Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
BACKGROUND OF THE INVENTION
This invention relates to a process for reducing sulfur
dioxide. More particularly~ this invention relates to a process
for reducing the sulfur dioxide content of gases by reaction with
coal.
Appreciable amounts of sulfur dioxide are contained in
many industrial gases vented in~o the atmosphere from plants
involved in roasting, smelting and sintering sulfide ores, in stack
gases from power plants burning sulfur-bearing coal, or in exit
~` 10 gases from other industrial operations involving the combustion of
sulfur-bearing fuels, such as fuel oilO Removal of the sulfur
dioxide from these off ~ases is important for two reasons, first,
the sulfur dioxide is an environmental pollutant and secondly, the
sulfur dioxide emitted into the atmosphere results in the loss of
sulfur values, a natural resource. Sulfur dioxide which is
res:ov~red from these off gases can be converted to sulfur.
There are, of course, many processes for reducing the
sulfur dioxide content o gases, most of which involve the use of
reclucing agents such as natural gas or other gaseous reductants
~uch as methane, carbon monoxide, liqui.d petroleum gases and the
like. The supply of these gaseous reductants is limited and
therefore their use becomes more and more expensive, when and if
they are available.
:.
It is known that coal can be a reductant for sul~ur
dioxide resulting in the production of elemental sulfur and rela-
tively harmless carbon dioxide~ Coal is also in abundance and,
therefore, relatively inexpensive. As such~ it is a very desirable
reductant for sulur dioxide-containing gasesO
; 8ritish Patent 1,390,694 is directed to a process for
the reduction of sulfur dioxide in gases with the use of coal. The
sulfur dioxide gases are introduced into a moving bed of particu-
late coal and reduced to elemental sulfur and/or hydrogen sulfideO
,,, ~L ~
. ,
`~
A potential problem with the use of such a moving bed is the
agglomeration of the coal particles at the temperatures required
for the reduction of the sulfur dioxide. Coal occurs naturally
in various forms, some of which are agglomerating and some of
which are non-agglomerating. With the high temperatures required
for the reduc~ion of sulfur dioxide with coal, potential problems
; are foreseen in the use of a movin~ bed with a~glomerating coals
unless the coal is first pretreated by, e.g., calcination, or some
other method to prevent agglomexation in the bed. It is apparent
that agylomeration of the coal in the bed would cause operating
difficulties and frequent shutdowns. Additionally, since this
process involves the use of a large excess of coal in the reaction
zone; the purity of the product sulfur can be adversely affected
by tar carry-over in the exiting ~ases. Some coals contain sub-
stantial amounts of tars, and when heated to reaction temperatures
these tars can be driven off with the product gases. Separation
of these impurities from the sulfur is difficult.
The reduction of sulfur dioxide in flue gases employing
coal in a fluidized bed is reported by Sinah and Walker in "Air
Pollution and its Control" AIChE Symposium Series, No. 126,
Vol. 68, pages 160 to 167 (197~). The authors report that the
coal employed in ~he fluidization process was anthracite ~hich
was calcined at 650C. for 2 hours prior to its use. The step of
calcination prior to use is, of course, undesirable in the sense
that this adds a step to the process and consumes substantial
amounts of energy~ rendering the process less economical.
` It is an object of this invention to provide an economi-
cal process for the reduction of SO2 with coal. It is a further
object of this invention to provide a process for reducing sulfur
dioxide with coal in a fluidized bed. These and other objects
will become apparent from the description which follows.
-2-
.~
:- 5UMMARY OF THE INVENTION
_ _
In accordance with this invention there is provided a
continuous process for reducing the sulfur dioxide content of gases
which comprises continuously introducing a sulfur dioxide-containing
gas stream into a reaction zone containing a bed of particulate
:.
material comprising coal and a solid diluent, the velocity of the
gas stream being maintained at least about 1 foot per second greater
than the minimum fluidization velocity of the bed of particulate
material and insufficient to carry substantial amounts of particu-
late material from the reaction zone, continuously introducingcoal into the bed while maintaining the temperature of the reac-
tion zone between about 1100~o and 2000F., continuously removing
product gases from the reaction zone and recovering sulfur from
the product gasesO
When operating in accordance with the process of this
invention, sulfur dioxide is effectively and efficiently reduced,
bed agglomeration is prevented regardless of the type of coal
employed in the process, the process may be run continuously by
regulating the introduction of additional coal into ~he reaction
zone, and high purity sulfur can be readily recovered.
DETAILED DESCRIPTION OF THE INVENTION
.. . _ . . . .
The use of coal to reduce sulfur dioxide pollutant to
elemental sulfur has several advantagesO The supply of coal is
not as limited as are other SO2 reductants such as natural gas.
Therefore, the use of coal as a reductant not only is more eco-
; nomical, but also conserves natural gas. It also represents an
additional ecological use for high sulfur coal.
Reduction of sulfur dioxide by crushed coal in a fluidbed of inert material also offers several practical advantages over
other types of reactors. The Eluid bed reactor is well suited for
the reaction because of the excellent gas solid contact. Thus,
~'
although the reaction is exothermic, ~he heat transfer is very good
resulting in good temperature and reaction control. As a conse-
quence, the reaction can be readily conducted in a fluid bed
- reactor at conditions favorable to high sulfur yields.
The source of the sulfur dioxide-containing gas is not
believed to be critical to the process of the present invention.
:.,
Stack yases which contain sulfur dioxide are generally treated to
concentrate the sulfur dioxide before the reduction process is
institutedO The use of gases containing a low concentration of
sul~ur dioxide would not be as efficient as the use of ~ases con-
taining higher concentrations and for this reason it is preferred
to employ a sulfur dioxide-containing gas having a concentration
above 5~ by volume, more preferably above 50%, and most preferably
above 80~ sulfur dioxide in the process of this invention. The
feed gases should be low in oxygen because it can compete with the
; sulfur dioxide for reaction with the coal. In some instances,
however, small amounts of oxygen may be desirable to maintain the
.;
desired heat balance in the reactor.
,
The process of this invention may be carried out utiliz-
ing various types of coal to reduce the sulfur dioxideO Both
agglomerating and non-agglomerating coals may be used at the pre-
ferred operating conditions without bed sticking problems. Among
the coals that may be used are lignite, subbituminous, bituminous
and anthracite. Calcination or other similar pretreatment steps to
render the coal nonagglomerating are not required for the proc~ss of
the present invention. Thus, the coal is merely dried if necessary,
pulverized to the desired particle size and fed to the fluidized bed
in the amounts necessary for efficient reduction of the sulfur
, di~xideO
The reaction of sulfur dioxide with coal may be illustrated
`~ by the equation:
ClH0.8OO 2(1ignite) ~ 1.1 SO2 ~ C 2 2
"
.
e ~ L'7~7
Side reactions can also occur producing COS, CS2 and H2S as
well as H2, CO and hydrocarbons. The formation of side products
will depend on the reaction conditions. It is of course most
. desirable to maximize the direct conversion of SO2 to sulfur.
The amount of SO2 conversion will be affected by the
feed ratio of coal to sulfur dioxide. ~igh ratios increase the
amount of sulfur dioxide conversion. However, high ratios also
tend to produce increased amounts of sulfur containing by-products
such as H2S and COS in the product gas. The optimum feed ratio
10 will depend on the type of coal used in the process. For example,
for lignite, it has been found that the conversion of sulfur dioxide
does not increase significantly above a 130% stoichiometric feed
ratio (based on the above equation) of coal to sulfur dioxide. In
some instances, sulfur cont~ining by-product formation is minimized
by employing less than stoichiometric ratios. Consequently, a feed
ratio between 80~ and 130% stoichiometric is preferred when lignite
is employed as the reductant.
The fluid bed in the reaction zone can be con-
trolled at a uniform reaction temperature by known means. Sulfur
20 dioxide is reduced by coal at any temperature between about 1100F.
and 2000F. However, reaction temperatures of 1300F. to 1800F.
are preferred to maximize SO2 reduction to sulfur. A temperature
of at least 1300F. appears necessary for practical, economical
conversion of sulfur dioxide. At temperatures above 1800F. there
is an increased danger of fusion in the bed which could result in
bed agglomeration.
Coals which are classified as "agglomerating coals" will,
as the classification indicates, tend to form lumps when heated
to the temperatures required for reaction with suflur dioxide.
30 Furthermore, regardless of the type of coal employed to reduce
sulfur dioxide, it is believeù, that at the reaction temperatures,
; .
certain low melting eutectics, which are formed from the coal
ash, build up as the reaction progresses and these tend to
ca~se bed agglomeration or sticking. It has been found,
according to this invention, that this tendency to agglomer-
ate can be combatted by providing a particulate, solid diluent
as a major portion of the fluidized bed. The use of such a diluent
ma~erial has also been found to result in stable bed operation
; and temperature uniformity. Thus, whether employing an agglomer-
ating or a non-agglomerating coal, the diluent solid material
is an essential part of this invention. The particulate, solid
diluent material must be one which will not adversely affect
the desired operation of the bed, i.e., should be substantially
unreactive with coal, sulfur dioxide and sulfur~ While inert
materials are satisfactory for this purpose, the solid diluent
may be composed in whole or in part of catalyst or other reaction
aidc Sand is a preferred diluent because of its abundance, its
cost and its abrasion resistance. Among other diluent materials
useful in the process of this invention are alumina, magnesium
oxide, alumino-silicates, quartz, silicon carbide, and the like.
The proportional make-up of the bed is not critical
except to the extent that there must be sufficient solid diluent
particles present to attain the desired result of preventing bed
sticking and there must be sufficient coal particles to react with
the 52 to the extent desired. Within these broad guidelines it
has been found advantageous to operate with a bed of particulate
solids containing between 0.1 and lOoO weight percent coal at any
given timec The relative amount of coal present in the bed will
also be affected by other factors, eOg., coal size, coal type,
i chemical composition of the coal and reactivity of the coal. Thus,
determination oi optimum be composition will vary accordin~ly.
. i
.. ~: '
The particle size of both the coal reductant and the
particulate material in the bed will depend on many fac~ors such as
the diameter of the reaction vessel, the abili~y of the particulates
to be fluidized and the efficiency of reaction which is required.
For example, obviously very large particle size ma~erial will not
be efficiently fluidized and very small material will be carried
~- away from the reaction zone with the gas. It has been found that
an average solid diluent particle size of between 150 and 1500
microns is appropriate for use in the process of this invention
but this could vary depending on the above-mentioned factors.
The coal particles may be the same size or different size than
the diluent. Since there usually is a relatively small amount
of coal present, compared to the diluent, the coal particle size
has little effect on the fluidization characteristics of the bed.
Of course, the coal size must be small enough to be fluidized
and large enough to remain in the bed for a time sufficient to
react with the sulfur dioxide. The depth of the bed will like-
wise be dependent on factors such as size of the reaction
vessel and the required contact time depending on the velocity
of the incoming gas. Further, bed depth is a factor which can
easily be selected based on the desired result.
It has been found, surprisingly, that mere ~luidization
of the coal even when combined with large amounts of particulate
diluent material is insufficient to provide an operating process
for the reduction of sulfur dioxide. At certain velocities,
even though the bed is in a fluidized state and the reaction
is proceeding as desired, bed sticking occurs and after a short
period of operation large lumps are formed which prevent fluidiza-
tion and cause a process shutdown. While not wishing to be bound
to any specific theory of operation it is believed that despite
the presence of the particulate diluent material, and despite the
fact that the particles in the bed are in continuous motion, the
-7~
6~
reaction of sulfur dioxide with the coal particles causes low
melting eutectics to form on the bed particles and when they come
in con~act with other bed particles, agglomeration occurs when
:: there is insufficient bed movement. It is thus a critical aspect
of the present invention to provide sufficient bed movement by
maintaining the velocity of the fluidizing gases at least about
1 foot per second, preferably 1.1 feet per second, greater than
the minimum fluidization velocity necessary to fluidize the
particular bed of particulate solids~
The minimum fluidization velocity (Vmf) of any specific
bed of particulate material is determined by using the correlation
of Wen and Yu (C~.P. Symposium Series No. 62, Vol. 62, pp.
100-111, 1966),
(Re)mf = ~33.7)2 + 0.0408 (Ga~ 1/2 _ 33.7
where (Re)~f is the particle Reynolds number at onset of fluidiza-
tion and Ga is the Galileo number, defined as,
. ( p) ~f (s - Pf)9
~ ~ ....
.'. (Y)
The minimum fluidization velocity (Vmf) is thusly
calculated by,
'~ V = ( ~f ( )
mf
where dp is the effective particle diameter in the bed and is
defined as,
dp
~ xi
i=l -
-~ 30 wherein
Pf - the density of the feed gas at reaction temperature
and pressure
. -8-
.
; Ps = the density of solid particle in bed
= the viscosity of the feed gas at reaction tempera-
ture
g = gravitational acceleration (32.2 feetjsec.2)
Xi = weight fraction of particles in th~ fluid bed
having diameter dio Xi and di are readily deter-
mined by standard sieve analysis
Operation of the process of this invention in accordance
with the parameter of velocity of feed gases defined above results
in the ability to conduct the process over long periods of time
without the detrimental bed sticking occuring. While this is the
minimum gas velocity for successful operation of the sulfur dioxide
. ~ recluction process, it will be apparent that the velocity of the
i incoming gas should not be so great as to result in loss of sub-
`~ stantial amounts of particulate materials from the reaction zone
by being carried away with the gas coming from the fluidized bed.
The gases leaving the reaction zone and the fluidized
bed may be treated to recover the elemental sulfur, e.g., by con-
densation. Furthermore, in some cases it may be desirable to
thereafter react the effluent gases, as in a Claus reactor.
A particularly advantageous feature of operating in
accordance with this invention lies in the ability to obtain
rather high purity sulfur in a relatively easy manner. Although
substantially all of the coal ash~ i.e., the ash remaining after
; the coal reacts with the sulfur dioxide, is carried out of the
bed with the exit gases, the majority of this coal ash may be
readily removed from the gases prior to sulfur condensation by
known means, for example, in a cyclone separator. Furthermore,
~ residual coal ash which may deposit in the liquid sulfur as the
.~ .
3Q product gases are cooled to allow the sulfur to condense may be
easily removed from the sulfur by simple filtration, resulting
., . - g_ .
~. :
~ t7~
in high purity sulfur product. Additionally, when operating in
accordance with this invention, product sulfur is essentially
uncontaminated with coal tars.
As indicated, the process of the present invention is
run in a continuous manner by continuously feeding coal to the
fluidized bed which may be accomplished in a known manner, for
example, by means of a ~eed screw, at a rate consistent with the
reaction rate of the coal in the bed. Preferably the coal is fed
~o a point near the bottom of the bed.
The following examples are given by way of illustration
only O
In the following examples a six-inch diameter fluid bed
reactor was used. The bed was primarily composed of sand and
usually had a static depth of about 2 ft unless otherwise noted.
Based on the fluid and solid properties in the bed, the minimum
fluidization velocity was calculated using the correlation of
Wen and Yu. SO2 and N2 gases were measured by rotameters,
preheated and continuously fed to the bottom of the bed at a
velocity to sustain bed fluidity. The superficial gas velocity
in the bed was based on the inlet gas feed rate at the operating
conditions. Dried pulverized coal was continuously fed into the
~luid bed at prescribed rates by means of a feedscrew. Reaction
temperatures were measured by thermocouples in the fluid bed.
The reaction temperature was controlled by the degree of feed
gas preheating. Because of the large heat losses from the six-
inch reactor~ it was sometimes necessary to include oxygen in
the feed gas (0-5 vol.%) to maintain the reaction temperature
(i.e., compensate heat losses by burning coal). Ash was
removed from the exit gas by cyclones. A~ter condensing the
sulfur (and water) the exit gases were sampled and analy~ed by
gas chromatography.
--10-
7~7
Example 1
A gas composed of 78 vol.~ SO2 and 22 vol.~ N2 was
preheated and fed to the bottom of a 2 foot deep fluid bed at
a rate to produce a superficial gas velocity of 1.27 ft/sec. in
the bed. The average particle size of the sand was 350 u. The
minimum fluidization velocity of the sand was 0.093 ft/sec. The
bed temperature was controlled at 1375F. The reactor was
continuously operated for 80 hours at these conditions without
bed sticking problems. During this time, the feed rate of dried
lignite (170 microns) was varied~ When the lignite feed rate was
; 102~ of the stoichiometric amount needed to convert the feed SO2
(to sulfur, carbon dioxide~ and water), the SO2 conversion was 80%.
83~ of the SO2 converted formed sulfur. After sulfur and water
condensation the reactor exit gas had the followin~ composition
(vol~%) 22.2% N2, 0.6% CH4, 0.6~ CO, 50.9% C02, 1.9% COS, 7.1
H2~ 0~ CS2, 15-4% SO2
Example 2
An attempt was made to operate at conditions similar
-to those in Example 1, except the average sand particle size was
550 microns. The minimum fluidization velocity of this sand was
, 0~28 ft/sec. The superficial gas velocity in the bed was 1.27
~t/sec. The bed was fluid initially, but bed sticking occurred
~fter only four hours of operation, causing loss of bed fluidity
and necessitating shutdown and discharge of the bed.
Example 3
Lignite from North Dakota was dried, pulverized to an
a~erage particle si2e of 170 microns, and used to reduce SO2 in the
fluid bed reactor, The lignite/SO2 feed ratio was 96% stoichio-
~etric. The gas fed to the reactor was composed of 80 vol.~ SO2,
19% N2 and 1~ 2 The fluid ~ed temperature was maintained at
1490F. The superficial gas velocity in the 2-foot deep bed was
. .
.
7~
1.3 ft/sec., and the average sand particle si~e in the bed was
350 microns~ The minimum fluidization velocity of the sand was 0.11
ft/sec. These conditions were maintained for 26 hours, and bed
sticking never occurred~ The exit gas analyzed (vol.%): 20.7%
N2, 0~9% CO~ 52.4~ CO2, 2.5% COS/ 6.9~ H2S, 0.7% CS2, 10-5% SO2-
88% of the SO2 in the feed yas was converted~ and the conversion
was 86% selective to sulfur.
Example 4
An agglomeratins subbituminous coal was fed to the
fluid bed reactor to reduce SO2. The coal was dried (220F)
and pulverized to an average size of 170 microns but received no
other pretreatment. The coal/SO2 feed ratio was 98~ stoichio-
metric. The feed gas contained 68~ SO2, 27% N2t 5% 2 and was
fed at a rate to effect 1.77 ft/sec. superficial gas velocity in
a 2-foot deep bed. The sand in the fluid bed had an average
particle size of 350 microns and a minimum fluidization velocity of
0.092 ft/sec. The bed temperature was 1500~F. These conditions
were maintained for 14 hours and bed fluidity was never lost. The
reactor exit gas analyzed 25.9 vol.% N2, 2.8~ CO, 33% CO2, 3.0%
20 COS, 3.6~ H2S, 1.9~ CS2 and 26O5~ SO2. 60% of the SO2 fed was
converted and the conversion was 74% selective to sulfur.
Example S
Dried pulverized (1~0 microns) anthracite coal was used
to reduce SO2 in the fluid bed reactor. A 2-foot deep sand bed
was usedO The sand had an average particle size~'of 350 u and a
minimum fluidization velocity of 0~095 ft/sec. The operating
super~icial gas velocity in the bed was 1.2 ft/sec. The feed gas
contained 86 vol.~ SO2, 6% N2, and 8% 2 The coal/SO2 feed ratio
was 165% stoichiometric, but the SO2 conversion was only 43~, which
illustrates the lower reactivity of anthracite. 96~ of the SO2
which was converted formed sulfur. On a water and sulfur free
12-
'
'7~7
basis, the reactor exit gas contained 6.9 vol.% N~, 0.6~ CO,41.8% CO2, 1.8% COS, 0.1% H2S, 0.1 CS2, and 50.6% SO2.
Example 6
~` Dried pulverized tl70 microns) lignite was fed to the fluid
bed reactor at a rate equivalent to 61% of the stoichiometric
amount needed to reduce the SO2 fed to the reactor (to sulfur,
carbon dioxide and watex). The feed gas contained 71% SO2,
26% N2 and 3% 2 and was fed at a rate to effect a superficial
gas velocity of 1.7 ft/sec in the 1500F fluid bed. 350 microns
sand with a minimum fluidization velocity of 0.09 ft/sec was used
in the bed. The reactor exit gas analyzed 27.5 vol.% N2, 0.7%
CO, 37.0% CO2, 1.0~ COS, 1.5% H2S, 1.1% CS2, and 33.7% SO2.
The SO2 conversion was 55%. 89% of the converted SO2 formed
sulfur.
Example 7
The fluid bed reactor was operated at the same condi-
tions as in previous Example 6, except the lignite/SO2 feed ratio
was 94~ stoichiometric. At this higher lignite feed rate, the
average SO2 conversion was 74~ and 84% of the converted SO2 formed
sulfur. The reactor exit qas composition was 24.5 vol.% N2,
0~1% CH4, 0.8~ CO, 50.9% CO2, 105% COS/ 4.2% H2S, 1.3~ CS2, and
17.3~ SO . The sulfur condensed from the exit gas was filteredO
The filtered product analyzed 99.9~ sulfur and was considered
high quality.
Example 8
Dried pulverized (170 microns3 lignite was fed to the
~luid bed reactor at essentially the same rate (94~ stoichiometric)
as in Example 7. Other reaction conditions were also the same,
except the fluid bed temperature was 1300F. At this lower tempera-
ture the average SO2 conversion was 66~, and the conversion was 89%
selective toward sulfur formation. The reactor exit gas analyzed
.
. ..
' : .
7~
2307 vol.~ N2, 0.5~ CH4, 0.3% CO, 45.1% CO2, 0.9~ COS, 1.1% H2S,
1.3% CS , and 20O7% SO .
2 2
Example 9
The fluid bed reactor was operated at conditions
similar to those in Example 8, except a shorter gas contact
time was effected by a higher gas velocity in the bed o~ 2.4
- ft/sec. A 2 foot deep bed of sand was used~ The average
particle size of the sand was 550 microns, and the minimum fluid-
ization velocity was 0.26 ft/sec. ~ried 170 microns lignite was
continuously fed at a 95% stoichiometric feed ratio. Because of
the shorter contact time, the average SO2 conversion was a lower
4B%. 80% of the converted SO2 formed sulfur. The reactor exit
gas contained 20.4 vol.% N2, 0.6% CH4, 0.7% CO, 29.9% CO2, 0.7%
COS, 3.5% H2S, 1.5% CS2, and 38.6~ SO2.
Example 10
` The mean gas contact time was increased by increasing
the sand bed depth to 3.5 feet (static). The sand had an average
., .
,~ particle size of 350 microns and a minimum fluidization velocity
of 0.09 ft/sec. The feed gas, composed of 70 vol.% SO2 and
and 30 vol.~ N2, was fed at a rate to effect a gas velocity
of 2.2 ft/sec in the bed. Dried lignite, with an average particle
size of 170 microns, was continuously fed at 104% of the stoichio-
metric amount needed to reduce the SO2 fed. The bed temperature
was maintained at 1500F. Because of the longer gas contact time,
94% of the SO2 feed was converted, and 79% of the SO2 feed was
reduced to sulfurO The reactor exit gas contained 33.5 vol.%
N2, 2.1% CO, 47.6% CO2, 4.6% COS, 6.1% H2S, 0.4% C5~ 4.7% SO2.
';,
,.,~
; -l4-
'
,
