Note: Descriptions are shown in the official language in which they were submitted.
CA 02334505 2001-02-07
TITLE: REMOVING HYDROGEN SULFIDE FROM A
GASEOUS MIXTURE USING IRON HYDROXIDE
BONDED TO CALCINED DIATOMITE
FIELD OF THE INVENTION
The present invention relates to the removal of hydrogen sulfide
(HZS) from various gases, and more particularly it relates to the removal of
hydrogen sulfide from a gaseous mixture using a filtering media containing
iron hydroxide intimately bonded to calcined diatomite.
BACKGROUND OF THE INVENTION
Hydrogen sulfic~le can be present in various air streams and is often
found in gas streams associated with petroleum storage and transfer
facilities, anaerobic digesters, sewage treatment plants and pulp and paper
mills. In many cases, the hydrogen sulfide has to be removed because of
its toxicity, corrosive properties, and unpleasant odour.
Several methods are known and have been used in the past for
removing hydrogen sulfide from a gas stream. Perhaps the most popular
method is one which consists in passing the gaseous mixture through an
iron sponge bed. The iron sponge bed is a type of filter which comprises
essentially iron salts adsorbed on a wood chip support media. Also, there
1
CA 02334505 2001-02-07
are known processes in which a sulfurous gas is passed through a bed of
iron oxide particles. These processes and various others are described in
US Patents which can be found in particular in the US Classification
423/231 entitled: Rerrioving Hydrogen Sulfide from a Gaseous Mixture
Utilizing Iron Oxide or Hydroxide.
Problems associated with the prior art processes are numerous and
include the facts that some media are self igniting when exposed to air and
therefore are not renewable. Because of their weak sulfur retention, some
of these non-renewable media must be treated as hazardous waste. Other
known filtering media have a relatively low HZS adsorption capacity or a
low HZS adsorption -performance in a single pass process. Another
drawback of some commercial H2S filtering systems is that the filtering
media must be disposed of after a single use.
As such, it will lbe appreciated that there continues to be a need for
a filtering process in which the filtering media is capable of removing
hydrogen sulfide from a gaseous mixture with a high single pass
performance. Further, it is believed that there continues to be a need for
a filtering process wherein the filtering media is easily renewable and does
not generate any hazardous waste when disposed of after multiple reuses.
SUMMARY OF THIC INVENTION
The present invention provides for an effective process for
removing hydrogen sulfide from a gas stream. Essentially, the process
according to the present invention uses a filtering media which has a large
single-pass adsorptionperformance, which is renewable several times and
which has a considerable lifetime adsorption capacity.
2
CA 02334505 2001-02-07
In a first aspect of the present invention, there is provided a process
for removing hydrogen sulfide from a gas stream wherein the gas stream
is passed through a filtering media consisting essentially of calcined
diatomite and between 5% and 30% by weight of ferric ions bonded to the
calcined diatomite.
This process is particularly efficient due to the fact that the removal
of hydrogen sulfide from the gas stream is effected with a single pass
adsorption performance of up to 45 mg of H2S per gram of filtering media.
Other advantages include the fact that the process has the ability to remove
HZS from a gaseous mixture, from a concentration of 30,000 ppm down to
non-detectable levels of less than 0.2 ppm in a single pass. Further, the
lifetime adsorption capacity of the filtering media is about one half or more
of the weight of the filtering media.
In another aspect of the present invention, there is provided a
process for removing hydrogen sulfide from a gas stream, comprising the
steps of passing the gas stream through a filtering media consisting
essentially of calcined diatomite, and between 5% and 30% by weight of
ferric ions bonded by c:hemisorption bonds to the calcined diatomite. The
process further includes the steps of renewing the filtering media several
times when the filtering media is saturated with sulfides by blowing
ambient air through the filtering media. The regeneration of the filtering
media can be accomplished while maintaining an average hydrogen sulfide
adsorption performance thereof of about 32 mg of hydrogen sulfide per
gram of filtering media per cycle. Because the filtering media is non-
flammable, there is no risk of combustion due to the heat generated during
the regeneration proce~~,s.
3
CA 02334505 2003-09-11
In yet another aspect of the present invention, there is provided a
process for removing hydrogen sulfide from a gas stream, comprising the
step of passing the gas stream through a filtering media consisting
essentially of calcined diatomite having particles ranging in sizes between
about 30 mesh and about 60 mesh, and between 5% and 30% by weight of
ferric ions bonded by chemisorption bonds to the calcined diatomite. The
process is particularly advantageous for filtering moist gases, due to the
fact that the filtering media remains porous when wet. The eff ciency of
the process increases with the adsorption of a certain amount of moisture
in the filtering media.
Still another feature of the process according to the present
invention is that it is susceptible of a low cost of material, installation
and
operation, and accordingly is then susceptible of low price of sale to the
industry, thereby making such H2S filtering process economically available
to the public.
Other advantages and novel features of the invention will become
apparent from the following detailed description.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT
While this invention is susceptible of embodiments in many various
forms, there will be described in details herein a specific embodiment, with
the understanding that the present disclosure is to be considered as an
example of the principles of the invention and is not intended to limit the
invention to the embodiment described.
The filtering media used in the process according to preferred
embodiment was the subject of an earlier patent, US Patent 6,200,482
4
CA 02334505 2003-09-11
issued on March 13, 2001, in which it is described as a filtering media for
removing arsenic from ground water. The filtering media is known under
the trademark MEDIA G2~ and is available from ADI International Inc.,
a corporation having its principal place of business in Fredericton, New
Brunswick, Canada. Although the initial purpose of the filtering media
was for use as a water filter, it was discovered that the same filtering media
has advantageous properties in removing hydrogen sulfide from a gas
stream.
The filtering media contains iron hydroxide Fe(OH)3 intimately
bonded to calcined diatomite. The ferric ions content in the filtering media
is between about 5% and about 30% by weight. The preferred calcined
diatomite material has particles ranging in sizes from about 30 mesh to
about 60 mesh. This size and type of diatomite particles has been found
to be advantageous for use in filtering arsenic from water as well as for use
in gas filtering columns, particularly for allowing intimate contact between
the gas and the ferric ions bonded to the diatomite particles. The calcined
diatomite particles do not offer substantial resistance to the flow of a gas
passing through it and do not expand in contact with a moist gas.
The calcined type of diatomite is believed to be an important
element also contributing to the performance of filtering media in
removing hydrogen sulfide from sulfurous gases. Calcined diatomite
particles have multiform shapes and a greater porosity than ordinary
diatomite particles. The heat treatment applied to the diatomite particles
during the calcination process, increases the porosity of the particles by
breaking their surfaces and forming pores, cracks, crevices, cavities,
hollows and protrusions. These pores, cracks, crevices, cavities, hollows
and protrusions offer additional surfaces on each particle to adsorb and to
retain ferric ions.
5
CA 02334505 2001-02-07
During the manufacturing of the filtering media, the calcined
diatomite is impregnated with ferric ions in a liquid form, such as a ferric
chloride solution. The solution entrains the ferric ions over the entire
surface of the calcined diatomite particles and deep inside the pores,
cracks, crevices and cavities of the particles. Then, the ferric chloride is
converted into iron hydroxide in-situ within the diatomite particles, using
sodium hydroxide for example, to better bond the ferric ions over and into
the entire structure of e-ach diatomite particle. The diatomite material has
negative charges and attracts the positively charged ferric ions, thereby
contributing to the formation of strong ionic impregnation bonds between
the diatomite material and the ferric ions.
During the manvufacturing process, the sodium hydroxide is added
slowly to bring the p~( of the slurry to a final value of at least about 9,
ensuring a complete and unhasty conversion of the ferric chloride to iron
hydroxide. This manufacturing process is advantageous for yielding a
ferric ion content, in the form of iron hydroxide, of between about 5% and
about 30% by weight of the media. During this manufacturing process, it
is believed that irregu'.Lar clusters of ferric ions are formed and become
entrapped or otherwise: interlocked inside the pores, cracks, crevices and
cavities of the calcined diatomite particles, and therefore become strongly
bonded to the calcined diatomite particles. The bond described above is
believed to be a chemisorption bond produced by an impregnation-
oxidation process which is characterized by its irreversible chemical
forces. It is also believed that these chemisorption bonds between the
ferric ions and the calcined diatomite particles contribute greatly to the
abilities of the filtering media to retain its ferric ions such that it is
renewable several times after being saturated with sulfides. Also it was
found that when the filtering media is saturated with water or subj ected to
6
CA 02334505 2001-02-07
a flow of water through it, the iron component is not released from the
filtering media. This .Finding supports the fact that iron hydroxide is
intimately bonded to the calcined diatomite particles.
Other advantagca of using the filtering media for removing H2S
from a gas stream include the fact that it has a pH of about 10 - 11, which
favourably affect the reaction of HZS with the ferric ions in the filtering
media.
Testing of the filtering media was carried out to determine its
optimum performance in removing hydrogen sulfide from a sulphurous
gas. In a first serifs of tests to determine adsorption capacities,
comparative results were obtained from similar tests carried on a variant
of the filtering media, hereinafter referred to as the variant media, wherein
the calcined diatomite was replaced by vermiculite. The filtering media and
the variant media are generally or j ointly referred to as the media sample
or both media samples..
Both media sannples were subjected to testing in dry and moist
conditions. Dry columns were filled with media samples that were not
rinsed or pre-moistened. All the fines were present in the media samples.
Moist columns were prepared in three different ways: 1) washing the
media samples with wGvter until all the fines were removed; 2) soaking the
media samples in water overnight and placing them in the columns with
minimal removals of the fines, and 3) placing the media samples into the
columns and then pouring water down through them. In all three cases, the
moist columns were allowed to drain out all excess water for at least one
day before testing began.
7
CA 02334505 2001-02-07
All columns were fed a gas stream from an anaerobic digester,
containing approximately 30,000 ppm of H2S. The outlet H2S
concentration was measured several times per day, using gas testing tubes
known under the trade name DRAEGERTM, and having a minimum
readable value of 0.2 ppm. Saturation of the media samples was
determined when HZS concentration in the treated gas exceeded 500 ppm.
The results of the tests were as follows. The adsorption
performance of the column containing the washed filtering media
according to the preferred embodiment was approximately 30 mg of HZS
per gram of filtering media before it was considered saturated. The
filtering media which had been soaked but not rinsed of fines had an
adsorption performance of 45 mg of HZS per gram of filtering media. The
dry filtering media ad:>orbed about 40 mg of HZS per gram of filtering
media. The variant mf;dia was able to adsorb almost 70 mg of HZS per
gram of variant media.
It is believed that during the filtration process, the media samples
adsorb HZS and form ferric sulfide by the oxidation of HZS and the
dissociation of the iron hydroxide species present in the media samples.
Regeneration of the filtering media was accomplished by the
oxidation of the FeS produced during HZS removal. FeS was oxidized by
simply blowing ambient air through the column to form different species
of iron hydroxide, elemental sulphur and water. The regeneration process
reconverts the iron hydroxide to its original bond to the diatomite material
such that the filtering; media is usable again to remove HZS from a
sulfurous gas stream.
8
CA 02334505 2001-02-07
Ferric sulfide is grey to brownish black in colour and agglomerates
into lumps, rods or granular powder during the filtration process. As a
sulfurous gas is passed upward through the filtering media, the filtering
media gradually turns 'black, beginning at the bottom of the column, and
indicates that H2S removal is taking place. During the process, the
formation of lumps and fine clay-like gray powder can also be noticed.
Some of the test columns were transparent. During regeneration of
the filtering media, a colour change was also noted. The filtering media
was seen to change from completely black to almost its original orange
colour. After regeneration, the filtering media may have a lighter shade of
orange, possibly due to the elemental sulphur, being yellow, produced in
the filtering media. In ~~ome cases, black specs may remain in the filtering
media. These black specs indicate that total regeneration has not been
attained.
Attempts to regenerate the variant media were unsuccessful.
Blowing air through tile column did not return the variant media to its
original colour. Also, it was found that as this variant media becomes wet,
it expands to such a degree that the filter column becomes completely
plugged, and the flow ~f gas there through becomes almost impossible.
The testing of the filtering media has indicated that the adsorption
performance of the dry filtering media was similar to that of the pre-
moistened filtering media, being 40 mg of H2S and 45 mg of H2S per gram
of filtering media respesctively. The performance of the dry filtering media
has been shown to increase to a same level as for the moist filtering media,
after it had adsorbed moisture from the gas stream passing through it. It
9
CA 02334505 2001-02-07
is believed that the slight difference in initial performance is compensated
for by the advantages. in eliminating the need for pre-moistening the
filtering media.
The testing of the filtering media and the variant media also
indicated that although the variant media had the ability to absorb more
HZS in a first run, it is not renewable and therefore, the adsorption capacity
of the filtering media according to the preferred embodiment exceeds that
of the variant media in only two cycles. One cycle is referred to as a
saturation of the filtering media with sulfides and the regeneration of the
filtering media.
A second part of the testing program was focussed on the
mechanical characteristics of the filtering process, and more particularly
it was focussed on finding an optimum empty bed contact time (EBCT).
The EBCT is defined as the residence time of the gas inside the filtering
column.
Three different columns of one half inch in diameter and five feet
tall each were set up to test the effect of EBCT. These three columns were
set up to run at 40, 130 and 200 ml/min, yielding EBCT of 300, 90 and 60
seconds, respectively. Again, all columns were fed a gas stream from an
anaerobic digester, containing approximately 30,000 ppm of HZS. The
filtering media in all three columns were pre-moistened.
The results indicate that the EBCT of 60 and 90 seconds worked
better than the 300 second column. The 60 second column adsorbed 45
mg of H2S per gram of filtering media during its first cycle, gradually
declining to an average 32 mg of HZS per gram of filtering media per cycle
CA 02334505 2001-02-07
and a total removal of 560 mg of HZS per gram of filtering media in 18
cycles. Testing on the 60 second EBCT column was stopped when the
removal was only 26 mg of HZS per gram of filtering media for the last two
cycles. Average outlet H2S concentration prior to saturation was about 30
ppm, with several readings as low as 0.2 ppm.
The 90 second EBCT had similar results, averaging 30 mg of H2S
removed per gram of filtering media, and a total removal of 326 mg of HZS
per gram of filtering miedia in 11 cycles. Although this column did not
remove as much H2S per cycle, it has achieved a lower effluent
concentration of HZS, averaging 23 ppm.
The 300 second EBCT column did not work as well as the other
two columns. Its adsorption performance was only 20 mg/g per cycle, and
it removed 115 mg/g in 6 cycles. Average outlet H2S concentration was
40 ppm.
Further testing was carried out to measure the effectiveness of the
filtering media in removing HZS at different linear velocities. Linear
velocity is the speed at which the gas flows vertically through the filtering
column. Two columns were set up to operate at 1 ft/min and 3 ft/min
respectively, with a common EBCT of 60 seconds. The results shown
below are compared to~ the 60 second EBCT column mentioned before,
which was operating ai: 5 ft/min.
The 3 ft/min column yielded an average removal rate of 18 mg/g per
cycle. The filtering column was tested through 6 cycles and adsorbed 110
mg/g in total. The 1 ft/min column averaged only 17 mg/g per cycle. The
column was tested through 6 cycles, and removed 100 mg/g in total. The
results show that the 1 ft/min and the 3 ft/min velocities are not as
effective
11
CA 02334505 2001-02-07
as the 5 ft/min test which gave an average 32 mg of H2S removal per gram
of filtering media per cycle, and a total lifetime removal capacity of 560
mg of HZS per gram oar filtering media, in 18 cycles.
In view of thesf: results, a new column, 10 feet in length was built
to find the maximum linear velocity which the filtering media can handle.
This column was set up to run at 60 second EBCT with a linear velocity
of 10 ft/min, through a. filtering media in a dry state. During three cycles,
the average adsorption performance was 32 mg of HZS per gram of
filtering media. It was observed, however, that during the first cycle, the
outlet H2S concentration never went below 50 ppm. During the second
and third cycles, it was much lower, averaging less than 5 ppm prior to
saturation. This may be explained by the higher flow rate and therefore by
a longer time requiref~ for wetting the filtering media. However, once
properly wetted, performance was found to be excellent.
In commercial and industrial applications it is recommended to
contain the filtering media in a filtering column having a window or sight
glass, such that users can develop certain visual skills for evaluating at a
glance, the conditions of the filtering media. During the regeneration of
the filtering media, it is recommended to pass the purging air exiting one
filtering column into another filtering column to capture any hydrogen
sulfide that may be released from the filtering column being regenerated.
During regeneration of the filtering media, a small amount of the sulphur
on the ferric sulfide rnay be reconverted to hydrogen sulfide gas and
stripped off the filtering media by the regeneration air. The amount of
hydrogen sulfide exiting the filter in the regeneration air is less than 0.02%
of the hydrogen sulfidle which was originally adsorbed by the filtering
media. For environmental reasons, it is therefore recommended to pass the
12
CA 02334505 2001-02-07
purging air exiting a :Filtering column being regenerated into a second
filtering column before releasing the purging air into the atmosphere. For
convenience, this second filtering column may be an adjacent filtering
column in a bank of f ltering columns or a secondary filtering column
provided for this purpose.
It will be appresciated that where the application can tolerate the
injection of ambient air in the gas stream, the regeneration of the filtering
media can be effected automatically on a continuing basis. It is also
believed that the filtering media can be made new again by separating the
sulphur particles from it, by washing, sifting or otherwise.
As to additional details related to the manufacturing, installation
and use of the filtering media, the same should be apparent from the above
description, and accordingly further discussion relative to the manner of
making, using and re~aewing the filtering media would be considered
redundant and is not provided.
While one embodiment of the present invention has been described
herein above, it will be appreciated by those skilled in the art that various
modifications, alternate compositions, alternate methods and equivalents
may be employed without departing from the true spirit and scope of the
invention. Therefore, the above description should not be construed as
limiting the scope of the invention which is defined by the appended
claims.
13
