-------
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
-------
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
-------
-------