RESEARCH REPORT
FUNDAMENTAL STUDY OF SULFUR FIXATION
BY LIME AND MAGNESIA
to
ROBERT A. TAFT SANITARY ENGINEERING
CENTER
PUBLIC HEALTH SERVICE, BSS-EH
DEPARTMENT OF HEALTH, EDUCATION,
AND WELFARE
June 30, 1966
BATTELLE MEMORIAL INSTITUTE
COLUMBUS LABORATORIES
-------
BATTELLE MEMORIAL INSTITUTE
COLUMBUS LABORATORIES* 505 KING AVENUE • COLUMBUS, OHIO 43201
.V* *«».
DIDICATCO TO T^f OOVANC^HIMT OF SCICNCI
FIELDS OF RESEARCH
Aeronautics — Astronautics
Agricultural Chemistry
Agricultural Economics
Alloy Development
Applied Mathematics
Area Economics
Biochemistry
Biophysics — Bionics
Catalysis — Surface Chemistry
Ceramics
Chemical Engineering
Chemical Processes
Communications Science
Computer Technology
Corrosion Technology
Earth —Atmospheric Sciences
Electrochemistry
Electronics
Energy Conversion
Engineering — Structural Materials
Environmental Systems
Extractive Metallurgy
Extreme-Temperature Technology
Ferrous Metallurgy
Food Technology
Foundry Practice
Fuels — Combustion
Glass Technology
Graphic Arts Technology
Immunology — Cancer Studies
Industrial Economics
Industrial Physics
Information Research
Inorganic Chemistry
Instrumentation
Light Alloys— Rare Metals
Lubricant Technology
Materials Separation —Concentration
Mechanical Engineering
Metal Fabrication Engineering
Metal Finishing
Metallurgical Processes
Microbiology
Microscopy — Mineralogy
Nondestructive Evaluation Technology
Nonferrous Metallurgy
Nucleonics
Ocean Engineering
Organic Chemistry
Organic Coatings
Packaging Research
Particle Dynamics
Petrochemicals
Petroleum Engineering
Pharmaceutical Chemistry
Physical Chemistry
Production Engineering
Psychological Sciences
Pulp— Paper Technology
Radioisotopes— Radiation
Reactor Technology
Refractories
Reliability Engineering
Rubber—Plastics
Semiconductors — Solid-State Devices
Sound — Vibration
Systems Engineering
Textiles— Fibers
Theoretical —Applied Mechanics
Thermodynamics
Transportation
Welding— Metals-Joining Technology
Wood — Forest Products
-------
FINAL REPORT
on
FUNDAMENTAL STUDY OF SULFUR FIXATION
BY LIME AND MAGNESIA
to
ROBERT A. TAFT SANITARY ENGINEERING
CENTER
PUBLIC HEALTH SERVICE, BSS-EH
DEPARTMENT OF HEALTH, EDUCATION,
AND WELFARE
June 30, 1966
Contract No. PH 86-66^108
BATTELLE MEMORIAL INSTITUTE
Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
Battelle is not engaged in research for advertising, sales promotion, or publicity
purposes, and this report may not be reproduced in full or in part for such purposes.
-------
TABLE_OF CONTENTS
Page
OBJECTIVES 1
INTRODUCTION 1
CHEMICAL THERMODYNAMIC CONSIDERATIONS 4
Desulfurizing Action of CaO(c) and MgO(c) in Flue Gas 6
Calcination of Limestone . 6
Stability of Dolomite 7
Sulfite Formation 9
SO2-O2-SO3 Equilibria 9
Sulfide Formation 10
Silicate Reactions 10
Hydration of CaO(c) and MgO{c) 11
Sodium Carbonate and Alumina . 11
Ferrous Sulfate Formation 11
KINETICS OF THE SIMPLIFIED BOILER FURNACE MODEL 12
Qualitative Chemical Kinetics 12
Simplified Furnace Model 15
Heat Balance on a Particle 16
Mass Transfer to a Particle 19
Rate Estimate Based on the Model for Heat- and Mass-Transfer Control . 20
RECOMMENDATION FOR THE USE OF LIMESTONE AND DOLOMITE IN
BOILER FURNACES 23
Basic Conditions 23
Composition of Flue Gas 24
Composition of Typical Coal Ashes 24
Composition of Oil Ash 25
Occurrence of Sulfur 2,6
Temperature and Time 26
Reactions With Ash 28
Calcination 32
Comparison of Limestone and Dolomite 32
Catalytic Activity of Impurities in Limestone 32
Preferred Location in Furnaces for Addition of Limestone or Dolomite . 33
Quantity of Additive Required 34
CONCLUSIONS 34
APPENDIX
THERMOCHEMICAL VALUES FOR BASIC REACTIONS IN FLUE-GAS SULFUR
FIXATION A-l
8ATTEULE MEMORIAL INSTITUTE
-------
FUNDAMENTAL STUDY OF SULFUR FIXATION
BY LIME AND MAGNESIA
June 30, 1966
OBJECTIVES
This study was undertaken by Battelle to identify the basic factors involved in the
capsure of SOz by limestone or dolomite added with the fuel or blown separately into the
hot flue gas of central-station boiler furnaces. It has been known for many years that
lime and magnesia will react with SO2 to form calcium and magnesium sulfates in boiler
furnaces, but the basic limiting conditions under which these reactions can occur have
never been adequately defined. It was the objective of this brief study to provide that
information.
The report is in three parts, following a brief introduction: thermodynamic con-
siderations, kinetics, and recommendations for use in the field. An extensive appendix
tabulates the results of the many thermochemical calculations made during this study.
INTRODUCTION
Presently the major consumer of solid fuels in the United States, public utilities in
1965 burned 243 million tons of bituminous coal to generate the largest part of the elec-
tricity that powers our economy. Residual fuel oil provided about one-ninth as much
equivalent energy for electrical generation. Although the stacks of large public-utility
power plants usually are high to distribute the products of combustion over a wide area -
the recently announced Conemaugh station will have 1000-foot stacks - sulfur dioxide
present in flue gas can still contribute markedly to wide-scale air pollution. For
instance, if the sulfur content of the utilities' coal supply and their residual fuel in 1965
averaged 3 percent, the amount of SO£ emitted by power plants over the year would have
come to more than 15 million tons, assuming 95 percent conversion of the sulfur in the
fuel to SO2 appearing in the stack gas. Although more troublesome from some stand-
points, 503 emission is much less, usually amounting to only 1 percent of the SO2-
Burning oil with low excess air reduces even this small amount of SO3 essentially to
zero, but low excess air does not decrease the problem with SO;?.
Loading the air we breathe with this tremendous tonnage of SO;? is causing increasing
concern among those who watch over public health, particularly because the utilities
double their output of electricity every 10 years. Sulfur dioxide from power plants, too,
would then double every 10 years. Coupled with a gradually increasing sulfur content in
fuels, great concern is being expressed for the quality of our atmosphere in the years
ahead. As a result, laws restricting the sulfur content of fuels have begun to appear in
many localities. Coal producers and public utilities have teamed up to seek feasible and
economically suitable procedures for removing sulfur from coal, and oil refiners have
sought practical means of desulfurizing residual fuel. So far, no economically attractive
schemes have been fcund. Generally, costs of removing sulfur from fuels are excessively
BATTELLE MEMORIAL INSTITUTE
-------
high, and no assured scheme for providing low-sulfur fuels has been turned up. Limiting
the sulfur in fuels provides one way of controlling SC>2 emission, and this has been the
objective of many ordinances. It may be a relatively costly one, however, if limiting the
sulfur content calls for expensive processing of fuels or for radical changes in sources
of fuel supply.
An alternative method of decreasing the SC>2 output from stacks is to remove a
large part of the sulfur oxides from the flue gases. Considerable attention has been
given this problem for many years, ranging from wet scrubbing of the flue gas as at
Battersea and Bankside stations in London beginning three decades ago to the catalytic
conversion of SO2 to SO3 and its removal from the flue gas as an acid mist at normal
stack temperatures. Adsorption on solid particles, reaction with ammonia, chemical
scrubbing, and bag filters combined with an additive are examples of other systems pro-
posed to remove sulfur oxides from flue gases.
The procedure considered here, one that has been brought up from time to time
over the past 30 years, is the addition of limestone or dolomite to a furnace such that
CaSC>4 or MgSC>4 and other solids would be the end product rather than SO2- It has long
been known for example, that coals containing large amounts of calcite and pyrite give
trouble in the analytical determination of ash in coal. In such cases, the ash appears
high because appreciable quantities of calcium sulfate are formed when the coal is heated
in a muffle. Hence ASTM specifies that such coals are to be ashed by a procedure where
the pyritic sulfur is oxidized and the SOz is expelled before the calcite is decomposed. U)
Lignite and brown coal, usually containing large amounts of calcite, generally trap
sulfur when burned, so that the ash from these fuels frequently contains as much as
30 percent 803. (2) It is this ability of lime to react with SC>2 under some conditions that
makes it so interesting as a means of decreasing the SC>2 in flue gas.
The chemical reactions involved here are relatively simple. Yet no clear insight
has been available as yet on which reactions predominate or how the reactions are in-
fluenced by conditions within the furnace. In particular, there has been little or no
information available on the relative effectiveness of limestone and dolomite, on the in-
fluence of temperature, on the presence of ash, and on the time required for trapping
SO2 by these solids under furnace conditions. Hence this study was made to evaluate the
potential usefulness of this means of decreasing SO2 emission from power plants.
Three tasks were involved:
* Thermodynamic calculations to show the course of the probable
chemical reactions
* Kinetic factors as far as they can be deduced without experimentation
• Recommendations for the use of limestone and dolomite most
effectively in large central-station boiler furnaces.
This report contains th* results of these calculations and evaluations to aid those
seeking practical and economical means of minimizing SC>2 emissions. It will also serve
(1) Sampling and Analysis of Coal and Coke, ASTM, D-271, Part 19, p 23, 1966.
(2) .Sondreal, E. A,, Kube. W. R., and Eider, J. L., "Characteristics and Variability of Lignite Ash From the Northern Great
Plains Province", 1965 Lignite Symposium. Bismarck, N. D. , p 12.
BATTELLE MEMORIAL INSTITUTE
-------
as a broad guide in evaluating proposed large-scale tests in boiler furnaces where con-
ditions cannot always be accurately defined. The calculations here are based on single
systems. A much more complicated procedure would be necessary for interrelated
systems where one process would be interacting with another. For the present state of
the art, the simplified methods used here should be entirely satisfactory and can serve
later as the starting point if a more elaborate treatment should ever be desirable.
It should be emphasized at the outset of this report that the conclusions reached
here are based on sound but limited theoretical considerations, and that theory at best
can point only to expected results. Experimentation, eventually on a large scale, will
be necessary to establish the conditions where this method of decreasing SC>2 emission
is most effective.
BATTELLE MEMORIAL INSTITUTE
-------
4
CHEMICAL THERMODYNAMIC CONSIDERATIONS
by
J. J. Ward and D. A. Pettit
A chemical thermodynamic study was made, as the first step in this study, of
reactions of CaCO3, MgCO3, and dolomite with SO2(g) and O2(g) in boiler atmospheres
at temperatures up to 3000 F. The purpose here was to collect from the literature
basic data required to evaluate schemes for fixing sulfur with limestone, dolomite,
CaCO3, CaO, MgCO3, and MgO. Interfering reactions that may occur from silica or
alumina in the coal ash were included. All these basic thermochemical data have now
been collected and interpreted as they apply to the sulfur-fixing problem.- "Thermo-
chemical" in this report is used in the same sense as "Chemical Thermodynamics".
As a first step in this work, standard free-energy changes of reactions of interest
were tabulated from literature sources at 100 F intervals from 300 to 2000 F, at
2500 F, and at 3000 F. From the value of the standard free energy change of a reac-
tion, A FR , the logarithm of the equilibrium constant of the reaction was calculated.
Thermochemical data for the pertinent reactions are collected in the Appendix in
Tables A-l through A-19 with references to the literature sources. A listing of the
reactions given in the tables is shown on page A-l,
These calculations are based on bulk properties. Some surface effects might be
expected because the solids would be pulverized, but the difference attributable to this
increase in surface area would be negligibly small.
A negative value of the standard free energy change of a reaction, AFR°, shows
the tendency or driving force for the reaction to occur. The AFj^° value, however,
gives no information on the rate of reaction nor the time necessary for its completion.
A review of that part of the problem is given in the next section of this report. The
equilibrium constant, also a measure of extent of reaction, is useful in calculating the
conversion under non-standard-state conditions, such as those existing in boiler at-
mospheres. A somewhat detailed example follows on the application of these data to
SO2
-------
5
where
AFj^° = calories for the reaction
T = absolute temperature, °K.
The equilibrium constant, K, may be written for Reaction (1) as,
aCaS04
K = - • - - - , (3)
where aCaSC>4 and aQaQ are Raoultian activities of the solids calcium sulfate and
lime. These solids are usually assumed to be at unit activity. As will be discussed
later, this assumption is not valid in the decomposition of dolomite nor in the inter-
action of CaO with ash components.
Here pg_ (g) and p_ (g) are the partial pressures of SC>2(g) and C>2(g) in
atmospheres.
Equation (3) can be expressed as
aCaSO\
K =
. f
_ v Y ' 13 '
aCaO / S02(g) ' X02 P
where X__ . . and X^ . . are the mole fractions in the gas phase of sulfur dioxide and
SO2(g) O2(g)
oxygen gas. Equation (4) can be written at 1 -atmosphere pressure as:
XS02(g) =
4
aCaO
1
' KX1/2
°2
The mole fraction, X . . can be expressed as:
percentage by volume SO2 =10 X
and
Similar equations can be derived for the reactions of MgO, SO2, and O2 to form
BATTELUE MEMORIAL INSTITUTE
-------
D_esulfur!zing Action of CaO(c)
and MgO(c) in Flue Gas
By applying the data of Tables A-l and A-2 according to the above method, the
equilibrium concentration of SO2(g) in flue gas can be calculated as a function of tem-
perature at a level of 2.7 percent O2 in the flue gas for CaO and MgO, assuming unit
activity of solids.
TABLE 1. EQUILIBRIUM CONCENTRATION OF SO2(g) IN PARTS PER MILLION
(PPM) OF FLUE GAS CONTAINING 2. 7 PERCENT O2, AND WITH
UNIT ACTIVITY OF SOLIDS
5O2, ppm
Temperature, CaO Fixation MgO Fixation
F of Sulfur of Sulfur
1000
1200
1400
1600
1800
2000
2500
3000
1.
1.
1.
2.
6.
(:
7 x 10'11
1 x 10-7
0 x 10-4
3 x ID'2
1.95
74. 8
3 x 10+5(a)
>10+6)(b)
1. 3 x 10-3
1. 0
180
1.1 x 10+4
3. 2 x 10+5(
(>10+6)(b)
a)
(a) Insufficient sulfur in coal to approach this equilibrium value of SC>2(g) in flue gas.
(b) Implies greater than 100 percent SCVj required for equilibrium.
An examination of the above table indicates that, thermochemically, sulfur removal by
CaO is limited to temperatures below about 2250 F and by MgO to less than about
1550 F under boiler atmosphere conditions. The assumption that the activity ratios,
are unity is a good one if limestone or magnesite are used as reactants, and if exces-
sive ash constituents are not included. Surface effects would appear in the term a
which would be greater than unity, resulting in a decrease in the equilibrium level
of SO2 in contact with the solid. As noted earlier, however, the effect is so small as
to be disregarded for solids ground to a practical particle size.
Calcination of Limestone
Some consideration has been given to the thermochemical calcination of lime-
stone, and to the desulfurizing action of limestone directly without formation of CaO.
Thermochemical data are reported for the calcination of CaCOs. From these data, the
dissociation pressure of CaCO3 can be calculated at 1-atmosphere pressure as shown
in Table 2.
eATTELt. E MEMORIAL INSTITUTE
-------
7
TABLE 2. DISSOCIATION OF LIMESTONE IN FURNACE FLUE GAS
[CaC03(c)^iCaO(c) + CO2(g)]
Equilibrium Concentration
Temperature, of CO2 *n Flue Gas,
F percentage by volume
1000 0,05
1200 3.5
1400 12. 1
Because flue gas contains typically about 14, 5 percent CO2 by volume, these data show
that CaO will tend to recombine with CO2 in boiler atmospheres at temperatures below
about 1415 F. Figure 1 shows this situation graphically. In other words, below 1415 F,
CaO is not stable because of the combustion atmosphere, and CaCO3 would be formed at
the expense of any CaO.
For this reason, the desulfurizing action of CaCO3 below 1400 F becomes impor-
tant. Table 3 shows the extent of this factor as calculated from basic data taken from
Table A-10.
TABLE 3. DESULFURIZING ACTION OF UNCALCINED LIMESTONE
[CaC03(c) + S02(g) + \ 02(g)^CaS04(c) + CO2(g)]
Equilibrium Concentration of
SOZjPP01* With 2. 7 Percent O2
Temperature, in Flue Gas and Unit Activity
F of Solids
1000 4. 5 x 10-9
1200 1. 4 x 10-6
1400 1. 2 x 10-4
These calculations predict that limestone can desulfurize flue gas up to its calcina-
tion temperature. However, the data indicate thermodynamic feasibility only; no inter-
pretation can be made concerning reaction kinetics nor the time required for the equilib-
rium values to be reached.
Stability of Dolomite
The thermochemical stability of dolomite is shown in Table A-13. These data indi-
cate that dolomite dissociates at 865 F to exert a partial pressure of 1 atmosphere ac-
cording to the reaction: •
CaC03-MgCO3(c)^iCaC03(c) + MgO(c) + CO2(g).
BATTELLE MEMORIAL INSTITUTE
-------
m
r
r
m
x
m
z
-------
In a flue gas containing 14. 5 percent CO2 , dissociation would tend to occur approxi-
mately at 730 F. Phase equilibrium of the CaO-MgO system indicates that a double
oxide forms. If this double, oxide were formed in the decomposition of dolomite, then
aCaO would be lowered, and the desulfurizing action of lime would be decreased pro-
portionately, as can be seen from Equation (3). Further examination of Equation (3)
suggests that the desulfurizing action of CaO can be increased by lowering the activity
of CaSO4, as can be done by dissolving CaSCXj in ash without dissolving CaO(c), This
is most unlikely. The extent of desulfurizing flue gas may be affected through this means
by a factor of at least four, depending on the phase equilibrium of the reaction
CaO(c) + S02(g) + | 02(g)^CaS04(c).
Sulfite Formation
Calcium sulfite and magnesium sulfite have a thermochemical tendency to oxidize
to the corresponding sulfate with excess oxygen as the data in Tables A-4 and A-5
show.
Calcium oxide will react directly with SO-,(g) to form CaS3 has no advantages.
Equilibria
A driving force exists for SO2(g) to react with O2(g) to form SC>3(g) up to 1400 F
under standard-state conditions. This conclusion can be reached by examining the data
of Table A-ll. Table 4 shows calculated values of the conversion of SO2(g) to SC"3(g) at
equilibrium with 2.7 percent oxygen with a preponderance of SC>3(g).
TABLE 4. EQUILIBRIUM CONVERSION OF SO2(g) TO SO3(g) IN FLUE GAS
CONTAINING 2. 7 PERCENT OXYGEN
Approximate Equilibrium
PSQ- Conversion of SO2(g) to
Temperature, p SO^(g), With 2. 7 Percent
F SO3 Excess O2(g)
748 0.01 99
876 0.10 90
1050 0.33 75
BATTELLE MEMORIAL INSTITUTE
-------
10
Again, it must be stressed that these are equilibrium levels. In boiler practice,
nearly all the sulfur occurs as SO^fg) in the flue gas. This difference between equilib-
rium and real concentrations indicates a frozen equilibrium from a higher temperature
where SO2(g) formation is favored. Catalysis is required to speed up the reactions
sufficiently to convert the SO2(g) to SO3
essentially yielding SO3, is the stronger acidic oxide.
TABLE 5. COMPARISON OF THE DESULFURIZING ACTION OF CaO AND
CALCIUM SILICATES WITH 2. 7 PERCENT O2 IN FLUE GAS
(Unit activity assumed for condensed phases. )
Equilibrium Concentration
of SO2(g), ppm
Reaction 1400 F 1800 F
CaO(c) + S02(g) +| 02(g)^CaS04(c)
Ca2Si04(c) + S02(g) + | 02(g)^CaSi03(c) + CaSO4(c)
CaSi03(c) + S02(g) + 7 02(g)^CaS04(c) + SiO2(c)
0. 0001
0. 15
3.0
1.95
135.0
9370
BATTELLE MEMORIAL INSTITUTE
-------
11
Thermochemically, CaSiOj could lower SO^ to 3 ppm at equilibrium, but the kinetics of
the reaction undoubtedly would be slow at 1400 F. It is conceivable that blast-furnace
slag containing CaS in a fluidized bed could desulfurize flue gas at temperatures below
1500 F. The rates of the reaction might preclude serious consideration of such a
desulfurizing scheme.
Hy drat ionof CaO(c) and MgO(c)
Basic thermochemical data for the hydration of CaO and MgO are given in
Table A-16. At a water-vapor concentration of 7. 1 percent by volume, as in flue gas,
CaO would hydrate below 700 F, and MgO below 400 F. However, at a flue-gas concen-
tration of 14. 5 percent CO;?, CaO would already have tended to react with COz to give
CaCO3 at 1400 F, as discussed earlier. The conclusion can be reached here that hydra-
tion or slaking of lime by steam in the flue gas will not occur.
Sodium Carbonate and Alumina
Sodium sulfate (Table A-17) is stable thermochemically above 3000 F although the
table covers only the range up to 1300 F. It can be concluded that Na£CO3, reacting to
form the sulfate, would desulfurize flue gas throughout the range of boiler temperatures.
Other problems including slagging and corrosion caused by this salt eliminate further
consideration of it.
Alumina could reduce SO2(g) to 1 ppm at temperatures below 800 F. This conclu-
sion can be reached from a study of the data in Table A-17.
Ferrous Sulfate Formation
There is no thermochemical benefit of desulfurization by the addition of ferric
oxide to the flue gas. Likewise, the addition of ferro-ferric oxide will not form any
ferrous sulfate, except at low temperatures. These conclusions are drawn from
Table A-18.
The data in Table A-19 indicate that ferrous oxide will combine with SO2 to form
ferrous sulfate at temperatures up to approximately 1700 F. Because the furnace atmo-
sphere contains 2. 7 percent oxygen, ferrous oxide will be unstable and will tend to go to
the more stable ferric oxide. Therefore, ferrous oxide does not appear to be a good
agent for capturing SO2-
The data used in compiling the tables in the Appendix were taken from the most
recent sources, listed in a bibliography at the end of the Appendix. The higher tem-
perature values reported in most of the tables are extrapolated values. Because of the
need for arithmetical consistency in this type of study, the significance of the values
reported is beyond that which actually exists.
BATTELLE MEMORIAL INSTITUTE
-------
12
KINETICS OF THE SIMPLIFIED BOILER FURNACE MODEL
by
J. F. Walling, R. H. Cherry, Jr. , and A. Levy
The objective of this part of the study was to attempt to identify products of the
reaction of SCs with limestone and dolomite during the residence time available in
boiler furnaces, and to estimate the rates of reaction of these substances under an as-
sumed set of temperatures and time. The course of this task was determined largely
by three facts:
(1) No pertinent chemical kinetic data on these systems which could be
interpreted quantitatively were found in the literature.
(2) It is presently impossible to calculate heterogeneous reaction rates
from first principles.
(3) Also, it is presently impossible to obtain reliable estimates about
heterogeneous reaction rates by using data from "similar" systems.
Therefore, determination of chemical reaction rates between SC>2 and limestone or
dolomite will require a laboratory effort which is not within the scope of this study.
Since it is not possible to estimate rates of reactions to calculate the probable
outcome on the basis of an assumed model, shown later as Figure 2, the task was
turned around. Instead, an attempt was made to provide an estimate of what the lower
limit of the rate, when measured, should be in order to produce a result no poorer
than heat or mass transfer control of the process if the model is adequate. Therefore,
this part of the report is divided into three sections:
(1) Qualitative comments about chemical kinetics
(2) A simplified model of the furnace for purposes of kinetic calculations
(3) Estimation of reaction rate based on model for heat- and mass-transfer
control.
Qualitative Chemical Kinetics
No quantitatively interpretable data on chemical kinetics for these systems were
found. Information sought would have included, for example, specification of the
method of chemical and physical preparation of the solid material (or other evidence to
substantiate uniqueness and reproducibility), its specific surface area and density, the
amount of reaction products as a function of time or flow rate, gas composition, tem-
perature, and pressure. Some evidence of chemical stability or reproducibility of the
surface of the solid with time during the reaction would also be needed.
BATTELLE MEMORIAL INSTITUTE
-------
13
These requirements for quantitatively interpretable data may appear to be exces-
sively stringent. However, they have been dictated by difficult problems, well known to
chemical kineticists, which are understood only in a qualitative way. (3) It may suffice
to indicate here that, in heterogeneous reactions, reproducibility of the "surface phase"
is critical. Any change which might alter its composition in an important average way —
for example, by impurity atoms present or absent in the crystal surface, by defect
structure changes, or by a strongly adsorbed species from the gas phase which may be
present or absent - may profoundly alter the course of a desired heterogeneous reac-
tion, This list is not exhaustive, and in a given instance, some changes in the surface
phase may produce no observable effects on the desired reaction. However, to get
reliable results, each factor must be held constant or its effect investigated in each
separate instance. Some documentation of the importance of some of the effects just
mentioned is available from literature pertinent to this problem.
Qualitative information found for the interaction of SO^ with solids comes from
three papers. \^> $> ") Perhaps the most certain conclusion to come from all three is
the common report that S~, SO3=, and SO^3 were all observed after reaction with SO£
when Ca"*"^ was present. The situations in all three papers are different. In all cases,
temperatures were somewhere between 750 F and 1800 F, but gas composition ranged
from pure SO£ to various air-SC>2 mixtures, and the solid varied from pure CaO or
CaCO^ to a rock of uncertain composition. Reaction times were much longer (minutes
to tens of minutes) than those of interest in boiler furnaces (seconds). However, it
seems likely that because all three anions were observed in all of the studies with their
sizable diversity of conditions, it is reasonable to expect all three anions as products
in boiler furnaces as well. This is the only reasonably firm conclusion which can be
drawn from all these qualitative data.
However, some other inferences from these studies which are considerably more
speculative are not unreasonable and may be instructive. Moreover, some of the points
serve to illustrate the general statements made earlier about the kinds of factors which
need control. Therefore some of this more speculative material is of interest here.
In his study of the reaction of pure SC>2 with CaO and CaCC^, ^' Pechkovskii ob-
serves that;
(I) Reaction with CaO occurs appreciably at 750 F but reaction with CaCO3
begins at 930 F.
(2) Reaction with CaO is nearly complete in 15 minutes, and above 1100 F
the fraction of CaO reacted is much less than the fraction of
He concludes from these observations that SO2 reacts with O~" rather than CO3~, and
that the crystal perturbations caused by CO£ evolution in CaCOj foster more rapid and
complete reaction than is possible with the more stable CaO lattice.
(3) Benson, S. W. , The Foundations (^Chemical Kinetics. McGraw Hill, New York(1960). Chapter XVII.
(4) Lugowska, M., "Possibility of Obtaining SC>2 From Indigenous Sulfur Ore", Przemysl Chem. , 40, 95(1960).
(5) Ketov. A. N.. and Pechkovskii. V. V.. "Interaction of 50% With CaO and CaCOa", Zh. Prikl. Khim. , 31 (12), 1783
(1958).
(6) Pechkovskii, V. V,, "Reactions of SO2 With Metal Oxides in an Oxidizing Atmosphere", Zh. Prikl. Khim., 30(11), 1580
(1957).
BATTELLE MEMORIAL INSTITUTE
-------
14
Pechkovskii warns, however, that if 02 is present in the gas and impurities of
transition- metal oxides are present in the oxide lattice, the rates may be altered pro-
foundly. (") For example, he carried out two experiments identically except that in one,
pure MgO was used and in the other, MgO with 9 percent FezO^ was used. The per-
centage conversion of the pure MgO was 5 percent and the mixed oxide conversion was
94 percent. Pechkovskii suggests as a result of such observations that if a transition-
metal oxide impurity which catalyses the conversion of SO-, to SO^ is present, more
than one path is available for an SO£- solid reaction. The particular data on which this
generalization was based involved MgO, not CaO or dolomite. However, if the inter-
pretation is correct, it is difficult to see why it would not apply to dolomite as well, at
least in a qualitative way.
It must be emphasized once again that the foregoing remarks are qualitative and
speculative with respect to dolomite. Another way to support this contention is to ex-
amine factors reported to affect the rate of decomposition of limestone and dolomite.
Grain size, particle surface area (roughness), heating conditions, and partial pressure
of CO£ and r^O all affect the decomposition of CaCO3_ Small grains decompose more
rapidly than large ones, and the larger the surface area, the greater the decomposition
rate. C?> Here CO£ reduces the rate of decomposition(°' and H^O accelerates it'''.
Similarities are reported between the decomposition of limestone and of dolomite.
For example, CO2 is reported to retard decomposition and H2O to accelerate itl*™),
just as with CaCO,. However, there is at least one report of differences in rates of
decomposition of one dolomite and its constituent carbonates, and of the sizable effect
of small quantities of ionic impurities on the decomposition behavior of dolomite. Im-
purities are reported to facilitate the onset of decomposition at lower temperatures and
at a generally greater rate(^). The behavior in gas- solid systems is obviously com-
plex. The foregoing examples provide some evidence for the contention that it is im-
possible to estimate heterogeneous reaction rates by deduction or by "extrapolation" of
known results.
In summary, it appears that:
(1) SO2 is likely to appear in the solid phase in some distribution between
S~ , SO^ = , and SO^~ as the result of reactions in the furnace between
SO2 and limestone or dolomite.
(2) SO£ appears to react with CO^~ after the carbonate has decomposed
into O= and CO2.
(3) If Q£ is present in the gas phase and transition- metal oxides are pres
ent in the limestone or dolomite, SO2 may be oxidized to 803. Thus
two paths may be available for reacting SO2 with the solid phase.
(8) Hyatt, E. P.. Cutler, I. B , and Wadsworth. M. E. , "CaCO3 Decomposition in CC>2 Atmosphere", J. Am. Ceram. Soc. ,
41, 70 (1958).
(9) Hyatt, E P., "Thermogravimetric Study of CaCOg Decomposition", Dissert. Abstracts, _17, 828(1957)
(10) Bischoff, F., "On the Kinetics of the Thermal Dissociation of Dolomite and Limestone in Various Atmospheres", Z. Anorg.
Chem., 262. 288 (1950).
(11) Esin, O. A., Gel'd, P. V., and Pope 1, S. I.. "Redistribution of Ions in the Thermal Dissociation of Double Salts", Zh.
Pfikl. Khim.,_22, 354(1949).
BATTEULE MEMORIAL INSTITUTE
-------
15
(4) All of these conclusions must be speculative with respect to the SO^-
solid reaction since they are based on different chemical systems
than the one of interest, on reaction times much longer than those
available in boiler furnaces, and on systems which were not adequately
characterized for quantitative analysis.
Simplified Furnace Model
In analyzing this system, it was necessary to model the boiler furnace after the
fashion of a simple catalytic chemical reactor because heat- and mass-transfer phe-
nomena are important. This is easily seen when it is recognized that from a practical
viewpoint it would probably be desirable to inject limestone or dolomite into the system
at a moderate temperature. However, the reaction with S(>> probably does not occur
appreciably below about 1000 F. Also, the contact time is short. Thus heat-up time is
important. The fact that limestone and dolomite particles would not be stationary but
would move with the gas stream suggests that mass transfer must be considered as
well.
Two extreme cases can be envisioned. In one, heat and mass-transfer phenom-
ena will control the rate of the process; in the other, the chemical reaction rate con-
trols. The purpose of the model to be described presently is to estimate the effects of
heat and mass transfer on the rate of the process. Since it has not been possible to
estimate the chemical reaction rate, it could be assumed that heat and/or mass trans-
fer do control the rate of the process. To the extent that the model reflects the prac-
tical situation, it can be assumed that this is the best which can be expected.
To insure that the best possible performance is obtained, it is necessary for the
chemical reaction rate to be greater than that estimated on the basis of heat-mass
transfer control. Therefore, within the limits of this model, an attempt was made to
estimate the smallest value for the overall chemical reaction rate that would achieve
the calculated performance. The procedure used was to model the gross physical situ-
ation, derive a relationship for heat transfer to an individual particle, obtain the mass
transfer relationship to a particle by analogy, and solve the resulting equations.
The model assumes that the hot flue gases are flowing in a unidirectional rectan-
gular duct that has the same cross section as the actual boiler, but which is empty and
does not contain superheater or reheater tubes or other flow obstructions. The lime-
stone or dolomite is assumed to be crushed to a uniform particle size small enough to
ensure a Stokes' law flow field. That is, spherical nonporous particles are assumed
with diameters less than approximately 200 microns. The assumption further specifies
essentially the relative velocity between the gas and the solid particles. This relative
velocity is insensitive to actual gas-flow velocity, turbulence, flow obstructions in the
actual furnace, and duct geometryll^)^ it is assumed that the solid particles are es-
sentially at ambient temperature when fed to the flue gas.
(12) Torobin, L, B. , and Gauvin, W. H., "The Effects of Fluid Turbulence on the Particle Drag Coefficient", Can. J. Chem.
Engr., 38. 189 (1960).
BATTELLE MEMORIAL INSTITUTE
-------
16
Heat Balance on a Particle
Since the dolomite particles are injected at ambient temperature, there is initially
a large difference between the local gas temperature and the mean particle temperature.
Work on radiative heat transfer to small spherical solid particles suspended in a gas
led Sleicher and ChurchillU3) to the result that the temperature gradient between the
surface and the center of the particle is small at the temperatures of interest here. It
is assumed then that no temperature gradient exists within the particle. This assump-
tion has been used successfully by Themelis and Gauvim^).
The differential heat balance on a single particle of diameter D is of the form:
Accumulation = input — output + generation . (6)
Individual terms in the above expression are:
Accumulation = (WC)p (dt/d0)p = (7tpCD3/6)p (dt/d9)p , (7)
where . D3
W = particle weight, Ib = — 7T — ^ p = (7T/6) Dp
3 c> r p t*
C = solid specific heat, Btu/(lb)(F)
t = temperature, F
8 - time, seconds
p = solids density, Ib/cu ft
and subscript p denotes the particles.
Heat input occurs primarily to the particle by convection (qc).
qc = hcAP W - M Dp W ' (8)
where h_ is the convective heat- transfer coefficient in Btu/(hr)(ft2)(F), A is the area
C ry • ' C
of the particle in ft*; and subscript g denotes the gas.
Zenz and Othmer (i5) have shown that the limiting Nusselt number for heat trans-
fer to single spheres in a Stokes1 flow field is given by
Nu » hcDp/kg * 2 , (9)
where kff is the thermal conductivity of the gas.
&
The data of Johnstone et al. 0^) and Joukovski^1?) show a much lower limiting
Nusselt number for heat transfer to clouds of small particles. Othmer and Zenz
(13) Sleicher, C. A., and Churchill, S. W., "Radiant Heating of Dispersed Particles". IEC, 48, 1819(1956).
(14) Themelis, N. J. , and Gauvin, W. H., "Heat Transfer to Clouds of Particles", Can. J. Chem, Engr., £1, 1(1963).
(15) Zenz, F. A., and Othmer, D. F., Fluidization and Fluid-Particle Systems, Reinhold Publishing Corporation, New York
(1960).
(16) Johnstone, H. P., Pigford, R. L., and Chapin, J. H., "Heat Transfer to Clouds of Falling Panicles", Trans. Am. Inst.
Chem. Engr., 37, 95 (1941).
(17) Joukovski, D. N., "Convective Heat Transfer Between Gases and Suspended Particles", J. Tech. Phys. (USSR), 10, 999
(1940).
BATTELLE MEMORIAL INSTITUTE
-------
17
concluded that the difference between the single-sphere and particle-cloud data is real,
and is probably caused by interparticle hindrance effects. Consequently, the limiting
value for Nusselt number is taken to be
Nu (particle clouds) = hcD /k = 0. Z. (10)
After using Equation (10) in Equation (8), the convective heat transfer becomes '
qc = o.27rkgDp2, and CO in the flue gas has a negligible
effect on the radiation rate; that is, the gas is completely trans-
parent to all radiation.
(2) The effect on radiation transfer rates of the nonisotherrnal surround-
ings, shape factor, etc. , have been ignored so that the results will
be conservative.
(3) The particle sees a portion of the duct wall at some temperature
below the bulk-gas temperature at a given location in the duct, and
this difference between surface and gas temperatures is nearly con-
stant throughout the duct.
It is common practice to eliminate the fourth-power temperature difference in the radia-
tion expression by defining a radiation transfer coefficient:
qr= hrAp <*.-tp) = hr* Dp" (t.-tp) . (13)
Values of hr calculated on the basis of assumptions similar to the above are available
in McAdams book(18).
The generation term in Equation (6) represents the net energy released or ab-
sorbed by the overall chemical reaction as it occurs on the particle surface.
For the purposes of these calculations, the heat of the reaction will be taken as
zero. This is an erroneous but not an absurd assumption. If the reaction proceeds by
(18) McAdams, W. H., Heat Transmission. Third Edition, McGraw-Hill Book Company, New York (1954).
BATTEULE MEMORIAL INSTITUTE
-------
18
decomposition of the CO^~ and then reaction of SO^ with O~, the first step will be endo-
thermic and the second exothermic. Up to this point at least, the net heat effect may be
nearly zero. However, the reaction may not stop at this point, and whatever actions
occur are likely to be on the particle's surface. The usual tabulated bulk thermody-
namic properties will not be appropriate for estimating a heat of reaction on the surface
because of the difference in bulk and surface properties. Therefore, since the product
distribution between S~, SO-j~ and SO^" and applicable thermodynamic properties are
unknown, the mathematical simplicity afforded by assuming zero heat of reaction has
been used.
The energy balance of Equation (6), combined with Equations (7), (11), and (13)
becomes
* " ('> - » " «t.-> - (14)
It is desirable now to change the independent variable, time, d, to gas temperature, tg,
because temperature can be specified, at least approximately as a function of distance
or time along the duct,
(dtp/d0) = (dtp/dtg) (dtg/dS) . (15)
The value of (dtg/df?) can be estimated from the temperature -time profile assumed for
the model boiler furnace and shown later in Figure 2. It is constant for each of two seg-
ments along the duct.
In addition, it is assumed that the difference between the duct wall temperature,
tg, and the local bulk-gas temperature, tg, at any point in the duct is given by a con-
stant, At = tg-ts. Using these assumptions, Equation (14) becomes:
p v v - *"> 'v*- v •
which has the form
qj (dtp/dtg) = q2 (tg-tp) - q3
-------
19
Mass Transfer to a Particle
To estimate the naass-transfer rates from the gas to a particle, it is necessary to
use one of the analogies between heat and mass transfer. Because of the lack of data,
it is assumed that the Colburn j-factor for mass transfer is equal to that for heat trans-
fer. Thus
(19)
therefore,
In the above,
JH= (Nu/RePr) (Pr)2/3
j = (Gz'/RePr) (Sc)E/3
Gz1 = Nu (Pr/Sc)2/3
Nu = Nusself number, hcD /k
Re = Reynolds number, D V p /JU0
" r 5 o
Pr = Prandtl number, C^i /kg
Sc = Schmidt number, /^/PpD
O O o
= modified Graetz number =
(20)
MmPmCP
KG
P = mean pressure of nontransferring gas, atm
M = mean molecular weight of gas
jU = viscosity of the gas
6
KQ = overall mass-transfer coefficient, lb-moles/(hr)(ft2)
(atm driving force).
Both the Prandtl (Pr) and Schmidt (Sc) numbers for a given gas are nearly constant
over a considerable temperature range. Therefore, evaluation of the ratio (Pr/Sc) '
at the mean of the terminal gas temperature in the duct will provide a sufficiently good
approximation for this simplified model. The Nusselt number (Nu) has been assumed to
be approximately 0, 2 throughout the duct. Consequently, the modified Graetz number
(Gz1) is approximately constant along the duct:
Gz1 = 0. 2 (0. 70/0. 60)2/3 = 0. 115.
The mass-transfer coefficient at any point in the duct is given by
KG=(Gz'/TTDp){k/MmPmCp)g.
(21)
(22)
BATTELLE MEMORIAL INSTITUTE
-------
20
The overall rate of mass transfer of SC>2 from the gas to a single particle at any
point in the duct is
where
N(8ingle particle) = KQ (Pg-Pp), lb-moles/(hr)(ft2), (23)
P = local partial pressure of SC>2 in gas, atm
o
P si partial pressure of SOT on solid surface,
taken as zero.
The total transfer of SC>2 to all the particles near a given location in the duct is
approximately
NT = (KGAT) P lb-moles/hr2, (24)
O
where A-J- is total surface area of all particles in solid:
AT = 6 Wp/ppDp, sq ft/hr (25)
Wp = rate of limestone or dolomite
injection by weight into flue -gas
duct, Ib/hr.
From Equations (22), (24), and (25), the mass-transfer [Equation (26)] emerges.
Rate Estimate Based on the Model for Heat- and
Mass-Transfer Control
The proper method of calculation requires simultaneous solution of the heat
[ Equation (18)] and mass [ Equation (26)] transfer equations for an increment of dis-
tance or time down the furnace. The overall result can be obtained by summing partial
results from each segment.
A simpler procedure has been used here. Conservative assumptions have been
made about the effects of heat transfer, and attention is focused on the mass-transfer
equation. This may have introduced some inconsistencies in the model. However, it
is considered that more elaborate computations are not justified.
The decrease in the SC>2 partial pressure was estimated for the following operat-
ing conditions using dolomite as an example:
(1) A total mass flow of 5475 tons/hr of which 5400 tons/hr is flue gas.
(2) The mole fraction of SC>2 is 3 x 10"^, its weight fraction is 6. 5 x
10"3, and therefore the SC>2 mass flow is about 7000 Ib/hr.
BATTELLE MEMORIAL INSTITUTE
-------
21
(3) Total system pressure is 1 atmosphere. The average molecular
weight of the gas, 29. 5, and the partial pressure of inert substances
is assumed to be constant throughout the furnace.
(4) At steady-state operation, dolomite equal to 1 percent of the total
flue-gas-weight flow rate is charged, so that W = 54 tons/hr.
(5) Spherical particles of 74-micron diameter are assumed.
D = 2. 4 x 1CT4 ft.
(6) Dolomite has a specific gravity of approximately 3. 0, so that
p » 187 lb/ft3.
(7) The temperature-time profile in the duct is given by Figure 2 in the
next section of this report. Two linear segments are assumed to be
adequate for this simplified calculation.
Mean values of specific heat and thermal conductivity of the gas are
used for each segment of the duct.
Then from Equation (26)
e
A0 9
N d9 = moles SC>2 consumed/hr
= Z(0.74x!08)
P (5. 56xl(r4) .
=>
(27)
Integrating over two segments using mean values for k/C in each segment yields the
results shown in Table 6.
TABLE 6. SAMPLE FURNACE CALCULATION
Mean
Temperature
Segment F
1 1600
2 900
Total
, (k/Cp)g, NT'
lb/(hr)(ft) moles/hr2
0. 154 3. 4 x 104
0, 123 2. 3 x 104
Residence SO2 Consumed
Time, hr Moles/Hr
5. 56 x 10-4 19
Ditto 13
32
Lb/Hr
1200
830
2030
The SO2 partial pressure at the end of the first segment is (0.003) (7000- 1200)/7000 or
0. 0025 atm. The exit partial pressure of SO2 in the flue gas is (0. 003) (4970/7000) or
0. 002 atm.
It is concluded, then, that roughly 2/3 of the SO^ would remain unreacted at the
exhaust point on the basis of mass-- transfer control of the process. This leads to the
BATTELLE
MEMORIAL
INSTITUTE
-------
22
expectation that considerations of chemical kinetics are secondary if the process is to
be successful. Unless the poor result limited by mass transfer can be improved, con-
siderations of chemical kinetics are pointless. Therefore, no attempt to estimate a
lower limit to a chemical reaction rate is considered worthwhile at this point. Also, no
attempt will be made to compute a temperature profile from the heat balance.
Examination of Equation (26) seems to suggest three ways in which the result
might be improved:
(1) Increase the amount of limestone or dolomite added to the furnace
(2) Decrease particle size
(3) Increase contact time.
Equipment modifications needed to effect Item (3) are obvious and probably unac-
ceptable. But at this stage, the validity of Items (!) and (2) as effective alternatives
must be considered questionable. This is because of the possibility of overstepping the
limits of validity of the model. For example, to achieve Item (1), a limestone or dolo-
mite flow rate greater than 1 percent of the flue-gas flow rate would be required. Under
these conditions, the assumption of a Stokes flow field is questionable. It is not known
whether this would result in a net increase in mass transfer, since the effect of in-
creased surface area may be more than offset by hindrance effects.
If the limestone or dolomite were ground to a finer particle size, the assumption
about the limiting value for the Nusselt number would become even more critical. The
value of Nusselt number used here was obtained by extrapolating data that are 25 years
old. This complicates not only the heat-transfer problem but the mass-transfer analogy
as well. Therefore, under these conditions, the model also becomes questionable. To
attempt to clarify the situation, it would be necessary to carry out a carefully designed
experimental program. Since this is beyond the scope of this task, no refinement of
these calculations appears justified at this time.
Refinement of the calculation and model would require an experimental investiga-
tion of the heat- and mass-transfer behavior of solid-gas dispersions in the particle-
size range below 200 microns. In the model presented here, the Nusselt number was
obtained by extrapolation of data, giving a Nusselt number one-tenth that for the limit-
ing theoretical value for heat transfer to a single sphere. The rate of mass transfer is
taken from the analogy between heat and mass transport, and is based on this low value
of Nusselt number. Thus the values calculated for mass transport are probably low if:
(1) The extrapolation is unwarranted
(2) The data are qualitative rather than quantitative
(3) The analogy breaks down.
For example, if the actual rate of mass transfer in a gas-solid system of this kind were
5 to 10 times the rate used here, and provided the reaction kinetics were sufficiently
rapid, then most of the SO^ in the flue gas would probably be converted. It is obvious
that this question can be answered only by a well-planned and carefully conducted ex-
perimental program.
BATTELLE MEMORIAL INSTITUTE
-------
RECOMMENDATION FOR THE USE OF
LIMESTONE AND DOLOMITE IN BOILER FURNACES
by
! William T. Reid
Many interesting questions come up in considering the most effective way in which
lime or magnesia can fix the sulfur in flue gas. Must the additive be calcined before it
is injected into the flue gas? Is there a difference in the behavior of lime and magnesia,
and how does dolomite compare in effectiveness with limestone? At what temperature
is the fixation of sulfur most likely to be highest? What is the effect of a shift in tem-
perature as the additive and flue gas cool in passing through superheaters, reheaters,
economizers, and air heaters? Where is the best point for injecting the additive?
What interaction can be expected between added limestone or dolomite and the ash in
the fuel?
Intensive experimentation will be required to resolve these questions. First,
however, to place such tests on a firm, basis, the fundamental thermochemical param-
eters have been considered as they affect the overall chemical reactions between lime-
stone and dolomite with flue gas. These calculations of the many possible reactions
show which are most likely to take place and the effect on the reactions of changes in
temperature and composition. Kinetic considerations based on the best available in-
formation show which factors influence the rate of 50% capture, and, in turn, demon-
strate that mass transfer rather than chemical reaction rates probably will be the
dominating factor. Such basic considerations are held as highly important in guiding
later experimentation in large central-station boiler furnaces.
This section of the report relates the thermochemical and kinetics sections to the
practical problems of using limestone and dolomite most effectively in boiler furnaces.
It is intended mainly as a guide to future tests, either on a relatively small scale in the
laboratory, or in experimentation in operating boilers. Results of the many thermo-
chemical calculations listed in the Appendix will be helpful to others interested in ex-
ploring this problem further.
Basic Conditions
Establishment of ground rules were necessary to define the conditions on which
the calculations were based. These limiting rules were fixed by factors related both to
boiler furnaces and to fuels. For example, they relate to temperature, residence time
as affected by mass flow rate, heat absorption, fuel characteristics, and ash analyses.
It was not the intent here to analyze these individually in every case, but to arrive at
overall conditions to establish broad limits for the calculations. Large deviations from
these conditions probably will affect the conclusions of this study, but the results here
nevertheless should provide a useful yardstick.
BATTELUE MEMORIAL INSTITUTE
-------
24
Composition of Flue Gas
Ultimate analyses of coal and of residual fuel will vary widely, and flue-gas com-
position will change with the fuel and with such operational factors as excess air.
Table 7 shows the composition of flue gases chosen as typical for coal and for oil firing.
Water is shown in these analyses because, of course, the temperature of the gas is
above the dew point and water is uncondensed, unlike in an Orsat apparatus. For re-
sidual fuel, the lower limit for oxygen when operating with low excess air can be taken
as 0. 2 percent; the other constituents are shifted proportionately.
TABLE 7. TYPICAL COMPOSITION OF FLUE GAS
CO 2
°2
N2
H2O
so2
Pulverized
14. 5
2. 7
75. 4
7. 1
0. 3
100. 0
Percentage Composition
Residual
Coal Normal Excess Air
12. 2
2. 7
74.9
9.9
0. 3
100. 0
Fuel
Low Excess Air
12. 5
0. 2
76.9
10. 1
0. 3
100.0
Although Table 7 shows the flue-gas composition when combustion is complete,
other conditions occur within a flame. While combustion is occurring, some regions
are overrich in fuel, while O2 can be in great excess elsewhere. These transient
conditions are most difficult to measure, and it is not within the scope of this study to
consider such nonequilibrium conditions.
Composition of Typical Coal Ashes
The impurities in coal vary widely, with at least seven constituents affecting ash
fusibility. Table 8 gives the composition of the ash from three widely differing coals,
and the typical limits of each component.
Iron in these ash analyses is reported as Fe2,O$. In coal-ash slags, where the
ash has been melted to form a silicate "glass", 90 percent of the iron usually is
present as FeO and 10 percent as Fe2O3- The state of the iron in fly ash varies widely.
It is often about half FeO and half
Sulfur trioxide shown in these analyses represents the state of the mineral matter
in the coal when it has been heated to no more than 1400 F. At higher temperatures
when the ash melts, all sulfur is expelled; coal-ash slags seldom contain more than
0. 1 percent SO3.
BATTELLE MEMORIAL INSTITUTE
-------
25
TABLE 8. COAL-ASH ANALYSES
Percentage Composition
Component
SiO2
A1203
Fe2°3
CaO
MgO
Na20
K2O
SO3
Pennsylvania ,
Pittsburgh Bed
44.7
22. 7
17. 3
6.5
0.9
(2-3)
6. 6
100.0
Kentucky,
Straight Creek
15.9
16. 3
52.4
5. 7
1. 0
(1.9)
6.8
100. 0
Illinois ,
No. 6
46. 2
22.9
7. 7
10. 1
1.6
0.7
0.8
8.9
98.9
Typical
Limits
20-60
10-35
5-35
1-20
3-4
1-4
1-12
Composition of Oil Ash
Ash from residual fuel is entirely different from coal ash. Table 9 gives the
analyses of the impurities in two residual fuels, one low and the other high in vanadium.
TABLE 9. OIL-ASH ANALYSES
Component
Si02
Al2O3+Fe2O3+TiO2
CaO
MgO
Na2O
V2°5
SO3
Percentage
Low Vanadium
(Mid-Cont.)
31.7
31.8
12.6
4. 2
6.9
Trace
10.8
98.0
Composition
High Vanadium
(Iran)
12. 1
18. 1
12, 7
0, 2
_-
38. 5
7,0
88, 6
BATTEULE
MEMORIAL INSTITUTE
-------
26
Occurrence of Sulfur
Sulfur in fuels varies widely, with the upper limit nominally 5 percent. In coal,
sulfur occurs in three forms, with seldom more than a tenth of the total sulfur occurring
as a sulfate. The remaining 90 percent or more of the sulfur is usually about evenly
divided between pyritic sulfur (FeS2) and organic sulfur associated with the coal sub-
stance. Pyrite occurs in discrete particles ranging in size from a few microns to
"sulfur balls" weighing many pounds. Although pulverizers are set to reject such large
chunks of pyrite, the smaller particles are not liberated when coal is ground only to
200 mesh (74 microns). Hence, much of the pyrite remains in the pulverized coal going
to the burners. Irrespective of source, the major part of the sulfur in coal appears
eventually in the flue gas, nearly all of it as SOz, as noted earlier. Small amounts of
sulfur may be trapped in fly ash, but usually this will not exceed 5 percent of the total
sulfur originally present in the coal.
Essentially all the sulfur in oil is present as organic compounds, for example
mercaptans, thiophenes, and complex sulfides. It can be removed only by such methods
as hydr ode sulfur ization, a relatively difficult process costing at least 50 cents per
barrel.
Temperature and Time
Defining gas temperatures in a system as complicated as a large boiler furnace
is difficult because the system is a dynamic one; the temperature of the gas stream de-
creases in a complex fashion as the products of combustion move past heat-receiving
surfaces. Being a dynamic system, time is also important, and it is this temperature-
time factor that poses some of the problems in evaluating the fixation of 50% by lime-
stone or dolomite. Corner-fired furnaces and cyclone-fired ones are quite different,
and the gas temperature-time relationship also differs considerably. Similarly, the
temperature pattern in a slag-tap furnace is not the same as that in a dry-bottom
furnace. With oil firing, the situation ie still different.
Figure I gives the time-temperature pattern assumed for this study. It repre-
sents a fair mean of the conditions in large boiler furnaces, although individual units
might differ considerably. With it as a guide, however, the temperature range was
established for the thermochemical calculations, and available reaction times were
provided for the kinetic evaluations. The figure is based on the assumption that the
flame reactions last 100 milliseconds, and that the products of combustion are in the
furnace, in the superheat and reheat sections, and in the economizer sections for
2 seconds each, with a final 100 milliseconds in the air heater before passing to the
stack. The temperatures shown are based on fair estimates expected for large steam
generators.
Deposits on the heat-receiving surfaces over which the gases flow will be at
some temperature intermediate between the metal temperature and the gas temper-
ature. Furnace-wall deposits will range from about 2000 F to 2800 F, depending on
position and whether the unit is slag tap or dry bottom; exposed-metal temperatures
in the furnace will seldom exceed about 800 F. In superheaters, the deposits will
have an average temperature of perhaps 1500 F, with exposed metal at about 1200 F.
Ash deposits in superheaters and reheaters will have a maximum surface temperature
of about 2000 F.
BATTELUE MEMORIAL INSTITUTE
-------
3£UU
2800
2400
2000
LL.
€
2 I60O
o>
a.
E
|2
I2OO
800
400
0
_ ^ Flame
\
>
^ *"
N
Superh
eaters
and
reheat ers
\
\|
\
[^f
!-* — Economizer — »•
Air
\
ft
-------
28
Reactions With Ash
As was shown in Table 2, both CaO and MgO are commonly present in coal ash
and together with iron oxides make up the major fluxing constituents. Generally, more
CaO is present than MgO. These oxides are also present in oil ash, usually as a
greater percentage of the ash than in coal.
Although almost no work has been done on the relationship between composition
and fusibility of oil ash, a great deal of effort has been spent learning how the fusibility
of coal ashes depends on their composition. Particularly good relationships have been
developed between the composition and the viscosity of coal-ash slags at different
temperatures.i *'' In addition to Newtonian viscosity, the behavior of coal-ash slags
also has been investigated at lower temperatures, when the separation of a solid phase
radically affects the flow behavior of the slag.'^"' The findings of those studies, made
more than two decades ago, have been confirmed by recent work in this country and
in England.
A basic concept developed in those early studies was that the viscosity of a coal-
ash slag at a given temperature, say 2600 F, depends only upon the "silica percentage"
of the slag, defined as
SiO2
Silica percentage = x 100 , (28)
* B SiO2 + Fe2O3 + CaO + MgO
where SiO2, CaO, and MgO are the percentages of these components shown by a
chemical analysis of the slag, and Fe-O, is the percentage of iron oxides recalculated
as Fe^Oo.
Figure 3 shows how closely experimental measurements of viscosity at 2600 F
are related to this silica percentage. In equation form, the relationship is
log(p-l) = (0. 066)(Si02) - 1. 4 , (29)
where p is the viscosity in poises at 2600 F, and SiO^ is the silica percentage. Large
variations in the ratio of SiO£/Al2O3 and of Fe2O3/CaO + MgO have almost no effect on
the viscosity. Similarly, normal variations in Na2O and K2O are not important. Also
of interest here is the fact that CaO and MgO can be taken as equivalent fluxes; they
can be considered as a single factor in affecting viscosity, at least when the percentage
of MgO is low as it usually is in coal ashes.
The major significance of the silica-percentage concept is that it shows that in-
creasing Fe2O-j or CaO + MgO leads to a more fluid slag. As the flux content increases,
the viscosity of the slag goes down markedly, being only 1 poise at 2600 F when the
silica percentage is 30. If such a slag contained 15 percent equivalent Fe2O3 and 2 per-
cent alkalies, with a SiO2/Al2O-j ratio of 2, the CaO + MgO content of the original slag
would have been 44.6 percent.
(19) Nicholls. P., and Reid, W. T.. "Viscosity of Coal-Ash Slags". Trans. ASME, 6£. 141-153(1940).
(20) Reid. W. T.. and Cohen, P., " The Flow Characteristics of Coal -Ash Slags in the Solidification Range", Ttans. ASME,
66. 83-97(1944).
BATTELLE MEMORIAL INSTITUTE
-------
30
40 50 60 70
Silica Percentage
80
A-53851
90
FIGURE 3. VISCOSITY OF COAL-ASH SLAGS AT 2600 F IN AIR
BATTELLE MEMORIAL
N S T I T U T E
-------
30
At some upper limit, CaO and MgO no longer serve as fluxes. Still greater
amounts in coal ash result in an increase in fusibility rather than a decrease. Fig-
ure 4 illustrates how adding CaO to one coal ash decreased the "flow temperature",
a measure of fusibility, until the CaO content was about 40 percent and increased the
flow temperature with further additions of
270O
2100
20 3O
CoO in Ash, percent
40 50
A-53852
FIGURE 4. EFFECT OF LIME CONTENT ON FUSIBILITY OF COAL ASH<21)
The point of this argument is that moderate amounts of lime or dolomite added
with the fuel so that there would be an intimate mixture of additive and the ash in the
fuel in the furnace during combustion could lead to serious problems with slagging,
particularly because temperatures would be so high in the flame that the slagging re-
actions could occur rapidly. If a great excess of lime or dolomite were used, so that
the CaO content in the ash was greater, say, than 50 percent, the resulting slag might
not cause troubles. If, however, only part of the lime or dolomite reacted with the ash
so that the slag contained but 30 or 40 percent CaO, the fluidity of the resulting slag
would be excessively high. This would almost certainly lead to troubles through
slagging or the accumulation of sticky deposits in superheater sections.
To complicate the picture still more, Barnhart and Williams'^^', in studying the
sintering characteristics in the laboratory of mixtures of various additives with fly ash,
showed that the sintering strength was decreased most by dolomite, CaO, and MgO in a
mixture of 20 percent additive and 80 percent fly ash. It should be noted, however,
that these mixtures were not heated above 2000 F. Hence, slagging reactions between
(21) Nicholls, P.. and Reid, W. T., "Fluxing of Ashes and Slags as Related to the Slagging Type Furnace" , Trans. ASME,
54, 167-190(1932).
(22) Barnhart, D. H.. and Williams. P. C., "The Sintering Test. An Index to Ash-Foul ing Tendency". Trans. ASME. T£.
1229-1236(1956).
BATTELLE MEMORIAL INSTITUTE
-------
31
the ash and the additive were unlikely, and the additive remained as a loose unreacted
solid to separate the fly-ash particles and lower their sintering strength.
Further questions come from comments such as by Baley in a discussion of
Barnhart and Williams 's paper. He recounts how adding only 1/3 pound of MgO per
ton of coal fired apparently decreased the amount of hard-bonded "eutectic" on tubes
in the 1500 F zone. Assuming 10 percent ash in his coal, this amount of additive would
have increased the MgO level by only 0. 2 percent, an insignificant amount.
Until more experimental evidence is in hand, it must be concluded that limestone
or dolomite added through the burner with the fuel, or injected directly into the flame in
boiler furnaces, can increase the fouling tendency of the coal ash by reducing the vis-
cosity of the coal-ash slag. Great excesses of additive may overcome this slagging
action, but as yet the exact action cannot be predicted. Caution should be used in
planning full-scale experiments where limestone or dolomite is added to high-
temperature zones. At least it should be recognized that fouling might be accentuated
by the presence of CaO and MgO that has reacted, at least in part, with coal ash.
The problem is even more complicated with oil ash, lacking information on how
lime or magnesia affect its fusibility. But with its much lower ash content, residual
fuel is less likely to produce a low-melting slag by addition of lime or magnesia be-
cause the amount of additive necessary to capture SO2 in flue gases would be greatly in
excess of that required to "flux" the oil ash. Here, the ash in the fuel would be such a
small fraction of the total deposit that the problem of increased fusibility would not
arise, and the great excess of CaO would lead to refractory rather than to fusible
deposits.
The thermochemical calculations clearly show in Table 5 that the conversion of
CaO to CapSiO^ or CaSiO?, as would be typical of reactions between added CaO and the
silica in coal ash, decreases the ability of the lime to capture SOz by as much as four
orders of magnitude. This is a great enough difference to suggest that any lime or
dolomite added at temperatures high enough to produce slagging essentially would re-
move that additive as far as reaction with SO is concerned.
Hence, from two points, the possible increase in slagging and the inability to re-
act with SO£, any lime or magnesia going into a coal-fired boiler furnace should be
added at a temperature low enough that it will not react with the silica in the coal ash.
This would eliminate adding limestone or dolomite in direct admixture with the fuel.
It would also suggest that blowing the additive into the furnace at the burner level to
take advantage of the high level of turbulence near the flames would be undesirable.
With oil firing, the situation would be quite different. Here the limestone or
dolomite could be added directly to the fuel, although this might cause practical prob-
lems in the wear of pumps and burners. Instead, injection of the additive directly into
the flame region might be more useful. Considering the low level of ash in residual
fuel oil, any silica present to tie up lime and magnesia would be so small that it could
be disregarded. A possible shortcoming could be that limestone and dolomite might be
overburned by exposure to temperatures above 2600 F, even momentarily. If heated
excessively, the resulting CaO or MgO might be physically modified so as to reduce its
reactivity, as occurs with slaked lime. Because this property varies with different
limestones and dolomites, and because neither the thermochemical calculations nor the
kinetic evaluations give any clue to the effect of overburning, it can be considered only
BATTECLE MEMORIAL INSTITUTE
-------
32
as a possible shortcoming to injecting the additive directly into the hottest zones of oil-
fired equipment. In general, limestone heated about 2450 F and dolomite heated to over
2250 F will be overburned and have poor reactivity.
Calcination
The desulfurizing action of limestone and dolomite before calcination is possible
thermodynamically as shown by the calculations. Limestone, for example, as shown
by calculations based on data in Table A- 10, may be only slightly less active at 1400 F
than CaO, shown in Table A-l. Even at 1000 F, limestone is not appreciably less ef-
fective than CaO, both of these materials having an extremely low level of SO^ in the
surrounding gas phase at equilibrium. Hence, at least by these calculations, it is not
necessary that the limestone or dolomite be calcined to capture SO2- Practically,
however, it may prove necessary to convert the additive to CaO or MgO to achieve a
feasible rate of reaction.
This same point may serve to fix the lower temperature at which CaO will be ef-
fective. Figure 1 indicates that CaO produced by heating limestone to a higher temper-
ature will revert to CaCO3 when cooled to 1415 F in flue gas containing 14. 5 per-
cent CO^. Again, it remains to be proven by experiment whether this reversion to
CaCO3 will appreciably decrease the rate at which SO£ can be captured.
Comparison of Limestone and Dolomite
Thermodynamically, at least, CaO is appreciably more effective than MgO in
reacting with SO2 under furnace conditions. As was shown in Table 1, the equilibrium
concentration at 1400 F of SO^ over CaO is only 0. 001 ppm, whereas it is 180 ppm for
MgO. At 1800 F, the equilibrium SO2 level over CaO is 1.95 ppm, while it is upward
of 500 ppm with MgO. Stated another way, equivalent equilibrium SO£ levels are ob-
tained about 400 F higher with CaO than with MgO. Because the rate of the reaction
between solid and gas must be some function of temperature, even though the kinetics
calculations could not develop this relationship, it follows that CaO at any given tem-
perature could be appreciably superior to MgO for tying up SO^- On this same basis,
limestone should be better than dolomite for removing SO2 from flue gas.
Catalytic Activity of Impurities in Limestone
The formation of CaSO,^ from the reaction of CaO with SO2 implies oxidation, as
is shown in Table A-l. Lime can also react with SO2 alone to form CaSO3, as il-
lustrated by Table A-4, but the equilibrium level of SO2 to form CaSO4 at 1000 F is
only 1. 7 x 10~ ^ ppm, whereas to form CaSO-j, it is 16 ppm. On this basis, the oxida-
tion step would be greatly preferred. Hence, any operation that enhances the formation
of SO3 from SO2 will help to increase the amount of SO2 captured, and possibly the rate
as well.
BATTELLE MEMORIAL INSTITUTE
-------
33
Probably the simplest way of increasing the formation of CaSO^ over
would be to incorporate a small quantity of Fe2O3 in the limestone. This Fe2O3,
serving as a catalyst to promote the approach to equilibrium of 803, also would
increase the rate of the overall reaction. Wahnschaffe in Gerrnany(^3)j working with
dolomite containing 0. 7 percent Fe£O3, showed that desulfurization increased as ad-
ditional Fe2C>3 was added, with essentially no further appreciable gain above 2 percent
total Fej>O3. Pechkovskiit^) also shows that transition-metal oxides present in CaCC>3
catalyzed the conversion of SO^ to 803 and increased the rate at which sulfur was
captured.
Hence, limestone or dolomite containing up to perhaps 3 percent Fe^C^ would be
preferred over purer stone containing less than 0. 5 percent Fe£O3. Addition to lime-
stone or dolomite of mill scale or other inexpensive sources of Fe^C^, preferably
before grinding so as to achieve the best possible mixture, should improve the capture
of SO2.
Preferred Location in Furnaces
for Addition of Limestone or Dolomite
As shown earlier, limestone or dolomite intended for the fixation of SOz probably
should not be added with the fuel, particularly with pulverized coal containing silica in
the ash. The preferred point of addition should be where the gas temperature is as
high as possible consistent with minimum reaction with ash. For most coals, this tem-
perature will be about 2000 F. Typically, this is about at the furnace outlet, or roughly
at about the point where the flue gases enter the superheater and reheater sections.
The minimum temperature at which any appreciable capture of SC>2 will still occur
is less easily defined, for there will be a gradual fall-off in activity as the gases cool.
Practically, however, the transition of CaO to CaCC>3 on cooling to 1400 F in flue gas
may set the lower limit. However, since CaCO3 may be nearly as effective in capturing
SO2 as is CaO, as was discussed under Calcination, it is possible thermodynamically
that there is no lower temperature limit. This will be an important point to check ex-
perimentally. Lacking such verification, it seems logical to conclude that the rate of
SO? pickup will be unpractically low at less than perhaps 1000 F, and that this be con-
f, fr r *- I *r i- f
sidered as a reasonable lower limit, as was demonstrated by Pechkovskii.*-*'
This temperature analysis means, of course, that the limestone or dolomite must
be added to the gas stream at the entrance to the superheater section, and that 2 or 3
seconds at most will be available before the gas stream will have been cooled below
1000 F. Table 6 shows, however, that if mass transfer controls the process as the
kinetic study indicates it will, then only a third of the SO2 would be captured in
4 seconds. As noted in that section, the situation might be improved by using more
additive, grinding the additive to a smaller particle size, or increasing contact time.,
but how far these alternatives can be carried cannot be predicted now.
(23) Wahnschaffe, E.. " Desulfuiization of Flue Case* by the Dolomite Process" (PHS file*, source documentation not available).
BATTELLE MEMORIAL INSTITUTE
-------
34
Quantity of Additive Required
For coal firing, relatively large amounts of limestone or dolomite would be re-
quired to fix all the SOz in the flue gas. For example, in a typical pulverized-coal-
fired boiler furnace serving a 500-megawatt turbine, the amount of coal burned per day
can be taken as roughly 5000 tons. With 3 percent sulfur, the total quantity of 50%
produced per day would be 300 tons. Also, if the ash content of this coal were 10 per-
cent, and if the ash contained 40 percent SiO£, there would be ZOO tons of SiO^ per day
passing through the furnace.
Assuming that the limestone was added at the furnace outlet so that there would
be no reaction with ash, the amount needed would come to 465 tons of limestone daily
to combine stoichiometrically with the SO2» The maximum quantity of CaSO^ formed
would be about 640 tons daily.
If, besides the SO^, even half the available SiO2 in the ash were to react with the
added CaO, assuming that the limestone was admixed with the fuel, then 166 more tons
of limestone would be required daily to form
These are extraordinarily large quantities even of such an inexpensive material
as limestone or dolomite. Handling this large quantity of material economically would
pose some troublesome engineering problems. Hence, before any serious attempts are
made to use limestone or dolomite on this huge scale, it would seem worthwhile to
make smaller plant tests where these theoretical considerations could be checked.
Settling for capture of half the SO£ in the flue gas would ease the problem appre-
ciably, but there is no assurance as yet that stoichiometric ratios of limestone or
dolomite to SO^ would indeed fix all the SC>2 as CaSO* or MgSC>4. That conclusion, too,
would have to be based on large-scale testing.
CONCLUSIONS
The results of this study can be summarized as follows:
Thermochemical calculations based on the best available data show that CaO
is capable, theoretically, at equilibrium of removing all but 1 ppm of SO£
from flue gas at 1770 F.
With MgO, the temperature can be no higher than 1200 F for an SO2 level of
1 ppm at equilibrium.
Increasing the temperature increases the concentration of SO2 remaining in
the flue gas, so that CaO is incapable of removing any SO2 from a typical flue
gas above 2250 F, and MgO is incapable of removing SO^ above 1550 F.
Limestone theoretically converts to CaO in boiler atmospheres at about 1415 F;
at lower temperatures, CO^ in the flue gas will tend to reconvert CaO into
BATTELLE MEMORIAL INSTITUTE
-------
35 and 36
Uncalcined limestone should be capable of desulfurizing flue gas.
Theoretically, CaCO3 can be in equilibrium with well under I ppm of
at 1400 F. How rapidly equilibrium can be approached is not known.
Limestone or dolomite added with the fuel in pulverized-coal-fired boiler
furnaces is apt to cause serious slagging problems, while the probable
reaction of CaO with SiO^ at furnace temperatures essentially may eliminate
CaO as a means of capturing SO^ in hot flue gases carrying suspended coal ash.
For maximum effectiveness in coal-fired equipment, limestone or dolomite
should be added at the point where the flue gas is no hotter than 2000 F,
probably at the entrance to the superheater sections.
With oil-fired furnaces where the ash content of the fuel is low, the upper
temperature limit will be fixed by the tendency to overburn limestone and
dolomite. Because CaO is not effective as a desulfurizer above 2250 F,
and MgO above 1550 F, it would appear pointless to add limestone or
dolomite at appreciably higher temperatures.
The lower temperature limit where desulfurizing essentially would stop is
not known, but 1000 F appears to be a reasonable temperature below which
no further reactions may occur.
No information is available on the rates at which limestone or dolomite
might fix SO2 in flue gas, and it does not appear possible to extrapolate
such data from other systems.
Mass transfer may be the controlling process in removing SO2 from flue gas
with suspended particulate CaO and MgO. Hence, for maximum effectiveness,
the quantity of additive should be as high as is practical, its particle size
should be small, and the contact time should be long.
Large amounts of limestone or dolomite would be necessary to remove
from the flue gas of a typical boiler furnace. On the basis of stoichiometric
calculations, about 230 tons of limestone would be needed daily to capture
half the SO2 emitted by the boiler furnace of a 500-megawatt unit burning a
fuel with 3 percent sulfur.
Extensive experimentation will be necessary to convert these calculations
into practical design data. Most badly needed will be measurements of the
rate at which SO£ is captured at different temperatures by CaO, MgO, and
raw limestone and dolomite. Lacking those experimental measurements,
the calculations made here can serve only as a guide to the actions occurring
in boiler furnaces.
JJW : DAP : JFW : RHC : AL : WTR/mln:js;sel
BATTELLE MEMORIAL INSTITUTE
-------
APPENDIX
THERMOCHEMICAL VALUES FOR BASIC
REACTIONS IN FLUE-GAS SULFUR FIXATION
BATTELLE MEMORIAL INSTITUTE
-------
A-l
REACTIONS GIVEN IN TABLES
TABLE A-l. CaO(c) + - S^g) + O2(g) = CaSO4(c)
CaO{c) + S02(g) + | 02(g) = CaS04(c)
TABLE A-2. MgO(c) + ~ S2(g) + O2(g) = MgSO4(c)
MgO(c) + S02(g) + 02(g) = MgS04(c)
TABLE A-3. CaMg(CC>3) (c) = CaCO (c) + MgCO (c)
CaMg(C03>2 = CaC03(c) + MgO(c)
TABLE A-4. CaSO^c) + O2(g) = CaSO^c)
CaS04(c) = CaS(c)
TABLE A-5. MgSC>4(c) = MgS{c) + 2O (g)
MgSCUc) + \ O, = MgSO .(c)
•3 C* £* *t
TABLE A-6. MgC03(c) + SO2(g) = MgSO3
-------
A-2
REACTIONS GIVEN IN TABLES
(Continued)
TABLE A-9. MgC03(c) + SO^g) = MgSO^c) + CC>2(g)
MgC03(c) + S02(g) + | 02(g) = MgS04(c)
TABLE A-10. CaC03(c) + SO2(g) + i O2(g) = CaSO^c) + CO2(g)
CaC03(c) + S03(g) = CaS04(c) +
TABLE A-ll. i S2(g) + 02(g) + SO2(g)
S03(g) = S02(g) + | Oz(g)
TABLE A-12. CaO(c) + | S2(g) = CaS(c) + | O
MgO(c) 4- S2(g) = MgS(c) + 02(g)
TABLE A-13. 4C(gr) 4- CaSO.(c) = CaS(c) + 4CO(g)
4CO(g) + CaS04(c) = CaS(c)
TABLE A-14. 4C(gr) + MgSO4(c) = MgS(c) + 4CO(g)
4CO(g) + MgS04(c) = MgS(c) -I- 4C02(g)
TABLE A-15. 2CaO(c) + SiO (c) = Ca SiO.(c)
^ L. 4
CaO(c) + SiO = CaSiO (a)
£ J
TABLE A-16. CaO(c) + H2O(g) = Ca{OH)2(c)
MgO(c) + H20(g) = Mg(OH)2(c)
BATTELLE MEMORIAL INSTITUTE
-------
A-3
REACTIONS GIVEN IN TABLES
(Continued)
TABLE A-17. Na2SO4(c) = Na2O(c) -I- SO2(g) + ~ O2(g)
Hh 3S02(g> 4-
TABLE A-18. FeSO4(c) = Fe2O3(c) + SO^g) + O2(g)
FeS04(c) = | Fe304(c) + SO2(g) + ~ O2(g)
TABLE A-19. FeSO^c) = FeO(c) + SC>2(g) + O2(g)
BATTELLE MEMORIAL INSTITUTE
-------
A-4
TABLE A-l. THE THERMOCHEMISTRY OF SULFUR FIXATION BY CALCIUM OXIDE
Temp,
F
300
400
500
600
700
800
900
1000
1100
. 1200
1300
1400
1500
1600
1700
1800
1900
2000
2500
3000
CaO(c)+ * S2(g) + |.O2(g)»CaSO4(c)(a) CaO(c)+SO2(g)-t-i. O2(g)*CaSO4(c)(a)
AF°R, cal mole"1
-170,095
-165,381
-160, 683
-155,987
-151,317
-146, 658
-142,023
-137,421
-132. 839
-128,295
-123,776
-119,298
-114, 848
-110,443
-106,068
-101,734
-97 ,-446
-93, 192
-72. 60o(b)
-53, 152
L°8lO K
88.045
75.644
65.846
57.887
51.317
45.787
41.077
37.026
33.496
30.403
27.664
25.231
23.050
21.091
19.317
17.707
16.243
14.902
9.648
6.043
AF*D, cal mole'1
tv
-91,036
-87.291
-83,566
-79,843
-76.154
-72,464
-68,816
-65, 185
-61,587
-58,018
-54.476
-50, 977
-47,496
-44,073
-40.658
-37,311
-33. 985
-30,711
-14.949(b)
-276
Log10K
47.122
39.926
34.245
29.630
25.827
22.623
19.904
17.563
15.529
13.749
12.175
10.782
9.532
8.416
7.404
6.494
5.665
4.911
1.987
0.031
Temp,
K
422.2
477.8
533.3
588.9
644.4
700.0
755.6
811.1
866.7
922.2
977.8
1033.3
1088.9
1144.4
1200.0
1255.6
1311.1
1366.7
1644.4
1922.2
(a) Data based on Appendix References (1. 7).
(b) Melting
point of CaSO4 = 1723 K.
TABLE A -2. THE THERMOCHEMISTRY OF
Temp,
F
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
2500
3000
MgOfcJ+SO^g)*1 02(g) =
£.
AF*R, cal mole"1
-62, 020
-58.245
-54,488
-50, 733
-47,003
-43,271
-39,571
-35,885
-32,224
-28,588
-24, 973
-21,392
-17*823
-14,306
-10,793
-7,339
-3,903
-511
15,905
31,332
*MgS04(c)
L°g10K
32.103
26.641
22.329
18.827
15.941 '.
13.509
11.445
9.669
8.125
6.775
5.581
4,524
3.577
2.732
1.966
1,277
0.651
0.082
-2.114
-3.562
SULFUR FIXATION BY MAGNESIUM
MgO(c) + Is2(g)^02
-------
A-5
TABLE A-3. THE THERMOCHEMISTRY OF DOLOMITE DISSOCIATION
Temp,
F
300
400
500
600
700
800
900
CaMg(CO3)2(c) = CaCO3(c) +
AF"R, cal mole"1
2. 602
2,558
2.513
2,469
2.424
2,380
2.336
MgC03(c)
Log10K
27,011
23.269
20.291
17.855
15.837
14.136
12.984
11.781
10.747
9,876
9.074
8.368
7.739
7.026
6.528
6.084
5.680
5.318
3.966
2.762
CaSO4(c) = CaS(c)+2
AF*R, cal mole"1
191,634
186.892
182.165
177,444
172.745
168,059
163,423
158,809
154, 172
149,573
144, 998
140,498
136, 048
131. 599
127. 168
122.834
118.536
114. 225
93, 600
02(g)(b)
Login K
-99.195
-85.482
-74. 649
-65.849
-58.585
-52.468
-47.269
-42.789
-38.875
-35.445
-32.407
-29.715
-27.305
-25.131
-23. 159
-21.380
-19.758
-18.265
-12.439
Temp.
K
422.2
477.8
533.3
588.9
644.4
700.0
755.6
811.1
866.7
922.2
977.8
1033.3
1088.9
1144.4
1200.0
1255.6
1311.1
1366.7
1644.4
1922.2
(a) Data based on Appendix References(l-3, 6, 7. 9). Data estimated by Brewer's approximation, Appendix Reference (13).
(b) Data based on Appendix References (1, 4. 7). Data estimated by Brewer's approximation, Appendix Reference (13).
Note: Melting point of CaS = 1123 K.
BATTELLE
MEMORIAL
INSTITUTE
-------
A-6
TABLE A-5. THE THERMOCHEMISTRY OF MAGNESIUM SULFIDE AND MAGNESIUM SULFATE FORMATION
Temp,
F
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
2500
3000
MgS04(c) » MgS(c) + !
AF*R. cal mole'1
185.710
180.870
176.053
171.236
166.452
161. 665
158,011
152,051
147. 348
142.82l(c)
138. 172
133.565
128. 968
124.431
119,897
115,426
110.973
106, 565
85,085
64.589
.«*«
LoUloK
-96.128
-82.728
-72.144
-63.546
-56.450
-50.472
-4S.701
-40.968
-37.154
-33.845
-30.882
-28.249
-25.884
-23.762
-21.835
-20. 090
-18.498
-17.040
-11.308
-7.343
MgSOg(c) + i Oa • MgS04(c)(b>
2
*f*R. caJ mole"1 Log10 K
-53.711 27.802
-52.224 23.887
-50.719 20.784
-49,199 18.258
-47.662 16.164
-46.117 14.398
Temp.
K
422,2
477.8
533.3
588.9
644.4
700.0
755.6
811.1
866.7
922.2
977.8
1033.3
1088.9
1144.4
1200.0
1255.6
1311.1
1366.7
1644.4
1922.2
(a) Data based on Appendix Reference (2).
(b) Data bated on Appendix References (1-3. 9).
(c) Melting point of MgS • 923 K.
TABLE A-6. THE THERMOCHEMISTRY OF SULFUR DIOXIDE FIXATION BY CALCIUM CARBONATE
AND MAGNESIUM CARBONATE
Ten*, MgCO^SO^.MgSO^CO^')
F AF'R, cal mole"1 ^lO K
300 2.110 -1.092
400 2.091 -0.956
500 2.056 -0.842
600 2,016 -0.748
700 1.957 -0.664
800 1.908 -0.596
900
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
2500
3000
CaC03(c)+S02(g]
tf *R. cal mole'1
-12.453
-12,103
-11.806
-11.539
-11.298
-11,051
-9,796
-9,317
-8.799
-8.130
-7.606
-7.069(c>
-6,513
-6,766
-6. 177
-5, 601
-5.013
-4.409
-1,131
-801
|"CaS03(c)+C02(g)(b)
LoglflK
6.446
5.536
4.838
4.282
3.832
3.450
2.833
2.510
2.219
1.927
1.699
1.495
1.307
1.292
1.125
0.975
0.835
0.705
0.150
0.091
Temp,
K
422.2
477.8
533.3
588.9
644.4
700.0
755.6
811.1
866.7
922.2
977.8
1033.3
1088.9
1144.4
1200.0
1255.6
1311.1
1366.7
1644.4
1922.2
(a) Data based on Appendix References (1-3. 7, 9).
{b) Data based on Appendix References (1-3, 6, 7. 9).
(c) CaCO3 theimochemlcally unstable in furnace atmosphere.
BATTEULE
MEMORIAL
INSTITUTE
-------
A-7
TABLE A-7. THE THERMOCHEMISTRY OF CALCIUM OXIDE WITH SULFUR DIOXIDE AND CARBON DIOXIDE
Temp,
P
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
2500
3000
CaO(c) + SO2(g) = Ca
AF*R> cal mole"1
-38.854
-36.416
-34, 051
-31,729
-29,454
-27,185
-23, 925
-21,459
-18.967
-16,343
-13,874
-11.409
-8,936
-7.289
-4.810
-2,357
92
2,548
14.896
24.017
S03(C)<*>
LogloK
20.112
16.656
13.954
11.775
9.989
8.487
6.920
5.182
4.783
3.873
3.101
2.413
1.793
1,392
0.876
0.410
-0.015
-0.407
-1.980
-2.731
CaO(c) + C02(g)
AF'R. cal mole"1
-26.401
-24,313
-22.245
-20, 190
-18.156
-16. 134
-14,129
-12.143
-10,168
-8.213
-6.268
-4. 340
-2.423
-523
1,367
3.244
5,105
6,957
16,027
24, 818
= CaCO3(c)(b>
t-SloK
13.666
11.120
9.116
7.492
6.157
5.037
4.086
3.212
2.564
1.946
1.401
0.918
0.486
0.100
-0.249
-0.565
-0.851
-1.112
-2.130
-2.822
Temp,
K
422.2
477.8
533.3
588.9
644.4
700.0
755.6
811.1
866.7
922.2
977.8
1033.3
1088.9
1144.4
1200.0
1255.6
1311.1
1366.7
1644.4
1922.2
(a) Data based on Appendix References (1-3, 6, 9).
(b) Data based on Appendix References (1, 7).
Data estimated by Brewer's approximation, Appendix Reference (13).
TABLE A-8. THE THERMOCHEMISTRY OF MAGNESIUM OXIDE REACTING WITH SULFUR DIOXIDE AND CARBON DIOXIDE
Temp,
F
300
400
500
600
700
800
900
MgO(c) + S02(g) s
AF*R, cal mole'1
-8.309
-6.021
-3,769
-1 . 534
659
2,846
i MgS03(c)(a>
Log10K
4.301
2.754
1.544
0.569
-0,223
-0.888
MgO(c) + COgtg
&F°R. cal mole"1
-10,419
-8.112
-5,825
-3,550
-1.298
938
3,153
) •= MgC03(c)
LoglflK
5.393
3.710
2.387
1.317
0.440
-0.293
-0.912
Temp,
- K
422.2
477.8
533.3
588.9
644.4
700.0
755.6
(a) Data based on Appendix References (1-3, 9). Data estimated by Brewer's approximation, Appendix Reference (13).
(b) Data based on Appendix References (1, 7).
BATTELLE
MEMORIAL
INSTITUTE
-------
A-8
TABLE A-9. THE THERMOCHEMISTRY OF MAGNESIUM CARBONATE REACTING WITH SULFUR OXIDES
Temp,
F
300
400
500
600
700
800
MgCO3(c) + SG
MgSO4(c) + CC
AF°Rl cal mole'1
-37,615
-37,389
-37,151
-36.898
-36, 625
-36, 331
l,~/o) m
>2
L°8lOK
19.470
17.101
15.224
13.693
12.421
11.343
MgS04(c) + CO,
AF*£, cal mole"1
-51,601
-50. 133
-48, 663
-47,183
-45.705
-44,209
tef»
Log^Q K
26.710
22.930
19.942
17.510
15.500
13.802
Temp.
K
422.2
477.8
533.3
588.9
644.4
700.0
(a) Data based on Appendix References (2, 7, 9).
(b) Data based on Appendix References (2, 7).
TABLE A-10. THE THERMOCHEMISTRY OF CALCIUM CARBONATE REACTING WITH SULFUR OXIDES
Temp,
F
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
1500
1600
noo
1800
1900
2000
2500
3000
CaC03(cJ+S02(g) + io2,
CaS04(c) + C02(g)
AF'R, cal mole"1
-64, 635
-62.978
-61,321
-59, 654
-57,997
-56,330
-54, 687
-53,043
-51,419
-49. 805
-48,208
-46, 637^°)
-45, 073
-43,551
-42.025
-40,555
-39,090
-37,667
-30,976
-25,094
fo) •
Log10K
33.457
28.805
25.129
22.138
19.669
17.586
15.817
14.292
12.965
11.803
10.775
9.864
9.046
8.317
7.653
7.059
6.516
6.023
4.117
2.853
CaC03(c) -
CaSO4(c) <
AF*R. cal mole'1
-50.649
-50.234
-49, 809
-49,369
-48.917
-48.452
-49,208
-47,621
-47.216
-46.955
-46, 552
-46, 176^c>
-45.810
-45,462
-45.107
-44, 782
-44,464
-44,216
i- S03(g) »
C02(g)(b)
Log10K
26.217
22.976
20.411
18.321
16.589
15.127
14.232
12.831
11 . 905
11.127
10.404
9.766
9.194
8.68-2
8.215
7.794
7.411 '
7.070
Temp,
K
422.2
477.8
533.3
588.9
644.4
700.0
755.6
811,1
866.7
922.2
977.8
1033.3
1088.9
1144.4
1200.0
1255.6
1311.1
1366.7
1644.4
1922.2
(a) Data based on Appendix References (1, 7).
(b) Data based on Appendix References (2, 7, 9).
(c) CaCO3 thermochemicaUy unstable in furnace atmosphere.
BATTELLE
M E.M O R I A L
INSTITUTE
-------
A-9
TABLE A-ll. THE THERMOCHEMISTRY OF SULFUR AND OXYGEN AS GASES
Temp,
F
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
2500
3000
(a) Data
(b) Data
| S2 (g> + 02 (g)
AF*R, cal mole"1
-71,942
-71,928
-71,879
.-71,805
-71,689
-71.562
-73,216
-72, 240
-71,262
-70,434
-69,460
-68,490
-67,520
-66,552
-65, 582
-64, 614
-63, 647
-62, 681
-57.867
-53,070
based on Appendix Reference
- S02(g)
Log10K
37.239
32.899
29.455
26.647
24.313
22.342
21.176
19.464
17.969
16.691
15.524
14.485
13.551
12.709
11,944
11.246
10.609
10.023
7.691
6.034
(2).
SO3(g) a SOg(
AF °R, cal mole"1
13,986
12,744
11,512
10,285
9,080
7,878
5,479
5.422
4,203
2.850
1,656
461
-737
-1,911
-3.082
-4. 227
-5.374
-6,549
g)*!
-------
A- 10
TABLE A-13. THE THERMOCHEMISTRY OF THE REDUCTION OF CALCIUM SULFATE BY CARBON AND CARBON MONOXIDE
Temp,
F
800
1700
2600
4C (gr) + CaSO4(c) = CaS(c) +
AF*R, cal mole"1
1.939
-81,432
-160,228
4CO(g)
L°glO*
-0.605
14.830
20.598
4CO(g) + CaSO4(c) - CaS(c) n
AF*R, cal mole"1
-43,821
-42,872
-39,428
h 4C02(g)(a)
Log10K
13.681
7.808
5.069
Temp.
K
700
1200
1700
(a) Data based on Appendix References(2, 4, 7).
TABLE A-14. THE THERMOCHEMISTRY OF THE REDUCTION OF MAGNESIUM SULFATE BY CARBON AND CARBON MONOXIDE
Temp,
F
800
1700
(a) Data
Temp,
F
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
2500
4C(gr) + MgSO^c)
AF"R, cal mole"1
-4.207
-88,299
=»MgS(c)+4CO(g)(a)
Log10K
1.313
16.081
4CO(g) + MgS04(c)
AF°R, cal mole"1
-50.503
-50. 631
= MgS(c) + 4C02(g)
LoglO K
15.767
9.221
Temp,
K
700
1200
based on Appendix Reference (2).
TABLE A -15
2CaO(c) + SiO2(c)
AF*R, cal mole'1
-30, 710
-30,770
-30. 840
-30, 910
-30. 970
-31.040
-31,110
-31,170
-31,240
-31,310
-31,370
-31,440
-31,510
-31.570
-31, 640
-31,710
-31,770
-31,840
-32,173
. THE THERMOCHEMISTRY
= Ca2Si04{c)
Log1Q K
15.896
14.074
12.638
11.471
10.503
9.691
8.998
8.398
7.877
7.419
7.011
6.649
6.324
6.028
5.762
5.519
5.296
5.091
4.276
OF FORMATION OF CALCIUM SILICATES
CaO(c) + Si02
AF*R, cal mole"
-21,249
-21,243
-21, 236
-21.229
-21. 223
-21.216
-21,209
-21.203
-21,196
-21,189
-21.183
-21.176
-21.169
-21,163
-21.156
-21, 149
-21.143
-21.136
« CaSiO3(a)(a)
Log10 K
10.999
9.716
8.702
7.878
7.198
6.624
6.134
5.713
5.345
5.021
4.734
4.479
4.248
4.041
3.853
3.681
3.524
3. 380
Temp,
K
422.2
477.8
533.3
588.9
644.4
700.0
755.6
811.1
866.7
922.2
977.8
1033.3
1088.9
1144.4
1200.0
1255.6
1311.1
1366.7
1644.4
(a) Data based on Appendix Reference (5).
BATTELLE
MEMORIAL
INSTITUTE
-------
A-11
TABLE A-16. THE THERMOCHEMISTRY OF HYDRATION OF CALCIUM OXIDE AND MAGNESIUM OXIDE
Temp,
F
300
400
500
600
700
800
900
1000
CaO(c) + H20(g) = Ca(OH)2(c)(a)
AF*R, cal mole"1
-10,882
-8.956
-7.061
-5. 190
-3,347
-1.526
272
2.042
LoglflK
5.633
4.096
2.894
1.926
1.135
0.476
-0.079
-0.550
MgO(c) + H20(g) * Mg(OH)2(c)(b)
-1
AF £, cal mole * ^°8lO ^
-4.165 2.156
-2.197 1.005
-261 0.107
1.658 -0.615
Temp,
K
422.2
477.8
533.3
588.9
644.4
700.0
755.6
811.1
(a) Data based on Appendix References (1, 7).
(b) Data based on Appendix Reference (2).
TABLE A-17. THE THERMOCHEMISTRY OF DISSOCIATION OF SODIUM SULFATE AND ALUMINUM SULFATE
Temp,
F
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
Na2S04(c) = Na2O(c) + SO.
AF0-, cal rnole"1
132, 791
125,436
118,081
110,719
103.365
96,010
88,655
81,300
73,945
66, 591
59,236
•-
T
Log-^Q K
-68.736
-57.373
-48.388
-41.088
-35.055
-29. 974
-25. 642
-21.905
-18.645
-15.781
-13.239
--
Al2*fWb)
Log10 K
-62.571
-50.025
-40. 146
-32. 148
-25. 567
-20.046
-15.360
-11.342
-7.851
-4. 801
-2. 108
.0.278
Temp,
K
422.2
477.8
533.3
599.9
644.4
700.0
755.6
811.1
866.7
922.2
977.8
1033.3
(a) Data based on Appendix Reference (1). Data estimated by Brewer's approximation. Appendix Reference (13).
(b) Data based on Appendix References (1, 7).
BATTELLE
MEMORIAL
N S T I T U T E
-------
A-12
TABLE A-18. THE THERMOCHEMISTRY OF FERRIC SULFATE TO FERRIC OXIDE AND FERRO-FERRIC OXIDE
Temp,
F
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
2500
3000
1
FeSO4(c) 3 -^Fe^O^c) + S
AF°R, cal mole"1
27,551
24, 399
21,252
18, 099
14,953
11, 800
8,648
5,501
2,348
-799
-3.951
-7, 098
-10,251
-13,397
-16,550
-19,702
-22, 849
-26,002
-41,747
-47,499
02(g) + 4-02(g)(a)
1*810 K
-14.261
-11. 160
-8.709
-6.717
-5.071
-3. 684
-2.501
-1.482
-0. 592
0.189
0.883
1.501
2.057
2.558
3,014
3.429
3.809
4.158
5. 548
6.537
1
FeSO4(c) « gFe3O4(c) + S(
AF*R, cal mole"1
34, 390
30, 960
27,535
24, 105
20,681
17,250
13, 820
10,395
6,965
3,540
110
-3,315
-6,745
-10,169
-13, 600
-17,030
-20,455
-23, 885
-41,020
-58,160
Log10K
-17.801
-14. 161
-11.284
-8. 945
-7.014
-5.385
-3.997
-2.801
-1.756
-0.839
-0.024
0.701
1.354
1.942
2.477
2.964
3.409
3.819
5.452
6.612
Temp,
K
422.2
477.8
533.3
588.9
644.4
700.0
755.6
811.1
866.7
922.2
977.8
1033.3
1088.9
1144.4
1200.0
1255. 6
1311.1
1366.7
1644.4
1922.2
(a) Data based on Appendix References (1,7). Data estimated by Brewer's approximation. Appendix Reference (13).
TABLE A-19. THE THERMOCHEMISTRY OF DISSOCIATION OF FERROUS SULFATE
TO FERROUS OXIDE
Temp,
F
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
1500
1600
1700
FeSO4(c) « FeO(c) + SO;
AF*R, cal mole"1
65, 608
gl.600
47,600
43,592
39,592
35, 584
31.476
27,576
23,569
19, 568
15,560
11.560
7,552
3.552
-456
*g> + |o2(g)C*>
L°glO*
-28.784
-23.601
-19.506
-16.177
-13.427
-11.109
-9.133
-7.430
-5.943
-4.637
-3.478
-2.445
-1.516
-0.678
0.083
Temp,
K
422.2
477.8
533.3
588.9
644.4
700.0
755.6
811.1
866.7
922.2
977.8
1033.3
1088.9
1144.4
1200.0
(a) Data based on Appendix References (1, 7). Data estimated by Brewer's approximation,
Appendix Reference (13).
BATTELLE MEMORIAL INSTITUTE
-------
A- 13 and A- 14
REFERENCES FOR THERMOCHEMICAL CALCULATIONS
(1) Latimer, Wendell M., Oxidation Potentials, Second Edition, Prentice Hall, Inc.,
New York (1952).
(2) JANAF Thermochemical Data, issued by the Dow Chemical Company, Midland,
Michigan.
(3) Stull, D. R. , and Sinke, G. C. , Thermo dynamic Properties of the Elements,
American Chemical Society, Washington, D. C. (1956).
(4) Elliott, J. F., and Gleiser, M. , Thermochemistry for Steel Making, Addison-
Wesley, Reading, Massachusetts (I960), Vol I.
(5) Kubaschewski, O., and Evans, E. L. , Metallurgical Thermochemistry,
Pergamon Press, New York (1958).
(6) Maslov, P. G. , "Thermodynamic Characteristics of Calcium, Gallium, Indium,
and Thallium Compounds", Journal of General Chemistry of the U. S. S. R. ,
V ^9 (5), 1387-1397(1959),
(7) Kelley, K. K., "Contributions to the Data on Theoretical Metallurgy", U. S.
Bureau of Mines Bull. No. 584 (I960).
(8) Kelley, K. K. , and King, E. G. , "Contributions to the Data on Theoretical
Metallurgy", U. S. Bureau of Mines Bull. No. 592 (1961).
(9) Coughlin, J. P., "Contributions to the Data on Theoretical Metallurgy", U. S.
Bureau of Mines Bull. No. 542 (1954).
(10) Wagman, D. D., Evans, W. H., Halow, I., Parker, V. B., Bailey, S. M., and
Schumm, R. H., "Selected Values of Chemical Thermodynamic Properties",
National Bureau of Standards Technical Note 270-1, Superintendent of Documents,
Washington, D. C. , 1965.
(11) Stout, J. W., and Robie, R. A., "Heat Capacity from 11 to 300° K, Entropy,
and Heat of Formation of Dolomite", J. Phy. Chem. , 6J, 2248-2252 (1963).
(12) Halla, F., "Note on the Thermodynamics of Formation of Dolomite", J. Phys.
Chem., 6j> (3), 1065 (March, 1965).
(13) Quill, L. L., Ed., The Chemistry and Metallurgy of Miscellaneous Materials:
Thermodynamics, Paper 4, McGraw-Hill Book Company, Inc., New York (1950),
(14) Wicks, C. E., and Block, F. E. t "Thermodynamic Properties of 65 Elements -
Their Oxides, Halides, Carbides, and Nitrides", U. S. Bureau of Mines Bull.
No. 605 (1963).
(15) Kelley, K. K. , and Anderson, C. T. , "Contributions to the Data on Theoretical
Metallurgy", U. S. Bureau of Mines Bull. No. 384, 1935.
(16) Kelley, K. K., "Contributions to the Data on Theoretical Metallurgy", U. S.
Bureau of Mines Bull. No. 406, 1937.
BATTELLE MEMORIAL INSTITUTE
-------
ADDENDA
Desulfurizing Action of Hydrated Lime
Since issuing this report, information has been received suggesting that Ca(OH)j>
is being used with considerable promise in Germany for the fixation of SO2 in flue gas.
Slaked lime had been omitted from the original calculations here because of its instability
above quite moderate temperatures. However, in light of this demonstrated ability of
Ca{OH)2 to capture SO2> the same sort of calculations were made for the reaction
Ca(OH>2 + S02 - CaS03
as were included initially in the report for other lime and magnesia reactions.
«
Table A- 20 gives the results of thermochemical calculations for this reaction.
The free-energy change, AFj^, suggests that there is a good likelihood for the reaction
TABLE A-20. THE THERMOCHEMISTRY OF CALCIUM HYDROXIDE
REACTING WITH SULFUR DIOXIDE TO FORM
CALCIUM SULFITE AND WATER VAPOR(a>
Ca{OH)2(c) + S02(g) ^ CaS03(c) + H2O{g)
Temperature,
F
cal mole~
Log1Q K
Temperature,
K
300
400
500
600
700
800
900
-------
Both compounds are extremely good desulfurizing agents if equilibrium can be
approached. However, Ca(OH);> has a lower thermodynamic efficiency than an equiva-
lent amount of CaO, as is evident from the data in Table A-21. If Ca{OH)2 proves to be
more effective as a desulfurizing agent than CaO, the behavior can be explained best
on the basis of a more reactive physical state of CaO resulting from dehydration of
Ca{OH)2, not one of greater thermodynamic efficiency.
The likely reaction product of CaO or Ca(OH)2 with SO2 is CaSOj. As was shown
in Table A-4, CaSOj will be thermochemically reactive with the excess O2 in flue gas to
form
TABLE A-21. COMPARISON OF CALCULATED EQUILIBRIUM CONCENTRATION
OF SO2(g), ppm IN FLUE GAS, IN THE PRESENCE OF EXCESS
Ca(OH)2 OR CaO
Ca(OH)2 + S02(g) ~ CaS03(c) + H2O CaO(c) + SO2(g) ^ CaSO3(c)
Temperature, With 7. 1% H2O in Flue Gas and Unit Activity of Solids Assumed
F Activity of Solids to be Unity
300
500
700*
800*
2. 4 x 10-10
2.0 x 10-8
1. 0 x 10-4
6. 9 x ID'4
7. 7 x 10-15
1. 1 x 10-8
1. 0 x 10-4
3. 3 x 10-3
•Above dissociation temperature of CafOH)^ in flue gas.
Reliability of Calculations
A question has been raised about the reliability of these thermochemical calcula-
tions. How precisely do they define the actual equilibrium concentrations, and how
reliably can the calculations be used?
The thermodynamic working Tables A-l through A-20 were compiled and cal-
culated from the most reliable data given in recent sources, as listed under the
references given on pages A-13 and A-14. Because of the need for arithmetical con-
sistency in this type of study, the number of significant figures in the values reported
is beyond that which actually exists.
In several of the tables, high-temperature heat-capacity data were not available.
In these cases, Brewer's Approximation (Reference 13) was used. Brewer's Approxi-
mation gave data comparable with that reported in Table A-l, which is based on ex-
perimental values to 2000 F. It is believed that the calculations listed in the tables can
be rated as "good" for evaluating practical temperature limits of the desulfurizing action
of the compounds under consideration. By estimating the limits of error for several
important cases, it was concluded that the SO2 concentration calculated at a specific
temperature should not vary from the actual value by more than 10 ppm SO2. Consid-
ering that flue gas nominally contains 2000 ppm to 3000 ppm SO2 and that a good de-
sulfurizer would lower the SO2 only to 200 ppm to 300 ppm, these calculations are con-
sidered to be well within reasonable practical limits.
BATTELLE MEMORIAL INSTITUTE
------- |