt
J3
E
.0
r-f
O
B
^ ^
w
o
H
I Ci
:) o
eg
o
o^
CXI
CM
O
CS
O
3
o
t/]
c
o
CJ
1047
-------
w
w
CO
M
CO
w
CO
O
M
O
CO
o
CO
EH
v:
K
CO
CM
OO
Pi
4-J
C
r-t
•U
CJ
^3
r-t
O
r-:
:s«
IS
O OO
S33
<-4 OO *4 [
ox".s" S*
1048
-------
Radian Corporation
8 . 0 USE OF THE RAW DATA
It was mentioned earlier that the results of the
chemical analyses have no value per se. They gain their value
in the chemical engineering framework within which they are used.
Dominant points of interest are:
mass transfer characteristics in the
scrubber
solid- liquid mass transfer rates
scaling potential
There Core, the da La presented in Tables 10 and 11 must be
processsecl further, As r-n example, suppose one wishes to predict
the scaling tendency of the scrubber effluent given in Table 10.
This task is solved by considering the ionic equilibria in the
aqueous phase. The results of the chemical analysis listed in Table
10, the pH value and the temperature were used as inputs for compu-
ter calculations. The resulting activities of the individual ionic
specJes for the scrubber effluent are listed in Table 12.
The activities of Ca, SO^" , and SO^ are given as
7.25xlO"3, 1. 22x10-*, and 5.84xl(T3 respectively. The ratios
of activity product to solubility product constant at 38.34°C
for CaS03- ^H., 0 and CaS04 • 2Ha 0 are 10.6 and 1.79 respectively.
This shoxtfs that the solution is highly supersaturated with
respect to CaSO:; ' %H=0 and moderately supersaturated with respect
to CaS04 • 2EP0. These numbers will be of value in conjunction
with scaling studies to define scaling tendency.
1049
-------
LJ
co
w
M
O
M
Pi
co
h
[_•-!
i--- 1
iv ,
f— -1
C
CO
K
I-!
C-,'
(•" '
^fS
\- '
H
O
"^-v
ki
.' •>,
^. *
U '
n p
r
C."1
t —
r ' •
M
Pi
I, I
co
M
Q
l
CNl
r— 1
K
i~3
r/"
r'-
<
10
UJ
O
0
a-
1*1
•
CC
Ki
UJ
if
ZJ r-j
f- C)
•3 1
rr L^
i '
c c
-C j
^-s u- ( -
4-J l~ •
C "'
-• O O
c$ — s «"7 fv
H a <-> x
2-
-
CU
CO
•• — ' h'; f-,
f i ( j
1 1
r D
c~ a
c rj
t o r_>
r- f- .
• •
— < cr
1 1 ii
r • L , Cj c;
L f" c; t.
t < r^ u c
O r- a tn
O C' f~> '^
h- •-< f-'i 1 T r--'
H II II II
2
D C-
~3 (M t v' r i f .•
c o c- --
K— •
r^
'z
U'
t-4
LJ
f-<
U
U.
Ii . l-H f— 1 t— * r-l *-•< «-H *— ^ t-H ^-( f-4 ^ } 'C\ r— 4 r-* ,"_) ^ J r— 1 f ; ^- 4 ,— « ^^ t ^ •— ' r"' ', J .— 1 O •"
c r. ( c c r r C i r, t c c c. c- c f. f- r c r c r: r t r '_ c r
LJ(llllllll|*+ll + *l + l||-»|4+|+(
F- r~ r-: r" C1 i . " 13 ^. f>- n- C^ i." c- fi-. r .-r c: CO n rt; 0^ cr 7^ r ^ ;•-. re
y- .i iT c_; i i c — < ^ , j- r • L" •-« *-, f L: •-> <-• i. r-i C-' r t". •— « i — • •-< i •-< (.
i_ iv o r- M cr rj v: c; o r- t" O >..• f- o t" r- c • r- i<< »— C r-- f_-. «.. " i" r
--,.-<-'•- r~- r~ • -. - r- • •> r~ r- ^-c t— < • r- — < ^-- i— ^-< r-- ••", r- —* r- ^-< ,~< r- ,j r-
t—i
i
u
<3
- u u. i , J f. ^ u. i i i.. i r- r- C". r r- ir. r\ i.; .^, o' 3 1.1 c. r- J f- t".
D 1- 'I Cf O C.J IT C) (D -J) O Ci 'J3 O d '_3 C-1 O LJ O CD O O O rl> C3 C.^ "« —
' "' * ' 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 i I 1 1 l
ij. > r-i: r^ i- C" Lf- — < !•'• >. i •••-- .-< c "~ ;r ; v, t i LI t-~ i > L- ;:• — i c ^ C' •"• u-
>-i > • f , r- ( >- 1 ( - r- LJ t r^i '- -j i. ;--«;••' r j .-i T--, r j i>.. — 1 ^ r . t- t • • a u
n i— — ' f ' — < ' •-' f- i j a r- •—> ' •• ( .. " < ^ <~ f. : (_.~ 3 i • n" i i " j- —4 f • i , •-* r-
^ H «? f f~^ t ' r~* I I ' »-4 f ^- ( r-^ p ' fv_ r . f i^. 4 r^_ ^ _. j , • r, j f , | —^ t t ^ '• * ^^
(—
ID
_J
C
f'
1. -
— .
<_
UJ
n
c: >- u r-: •- n <>j j- 'r r-i i. t r-- r-. <--. r-- r- - , i ^ f. , <-• ^ L' ^ f. ^r r>, c")
<; l_ C-. C^ C . ^ - C~ i . 1 [_, C- C^ t "> C^ C^' C_, L " O C t ' ^J C. - t J C- * ^ Cv C ? '—» •— '
'- 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 | 1 1 1 1 1 1 1 1 1 1
.j c i " ^- . ' - i " J L_ t L. r-- c i "• v. .r r. i c. a j ^ •• j .-.—.( r- • t
-J C_' c, OT C
: o i o f --!(-• '-'- ' c i. t. + c (.- c' 3 1.1 r •* c i , T n c> + e o
r, ( ! •* D" i , c" c i - c~ c i f ' t ., «--; «•* « «j «.-t .= «j cr c i' i' i ' LI- *= <= *-
u T. 2r c: ;r L-- i.'i 3 <_ : n: j J t e..' c> i • e> c i^ r ;-.:.:> i > r: ;: i'
Cj
-------
4 I
C —>
c;
I
(Nl
01
r\)
i
LL)
tt
r- r~>
o o
I I
1.1
-J
•X
Cj
!
I
r-
r-
o
vr
r-!
CO
M
Q
o
c. o
C' r-
o .—i
r-i r-
X
d
c_
u.
W
E-i
.-i Ci
ii n
CM (si
c. o
l/l O
a a
LJ
a
I/-.
1051
-------
IsUrpUTdllOn SBOO SHOAL CREEK BLVD • PO BOX 9949 • AUSTIN, TEXAS 73757 • TELEPHONE 512- 454.9535
Another number of importance for engineering calculation
is the partial pressure of S0a. From Table 12 it is seen that
PCQ in the scrubber effluent is 1.46xlO~s atm or about 1.5 ppm.
This is a necessary input for SOS vapor-liquid mass transfer
calculations. In similar fashion other activities will be re-
quired for solid-liquid mass transfer rates.
ACKNOWLEDGEMENTS
The work presented here was sponsored by the Office
of Air Programs, Environmental Protection Agency under Contract
CPA 70-143, Mr. Julian Jones, Project Officer.
1052
-------
ON-STREAM CHARACTERIZATION OF THE LIMESTONE/DOLOMITE
WET SCRUBBER PROCESS
E. A. Burns and A. Grunt
TRW SYSTEMS
Chemistry and Chemical Engineering Laboratory
Redondo Beach, California
1053
Second International Lime/Limestone Wet Scrubbing Symposium
November 8-12, 1971
New Orleans, Louisiana
-------
ON-STREAM CHARACTERIZATION OF THE LIMESTONE/DOLOMITE
WET SCRUBBER PROCESS
by
E. A. Burns and A. Grunt
TRW SYSTEMS
Chemistry and Chemical Engineering Laboratory
Redondo Beach, California
ABSTRACT
The development of control methodology for sulfur oxide
and particulates from power plant emissions by wet scrubbing
requires accurate and reliable measurements of process vari-
ables. Planned OAP process demonstration studies will result
in a requirement for a large number of chemical analyses re-
quiring 1) automatic instrumental methods and 2) associated
data acquisition and processing capabilities which exceed
current instrumental capabilities. This paper describes acti-
vities undertaken at TRW Systems under Contract 68-02-0007
toward the development of methods suitable for optimization
and control of the wet limestone and dolomite scrubbing pro-
cesses 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 con-
tinuous or slug flow analysis within two minutes were identi-
fied for characterization of the separated solid matter and
liquor. Analytical methods were identified which permit con-
tinuous 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, sul-
fate, 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.
1054
-------
INTRODUCTION
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 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 is not a requisite of the needed methods but rather they
must be adaptable to instrumental techniques that will be reliable, repro-
ducible, cost effective and employ hardware requiring little maintenance.
This paper describes activities undertaken at TRW Systems under Contract
68-02-0007 for development of methods suitable for optimization and control
of the wet limestone/dolomite scrubbing processing by continuous on-stream
analytical methods.
SAMPLING AND ANALYSIS REQUIREMENTS
As described in earlier papers presented at this symposium, the po-
tential chemical species in limestone/dolomite wet scrubbing processes are
numerous. Table I lists the major and minor species that could be present
in the limestone/dolomite wet scrubber slurry. The number of chemical vari-
ables studied were limited to only the key parameters affecting the oper-
ation of the limestone/dolomite scrubber in order to scope the present
program at a manageable size. As a result, activities were focused on es-
tablishing instrumental criteria and identification of instruments suitable
for on-line monitoring of the following chemical species or characteristics:
1055
-------
TABLE I
POSSIBLE LIMESTONE/DOLOMITE SLURRY COMPONENT DISTRIBUTION3
Major Components
Liquid Phase Solid Phase
Ca2+ CaO
Mg2+ "9°
HS03" Ca(OH)2
S042' Mg(OH)2
HC03" CaS03
S032" CaS04
C032" CaC03
MgC03
Minor Components
Liquid Phase Solid Phase
K+ MP04
PO/ SiO,
4 <-
N03" CaF2
Na+ PbS
Fe3+ A1203
Fe2+ S8
Mn2+ ZnS
Cl' Na20
N02" MS04
N03" FeS2
Ti02
C02°3
Distribution of components is dependent on pH and temperature.
• Calcium ion concentration
• Magnesium ion concentration
t Sulfite ion concentration
• Sulfate ion concentration
• Carbonate ion concentration
• pH
• Ionic strength
The analysis methods were selected for evaluation based on their applica-
bility for characterizing slurries, filtered solutions and dry and/or separ-
ated solids on a continuous basis.
On review of the scrubbing process variables several sampling require-
ments were identified relating to characterization of the scrubbers mixture.
These sampling requirements are shown in Table II. During a literature re-
view phase, sampling, separation and quenching of reactants were identified
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
1056
-------
TABLE II
LIMESTONE SLURRY SAMPLING REQUIREMENTS
• Slurry Solids Content - 0 to 15% w/w
• Slurry Sample Quantity - <_!% of stream flow
• Instream Sampling Rate - <2X stream velocity
• Phase Separation
• 100% removal of >0.5y particles in liquid
• Lag Time - <30 seconds
• Sampling Rate - 30 samples/hr, min
• Analysis Time - 2 min, max.
• Easily Maintained
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 and analysis
location and sample frequency. An example which fulfills this requirement
is three different scrubber design processes sampled every 30 minutes at
five different locations. Variations of sampling locations up to eight and
sampling frequency of up to 15 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-loop analysis requirements a total of 30 samples per
hour was taken as a nominal value.
The sampling requirements, process characteristics and the need for
rapid analysis established the instrument requirements identified in
Table III. During the early phases of the program, analysis error require-
ments for the key chemical constituents present in the limestone/dolomite
scrubber were based on estimates provided by the OAP Project Monitor of con-
centration 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 IV and were used to guide the direction of the pro-
gram pending updating of these requirements in concurrent programs by the
Radian Corporation and Bechtel Corporation. It is interesting to compare
the cost associated with analyzing this number of samples by alternative
1057
-------
TABLE III
INSTRUMENT REQUIREMENTS
• Selective Ca +, Mg++, H+, S03=, SO^
t Continuous or plug flow analysis
• Minimum analytical lag time
• Facile calibration
t Routine operation
t Rugged construction
• Low maintenance
t Acceptable accuracy
TABLE IV
LIQUID ANALYSIS REQUIREMENTS
Concentration
Range mM
Mg++
Ca++
so3=
so4=
co3=
Na+
K+
cr
i
i
i
i
i
i
i
i
- 1000
250
150
500
20
500
500
500
Maximum Allowable
Relative Error*
3
3
3
15
15
15
15
15
*For 20% sulfur mass balance closure
laboratory procedures as opposed to on-line instrumentation. Table V lists
the estimated labor for laboratory analyses of sulfite, calcium, magnesium,
sulfate and pH and solids analysis for calcium, total sulfur and magnesium.
The labor hours per sample are estimated to be between 2.1 and 5.2 hours.
The extrapolation of these values to an hour, daily, and monthly basis
show that a large amount of labor is required for minimum characteriza-
tion of a limestone/dolomite scrubber. When the cost of this labor is
1058
-------
TABLE V
ESTIMATED LABOR FOR LABORATORY ANALYSES
Liquid Analyses
Sampling, Hours
S03=, (Titr), hrs
Ca , (AA) Hours
Mg++, (AA) Hours
S04~, (Grav) Hours
pH, Hours
Solids Analyses
Ca
S (X-ray) Hours
Mg
Total Labor Manhours/Sample
For 30 samples/hour, manhours
Per 8-hour day, manhours
Per 30 days, manhours
Minimum
0.1
0.2
0.2
0.2
0.7
0.1
0.6
2.1
63
504
15,120
Maximum
0.1
0.5
0.3
0.3
3.0
0.1
0.9
5.2
156
1,248
37,440
projected together with its necessary supervision, the use of automated on-
line analyses is readily justified on a cost saving basis alone without con-
sideration of advantages of reproducible sampling, calibration and sample
representation.
SLURRY SAMPLING AND SEPARATION
During the course of this program several vendors were contacted to
determine whether they had equipment available which could separate a lime-
stone/dolomite slurry meeting the following operating parameters:
• Flow rate - to 300 Ib/minute (a portion of this flow could
be diverted prior to the separator)
• Solids, % - 0.5 - 15
0 Particle size, micron - 5 - 300
• Density of solids (unpulverized) g/ml - 2.7 - 2.9
0 Density of liquid, g/ml - 1.005 - 1.080
1059
-------
• Temperature, °F - to 150
• System to exclude air during and after separation -
both phases
• Time to effective separation - 15 seconds
Ten companies replied positively that they had equipment which might
fit these operating parameters. The separation principles identified in-
cluded continuous 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 VI. It was found that neither
TABLE VI
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-Sharples solid bowl/Demco
Performance - minimum one hour continuous operation
Results - slight turbidity
Polishing filter - Accu-flow in-line convoluted cartridge
Performance - high capacity - quick interchange
Results - optically clear output
the cone centrifuge nor a combination of the solid bowl centrifuge-centri-
fugal cone provided clear-cut separation as indicated by slight cloudiness
in the discharge fluids. An optically clear fluid would demonstrate excel
lent solids rejection and is needed for any subsequent colorimetric
1060
-------
INLET
OVERFLOW
characterization of the liquid phase. However, inclusion of a polishing
filter, such as an Accu-Flow in-line convoluted cartridge filter downstream
resulted in a high capacity unit providing continuous transparent liquid
for periods as long as several hours depending on the initial solid loading.
The cyclone cone separator was fabricated by Demco to meet TRW's design
requirements and is shown schematically in Figure 1. The device consists
of an 18-mm cone fabricated from 316
stainless steel and possesses an ad-
justable orifice control. The unit
operates with a 35 psi minimum pressure
differential with an inlet feed velocity
of 46 ft/second and a volume demand of
1 gpm. Throttling the underflow to
cause an overflow to underflow ratio
of 45, resulted in an overflow to under-
flow solids content ratio of 0.0204.
Consequently, operation of the Demco in
this mode permitted rejection of approx-
imately 98% of the original solids con-
tent. The solution containing 2% of
the original solids is readily handled
through banked parallel polishing
filters to provide optically clear
liquid. The life time of the filters
are at least one hour and use of a
parallel bank system permits back
flushing to reactivate a spent filter
when it is isolated from the flow loop.
A continuous stage separation con-
cept has been devised which is capable
of achieving "instantaneous quenching
of reaction" within an arbitrary allotted
time of 15 seconds in such a manner as
to present "dry" stream of slurry solids
SEAL
•UNDERFLOW
Figure 2. DEMCO Centrifugal Separator
1061
-------
for continuous analysis. In the conceptual design shown in Figure 2, the
first stage utilizes a liquid/liquid/solid centrifugal separator, such as
one of the deLaval PX solids ejecting centrifuges. The separator would be
fed by the slurry stream, from the point in the scrubber process under
scrutiny. A second heavy liquid phase such as a Freon, trichlorethylene or
other heavy inert solvent would be added to the slurry as it entered the
separator. As shown in the schematic drawing, the light, clear aqueous
phase is separated from an annular zone near the center, the denser non-
aqueous phase is ejected from an intermediate zone while the solids, essen-
tially free from aqueous liquid contamination are continuously discharged
from the outermost zone and transferred to the quartz filter carrier belt.
In the filter/drying housing, which is the second stage, residual inert sol-
vent is volatilized in a heated high pressure dry nitrogen stream before the
solids pass into the solids analyzer.
ANALYSIS METHODS
During the course of this program many alternative instrumental
methods were considered as candidates for adaptation to continuous on-line
analysis. The details of these evaluations will be documented in the final
report to Contract 68-02-0007. A summary of our recommendations is provided
in Table VII which identifies methods of analysis for liquid phase and solid
phase components. The applicability of the recommended methods have been
confirmed through analysis of known simulated slurry mixtures and actual
slurry mixtures and constituents obtained from operating wet limestone scrub-
bing units at the Key West Electric Company, Kansas Power and Light Company
and Shawnee Power Plant. As can be seen from Table VII, considerable use is
made of X-ray and atomic absorption methods which handle samples with mini-
mum pretreatment. Colorimetric methods have been recommended for sulfur (IV)
content in solution and tentatively for nitrite and nitrate. The turbidi-
metric method for dissolved sulfate is a standard method which requires the
addition of barium salts after acidification and heating to remove carbon-
ate and bisulfite interferences. Brief discussions are provided below on
the continuous X-ray analysis methodology and the colorimetric dissolved
sulfur (IV) analysis method. This information is provided because of the
special consideration and testing that were required to ensure acceptable
results for analysis in a limestone slurry environment. The other methods
1062
-------
x
*
o
l_>
=>
h-
**
1
0
s
o
z
13
1/1
O
to
oo
-o
C
13
g
g.
g
cn
(T3
£
S-
CD
1063
-------
TABLE VII
RECOMMENDED METHODS FOR ON-LINE ANALYSIS OF LIQUID
AND SOLID PHASE WET SCRUBBER COMPONENTS
Liquid Phase
• S (IV) (HS03~ + SO,
f
t
Ca
Mg
Fe
K+
++
+ Fe
• NaT
• Total Solids
t pH
• C03=
t Total Sulfur
t N02"
• N03"
Solid Phase
t
t
Total Sulfur
Ca
Mg
Fe
Si
Al
so2, co2
Colorimetric (Furfural Bleaching)
Atomic Absorption
Atomic Absorption
Atomic Absorption
Atomic Absorption
Atomic Absorption
Conductivity
Electrometry
Acidification, Heat + IR Determination
of Evolved Gas
Acidification, Heat and Turbidimetric
X-ray (limit 9.4 mM)
Colorimetric* (Brucine)
Colorimetric* (Brucine)
X-ray
X-ray
X-ray
X-ray
X-ray
X-ray
Pyrolysis + IR Determination*
tentative projected method
are relatively straightforward and are considered routine by those experi-
enced in process control analysis and monitoring.
X-RAY ANALYSIS EQUIPMENT
Commercially available analysis equipment was evaluated with the identi-
fication of the Applied Research Laboratories (ARL) process control X-ray
quantometer (PCXQ) as being suited for continuous on-line analysis of
1064
-------
selected species for both liquid, solid or slurry phases. This instrument
was evaluated for applicability of analysis of these mixtures by determina-
tion of synthesized simulated mixtures. The equipment can be obtained with
a slurry presenter and can handle up to 15 slurry streams sequentially in an
automative mode for elements from magnesium upwards in the periodic table.
Nine spectrometric channels of information are available and nine elements
in a slurry stream can be simultaneously detected and analyzed. A bulk
density monitor is incorporated into the system along with a fixed external
standard.
In the ARL unit the limit of detection for sulfur is 0.03%. The sensi-
tivity of sulfur for on-line slurry units utilizing helium X-ray path and
a Kapton cell window was determined to be significantly less than the 0.25%
absolute value considered to be the lowest reasonable value likely to be
encountered in the slurry mixture. This detection limit corresponds to a
sulfur content of 9.4 mM in the liquid phase. Consequently, X-ray cannot
be used for determining dissolved sulfate and sulfite concentrations which
total less than 9.4 mM. The determined repeatability of the unit was 0.4%
relative, far less than the 3% which has been viewed as a requirement.
As expected, the key for obtaining good X-ray information is to estab-
lish elemental calibration curves using comparable matrix materials which
will be present in the analysis sample. This requires incorporation of both
limestone and flyash to ensure comparable matrices. The effect of particle
size on analytical accuracy is most striking when the size if greater than
40 microns. However, the case of spent limestone/flyash solids, approxi-
mately 90%, have a particle size below 30 microns. For the solids analyzed
from Key West Electric Company, Shawnee Power Plant and Kansas Power and
Light Company, more than 95% of the particles were less than 40 microns.
Consequently, the variability that can be introduced by particle size will
not play a significant role in the analysis of these mixtures.
Comparison of the on-line analysis capability of the ARL unit with re-
presentative laboratory analysis X-ray equipment shows that considerably more
analyses can be accomplished using the ARL PCXQ as is seen in Table VIII.
Extrapolation of these data to obtain operating costs show the ARL unit has
a 4.5-fold advantage over the best competitor oer analysis ($1.33 vs $6.00)
and a 3.4-fold cost per element analyzed advantage (18<£ vs. 62<£).
1065
-------
TABLE VIII
ELEMENTAL ANALYSIS CAPABILITIES OF CANDIDATE X-RAY UNITS
X-Ray
A.R.L.
A.R.L.
G.E. (
Kevex
Unit
(on-line)
(lab)
lab)
(lab)
Estimated Maximum
Number of Analyses
Per 8-hour Shift
240a
160a
160a
45b
Number of
Elements
Per Analysis
8
7
1
a-50 (16)
Total
Elemental
Analyses
^1800
'vlOOO
-v 160
* 640
aTwo-minute residence period in spectrometer
Ten-minute residence period in spectrometer
DETERMINATION OF DISSOLVED SULFUR DIOXIDE
During the review of candidate analytical methods for the determination
of dissolved sulfur dioxide (HS03~ and SO-~) it was determined that no sat-
isfactory methods existed for determining concentrations in the range to be
found in the limestone slurry mixture (see Table IV). Consequently, a new
method based on bisulfite bleaching of the furfural UV absorption was de-
veloped to facilitate this analysis. This method is based on the chemical
equation b in Equations 1 - 4 and depends on the bleaching of the 276 nM
absorption of furfural by reaction with bisulfite.
C4H3OCHO + HS03" t C4H3OCHOHS03" (1)
" + H+ (2)
(3)
(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 Lambert-Beer-Bouguer Law.
A = abcp (5)
where a = molar absorptivity of furfural
b = optical path length
Cp = concentration of uncombined furfural
1066
4. "^
_/
HS03- 1
/I 1 Ov/ o 1 ' ~
4.
^ H +
+
^ H+ +
4
HS03
S03~
-------
The equation of the bleaching reaction (Equation 1) is governed by the
formation constant, K
cF[HS03~]
where c. = concentration of furfural -sulfite adduct
(CA = co - CF}
"] = concentration of uncombined bisulfite
Combining Equations 5 and 6 results in a relationship of absorbance and bi-
sulfite ion as shown in Equation 7.
= _ [HSO "] + -—
A abcQ 3 abcQ
(7)
It is interesting to note that this method was first developed for the de-
termination of furfural and prior to this study has not been used for the
determination of bisulfite. The reason for this is because in most situ-
ations colorimetric procedures are used for determining low concentrations
of chemical species but in the limestone scrubber case the concentration of
bisulfite (1 - 150 mM) is too large for trace analysis methods (without mass-
ive dilution) and not readily adaptable to common macro titrimetric proce-
dures (without using large volumes and dilute titrants).
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 remove trace concentrations of nitrite interference). The reproducibility
of the method has been determined to be better than 2% relative or 0.2mM
absolute whichever is higher. This method is currently being adapted to a
plug flow analyzer system.
LABORATORY BENCH SCALE SCRUBBER
A basic modular designed bench scale test loop wet scrubber (see
Figure 3) 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) ac-
curate control, and 3) producing stable (equilibrium) and unstable
1067
-------
GASEOUS NITHOGEN
MAKEUP CONNECTION
V]GAS RELIEF VALVE Y
GAS BLOCK VALVE 1
t"4
|| f"| ||ROTOMETERS
1 [ 1 ] |_J 0 - 5 CFM
V, V, Vi WASHOUT
£? Q* q? CONNECTION
(\ (l l\ ^V
I IsOj I koJ Io2 jr
^ >
rr^—\2J*-m
10 CFM BLOWER,
W/ DRIVER
6P - 15" H,O MAX
1/3 HP
WASHOUT
CONNECTION
^
X ^ <0
) PISTON TYPE
— - -or:04 v " dh
MINOR GAS ADDITION FACILITY ~*~ - - c«rlP?-t - • , • -r
V-103
\
,T£O—
1/20 HP ELECTRIC STIRRER<^\
M-IOI vf
V
TANK HEATER WITH ^3
THERMOSTAT (1KW) J
E-IOI -J
(INSULATED)
_) t
^s V^
.X* SAMPLE /WASHOUT AL
P*^ CONNECTION
~~ O?«S
15 GALLON - 304 SS TANK
SOLIDS H
10 LB CAP
v-ic
TV" 1
V/ioi
Y TANK BLOCK VALVE
,... SAMPLE /WASHOUT J[
'•*•' CONNECTION f^>
VA,
_. STAR VALVE
r \ AR yj' DRIVER
\^_J 1 01
CARTRIDGE FILTER |
'ijo-
!02Vy
TANK HEATER WITH 13
THERMOSTAT (1 KW) ^J
E - 102
j
u
O,ri
LINES 1 "." SS TUBING
0.028" WALL
OPPER
ACITY
2
/ SLURRY FLOWMETER
CALIBRATION SY-PASS
til
PISTON TYPE
FLOW METER Q
EMERGENCY
BY -PASS VALVE
b"b
' / ELECTHIC STIRKER
/ M-102
+1
15 GALLON,
304 SSTANK
TANK BLOCK VALVE
»X. SAMPLE WASHOUT
~V*~ CONNECTION
OAR ,03
1 \MC
PROCESS FEED TANK y K
T - 102
(INSULATED)
3-
-OlS,
H RUPTURE
DISC
3YNO VARIABLE SPEED PUMP
P-101
1 - .3 GPM
Figure 3. Bench Scale Scrubber Analysis Loop
(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 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
1068
-------
associated with the ducting is 0.4-inch of water. The packed bed is 9-inch
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 SO^, COp and 0~ are controlled in
this manner. Nitrogen is occasionally bled into the system to make up that
amount which has been absorbed by the circulating slurry. 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 S0?, CCL and 0? in the circulating stream. Flow of
the gas stream is given by a differential pressure cell, FR-101.
The liquid slurry exits the Venturi scrubber via the liquid separator
V-101. The temperature in the downcomer is measured and recorded by
TR-105. Analyses of the slurry is also provided in the downcomer by
analyzer AR-105 (type of instrument to be determined during Contract
68-020-0007. The liquid stream from the liquid separator dumps into a
15-gallon delay tank, T-101, where it is agitated with a 1/20 horsepower
electric laboratory stirrer (M-101) and the temperature is adjusted and
controlled by a tank heater E-101. The temperature is measured and re-
corded 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 exit-
ing 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 stirrer, is adjusted by adding limestone from the solids
hopper via a star valve. The composition of the tank is monitored and
1069
-------
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. The characteristics of this stream are given in Table VII. Ac-
curate flow of the pump output is adjusted via the recycle stream to T-102.
The flow is monitored and recorded on FI-103 which receives a signal from
a positive displacement piston type flow meter. The temperature in this
section of line is recorded on TR-103. The flow then splits, part going
through a counter-current flow section packed with 1/4-inch Raschig
rings. The other part of the flow goes to the throat of the Venturi.
The flow which goes to the packed section is measured and recorded on
a flow indicator FI-104, which receives its signal from a positive
displacement flow meter.
All liquid lines present in the bench loop simulator are of 1/4-inch
polypropylene, with an .028-inch wall. Utilizing this type of tubing,
the flow velocity will be about 1.8-feet per second.
This unit has been used to test the applicability of the recommended
methods under controlled conditions. These current studies have confirmed
the fact that 1) the constituents of flyash catalyze the oxidation of sul-
fite to sulfate and 2) the presence of dissolved oxygen in the slurry (al-
though only 0.5 mM at 125°F) contributes significantly to sulfite oxidation.
SUMMARY
Methods have been identified which are suitable for rapid sampling and
on-line chemical analysis of the principle constituents limestone/dolomite
wet scrubber solutions. Utilization of the identified methods in process
demonstrations will permit rapid chemical changes in the scrubber slurry
as a function of process variables and provide needed basic information for
subsequent process optimization.
1070
-------
ACKNOWLEDGMENT
The authors wish to acknowledge the assistance of J. Craig (Zurn
Engineering), Lee Bruton (Kansas Power and Light Company), Jim Martin
(Combustion Engineering), and Joe Barkley (Tennessee Valley Authority) for
their cooperation and assistance in acquisition of limestone scrubber slurry
and solid samples as well as Bob Statnick, OAP Project Monitor for his guid-
ance and encouragement. In addition, the authors wish to acknowledge mem-
bers of the TRW Systems Chemistry and Chemical Engineering Laboratory for
their efforts on this program.
1071
-------
-------
PARTICULATE EMISSIONS PROM TWO LIMESTONE WET SCRUBBERS
Terry Smith
Ronald Draftz
Walter C. McCrone Associates, Inc.
493 East 31st Street
Chicago, Illinois 60616
Prepared for
Second International Lime/Limestone
Wet Scrubbing Symposium
New Orleans, Louisiana
November 8-12, 1971
1073
-------
PARTICULATE EMISSIONS FROM TWO LIMESTONE WET SCRUBBERS
Terry Smith and Ronald Draftz*
Walter C. McCrone Associates, Inc.
493 East 31st Street
Chicago, Hlinois 60616
Abstract
During field tests on a full-scale flooded bed scrubber and a pilot plant
scrubber, data were collected on the variation in mass emission levels and the
size distribution and chemical composition of the particle emissions.
A filter sample and three runs with an Andersen stack sampler were
taken at the outlet of the flooded bed scrubber. We found thai half of the mass
emissions from this scrubber are smaller than 1. 6 /.'tn in diameter. Using elec-
tron diffraction, we determined that 70% of these emissions are hydrated crystals
of calcium sulfate.
Filter samples taken from the pilot plant scrubber showed that the parti-
cles emitted from this scrubber are even smaller than those from the flooded
bed scrubber: The mass average diameter is 0.8 /zm. Here, again, a large
majority of the particle is calcium sulfate.
The size, shape and quantity of small calcium sulfate crystals indicate
that much of the emissions from limestone wet scrubbers are being produced by
the evaporation of fine droplets containing dissolved solids.
* Presented by Ronald G. Draftz
1074
-------
Introduction
During our program (1) to evaluate particulate sampling methods for wet
scrubbers, field tests were conducted on a full-scale flooded bed scrubber at
Kansas Power and Light (KPL) Lawrence, Kansas, and the Zurn pilot plant
Dustraxtor® scrubber at TVA's Shawnee plant. Both scrubbers were operating
on coal-fired power plants.
Although our primary task was to evaluate the effectiveness of various
methods of determining particle mass concentration and size distribution, in the
outlet and inlet of wet scrubbers, we were also interested in obtaining some data
on the composition and size distribution of the particulate in the scrubber outlet.
We did not attempt to determine the exact mass loading from either unit.
Sampling Methods
With the exception of the particulate collection device and the diameter of
the sampling probe, all samples were collected using a standard EPA sampling
train (2). A summary of the sampling conditions of the various particulate col-
lectors is shown in Table 1.
TABLE 1
Summary of Sampling Conditions
Conditions
KPL Flooded-bed
scrubber outlet
Zurn pilot
scrubber outlet
particulate collector
sampling time (rain)
flow rate (SCFM)
collector temperature (° R)
probe diameter (in.)
filter
30
3.68
3/8
Andersen stack sampler
1 4 10
1.0 1.25 1.25
760 760 760
1/4 1/4 1/4
cyclone + filter
45
1.07
695
1/4
1075
-------
Tests at the outlet of the flooded-bed scrubber included collecting parti-
cles with glass-fiber filters, an improved Andersen Stack Sampler, and a 7-stage
cascade impactor. The Andersen Stack Sampler tests were designed to evaluate
particle re-entrainment using three sampling times, 1, 4, and 10 minutes. An
experimental cascaded cyclone and high-temperature membrane filter were used
for sampling the outlet of the Zurn scrubber.
None of the samples were collected under isokinetic conditions because
the particulate matter encountered in these tests is so small that particle size bias
was assumed to be negligible. Since the samples were collected from only one
point they are not completely representative of the stack emissions. However,
the samples from KPL were taken 24 inches inside one stack, and the velocity
profile was very flat, leading us to believe that the particulate was evenly dis-
tributed. The outlet of the Zurn scrubber is only 8 in. in diameter so that fixed
point sampling should be representative.
Analysis Methods
Particulate collected on the glass-fiber filter at the flooded bed scrubber
was sized using an optical microscope in conjunction with a Millipore IIMC auto-
matic image analyzer, set to measure Feret's diameter Transmitted light il-
lumination was used with a 100X oil-immersion objective on the microscope. Five
samples of particulate were removed from different locations on the filter and
mounted in glycerol on glass slides. Glycerol was chosen because its refractive
index is significantly different from those of the major components found in the
samples, calcium sulfate and glass spheres. A total of 2,919 particles were
counted in 10 size intervals from 0. 2 fj.m to 4.5 nm.. Number fractions in each
interval were converted to mass fractions using the cube of the geometric mean
of the interval.
1076
-------
The size distributions of the particles collected by the Andersen Stack
Sampler were obtained by first determining the fraction of the total mass col-
lected on each stage. The particulate collected on each plate was weighed to the
nearest 0.01 milligrams using a semi-micro analytical balance. The characteris-
tic cut-off point, d , for each stage was calculated using the data supplied by
3
2000, Inc., for operating conditions of 760°R and a particle density of 2. 5 g/cm
(3). The cumulative distribution was produced by plotting the d for each stage,
ou
against the sum of the fractions collected below that stage. More precise methods
for determining size distributions have been reported which take into account the
variation of d with the size distribution (4, 5). However, for our purpose, such
O \}
precision was unnecessary.
The particles collected by the experimental cascaded cyclone and filter
were sized with a different method because of the extremely small size of the
particle sampled. Particles from several samples removed from the cyclone and
filter were photographed at magnifications of 10,OOOX and 5,OOOX repsectively,
with a Cambridge Stereoscan IIA scanning electron microscope (SEM). The par-
ticles were then sized using an epidiascope attachment to the IIMC automatic
image analyzer. Again Feret's diameter was used and weight fractions were
obtained in the manner described above.
The fractional efficiency curve was obtained from the fractional mass
distributions of the cyclone and filter catch. The mass collection efficiency of the
cyclone is defined as:
_ Mi - Mo _ (Me + Mo) - Mo
Mi Me + Mo
where Mi, Me, and Mo are the inlet mass, mass of the cyclone catch, and out-
let mass respectively.
1077
-------
It is easy to show that the fractional collection efficiency of size X is given by
C
E =
x F (1 - K)
C +
x K
where C and F are the mass fractions at size X in the cyclone catch and out-
xx
let or filter catch.
Extensive electron diffraction analysis of individual particles from the
filter sample from the flooded-bed scrubber was performed using an RCA EMU 4
transmission electron microscope and later confirmed for the total sample using
x-ray diffraction analysis. Elemental analysis of the cyclone samples were
performed using the SEM and an energy dispersive analyzer for x-ray fluorescence
analysis.
Results
The four particle-size distribution obtained at the flooded-bed scrubber
are shown in Figure 1. The microscopically determined size distribution shows
that 50% of the mass emissions are smaller than 1.6 /urn in diameter. Variations
in mass loading in the stack resulted in the difference between the three distri-
butions obtained with the Andersen Stack Sampler.
Figure 2 is a transmission electron micrograph of particles from the
filter sample. The large cubic particle measures 0. 38 jim on a side, and the
numerous small particles of y-calcium sulfate and gypsum are less than 0.1 jum
in diameter. The complete results of the selected area electron diffraction analy-
sis are given in Table 2.
1078
-------
TABLE 2
Approximate Composition of Particles Collected from a
Flooded-Bed Scrubber Outlet
Chemical species Concentration
CaSO, • 2H O ~ 70%
4 2
CaSO, ~ 20%
4
CaCO < 5%
O
Fe O and Fe O < 5%
o 4 2* o
SiO (glass spheres) < 5%
CaO < 5%
The analysis of the cyclone sample collected from the Zurn scrubber
showed that 83% of the mass of particulate emitted from that scrubber are
smaller than 0. 74 p,m in diameter. X-ray fluorescence analysis on these samples
indicated that the major elements are calcium and sulfur with only minor amounts
of silicon and iron, which is similar to the particle composition of the flooded-
bed scrubber.
Conclusions
Because short sampling times are necessary to avoid particle re-
entrainment, the Andersen Stack Sampler is not useful for determining particle
size-distribution from wet scrubbers.
The high concentration of small hydrated calcium sulfate crystals indicates
that much of the particulate emissions from limestone wet scrubbers are being
produced by evaporation of droplets containing dissolved calcium and sulfate ions.
However, scrubber efficiency for solids seems very good since very little flyash
was found in our samples.
1079
-------
References
This work is supported by Environmental Protection Agency contract
EHS-D-71-25.
2
Federal Register, Standards of Performance for New Stationary Sources,
EPA, Volume 36, Number 159, Part 11, 17 August 1971.
3
2000, Inc., Instructions for Andersen Stack Sampler, Salt Lake City, Utah.
4
Kubie, G., A note on a treatment of impactor data for some aerosols,
Aerosol Sci. 2, 23-30 (1971).
5
Soole, B. W., Concerning the calibration constants of cascade impactors,
with special reference to the Casella MK. 2, Aerosol Sci. 2, 1-14 (1971).
1080
-------
1081
(uir/)
9c-7
-------
FIGURE 2 Transmission electron micrograph of particles emitted
from the wet scrubber, (50, OOOX).
1082
-------
DESIGN CRITERIA FOR A SIZE-SELECTIVE SAMPLER
FOR LIMESTONE WET SCRUBBERS
Terry Smith
Hsing-Chi Chang
Walter C. McCrone Associates, Inc.
493 East 31st Street
Chicago, Illinois 60616
Prepared for
Second International Lime/Limestone
Wet Scrubbing Symposium
New Orleans, Louisiana
November 8-12, 1971
1083
-------
DESIGN CRITERIA FOR A SIZE-SELECTIVE SAMPLER
FOR LIMESTONE WET SCRUBBERS
by
Terry Smith and Hsing-Chi Chang
Walter C. McCrone Associates, Inc.
493 East 31st Street
Chicago, Illinois 60616
SUMMARY
The reasons for and difficulties in developing a gravimetric size-selective
sampler for use with a limestone wet scrubber are outlined. The expected gas
stream conditions at the sampling site are described.
A parallel cyclone sampler which meets the design requirements is
discussed. Methods of obtaining representative samples of particulates from
the gas stream, of accurately sizing the samples, and of determining the max-
imum number of size cuts which can be obtained have been developed for this
sampler and are described. Also included is a discussion of how data from
the sampler can be used to determine the particle-size collection efficiency
curves for a scrubber.
INTRODUCTION
Particle-size information is useful in the design of all particulate col-
lection devices. The effects of size distribution of limestone moeties on gas-
solid reaction kinetics makes particle-size data vital to the development of
the limestone wet scrubber. As reported in a previous paper, the only com
mercially available size-selective sampler for use at stack conditions, the
1084
-------
Andersen stack sampler, does not perform satisfactorily. As a result, the
development of a parallel cyclone sampler for use at the EPA Alkali Scrubbing
2
Test Facility at Shawnee was begun.
Conceptual Design of a Size-Selective Sampler
A sampler should, of course, provide accurate results and avoid the
particle reentrainment problem encountered with the Anderson stack sampler.
It should also measure those segments of the particle-size distribution which
will provide information about the fractional removal efficiency of the wet
scrubber and the specific surface of the particulate. The sensitivity and res-
olution of the sampler should be adequate to detect significant changes in the
particle-size distribution.
A small cyclone followed by a filter meets these requirements, and,
by using several of these in parallel, a gravimetric size-selective sampler is
obtained. The filter provides a stable low terr weight collection media upon
which the particulate matter which passes the cyclone can be accurately weighed.
The cyclone thereby acts as a particle-size selector for the filter.
Since a single sampling probe must transport an unbiased sample of
particulate from the stack to the cyclone for fractionation, the first task in
designing the sampler is determination of the conditions that produce a mini-
mum of sample bias by particulate deposition in the transpost tube. Laboratory
experiments later confirmed during a field test, proved that dust deposition
can be reduced to 2% by mass with proper selection of transport tube diameter
and a transport nozzle having a radius of curvature of 4 diameters. Transpost
tube diameter should be selected to obtain a Reynold's number of approximately
1085
-------
15,000 for the tube.
The next task in the design of the cyclone sampler is the determination
of the number of stages that can be used and the selection of the particle-size
cut-off point (d ) for each stage. Each stage of the sampler views one portion
o U
of the particle-size spectrum: the number of stages that can be used, then, is
limited by the line width, or resolution of each stage, the line width being the
uncertainty in knowing the collection efficiency of the device. At best, the
cut points for the stages can be one line width apart over the entire size range.
A more reasonable spacing would be three line widths.
Two factors affect the resolution of the collector: the error in controlling
the collection parameters and the error in the calibration method. It is
possible to estimate these errors from theoretical considerations and thereby
determine the approximate resolution of a cyclone. By using the method of
3
sensitivity analysis to analyze the mathematical prediction equation of the
cyclone collection efficiency, we found that reasonable errors in controlling
the collection parameters (flow rate, viscosity, temperature, etc.) lead to
variations in the cut point of 2-10%, If a scanning electron microscope is
used for calibration, the calibration error can easily be kept below 0.1 /urn.
The sum of the calibration error and the variance in the collection parameters
reduce the size resolution of a cyclone having a cut point of 1 p.m to a resolution
of 0.1-0.14/^m. For larger cut points, the size resolution is limited by the
variance of the collection parameters; for a cut point of 4.5 Mm the size
resolution is about 0.14-0.49 fiin. By applying these limitations to the size
distribution obtained at the outlet of the Kansas Power and Light Scrubber, some
1086
-------
logical choice of the number of stages can be made. Table 1 shows that six
cut points can be placed, three resolution elements apart, along the mass dis-
tribution.
TABLE 1
Cut Points and Size Resolution of Cyclone Stages
Stage
1
2
3
4
5
6
Cut point
4.48
3.16
2.10
1.35
0.80
0.52
3 Resolution
elements (/nm)
1.47
0.96
0.70
0.5
0.39
0.3
% of particulate
mass below d50
90
75
60
39
15
3
% of particulat
area below dg0
98
90
86
65
35
10
The outlet of the precollector functions as a gas manifold to divide the
gas into seven branches. Since the gas flow in the precollector outlet is a
vortex, the best aerodynamic method of dividing the flow is the use of 7
tangential outlets. An inverted cone in the middle of the outlet manifold is
used to maintain the vortex motion all the way to the top of the manifold. By
incorporating the filter holders into the cover plates of the small cyclones,
particle losses due to deposition in the connecting tubing are minimized.
1087
-------
When the sampler is used at the scrubber inlet where large particles
are present, a precollector to scalp large particles is used in front of the parallel
stages. It is necessary to prevent large particles from entering the small
cyclone where gas velocities are high and particle bounce can lead to the escape
of large particles. The cut point for the precollector then should be small
enough to remove most of the large particles but not so small that there is sub-
stantial overlap between the collection efficiency curve of the first stage and
that of the precollector. We found that a cut point for the precollector of 6.75 Mm
meets the requirements.
By addition of a filter as one of the parallel stages, the mass concentration
of the particulate can be determined and the total flow rate of the sampler can
be easily varied to maintain isokinetic conditions.
Methods of Designing Cyclones
So that the parallel cyclones can be systematically designed, an adequate
understanding of cyclone performance is necessary. A design technique which
optimizes and adjusts all cyclone parameters has been developed based on the
work of several German researchers. ' ' This design method not only
predicts the performance of the cyclone but also adjusts the cyclone geometry
so that a minimum amount of energy in the form of a pressure loss is used to
collect particles of a given size. The result of the optimization process is
reduced turbulence in the cyclone, which should lead to sharp collection ef-
ficiency characteristics.
1088
-------
With a computer to aid in the calculations, three types of cyclones have
been designed for the sampler. One type of cyclone serves for stages 1, 2, and
3 while another is used for stages 4, 5, and 6. The third type is used as the
precollector. We estimated that the flow rate through stage 6 would have to
be greater than 0.75 cfm to obtain an accurately weighable sample of par-
ticulate in 30 minutes at the lowest mass concentration expected. Table 2
shows the operating conditions for each stage and the precollector.
TABLE 2
Operating Conditions for Each Cyclone
Stage
1
2
3
4
5
6
precollector
Flow rate
(cfm)
0.475
0.675
1.00
0.400
0.650
0.875
5.025
Pressure drop
(in. of HO)
Lt
0.114
0.248
0.605
2.63
7.79
15.31
0.165
Maintaining a constant cut point for each cyclone is a complex task
since the cut point depends on the gas velocity, density, viscosity, and wall
friction in the cyclone. In practice it has been found that as particulate de-
posits on the walls of the cyclone, the pressure loss in the cyclone drops,
1089
-------
indicating a reduction in gas velocity. This leads to the idea that perhaps the
best way of maintaining a constant cut point would be maintaining a constant
pressure loss in the cyclone. Therefore the pressure drop across each stage is
monitored by magnihelic differential pressure gages.
A conceptual design for the entire sampling train shows in Figures 1 and 2.
The sampling train consists of three units: the sample box which contains the
cyclone, a control unit which contains pumps, and gas metering equipment and
a cooling supply system for the water vapor traps.
Results of Preliminary Testing
A cyclone having near optimum geometry was designed and constructed.
A field test on a pilot-plant wet scrubber, discussed in a previous paper, was
carried out to determine the performance of the cyclone. We had calculated
that for a flow rate of 0.8 cfm the small cyclone would have a pressure drop
of 27 in water; however, we found its actual drop to be 17 in. water, indicating
that the frictional losses in the cyclone walls had been underestimated. The
design equations also predicted that 50% collection efficiency would occur at
0.44-Mm particle diameter for a 1-cfm flow rate. As is seen in Figure 3,
the actual 50% collection efficiency occurred for 0.74 Mm. This is, again,
due to the increased wall losses causing reduced velocities and collection ef-
ficiency in the cyclone.
A measure of the steepness of the collection efficiency curve for a
device is given by the geometric standard derivation of the collection efficeincy
S where
1090
-------
Particle diameter at 50% efficiency
Particle diameter at 84.13% efficiency
We are gratified to note that the optimization procedure for the cyclone design
produced a very steep efficiency curve having a geometric standard deviation
789
of 0.94. This surpasses the performance reported for several cyclones ' '
10
as well as for inertial impactora
Conclus ions
Undoubtedly a parallel cyclone sampler can be built using the design
techniques developed thus far. The inability to accurately predict performance
merely means that emperical calibrations methods will have to be used. Before
better predictions can be made, however, a better understanding of friction
losses in cyclones will have to be obtained.
1091
-------
Footnotes
Smith, T.M., and R.G. Draftz, Particulate Emissions from two limestone wet
scrubbers, delivered at 2nd International Symposium on Limestone Wet Scrubbers
New Orleans, La. (8-12 November 1971).
2
This work is supported by EPA contract EHS-D-71-25.
g
Schenck, H., Theories of Engineering Experimentation, pp. 50-51, McGraw-Hill,
New York, 1968.
4
Barth, W., Calculation and design of cyclone separators on the basis of recent
investigations, Brennstoff-Warme-Kraft 8, 1-9 (1956).
g
Muschelknautz, E., Design of cyclone separators in the engineering practice,
Staub-Reinholdt Luft 30, 1-12 (1970).
/»
Muschelknautz, E., and W. Krambrock, The aerodynamic coefficients of the
cyclone separator as based on recent, improved measurement, Chem,-Ing.-Tech.
42, 247-255 .(1970).
17
Statrmand, C. J., The design and performance of cyclone separators, Trans.
hist. Che IT. Engrs. 29, 356-383 (1951).
g
Lipman, M., and A. Kydonieus, A multistage aerosol sampler for extended
sampling intervals, Am. hid. Hyg. Assoc. J., 730-7 (1970).
g
Freudenthal, P., High collection efficiency of the Aerotec-3 cyclone for
submicron particles, Atmos. Environ.5, 151-4 (1971).
Cocchman, J. C., and H. M. Moseley, Simplified method for determining
cascade impactor stage efficiencies, Am. Lid. Hyg. Assoc. J., 62-67 (1967).
1092
-------
1093
-------
-------
FIGURE 3 Size distribution of TVA samples collected
with small cyclone
»0 .1
-------
-------
INSTRUMENTAL METHODS FOR FLUE GAS ANALYSIS
R.M. Statnick and J.A. Dorsey
Process Measurements Section
Control Systems Division
Environmental Protection Agency
For presentation at
Second International Lime/Limestone
Wet Scrubbing Symposium
New Orleans, Louisiana
November 8-12, 1971
1097
-------
The application of continuous monitoring instrumentation to
pilot plant and fuel scale evaluations of control technologies
require careful consideration of the effects of the source and
sampling system on overall accuracy of the measurement. In
general, to monitor the mass flow rate of pollutants in flue
gas, the following three problem areas must be considered.
1. Sample Acquisition
2. Sample Handling
3. Instrument Selection
1. Sample Acquisition;
The major problem in the precision determination of pollutant
mass flow rate (Ibs/hr) is the variation of the species
concentration which may exist spacially in a large duct as
a result of air infiltration and poor mixing (stratification).
In the course of OAP's extensive studies of coal fired power
plant effluents, a large number of carbon dioxide concentration
profiles within large ducts were obtained at various sampling
locations.
Typical sampling locations are illustrated in Figure 1 for a
power plant. The two most common sampling locations are at
the inlet and at the outlet of control equipment. Infiltration
of air generally occurs within and post the air pre-heater.
In Figure 2 and Table I, examples of homogeneous and stratified
carbon dioxide profiles are shown. As can be seen in Figure 3,
the probability of determining the concentration of a species
within 15% with a single probe is fifty percent. With nine
probes, one achieves a 99+% probability of determining the
pollutant concentration within 15%.
Stratified flow in pilot plant operations can be avoided by
installation of gas mixing equipment such as venturi, perforated
plate, vanes, etc. at a cost of about $500-$1500 @ 3000 acfm.
This expedient will assure the most accurate measurement
possible of the concentration of the pollutant with a single
point sample probe. To determine the total gas flow, a gas
metering venturi is acceptable.
At a full scale power plant, it is unrealistic to modify the
plant; therefore, the mass flow rate can best be determined by
careful selection of the sampling site and verification of
1098
-------
sampling conditions by complete characterization of the
velocity profile and determination of whether stratification
exists.
The best sites for pollutant concentration and determination
of the velocity profile are not necessarily identical; one
approach which can be used is:
a. Determination of SC>2 or NOx concentration along with the
CC>2 concentration post the economizer but prior to the
air preheater.
b. Determination of the velocity profile at a location move
amenable to a velocity traverse. This will yield the
total gas volumetric flow at the location.
c. The volumetric flow of gas at the economizer is given by
volumetric flow at economizer = volumetric flow (from
velocity data) x CO? (at traverse)
CC>2 (at economizer)
2. Sample Handling
Having selected and evaluated a site, the next consideration
is extraction of samples.
Probes;
The sampling probe design is dictated by temperature. It
can be as simple as a 1/2-inch O.D. stainless steel tube for
temperature 350°P and above, or as complex as a shielded in
the stack filter-probe combination for temperatures below
350°F to minimize reactive losses of SC>2
Filters;
All instruments which utilize optical principles to determine
the pollutant concentration require the removal of particulate
matter. Particulate matter is removed by passing the particulate
laden-gas stream through a positive filter. A silicon carbide
filter which has a 90% collection efficiency at 5/1 have been
found to be practical. These filters can be mounted internal
or external of the ducting; since the average working life of
the filters is 3-4 weeks (at 5-7 grains/scf and 2 cfh flow),
the external stack filter is recommended for ease of replacement
1099
-------
of the filter and minimum down time. The filter assembly
should be maintained at 300-350°F to eliminate reactive S02
losses with the filtering media.
A potential problem particular to wet scrubbers might be
observed post the device. If there is substantial liquid
re-entrainment and poor mist eliminator efficiency, the
saturated scrubber liquor droplets will enter the probe and
be collected on the filtering media. At 300-350°F, evaporation
of the water will leave a residual deposit of CaSC>4, CaSC^J^O,
and CaCOs platlets which will plug the filter. Filter pluggage
was observed during manual particulate sampling at Kansas
Power and Light. It will also occur using the filters describe
above; to reduce the probability of this type of pluggage, high
turbulence in the probe to promote droplet evaporation is
recommended.
SamplingLines and Water Vapor Condensors
The sampling lines can be constructed of heat traced teflon
or stainless steel. The sampling lines should be maintained
above the dew point of the flue gas, 300-350°F. Stress
corrosion has been observed in the sampling lines (304 stainles
steel) used at the TVA dry limestone injection tests. The need
for frequent replacement of the stainless steel tubing makes th
heat traced teflon, although initially more expensive, the most
desirable material of construction. The use of rubber on PVC
tubing is not recommended since absorption of SC>2 on tube walls
occurs in these materials.
For those who select non-dispersive infrared (NDIR) as the
pollutant monitor, a condensing system is required to remove
water vapor which is a positive interference. A sample NDIR
has a rejection ratio of 100; that is, 100 ppm of water vapor
yields a signal equivalent to 1 ppm of SC>2 • A condenser held
at 0° .+ 1°C will contribute a _+ 4 ppm signal of SC>2 equivalent
water vapor. At 1000 ppm SO2 or greater, typical of the
scrubber inlet this error is insignificant, but with 90% contro
of a 1000 ppm inlet S02 concentration, it will yield a ± 4%
error in the SO2 level at the outlet.
Gas Pumps
Either bellows type or other leakless air moving pump is
acceptable. The pump is located prior to the detection system.
1100
-------
Response Time
All of the components of a sampling system have been covered,
these elements should be so constructed such that the desired
system response time is achieved. The system response time
is defined as:
The time interval from a step change in pollutant
concentration at the probe inlet to a recording of
90% of the ultimate recorded output.
By suitably adjusting the volume of the sample handling
system and/or the volumetric flow through the sampling system
a wide range of system response times are achievable. For a
typical NDIR sampling analysis system, the sample handling
system has a total volume of 0.5 cu ft (assuming 100' 1/2-inch
I.D. tubing, filter, and 0.3 cu ft cooler volume). The
response time of the system, assuming plug flow and 2 cfh
pumping capacity, normally supplied with instrument, is 15
minutes.
In Figure 4, a high volume pump about 20 cfh is used to
extract the sample and draw it through the filter and cooler.
The water and particulate free stream is then sampled by the
2 cfh pump. This will reduce the system response time to 1.5
minutes with a sample system of 0.5 cu ft.
3. Instrumentation
The Office of Air Programs had funded field evaluation of
commercially available sulfur dioxide and nitrogen oxide
monitors. These studies were conducted at coal fired power
plants; the sulfur oxide study at a steam station burning
0.5% sulfur coal; the nitrogen oxides study at a coal fired
station with approximately 200 ppm NOx emissions. Preliminary
results of these studies are shown in Table II.
The table also shows that for sulfur dioxide NDIR's or DNUV1s are
of comparable accuracy and reliability. For NO, NDIR; for
NO2 NDUV; presently only NDUV for total NOx; however, this is
batch operation and the 10 minute reactor time must be added
to system response time. In addition to the instrumentation
described above, flue gas analysis could be performed by gas
chromatography, mass spectroscopy, dispersive infrared, etc.
None of these are presently available as commercially tested
units.
1101
-------
Instrument types, found to be effective for data collection
for engineering analysis for control systems development of
SC>2 and NOx flue gas control equipment, include:
S02—NDIR and NDUV,
NO —NDIR and NDUV, and
NOx—NDUV.
The NDUV is sensitive to NO2 and the commercially available
instrument provides integral catalytic oxidation of NO to NO2.
Overall Conclusions
The overall conclusions which can be reached are:
1. Extreme care must be exercised in the choice of sampling
location and each location should be completely characteriz
2. The probe, filter, and sampling lines must be held above
the dew point of the gas stream.
3. Using vendor supplied sample conditioning equipment and
gas pumps, system response times are greater than 10 minute
4. For sulfur dioxide monitoring, NDIR's or NDUV's are
effective.
5. For nitrogen oxides, NDUV is effective if 10 minute cycle
times are acceptable; for continuous analysis NDIR's are
effective (note: total nitrogen oxides will be NO; NDIR's
are insensitive to NO2).
1102
-------
Low level
economizer
Sampling points B
Sampling points A
Burners
Figure 1 Boiler Outline for Corner-Fired Unit Showing Sampling Posit
1103
ions
-------
HOMOGENEOUS
1.00
1.03
1.00
1.00
1.01
1.01
0.98
0.98
0.98
1.00
1.00
1.01
1.01
1.01
1.01
t
3 '3"
if
A ??' kl
Avg C02 = 11.7%
CV = 1.9%
STRATIFIED
0.
0.
0.
0.
74
82
75
82
0
1
1
1
.92
.01
.03
.00
1.
1.
0.
0.
00
02
99
93
1.
1.
1.
0.
00
01
02
93
0
1
1
1
.97
.00
.01
.01
0.
0.
0.
0.
95
90
90
85
t
4'8"
I
Avg CO2 = 12.6%
CV = 9.3%
Random selection from six plants.
Figure 2. Normalized CO2 Traverse Data at Dust Collector of
Coal-Fired Power Plants.
1104
-------
TABLE I. OBSERVED COEFFICIENT OF VARIATION FOR CO2 TRAVERSE
FOR VARIOUS COAL-FIRED PLANTS
Plant
No.
1
2
3
4
5
Type of
Boiler Firing
Horizontally opposed
Cyclone
Spreader stoker
Corner
Vertical
Dust
Collection
Equipment
C
E
C
C, E
C, E
Sampling
Location
I
O
I
0
I
0
I
0
I
0
No. of
Traverse
Points
24
12
24
24
18
9
18
12
24
12
co2
%
9.3
2.3,
4.6
3.2
1.5
1.02
8.8
0.97
7.1
3.2
(CV)
1.4
]=Cyclone
]=Electrostatic precipitator
:=Dust collector inlet
)=Dust collector outlet
1105
-------
10
PROBABILITY FOR
3 PROBES ACROSS
CENTER OF DUCT*
PROBABILITY FOR A
•SINGLE PROBE IN THE
CENTER OF THE DUCT
01 2345678
NUMBER OF PROBES UTILIZED
Figure 3. Probability of Obtaining an Accuracy Within
15% of 9-Point Analysis for 02 in a Large Duct
1106
-------
PROBE
mt^m
FII
/TER
(
CONDENSER
1
2(
_JO_
2:
20 CFH
2 CFH
MONITOR
Figure 4. Fast-Response Sampling System.
1107
-------
TABLE II. SULFUR DIOXIDE MONITORS
Detection Principle
NDIR
NDUV
Conductrometric
Couldmetric
Ele c tr ochemlca-1
Instrument ^ '
Response Time
Good
Good
Poor
Good
Good
Reliability(2)
MTF
351 hrs.
322 hrs.
67.1 hrs.
569 hrs.
Poor
Accuracy
Good
Good
Good
Good
Good
Nitrogen Oxide Monitor
Detection Principle
NDIR
NDUV
Electrochemical
Response
Time
Good
Good
Good
Reliability, MTF- (2)
Very Good(3)
Very Good^3^
PrtrtT
Accurac'
Good
Good
Good
(1) Time interval from a step change in the pollutant in concentration
at the instrument inlet to a recording of 90% of the ultimate recorded
output (>3 sec.).
C2) Mean time between failure
(3) No failures during test
1108
-------
EPA RECOMMENDED SOURCE TEST METHODS FOR NEW SOURCE
PERFORMANCE STANDARDS TESTING
Gene W. Smith
Applied Technology Division
Environmental Protection Agency
Prepared for
Second International Lime/Lime stone
Wet Scrubbing Symposium
New Orleans, Louisiana
November 8-12, 1971
1109
-------
EPA RECOMMENDED SOURCE TEST METHODS FOR NEW SOURCE
PERFORMANCE STANDARDS TESTING
The text of this paper consisted of an explanation of
the Standards of Performance for New Stationary Sources
proposed by the Environmental Protection Agency and established
by the Clean Air Act as Arranended. Gene Smith used the Federal
Register, Vol. 36, No. 247—Thursday, December 23, Part II
as his reference material, which he handed out during the
symposium.
1110
-------
THURSDAY, DECEMBER 23, 1971
WASHINGTON, D.C.
Volume 36 • Number 247
PART II
ENVIRONMENTAL
PROTECTION
AGENCY
Standards of Performance for
New Stationary Sources
No. 247—Pt. H 1
mi
-------
24876
RULES AND REGULATIONS
Title 40—PROTECTION OF
ENVIRONMENT
Chapter I—Environmental Protection
Agency
SUBCHAPTER C—AIR PROGRAMS
PART 60—STANDARDS OF PERFORM-
ANCE FOR NEW STATIONARY
SOURCES
On August 17, 1971 (36 F.R. 15704)
pursuant to section 111 of the Clean Air
Act as amended, the Administrator
proposed standards of performance for
steam generators, Portland cement1
plants, Incinerators, nitric acid plants,
and sulfuric acid plants. The proposed
standards, applicable to sources the con-
struction or modification of which was
initiated after August 17, 1971, included
emission limits for one or more of four
pollutants (particulate matter, sulfur
dioxide, nitrogen oxides, and sulfuric
acid mist) for each source category. The
proposal included requirements for per-
formance testing, stack gas monitoring,
record keeping and reporting, and pro-
cedures by which EPA will provide pre-
construction review and determine the
applicability of the standards to specific
sources.
Interested parties were afforded an
opportunity to participate in the rule
making by submitting comments. A total
of more than 200 interested parties, in-
cluding Federal, State, and local agen-
cies, citizens groups, and commercial and
Industrial organizations submitted com-
ments. Following a review of the pro-
posed regulations and consideration of
the comments, the regulations, includ-
ing the appendix, have been revised and
are being promulgated today. The prin-
cipal revisions are described below:
1. Particulate matter performance
testing procedures have been revised to
eliminate the requirement for impingers
in the sampling train. Compliance will be
based only on material collected in the
dry filter and the probe preceding the
filter. Emission limits have been adjusted
as appropriate to reflect the change in
test methods. The adjusted standards re-
quire the same degree of particulate con-
trol as the originally proposed standards.
2. Provisions have been added whereby
alternative test methods can be used to
determine compliance. Any person who
proposes the use of an alternative
method will be obliged to provide evi-
dence that the alternative method is
equivalent to the reference method.
3. -The definition of modification, as it
pertains to increases in production rate
and changes of fuels, has been clarified.
Increases in production rates up to design
capacity will not be considered a modifi-
cation nor will fuel switches if the equip-
ment was originally designed to accom-
modate such fuels. These provisions will
eliminate inequities where equipment had
been put into partial operation prior to
the proposal of the standards.
4. The definition of a new source was
clarified to include construction which
is completed within an organization as
well as the more common situations
•where the facility is designed and con-
structed by a contractor.
5. The provisions regarding requests
for EPA plan review and determination
of construction or modification have been
modified to emphasize that the submittal
of such requests and attendant informa-
tion is purely voluntary. Submittal of
such a request will not bind the operator
to supply further information; however,
lack of sufficient information may pre-
vent the Administrator from rendering
an opinion. Further provisions have been
added to the effect that information sub-
mitted voluntarily for such plan review
or determination of applicability will be
considered confidential, if the owner or
operator requests such confidentiality.
6. Requirements for notifying the Ad-
ministrator prior to commencing con-
struction have been deleted. As proposed,
the provision would have required notifi-
cation prior to the signing of a contract
for construction of a new source. Owners
and operators still will be required to
notify the Administrator 30 days prior to
initial operation and to confirm the
action within 15 days after startup.
7. Revisions were incoporated to per-
mit compliance testing to be deferred up
to 60 days after achieving the maximum
production rate but no longer than 180
days after initial startup. The proposed
regulation could have required testing
within 60 days after startup but defined
startup as the beginning of routine
operation. Owners or operators will be
required to notify the Administrator at
least 10 days prior to compliance testing
so that an EPA observer can be on hand.
Procedures have been modified so that
the equipment will have to be operated
at maximum expected production rate,
rather than rated capacity, during com-
pliance tests.
8. The criteria for evaluating perform-
ance testing results have been simplified
to eliminate the requirement that all
values be within 35 percent of the aver-
age. Compliance will be based on the
average of three repetitions conducted in
the specified manner.
9. Provisions were added to require
owners or operators of affected facilities
to maintain records of compliance tests,
monitoring equipment, pertinent anal-
yses, feed rates, production rates, etc. for
2 years and to make such information
available on request to the Administra-
tor. Owners or operators will be required
to summarize the recorded data daily
and to convert recorded data into the
applicable units of the standard.
10. Modifications were made to the
visible emission standards for steam
generators, cement plants, nitric acid
plants, and sulfuric acid plants. The
Ringelmann standards have been de-
leted; all limits will be based on opacity.
In every case, the equivalent opacity will
be at least as stringent as the proposed
Ringelmann number. In addition, re-
quirements have been altered for three
of the source categories so that allowable
emissions will be less than 10 percent
opacity rather than 5 percent or less
opacity. There were many comments
that observers could not accurately
evaluate emissions of 5 percent opacity.
In addition, drafting errors in the pro-
posed visible emission limits for cement
kilns and steam generators were cor-
rected. Steam generators will be limited
to visible emissions not greater than 20
percent opacity and cement kilns to not
greater than 10 percent opacity.
11. Specifications for monitoring de-
vices were clarified, and directives for
calibration were included. The instru-
ments are to be calibrated at least once
a day, or more often if specified by the
manufacturer. Additional guidance on
the selection and use of such instruments
will be provided at a later date.
12. The requirement for sulfur dioxide
monitoring at steam generators was
deleted for those sources which will
achieve the standard by burning low-sul-
fur fuel, provided that fuel analysis is
conducted and recorded daily. American
Society for Testing and Materials
sampling techniques are specified for
coal and fuel oil.
13. Provisions were added to the steam
generator standards to cover those In-
stances where mixed fuels are burned.
Allowable emissions will be determined
by prorating the heat input of each fuel,
however, in the case of sulfur dioxide, the
provisions allow operators the option of
burning low-sulfur fuels (probably
natural gas) as a means of compliance.
14. Steam generators fired with lignite
have been exempted from the nitrogen
oxides limit. The revision was made in
view of the lack of information on some
types of lignite burning. When more in-
formation is developed, nitrogen oxides
standards may be extended to lignite
fired steam generators.
15. A provision was added to make it
explicit that the sulfuric acid plant
standards will not apply to scavenger
acid plants. As stated in the background
document, APTD 0711, which was issued
at the time the proposed standards were
published, the standards were not meant
to apply to such operations, e.g., where
sulfuric acid plants are used primarily
to control sulfur dioxide or other sulfur
compounds which would otherwise be
vented into the atmosphere.
16. The regulation has been revised
to provide that all materials submitted
pursuant to these regulations will be di-
rected to EPA's Office of General En-
forcement.
17. Several other technical changes
have also been made. States and inter-
ested parties are urged to make a careful
reading of these regulations.
As required by section 111 of the Act,
the standards of performance promul-
gated herein "reflect the degree of emis-
sion reduction which (taking into ac-
count the cost of achieving such reduc-
tion) the Administrator determines has
been adequately demonstrated". The
standards of performance are based on
stationary source testing conducted by
the Environmental Protection Agency
and/or contractors and on data derived
from various other sources, including the
available technical literature. In the com-
ments on the proposed standards, many
questions were raised as to costs and
FEDERAL REGISTER, VOL. 36, NO. 247—THURSDAY, DECEMBER 23, 1971
1112
-------
RULES AND REGULATIONS
24877
demonstrated capability of control sys-
tems to meet the standards. These com-
ments have been evaluated and investi-
gated, and it is the Administrator's
judgment that emission control systems
capable of meeting the standards have
been adequately demonstrated.and that
the standards promulgated herein are
achievable at reasonable costs.
The regulations establishing standards
of performance for steam generators, in-
cinerators, cement plants, nitric acid
plants, and sulfuric acid plants are here-
by promulgated effective on publication
and apply to sources, the construction or
modification of which was commenced
after August 17, 1971.
Dated: December 16, 1971.
WILLIAM D. RTJCKELSHATJS,
Administrator,
Environmental Protection Agency.
A new Fart 60 is added to Chapter I,
Title 40, Code of Federal Regulations, as
follows:
Subpart A—General Provisions
Sec.
601 Applicability.
60.2 Definitions.
60.3 Abbreviations.
60.4 Address.
605 Determination of construction or
modification.
60.6 Review of plans.
60.7 Notification and recordkeeping.
60.8 Performance tests.
60.9 Availability of information.
60 10 State authority.
Subpart D—Standards of Performance for
Fossil Fuel-Fired Steam Generators
60.40 Applicability and designation of af-
fected facility.
60.41 Definitions.
60.42 Standard for particulate matter.
6O.43 Standard for sulfur dioxide.
60.44 Standard for nitrogen oxides.
60.45 Emission and fuel monitoring.
60.46 Test methods and procedures.
Subpart E—Standards of Performance for
Incinerators
60.50 Applicability and designation of af-
fected facility.
60.51 Definitions.
60.62 Standard for particulate matter.
60.53 Monitoring of operations.
60.54 Test methods and procedures.
Subpart F—Standards of Performance for
Portland Cement Plants
60 60 Applicability and designation of
affected facility.
60.61 Definitions.
60.62 Standard for particulate matter.
60.63 Monitoring of operations
60.64 Teat methods and procedures.
Subpart G—Standards of Performance for Nitric
Acid Plants
60.70 Applicability and designation of af-
fected facility
60 71 Definitions.
60 72 Standard for nitrogen oxides.
60.73 Emission monitoring.
60.74 Test methods and procedures.
Subpart H—Standards of Performance for Sutfuric
Acid Plants
60.80 Applicability and designation of af-
fected facility.
60.81 Definitions.
Sec.
60.82
60.83
60,84
60.85
Standard for sulfur dioxide.
Standard for acid mist.
Emission monitoring.
Test methods and procedures.
APPENDIX—TEST METHODS
Method 1—Sample and velocity traverses for
stationary sources.
Method 2—Determination of stack gas veloc-
ity and volumetric flow rate (Type S
pitot tube).
Method 3—Gas analysis for carbon dioxide,
excess air, and dry molecular weight.
Method 4—Determination of moisture in
stack gases.
Method 5—Determination of parttculate
emissions from stationary sources.
Method 6—Determination of sulfur dioxide
emissions from stationary sources.
Method 7—Determination of nitrogen oxide
emissions from stationary sources.
Method 8—Determination of sulfuric acid
mist and sulfur dioxide emissions
from stationary sources.
Method 9—Visual determination of the opac-
ity of emissions from stationary
sources.
AUTHORITY: The provisions of this Part 60
issued under sections 111, 114, Clean Air Act;
Public Law 91-604, 84 Stat. 1713.
Subpart A—General Provisions
§ 60.1 Applicability.
The provisions of this part apply to
the owner or operator of any stationary
source, which contains an affected facil-
ity the construction or modification of
which is commenced after the date of
publication in this part of any proposed
standard applicable to such facility.
§ 60.2 Definitions.
As used in this part, all terms not
defined herein shall have the meaning
given them in the Act:
(a) "Act" means the Clean Air Act
(42 U.S.C. 1857 et seq., as amended by
Public Law 91-604, 84 Stat. 1676).
(b) "Administrator" means the Ad-
ministrator of the Environmental Pro-
tection Agency or his authorized repre-
sentative.
(c) "Standard" means a standard of
performance proposed or promulgated
under this part.
(d) "Stationary source" means any
building, structure, facility, or installa-
tion which emits or may emit any air
pollutant.
(e) "Affected facility" means, with
reference to a stationary source, any ap-
paratus to which a standard is applicable.
(f) "Owner or operator" means any
person who owns, leases, operates, con-
trols, or supervises an affected facility
or a stationary source of which an af-
fected facility is a part.
(g) "Construction" means fabrication,
erection, or installation of an affected
facility.
(h) "Modification" means any physical
change in, or change in the method of
operation of. an affected facility which
increases the amount of any air pol-
lutant (to which a standard applies)
emitted by such facility or which results
in the emission of any air pollutant (to
which a standard applies) not previously
emitted, except that:
(1) Routine maintenance, repair, and
replacement shall not be considered
physical changes, and
(2) The following shall not be consid-
ered a change in the method of
operation:
(i) An increase in the production
rate, if such increase does not exceed the
operating design capacity of the affected
facility;
(ii) An increase in hours of operation;
(iii) Use of an alternative fuel or raw
material if, prior to the date any stand-
ard under this part becomes applicable
to such facility, as provided by § 60.1,
the affected facility is designed to ac-
commodate such alternative use.
(i) "Commenced" means that an own-
er or operator has undertaken a con-
tinuous program of construction or
modification or that an owner or opera-
tor has entered into a binding agree-
ment or contractual obligation to under-
take and complete, within a reasonable
time, a continuous program of construc-
tion or modification.
(j) "Opacity" means the degree to
which emissions reduce the transmission
of light and obscure the view of an object
in the background.
(k) "Nitrogen oxides" means all ox-
ides of nitrogen except nitrous oxide, as
measured by test methods set forth In
this part.
(1) "Standard of normal conditions"
means 70° Fahrenheit (21.1" centi-
grade) and 29.92 in. Hg (760 mm. Hg).
(m) "Proportional sampling" means
sampling at a rate that produces a con-
stant ratio of sampling rate to stack gas
flow rate.
(n) "Isokinetic sampling" means
sampling in which the linear velocity of
the gas entering the sampling nozzle is
equal to that of the undisturbed gas
stream at the sample point.
(o) "Startup" means the setting hi
operation of an affected facility for any
purpose.
§ 60.3 Abbreviations.
The abbreviations used in this part
have the following meanings in both
capital and lower case:
B.t.u.—British thermal unit.
cal.—calorie (s).
c.f.m.—cubic feet per minute.
COa—carbon dioxide.
g—gram(s).
gr.—grain(s).
mg —milligram (s).
mm.—millimeter(s).
1.—liter (s).
nm—nanometer(s), —10-' meter.
Pg.—microgram(s), 10-« gram.
Hg.—mercury.
in—inch(es).
K—l ,000.
lb.—pound (s).
ml —milliliter(s).
No.—number.
%—percent.
NO—nitric oxide
NOj—nitrogen dioxide.
NOX—nitrogen oxides.
NM.1—normal cubic meter.
s.c.f.—standard cubic feet.
SO,—sulfur dioxide.
H2SO,—sulfuric acid.
SO,—sulfur trioxide.
FEDERAL REGISTER, VOL 36, NO 247—THURSDAY, DECEMBER ?3, 1971
1113
-------
24878
RULES AND REGULATIONS
ft.3—cubic feet.
ftJ—square feet.
mm.—minute(s).
hr.—hour(s).
§ 60.4 Address.
All applications, requests, submissions,
and reports under this part shall be sub-
mitted in triplicate and addressed to the
Environmental Protection Agency, Office
of General Enforcement, Waterside Mall
SW., Washington, DC 20460.
§ 60.5 Determination of construction or
modification.
When requested to do so by an owner
or operator, the Administrator will make
a determination of whether actions taken
or Intended to be taken by such owner or
operator constitute construction or modi-
fication or the commencement thereof
within the meaning of this part.
§ 60.6 Review of plans.
(a) When requested to do so by an
owner or operator, the Administrator will
review plans for construction or modifi-
cation for the purpose of providing
technical advice to the owner or operator.
(b) (1) A separate request shall be
submitted for each affected facility.
(2) Each request shall (i) identify the
location of such affected facility, and (ii)
be accompanied by technical information
describing the proposed nature, size,
design,' and method of operation of such
facility, including information on any
equipment to be used for measurement or
control of emissions.
(c) Neither a request for plans review
nor advice furnished by the Administra-
tor in response to such request shall (1)
relieve an owner or operator of legal
responsibility for compliance with any
provision of this part or of any applicable
State or local requirement, or (2) prevent
the Administrator from implementing or
enforcing any provision of this part or
taking any other action authorized by the
Act.
§ 60.7 Notification and record keeping.
(a) Any owner or operator subject to
the provisions of this part shall furnish
the Administrator written notification as
follows:
(1) A notification of the anticipated
date of initial startup of an affected
facility not more than 60 days or less
than 30 days prior to such date.
(2) A notification of the actual date
of initial startup of an affected facility
within 15 days after such date.
(b) Any owner or operator subject to
the provisions of this part shall maintain
for a period of 2 years a record of the
occurrence and duration of any startup,
shutdown, or malfunction in operation of
any affected facility.
§ 60.8 Performance tests.
(a) Within 60*days after achieving the
maximum production rate at which the
affected facility will be operated, but not
later than 180 days after initial startup
of such facility and at such other times
as may be required by the Administrator
under section 114 of the Act, the owner
or operator of such facility shall conduct
performance test(s) and furnish the Ad-
ministrator a written report of the results
of such performance test(s).
(b) Performance tests shall be con-
ducted and results reported in accord-
ance with the test method set forth in
this part or equivalent methods approved
by the Administrator; or where the Ad-
ministrator determines that emissions
from the affected facility are not sus-
ceptible of being measured by such
methods, the Administrator shall pre-
scribe alternative test procedures for
determining compliance with the re-
quirements of this part.
(c) The owner or operator shall permit
the Administrator to conduct perform-
ance tests at any reasonable time, shall
cause the affected facility to be operated
for purposes of such tests under such
conditions as the Administrator shall
specify based on representative perform-
ance of the affected facility, and shall
make available to the Administrator
such records as may be necessary to
determine such performance.
Safe sampling platform (s).
(3) Safe access to sampling plat-
form (s).
(4) Utilities for sampling and testing
equipment.
(f) Each performance test shall con-
sist of three repetitions of the applicable
test method. For the purpose of deter-
mining compliance with an applicable
standard of performance, the average of
results of all repetitions shall apply.
§60.9 Availability of information.
(a) Emission data provided to, or
otherwise obtained by, the Administra-
tor in accordance with the provisions of
this part shall be available to the public.
(b) Except as provided in paragraph
(a) of this section, any records, reports,
or information provided to, or otherwise
obtained by, the Administrator in accord-
ance with the provisions of this part
shall be available to the public, except
that (1) upon a showing satisfactory to
the Administrator by any person that
such records, reports, or information, or
particular part thereof (other than
emission data), if made public, would
divulge methods or processes entitled to
protection as trade secrets of such per-
son, the Administrator shall consider
such records, reports, or information, or
particular part thereof, confidential in
accordance with the purposes of section
1905 of title 18 of the United States
Code, except that such records, reports,
or information, or particular part there-
of, may be disclosed to other officers, em-
ployees, or authorized representatives of
the United States concerned with cairy-
ing out the provisions of the Act or when
relevant in any proceeding under ths
Act; and (2) information received by the
Administrator solely for the purposes 01
§160.5 and 60.6 shall not be disclosed
if it is identified by the owner or opera-
tor as being a trade secret or com-
mercial or financial information which
such owner or operator considers
confidential.
§ 60.10 State authority.
The provisions of this part shall not
be construed in any manner to preclude
any State or political subdivision thereof
from
(a) Adopting and enforcing any emis-
sion standard or limitation applicable tj
an affected facility, provided that such
emission standard or limitation is not
less stringent than the standard appli-
cable to such facility.
(b) Requiring the owner or operator
of an affected facility to obtain permits.
licenses, or approvals prior to initiating
construction, modification, or operation
of such facility.
Subpart D—Standards of Performance
for Fossil-Fuel Fired Sfeam Generators
§ 60.40 Applicability and designation of
all'ected facility.
The provisions of this subpart are ap-
plicable to each fossil fuel-fired steam
generating unit of more than 250 million
B.t.u. per hour heat input, which is the
affected facility.
§ 60.41 Definitions.
As used in this subpart, all terms not
defined herein shall have the meaning
given them in the Act, and in Subpart
A of this part.
(a) "Fossil fuel-fired steam generat-
ing unit" means a furnace or boiler used
in the process of burning fossil fuel
for the primary purpose of producing
steam by heat transfer.
(b) "Fossil fuel" means natural gas,
petroleum, coal and any form of solid,
liquid, or gaseous fuel derived from
such materials.
(c) "Particulate matter" means any
finely divided liquid or solid material,
other than uncombined water, as meas-
ured by Method 5.
§ 60.42 Standard for particulate matter.
On and after the date on which the
performance test required to be con-
ducted by § 60.8 is initiated no owner
or operator subject to the provisions of
this part shall discharge or cause the
discharge into the atmosphere of par-
ticulate matter which is:
(a) In excess of 0.10 Ib. per million
B.t.u. heat input (0.18 g. per million calj
maximum 2-hour average.
(b) Greater than 20 percent opacity,
except that 40 percent opacity shall be
permissible for not more than 2 minutes
in any hour.
(c) Where the presence of uncom-
bined water is the only reason for fail-
ure to meet the requirements of para-
graph (b) of this section such failure
shall not be a violation of this section.
FEDERAL REGISTER, VOL. 36, NO. 247—THURSDAY, DECEMBER 23, 1971
1114
-------
RULES AND REGULATIONS
24879
§ 60.43 Standard for sulfur dioxide.
On and after the date on which the
performance test required to be con-
ducted by § 60.8 is initiated no owner
or operator subject to the provisions
of this part shall discharge or cause the
discharge into the atmosphere of sulfur
dioxide in excess of:
(a) 0.80 Ib. per million B.t.u. heat in-
put (1.4 g. per million cal.), maximum 2->
hour average, when liquid fossil fuel is
burned.
(b) 1.2 Ibs. per million B.t.u. heat input
(2.2 g. per million cal.), maximum 2-
hour average, when solid fossil fuel is
burned.
(c) Where different fossil fuels are
burned simultaneously in any combina-
tion, the applicable standard shall be
determined by proration. Compliance
shall be determined using the following
formula:
y(0.80)-l-z(1.2)
x+y+z
where:
x Is the percent of total heat input derived
from gaseous fossil fuel and,
y is the percent of total heat input derived
from liquid fossil fuel and,
z is the percent of total heat input derived
from solid fossil fuel.
§ 60.44 Standard for nitrogen oxides.
On and after the date on which the
performance test required to be con-
ducted by I 60.8 is initiated no owner or
operator subject to the provisions of this
part shall discharge or cause the dis-
charge into the atmosphere of nitrogen
oxides in excess of:
(a) 0.20 Ib. per million B.t.u. heat in-
put (0.36 g. per million cal.), maximum
2-hour average, expressed as NO2, when
gaseous fossil fuel is burned.
(b) 0.30 Ib. per million B.t.u. heat in-
put (0.54 g. per million cal.), maximum
2-hour average, expressed as NO2, when
liquid fossil fuel is burned.
(c) 0.70 Ib. per million B.t.u. heat in-
put (1.26 g. per million cal.), maximum
2-hour average, expressed as NOa when
solid fossil fuel (except lignite) is burned.
(d) When different fossil fuels are
burned simultaneously in any combina-
tion the applicable standard shall be de-
termined by proration. Compliance shall
be determined by using the following
formula:
x(0.20) +y(0.30) +z(0.70)
x+y+z
where:
x is the percent of total heat input derived
from gaseous fossil fuel and,
y is the percent of total heat input derived
from liquid fossil fuel and,
z is the percent of total heat input derived
from solid fossil fuel.
§ 60.45 Emission and fuel monitoring.
(a) There shall be installed, cali-
brated, maintained, and operated, in any
fossil fuel-fired steam generating unit
subject to the provisions of this part,
emission monitoring instruments as
follows:
(1) A photoelectric or other type
smoke detector and recorder, except
where gaseous fuel is the only fuel
burned.
(2) An instrument for continuously
monitoring and recording sulfur dioxide
emissions, except where gaseous fuel is
the only fuel burned, or where compli-
ance is achieved through low sulfur fuels
and representative sulfur analysis of
fuels are conducted daily in accordance
with paragraph (c) or (d) of this section.
(3) An instrument for continuously
monitoring and recording emissions of
nitrogen oxides.
(b) Instruments and sampling systems
installed and used pursuant to this sec-
tion shall be capable of monitoring emis-
sion levels within ±20 percent with a
confidence level of 95 percent and shall
be calibrated in accordance with the
method(s) prescribed by the manufac-
turer^) of such instruments; instru-
ments shall be subjected to manufactur-
ers recommended zero adjustment and
calibration procedures at least once per
24-hour operating period unless the man-
ufacturer^) specifies or recommends
calibration at shorter intervals, in which
case such specifications or recommenda-
tions shall be followed. The applicable
method specified in the appendix of this
part shall be the reference method.
(c) The sulfur content of solid fuels,
as burned, shall be determined in accord-
ance with the following methods of the
American Society for Testing and
Materials.
(1) Mechanical sampling by Method
D 2234065.
(2) Sample preparation by Method D
2013-65.
(3) Sample analysis by Method D
271-68.
(d) The sulfur content of liquid fuels,
as burned, shall be determined in accord-
ance with the American Society for Test-
ing and Materials Methods D 1551-68, or
D 129-64, or D 1552-64.
(e) The rate of fuel burned for each
fuel shall be measured daily or at shorter
intervals and recorded. The heating
value and ash content of fuels shall be
ascertained at least once per week and
recorded. Where the steam generating
unit is used to generate electricity, the
average electrical output and the mini-
mum and maximum hourly generation
rate shall be measured and recorded
daily.
(f) The owner or operator of any
fossil fuel-fired steam generating unit
subject to the provisions of this part
shall maintain a file of all measurements
required by this part. Appropriate meas-
urements shall be reduced to the units
of the applicable standard daily, and
summarized monthly. The record of any
such measurement(s) and summary
shall be retained for at least 2 years fol-
lowing the date of such measurements
and summaries.
§ 60.46 Test methods and procedures.
(a) The provisions of this section are
applicable to performance tests for de-
termining emissions of particulate mat-
ter, sulfur dioxide, and nitrogen oxides
from fossil fuel-fired steam generating
units.
(b) All performance tests shall be con-
ducted while the affected facility is oper-
ating at or above the maximum steam
production rate at which such facility
will be operated and while fuels or com-
binations of fuels representative of
normal operation are being burned and
under such other relevant conditions as
the Administrator shall specify based
on representative performance of the
affected facility.
(c) Test methods set forth in the
appenlfet to this part or equivalent
methods approved by the Administrator
shall be used as follows:
(1) For each repetition, the average
concentration of particulate matter shall
be determined by using Method 5.
Traversing during sampling by Method 5
shall be according to Method 1. The
minimum sampling time shall be 2 hours,
and minimum sampling volume shall be
60 ft.3 corrected to standard conditions
on a dry basis.
(2) For each repetition, the SO* con-
centration shall be determined by using
Method 6. The sampling site shall be the
same as for determining volumetric flow
rate. The sampling point in the duct
shall be at the centroid of the cross
section if the cross sectional area is less
than 50 ft.2 or at a point no closer to the
walls than 3 feet if the cross sectional
area is 50 ft.' or more. The sample shall
be extracted at a rate proportional to the
gas velocity at the sampling point. The
minimum sampling time shall be 20 min.
and minimum sampling volume shall be
0.75 ft.3 corrected to standard conditions.
Two samples shall constitute one repeti-
tion and shall be taken at 1-hour
intervals.
(3) For each repetition the NO, con-
centration shall be determined by using
Method 7. The sampling site and point
shall be the same as for SO». The sam-
pling time shall be 2 hours, and four
samples shall be taken at 30-minute
intervals.
(4) The volumetric flow rate of the
total effluent shall be determined by using
Method 2 and traversing according to
Method 1. Gas analysis shall be per-
formed by Method 3, and moisture con-
tent shall be determined- by the con-
denser technique of Method 5.
(d) Heat input, expressed in B.t.u. per
hour, shall be determined during each 2-
hour testing period by suitable fuel flow
meters and shall be confirmed by a ma-
terial balance over the steam generation
system.
(e) POT each repetition, emissions, ex-
pressed in Ib./lO* B.t.u. shall be deter-
mined by dividing the emission rate in
Ib./hr. by the heat input. The emission
rate shall be determined by the equation,
lb./hr.=Q.xc where, Q,=volumetric
Sow rate of the total effluent in ft.'/hr. at
standard conditions, dry basis, as deter-
mined in accordance with paragraph (c)
(4) of this section.
(1) For particulate matter, c=partic-
ulate concentration in lb./ft.3, at deter-
mined in accordance with paragraph (c)
(1) of this section, corrected to standard
conditions, dry basis
FEDERAL REGISTER, VOL. 36, NO. 247—THURSDAY, DECEMBER 23, 1971
1115
-------
24880
RULES AND REGULATIONS
(2) For SO, c=SO* concentration in
Ib./f t.3, as determined in accordance with
paragraph (c) (2) of this section, cor-
rected to standard conditions, dry basis;
(3) For NO*, c=NO, concentration in
lb./ft.s, as determined in accordance with
paragraph (c) (3) of this section, cor-
rected to standard conditions, dry basis.
Subpart E—Standards of Performance
for Incinerators
§ 60.50 Applicability and designation of
affected facility.
The provisions of this subpart are ap-
plicable to each incinerator of more than
50 tons per day charging rate, which is
the affected facility.
§ 60.51 Definitions.
As used in this subpart, all terms not
defined herein shall have the meaning
given them in the Act and in Subpart A
of this part.
(a) "Incinerator" means any furnace
used in the process of burning solid waste
for the primary purpose of reducing the
volume of the waste by removing com-
bustible matter.
(b) "Solid waste" means refuse, more
than 50 percent of which is municipal
type waste consisting of a mixture of
paper, wood, yard wastes, food wastes,
plastics, leather, rubber, and other com-
bustibles, and noncombustible materials
such as glass and rock.
(c) "Day" means 24 hours.
(d) "Particulate matter" means any
finely divided liquid or solid material,
other than uncombined water, as meas-
ured by Method 5.
§ 60.52 Standard for paniculate matter.
On and after the date on which the
performance test required to be con-
ducted by § 60.8 is initiated, no owner
or operator subject to the provisions of
this part shall discharge or cause the
discharge into the atmosphere of par-
ticulate matter which is in excess of 0.08
gr./s.c.f. (0.18 g./NM") corrected to 12
percent CO2, maximum 2-hour average.
§ 60.53 Monitoring of operations.
The owner or operator of any in-
cinerator subject to the provisions of this
part shall maintain a file of daily burn-
ing rates and hours of operation and any
particulate emission measurements. The
burning rates and hours of operation
shall be summarized monthly. The
record(s) and summary shall be retained
for at least 2 years following the date of
such records and summaries.
§ 60.54 Test methods and procedures.
(a) The provisions of this section are
applicable to performance tests for de-
termining emissions of particulate matter
from incinerators.
(b) All performance tests shall be
conducted while the affected facility is
operating at or above the maximum
refuse charging rate at which such facil-
ity will be operated and the solid waste
burned shall be representative of normal
operation and under such other relevant
conditions as the Administrator shall
specify based on representative per-
formance of the affected facility.
(c) Test methods set forth in the ap-
pendix to this part or equivalent methods
approved by the Administrator shall be
used as follows:
(1) For each repetition, the average
concentration of particulate matter shall
be determined by using Method 5. Tra-
versing during sampling by Method 5
shall be according to Method 1. The mini-
mum sampling time shall be 2 hours and
the minimum sampling volume shall be
60 ft.3 corrected to- standard conditions
on a dry basis.
(2) Gas analysis shall be performed
using the integrated sample technique of
Method 3, and moisture content shall be
determined by the condenser technique
of Method 5. If a wet scrubber is used,
the gas analysis sample shall reflect flue
gas conditions after the scrubber, allow-
ing for the effect of carbon dioxide ab-
sorption.
(d) For each repetition particulate
matter emissions, expressed in gr./s.c.f.,
shall be determined in accordance with
paragraph (c) (1) of this section cor-
rected to 12 percent CO,, dry basis.
Subpart F—Standards of Performance
for Portland Cement Plants
§ 60.60 Applicability and designation of
affected facility.
The provisions of the subpart are ap-
plicable to the following affected facili-
ties in Portland cement plants: kiln,
clinker cooler, raw mill system, finish
mill system, raw mill dryer, raw material
storage, clinker storage, finished prod-
uct storage, conveyor transfer points,
bagging and bulk loading and unloading
systems.
§ 60.61 Definitions.
As used in this subpart, all terms not
defined herein shall have the meaning
given them in the Act and in Subpart A
of this part.
(a) "Portland cement plant" means
any facility manufacturing Portland ce-
ment by either the wet or dry process.
(b) "Particulate matter" means any
finely divided liquid or solid material,
other than uncombined water, as meas-
ured by Method 5.
§ 60.62 Standard for particulate matter.
(a) On and after the date on which
the performance test required to be con-
ducted by § 60.8 is initiated no owner
or operator subject to the provisions of
this part shall discharge or cause the
discharge into the atmosphere of par-
ticulate matter from the kiln which is:
(1) In excess of 0.30 Ib. per ton of feed
to the kiln (0.15 Kg. per metric ton),
maximum 2-hour average.
(2) Greater than 10 percent opacity,
except that where the presence of uncom-
bined water is the only reason for failure
to meet the requirements for this sub-
paragraph, such failure shall not be a
violation of this section.
(b) On and after the date on which
the performance test required to be con-
ducted by 5 60.8 is Initiated no owner
or operator subject to the provisions of
this part shall discharge or cause the dis-
charge into the atmosphere of particulate
matter from the clinker cooler which is:
(1) In excess of 0.10 Ib. per ton of feed
to the kiln (0.050 Kg. per metric ton)
maximum 2-hour average.
(2) 10 percent opacity or greater.
(c) On and after the date on which the
performance test required to be con-
ducted by § 60.8 is initiated no owner
or operator subject to the provisions of
this part shall discharge or cause the
discharge into the atmosphere of partic-
ulate matter from any affected facility
other than the kiln and clinker cooler
which is 10 percent opacity or greater.
§ 60.63 Monitoring of operations.
The owner or operator of any portland
cement plant subject to the provisions
of this part shall maintain a file of daily
production rates and kiln feed rates and
any particulate emission measurements.
The production and,feed rates shall be
summarized monthly. The record (s) and
summary shall be retained for at least
2 years following the date of such records
and summaries.
§ 60.64 Test methods and procedures.
(a) The provisions of this section are
applicable to performance tests for de-
termining emissions-of particulate mat-
ter from Portland cement plant kilns
and clinker coolers.
(b) All performance tests shall be
conducted while the affected-facility is
operating at or above the maximum
production rate at which such facility
will be operated and under such other
relevant conditions as the Administrator
shall specify based on representative per-
formance of the affected facility.
(c) Test methods set forth in the ap-
pendix to this part or equivalent meth-
ods approved by the Administrator shall
be used as follows:
(1) For each repetition, the average
concentration of particulate matter shall
be determined by using Method 5. Tra-
versing during sampling by Method 5
shall be according to Method 1. The mini-
mum sampling time shall be 2 hours and
the minimum sampling volume shall be
60 ft.* corrected to standard conditions
on a dry basis.
(2) The volumetric flow rate of the
total effluent shall be determined by us-
ing Method 2 and traversing according to
Method 1. Gas analysis shall be per-
formed using the integrated sample tech-
nique of Method 3, and moisture content
shall be determined by the condenser
technique of Method 5.
(d) Total kiln feed (except fuels), ex-
pressed in tons per hour on a dry basis,
shall be determined during each 2-hour
testing period by suitable flow meters
and shall be confirmed by a material
balance over the production system.
(e) For each repetition, particulate
matter emissions, expressed in Ib./ton of
kiln feed shall be determined by dividing
ttie emission rate in Ib./hr. by the kiln
feed. The emission rate shall be deter-
mined by the equation, lb./hr.=Q«xc,
FEDERAL REGISTER, VOL. 36, NO. 247—THURSDAY, DECEMBER 23, 1971
1116
-------
RULES AND REGULATIONS
24881
where Q.=volumetric flow rate of the
total effluent in f t.'/hr. at standard condi-
tions, dry basis, as determined in ac-
cordance with paragraph (c) (2) of this
section, and, c=particulate concentra-
tion in lb./ft.*, as determined in accord-
ance with paragraph (c) (1) of this
section, corrected to standard conditions,
dry basis.
Subpart G—Standards of Performance
for Nitric Acid Plants
§ 60.70 Applicability and designation of
affected facility.
The provisions of this subpart are
applicable to each nitric acid production
unit, which is the affected facility.
§ 60.71 Definitions.
As used in this subpart, all terms not
defined herein shall have the meaning
given them in the Act and in Subpart A
of this part.
(a) "Nitric add production unit"
means any facility producing weak nitric
acid by either the pressure or atmos-
pheric pressure process.
(b) "Weafc nitric add" means add
which is 30 to 70 percent in strength.
§ 60.72 Standard for nitrogen oxides.
On and after the date on which the
performance test required to be con-
ducted by § 60.8 is initiated no owner
or operator subject to the provisions of
this part shall discharge or cause the
discharge Into the atmosphere of nitro-
gen oxides which are:
(a) In excess of 3 Ibs. per ton of acid
produced (1.5 kg. per metric ton),
maximum 2-hour average, expressed as
N02.
(b) 10 percent opadty or greater.
§ 60.73 Emission monitoring.
(a) There shall be installed, cali-
brated, maintained, and operated, in any
nitric add production unit subject to
the provisions of this subpart, an instru-
ment for continuously monitoring and
recording emissions of nitrogen oxides.
(b) The instrument and sampling
system1 installed and used pursuant to
this section shall be capable of monitor-
ing emission levels within ±20 percent
with a confidence level of 95 percent and
shall be calibrated in accordance with
the method(s) prescribed by the manu-
facturer (6) of such instrument, the
instrument shall be subjected to
manufacturers' recommended zero ad-
justment and calibration procedures at
least once per 24-hour operating period
unless the manufacturer (s) specifies or
recommends calibration at shorter In-
tervals, in which case such specifications
or recommendations shall be followed.
The applicable method specified in tbe
appendix of this part shall be the ref-
erence method.
(c) Production rate and hours of op-
eration shall be recorded daily.
(d) The owner or operator of any
nitric acid production unit subject to the
provisions of this part shall maintain
a file of all measurements required by
this subpart. Appropriate measurements
shall be reduced to the units of the
standard daily and summarized monthly.
The record of any such measurement
and summary shall be retained for at
least 2 years following the date of such
measurements and summaries.
§ 60.74 Test methods and procedures.
(a) The provisions of this section are
applicable to performance tests for de-
termining emissions of nitrogen oxides
from nitric acid production units.
(b) All performance tests shall be
conducted while the affected facility is
operating at or above the maximum acid
production rate at which such facility
will be operated and under such other
relevant conditions as the Administra-
tor shall specify based on representa-
tive performance of the affected facility.
(c) Test methods set forth in the ap-
pendix to this part or equivalent methods
as approved by the Administrator shall
be used as follows:
(1) For each repetition the NO, con-
centration shall be determined by using
Method 7. The sampling site shall be
selected according to Method 1 and the
sampling point shall be the centroid of
the stack or duct. The sampling time
shall be 2 hours and four samples shall
be taken at 30-minute intervals.
(2) The volumetric flow rate of the
total effluent shall be determined by
using Method 2 and traversing accord-
ing to Method 1. Gas analysis shall be
performed by using the integrated
sample technique of Method 3, and
moisture content shall be determined by
Method 4.
(d) Add produced, expressed in tons
per hour of 100 percent nitric acid, shall
be determined during each 2-hour test-
ing period by suitable flow meters and
shall be confirmed by a material bal-
ance over the production system.
(e) For each repetition, nitrogen
oxides emissions, expressed in Ib./ton
of 100 percent nitric acid, shall be de-
termined by dividing the emission rate
in Ib./hr. by the add produced. The
emission rate shall be determined by
the equation, lbyhr.=QsXc, where
Qa=volumetrlc flow rate of the effluent
In ft.'/hr. at standard conditions, dry
basis, as determined in accordance with
paragraph (c) (2) of this section, and
c=NO, concentration in lb./ft.', as de-
termined in accordance with paragraph
(c) (1) of tills section, corrected to stand-
ard conditions, dry basis.
Subpart H — Standards of Performance
for Sulfuric Acid Plants
§ 60.80 Applicability and designation of
affected facility.
The provisions of this subpart are ap-
plicable to each sulfuric acid production
unit, which is the affected facility.
§ 60.81 Definitions.
As used in this subpart, all terms not
defined herein shall have the meaning
given them in the Act and in Subpart A
of this part.
(a) "Sulfuric acid production unit"
means any faculty producing sulfuric
acid by the contact process by burning
elemental sulfur, alkylation acid, hydro-
gen sulfide, organic sulfides and mer-
captans, or acid sludge, but does not in-
clude facilities where conversion to sul-
furic acid is utilized primarily as a means
of preventing emissions to the atmos-
phere of sulfur dioxide or other sulfur
compounds.
(b) "Acid mist" means sulfuric acid
mist, as measured by test methods set
forth in this part.
§ 60.82 Standard for sulfur dioxide.
On and after the date on which the
performance test required to be con-
ducted by § 60.8 is initiated no owner or
operator subject to the provisions of this
part shall discharge or cause the dis-
charge into the atmosphere of sulfur
dioxide in excess of 4 Ibs. per ton of acid
produced (2 kg. per metric ton), maxi-
mum 2 -hour average.
§ 60.83 Standard for acid mist.
On and after the date on which the
performance test required to be con-
ducted by § 60.8 is initiated no owner or
operator subject to the provisions of this
part shall discharge or cause the dis-
charge into the atmosphere of acid mist
which is:
(a) In excess of 0.15 Ib. per ton of acid
produced (0.075 kg. per metric ton),
maximum 2-hour average, expressed as
(b) 10 percent opacity or greater.
§ 60.84 Emission monitoring.
(a) There shall be installed, cali-
brated, maintained, and operated, in any
sulfuric acid production unit subject to
the provisions of this subpart, an in-
strument for continuously monitoring
and recording emissions of sulfur dioxide.
(b) The instrument and sampling sys-
tem installed and used pursuant to this
section shall be capable of monitoring
emission levels within ±20 percent with
a confidence level of 95 percent and shall
be calibrated in accordance with the
RDERAl MGISlCt. VOL 36, NO. 147—THURSDAY, DECEMBER 23, 1971
1117
-------
24882
RULES AND REGULATIONS
^irPili^fU
»«ag|l§l8||sa|s5
*j!b3J|aM|«*3
lfSg|j)if|JSgI:i3g
* I'llilWf
2i!2?g||lr!r^I
«B3«MO
o
ui
o
2
O
i
3
II
5S
1118
-------
RULES AND REGULATIONS
24883
QJ CO
J3
E
tu s-
W JJ J- Q)
s- cr tu 03 -M
o> -I— .a at
> o E c: E
fd ex n o TO
i— CM VO
en ,u")
r*"» o
*3- O CM CO
i— CM CM CO
LD CO O)
*£ 10 CO
»-* M VQ CO »—
co i~ in
CM CM CO
63
ro cn r--. r-» en
CM <^0 O
U3 CM CO
i — «3" CO
LD cn
CM UD
LT) 1T> I—
co co en
en UD i —
f — CO CM
o in un o co CM ro
t— f-. in in
^- i— CM CO
in N- in f— r-
«— r- CO
co co en
r*v o o co
vo in in ro
10 10 s, co cn o t-
IU
U
tu
a
CM cd
i- CO
'" 12
*a o
is
^- m
0 >
•*- £
"^; c
0 0
o
"I
. '5
CM O
' ,C
•r- CO
o 2
CO £
CD
5 cn
2, to
.? CD
U. (Q
«
6
O
1119
-------
24884
RULES AND REGULATIONS
2.2.2 For rectangular stacks divide the
cross section into as many equal rectangular
areas as traverse points, sucll that the ratio
of the length to the width at the elemental
areas Is between one and two. Locate the
traverse points at the centroid of each equal
area according to Figure 1-3.
3. References.
Determining Dust Concentration in a Gas
Stream, ASME Performance Test Code #27,
New York, N.Y., 1957.
Devorkin, Howard, et al., Air Pollution
Source Testing Manual, Air Pollution Control
District, Los Angeles, Calif. November 1963.
Methods for Determination of Velocity,
Volume, Dust and Mist Content of Gases,
Western Precipitation Division of Joy Manu-
facturing Co., Los Angeles, Calif. Bulletin
WP-50, 1968.
Standard Method for Sampling Stacks for
Particulate Matter, In: 1971 Book of ASTM
Standards, Part 23, Philadelphia, Pa. 1971,
ASTM Designation D-2928-71.
METHOD 2 DETERMINATION OP STACK GAS
VELOCITY AND VOLUMETRIC FLOW RATE (TYPE
S PTTOT TUBE)
1. Principle and applicability.
1.1 Principle. Stack gas velocity is deter-
mined from tne gas density and from meas-
urement of the velocity head using a Type S
(Stausohelbe or reverse type) pitot tube.
1.2 Applicability. This method should be
applied only when specified by the test pro-
cedures for determining compliance with the
New Source Performance Standards.
2. Apparatus.
2.1 Pitot tube—Type S (Figure 2-1), or
equivalent, with a coefficient within ±5%
over the working range.
2.2 Differential pressure gauge—Inclined
manometer, or equivalent, to measure velo-
city head to within 10% of the minimum
value.
2.3 Temperature gauge—Thermocouple or
equivalent attached to the pitot tube to
measure stack temperature to within 1.5 % of
the minimum absolute stack temperature.
2.4 Pressure gauge—Mercury-filled TJ-tube
manometer, or equivalent, to measure stack
pressure to within 0.1 in. Hg.
2.5 Barometer—To measure atmospheric
pressure to within 0.1 in. Hg.
2.6 Gas analyzer—To analyze gas composi-
tion for determining molecular weight.
2.7 Pitot tube—Standard type, to cali-
brate Type S pitot tube.
3. Procedure.
3.1 Set up the apparatus as shown in Fig-
ure 2-1. Make sure all connections are tight
and leak free. Measure the velocity head and
temperature at the traverse points specified
by Method 1.
3.2 Measure the static pressure in the
stack.
3.3 Determine the stack gas molecular
weight by gas analysis and appropriate cal-
culations as indicated in Method 3.
PIPE COUPLING
TUBING ADAPTER
4. Calibration.
4.1 To calibrate the pitot tube, measure
the velocity head at some point in a flowing
gas stream with both a Type S pitot tube and
a standard type pitot tube with known co-
efficient. Calibration should be done in the
laboratory and the velocity of the flowing gas
stream should be varied over the normal
working range. It is recommended that the
calibration be repeated after use at each field
site.
4.2 Calculate the pitot tube coefficient
using equation 2-1.
*• o ,
Apt,,, equation 2-1
where :
Cp,esl=Pitot tube coefficient of Type S
pitot tube.
Cn,td=Pitot tube coefficient of standard
type pitot tube (if unknown, use
0.99) .
Ap,tt = Velocity head measured by stand-
ard type pitot tube.
Apte«t=: Velocity head measured by Type S
pitot tube.
4.3 Compare the coefficients of the Type S
pitot tube determined first with one leg and
then the other pointed downstream. Use the
pitot tube only if the two coefficients differ by
no more than 0.01.
5. Calculations.
Use equation 2-2 to calculate the stack gas
velocity.
where-
(Va)«
Equation 2-2
= Stack gas velocity, feet per second (f.p.s ).
C0=Pltot tube coefficient, dimenslonless.
(T8)avB.=Average absolute stack gas tempeiature,
°
= Average velocity head of stack gas, inches
H,0 (see Fig. 2-2).
P,=Absolute stack gas pressure, inches Hg.
Ma=Molecular weight of stack gas (wet basis),
Ib /Ib.-mole.
Md(l— B,o)+18B,0
Md=Dry molecular weight of stack gas (from
Methods).
Bwo= Proportion by volume of water vapor in
the gas stream (from Method 4).
Figure 2-2 shows a sample recording sheet
for velocity traverse data. Use the averages
in the last two columns of Figure 2-2 to de-
termine the average stack gas velocity from
Equation 2-2.
Use Equation 2-3 to calculate the stack
gas volumetric flow rate.
Q.=3600
Figure 2-1. Pitot tube-manometer assembly.
Equation 2-3
where:
Q.=Volumetric flow rate, dry basis, standard condi-
tions, ft.'/hr.
A = Cross-sectional area of stack, ft.'
T,td*=Absolute temperature at standard conditions,
630° R.
Patd^AbsoIute pressure at standard conditions, 29.93
inches Hg.
FEDERAL REGISTER, VOL. 36, NO. 247—THURSDAY, DECEMBER 23, 1971
1120
-------
RULES AND REGULATIONS
24885
6. References.
Mark, L. S., Mechanical Engineers' Hand-
book, McGraw-Hill Book Co., Inc., New York,
N.Y., 1951.
Perry, J. H., Chemical Engineers' Hand-
book, McGraw-Hill Book Co., Inc., New York,
N.Y., I960.
Shigehara, K. T., W. F. Todd, and W. S.
Smith, Significance of Errors in Stack Sam-
pling Measurements. Paper presented at the
Annual Meeting of the Air Pollution Control
Association, St. Louis, Mo., June 14-19, 1970.
Standard Method for Sampling Stacks for
Particulate Matter, In: 1971 Book of ASTM
Standards, Part 23, Philadelphia, Pa., 1971,
ASTM Designation D-2928-71.
Vennard, J. K., Elementary Fluid Mechan-
ics, John Wiley & Sons, Inc., New York, N.Y.,
1947.
PLANT,
DATE
RUN NO.
STACK DIAMETER, in.
BAROMETRIC PRESSURE, in.
STATIC PRESSURE IN STACK (Pg), in. Hg._
OPE R ATORS
SCHEMATIC OF STACK
CROSS SECTION
Traverse point
number
Velocity head,
in. H20
Stack Temperature
AVERAGE:
Figure 2-2. Velocity traverse data.
FEDERAL REGISTER, VOL. 36, NO. 247—THURSDAY, DECEMBER 23, 1971
1121
-------
24886
RULES AND REGULATIONS
METHOD 3 GAS ANALYSIS FOE CARBON DIOXIDE,
EXCESS AIR, AND DRY MOLECtTLAR WEIGHT
1. Principle and applicability.
1.1 Principle. An integrated or grab gas
sample is extracted from a sampling point
and analyzed for its components using an
Orsat analyzer.
1.2 Applicability. This method should be
applied only when specified by the test pro-
cedures for determining compliance with the
New Source Performance Standards. The test
procedure will indicate whether a grab sam-
ple or an integrated sample is to be used.
2. Apparatus.
2.1 Grab sample (Figure 3-1).
2.1.1 Probe—Stainless steel or Pyrex1
glass, equipped with a filter to remove partic-
ulate matter.
2.1.2 Pump—One-way squeeze bulb, or
equivalent, to transport gas sample to
analyzer.
1 Trade name.
2.2 Integrated sample (Figure 3-2).
2.2.1 Probe—Stainless steel or Pyrex1
glass, equipped with a filter to remove par-
ticulate matter.
2.2.2 Air-cooled condenser or equivalent—
To remove any excess moisture.
2.2.3 Needle valve—To adjust flow rate.
2.2.4 Pump—Leak-free, diaphragm type,
or equivalent, to pull gas.
2.2.5 Bate meter—To measure a flow
range from 0 to 0.035 cfm.
2.2.6 Flexible bag—Tedlar,1 or equivalent,
with a capacity of 2 to 3 cu. ft. Leak test the
bag in the laboratory before using.
2.2.7 Pitot tube—Type S, or equivalent,
attached to the probe so that the sampling
flow rate can be regulated proportional to
the stack gas velocity when velocity is vary-
ing with time or a sample traverse is
conducted.
2 3 Analysis.
2.3.1 Orsat analyzer, or equivalent.
PROBE
'FLEXIBLE TUBING
TO ANALYZER
LTER (G
FILTER (GLASS WOOL)
SQUEEZE BULB
Figure 3-1. Grab-sampling train.
RATE METE?
VALVE
AIR-COOLED. CONDENSER / PUMP
PROBE
FILTERlGLASSYIIOOL}
QUICK DISCONNECT
RIGID CONTAINER'
Figure 3-2. Integrated gas • sampling train.
3. Procedure.
3 1 Grab sampling.
3.1.1 Set up the equipment as shown in
Figure 3-1, making sure all connections are
leak-free. Place the probe in the stack at a
sampling point and purge the sampling line.
3.1.2 Draw sample into the analyzer.
3.2 Integrated sampling.
3.2.1 Evacuate the flexible bag. Set up the
equipment as shown in Figure 3-2 with the
bag disconnected. Place the probe In the
stack and purge the sampling line. Connect
the bag, making sure that all connections are
tight and that there are no leaks.
3.2.2"- Sample at a rate proportional to the
stack velocity.
3.3 Analysis.
3.3.1 Determine the CO2, O2, and CO con-
centrations as soon as possible. Make as many
passes as are necessary to give constant read-
ings. If more than ten passes are necessary,
replace the absorbing solution.
3.3.2 For grab sampling, repeat the sam-
pling and analysis until three consecutive
samples vary no more than 0.5 percent by
volume for each component being analyzed.
3.3.3 For integrated sampling, repeat the
analysis of toe sample until three consecu-
tive analyses vary no more than 0.3 percent
by volume for each component being
analyzed.
4. Calculations.
4.1 Cartoon dioxide. Average the three con-
secutive runs and report the result to the
nearest 0.1% CO-
4.2 Excess air". Use Equation 3-1 to calcu-
late excess air, and average the runs. Report
the result to the nearest 0.1% excess air.
%EA =
(%02)-0.5(%CO)
0.264(% N,)-(% 02)+0.5(% CO)X1UU
equation 3-1
where:
%EA=Percent excess air.
%O3=Percent oxygen by volume, dry basis.
%N3=Percent nitrogen by volume, dry
basis.
% CO=Percent carbon monoxide by vol-
ume, dry basis.
0.264=Ratio of oxygen to nitrogen in air
by volume.
4.3 Dry molecular weight. Use Equation
3-2 to calculate dry molecular weight an-mole.
% COn=Percent carbon dioxide by volume,
dry basis.
%Oa=Percent oxygen by volume, dry
basis.
%Ni=Percent nitrogen by volume, dry
basis.
0.44=Molecular weight of carbon dioxide
divided by 100.
0.32=Molecular weight of oxygen divided
by 100.
0.28=Molecular weight of nitrogen and
CO divided by 100.
FEDERAL REGISTER, VOL. 36, NO. J47—THURSDAt, DECEMBER 23. 1971
1122
-------
RULES AND REGULATIONS
24887
in
§
I
Q
C 0
*5
at
tz
"a.
rt
u>
a>
w
'o
S
„.:
••a-
a)
3
O)
LL
2
O
S te
3 P
u.
§
a
i
S3
K
fl_
O
OC
^ £
S 8
UJ £C
fe <
O co
UJ
cc
5
OC
UJ
^
S u-
uj a
H-
ec
UJ
£
S
(3
z
(-•
S
In.e
DC *E
S <
*~ CO
U; -
<
g*"
I
o
o
o ^
M
S ~-
aa-SB'^j.
SS jajSa
Ipfl1
cs
-»§"S«3§«aS^§i5 .
^iriiilig!i°ir
Ss" 0«
S§S¥S °
•o$33aa
3 1 » 2 £2
&= a fe &-g ,
s « t>>
SOP
o
o
_
I
> «
1123
-------
24888
RULES AND REGULATIONS
4.2 Oas volume.
17 71 -
'
in. Hg V Tm / equation 4-2
where:
Vmc =Dry gas volume through meter at
standard conditions, cu. ft.
Vm =Dry gas volume measured by meter,
cu. ft.
Pm = Barometric pressure at the dry gas
meter, Inches Hg.
P.td=Pressure»t standard conditions, 29.92
Inches Hg.
T»td=Absolute temperature at standard
conditions, 530° B.
Tm = Absolute temperature at meter (° F+
460), °B.
4.3 Moisture content.
T> *W9 I T3 • WO
-+(0.025)
equation 4-3
where:
Bwo=Proportion by volume of water vapor
in the gas stream, dimensionless.
Vwc =Volume of water vapor collected
(standard conditions), cu. ft.
Vmc =Dry gas volume through meter
(standard conditions), cu. ft.
BWM=Approximate volumetric proportion
of water vapor in the gas stream
leaving the implngers, 0.025.
5. References.
Air Pollution Engineering Manual, Daniel-
son, J. A. (ed.), U.S. DHEW, PHS, National
Center for Air Pollution Control, Cincinnati,
Ohio, PHS Publication No. 999-AP-40, 1967.
Devorkln, Howard, et al., Air Pollution
Source Testing Manual, Air Pollution Con-
trol District, Los Angeles, Calif., November
1963.
Methods for Determination of Velocity,
Volume, Dust and Mist Content of Gases,
Western Precipitation Division of Joy Manu-
facturing Co., Los Angeles, Calif., Bulletin
WP-60, 1968.
METHOD 5—DETERMINATION OF PARTICULATE
EMISSIONS FROM STATIONARY SOURCES
1. Principle and applicability.
1.1 Principle. Particulate matter is with-
drawn Isokinetioally from the source and its
weight Is determined gravimetrically after re-
moval of uncomiblned water.
1.2 Applicability. This method is applica-
ble for the determination of particulate emis-
sions from stationary sources only when
specified by the test procedures for determin-
ing compliance with New Source Perform-
ance Standards.
2. Apparatus.
2.1 Sampling train. The design specifica-
tions.of the particulate sampling train used
by EPA (Figure 5-1) are described in APTD-
0581. Commercial models of this train are
available.
2.1.1 Nozzle—Stainless steel (316) with
sharp, tapered leading edge.
2.1.2 Probe—Pyrex1 glass with a heating
system capable of maintaining a minimum
gas temperature of 250° F. at the exit end
during sampling to prevent condensation
from occurring. When length limitations
(greater than about 8 ft.) are encountered at
temperatures less than 600° F., Incoloy 825 *,
or equivalent, may be usedv Probes for sam-
pling gas streams at temperatures in excess
of 600° F. must have been approved by the
Administrator.
2.1.3 Pitot tube—Type S, or equivalent,
attached to probe to monitor stack gas
velocity.
2.1.4 Filter Holder—Pyrex» glass with
heating system capable of maintaining mini-
mum temperature of 225° F.
2.1.5 Implngers / Condenser—Four impin-
gers connected in series with glass ball Joint
fittings. The first, third, and fourth impin-
gers are of the Greenburg-Smitn design,
modified by replacing the tip with a 2-lnch
ID glass tube extending to one-half inch
from the bottom of the flask. The second 1m-
pinger is of the Greenburg-Smlth design
with the standard tip. A condenser may be
used in place of the impingers provided that
the moisture content of the stack gas can
still be determined.
2.1.6 Metering system—Vacuum gauge,
leak-free pump, thermometers capable of
measuring temperature to within 5° F., dry
gas meter with 2% accuracy, and related
equipment, or equivalent, as required to
maintain an isokinetic sampling rate and to
determine sample volume.
2.1.7 Barometer—To measure atmospheric
pressure to ±0.1 Inches Hg.
2.2 Sample recovery.
2.2.1 Probe brush—At least as long as
probe.
2.2.2 Glass wash bottles—Two.
2.2.3 Glass sample storage containers.
2.2.4 Graduated cylinder—250 ml.
2.3 Analysis.
2.3.1 Glass weighing dishes.
2.3.2 Desiccator.
2.3.3 Analytical balance—To measure to
±0.1 mg.
2.3.4 Trip balance—300 g. capacity, to
measure to ± 0.05 g.
3. Reagents.
3.1 Sampling.
3.1.1 Filters—Glass fiber, MSA 1106 BH1,
or equivalent, numbered for identification
and preweighed.
3.1.2 Silica gel—Indicating type, 6-16
mesh, dried at 175° C. (350° F.) for 2 hours.
3.1.3 Water.
3.1.4 Crushed ice.
3.2 Sample recovery.
3 2.1 Acetone—Reagent grade.
3.3 Analysis.
3.3.1 Water.
IMPINGER TRAIN OPTIONAL. MAY BE REPLACED
BY AN EQUIVALENT CONDENSER
HEATED AREA KILTER HOLDER / THERMOMETER CHECK
^VALVE
,VACUUM
LINE
PIT01 MANOMETER
ORIFICE
THERMOMETERS
IMPINGERS ICE BATH
BY-PASSVALVE
DRY TEST METER
AIR-TIGHT
PUMP
Figure 5-1. Particulate-sampling train.
3 3.2 Desiccant—Drierite,' indicating.
4 Procedure.
4.1 Sampling
41.1 After selecting the sampling site and
the minimum number of sampling points,
determine the stack pressure, temperature,
moisture, and range of velocity head.
4.1.2 Preparation of collection train.
Weigh to the nearest gram approximately 200
g. of silica gel. Label a filter of proper diam-
eter, desiccate2 for at least 24 hours and
weigh to the nearest 0.5 mg. in a room where
the relative humidity is less than 50 vr. place
100 ml. of water in each of the first two
Impingers, leave the third impinger empty,
and place approximately 200 g. of preweighed
silica gel in the fourth impinger. Set up the
train without the probe as in Figure 5-1.
Leak check the sampling train at the sam-
pling site by plugging up the inlet to the fil-
ter holder and pulling a 15 in. Hg vacuum. A
leakage rate not In excess of 0.02 c.f.m. at a
vacuum of 15 in. Hg is acceptable. Attach
the probe and adjust the heater to provide a
gas temperature ot about 250° F. at the probe
outlet. Turn on the filter heating system.
Place crushed ice around the impingers Add
1 Trade name.
1 Trade name.
'Dry using Drierite1 at 70° F.±10°
F.
more ice during the run to keep the temper-
ature of the gases leaving the last Impinger
as low as possible and preferably at 70° F
or less. Temperatures above 70° F. may result
in damage to the dry gas meter from either
moisture condensation or excessive heat.
4.1.3 Particulate train operation. For each
run, record the data required on the example
sheet shown in Figure 5-2. Take readings at
each sampling point, at least every 5 minutes,
and when significant changes in stack con-
ditions necessitate additional adjustments
in flow rate To begin sampling, position the
nozzle at the first traverse point with the
tip pointing directly into the gas stream
Immediately start the pump and adjust the
flow to isokinetic conditions. Sample for at
least 5 minutes at each traverse pjint, san:-
pling time must be the same for each pour.
Maintain isokinetic sampling throughout the
sampling period. Nomographs are available
which aid in the rapid adjustment of the
sampling rate without other computation.-
APTD-0676 details the procedure for vising
these nomographs. Turn off the pump at the
conclusion of each run and record the final
readings Remove the probe and nozzle frcm
the stack and handle in accordance with the
sample recovery process described in section
4.2.
FEDERAL REGISTER, VOL. 36, NO. 247—THURSDAY, DECEMBER 23, 1971
1124
-------
RULES AND REGULATIONS
24889
PLANT .
LOCATION
OPERATOR __,
DATE
BUN NO.
SAMPLE BOX N0j_
METER BOX NO,
METER AH.,
C FACTOR
AMBIENT TEMPERATUflf _
BAROMETRIC PRESSURE_
ASSUMED MOISTURE. *__
HEATtR BOX SETTING
PROBE LENGTH, »
NOZZLE DIAMETER, ln._
PflOBt HEATER SETTING,
SCHEMATIC Of STACK CROSS SECTION
TRAVERSE POINT
NUMBER
TOTAL,
SAMPLING
TIME
(•). min.
AVERAGE
STATIC
PRESSURE
IPS). fc H9
STACK
TEMPERATURE
ITS)."F
VELOCITY
HEAD
I*PS>.
PRESSURE
DIFFERENTIAL
ACROSS
ORIFICE
METER
(AH),
In, H2O
GASSAMPU
VOLUME
IVm) It3
GAS SAMPLE TEMPERATURE
AT DRY GAS METER
INLET
tT(H}n>1,*F
A«g.
OUTLET
IT-ou.l.-f
Avg.
Avg.
SAMPLE BOX
TEMPERATURE.
TEMPERATURE.
OF GAS
LEAVIHG
CONDENSER OR
LAST IMPINCER
Tm = Average dry gas meter temperature,
°R.
Pbir = Barometric pressure at the orifice
meter, inches Hg.
AH = Average pressure drop across the
orifice meter, inches H.O.
13.6 = Specific gravity of mercury.
Plld= Absolute pressure at standard con-
ditions, 29.92 inches Hg.
6.3 Volume of water vapor
4.2 Sample recovery. Exercise care in mov-
ing the collection train from the test site to
the sample recovery area to minimize the
loss of collected sample or the gain of
extraneous particulate matter. Set aside a
portion of the acetone used in the sample
recovery as a blank for analysis. Measure the
volume of water from the first three im-
pingers, then discard. Place the samples in
containers as follows:
Container No. 1. Remove the filter from
its holder, place in this container, and seal.
Container No. 2. Place loose particulate
matter and acetone washings from all
sample-exposed surfaces prior to the filter
in this container and seal. Use a razor Wade,
brush, or rubber policeman to lose adhering
particles.
Container No. 3. Transfer the silica gel
from the fourth impinger to the original con-
tainer and seal. Use a rubber policeman as
an aid in removing silica gel from the
Impinger.
4.3 Analysis. Record the data required on
the example sheet shown in Figure 5-3.
Handle each sample container as follows:
Container No. 1. Transfer the filter and
any loose particulate matter from the sample
container to a tared glass weighing dish,
desiccate, and dry to a constant weight. Re-
port results to the nearest 0,5 mg.
Container No. 2. Transfer the acetone
washings to a tared beaker and evaporate to
dryness at ambient temperature and pres-
sure. Desiccate and dry to a constant weight.
Report results to the nearest 0.5 mg.
Container No. 3. Weigh the spent silica gel
and report to the nearest gram.
5. Calibration.
Use methods and equipment which have
been approved by the Administrator to
calibrate the orifice meter, pitot tube, dry
gas meter, and probe heater. Recalibrate
after each test series.
6. Calculations.
6.1 Average dry gas meter temperature
and average orifice pressure drop. See data
sheet (Figure 5-2).
6.2 Dry gas 'volume. Correct the sample
volume measured by the dry gas meter to
standard conditions (70° F., 29.92 inches Hg)
by using Equation 5-1.
V -V /T....\(Fb"+ibV
'"d "VT,J\ P.ld /
(0.0474 51^) V,.
equation 5-2
where :
VwlU= Volume of water vapor in the gas
sample (standard conditions) ,
cu. ft.
Vi0 = Total volume of liquid collected in
impingers and silica gel (see Fig-
ure 5-3 ) , ml.
P«jO= Density of water, 1 g./rnl.
MH,O= Molecular weight of water, 18 lb./
Ib.-mole.
B = Ideal gas constant, 21.83 inches
Hg — cu. ft./lb.-mole-°R.
T,ta= Absolute temperature at standard
conditions, 530° R.
P,,4= Absolute pressure at standard con-
ditions, 29.92 inches Hg.
6.4 Moisture content.
V
'"id
,
13-6
equation 5-1
where :
Vm,td= Volume of gas sample through the
dry gas meter (standard condi-
tions) , cu. ft.
Vra= Volume of gas sample through the
dry gas meter (meter condi-
tions) , cu. ft.
T.,d= Absolute temperature at standard
conditions, 530* R.
equation 3-3
wheie'
Bwo
— Pioportkm by volume of watei vapor in the ^as
stieam, dimensionless.
^"btd =Volume of water in the gas sample (stand-aid
conditions) , cu. ft.
^"Vd = Volume of gas sample through the dry gas motcr
(standard conditions) , cu. ft.
6.6 Total particulate weight. Determine
the total particulate catch from the sum of
the weights on the analysis data sheet
(Figure 5-3) .
6.6 Concentration.
6.6. 1 Concentration in gr./s c.f .
c'.= 0.0154
equation 5-4
where:
c'.= Concentration of particulate matter in stack
gas, gr./s.c.f., dry basis.
M.=Total amount of particulate matter collected,
mg.
^matd=Volume of gas sample through dry gas meter
(standaid conditions), cu. ft,
FEDERAL REGISTER, VOL. 36, NO. 247—THURSDAY, DECEMBER 23. 1971
-------
24890
RULES AND REGULATIONS
PLANT.
DATE_
RUN NO.
CONTAINER
NUMBER
1
2
TOTAL
WEIGHT OF PARTICULATE COLLECTED,
mg
FINAL WEIGHT
:xi
TARE WEIGHT
X
WEIGHT GAIN
FINAL
INITIAL
LIQUID COLLECTED
TOTAL VOLUME COLLECTED
VOLUME Of LIQUID
WATER COLLECTED
IMPINGER
VOLUME.
ml
SILICA GEL
WEIGHT,
9
g* ml
CONVERT WEIGHT OF WATER TO VOLUME BY DIVIDING TOTAL WEIGHT
INCREASE BY DENSITY OF WATER. (1 g. ml):
= VOLUME WATER, ml
Figure5-3. Analytical data.
6.6.2 Concentration In Ib./cu. ft.
c ^(453^00 rZj:)M"^OOA,win_A M,,
where1
cfl=Concentiatioii of pattieulate matter in stack
gas, Ib./s.c.f., diy ba.sis.
453,600=Mg/lb.
^"ttd equation 5-5
Mn = Tot,U amount of paiticulate matter collected,
Vm,,j= Volume of gas sample through dry gas meter
(standard conditions), cu. ft.
6 7 Isokinetic variation.
T
iB2O
-xioo
when1'
I = Perccnt of isokinctic sampling.
Vjc=Total volume of liquid collected lii Impingets
and silica gel (Sec Fig. 5-3), nil.
pH2o=Density of water, 1 g./ml.
K=Ideal gas constant, 21.83 inches Hg-cu. ft./lb.
moio-°R.
MH,O =Molecular weight of water, 18 Ib /Ib.-mole.
Vm = Volume of gas sample through the di y gas meter
(metei conditions), eu. ft.
Tm= Absolute average dry gas meter temperature
(see Figiue6-2),°K.
Fbar=Baiomctue pressuie at sampling site, Indies
UK.
AH=Aveiagc pressuie drop across the orifice {see
Fig. 5-2), inches H2O.
T,=Absolute aveiagc stack gas temperature (see
Fig. 5-2),°It.
0=Total sampling time, min.
V,=8tack gas velocity calculated by Method 2,
Equation 2-2, ft /sec.
l\=Absolute stack gas piossure, inches Hg.
An = Cross-sectional area of nozzle, sq. ft.
6.8 Acceptable results. The following
range sets the limit on acceptable isokinetic
sampling results:
If 90%
-------
RULES AND REGULATIONS
24891
n 8« *~a •* '• 3 s
o ~S&* -S«~
o
1127
-------
24892
RULES AND REGULATIONS
nitrous oxide, are measure eolorimetrically
using the phenoldlsulfonic acid (PDS)
procedure.
1.2 Applicability. This method is applica-
ble for the measurement of nitrogen oxides
from stationary sources only when specified
by the test procedures for determining com-
pliance with New Source Performance
Standards.
2. Apparatus.
2.1 Sampling. See Figure 7-1.
2.1.1 Probe—Pyrex1 glass, heated, with
filter to remove particulate matter. Heating
is unnecessary if the probe remains dry dur-
ing the purging period.
2.1.2 Collection flask—Two-liter, Pyrex,'
round bottom with short neck and 24/40
standard taper opening, protected against
Implosion or breaKage.
1 Trade name.
2.1.3 Flask valve—T-bore stopcock con-
nected to a 24/40 standard taper Joint.
2.1.4 Temperature gauge—Dial-type ther-
mometer, or equivalent, capable of measur-
ing 2° F. intervals from 25' to 125' P.
2.1.5 Vacuum line—Tubing capable of
withstanding a vacuum of 3 inches Hg abso-
lute pressure, with "T" connection and T-bore
stopcock, or equivalent.
2.1 6 Pressure gauge—U-tube manometer,
36 inches, with 0.1-inch divisions, or
equivalent.
2 1.7 Pump—Capable of producing a vac-
uum of 3 inches Hg absolute pressure.
2.1.8 Squeeze bulb—Oneway.
2.2 Sample recovery.
2.2 1 Pipette or dropper.
2.2.2 Glass storage containers—Cushioned
for shipping.
PROBE
A
fILTER
GROUND-GLASS SOCKET,
g NO. M/S
f LASK SHIELD-, ,\
GROUNO-GLAS:
STANDARD TAPER,
J SLEEVE NO, 24/40
GROUND-GLASS
SOCKET, § NO. 12,5
PYREX
FOAM ENCASEMENT
BOILING FLASK -
2 LITER. ROUND-BOTTOM, SHOUT 1CCK.
WITH J SLEEVE NO. 24/40
Figure 7-1, Sampling uain, l)ask valve, and flask.
2.2.3 Glass wash bottle.
2.3 Analysis.
2.3.1 Steam Datn.
2.3.2 BeaKers or casseroles—250 ml., one
for each sample and standard (blank).
2.3.3 Volumetric pipettes—1, 2, and 10 ml.
2.3.4 Transfer pipette—10 ml. with 0.1 ml.
divisions.
2.3.5 Volumetric flask—100 ml., one for
each sample, and 1,000 ml. for the standard
(blank).
2.3.6 Spectrophotometer—To measure ab-
Eorbance at 420 urn.
2.3.7 Graduated cylinder—100 ml. with
1.0ml. divisions.
2.3.8 Analytical balance—To measure to
0.1 mg.
3. Reagents.
3.1 Sampling.
3.1.1 Absorbing solution—Add 2.8 ml. of
concentrated H,SO, to 1 liter of distilled
water. Mix well and add 6 ml. of 3 percent
hydrogen peroxide. Prepare a fresh solution
weekly and do not expose to extreme heat or
direct sunlight.
3.2 Sample recovery.
3.2.1 Sodium hydroxide (IN)—Dissolve
40 g. NaOH in distilled water and dilute to 1
liter.
3.2.2 Red litmus paper.
3.2.3 Water-—Deionized, distilled.
3.3 Analysis.
3.3.1 Fuming sulfurlc acid—15 to 18% by
weight free sulfur trioxide.
3.3.2 Phenol—White solid reagent grade.
3.3.3 Sulfuric acid—Concentrated reagent
grade.
3.3.4 Standard solution—Dissolve 0.5495 g.
potassium nitrate (KNOS) in distilled water
and dilute to 1 liter. For the working stand-
ard solution, dilute 10 ml. of the resulting
solution to 100 ml. with distilled water. One
ml. of the working standard solution is
equivalent to 25 /ig. nitrogen dioxide.
3.3.5 Water—Deionized, distilled.
3.3.6 Phenoldlsulfonlc acid solution—
Dissolve 25 g. of pure white phenol in 150 ml.
concentrated sulfurlc acid on a steam bath.
Cool, add 75 ml. fuming sulfuric acid, and
heat at 100° C. for 2 hours. Store in a dark,
stoppered bottle.
4. Procedure.
4.1 Sampling.
4.1.1 Pipette 25 ml. of absorbing solution
Into a sample flask. Insert the flask valve
stopper into the flask with the valve in the
"purge" position. Assemble the sampling
train as shown. In Figure 7-1 and place the
probe at the sampling point. Turn the flask
valve and the pump valve to their "evacuate"
positions. Evacuate the flask to at least 3
inches Hg absolute pressure. Turn the pump
valve to its "vent" position and turn off the
pump. Check the manometer for any fluctu-
ation in tne mercury level. If there is a visi-
ble change over the span of one minute,
check for leaks. Record the initial volume,
temperature, and barometric pressure. Turn
the flask valve to its "purge" position, and
then do the same with the pump valve.
Purge the probe and the vacuum tube using
the squeeze bulb. If condensation occurs in
the probe and flask valve area, heat the probe
and purge-.until the condensation disappears.
Then turn the pump valve to Its "vent" posi-
tion. Turn the flask valve to Its "sample"
position and allow sample to enter the flask
for about 15 seconds. After collecting the
sample, turn the flask valve to its "purge"
position and disconnect the flask from the
sampling train. Shake the flask for 5
minutes.
4 2 Sample recovery.
4.2.1 Let the flask set for a minimum of
16 hours and then shake the contents for 2
minutes. Connect the flask to a mercury
filled U-tube manometer, open the valve
from the flask to the manometer, and record
the flask pressure and temperature along
with the barometric pressure. Transfer the
flask contents to a container for shipment
or to a 250 ml. beaker for analysis. Rinse the
flask with two portions of distilled water
(approximately 10 ml.) and add rtnse water
to tne sample. For a blank use 25 ml. of ab-
sorbing solution and the same volume of dis-
tilled water as used in rinsing the flask. Prior
to shipping or analysis, add sodium hydrox-
ide (IN) dropwlse into both the sample and
the blank until alkaline to litmus paper
(about 25 to 35 drops in each).
4.3 Analysis.
4.3 1 If the sample has been shipped in
a container, transfer the contents to a 250
ml. beaker using a small amount of distilled
water. Evaporate the solution to dryness on a
steam bath and then cool. Add 2 ml. phenol-
disulfonlc acid solution to the dried residue
and triturate thoroughly with a glass rod.
Make sure the solution contacts all the resi-
due. Add 1 ml. distilled water and four drops
of concentrated sulfuric acid. Heat the solu-
tion on a steam bath for 3 minutes with oc-
casional stirring. Cool, add 20 ml. distilled
water, mix well by stirring, and add concen-
trated ammonium hydroxide dropwise with
constant stirring until alkaline to litmus
paper. Transfer the solution to a 100 ml.
volumetric flask and wash the beaker three
times with 4 to 5 ml. portions of distilled
water. Dilute to the mark and mix thor-
oughly. If the sample contains solids, trans-
fer a portion of the solution to a clean, dry
centrifuge tube, and centrifuge, or niter a
portion of the solution. Measure the absorb-
auce of eacn sample at 420 nm. using the
blank solution as a zero. Dilute the sample
and the blank with a suitable amount of
distilled water if absorbance falls outside the
range of calibration.
5. Calibration.
5.1 Flask volume. Assemble the flask and
flask valve and fill with water to the stop-
cock. Measure the volume of water to ±10
ml. Number and record the volume on the
flask.
5.2 Spectrophotometer. Add 0.0 to 16.0 ml.
of standard solution to a series of beakers. To
each beaker add 25 ml. of absorbing solution
and add sodium hydroxide (IN) dropwlse
until alkaline to litmus paper (about 25 to
35 drops). Follow the analysis procedure of
section 4.3 to collect enough data to draw a
calibration curve of concentration In /«g. NO»
per sample versus absorbance.
6. Calculations.
6.1 Sample volume.
FEDERAL REGISTER, VOL. 36, NO. 247—THURSDAY, DECEMBER 23, 1971
1128
-------
RULES AND REGULATIONS
24893
v..=-
P.ui
where:
Vsc= Sample volume at standard condi-
tions (dry basis), ml.
T
-------
24894
RULES AND REGULATIONS
51ili
-r* -w ^ '
1 3!
< g
f i
3 °
*
?
WED MOIS
5
•"
O
J
I « B
1 I !
S o ^
<£ £ S
1
£
2
i
S
*
I i
I
O ™
ill
g
m
S
SS SECTION
5
•t
S
i
K
' 6
! E
sj
C
^
i*
I
2 £
IS
Ii
i
i-
u.
t*.
3 5
^^
^~
-'i
III
Et*o =
"
Cj I ^
Ifi-
s
ill
I*i
r
I
i i
M
o
Ul
a
1130
-------
RULES AND REGULATIONS
24895
Bom, Jerome J., Maintenance, Calibration,
and Operation of Isokinetic Source Sam-
pling Equipment, Environmental Protection
Agency, Air Pollution Control Office Publi-
cation No. APTD-0576.
Shell Development Co. Analytical Depart-
ment, Determination of Sulfur Dioxide and
Sulfur Trioxide in Stack Gases, Emeryville
Method Series, 4516/59a.
METHOD 9 VISUAL DETERMINATION OF THE
OPACITY OF EMISSIONS FROM STATIONARY
SOURCES
1. Principle and applicability.
11 Principle. The relative opacity of an
emission from a stationary source is de-
termined visually by a qualified observer.
1.2 Applicability. This method is appli-
cable for the determination of the relative
opacity of visible emissions from stationary
sources only when specified by test proce-
dures for determining compliance with the
New Source Performance Standards.
2. Procedure.
2.1 The qualified observer stands at ap-
proximately two stack heights, but not more
than a quarter of a mile from the base of
the stack with, the sun to his back. From a
vantage point perpendicular to the plume,
the observer studies the point of greatest
opacity in the plume. The data required in
Figure 9-1 is recorded every 15 to 30 seconds
to the nearest 5 % opacity. A minimum of 25
readings is taken.
3. Qualifications.
3.1 To certify as an observer, a candidate
must complete a smokereading course con-
ducted by EPA, or equivalent; in order to
certify the candidate must assign opacity
readings in 5% increments to 25 different
black plumes and 25 different white plumes,
with an error not to exceed 15 percent on
any one reading and an average error not to
exceed 7.5 percent in each category. The
smoke generator used to qualify the ob-
servers must be equipped with a calibrated
smoke indicator or light transmission meter
located in the source stack if the smoke
generator is to determine the actual opacity
of the emissions. All qualified observers must
pass this test every 6 months in order to
remain certified.
4. Calculations.
4.1 Determine the average opacity.
5, References.
Air Pollution Control District Rules and
Regulations, Los Angeles County Air Pollu-
tion Control District, Chapter 2, Schedule 6,
Regulation 4, Prohibition, Rule 50,17 p.
Kudluk, Rudolf, Ringelmann Smoke Chart,
TJ.S. Department of Interior, Bureau of Mines,
Information Circular No. 8333, May 1967.
fil-rl, tnml.nn
Optc'W ;
Sum of iws. racarcfod
Total no. reading*
Figure 9-1. Field data,
[FR Doc.71 18624 Filed 12-22-71:8:45 am]
U. S. GOVERNMENT PRINTING OFFICE: 1972 746461/4IO2
FEDERAL REGISTER, VOL. 36, NO. 247—THURSDAY, DECEMBER 23, 1971
1131
-------
------- |