Note: Descriptions are shown in the official language in which they were submitted.
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-1-
CONTROL OF COMBUSTION SYSTEM EMISSIONS
TECHNICAL FIELD
' The present invention relates to processes for improved operation of coal-
fired and
other carbonaceous-fuel-fired electrical utility boilers, incinerators, and
high temperature
combustion reactors. More particularly, the present invention relates to cost
effective
processes for reducing fouling of carbonaceous-fuel-fired combustion system
components,
reducing corrosion within such systems, and reducing undesirable and noxious
stack
emissions.
BACKGROUND ART
In coal-fired power generating plants, as weli as in other industrial
processes
involving combustion of coal, a number of the products of the combustion
process include
compounds that have an adverse influence on boiler operation, or they are
environmentally
undesirable and the discharge of which into the environment is subject to
environmental
regulations. Such compounds include sulfur oxides (SOX), nitrogen oxides
(NOx),
hydrochloric acid, and such heavy metals as mercury, arsenic, lead, selenium,
and
cadmium. Additionally, a significant number of nations, including the European
Union and
Japan, have taken steps to further limit the emissions of carbon dioxide
(COZ), and similar
steps have been proposed in the United States but are currently being
implemented by few
of the 50 states.
In order to meet environmental limitations affecting the discharge into the
atmosphere of the most prevalent of the most widely regulated compounds,
sulfur dioxide
(SO2), combustion products from such piants and processes are commonly passed
through
flue gas desulfurization (FGD) systems. The treatment of flue gases to capture
SO2 is
often effected in lime- or limestone-based wet scrubbers, in which lime or
limestone slurries
contact the flue gases before they are discharged into the atmosphere. The
sulfur oxides
are thereby chemically converted into insoluble calcium compounds in the form
of calcium
sulfites or sulfates. The sulfur oxides contained in such combustion products
are thus
converted into less-environmentally-harmful compounds that are either disposed
of in
landfills, or, when suitably modified or treated, are sold as marketable
chemicals as a result
of their conversion into marketable gypsum.
Although useful for converting some sulfur oxides, the widely-used types of
lime/limestone scrubbers are not very effective in capturing the 1% to 1.5% of
the sulfur in
the fuel that is transformed during the combustion process into gaseous sulfur
trioxide
(SO3), which can escape from the scrubber. The SO3 poses operating problems
within the
boiler itself, in that it leads to corrosion and fouling of low temperature
heat exchange
surfaces. Additionally, it poses environmental problems in that unless it is
captured or
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-2-
transformed, the SO3 results in a persistent, visible plume and the discharge
of corrosive
and potentially hazardous sulfuric acid mist. Further complicating the matter,
selective
catalytic reactors (SCR's), which are available and because of high capital
costs are
installed primarily in the larger, newer such plants to comply with nitrogen
oxide emission
regulations, essentially cause a doubling of the amount of SO3 that is
generated.
Consequently the already serious operational and environmental problems caused
by the
presence of SO3 are magnified.
The SO3 emission problem has been addressed chemically using a variety of
alkaline chemicals (wet and dry) that are injected into the system at many
different points in
the flue gas flow path. Lime or limestone injected into the high temperature
region of the
boiler can be effective in capturing the SO3, but the commercial materials
that are generally
utilized tend to magnify boiler deposit problems and increase the quantity of
particulates
that can escape from the electrostatic precipitators (ESP's). The adverse
impact on the
precipitators is also encountered when lime or lime hydrate is injected as
powders into the
lower temperature region downstream of the SCR's. On systems with scrubbers
capable of
capturing particulates, the precipitator problem can be circumvented by
injecting the lime
downstream of the precipitator. However, fine powders tend to become
agglomerated
during the course of handling and result in relatively inefficient S02
capture, thereby
necessitating dosage at several times stoichiometric. Further, the injection
of slurries
downstream of the ESP pose serious problems relative to drying and deposit
buildup in the
ducts, because the low temperatures at that point do not provide the
evaporative driving
force that is needed to quickly flash off the water.
Sodium compounds, such as the bisulfite, carbonate, bicarbonate and
carbonate/bicarbonates (Trona) compounds, have also been injected into the
cooler
regions of the system and are effective in SO3 capture. However, they pose
material
handling, ash disposal, and potential deposit problems. They also tend to have
poor
utilization efficiencies unless they are ground to very fine particle sizes.
Relatively coarse
particles are prone to formation of an outer sulfate shell, thereby inhibiting
utilization of the
unreacted chemical inside the shell. Additionally, grinding of such materials
is expensive,
and it creates storage and handling problems because of the fineness and
hygroscopic
nature of the particles. Ash disposal issues arise because of the solubility
of sodium
compounds, and in some cases steps to insure containment in the disposal ponds
may be
required.
Commercially available, but relatively expensive, oil-based magnesium
additives
can be effective in SO3 capture. In that regard, one of the most effective
chemical
techniques for controlling both ash-related fouling in the boiler, and also
the corrosion and
emission problems associated with SO3 generated in solid-fueled boilers, is
the injection
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-3-
into the upper region of the boiler of oil slurries of MgO or Mg(OH)2. That
technology was
originally developed for use with oil-fired boilers in which the magnesium-
based oil
suspension was usually metered into the fuel. It was later applied to coal-
fired boilers. The
most widely accepted mode of application of such additives today is by
injection of slurries
of MgO or Mg(OH)2 into the boiler above the burners and just below the region
at which a
transition from radiant heat transfer to convective heat transfer occurs.
Another approach to SO3 capture involves the use of so-called "overbased"
organic-
acid-neutralizing additives of the type that are included in motor oils and as
fuel oil
combustion additives. Those additives are actually colloidal dispersions of
metallic
carbonates, usually magnesium or calcium. When burned with the fuel, they are
effective
at near stoichiometric dosage in capturing SO3 and in mitigating ash deposits
caused by
vanadium and/or sodium in the oil. The colloids are stabilized by carboxylic
or sulphonate
compounds and are known to provide mostly particles in the Angstrom range.
Though very
expensive, the "overbased" compounds are widely used at low dosages to capture
vanadium in heavy-oil-fired combustion turbines. Although they have been
utilized in SO3
capture efforts, there have been no prior reports of their use for capturing
either SO2 or
toxic metals.
Although emissions benefits can be obtained by the use of the so-called
"overbased" compounds, their much higher cost and combustibility make them a
less
attractive option for most applications. Additionally, the combustibility of
the overbased
materials requires hard piping as well as additional safety devices, each of
which involves
increased costs.
I In addition to oil-based slurries, Mg(OH)2 powders and water-based slurries
have
also been utilized as fireside additives in boilers, but because of their
generally coarser
particle size they are less efficient in capturing the SO3. Water slurries of
MgO have also
been injected through specially modified soot blowers installed on oil and
Kraft-liquor-fired
boilers, in which they moderated high temperature deposits but had only a
nominal impact
on S03-related problems because of an inability to apply the chemicals
continuously.
In addition to limitations on SOx emissions, regulations aimed at controlling
mercury
emissions from coal-fired boilers have been promulgated by regulatory
authorities, and
regulations applicable to other toxic metals are anticipated eventually. A
considerable
amount of research aimed at finding practical techniques for capturing such
toxic metals
has shown that high-surface-area solids can capture a significant portion of
mercury by
adsorption, if the mercury is in an oxidized form rather than in an elemental
form.
Oxidants, either added to or naturally present in the fuel, such as chlorides,
can facilitate
the oxidation. Although high-surface-area lime can be effective in mercury
capture, the
usual commercial products can result in operational problems in the form of
ash deposits
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-4-
and increased stack emissions. The most widely accepted way to achieve mercury
capture has been the injection of expensive activated carbons in the cooler
regions of the
boiler gas path.
Combustion systems requiring additional emission control generally fall into
two
broad groups. The first group includes those systems that are sufficiently
large and are
sufficiently new to justify the large capital investment in scrubbers for SOZ
and in selective
catalytic reactors (SCR's) for NOx. The second group includes those systems
that are
older and smaller, and for which scrubbers and/or SCR's are difficult to
physically retrofit
and involve a major capital investment that is often difficult in to justify
economically.
In the second group of systems, SO2 emission regulations are currently being
met
by switching to more costly, low-sulfur fuels and/or by utilizing market-based
emissions
credits. Combustion process modifications have also been used successfully to
reduce
NOX emissions, but the reduction is often insufficient to bring the systems
into compliance
with the latest regulations. Those systems may also generate a byproduct fly
ash that is
higher in unburned carbon as a result of combustion modifications that are
aimed at
minimizing NOX formation. The efficiency loss as a result of the increased
unburned
carbon is small, typically less than about 0.5% of the fuel carbon, but if the
amount of
unburned carbon in the ash is too high (>5% of the ash), the ash becomes
unmarketable,
thereby converting a potential revenue stream from the sale of ash into an
expenditure for
ash disposal. Considerable work has gone into optimizing the burners of such
systems,
but with limited success. Because limiting NOx emissions is an important
objective,
techniques for separating the carbon from the ash are being pursued as an
alternative.
The larger, newer systems can justify the major investment in SCR's, while the
smaller, older systems tend to use selective non-catalytic reduction (SNCR),
which
employs similar reactions to the SCR's using ammonia or an amine, but without
the
catalysts. Both of those control technologies result in a small amount of
ammonia in the
flue gas downstream of the SNCR or SCR systems. The ammonia can react with the
SO3
that results from the combustion process to form low-melting-point ammonium
bisulfate,
which can foul air preheaters that are further downstream in the flue gas flow
path.
Both groups of combustion systems are likely to be required to conform with
additional regulations that require the capture, of trace quantities of toxic
metals. Despite
gas scrubbing, the scrubber/SCR-equipped systems that utilize higher sulfur
content fuels
also face a new, stack opacity problem that results from a doubling by the
SCR's of the
SO2 that is catalyzed to SO3 and is emitted as a visible, sulfuric acid mist
plume. The acid
in the flue gas also results in system operating problems by plugging and
corroding lower
temperature components of the system.
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-5-
The sulfuric acid plume problem has resulted in major environmental public
relations issues for utilities, as evidenced by American Electric Power
Company's purchase
of the town of Cheshire, Ohio, because of acid mist discharge issues. The
Department of
Energy has spent millions of dollars in testing various SO3 control
techniques, and a variety
of acid-neutralizing systems are being installed. Some systems are currently
operated only
during the NOX season, that part of the year when NOx controls must be
employed
(currently May through September). Those SO3 mitigation systems utilize a
variety of
alkaline chemical compounds that are injected at various points in the flue
gas path to
effect the acid neutralization. Most of those chemicals, including Ca(OH)2,
Mg(OH)2,
Trona, and SBS (sodium bisulfite), are relatively coarse in particle size,
with the finest-
sized particles tested reportedly having a particle size of about 3 microns.
However, those
chemical compounds are difficult to deploy, they are utilized at rates that
are 3 to 12
multiples of stoichiometric, and their use involves significant costs.
Although the use of
furnace injection of those coarser particles as an emissions control vehicle
has been
evaluated extensively, most current installations feed chemicals for SO3
control in the
cooler section of the system at a point downstream of the SCR's, either as
powders,
slurries, or solutions.
It is likely that the remaining boiler systems and combustion systems without
scrubbers will soon need to meet more stringent SO2 regulations or face early
shutdown if
a practical, low capital cost, moderate operating cost, pollution control
system does not
become available. Those same power plants will soon also be required to
capture mercury
and other toxic metals, as well as to deal with more stringent SOX and year
round NOX
emission limitations.
Considerable research has been conducted on techniques for capturing the toxic
metal pollutants before they can escape from the combustion system and/or
damage the
SCR catalyst. That research has shown that the injection at various points in
the boiler of
finely sized, high-porosity, high-surface-area particulate materials, such as
specially
modified CaO, silicates, MgO, or activated carbon can help to capture most of
the metals.
Heavy metals (Hg, Se, and As) capture has been shown to be significant when
lime is
injected into the high temperature region at twice the sulfur stoichiometric
ratio, even
though the surface area of the injected materials is relatively modest, of the
order of about
1 to 4 m~/gm or more, and even though competition exists for that same
reagent/reactant
surface area by the acid-forming gases. The current regulatory focus is on the
capture of
mercury, and the current user focus is on injection into the cooler regions of
the boiler of
expensive, high-surface-activated carbon. However, the adverse operating and
environmental impacts of the other toxic metals will eventually lead to
emissions
regulations for the other toxic metals
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-6-
SO2 control utilizing powdered limestone injection into the high temperature
furnace, a technology known as LIMB (Limestone Injection Multistage Burner),
has been
investigated extensively since the 1970's. However, that approach has not been
widely
implemented because treatment rates twice stoichiometric with -325 mesh
powders (typical
mean particle size of about 20 microns) captured no more than about 65% of the
SO2.
That approach also dramatically increases the ash burden (as much as double).
And it has
posed deposit problems in the boiler convective pass, requiring near
continuous operation
of the soot blowers, and has overburdened the particulate control device. Much
of the
research effort has focused on the creation of high-porosity, high-surface-
area CaO by
flash calcination of the limestone in the furnace. However, the desired
improvement in
chemical utilization efficiency has not been achieved because of the plugging
of the pores
of the high-porosity particles with CaSO4, thereby reducing the accessible
surface area for
reaction and leaving a core of unreacted CaO. Some work with particle sizes in
the 5
micron range has been reported, but that approach also has not been utilized
commercially
because of the pore-plugging problem, along with what is perceived to be a
high cost of
grinding limestone into a fine particle size.
With regard to SO3- capture, the University of North Dakota Energy and
Environmental Research Center recently reported that the tiny fraction, less
than about
1.5%, of submicron-size ash particles that are present in fly ash have been
found to adsorb
SO3. It suggested that the fraction of those particles is important for
controlling the SO3
problem. The addition of fine alkaline materials (under 5 microns) was also
mentioned.
Other workers have reported that fly ash will absorb toxic metals, but its low
surface area
leads to poor capture efficiency.
Finally, the possibility of modifying combustion conditions to increase the
fraction of
submicron size particles has not been reported. Instead, the focus has been on
adding
excessive amounts of what are perceived as fine, ground limestone (-325 mesh)
having a
median particle size of around 20 microns. The reason for that focus is that
limestone is
inexpensive, and even if one were to desire smaller particles the normal
techniques for
providing very fine particle sizes having high-surface-areas have all been
judged to be too
expensive.
Some research has been conducted on what might be described as a multi-
pollutant control process simulating the furnace injection of calcium and
magnesium
compounds slurried in solutions of nitrogen compounds. Theoretically, the
combination
would address all the emission issues except CO2. Although the injection of
nitrogen
solutions to control NOx is in wide use on power boilers, the combination with
calcium
slurries for simultaneous SO2 capture has not been commercially adopted.
Reportedly, the
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-7-
failure to do so is the result of problems with settling and pluggage in the
slurry injection
systems.
Reducing CO2 emissions has thus far not been the subject of regulations in
much of
the world. Emphasis has been placed on improving efficiency of fuel use. And
research
on sequestering the CO2 is ongoing, with some CO2 captured, liquefied, and
used in
enhancing oil recovery. Most of the commercial SOx emissions control processes
for
fossil-fueled combustion systems employ limestone (directly or as lime) with
the net result
being a significant secondary emission of CO2. The scrubbers employing
limestone on the
larger, newer units are the lowest emitters (about 0.7 ton CO2/ton of SO2
captured) while
those using lime have net emissions at least twice as high because of the
thermal loss in
the calciners. Because the utilization of the limestone is so poor with
conventional Furnace
Sorbent Injection (FSI), the CO2released per ton of SO2 captured is nearly 14
tons/per ton.
It is therefore an object of the present invention to provide improved
processes by
which boiler operational and emissions problems can be reduced more
economically than
is attainable by presently utilized methods.
DISCLOSURE OF INVENTION
Briefly stated, in accordance with one aspect of the present invention, a
reduction of
undesirable combustion system emissions, primarily SOx and toxic metals, but
also CO2,
NOx,, and unburned carbon, is achieved by preparing an essentially agglomerate-
free,
high-solids-content aqueous slurry containing CaCO3, Ca(OH)2, MgCO3, or
Mg(OH)2
particles having a particle size of about one to two microns or smaller. The
method of the
invention includes injecting the slurry into the boiler furnace at
temperatures sufficient to
flash calcine the alkaline earth compounds to still smaller size oxide
particles. Both the
ratio of water to solids and the point of injection can be selected to limit
the temperature
exposure of the product oxide and thereby limit grain coalescence.
The present invention is also directed to:
1. the incorporation of alkali-type dispersants or high-melting-point alkali
compounds to offset the adverse impacts on the ESP of the alkaline earth
scavenging agents for capturing SO2;
2. the addition of soluble chlorides to freeze condition the dispersion and,
when needed, to oxidize elemental Mercury (Hg) to facilitate Hg capture;
3. the mitigation of ammonium bisulfate air heater fouling by preferentially
scavenging the SO3i
4. the provision of a simple, low cost way to extend the operating life of SCR
catalysts by scavenging the damaging toxic metal Arsenic;
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-8-
5. the provision of a simple, low cost way to minimize total CO2 emissions
from unscrubbed emissions control systems;
6. the provision of a means for recycling industrial byproducts and wastes
that
also contributes modestly to reducing NOx emissions;
7. the provision of a cost effective way to restore the marketability of high
LOI
fly ash; and
8. the provision of a simple, low cost way to capture SO3 and toxic metals
using fly ash as the capture medium.
In another aspect of the present invention, fly ash or other thermally stable
minerals, particularly those with significant alkali content, are ground,
either dry or wet
ground, to provide ash particles having a size of the order of about one to
two microns or
smaller. The finely-ground ash particles are injected, either alone or in
combination with
the alkaline earth slurries, in quantities and at locations in the flue gas
path appropriate to
the pollutant to be addressed, generally closer to the furnace outlet because
it provides
more reaction time. The injection location is not as critical for the capture
of toxic metals,
which should be effective if injected after the economizer. The use of fly ash
is a less
costly way to reduce undesirable emissions. Further, fly ash can be a partial
substitute for
calcium compounds or for activated carbon, and it does not contribute to
overall process
emissions of CO2.
BRIEF DESCRIPTION OF DRAWINGS
The structure, operation, and advantages of the present invention will become
further apparent upon consideration of the following description, taken in
conjunction with
the accompanying drawings in which:
Figure 1 is a photomicrograph showing the sizes of calcium carbonate particles
that
were produced by prior art processes;
Figure 2 is a photomicrograph showing the sizes of calcium carbonate particles
that
are produced by a process as disclosed herein for producing very-finely-sized
particles;
Figure 3 is a table showing the effect of particle size per unit weight upon
the
number of particles and their surface area;
Figure 4 is a table showing particle size and surface areas for a number of
commercially available hydrated limes;
Figure 5 shows the effect of particle size on pollutant capture effectiveness;
Figure 6 is a table showing the typical ash composition of representative U.S.
coals;
and
Figure 7 is a table showing a comparison of the ash composition of two
lignites with
wood ash.
CA 02628198 2008-05-01
WO 2007/053786 1 PCT/US2006/043219
-9-
BEST MODES FOR CARRYING OUT THE INVENTION
The processes herein disclosed relate to the capture of undesirable pollutants
that
result as products of the combustion process, particularly in coal-fired
combustion systems
such as those employed in industrial operations or electrical power generating
stations.
Among the pollutants that are more effectively captured as a result of
practicing the
processes of the present invention are SOZ, SO3, HCI, and toxic metals, such
as mercury,
selenium, lead, and arsenic. The present processes also relate to reducing the
overall
amount of CO2 released in the course of capturing the other pollutants.
The processes in accordance with the present invention primarily involve the
injection into the high temperature combustion zone of a furnace, where
temperatures are
in the range of from about 1090 C to about 1260 C, of a particulate slurry of
alkaline earth
metal compounds that flash dry or calcine to provide a large number of
smaller, very finely
sized calcium oxide or magnesium oxide particles having a particle size of
about I to 2
microns or finer, and that are present as distinct, individual, fine
particles, as opposed to
agglomerates, and that are available for capturing the undesirable pollutants.
The very finely sized alkaline earth metal oxide particles that are produced
within
the furnace are sufficiently small that they are carried along with and follow
the flow lines of
the flue gas that contains the combustion products. Because of their very
small size and
very low weight, most of the particles act like gas molecules and flow with
the flue gas
around the heat exchange tube surfaces within the furnace, rather than
impinging on the
surfaces, as a result of which very little ash buildup occurs on those
surfaces.
The primary sources of compounds for providing the metallic oxide particulates
include lime, limestone, and other calcium compounds that yield calcium oxide
when
exposed to high temperatures, such as dolomite. Also suitable are Mg(OH)2,
MgO, or
other magnesium compounds that yield magnesium oxide when exposed to high
temperatures.
A suitable suspension in the form of a slurry can be prepared by finely
grinding the
alkaline earth metal compounds in an aqueous medium. The fine grinding is
carried out in
a media mill or another type of fine grind mill to subject the solids to shear
forces and
reduce their size, to provide an aqueous slurry containing finely ground,
suspended solids.
The solids in such a slurry can typically have surface areas in the range of
from about 4
m2/gm to about 20 m2/gm. The dispersion concentrations can range from about
20% to
about 85% solids by weight, with a preferred concentration range of from about
50% to
about 75% solids by weight. Generally, a relatively high solids concentration
is preferred,
in order to minimize the amount of water that is introduced into the furnace
with the solids.
However, handling properties and site-specific operating considerations will
influence the
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-10-
solids concentrations that are selected for a particular application. An
important attribute of
the slurry is a very fine particle size and substantial dispersion of the
particles, and the
avoidance of as many as possible of larger-size agglomerates of the particles.
The desired very fine particle size and the substantial dispersion of the
particles
throughout the slurry can be achieved by grinding the alkaline earth metal
compounds in a
media mill in the presence of a dispersant. One type of suitable media mill is
available
from Union Process Inc., of Akron, Ohio. Suitable dispersants that can be
utilized in the
grinding process include anionic surfactants and sodium salts of a
polycarboxyic acid, or
polyacrylates. Commercially available surfactants that are suitable include
Acumer 9300,
available from Rohm and Haas Company, of Philadelphia, Pennsylvania, or Darvan
7 or
Darvan 811, available from R. T. Vanderbilt Company, Inc., of Norwalk,
Connecticut.
Dispersant concentrations in the slurry can be of the order of from about
0.20% to about
5% by weight. Acceptable results are obtainable with dispersant concentrations
within the
range of from about 0.25% to about 1.5% by weight.
Another way to obtain the desired high-solids-concentration slurry that
contains
particles having sizes of from one to two microns involves the preparation of
a suspension
having as raw materials metallic carbonate, hydroxide, or oxide particles that
are then
solubilized, as the bicarbonate, and reprecipitated as carbonates. The raw
material can be
CaCO3i MgCO3, Ca(OH)2, Mg(OH)2, as well as the oxides of calcium and
magnesium.
Further, 'dolomitic limestone and products derived from it, such as lime or
lime hydrate, as
well as corresponding magnesium compounds, can also be used. The suspension is
subjected to a flow of CO2 gas in the presence of an anionic surfactant or of
a salt of a
carboxylic acid or a sulphonic acid. The CO2is bubbled through the slurry to
react with the
suspended materials to form a soluble compound, the bicarbonate, which is then
decomposed to form the ultra-fine carbonates.
Those same particle size preparation steps, involving either grinding or
solubilization and subsequent reprecipitation, can also be employed to process
some high-
lime-content fly ash in which the lime content is of the order of at least
about 5% of the ash
by weight, industrial wastes, such as lime waste from beet.sugar processing,
or used
refractory brick, for use as a pollutant scavenger in a combustion system.
Some of those
materials typically have an organic content, generally less than about 10%,
and when
injected into the furnace provide some additional heating value to offset the
heat decrease
associated with the water contained in the aqueous reagents undergoing
calcination.
Because the material is added at the location in the gas flow path that is
normally the
introduction point that is selected to achieve NO,, reduction by what is known
as "reburn",
some additional NO, reduction will be realized. The NO, reduction can be
increased
further by increasing the combustible content of the injected slurry.
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-11-
The calcined, reduced size oxide particles that result when the slurry of
alkaline-
earth-metal compounds are injected into the high temperature region of the
furnace are of
a particle size of about 0.5 micron or finer. The external surface area of the
particles is
about 40 times that of a -325 mesh (20 micron) limestone particle and provides
about
61,000 times as many oxide particles per pound of material is provided. The
result of the
presence of such smaller particles in the high temperature region of the
combustion zone
will be the capture of SOX and up to 90+% of toxic metals at a stoichiometric
ratio of about
1.15 times, or less, of the fuel sulfur content. Moreover, because of their
small size and
light weight, the very finely sized particles will act within the furnace like
gas molecules.
They will therefore pass around the heat exchange tubes with substantially
fewer of the
fine particles impacting the heat exchange tubes, reducing particle deposition
on the tube
surfaces and thereby improving the effective heat transfer within the,furnace.
Figures 1 and 2 are photomicrographs that each represent an area having a
width
of 0.44 mm and show the vast difference in the sizes of CaCO3 particles
depending upon
the process utilized. Figure 1 shows the powder particles that result from the
prior art
processes that produce a product that is at the fine end of the range of
commercial
offerings, generally denoted as -325 mesh (5 micron median), and that result
from
commercially-ground powdered limestone. In that regard, the agglomerates that
can be
seen in Figure 1, although composed of a number of smaller particles, are
effectively
single, large, relatively heavy particles. The particles shown in Figure 1 are
of a finer size
(median size of 5 microns) than most commercially available ground particles,
which
generally fall within the median size range of from about 10 to about 20
microns. Although
the prior art refers to the surface area and the pores of such particulate
agglomerates, it
appears not to appreciate that in actual practice the pores of the
agglomerates get plugged
by reaction products or by other, smaller particles. Consequently, the
effective surface
areas of the agglomerates for pollutant capture purposes is reduced, leaving a
core of
unreacted lime.
Figure 2 shows the powder particles that result from practicing the processes
disclosed herein that provide sub-micron size particles that are essentially
agglomerate-
free dispersions having a median particle size of about 0.7 microns. The very
low fraction
of agglomerated stone particles is achieved by applying both shear and
dispersant
chemicals during the wet preparation of the sub- and micron-size stone, and by
using spray
injectors designed to minimize agglomeration of the dried/calcined particles.
The SO2 concentrations in flue gas are typically of the order of about 2,000
ppm.
But the presence of a large number of other gas molecules, roughly 500 per
molecule of
SO2, tends to limit the contact of a CaO particle with an SO2 molecule. The
toxic metals to
be captured, such as Hg, Se, and As, are present in flue gas at much lower
concentrations,
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-12-
of the order of parts per billion. Thus, reducing the particle size from 20
microns to 1
micron increases the number of CaCO3 particles per pound by about 7800 times,
which will
significantly increase the probability of a resulting CaO particle finding and
reacting with an
SO2i an SO3, or a toxic metal molecule in a space that is crowded with other
molecules.
Figure 3 is a table that dramatically shows the effect upon the surface area
of the
particles and upon the number of particles available for reaction of reducing
the particle
diameter. Because the large number of very finely sized particles that are
produced in
practicing the present invention act like gas molecules, they are carried
along with and are
subject to the same movements and random motions during flow as are other gas
molecules that are present, thereby increasing the probability of contact with
pollutant
molecules. Clearly, increasing the probability of contact increases the
probability of more
complete reaction. Moreover, the fine size of the particles reduces the
chances of a
significant unreacted core of lime.
Another approach to increase overall chemical utilization still further is to
incorporate in the milled dispersion a small quantity, less than about 5%, of
overbased
calcium or magnesium compounds, which at 0.005 microns (50 nanometers) contain
countless particles per pound. The quantity added will be governed by
economics. Since
most overbased products are oil based, a surfactant will be needed to make the
two
materials compatible.
The increase in surface area that results from particle size reduction is not
as
dramatic as is the increase in the total number of available particles, only a
40 fold
increase, but it is significant for pollutant capture performance since it is
essentially all
external surface area that is available for reaction. The problems reported in
most of the
prior work occurred where much of the available surface was internal surface
area, within
surface-connected pores of agglomerated particles. The pores were then plugged
by the
reaction with SOa, thereby preventing full utilization of the surface area of
the pollutant
capture material.
Figure 4 is a table that is adapted from a table presented in "Chemistry and
Technology of Lime and Limestone," by Robert S. Boynton, 2"d Ed., 1980, at
page 340,
and presents the specific surface in cm2/g rather than in m2/g for 25
commercial hydrated
limes (multiply by 1 x 10 -4 to convert). The relatively large particle sizes
and the relatively
low specific surface values for a number of commercially available hydrated
limes are
shown. Slurries of the hydrates can be reduced to the one to two micron range
and
deagglomerated by employing milling and surfactants as described above in the
context of
the present invention.
A further difference between the process in accordance with the present
invention
and the previous approaches resides in the fact that the larger effective
particle size of the
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-13-
agglomerated particles that are present in the previous approaches are likely
to result in a
significant fraction of those larger particles, having a size of over 4 or 5
microns, impacting
and attaching to the surface of a heat exchange tube. As a result, those
particles are
effectively removed from the gas stream, and they are thereby rendered less
available for
scavenging pollutant molecules during a significant part of the very short
time that particles
are within the high temperature region of the furnace. Additionally, the
particles that attach
to the outer surfaces of the heat exchange tubes result in a coating on the
surfaces of the
tubes, thereby reducing the heat transfer effectiveness, and consequently
reducing the
overall operating efficiency of the heat transfer system. In previous full-
scale trials,
operators found that to reduce heat exchange tube deposits it was necessary to
operate
the soot blowers nearly continuously, at a significant cost in the energy
required and in
greater equipment wear and tear.
Figure 5 is a graphical depiction of the effect of particle size on pollutant
capture
effectiveness. The larger particles characteristic of the previous approaches
react with the
combustion gases up to a point that is limited by the ultimate formation of a
CaSO4 shell
around a CaO core. The core portion is therefore no longer available for
reaction with the
combustion gases. The smaller particles that are provided by the processes
disclosed
herein, on the other hand, will result in a more complete reaction and
pollutant capture, by
exposure of the greater surface area of the smaller particles to a larger
volume of the
combustion gases, because the greater number of smaller particles present a
significantly
larger total surface area available for reaction before a reaction-limiting
CaSO4 shell is
formed.
Figures 6 and 7 are two tables adapted from Steam/lts Generation and Use, 39tn
edition, Section 15-2, published by The Babcock & Wilcox Company, 1978,
showing the
typical ash content and chemistry of some U.S. coals and lignites. They show
that some
forms of ash, particularly those from the Western states, can be quite high in
CaO.
Therefore, if the ash particle size is reduced and is activated as disclosed
herein, such
materials can potentially be useful and economic sources of pollution control
reagent both
for systems not having scrubbers as well as those having scrubbers. Clearly,
reinjecting
processed high calcium ash would be more appropriate for injection in systems
not having
scrubbers, while similarly processed low calcium ash would be useful to
capture SO3 and
toxic metals in systems having scrubbers.
Further in connection with SO2 capture, it is known that humidifying the flue
gas
ahead of the electrostatic precipitator to bring the dew point to within about
11 C to about
17 C of saturation significantly enhances SO2 capture. The added moisture
condenses on
the fine ash and lime particles under those conditions to yield a liquid film
that stimulates
the acid/base reaction. The multiplicity of finely-sized particles achieved in
the course of
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-14-
carrying out the processes of the present invention performs that function
much more
effectively than those relatively fewer, large agglomerates from dry powder
that do manage
to continue through the system to the cooler regions of the boiler.
As was earlier noted, utilizing the deagglomerated product that is produced by
carrying out the present invention reduces the potential for ash deposits in
the convective
section of the boiler. The dramatic increase in CaO particles thus also
significantly
increases the chances of preventing corrosive attack of the high temperature
tube surfaces
by semi-molten ash particles. The net effect is to make those ash particles
less sticky and
less prone to attach themselves to and to build up on the surfaces of the heat
exchange
tubes.
Other benefits of the agglomerate-free, fine particle chemical addition in
accordance with the present invention include the capture of almost all of the
SO3, which
allows lowering the boiler exit temperature without encountering the fouling
and corrosion
of cold end equipment by either ammonium bisulfate or sulfuric acid. Further,
in addition to
the net increase in system thermal efficiency and reduced maintenance, the
lower
temperatures and reduced gas flow will enhance both precipitator efficiency
and fan
capacity.
Additionally, utilizing the processes in accordance with the present invention
will
result in the dust collector burden to be about half that experienced in known
processes
utilizing -325 mesh lime or limestone. Because CaS04 is less detrimental to
precipitator
performance than is free CaO, the significant increase in CaO utilization
realized with the
present invention serves to minimize the adverse impact of a larger particle
size limestone
injection on the precipitator performance. As mentioned above, air heater
fouling and
corrosion is also minimized, permitting the lowering of flue gas exit
temperatures. That
step has the potential both to modestly enhance system heat rate and also to
reduce gas
volume to the precipitator, which enhances precipitator efficiency and reduces
fan power
demand. The total solids burden (ash + CaSO3 or CaSO4 + CaO) and the operating
specifications of the particular electrostatic precipitator will determine if
the advantages of
reduced gas volumes and reduced ash temperatures are sufficient to offset the
disadvantage of higher resistivity on the particular precipitator, and will
influence whether
supplemental dust treatment or collection capacity is needed.
It is known that increasing the submicron fraction of the dust burden can have
both
positive and negative impacts on the operation of the ESP. Moderate increases
in the
number of small charged particles increase the electric field close to the
electrode, which
will improve particle collection of all particle sizes. If the number of
particles smaller than 1
micron is too large, the ESP current will be reduced to the point where
charging and
collection is not as effective. Thus, for any given boiler-ESP combination, a
balance would
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-15-
have to be struck between the amount and size of the fine particles employed
to capture
SO2 in the ESP with maximum efficiency, while limiting the number of submicron
particles
to that amount that will enhance rather than inhibit particulate collection
performance.
Alternatively, to achieve those same positive impacts on the ESP, one can
target a more
optimal SOM capture that can be achieved with sub-stoichiometric injection.
Another
alternative is to employ a commercially available ash conditioning system that
agglomerates the fine particles, or to add humidification just before the
entrance to the
ESP.
The alkaline-earth-based chemicals that are injected into the combustion zone
to
capture SOx and toxic metal pollutants can impede the collection of
particulate emissions in
some electrostatic precipitators. But by slightly modifying the composition of
the injected
chemicals, the negative impact of the alkaline-earth-based chemicals can be
mitigated, and
the performance of the ESP can be enhanced. In that regard, co-injecting
modest amounts
of alkali compounds to counteract the adverse impacts of the alkaline-earth-
based
reagents was not previously recognized as an option to address that problem.
Instead, the
previous approach focused on providing additional systems, such as
humidifiers, and
ammonia and SO3 fly ash conditioning equipment, which present retrofit
problems and
increase the capital costs in older systems.
As earlier noted, the addition of finer particle sizes yields efficient SOX
capture at
treatment ratios that are much closer to stoichiometric than those achieved
previously,
thereby allowing the power plant operator to lower the furnace exit gas
temperature, to
decrease gas volume and velocity through the ESP, and to increase the capture
efficiency.
The lower exit gas temperature is dependent upon neutralizing any SO3 present
in order to
avoid corrosion, and that temperature reduction can be achieved by heat
exchange in new
installations, or by spray cooling in retrofits, such as disclosed in U.S.
Patent No.
4,559,211.
If the system temperature can be lowered, such as by turning off steam air
preheater coils, and utilizing an oversized ESP, those factors together with
the reduced ash
burden afforded by the more efficient SO2 capture technology disclosed above
can be
sufficient to ensure compliance with particulate emission requirements. If
such
temperature lowering capability does not exist in a particular power plant,
spray cooling can
be implemented by incorporating a humidification system and sacrificing some
energy
efficiency. That approach was used in the Limestone Injection Multiple Burner
(LIMB)
demonstrations funded by the U.S. Department of Energy and are of some value,
but that
technology has not been implemented, largely because of the poor reagent
utilization.
Another approach for addressing the ESP problems encountered when injecting
alkaline-earth-based chemicals includes the injection of various chemicals
into the lower
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-16-
temperature regions of the boiler, rather than into the higher temperature
regions. That
approach serves to modify the electrical characteristics of the ash, and is
disclosed in U.S.
Patent No. 4,306,885. It includes the injection of gases such as SO3 and NH3,
combinations thereof, and various ammonium and sodium phosphorus compounds.
The
addition of sodium compounds to the fuel is discouraged because most such
compounds
are relatively low melting and can cause serious slagging or fouling in the
high
temperatures regions of the boiler. Consequently, injection of such compounds
into the
flue gas stream at temperatures below about 900 C is specified. However, such
additions,
whether gaseous, liquid, or solid require investment in and maintenance and
operation of
separate feed equipment. The quantities of the various chemicals needed are
generally
quite small relative to the amount of ash or coal fired, i.e., in the range of
about 10 lb/ton
ash and about 1 lb per ton of coal, respectively, for a 10% ash coal.
Industry practice when injecting modest amount of chemicals, such as a
limestone
dispersion, is to dilute the product with water at the point of injection to
enhance distribution
in the flue gas stream. For limestone, that practice has the additional
advantage of
evaporative cooling and slightly delaying the calcination, which serves to
minimize
overburning of the CaO.
Ideally, the simplest mode of supplying the ash conditioning reagent is to
incorporate it in or to coinject it with the limestone dispersion at the
higher temperature.
There are a number of alkali compounds capable of modifying ash electrical
properties that
are high melting and suitable for incorporation in or co-injection with the
limestone
dispersion. They include sodium and potassium phosphate (tribasic), lithium
silicate, and
the aluminates of all three major alkali metals (Na, K, and Li). The
phosphates are
believed to provide an agglomerating effect downstream to complement the
resistivity
modifications that result from the introduction of the alkali metals.
Another way of supplying the needed alkali to condition the CaSO4 product of
the
capture of SOx by stone-derived lime is to employ alkali-containing
surfactants such as
sodium, potassium, or lithium polyacrylates as dispersants in the preparation
of the stone
slurry. The quantity of dispersant employed will vary with the solids loading
and particle
size of the suspension, but it should be in the range of from about 0.25% to
about 2.0% of
the product, by weight. The amount of alkali contributed to the CaS04
conditioning will vary
with the surfactant composition. It is likely to be sufficient to have a
conditioning effect, but
it can be supplemented with the other sources if more conditioner is needed.
The amount of the CaSO4 conditioning agent to be added is in the range of from
about 0.005% to about 5% of the weight of the injected stone. Because the
alkalis that are
the actual slagging agents represent less than 30% to 50% of the conditioning
reagent by
weight, the total deposit formers supplied is far less than the 3% or 4 %
sodium "rule of
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-17-
thumb" slagging threshold.for boilers. Treatment rates will be dependent on
the relative
quantities of sulfur and ash in the fuel, the stoichiometric efficiency of the
limestone, which
is a function of particle size, the effectiveness of the distribution upon
injection, and the
design and condition of the ESP. It should be noted that from a practical
operating
standpoint, the quantities of alkali added with the ash conditioning agents is
well below the
levels at which they actually pose slagging problems, and that the CaO
resulting from the
stone injection tends to mitigate any slagging problems.
A further beneficial result of utilizing the processes disclosed herein is
better carbon
burnout because of the resulting cloud of fine, reflective particles, which
also allow for
modest reductions in excess air, for reduced NOx formation, and for increased
unit heat
rate. A previous report by the inventor disclosed some such benefits from the
injection of a
coarser 3 micron Mg(OH)2 dispersion, but the use of the much finer size
particles and the
larger reflective cloud to restore the marketability of high LOI fly ash was
not envisioned.
The reduction of unburned carbon in the fly ash from as much as 16% to under
5%
restores the marketability of the ash.
Still another benefit of the present invention is a dramatically lower CO2
release per
unit of SO2 captured. The higher stone utilization of the present invention
brings the FSI
emission, 2.5 tons COZ per ton SO2, close to the 0.7 tons CO2 per ton SO2
achieved in a
wet limestone scrubber. That result compares with nearly 14 for the previous
furnace
injection approaches. Those numerical values take into account both the CO2
released
from the stone and that from the fuel used to calcine it. Clearly, to the
extent that finely-
ground fly ash can take the place of limestone, the CO2/SO2 ratio will be that
much closer
to that for limestone scrubbers.
Another difference between the invention as disclosed herein and previously-
disclosed approaches is the ability to provide a truly multipollutant FSI
process by grinding
the stone or ash in a weak ammonium solution, such as urea, along with a
surfactant, such
as ammonium polyacrylate or ammonium sulfonate, to facilitate high solids
loading. The
inherent stability of a fine-particle-size CaCO3 together with the reduced
settling afforded
by the high solids, overcomes the handling and pluggage problems that were
encountered
in previous attempts to combine -325 mesh stone and ammonium solutions. The
addition
of urea, even though in a small amount, could necessitate a substitution of
ammonium
polyacrylate for the sodium polyacrylate dispersant in order to assure a
stable, handleable
dispersion.
EXAMPLE
The following example is for a coal-fired boiler in an electrical power
generating
station, but the technology is applicable to the use of any fuel that results
in combustion
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-18-
products that contain undesirable sulfur oxide, nitrogen oxide, acids, or
toxic metal
compounds. Other types of combustion systems that can benefit from the
disclosed
processes include industrial boilers, trash incineration systems, residual
oil, and biomass-
fired combustion systems.
A 100 MW utility boiler burning 1.3% sulfur Midwestern coal purchases S02
credits
costing $575/ton to offset the emission of about 25 tons of S02/day. It
operates with flue
gas exit temperature of about 180 C to avoid SO3 corrosion of outlet duct and
ESP cold
end equipment. Station management is evaluating options for complying with
toxic metals
regulations. The station has tested furnace injection of powdered limestone
that has an
effective particle size of 25 microns at a Ca/S ratio of 2.4, but has been
able to capture only
50% of the SOa, even when humidifying the flue gas to boost performance. That
trial has
not resulted in commercial application because of the significant increase in
ash burden, of
the need for near continuous soot blowing to maintain heat transfer, and of
the adverse
impact on electrical properties of the ash, which resulted in the boiler being
out of
compliance on particulate emissions. During the NO,, generation season, the
station
employs a SNCR system and injects a urea solution, which effectively controls
the NO,,, but
the associated NH3 slip periodically fouls the air preheater, thereby
necessitating outages
every month or two to allow cleaning of the air preheater. Future regulations
will require
year-round NO, control, Hg emission control, and a further reduction in SO2
emissions.
A trial is then conducted utilizing a 75% CaCO3 aqueous dispersion that is
produced by wet grinding of limestone that is treated (by addition of a
surfactant?) to avoid
agglomeration of the reagent particles. The average -particle size is about 1
micron. The
dispersion is injected into the furnace through multiple injection ports that
are selected
using computer modeling in order to optimize the distribution of the
dispersion across the
combustion zone. The dispersion is at a Ca/S ratio of 1.15 and is injected
near the furnace
outlet, where the gas temperatures are within the range of from about 1000 C
to about
1100 C. The result is the capture of about 70% the SO2 with 50% less chemical
feed than
was required with the initial powdered stone with its larger, agglomerated
particles. That
result is accompanied by a virtual elimination of the SO3-related fouling and
corrosion
problems. Moreover, about 90% of the toxic metals are captured. Additionally,
about a
50% reduction in convective pass soot blowing is achieved because the ash that
is
produced is less sticky and more readily removed. The reduced ash quantity
relative to the
previous approaches allows the unit to meet stack emission regulations even
though some
modest back corona is still encountered in the first section of the
precipitator.
*********** ~*****
A boiler owner/operator might conclude that the approach disclosed in the
example
given above requires a relatively expensive reagent, because of the required
fine-grinding
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-19-
costs for the dispersions involved. However, the accompanying relatively-low
capital
investment, as well as the relatively short construction time required to
install boiler
injectors and system hardware in the minimal space required to retrofit this
new process at
a plant site, does provide an overall advantage as a multi-pollutant control
technology with
considerable operational flexibility. The technology benefits are
compared/evaluated to the
other emission-control system options available for a site-specific situation.
The new-art
technology is justified/selected based on availability and a life-cycle cost
that is
economically viable for the present as well as the future remaining life of
the generating
asset
On FGD-equipped systems where the capture of SO3 and toxic metals, rather than
SO2, are the primary objectives, magnesium chemicals are preferred because
they do not
compete for the SO2 pollutant. Similarly-sized, very fine fly ash particles
can be a better
option than, or a partial substitute for, magnesium compounds. For those
objectives, low
calcium fly ash or magnesium compounds can be injected after the economizers,
instead
of at the in-furnace transition from radiant to convective heat transfer near
what is
commonly referred to as the "nose" of the furnace. The "nose" is the region
adjacent to the
furnace outlet, where the heat transfer mechanism undergoes a transition from
radiant heat
transfer to convective heat transfer, as well as the superheaters and
reheaters that make
up the convective pass, and the economizer, where the incoming water is
preheated before
passing into the furnace heat exchange tubes for vaporization as steam. For
purposes of
this invention, the finely sized alkaline earth compounds are injected through
ports in the
upper furnace wall, generally close to the exit of the furnace exit.
The above are the preferred treatment options on systems that include
scrubbers,
because the total treatment reagent required is significantly less. However,
those
compounds do not operate to scavenge SO2 from the flue gas stream, which
instead is
carried out in the scrubber. If the fly ash has a significant calcium content,
which can be
present naturally or can be added, it can optionally be introduced into the
furnace
combustion zone to scavenge SOZ in systems that do not include scrubbers.
The preparation of suitable finely-sized fly ash particles is influenced by
the ash
chemistry, by the emissions control systems that are present, and by which
pollutants are
being addressed. If the plant has existing FGD capability and the objectives
are primarily
the control of SO3 and toxic metals, the preparation involves withdrawing a
small fraction of
the total fly ash, less than about 5%, from the collection hoppers, passing it
through a jet
mill or other type of mill to reduce the particle size of the ash to a micron
or less, and then
discharging the mill output back into the boiler or ducts. The solubilization
and
reprecipitation size reduction technique can also be applied to process the
high lime
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-20-
fraction of high calcium content ash, either with or without wet milling of
the insoluble
component.
The quantity of fly ash that is of reduced particle size can be small because
the
SO3 and toxic metals concentrations in the flue gas stream are in the low ppm
range, and
also because reducing a 20 micron particle to 0.5 micron size increases the
number of
particles by more than 61,000 and also multiplies the surface area by factor
of 40.
Because the increase in ash burden is small, the impact on the performance of
the
particulate control device will generally be manageable. Media mills (wet or
dry) are viable
alternatives to the jet mill, which can be powered by steam or air. In many
instances, a
steam-powered jet mill is preferable to a compressed air system because steam
is readily
available and is less expensive at a power plant, and also because the
moisture injected
would enhance acid adsorption and help ESP performance. Wet systems are
slightly more
cumbersome and could pose a very minor heat rate penalty.
The result of utilizing the finely-sized particles provided by the present
invention
include a dramatic increase in ash particles, and consequently in ash surface
area, thereby
providing enhanced adsorption of the SO3 and of toxic metals. That result is
achieved at a
low energy cost and at a relatively low capital investment, while
simultaneously eliminating
or significantly reducing the need for purchased chemicals. Increased fly ash
surface area
will not only enhance SO3 absorption, it will also stimulate more acid
neutralization as a
result of the alkaline materials present in the ash. In the few instances
where the adsorbed
acid inhibits constructive use of the ash, adding a modest limestone
supplement to the mill
input material would suffice.
Costs can be offset or converted into a net benefit because the capture of the
SO3
will reduce corrosion and fouling, and will make it feasible to lower the air
heater exit
temperature, thereby increasing fuel efficiency.
Grinding ash from the economizer hopper, or bottom ash, is preferable to
utilizing
ash from the precipitator hopper, because it avoids ash recirculation and
potential increase
of pollutant concentrations in the flue gas stream. For power plants without
high efficiency
FGD systems, the ground ash can be injected into the upper furnace region to
scavenge
SO2 if the ash has a significant CaO content. For a 1% Sulfur coal, ash having
a CaO
content of about 12.5% would provide the stoichiometric quantity that would be
needed.
Thus, many Midwestern coals with below stoichiometric CaO levels in the ash
can still be
beneficial, while Western coals with high CaO content could be more useful, if
they were
readily available.
Previous efforts at capturing undesirable pollutants have focused on adding
purchased chemicals that have or yield a high surface area needed to absorb
the SO3 and
toxic metals. In utilizing fly ash, the present invention, in contrast,
utilizes what is normally
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-21-
a waste product or low value byproduct of the power plant in order to reduce
the need to
purchase supplemental reagents.
Another aspect of pollution control of carbonaceous-fuel-fired combustion
systems,
primarily coal-fired facilities, is the matter of protection of the selective
catalytic reactors
(SCR's) from what has been referred to as "arsenic poisoning." Experience has
shown that
typical catalyst life in systems burning a wide range of coals is only about
three years. Not
only is catalyst replacement expensive (amounting to millions of dollars), but
NOx reduction
performance deteriorates steadily over that time period. Many coals,
particularly Eastern
U.S. coals, contain significant amounts of arsenic. The arsenic in the coal
forms gaseous
As203 as one of the products of combustion, which causes a reduction in the
activity of the
catalyst for its intended purpose of reducing NOX emissions
Because the replacement of SCR catalysts is an expensive undertaking, it is
therefore desirable to capture as much of the gaseous As203 as solid
particulates as is
possible before the arsenic-based vapor comes into contact with the catalyst,
and before it
is emitted to and pollutes the environment. Catalyst suppliers currently
recommend adding
sufficient limestone, either as rock or as powder, on the coal supply belt to
bring the CaO
content of the fly ash to at least 3%. Depending upon the mineralogy of the
ash, that can
involve a considerable limestone addition rate, a significant feed system
investment, and
significant ash disposal problems.
When alkaline earth carbonates, such as CaCO3 or MgCO3, are introduced into
the
furnace in the slurry form as described earlier herein to form the ultrafine
CaO and MgO
particles for capturing SO2, those particles are also useful for capturing the
As203. Other
compounds, such as those yielding ZnO upon thermal decomposition, can also be
effective, but are likely to be more costly. But when CaO is utilized as the
capture medium
a greater amount of reagent is required , because of the propensity of CaO to
also capture
the SO2, Thus, that portion of the CaO particles that effectively captures the
SO2 in the
combustion products by combining with the SOZ is therefore unavailable to
react with the
As203, resulting in a lower As2O3 capture effectiveness and a need to employ
higher
reagent dosages to protect the SCR catalyst than would be needed with MgO or
other high
surface adsorbents. Similarly, the "overbased" form of reagent can be much
more effective
than the ground or precipitated types in capturing the toxic metals that are
usually present
in the ppb range, because of the far greater number of particles per pound.
The choice of
reagent type, or combinations thereof, is an economic decision based on
whether the lower
dosage of the overbase is sufficient to offset its 4 or 5 times higher cost
per unit weight.
Unlike CaO, however, MgO does not readily capture SO2. Consequently, when
MgCO3 or Mg(OH)2 is introduced in slurry form to provide the fine MgO
particles, there is
little competition for the MgO between the SO2 and the toxic metals. As a
result, the
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-22-
quantity of reagent required to capture the arsenic is orders of magnitude
less with the
magnesium compounds. That is true even though the chemical has the additional
benefit
of reducing fouling and ESP problems. Further, much of the SO3 that is formed
in the SCR
will also be captured, which thereby serves to minimize the visible plume
problem often
incident to SCR/scrubber-equipped combustion systems.
In addition to the desirability of capturing the arsenic to minimize catalyst
poisoning,
it is also desirable to capture the toxic metal mercury. In that regard, the
mercury that is
present in the combustion products from Eastern coals is in the more readily
sorbed "salt "
form. In the Western coals, on the other hand, the mercury tends to be in the
more
difficult-to-capture elemental form. One way to capture the mercury present in
Western
coals, as well as in low-chloride-content Eastern coals, is to introduce with
the coal modest
amounts of chloride or of other oxidizing compounds, such as ozone or
peroxides, to
facilitate the mercury capture. The chloride addition allows the conversion of
elemental
mercury to a salt that can be captured in the dust collectors of the coal-
fired plants.
The introduction of chlorides or other oxidizing compounds is especially
desirable
when Western U.S. coals are utilized, because they have a lower intrinsic
chloride content.
To date, most applications have involved adding the chloride on the coal
supply belt. But
the additional chloride can be more effectively provided by adding it to the
carbonate
dispersion discussed earlier, which is injected into the combustion zone of
the furnace.
Chloride can be introduced to increase the equivalent effective chloride level
to
approximate that in Eastern coals, about 0.5% chloride by weight of coal, but
better
distribution of the chloride in the gas stream in the combustion zone should
achieve the
mercury oxidation with less chemical than is needed to be applied to the coal
supply belt.
The chloride can also be added separately at the point of the slurry injection
and in the
form of a salt solution. For example, at the equivalent of a MgC12 addition
rate of 500 ppm
on coal, the MgC12 would be only about 0.8% of the weight of the CaCO3 that
would be
needed to capture a significant amount of S02 at a Ca/S ratio of 1.5. Such a
small amount
could be sufficient to allow incorporation of the chloride salt in the
dispersion without
adversely impacting the stability of the carbonate dispersions. Including the
chloride salt in
the dispersion can also provide a modest degree of freeze protection.
On the other hand, the stability of the dispersion composition can be
adversely
impacted by the addition of chloride to the dispersion. One way to avoid a
dispersion
stability problem is not to incorporate the chloride in the dispersion
directly, but to provide
dual feeds to the injection ports, one feed being the carbonate dispersion and
the other
feed being the chloride. Another approach is to combine the carbonate
dispersion with
urea to be able to provide in a single solution a true multi-pollutant-
capturing product. The
combination will require a reformulation of the slurry to adjust for the
incompatability of the
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-23-
sodium-based dispersants and urea, and will necessitate a change from a sodium
polyacrylate dispersant to another dispersant, such as ammonium polyacrylate,
in order to
assure a stable dispersion.
It will therefore be apparent that the present invention provides significant
flexibility
and advantages in connection with the control of undesired SOx,, NOx, and
acids, as well
as toxic metals emissions, when it is utilized in carbonaceous-fuel-fired
combustion
systems.
A first advantage of the present invention is that the practice of the
disclosed
invention results in a significant reduction of the discharge of undesirable
pollutants. The
reduction is achieved with relatively simple modest cost equipment and is
effected within
the furnace by the reaction of the pollutants with a particulate material, or
the adsorption of
the pollutants on the surface of a particulate material. The disclosed
particulate materials
that achieve the control of undesired emissions can be derived from limestone,
magnesite,
alkaline-earth-containing wastes or byproducts, fly ash, and combinations
thereof. Those
materials are relatively inexpensive, and their preparation as finely-sized
dispersions in the
manner described herein avoids the high costs associated with ultra-fine, dry
grinding of
limestone. Moreover, the pollutant removal efficiency provided by a large
number of
unagglomerated, finely-sized particles is not dependent upon the existence of
particles
having pores, as in some previously disclosed approaches. And as earlier
noted, the pores
of such porous particles can easily become plugged, which thereby reduces the
effective
surface are of the particles and limits the access of the gaseous or vaporized
pollutants to
the inner core regions of the particles. Because of the resultant reduction of
available
effective particle surface area, the pollutant capture effectiveness of such
previous
processes is significantly reduced.
A second advantage of the disclosed invention is that it results in increased
combustion efficiency and reduced CO2 emissions from combustion systems
lacking
environmental controls. Consequently, the quantity of unburned carbon that is
otherwise
normally contained in the ash is reduced by the reflectivity of the cloud of
oxide particles
released by injection of the finely ground reagent dispersion. And the
combustion process
requires less excess combustion air, thereby reducing NOx emissions. It also
results in
minimizing ash fouling in the furnace, as well as reducing the deposition of
ammonium
bisulfite or sulfuric acid in the air preheater.
A third advantage of the disclosed invention is that it provides a low-capital-
cost
process that permits economical stripping of SOX and toxic metals pollutants
from
combustion gas streams. It also enables the capture of NOX along with the
other
pollutants. Furthermore, the resulting combustion gas stream cleanup can be
achieved
with fewer undesirable side effects, such as hard-to-handle ash deposits,
overburdened
CA 02628198 2008-05-01
WO 2007/053786 PCT/US2006/043219
-24-
dust collectors, visible stack plumes, and the like. And the hard piping and
safety devices
required when overbased compounds (oil-based magnesium compounds) are utilized
are
therefore rendered unnecessary.
In summary, the disclosed invention provides a new, more effective FSI
technology
that can be combined with other pollution control techniques as process
enhancements.
Although particular embodiments of the present invention have been illustrated
and
described, it will be apparent to those skilled in the art that various
changes and
modifications can be made without departing from the spirit of the present
invention. It is
therefore intended to encompass within the appended claims all such changes
and
modifications that fall within the scope of the present invention.
INDUSTRIAL APPLICABILITY
The present invention is applicable to the control of undesirable pollutant
materials
arising as a result of combustion in carbonaceous-fuel-fired combustion
systems, such as
electrical utility boilers, incinerators, and high temperature combustion
reactors.
