Chapter 8

8 Desulfurization of Natural Gas

8.1 Introduction
The effect of contaminants on fuel cells is one of the most important issues in fuel cells
operation and applications [93]. It has been identified that out of fuel cell components the
most affected by the contamination process is the membrane electrode assembly
(MEA).Three major effects have been identified such as kinetic effect (poisoning of the
electrode catalyst), conductivity effect (increase in the PEM resistance, including that of
the membrane and catalyst layer, and mass transfer effect (catalyst layer structure and
hydrophobicity changes causing a mass transfer problem) [94]. However, impurities in
hydrogen fuel, such as CO, H2S, NH3, organic sulfur-carbon, and carbon-hydrogen
compounds, and in air, such as NOx, SOx, and small organics, are brought along with the
fuel and air feed streams into the anodes and cathodes in a PEM fuel cell stack, causing
performance deterioration, and sometimes permanent damage to the MEAs [95, 96]. The
hydrogen fuel impurities mentioned above mainly arise from the manufacturing process,
in which natural gas (CH4) or other organic fuels are reformed to produce hydrogen gas.
Park and O’Brien [95] and Hayter et al. [96] briefly reviewed the effects of
contaminations in the fuel and air streams on cell performance. They discussed CO, CO2,
NH3, H2S and inert diluents such as N2 with respect to anode contaminations; SOx and
NOx (cathodic contamination); and NH3 and metallic species leached from the cell
components (membrane contamination). Since then, a considerable amount of work has
been carried out by many researchers who focused on the impacts of contaminations on
fuel cell performance and lifetime.
It has been investigated extensively by many workers that hydrogen sulphide is
an even more severe fuel contaminant than CO [97-100]. A trace amount of H2S when
exposed to an anode or cathode of the cell, was found to degrade the cell performance
significantly mainly through the poisoning effect of the platinum catalysts [99]. Knights

289
et al. [94] found that at 100 mA cm-2, 80C, 1.2 ppm H2S could cause a cell voltage drop
greater than 300 mV with in 25 hour. This poisoning effect was also sensitive to the
temperatures and load levels of the fuel cell in operation [94, 101]. Large performance
losses were measured at H2S concentrations as low as 50 ppm at 70C when the fuel cell
anode was exposed to the H2S-containing fuel for 3.8 hours. Unlike the case of CO
poisoning, the presence of Ru in the platinum catalyst could not provide sufficient
tolerance to H2S poisoning [101].
Natural gas (the major part is methane) is one of the lightest hydrocarbon and
promising candidate for hydrogen production. It is also available abundantly and being
utilized as the preferred fuel in most of the countries. But the inherent problem with the
natural gas is that it contains sulfur generally in the form of H2S and some sulfur-
containing THT for safety reasons. These compounds have poisoning effects for the
hydrogen reforming catalysts and the catalysts deposited on the fuel cell electrodes.
These also shorten the life of gas handling, storing and piping systems. Hence, it is
necessary to remove these sulfur compounds from the raw natural gas or feedstock prior
to reforming as well as from the product hydrogen gas due to multidimensional poisoning
effects [102, 103].
Natural gas in pure form is an odourless gas and can not be detected in case of
leakage. THT is an induced impurity in the natural gas which is added in the pipeline
before distribution to make it detectable by its bad smell. But for hydrogen fuel reforming
the reaction takes place in the temperature range of 700C to 900C while the thermal
decomposition of THT takes place into H2S, thiophene and other sulfur compounds ( So
to S14) in the temperature range of 450C to 500C [104].

The removal of sulfur from the natural gas is not an easy task and lot of efforts
have been made with various techniques by many researchers and still the research work
is in progress due to the limitations of every process. A number of sulfur-removing
systems are and have been in use in different industrial as well as laboratory applications.
As every process employed can remove one or more sulfur compounds and not all of
them, the initial selection of a particular process may depend on feed parameters such as
composition, pressure, temperature and nature of impurities as well as product

290
specifications. The final selection of a particular process may be based on acid/sour gas
percentage in the feed, whether only H2S, or mixed [105, 181]. Catalytic oxidation of H2S
is also one of the processes to convert it into elemental sulfur which then get adsorbed on
the internal surface of the activated carbon, thus, leading to a sulfur load of 120% by
weight [106-110].
Another way to remove sulfur compounds is to convert the organic sulfur species
to H2S over a hydro-desulfurization catalyst. The next step is sulfur removal with an
absorbent. The same catalyst similarly converts any organo-chloride species to give HCl
and also acts as an absorbent for most problematic metal species. For chloride removal a
second absorbent is used [111-114]. For the removal of sulfur-containing odorants from
natural gas for fuel cell-based combined heat and power applications, Beavon sulfur
removal process, Co-capture of H2S and CO2 in pressurized gasifier-based process and
calcium-based sorbents are employed [115]. The high sorption capacity of the sorbents
has also been utilized for the removal of sulfur containing odorants from natural gas.
Impregnated carbons containing several types of inorganic impregnant are also used for
the adsorption of H2S and mercaptans. It was reported that H2S can be adsorbed up to
50% by weight [116-117]. In one of the processes the gas containing H2S is contacted
with the solid sorbent to remove H2S producing a product which has less than 0.1 ppm
sulfur. This product is contacted with a highly selective reforming catalyst which is more
sulfur sensitive under severe reforming conditions in subsequent reactors [117].
In the present research work, efforts were made to remove the sulfur compounds
(H2S and THT) from natural gas by the utilization of commercially available sulfur
removing sorbents, manufactured by renowned chemical companies. The other aspect of
the present investigation was to develop an economical and efficient desulfurization
system for fuel cell applications. The main feature of this work was the replacement of
costly sorbent by cheap molecular sieves, which shows an over all economical
replacement. All results pertaining to the desulfurization of H2S and THT were
encouraging and have been reported.

291
8.2 Overview of Desulfurization
8.2.1 Selection of Process

Different available technologies and possible reaction paths are in practice to remove the
sulfur, sulfur compounds and mercaptans as odorants from the natural gas of pipe line
quality for onward utilization for hydrogen fuel reforming for the utilization in PEMFCs.
A number of desulphurization processes have been designed, developed, plants built and
are in operation as per feed stocks available and product gas requirements. Here is an
overview of selective desulphurization processes:;

8.2.1.1 Sulfint HP

This process can be used for selective hydrogen sulphide (H2S) removal from high
pressure gases with direct conversion of H2S into elemental sulphur (S) and the residual
H2S levels in the treated gas can be lower than 1 ppm by volume. This process uses the
well known concept of redox desulfurization. However, this process has been specifically
designed to treat high pressure gases. The feed gas is contacted with the redox catalytic
solution which is an aqueous iron chelate-based solution in a co-current absorber. The
H2S is absorbed by the catalytic solution, reacts with the catalyst and is converted in to
elemental sulfur. Desulfurized gas is recovered from the top of the separator vessel. The
sulphur loaded solution is then pumped through a multi-cartridge, high pressure filter.
The filtered solution can be partly recycled to the absorber and partially expanded for
regeneration.
The expanded solution, after separation of the dissolved gases is regenerated with
air in the oxidizer vessel. The whole process operates at near-ambient temperature and no
thermal regeneration nor chilling is required. High pressure filtration mitigates foaming
and plugging problems, especially after the expansion of the solution. Due to direct
recycle of the filtered solution to the absorber, pumping costs are minimized. This
process is highly selective; only H2S is removed [94].

292
8.2.1.2 Eliminator

This technology removes H2S and light mercaptan species from a variety of process gas
in a safe, efficient and easy to operate manner. A number of different treatment
methodologies may be used to apply liquid Eliminator technology to a natural gas stream. Commented [Z9]: Only, if this information is NOT
confidential: Is is possible to specify the reacting species?
In Line injection methodology, liquid Eliminator product is sprayed into a gas
stream and the spent eliminator is then separated in either a vertical knockout pot or a
horizontal separation vessel.
In sparge tower technology, sour gas is bubbled up through a static solution of
liquid eliminator.
In packed tower technology, sour gas is contacted with circulating liquid
Eliminator in a counter current, packed bed scrubber tower.
In Venturi Scrubber technology, the sour gas is pulled into a venture scrubber by
the motive force of a circulating Eliminator stream and is used mostly for low pressure
systems.
In the entire above-mentioned technologies, liquid Eliminator removes H2S and
light hydrocarbons through conversion into a safe, easy-to-remove and dispose of amine
resin compound. Following contact with the Eliminator product, further treatment is
required. Eliminator may be used in water-saturated streams or dry streams, and works in
almost any operating conditions; other gas components, such as carbon mono-oxide, do
not effect media cost and or performance to an appreciable level.
Custom designed eliminator units matching design and operating conditions are
available to remove H2S and light mercaptans. The Eliminator system is capable of
removing performance < 0.1 ppm of sulphur at any stream pressure (vacuum), inlet H2S
level and for system temperatures up to 300°F [95].

8.2.1.3 Sulfur-Rite

This technology removes H2S and light mercaptan species from a variety of process gas
streams in a safe, efficient and easy-to-operate manner. Sour gas enters the treatment
vessel near the top, and is distributed evenly throughout the cross-sectional area, in order

293
to fully contact the granular, solid media bed. The sour gas then flows downward through
the bed, contacting the Sulfur-Rite media, which removes H2S and light hydrocarbons Commented [Z10]: Only, if this information is NOT
confidential: Is is possible to specify the reacting species?
through conversion into a safe, easy-to-remove and dispose of iron pyrite compound. Is Sulfur-Rite a trade mark? If so, please add “TM”. Please also
check, if this may apply for other names/substances in this chapter.
Following contact with the media, the sweetened gas exits the vessel near the bottom and
no further treatment is required.
Sulfur-Rite media prefers high humidity to water-saturated gas streams; other gas
components such as CO do not effect media cost and or performance to an appreciable
level.
Customized units matching a particular design and or operating conditions are
available. Typically, bed pressure drop requirements are only a few inches of water. The
Sulfur-Rite system is capable of providing removal performance < 0.1 ppm of sulphur at
any stream pressure or vacuum, inlet H2S level and for system temperature up to 300 °F.
Following are Sulfur-Rite materials and their general uses;
ARI-710 treats anaerobic gas streams where only H2S must be removed.
ARI-720 is for the removal of both H2S and light mercaptans from anaerobic gas
streams.
ARI-750XL is for the removal of both H2S and light mercaptans from anaerobic
gas streams where an extended media bed life cycle is required.
The operating costs are less then liquid scavengers, and for low-pressure
treatment systems, and have much greater reliability. In addition, there is minimal
operating and utility requirements; vessel design is quite simple thereby lowering capital
costs [96].

8.2.1.4 Gas-To-Liquids (GTL)

This process is used to produce a zero-sulfur, pump able syncrude from remote gas fields
or from associated gas. The process can operate on a range of natural gas feedstocks
including fields containing high carbon dioxide levels. The process is suited for remote or
offshore locations where space and weight are of particular concern.

294
 Natural gas is pre-treated to remove sulfur using conventional desulfurization
technology. Steam and recycle gases are added and the feed is further heated before
passing to the CRG pre-reformer. Using a nickel catalyst, the CRG pre-reformer converts
heavier hydrocarbons to methane and partially reforms the feedstock. Addition of steam
and further preheating are completed before the mixed gas passes to the compact
reformer.
The Davy/BP compact reformer is a multi-tubular counter-current reactor, which
in the presence of a nickel catalyst, produces a mixture of carbon oxides and hydrogen.
Heat for this endothermic reaction is provided by external firing of excess hydrogen
produced by the process with supplementary natural gas as required. Gas leaving the
reformer is cooled and generates sufficient steam to satisfy process heating requirements.
Excess condensate is then removed.
Dry gas is compressed in a single-stage centrifugal compressor and passes to a
membrane separation package where the surplus hydrogen is recovered and reused as
fuel. The non-permeate product from the membrane separation is fed to the conversion
section where the syngas is converted into a mixed paraffin and wax product using a
cobalt catalyst.
The reaction system can either be a fixed bed or slurry type depending on the unit
size and project needs. Unconverted syngas is recycled to the compact reformer feed. The
wax products from the conversion section can be hydro cracked to produce a pump able
syncrude using conventional hydro cracking technology. A wide range of reformer
operating conditions are possible to optimize the process efficiency.

8.2.1.5 Deep Desulfurization

Deep desulfurization of different fuels can be made by adsorption of sulfur on Nickel

based adsorbents, Cu(I) Y-Zeolite by Selective Adsorption for Reforming Sulfur (SARS),
deep oxidative desulfurization of fuels catalyzed by ionic liquids in the presence of H2O2
and by using calcium based sorbents [96-99].
Moreover, for the removal of sulfur containing odorants from natural gas for fuel
cell based combined heat and power applications Beavon sulfur removal process, Co-

295
capture of H2S and CO2 in pressurized gasifier process by calcium based sorbents can be
applied [100-103].

8.2.1.6 Molecular Sieve

A molecular sieve is a material containing tiny pores of a uniform size that is used as an
adsorbent for gases and liquids. Molecules small enough to pass through the pores are
adsorbed while larger molecules are not adsorbed. It is different from a common filter in
that it operates on a molecular level. A water molecule may be small enough to pass
through and larger molecules may not pass through these pores that’s why they often
function as desiccant. A molecular sieve can adsorb water up to 22% of its own weight.
Molecular sieves generally consist of Aluminosilicate minerals, clays, porous
glasses, micro porous charcoals, zeolites, active carbons or synthetic compounds that
have open structures through which small molecules can diffuse.
Molecular sieves can be regenerated by pressure change, heating and purging with
a carrier gas or by heating under high vacuum.

8.3 Experimental Setup

8.3.1 Materials

The experimental set up consists of mainly 5 reactors and the associated accessories. The
reactor no. 1 and 2 as shown in the Figure 8.4 were filled with the commercial sorbents
made by company A as follows ;(1). One of the sorbents is first responsible for
decomposition of the sulfur compounds and then the decomposed components are
absorbed by that material which is filled in the reactor-1 designated as 1.1 with mass M1.1
= 284 g and volume as 0.4 litre. (2) The 2nd sorbent filled in the reactor-2 was 1.2 along
with the filler material. Its mass was M1.2 = 93 g and volume, 0.1 litre. The inert material
was zirconium silicate type Z which was mixed with 1.2 and its volume was 0.3 litres.
The reactor no. 3 was loaded with the sorbents made by another company B as follows;
(3) The material 2.1 was filled in the reactor no. 3 whose mass was m2.1 = 159 g and
volume was 0.17 litre. The volume of the inert material zirconium silicate mixed with 2.1

296
was 0.23 litres. The reactor no. 4 was filled with the sorbent material 2.2 whose mass was
m2.2 = 1037.9 g and its volume was 1.11 litre. The inert material was used to fill and
provide homogeneous space utilization. The zirconium silicate was in the form of silly
beads type Z.
Table 8.1: Specifications of 13X molecular sieve.
Properties Values
Physical shape Spheres
Size of spheres (mm) Ø2.0-2.8
Wear ratio %≤ 0.15
Bulk density g/ml 0.66
Ratio % 96
Static water adsorption % 24
Ethylene adsorbed % 18
2
Crushing strength (N/mm ) 40
Coefficient of variation 0.3
Packing water % 1.5
Pore diameter (A) 10
 200-315
Regeneration temperature ( C)

The adsorbent, 13X molecular sieve whose chemical formula is Ca86

[(NaAIO3)86(SiO2)106]·XH2O was also utilized for the cleaning of natural gas as per
previous practice and its specifications are listed in Table 8.1.

8.3.2 Instrumentation and analysis procedure

The data acquisition system presented in the Figure.8.6 was designed in the LAB View-6
program and installed on a dedicated P-4 computer. The computer was hooked up with
the process controller which received inputs and out-puts from the pressure controllers,
temperature controllers and flow controllers installed on the incoming lines of natural
gas, combustion air, cooling air and cooling water. To monitor and control the
combustion of natural gas after cleaning, it was fed to the FLOX-burner whose exhaust
was equipped with a gas analyzer. The parameters of the system were optimized and

297
feed-back of the gas analyzer was utilized to control and monitor the flow of all of the
supplies i.e., natural gas, air and water through sensors, solenoids and controllers. The
respective parameters were adjusted to keep the combustion products e.g., CO2, CO, NOx
and unburned CH4 and O2 within permissible limits as recommended by the
environmental protection agency, Germany.

8.3.3 Dräger sampling tubes

Dräger sampling tubes as shown in Figure 8.1, 8.2 and Dräger pump system shown in
Figure 8.3 were used to perform identification and measurement of H2S and THT present
in the natural gas after the cleaning process. The contaminants in the natural gas were
first collected by adsorption or chemisorption on a suitable medium such as activated
charcoal or silica gel. The sample was then analyzed both qualitatively and quantitatively
by observing and comparing the change in colors of the adsorbed/ chemisorbed species in
different sections of the Dräger tube. The measuring range or detection limit of these
tubes was 1-10 ppm (4-40 mg/m3) for THT and 0.2-5 ppm for H2S.
Following reaction occurs with the change of color;
THT + KMnO4 → Yellowish brown reaction product. Commented [Z11]: Is it known and not confidential, which one
it is. If so, please specify.
Flexible
Rubber
connector
ppm
3

10
1
2

4
5
6
7

9
8

White Pre-layer Violet indicating-layer

Figure 8.1: Dräger tubes for the determination and measurement of THT in the
product gas.
ppm
0.5
0.2

3
1

W h ite -in d ica tin g lay e r

Figure 8.2: Dräger tube for the determination and measurement of H2S in the
product gas.

298
 It sucks a calibrated 100 ml sample of gas through the Dräger-Tube® with each
stroke. A built-in stroke counter displays stroke numbers and they have been recorded
accordingly. The completion of each pump stroke was indicated by the visual end-of-
stroke indicator.

Figure 8.3: Dräger Accuro 2000 pump and tube system.

8.4 Development of Desulfurization System

The line diagram of the experimental desulfurization system shown in Figure 8.5 consists
of a compressor to store the natural gas in a small storage cylinder with a capacity of 25
litres and then this gas was supplied to two sets of sorbing reactors at a pressure of 1 bar
and an average flow of 450 M3/hour (at an ambient temperature of 25C); each set having Commented [Z12]: m³?

two reactors for different reactions.

The 5 reactors have the following dimensions; length of reactor no.1, no. 2 and
no. 3 was 260 mm where as 720 mm is the length of reactor no. 4 and no. 5. The outer
diameter of all the reactors was 60 mm and inner diameter was 50 mm as shown in Figure
8.4. Each reactor was a tubular vessel of stainless steel with a thickness of 5 mm, having
two strainers on each side, one at the top side and the other on the bottom side with holes
of diameter 1.5 mm to keep the catalyst beads of diameter Ø = 2-2.8 mm with in the
reactor by allowing the natural gas to pass through the sorbent bed and two removable
threaded end caps were provided on each side for sorbent loading and unloading.
The display of the experimental gas cleaning detection and monitoring system
was developed in Lab View-6 program as presented in the Figure 8.6. The colour coding
of the piping system used in this display has been indicated as follows:

299
 Natural gas : Yellow (supplied through the compressed natural gas storage).
Air : Light blue (Red colour blower supply the combustion air and
brown colour blower supplies the cooling air).

To Flox Burner To Flox Burner To Flox Burner

To To To
Drager Drager Drager
Pump Pump Pump

M1.2
+
Zirconium Reactor-2
silicate

Reactor-5 Reactor-4
Molecular m2.2
seive

Reactor-1
M1.1

Reactor-3
m2.1
Strainer Shell End cap +
Zirconium
silicate

Figure 8.4: The reactors 1-5 along with assembly configuration and sorbent
loadings.

Cooling water: Dark blue (water pump was used to draw the cooling water).
The reactors were assembled as shown in Figure 8.5 and operated by passing the
pressurized natural gas into the reactor no. 2 through reactor no.1 as presented in the
Figure 8.4 and reactor no. 4 through reactor no. 3 in parallel and after desulfurization the
gas was supplied to a FLOX burner for safe disposal. In the meantime gas samples were
taken from the outlet of the reactors no.2 and reactor no. 4 separately for making
measurement and analysis of THT and H2S present in the product stream at regular
intervals of 4 hours.

300
 At the outlets of the reactor no.2 and reactor no.4 a nozzle of diameter 6 mm with
gate valve as shown in the Figure 8.4 was installed to get the sample of the natural gas for
analysis after cleaning. The Dräger pump of model number Accuro 2000 as shown in

301
Figure 8.5: Line diagram of the desulfurization system

302
Figure 8.6: The display of the experimental monitoring system in LabView-6 at ZSW Lab. in Stuttgart, Germany.

303
Figure 8.3 was used for the suction of sample from the desulfurised gas stream. The
pump was pressed 5 times to make it clear of the previous readings. Then a Dräger pre-
tube as shown in the Figure 8.1 for H2S adsorption was taken and its both ends were
broken by inserting and twisting into the tube opener and the tube was installed into the
nozzle on the outlet of the reactor no.2 and reactor no. 4 as shown in Figure 8.4 through
the flexible rubber tube, one by one, by keeping the arrows towards the Dräger pump.
This tube was joined with the THT determination graduated tube by keeping the arrows
towards the pump and thirty strokes were applied through the Dräger tube as each pump
stroke gets 100 ml of sample gas. Thus 3 litre of natural gas was passed through the
Dräger tube. The Dräger tube was removed from the pump and after 10 minutes it was
examined visually if there was any change in violet colour of the graduated portion of the
Dräger tube. However careful observation conformed that there was no identifiable
change in colour for the molecular sieve up to 580 hours run time in case of THT. The
result was that THT was 1 ppm after 580 hours run time as shown in Table 8.4.
Similarly, the detection and measurement of H2S in the natural gas, was done by
using H2S specific Dräger graduated tube after the completion of cleaning process. Both
of the ends of this tube were opened by the help of tube opener and inserted into the
pump, the tube was connected to the sample dispensing nozzle through the flexible
rubber tube and 15 strokes were applied to get 1.5 liter of sample gas. After 15 minutes
the tube was carefully checked for any change in color and the presence of H2S in the
sample. Hence, after 580 hours of run time the value of H2S was less then the minimum
detection limit of 1 ppm as shown in Table 8.4.
In this way measurement was made every 4 hours run time of the system and
recorded accordingly for 2000 hours of natural gas cleaning process to determine and
measure the affective life cycle assessment of the sorbent materials and degradation
behavior as well. The results were tabulated to check and evaluate the performance and
life of the sorbents from the Company-A, Company-B and for the molecular sieve as
shown in the Table 8.4.

303
8.5 Results and Discussion
The Key parameters affecting the selection of the gas cleaning processes were considered
and optimized accordingly.
1. Selection of gas preconditioning process upstream or final gas conditioning
downstream of the gas treating unit based on nature of impurities.
2. Gas pressure and temperature.
3. Feed gas composition including H2S content, CO2, hydrocarbons and other
contaminants.
4. Process configuration.
5. Selection of the dehydration process.
6. Product specification, such as H2S, CO2, H2O, Hydrocarbons and mercaptans.
7. Optimization of the existing equipment.
8. Required recovery efficiency.
9. Concentration of sulphur species in the stack gas.
10. Ease of operation.
11. Remote location.
12. Sulfur product quality.
13. Costs (Capital and operating).
With the considerations of above-mentioned factors selection of the right tool is very
important. A well-defined and economical way was adopted to remove sulfur contents
from the feedstock natural gas for the production of clean gas which is acceptable to the
fuel reforming systems and later on as a feed gas for fuel cell systems.
The cost of the materials used initially in the reactor no.1, reactor no. 2 (from
Company-A) and reactor no. 3, reactor no. 4 (from Company-B) and the reactor no.5
along with relevant parameters have been presented in Table 8.2 for comparison.
Similarly the 13 X molecular sieve costs at the rate of Euro 1.75 per litre and only one
litre was loaded into the reactor no. 5 which shows a significant running cost savings for
the same flow of natural gas, same composition of gas and same H2S and THT loadings.

304
 Table: 8.2 Mass, volume and cost comparison of the 3 different material combinations.
Material Company A Company B Molecular
Designation Sieve
Reactor No. 1 2 3 4 5
Material M1.1 M1.2+ZrSiO4 M2.1+ ZrSiO4 M2.2 13 X Commented [Z13]: Does 13X specify a definite material from a
defnite manufacturer? If so: Is this information confidential? If so,
Mass 284 g 93 g 159 g 1038 g 660 g please use a general term which describes the physical properties
without definite product information.
Volume 0.4L 0.1L + 0.3L 0.17L + 0.23L 1.11 L 1L
Cost $ / L 14.18 48 1.75

A desulfurization test system was developed and its line diagram have been
presented in the Figure 8.5 to check and evaluate the efficiency of the desulfurization
sorbents and also the design parameters of the reactors etc. The natural gas having
various components and sulfur impurities were analyzed by using the portable gas
analyzer model Lancom III, AMETEK Inc. USA and percentage of components have
been presented in the Table 8.3.

Table 8.3: Composition of the natural gas supply in ZSW lab. Germany.
S.NO COMPONENT VOLUME %
1. Methane 97.88
2. Ethane 0.79
3. Propane 0.24
4. Butane 0.06
5. Pentane 0.01
6. Carbon-di-oxide (CO2) 0.012
7. Oxygen 0.9
8. Hydrogen sulphide (H2S) < 1 mg/m3
9. Tetrahydrothiophene (THT) 18.1 mg/m3
10. Sulfur (in the form of H2S & THT) 6.5 mg/m3

The amount of H2S and THT was measured in the product gas stream with the
help of Dräger Accuro 2000 pump and tube system as shown in Figure 8.3. The values of

305
both of these sulfur compounds were found less than the detection limit of 1 ppm up to
576 hours run time (Table 8.4), in the case of molecular sieve loaded in the reactor no. 5
as presented in the (Table 8.2). And after 580 hours the H2S was found still less then the
minimum detection limit of 0.2 ppm but the THT was measured as 1 ppm which was
increased to 10 ppm after completing 600 hours run time. This showed that the molecular
sieve was becoming saturated and loosing the efficiency of sorption of THT, so
measurement was discontinued for regeneration process. It was noted down that the
desulfurization cycle of 13X molecular sieve without regeneration was approximately
500 hours and cleaning 225,000 M3 of natural gas by applying the above mentioned Commented [Z14]: m³?

parameters in Para 8.2.3.

Table 8.4: Testing data for the sorbents from Company-A, Company-B and molecular sieve.

Time Reactor no. (1+2) Reactor no. (3+4) Reactor no. 5

(Company-A) (Company-B) 13X Molecular Sieve
Hour H2S THT H2S THT H2S THT
s (ppm) (ppm) (ppm) (ppm) (ppm) (ppm)
1 < MDL* < MDL < MDL < MDL < MDL < MDL
50 < MDL < MDL < MDL < MDL < MDL < MDL
100 < MDL < MDL < MDL < MDL < MDL < MDL
576 -do- -do- -do- -do- -do- -do-
580 1
600 10
946
950 1
1000 8
1576 -do- -do- -do- -do-
1580 1
1600 1.5
1816 2
1864 3
1900 5

306
 2000 6
< MDL (Less then the minimum detection limit, for H2S: 0.2 ppm and THT: 1 ppm).
It was observed that the channel diameters of 13X molecular sieve cages (2-9 Å)
were good enough to adsorb the H2S and THT compounds with reasonable capacity
which could be utilized for the desulfurization of natural gas for fuel reforming processes
for ultimate utilization of hydrogen in the fuel cells.
Similarly, the sorbents loaded in to the reactors no. 1 and no. 2 (M1.1 and M1.2)
performed satisfactory up to 946 hours run time (Table 8.4) for H2S and THT compounds
and after 950 hours run time the H2S was still less then the minimum detection limit of
0.2 ppm but THT was found 1 ppm and gradually it was increased to 8 ppm after the
completion of 1000 hours. This showed that the sorption capacity of these materials was
degrading continuously for THT after completing 950 hours run time, so the
measurements for these materials were discontinued. Hence, it was noted that these
materials were good enough for desulfurization of 405,000 M3 m3 of natural gas at the
operating conditions mentioned above.
Moreover, the sorbents loaded into the reactors no. 3 and no. 4 (M2.1 and M2.2)
were giving measurements for H2S and THT in the product gas stream less then the
minimum detection limits up to the run time of 1576 hours and after 1580 hours run time
as presented in the Table 8.4 the H2S was still less then the minimum detection limit of
0.2 ppm but THT was found 1 ppm and gradually it was increased to 6 ppm after the
completion of 2000 hours. Hence the further measurements were discontinued and it was
noted that M2.1 and M2.2 are quite good for 1500 hours run time and cleaning of 675,000
M3 m3 of natural gas along with above-mentioned operating conditions.
In this way it was determined that with the operating conditions mentioned above
M2.1 and M2.2 have the desulfurization capacity of 675,000 M3 m3 without regeneration
and it costs 48 Euro (Table 8.2) and the volume of natural gas cleaned per Euro comes
out to be 14000 M3 m3 while M1.2 and M2.2 have the desulfurization capacity of 405,000
M3 m3 before regeneration and it costs 14.18 Euro (Table 8.2), giving volume of natural
gas cleaned per Euro as 28561 M3m3. But the 13X molecular sieve has the
desulfurization capacity of 225,000 M3 m3 before regeneration and it costs 1.75 Euro

307
(Table 8.2). Hence, the volume of natural gas cleaned per Euro was 128571 M3 m3 which
is maximum as compared to the other two set of sorbent materials.
Therefore, it can be stated that 13 X molecular sieve is an economical and
competitive material for the desulfurization of natural gas for fuel cell applications.

308