U.S. patent application number 11/078509 was filed with the patent office on 2006-09-14 for catalytic adsorbents for mercury removal from flue gas and methods of manufacture therefor.
Invention is credited to Chien-Chung Chao, Steve J. Pontonio.
Application Number | 20060205592 11/078509 |
Document ID | / |
Family ID | 36971150 |
Filed Date | 2006-09-14 |
United States Patent
Application |
20060205592 |
Kind Code |
A1 |
Chao; Chien-Chung ; et
al. |
September 14, 2006 |
Catalytic adsorbents for mercury removal from flue gas and methods
of manufacture therefor
Abstract
The present invention provides catalytic adsorbents formed from
doping activated carbon with a dispersed halide salt. The catalytic
adsorbents provided herein are stable and harmless at room
temperature, yet allow for chemical adsorption at elevated
temperatures typical of those for flue gas streams. The present
invention also provides methods of manufacturing the doped
activated carbon adsorbents.
Inventors: |
Chao; Chien-Chung;
(Williamsville, NY) ; Pontonio; Steve J.;
(Hamburg, NY) |
Correspondence
Address: |
PRAXAIR, INC.;LAW DEPARTMENT - M1 557
39 OLD RIDGEBURY ROAD
DANBURY
CT
06810-5113
US
|
Family ID: |
36971150 |
Appl. No.: |
11/078509 |
Filed: |
March 14, 2005 |
Current U.S.
Class: |
502/417 ;
423/210 |
Current CPC
Class: |
B01J 20/3236 20130101;
B01D 53/8665 20130101; B01J 20/046 20130101; B01J 20/20 20130101;
B01J 20/3204 20130101; B01D 53/64 20130101; B01D 2253/102 20130101;
B01D 2251/60 20130101; B01D 2257/602 20130101 |
Class at
Publication: |
502/417 ;
423/210 |
International
Class: |
C01B 31/08 20060101
C01B031/08 |
Claims
1. A catalytic adsorbent composition for removing mercury from a
flue gas stream at elevated temperatures, comprising: an activated
carbon having a halide salt dispersed thereon, the halide salt
having a cation and an anion.
2. The composition of claim 1, wherein the cation is selected from
the group comprising: an alkaline metal, an alkaline earth metal,
and a transition metal.
3. The composition of claim 2, wherein the cation is selected from
the group consisting of: Na, Mg, Ca, Cu and K.
4. The composition of claim 1, wherein the anion is selected from
the group comprising: bromide and chloride.
5. The composition of claim 1, wherein the halide salt is selected
from the group consisting of: NaCl, KCl, CaCl.sub.2, CuCl.sub.2,
CuBr.sub.2, NaBr, KBr, CaBr.sub.2, MgBr.sub.2 and mixtures
thereof.
6. The composition of claim 5, wherein the halide salt is NaBr, KBr
or mixtures thereof.
7. A method of making a catalytic adsorbent for use in the
adsorption of mercury from flue gas streams at elevated
temperatures, comprising: placing a powder activated carbon in an
aqueous solution containing a halide salt having a cation and an
anion to form a mixture; stirring the mixture until a homogeneous
slurry is formed; drying the powder activated carbon such that
water from the aqueous solution evaporates and the halide salt is
dispersed on the surface of the powder activated carbon.
8. The method of claim 7, wherein the cation is selected from the
group comprising: an alkaline metal, an alkaline earth metal, and a
transition metal.
9. The method of claim 8, wherein the cation is selected from the
group consisting of: Na, Mg, Ca, Cu and K.
10. The method of claim 7, wherein the anion is selected from the
group comprising: bromide and chloride.
11. The method of claim 7, wherein the halide salt is selected from
the group consisting of: NaCl, CaCl.sub.2, CuCl.sub.2, CuBr.sub.2,
NaBr, KBr, CaBr.sub.2, MgBr.sub.2 and mixtures thereof.
12. The method of claim 11, wherein the halide salt is NaBr, KBr or
mixtures thereof.
13. A method of making catalytic adsorbent for use in the
adsorption of mercury from flue gas streams at elevated
temperatures, comprising: injecting a presoaked carbonaceous
feedstock into a reaction chamber; and injecting at least one
oxidizing gas into the reaction chamber; injecting steam into the
reaction chamber, wherein the carbonaceous feedstock, the air and
the steam are injected into the reaction chamber under conditions
and for a residence time sufficient to form an activated carbon
having a halide salt having a cation and an anion dispersed on the
surface of the activated carbon.
14. The method of claim 13, wherein the oxidizing gas comprises:
air, oxygen, steam, nitrogen or combinations thereof.
15. The method of claim 13, wherein the reaction chamber is a tube
furnace.
16. The method of claim 13, wherein the reaction chamber is a
fluidized bed reactor.
17. The method of claim 13, wherein the cation is selected from the
group comprising: an alkaline metal, an alkaline earth metal, and a
transition metal.
18. The method of claim 17, wherein the cation is selected from the
group consisting of: Na, Ca, Cu and K.
19. The method of claim 13, wherein the anion is selected from the
group comprising: bromide and chloride.
20. The method of claim 13, wherein the halide salt is selected
from the group consisting of: NaCl, KCl, CaCl.sub.2, CuCl.sub.2,
CuBr.sub.2, NaBr, KBr, CaBr.sub.2, MgBr.sub.2 or mixtures
thereof.
21. The method of claim 19, wherein the halide salt is NaBr, KBr or
mixtures thereof.
22. A method for removing mercury from a gas stream at an elevated
temperature, the method comprising: injecting a catalytic adsorbent
containing an activated carbon and a dopant, the dopant having a
cation and an anion into the gas stream; adsorbing mercury onto the
catalytic adsorbent; and removing the mercury containing catalytic
adsorbent from the gas stream.
23. The method of claim 22, wherein the gas stream contains
oxidative gas, acidic gas or a combination thereof.
24. The method of claim 22, wherein the gas stream contains an
inert gas.
25. The method of claim 24, wherein the inert gas comprises
nitrogen.
Description
TECHNICAL FIELD
[0001] The present invention relates generally to catalytic
adsorbents for use in the removal of mercury from flue gas streams
and methods of manufacturing such catalytic adsorbents.
BACKGROUND OF THE INVENTION
[0002] The toxicity of mercury to humans and the environment has
long been known. It is known for example that mercury exposure can
cause neurological damage in humans. A particularly devastating
example of the harmful effects of mercury occurred in Minamata,
Japan in the 1950's where organic mercury byproducts of
acetaldehyde production were discharged into the local bay. The
byproducts were consumed and metabolized by fish. By consuming fish
in the bay, wide spread neurological damage and birth defects among
the local population were reported.
[0003] Coals used for generating electric power often contain about
0.1 ppm mercury. In the United States alone, about 50 tons of
mercury are discharged as vapor in stack gas every year. Through
chemical and biological processes, this mercury can become
concentrated in fish by many thousand fold, thereby entering human
food supplies at harmful levels.
[0004] The effort to remove trace mercury from air, water, natural
gas, and other industrial streams has a long history, however;
removing mercury from coal burning flue gas streams is a very
different problem.
[0005] Prior art techniques for removing mercury from air or
hydrocarbons at room temperature generally have limited relevance
to removing mercury from flue gas streams. Mercury has a high
atomic weight and adsorption temperature is a significant issue. At
room temperature, the dispersion interaction with carbon is
sufficient to immobilize mercury atoms. At about 300.degree. F.
(the temperature of many flue gas streams), however, physical
adsorption is no longer able to hold down the volatile elemental
mercury.
[0006] In addition, sufficient contact time with rapidly moving
flue gas streams is another issue for mercury removal. The total
time for flue gas, from generation by combustion to exit through
the stack, is often less than 10 seconds. Either as injected
powder, where adsorbent fly amid flue gas is for about 2 seconds,
or as filter cake on bags in a bag house, the contact time between
flue gas and activated carbon captured by the filter is less than
one second.
[0007] The demand on reactivity and reaction kinetics by flue gas
cleaning can not be properly tested by conventional packed beds.
Conventional packed beds are insufficient for flue gas cleaning
because the volume of flue gas is so large, the cost for
compressing it to push it through a packed bed is prohibitive.
[0008] Further issues relating to the removal of mercury from flue
gas include the small, yet potentially toxic, concentration levels
of mercury in the flue gas streams. The concentration of mercury in
flue gas streams is in .mu.g/m.sup.3 whereas the concentration of
mercury in many other industrial processes is on the order of
mg/m.sup.3. Much early work considered effluents containing mercury
in the 5 .mu.g/m.sup.3 range (that is not much lower than the
initial concentration of mercury in the flue gas) as fully
purified.
[0009] Above all, prior art techniques consider the adsorption of
mercury as an event between the adsorbent and the mercury. While
this is true in air or hydrocarbon streams at room temperature,
flue gas contains highly polar and reactive components that can
play both an interfering and enabling role for mercury removal. One
model composition used for flue gas contains about: 6% O.sub.2, 12%
CO.sub.2, 8% H.sub.2O, 1600 ppm SO.sub.2, 400 ppm NO, 50 ppm HCl,
20 ppm NO.sub.2, and 12 .mu.g/m.sup.3 elemental Hg.
[0010] Prior art attempts to remove mercury from flue gas have
included various techniques. One approach has focused on adding
halogen salts into coal prior to combustion such that the
combustion process generates hydrogen halide gases and then
injecting powder carbon downstream into the flue gas at a lower
temperature. Some mercury is captured by interaction between the
hydrogen halide gases, activated carbon and mercury. Another
approach has been to add hydrogen halides or elemental halogen
together with activated carbon to a lower temperature flue gas.
[0011] U.S. Pat. No. 1,984,164 to Karlsruhe proposes carbon or
silica gel or other adsorbents impregnated with elementary halogen
for removal of mercury from room air. Other prior art attempts have
included adding halide salts to coal before combustion since these
salts are known to be very stable. The combustion process oxidizes
halides to halogen and further reacts with hydrogen to yield
hydrogen halides. For example, U.S. Pat. No. 5,435,980 to Felsvang
et al. suggest adding chloride or a chlorine containing material
into the coal before or during combustion or adding HCl into flue
gas upstream of or in the drying-absorption zone.
[0012] U.S. Patent Application No. 2004/0003716 Al to Nelson, Jr.
discloses a method for removing mercury and mercury containing
compounds from combustion gas by injecting an adsorbent into the
flue stream. The sorbent is prepared by treating a carbonaceous
substrate with a bromine containing gas. Bromine gas is known to be
highly toxic by inhalation, ingestion or skin contact. HBr is also
known to be corrosive. In addition, bromine and HBr compounds are
reactive and can easily be added onto alkenes. Further, bromine is
reactive with aromatics.
[0013] U.S. Pat. No. 6,533,842 B1 to Maes et al. disclose powder
adsorbents which contain about 40% carbon, 40% calcium hydroxide,
10% cupric chloride and 10% KI.sub.3 impregnated carbon to remove
mercury from a high temperature, high moisture gas stream.
[0014] In December 2000, the United States Environmental Protection
Agency (EPA) made its regulatory decision that mercury emissions
from coal-fired electric generating plants need to be
controlled.
[0015] In the field of the mercury removal from flue gas streams,
it would therefore be desirable to provide adsorbents having
improved adsorbent characteristics in the flue gas temperature
range and that can be economically and efficiently
manufactured.
BRIEF SUMMARY OF THE INVENTION
[0016] The present invention provides catalytic adsorbents in which
a halide salt is dispersed on activated carbon and the oxidation
catalytic activity of the activated carbon promotes the formation
of mercury halide. At the same time, the adsorbent qualities of
activated carbon retain the mercury halides thus formed. The
present invention recognizes that while the halide salts are stable
and harmless at room temperature, these doped activated carbon
compounds form mercury halogen compounds at elevated temperatures
typical of those found in flue gas streams, and in the presence of
reactive components typical of flue gas. These mercury halogen
compounds are retained on the surface of the activated carbon.
Moreover, the increased adsorbent capacity and faster rate of
adsorption result in a need for smaller quantities of adsorbent
relative to an undoped activated carbon formed from the same
starting material.
[0017] A catalytic adsorbent composition for removal of mercury
from a flue gas stream thus includes an activated carbon having a
dopant (i.e, a halide salt) dispersed thereon. The cation of the
dopant used for the halide salt in accordance with the present
invention can be an alkaline, alkaline earth, or transition metal
(e.g., Na, Ca, Mg, Cu and K). The anion involved can be bromide or
chloride. Particularly preferred dopants include, but are not
limited to, NaCl, CaCl.sub.2, CuCl.sub.2, CuBr.sub.2, NaBr, KBr,
CaBr.sub.2 and MgBr.sub.2.
[0018] The halide salt is inert with respect to mercury and the
activated carbon at room temperature. At elevated temperatures
(e.g., 200-570.degree. F.) and in the presence of typical flue gas
compositions, mercury halogen compounds are formed and retained on
the activated carbon. While not intending to be bound by any
theory, it is believed that any or all of the following or a
combination of the following may occur. An oxidant (for example,
oxygen form the flue gas or oxidant on the activated carbon)
oxidizes the mercury and the anion of the dopant provides a counter
ion for the mercury ion as oxidized by the oxidant. Alternatively,
the oxidant oxidizes the anion in the salt and the oxidized anion
in turn oxidizes the mercury to form a mercury halogen compound on
the activated carbon. In addition or in the alternative, acidic
gases present in the flue gas react with the dopant salt to yield a
hydrogen halide. The hydrogen halide is then oxidized by an oxidant
and yields a halogen compound. The halogen compound then reacts
with the mercury to form a mercury halogen compound that are then
adsorbed by the activated carbon.
[0019] The present invention also provides methods of manufacturing
such doped activated carbon adsorbents that are both economical and
safe. The catalytic adsorbents of the present invention can be made
from a variety of methods. In one embodiment, the catalytic
adsorbents can be formed by placing an activated carbon in an
aqueous solution containing a halide salt to form a mixture,
stirring the mixture until a homogeneous slurry is formed and
drying the activated carbon such that water from the aqueous
solution evaporates and the halide salt is dispersed on the surface
of the activated carbon.
[0020] In another exemplary method of manufacture, the catalytic
adsorbents can be made by injecting a presoaked carbonaceous
feedstock into a reaction chamber together with oxidizing gases
such as air and/or steam. The carbonaceous feedstock and the
oxidizing gases are injected into the reaction chamber under
conditions and for a residence time sufficient to form a powder
activated carbon having a dopant dispersed on the surface of the
powder activated carbon. In this method, the reaction chamber can
be a batch type reactor such as a tube furnace or a reactor
designed for continuous mode operation (e.g., a fluidized bed
reactor). The dopant is formed of a cation selected from the group
including an alkaline metal, an alkaline earth metal, and a
transition metal (e.g, Na, K, Mg, Ca and Cu) while the anion is
selected from bromide and chloride. In some embodiments, the dopant
may be selected from the group including: NaCl, KCl, CaCl.sub.2,
CuCl.sub.2, CuBr.sub.2, NaBr, KBr, CaBr.sub.2 and MgBr.sub.2.
[0021] The catalytic adsorbents of the present invention are
suitable for use in the removal of mercury from a gas stream
containing an oxidant and/or acidic gases at an elevated
temperature such as a flue gas stream exiting a boiler or
combustion process. In this process, the catalytic adsorbents of
the present invention are injected into the flue gas stream for an
in-flight mode of mercury capture. As discussed above, the dopant
is inert with respect to the mercury at room temperature. At flue
gas temperatures and in the presence of the activated carbon,
oxidant and/or acidic gases, however, the dopant effectively
removes mercury from the flue gas stream. The mercury is retained
on the activated carbon in the form of mercury halogen compounds
and can be separated from the flue gas stream together with the
flyash.
BRIEF DESCRIPTION OF THE DRAWINGS
[0022] For a more complete understanding of the present invention
and the advantages thereof, reference should be made to the
following Detailed Description taken in conjunction with the
accompanying drawings in which:
[0023] FIG. 1 illustrates one embodiment for manufacturing
catalytic adsorbents in accordance with the present invention;
[0024] FIG. 2 illustrates a method of using the catalytic
adsorbents in accordance with the present invention;
[0025] FIGS. 3-6 illustrate graphs relating to Example 1;
[0026] FIGS. 7-12 illustrate graphs relating to Example 2;
[0027] FIG. 13 illustrates a graph relating to Example 3; and
[0028] FIGS. 14-20 illustrate graphs relating to Example 4.
[0029] Similar reference characters refer to similar parts
throughout the several views of the drawings.
DETAILED DESCRIPTION
[0030] The present invention provides catalytic adsorbents suitable
for use in the removal of mercury from flue gas streams at elevated
temperatures. The catalytic adsorbents of the present invention
include compositions having an activated carbon with a dopant
dispersed on the activated carbon. The dopant is a halide salt. The
cation of the dopant can be an alkaline, alkaline earth, or
transition metal while the anion of the dopant can be bromide or
chloride. The catalytic adsorbents of the present invention can be
formed from a variety of methods.
[0031] The present invention also provides methods of using these
compositions for mercury capture at elevated temperature in the
presence of acidic gases and/or oxidative gases that are commonly
found in flue gas streams generated by coal burning.
[0032] The mercury capture action is a synergistic combination of
components in the adsorbent compositions, the flue gas stream as
well as the flue gas stream temperature. Activated carbon doped
with bromide salts may be particularly preferred adsorbents as the
bromide salts appear to require less assistance from acidic and/or
oxidative gases in the flue gas stream and appear to be
particularly effective at removing mercury from the flue gas
stream.
[0033] As discussed hereinabove, alkaline, alkaline earth and
transition metal halides are harmless salts and inert to mercury
and activated carbon at room temperature. At about
200.degree.-570.degree. F. (e.g., 270.degree. F.), however, and in
the presence of acidic gases and/or oxidative gases of flue gas,
these doped activated carbon compositions are capable of capturing
mercury with high efficiency. Unused halide salts remain in their
salt form.
[0034] The catalytic adsorbents of the present invention also
perform well in flue gas streams generated by burning low chloride
coal (e.g., Powder River Basin (PRB) coal from Wyoming) where
current adsorbents such as FGD carbon do not function
efficiently.
[0035] The present invention thus provides for halide salts to be
dispersed on activated carbon such that the salts retain their
chemical inertness at room temperature, but react with mercury in
hot flue gas to yield non volatile mercury halide. More
particularly, at temperatures in the range of about 200-570.degree.
F., and in the presence of acidic and/or oxidative gas from the
flue gas, halide salts react with mercury and assist the activated
carbon to capture the mercury, which is present in very low
concentrations in flue gas streams. The catalytic adsorbents of the
present invention utilize the very fast kinetics at elevated
temperatures to optimize both physical adsorption as well as
chemical adsorption. The reactivity of the halide salts as used
herein is thus a cooperative phenomenon.
[0036] As discussed hereinbelow, the catalytic adsorbents of the
present invention can be made from a variety of methods. The
adsorbents can be made from commercially available powdered
activated carbon (PAC) or from raw carbonaceous material. Exemplary
PACs suitable for use in the invention include, but are not limited
to, FGD (available from Norit America, Inc.), ashless activated
carbon powder made from purified petroleum coke and carbon fiber
powder made by carbonization of rayon fiber. It will be appreciated
that other activated carbons can also be used in the present
invention.
[0037] The catalytic adsorbents of the present invention can be
made from various techniques. In one embodiment of the invention,
the adsorbents can be manufactured by soaking activated carbon in
an aqueous solution of halide salts. This approach is an economical
and safe process relative to treating activated carbon with
hydrogen halides or halogen gases.
[0038] In this embodiment, the minimum amount of water necessary to
make a solution of the salt is utilized. The cation of the dopant
can be an alkaline, alkaline earth, or transition metal. The anion
involved can be bromide or chloride. Suitable salts for use in the
invention therefore include, but are not limited to, NaCl,
CaCl.sub.2, CuCl.sub.2, CuBr.sub.2, NaBr, KBr, CaBr.sub.2 and
MgBr.sub.2. In some embodiments, KBr, NaBr or CaBr.sub.2 may be
preferred and in some embodiments, NaBr or KBr may be the most
preferred salt.
[0039] The PAC, preferably in powder form, is placed in the aqueous
solution and the mixture is stirred until it becomes a homogeneous
slurry and such that there is sufficient contact time between the
salt solution and PAC that the salt solution becomes dispersed on
the PAC. It will be appreciated by those skilled in the art that
the PAC has porosity such that the solution and hence the halide
salt will disperse into the PAC.
[0040] In this approach, the amount of salt necessary for the
aqueous solution is determined based on the amount of PAC and the
ratio of the salt to PAC that is desired for a particular adsorbent
(i.e., the dopant level in the desired PAC determines the
concentration of the salt solution). In some embodiments, the ratio
of the dopant level to that of the PAC is 1:10,000 to 30:100. In
more preferred embodiments, the ratio of dopant to PAC is 1:1000 to
10:100 and in other embodiments, the ratio of dopant to PAC is
0.5:100 to 7:100.
[0041] The salt solution containing the PAC is allowed to soak and
then allowed to sufficiently dry such that the PAC is free flowing.
During this time, the water evaporates and the salt enters the pore
volume of the PAC and becomes dispersed on the surface of the PAC.
After the PAC is dried, it is in powder form. It may be ground and
passed through an appropriate size desired mesh. While not to be
construed as limiting, the PAC may be passed through a 200 mesh. In
this manner, the PAC can be used for mercury removal at less than
or equal to a 200 mesh material. It will be appreciated by those
skilled in the art that the adsorbent can be treated for
appropriate size depending on the intended use of the adsorbent.
For example, smaller mesh (e.g., 400 mesh) may be desirable in some
applications.
[0042] It is believed that the catalytic adsorbents of the present
invention will perform well for mercury removal from flue gas
streams at elevated temperatures given the dispersed salts on the
surface of the PAC. While not intending to be bound by any theory,
it is believed that the salt is inert with respect to elemental
mercury at room and high (i.e. in the range of combustion zone)
temperatures. At elevated temperatures of about 200-570 F (for
example, at about 270-300 F), however, and in the presence of
oxidative and/or acidic gases in the flue gas, and the doped
activated carbon, mercury in the flue gas stream can be oxidized
and effectively removed therefrom.
[0043] An alternative method to soaking a PAC in an aqueous
solution as described above is to spray water droplets containing
the desired halide salt on the PAC in a manner such that the halide
salts become dispersed as discussed above. Such an approach can be
used in connection with the activated char produced in commonly
owned U.S. patent application Ser. No. ______ entitled "Production
of Activated Char Using Hot Gas" to Bool et al., filed on even date
herewith. The entire contents of U.S. patent application Ser. No.
______ are incorporated herein by reference.
[0044] An alternative method for manufacturing catalytic adsorbents
suitable for use in the present invention is shown in FIG. 1. In
this embodiment, the catalytic adsorbents can be manufactured by
presoaking a prepulvurized carbonaceous feedstock in an aqueous
solution of an alkaline, alkaline earth or transition halide salt.
Alternatively, the prepulverized carbonaceous feedstock may be
soaked in an alcohol (e.g., ethanol) solution containing the
alkaline, alkaline earth or transition halide salt. The presoaked
feedstock in then exposed to an oxidizing gas mixture such as air
and steam at an elevated temperature in a reaction chamber to
produce catalytic adsorbents and an exhaust gas.
[0045] The final concentration of the catalytic adsorbent is
determined as in the prior embodiment (i.e. the ratio of the dopant
to activated carbon is predetermined in order to determine the
concentration of the salt solution), except that in this
embodiment, the loss of carbon due to combustion in the reaction
chamber must be taken into account. One can therefore determine the
concentration based on the yield of the final product to account
for the loss of carbon due to combustion.
[0046] As illustrated in FIG. 1, carbonaceous feedstock 16 is
injected into reaction chamber 10. In some embodiments,
carbonaceous feedstock 16 is not yet activated and can be selected
from various types of feedstock such as coal or biomass materials.
The feedstock can be prepulverized to an appropriate size, for
example from about 5-200 microns.
[0047] The carbonaceous feedstock 16 is also presoaked as discussed
above prior to injection into reaction chamber 10 with a solution
containing the desired halide salt. In this embodiment, the
solution can be formed from water or ethanol, although water may be
preferred.
[0048] Oxidizing gases 12 and 14 (e.g., air 12 and steam 14) are
injected into reaction chamber 10 simultaneous with or nearly
simultaneous with carbonaceous feedstock 16. Preferably, the steam
is preheated and is injected at a temperature of about 1800 F.
[0049] Reaction chamber 10 may be selected from a variety of
reactors such as single batch reactors where the feedstock is
suspended on a filter media and reactant gases pass through the
feedstock (e.g, a tube furnace) or continuous reactors whereupon
the gas temperature, composition and feedstock residence time can
be controlled for optimal conditions (e.g., a fluidized bed
reactor). One type of a continuous process reactor may be a Plow
Mixer, available from Scott Equipment Company.
[0050] Heat for reaction chamber 10 can be provided by from various
sources, for example, the reaction chamber can be electrically
heated or heated by a flame. Alternatively or in addition to such
heat, reaction chamber 10 may be heated from the temperature of the
feedstock and/or steam. It will be appreciated by those skilled in
the art that the desired temperature within the reaction chamber
depends on several factors, including temperature of the air and/or
oxidizing gases, amount of oxygen, stoichiometric ratio of oxygen
to feedstock and/or temperature of the feedstock. The heat may be
provided from any source so long as it is sufficient to generate
flue gas 18 and adsorbent 19. Typically, the temperature within the
furnace will be between about 1450-2700.degree. F., and more
preferably between about 1650-2200.degree. F. When the
stochiometric ratio of oxygen to feedstock is greater than one, the
contact time between the oxidizing gas and the feedstock becomes
more significant because more of the feedstock will be consumed and
therefore impact product yield. When the stoichiometric ratio is
less than one, the contact time will be less critical.
[0051] The residence time of the carbonaceous feedstock 16,
reactive oxidizing gases (such as air 12 and steam 14) within
reaction chamber 10 is long enough such that flue gas 18 and
adsorbent 19 are generated within chamber 10. The residence time of
the carbon is independent of the gas and can be independently
controlled. This can be significant because sufficient time is
necessary to devolatilize and partially oxidize the feedstock.
While the residence time is short, it is important that it be long
enough to adequately activate the carbon. In some embodiments, the
residence time may be on the order of minutes. It will be
appreciated that if the residence time is too long or there is too
much oxygen or steam, adsorbent yield will be negatively
impacted.
[0052] Adsorbent 19 is removed from reaction chamber 10 and is
ready for use as a mercury removal adsorbent from flue gas streams
at elevated temperatures. Flue gas 18 typically includes combustion
gases such as CO.sub.2, CO, N.sub.2 and H.sub.2O. Any unreacted,
partially combusted (e.g., CO) or volatile gases in gas stream 18
can be further combusted.
[0053] Yet another alternative embodiment for manufacturing
catalytic adsorbents for use in accordance with the present
invention can be found in commonly owned U.S. patent application
Ser. No. ______, entitled "Production of Activated Char Using Hot
Gas" to Bool et al., filed on even date herewith. The entire
contents of U.S. patent application Ser. No. ______ are
incorporated herein by reference.
[0054] In this embodiment, the feedstock is presoaked with an
aqueous or ethanol solution as discussed above. The presoaked
feedstock is then treated to produce activated char as discussed in
commonly owned U.S. patent application Ser. No. ______, entitled
"Production of Activated Char Using Hot Gas".
[0055] Catalytic adsorbents of the present invention can also be
formed by dry mixing a prepulverized raw carbonaceous material with
a halide salt powder. In this embodiment, the raw carbonaceous
material and halide salt powder are mixed together in dry form. The
mixture can then be injected and processed as discussed hereinabove
with regard to FIG. 1 or as shown in commonly owned U.S. patent
application Ser. No. ______, entitled "Production of Activated Char
Using Hot Gas". The temperature within the reaction zone will be at
or above the melting point of the halide salt such that the halide
salt melts and wets the surface of the carbonaceous material.
Consequently, the salt can be dispersed in the carbonaceous
material.
[0056] Referring now to FIG. 2, an exemplary system for using the
catalytic adsorbents of the present invention is shown. Flue gas 22
is formed as a result of combustion in a furnace or boiler 20.
While flue gas 22 can vary in composition and temperature, a
typical composition can include: 6% O.sub.2, 12% CO.sub.2, 8%
H.sub.2O, 1600 ppm SO.sub.2, 400 ppm NO, 50 ppm HCl, 20 ppm
NO.sub.2, and 12 .mu.g/m.sup.3 elemental Hg and can be in the
temperature range of about 200-570 F. Catalytic adsorbent 30a,
which can be formed from any of the methods described hereinabove,
can be injected upstream of particulate collection device (PCD) 24.
Particulate collection device 24 is typically a baghouse or
electrostatic precipitators (ESPs). Adsorbent 30a is injected into
flue gas stream 22 upstream of PCD 24 such that there is sufficient
residence time for the catalytic adsorbent to capture and remove
mercury from flue gas 22.
[0057] Particulates and adsorbent containing mercury are removed
from PCD 24 by stream 28. Flue gas 26 thus contains less mercury
than flue gas 22 and may be sent to the stack.
[0058] In some embodiments, it may be desirable to inject the
catalytic adsorbent into the flue gas downstream of the PCD. Such
processes are currently being investigated by others.
[0059] As discussed above, it is believed that the catalytic
adsorbents of the present invention will perform well for mercury
removal from flue gas streams at elevated temperatures given the
dispersed salts on the surface of the PAC.
EXAMPLES
[0060] As will be seen hereinbelow, physical adsorption of PAC at
about 270.degree. F. is not sufficient to retain elementary mercury
without HCl as a promoter. In contrast, doped PAC function well
without HCl; however, the presence of HCl, O.sub.2 and/or SO.sub.2
function as promoters for a doped PAC.
[0061] In some examples, doped PAC were prepared by treating three
types of commercially available PAC. In other examples, doped PAC
was prepared by activation of halide salt treated coal.
[0062] The first commercial PAC used is FGD carbon, available from
Norit America, Inc. It is made from lignite coal and contains about
30 weight percent ashes. In powder form, it is widely tested and
accepted as a bench mark for activated carbon for mercury removal
from flue gas. The second PAC was ashless activated carbon
available from Carbon Resource, Inc. It is typically made from
purified petroleum pitch and contains a trace amount of ash. It is
generally sold in bead form. For mercury removal in the following
examples, it was ground, sieved and the -400 mesh portion was used.
The third PAC that was used was activated carbon fiber
ACF-1300/200, also available from Carbon Resources, Inc. It is made
from rayon and typically received in cloth form. This material was
ground and screened through 400 mesh sieve before use.
[0063] To prepare halide salt doped PAC from coal, coal was soaked
in an aqueous or ethanol halide salt solution. The doped coals were
activated in a stream of oxygen, nitrogen and steam in temperature
range of about 1800.degree. F.
[0064] Two tests were used to evaluate the adsorbents: a fixed bed
test and a residence chamber test. In the fixed bed test, the fixed
bed consisted of 150 mg adsorbent supported on a quartz filter of
about 63.5 mm in diameter. The details of the test setup are
described in papers published by EERC, as published for example at
the Mercury Control Technology R&D Program Review Meeting on
Aug. 12-13, 2003 at Pittsburgh, Pa. Gas streams containing mercury
as well as components of flue gas were passed through the thin bed.
The break through of mercury was monitored and spent adsorption
beds were collected and analyzed.
[0065] In the residence time chamber test, a slip stream from a
power plant at Pleasant Prairie, Wis. was made to pass through
chambers of different length. Adsorbent was injected at one end of
the chamber to flight with the flue gas stream. At the other end of
the chamber, the adsorbent was separated from the flue gas stream
and the cleaned flue gas was analyzed for Hg content to determine
the efficiency of the adsorbent. The chamber length was used to
determine the contact time between the flue gas and the adsorbent.
Details of the residence time chamber apparatus (designed by the
Electric Power Research Institute (EPRI)) can be found in published
papers (see e.g., "Assessment of Low Cost Novel Sorbents for Coal
Fired Power Plant Mercury Control", Combined Power Plant Air
Pollutant Control Mega Symposium (Washington, D.C., Aug. 30-Sep. 2,
2004).
Example 1
[0066] This example demonstrates that at room temperature, undoped
PAC is a good adsorbent for elemental mercury and a promoter would
appear to provide no additional benefit. At 270.degree. F.,
however, physical adsorption is overwhelmed by kinetic energy and
adsorption by undoped PAC and without a promoter was
inadequate.
[0067] Fixed bed tests were conducted on four samples in a stream
which contained nitrogen and about 13 .mu.g/m.sup.3 of elemental
mercury. The tests conditions and results are summarized in Table 1
and FIGS. 3-6. The undoped FGD carbon sample was tested at room
temperature and achieved 100% mercury removal for more than 15
hours with no sign of mercury breakthrough. For samples tested at
270.degree. F., all three types of activated carbon reached almost
100% breakthrough immediately (0% removal). TABLE-US-00001 TABLE 1
Test gases Sample Sample Test composition and Comments on Test
Sample # Name treatment Temp sequence Results FGD As received 72 F.
N.sub.2 + Hg 100% Hg removal for carbon 15 hrs; no any sign of
breakthrough 17297-89 FGD Vacuum 270 F. N.sub.2 + Hg Breakthrough
carbon activated at occurred immediately 1100 F. 17343-13 Carbon 6
N HCl 270 F. N.sub.2 + Hg Breakthrough fiber extraction and
occurred heating at 1800 F. immediately. in N.sub.2 17297-99
Ashless 6 N HCl 270 F. N.sub.2 + Hg Breakthrough carbon extraction
and occurred immediately heating at 1800 F. in N.sub.2
[0068] 6 N HCl extraction was used in Sample Numbers 17343-13 and
17297-99 to remove any trace ashes. Heating in N.sub.2 at
1800.degree. F. is intended to remove oxidizing species on the
commercially obtained PAC. Neither treatment changed the adsorption
behavior of the PAC.
Example 2
[0069] This example demonstrates how halide salts as a dopant alter
the flue gas, mercury and carbon interaction so as to promote
mercury adsorption from the flue gas stream. In this Example, thin
fixed beds of PAC samples were exposed to different gas mixtures in
sequence. All experiments started with nitrogen and mercury (about
13.mu. gm/cubic meter). Other components of the flue were added
into the stream sequentially or in sequential combination toward a
composition of synthetic flue gas, which is typified as: 6%
O.sub.2, 12% CO.sub.2, 8% H.sub.2O, 1600 ppm SO.sub.2, 400 ppm NO,
20 ppm NO.sub.2, 50 ppm HCl, 12-14 .mu.g/m.sup.3 Hg, with the
balance being N.sub.2.
[0070] Two type of PAC (ashless and FGD) and three dopants (KBr,
NaBr, and NaCl) were used in the experiments. Detail of the
experiments are summarized in Table 2. The breakthrough curves are
given in FIGS. 7-12. TABLE-US-00002 TABLE 2 Test gases Sample
Sample composition and Comments on Sample # Name Description
sequence Test Results 17297-99 Ashless 6 N HCl extraction 1.
N.sub.2 + Hg; Removed Hg carbon and heating at 2. 6% O.sub.2 + 8%
H.sub.2O only after HCl 1800 F in N.sub.2 added; was added to the
3. 1600 ppm SO2 test gas added; 4. 50 ppm HCl added 17297-99
Ashless 6 N HCl extraction 1. N.sub.2+; Hg; HCl promoted Hg carbon
and heating at 2. 50 ppm HCl adsorption, SO.sub.2 1800 F in N.sub.2
added; caused decline of 3. 1600 ppm SO.sub.2 Hg removal added
17343-15 KBr doped 15:100 ratio of 1. N.sub.2 + Hg; Adsorbed Hg in
Ashless KBr:Carbon 2. 6% O.sub.2 added; N.sub.2 stream. Both carbon
3. 8% H.sub.2O added; O.sub.2 and SO.sub.2 4. 1600 ppm SO.sub.2
promoted Hg added removal 17297-89 FGD carbon Vacuum activated 1.
N.sub.2 + Hg; Removed Hg at 1100.degree. F. 2. 8% H.sub.2O + 50 ppm
only after HCl HCl added; was added to the 3. 6% O.sub.2 added;
test gas 4. Full Flue added 17297-93 NaBr doped 15:100 ratio of 1.
N.sub.2 + Hg; Adsorbed Hg in FGD NaBr:Carbon 2. 8% H.sub.2O + 12%
N.sub.2 stream. Both CO.sub.2 + 6% O.sub.2 O.sub.2 and SO.sub.2
added; promoted Hg 3. 1600 ppm SO.sub.2 removal added; 4. Full Flue
added 17297-91 NaCl doped 15:100 ratio of 1. N.sub.2 + Hg; CO.sub.2
and SO.sub.2 FGD NaCl:Carbon 2. +8% H.sub.2O + 12% CO2 + 6% were
weak O.sub.2 added; promoters. The 3. 1600 ppm SO.sub.2 presence of
HCl added; was important for 4. Full Flue added; Hg removal 5.
--HCl added
Example 3
[0071] This example demonstrates that a physical adsorbent such as
silica gel, doped with KBr, did not remove mercury from the flue
gas.
[0072] The same thin fixed bed method as in Examples 1 and 2 was
used in this Example. The details of sample preparation, test
conditions and results are given in Table 3 and FIG. 13.
TABLE-US-00003 TABLE 3 Test gases Comments Sample Sample
composition and on Test Sample # Name Description sequence Results
17297-69 KBr The weight 1. O.sub.2 6% + CO.sub.2 The con- doped
ratio of 12% + H.sub.2O 8% + SO.sub.2 centration of silica
KBr:Silica 1600 ppm + NO Hg reduction gel gel = 400 ppm + HCl was
less 15:100 50 ppm + NO.sub.2 than 10%. 20 ppm + Hg 14 micro gram +
N.sub.2 (full synthetic flue)
Example 4
[0073] This example analyzed the effectiveness of various halide
salts as dopants. Doped ashless carbons were tested by thin fixed
bed methods as in Examples 1-3 in synthetic flue. The results are
compared with undoped FGD.
[0074] The thin fixed bed test is to simulate the function of a bag
house in a power plant. The efficiency of adsorbent is analyzed by
the percent of mercury removal from the flue gas. All doped samples
reached higher mercury removal than undoped FGD carbon. The results
are given below in Table 4 below. The breakthrough curves are shown
in FIGS. 14-20. TABLE-US-00004 TABLE 4 Test gases Sample
composition and Comments on Sample # Sample Name Description
sequence Test Results 17343- KCl doped 15:100 ratio of 1. O.sub.2
6% + CO.sub.2 12% + H.sub.2O Best removal at 02D ashless carbon
KCl:Carbon 8% + SO.sub.2 about 95% level 1600 ppm + NO 400 ppm +
HCl 50 ppm + NO.sub.2 20 ppm + Hg 14 .mu.g/m.sup.3 + N.sub.2 (full
flue) 17343- NaCl doped 15:100 ratio of 1. O.sub.2 6% + CO.sub.2
12% + H.sub.2O Best removal at 01C ashless carbon NaCl:Carbon 8% +
SO2 about 93% level 1600 ppm + NO 400 ppm + HCl 50 ppm + NO.sub.2
20 ppm + Hg 14 .mu.g/m.sup.3 + N.sub.2 (full flue) 17343- NaBr
doped 15:100 ratio of 1. O.sub.2 6% + CO.sub.2 Best removal at 02A
ashless carbon NaBr:Carbon 12% + H.sub.2O 8% + SO2 about 98% level
1600 ppm + NO 400 ppm + HCl 50 ppm + NO.sub.2 20 ppm + Hg 14
.mu.g/m.sup.3 + N.sub.2 (full flue) 17297- KBr:Carbon = 15:100
15:100 ratio of 1. O.sub.2 6% + CO.sub.2 Best removal at 75A
KBr:Carbon 12% + H.sub.2O 8% + SO2 about 100% level 1600 ppm + NO
400 ppm + HCl 50 ppm + NO.sub.2 20 ppm + Hg 14 .mu.g/m.sup.3 +
N.sub.2 (full flue) 17343- CaBr.sub.2 doped 15:100 ratio of 1.
O.sub.2 6% + CO.sub.2 Best removal at 02B ashless carbon
CaBr.sub.2:Carbon 12% + H.sub.2O 8% + SO2 about 100% level 1600 ppm
+ NO 400 ppm + HCl 50 ppm + NO.sub.2 20 ppm + Hg 14 .mu.g/m.sup.3 +
N.sub.2 (full flue) 17343- MgBr.sub.2 doped 15:100 ratio of 1.
O.sub.2 6% + CO.sub.2 Best removal at 02C ashless carbon
MgBr.sub.2:Carbon 12% + H.sub.2O 8% + SO2 about 95% level 1600 ppm
+ NO 400 ppm + HCl 50 ppm + NO.sub.2 20 ppm + Hg 14 .mu.g/m.sup.3 +
N.sub.2 (full flue) 17297-62 FGD carbon From Norit 1. O.sub.2 6% +
CO.sub.2 Best removal at Amercia, a 12% + H.sub.2O 8% + SO2 about
90% level reference 1600 ppm + NO 400 ppm + HCl 50 ppm + NO.sub.2
20 ppm + Hg 14 .mu.g/m.sup.3 + N.sub.2 (full flue)
Example 5
[0075] This example used a residence time chamber test to
demonstrate the effectiveness of bromide salt doped PAC in an "in
flight adsorption" and the quality of PAC made by direct activation
of bromide salt doped coal.
[0076] Residence time chamber test. The residence time chamber used
in this Example was an EPRI 8-inch diameter tube setup as discussed
above. A slip stream of 30 acfm flue was taken out from a coal
burning boiler duct for flow through this tube. Adsorbent is
injected at one end of the tube. At each of the middle section and
exit end of this tube, there are one outlet sampling tubes to allow
measurement at two different residence times. The mercury
concentrations were measured at the inlet as well as the sampling
outlets to determine the mercury removal efficiency of the
adsorbents.
[0077] The residence time chamber simulates the situation of a
plant which has only an electrostatic precipitator (ESP), therefore
mercury removal depends on inflight adsorption. Typical inflight
time is about 2 seconds. In the example, the sampling outlets allow
about 2 and 4 seconds of residence time. Three groups of adsorbents
were tested. The first group of samples were prepared by doping FGD
PAC with an aqueous bromide salt solution. The second group of
samples were prepared by activation of halide salt doped coal in a
tube furnace at 1650.degree. F. to 2000.degree. F. in a stream
containing, oxygen, nitrogen and water. The third group of samples
were prepared by activation of halide salt doped coal by a burner
as in commonly owned U.S. patent application Ser. No. ______,
entitled "Production of Activated Char Using Hot Gas" to Bool et
al., filed on even date herewith, with or without further steam
activation at 1800.degree. F.
[0078] Undoped FGD carbon samples were also tested to serve a as
reference. The test results are given in Table 5. The percentage of
Hg removal is calculated by dividing outlet mercury concentration
with the inlet mercury concentration. Since there is no way to
determine how much mercury is removed by the cylinder wall, the
reported number is the sum of inflight removal plus removal by wall
effect. TABLE-US-00005 TABLE 5 Injection Outlet % Hg % Hg Sample
Sample rate Temp Inlet Hg Hg Removal Removal Sample # Name
Description lb/mmacf (.degree. F.) .mu.g/Nm.sup.3 2/4 sec (2 sec)
(4 sec) 17297- FGD/ FGD 5.8 300 9 1.75/0.9 81 90 22 KBr carbon
doped with 7:100 ratio of KBr:FGD 17297- FGD/ FGD 5.7 300 8.7
1.3/0.7 85 92 23 KBr/ carbon CuBr2 with 6:1:100 ratio of
KBr:CuBr.sub.2:FGD FGD No 6 300 6.4 3.3/2.8 48 56 doping 17343-
Activated 7:100 6 300 10.3 3.0/2.0 71 80 76 PRB coal ratio of
predoped CaBr.sub.2:Coal. with Activated CaBr2 in tube furnace
17343- Activated 7:100 6 300 9.5 2.1/1.5 78 84 77 PRB coal ratio of
predoped NaBr:coal. with Activated NaBr in tube furnace 17343-
Activated 5:100 6 300 9.9 2.8/1.8 72 82 83B PRB coal ratio of
predoped (CaBr.sub.2 with 1/2H.sub.2O):coal. CaBr2 Activated In
tube furnace 78B Activated 7:100 6 300 9.8 3.5/2.2 64 78 PRB coal
ratio of predoped KBr:coal. with KBr Activated in burner 42A-15-
Activated 7:100 6 300 9 1.7/1.2 81 87 1000 PRB coal ratio of
predoped KBr:coal. with KBr Activated in burner, then steamed at
1800 F. (15 min)
Example 6
[0079] The chemical form of dopant in PAC. Activation of bromide
salt doped coal is at a temperature close to 1800.degree. F. This
raises the question whether the bromide salt retains its ionic
form. Chemical analyses of bromide salt doped coal before and after
activation are shown in Table 6. Bromide salt maintains its inert
ionic form. This may be particularly advantageous because
bromination of carbon can create unknown and undesirable organic
bromide compounds. It is therefore desirable to avoid the formation
of such compounds. TABLE-US-00006 TABLE 6 Ionic Total Sample Sample
bromine bromine Na Sample # Name Description mmol/gm mmol/gm
mmol/gm 17343- NaBr NaBr:PRB coal = 5:100 0.43 0.48 0.17 85A doped
PRB Before coal activation 17343- 17343-85A tube furnace 0.65 0.70
0.26 85B after at 1800 F, activation purged with 10% O.sub.2, 90%
N.sub.2 saturated with water vapor at 194 F. 17343- NaBr NaBr:PRB
coal = 1:100 0.09 0.08 0.03 88A doped PRB before coal activation
17343- 17343-88A tube furnace 0.11 0.12 0.05 88B after at 1800 F,
activation purged with 10% O.sub.2, 90% N.sub.2 saturated with
water vapor at 194 F.
[0080] It should be appreciated by those skilled in the art that
the specific embodiments disclosed above may be readily utilized as
a basis for modifying or designing other structures for carrying
out the same purposes of the present invention. It should also be
realized by those skilled in the art that such equivalent
constructions do not depart from the spirit and scope of the
invention as set forth in the appended claims.
* * * * *