Ammonia Destruction in a Claus Tail Gas Treating Unit Abstract by gjjur4356

VIEWS: 142 PAGES: 12

									                Ammonia Destruction in a Claus Tail Gas Treating Unit

                             RAC™ (Rameshni Ammonia Combustion)

                                         Mahin Rameshni, P.E.
                     Technical Director, Sulphur Technology and Gas Processing


                       181 West Huntington Drive, Monrovia, California 91016, USA


This paper presents the basis of a new process for ammonia destruction in a Claus Tail Gas process in
which the ammonia acid gas combustion is carried out in a Tail Gas Furnace.

Two preferred methods are described for the partial oxidation of an ammonia (NH3) bearing gas stream,
potentially containing minor but significant quantities of hydrogen sulphide (H2S), in a conventional Claus
sulphur recovery tail gas treating unit. Historically, byproduct NH3 has normally been combusted in the
Claus reaction furnace, where the amount of NH3, which can be processed, is generally considered to be
limited to 30-35% of the total Claus feed on a wet basis. Typically, additional NH3 not processed in Claus
units is typically converted to ammonium thiosulphate fertilizer, or purified in order to be suitable for mar-
keting or incineration. In either case, the cost of such alternatives will normally be substantially greater
than with the process described in this paper.

A new process for ammonia destruction in a Claus Tail Gas Treating Unit has been developed by
WorleyParsons. A patent is pending for this new process. (Serial No. 60/910,074 with the filing date of
April 04, 2007) with the Trade Mark of RAC™ (Rameshni Ammonia Combustion).

This process is not commercialized yet, but WorleyParsons is looking for possible commercial applica-
tions. The critical components of this idea have already been industrially practiced.


With the sulphur content of crude oil and natural gas on the increase and with the ever tightening sulphur
content in fuels, the refiners and gas processors will require additional sulphur recovery capacity. At the
same time, environmental regulatory agencies of many countries continue to promulgate more stringent
standards for sulphur emissions from oil, gas and chemical processing facilities. It is necessary to de-
velop and implement reliable and cost-effective technologies to cope with the changing requirements.

                                                                                                                 1 of 12
Crude oil from several areas, including Brazil and Venezuela, contains high levels of nitrogen, which re-
sults in a very high level of ammonia in the sulphur recovery unit feed. We have seen if the ammonia con-
tent of the total acid gas as high as 55% by volume recently.

Historically, refinery byproduct NH3 has normally been combusted in the Claus reaction furnace, where
the amount of NH3, which can be processed, is generally considered to be limited to 30-35% of the total
Claus feed on a wet basis. Additional NH3 not processed in Claus units has typically been converted to
ammonium thiosulphate fertilizer, or purified in order to be suitable for marketing or incineration. In either
case, the cost of such alternatives will normally be substantially greater than with the new process de-
scribed here. In refineries where additional ammonia is produced as a result of a change in feedstock or
operation, the additional ammonia that is beyond the amount the sulphur plant can process could be
processed according to this new process with a simple retrofit to the existing tail gas unit.

Some refiners, including several in California, sell anhydrous ammonia for agricultural purposes on sea-
sonal basis through out the year. However, there are times that the ammonia needs to be processed in
the sulphur plants and amount of ammonia to be processed exceeds the capability of the sulphur plants
to handle it so the refineries are forced to store the sour water outside of the plant for months until it can
be processed.

In some refineries, a purified NH3 stream is directly routed to a forced draft incinerator for instance, the
NOxIdizer system from John Zink. The NOxIdizer system employs a 2 zone furnace with a quench sec-
tion between the two furnaces. The first zone operated under reducing conditions and the second zone
operates under oxidizing conditions at a relatively low temperature to limit NOx formation.

It is well known that using Oxygen Enrichment process will help with ammonia destruction in the Claus
process. Oxygen enrichment raises the reaction furnace temperature which ensures complete destruction
of heavy hydrocarbons and ammonia; reduces formation of COS and CS2, and shortens gas residence
time requirements for contaminants destruction. However, for a very high level of ammonia the Claus unit
has to operate on oxygen all the time and high provision is required to prevent any plugging through out
the unit.

Ammonia Destruction Background

The refining processes conducive to formation of byproduct H2S from more complex organic sulphur
compounds also tend to convert nitrogen compounds to ammonia (NH3). Subsequent recovery of the NH3
will typically yield a roughly equimolar gaseous mixture of NH3 and H2S (and water), which usually can be
conveniently fed to the Claus reaction furnace in combination with a larger H2S stream containing negligi-
ble NH3. The NH3 is ostensibly oxidized as follows:

        2 NH3 + 3/2 O2 → N2 + 3 H2O

                                                                                                                 2 of 12
However, since the O2 is nominally limited to that required for oxidation of only 1/3 of the H2S, H2S and
NH3 are in competition for available O2 such that initial SO2 generation likely exceeds that required by the
Claus stoichiometry and much of the NH3 is subsequently oxidized by reaction with SO2 as follows:

        4 NH3 + 3 SO2 → 2 N2 + 6 H2O + 3/2 S2

Prior to the 1970s, it was generally believed that complete NH3 destruction required combustion in an
oxidizing environment, followed by subsequent combination with the segregated amine acid gas stream.
This worked fairly well when the NH3-bearing acid gas stream was minor. However, as hydrotreating /
hydrocracking activity increased, the amount of sour water stripper gas to be processed became increas-
ingly significant, with attendant operational problems attributable to the formation of SO3 in the reaction
furnace - catalyst sulphation (including the TGU hydrogenation reactor), condenser corrosion from sul-
phuric acid condensation and equipment fouling with iron sulphate. Particularly susceptible to such acid
corrosion are the bottom tubes in the final condenser, and the dip leg inlet just above the sulphur level
where acid accumulates.

These problems prompted re-examination of the reaction mechanisms. In the late 1960s, the Ralph M.
Parsons Company determined that NH3 destruction under nominally reducing conditions in the first zone
of a split-flow reaction furnace greatly reduced SO3 formation. This became the basis for David Beavon's
1976 U. S. Patent #3,970,743, which recommends splitting the amine acid gas such that 33-66% of the
total H2S is routed to the first (NH3-burning) zone, and preferably 33-50%.

Today, WorleyParsons nominally targets for 45-65% of the total H2S routed to zone 1 based on measured
flows and assumed concentrations, which includes a reasonable safety margin for typical accuracy limita-
tions. For equimolar H2S/NH3 consistently accounting for 0-10% or 10-20% of the total H2S, Options 1a
or 1b (below) are the simplest. Otherwise, Option 2 can be applied over the entire practical range of NH3
relative to amine acid gas.

    Option 1a - For NH3 acid gas H2S = 0-10% of the total:
    Route 45% of the amine acid gas to zone 1. (For unsophisticated control schemes, a simple 50/50
    split is sufficient.)

    Option 1b - For NH3 acid gas H2S = 10-20% of the total:
    Route 35% of the amine acid gas to zone 1.

    Option 2 - For NH3 acid gas H2S = 0-65% of the total:
    Calculate the split based on typical compositions as follows:

        A*X + C*Y = 0.45*(A*X + B*Y)
        C = 0.45*B - 0.55*A*X/Y

When NH3 gas H2S exceeds 45% of total H2S, all amine acid gas will be routed to zone 2.

For the atypical case of a relatively pure NH3 stream containing negligible H2S, the amine acid gas can
simply be split 50/50.

                                                                                                               3 of 12
The benefit of the amine acid gas split, as generally cited, was maximization of the zone 1 temperature
for thermal dissociation of un-oxidized NH3. While true, it overlooks the equally important fact that split-
flow also maximizes the zone 1 oxidant concentration.

NH3 destruction was usually considered to be the result of simple oxidization:

        2 NH3 + 3/2 O2 → N2 + 3 H2O

As such, many viewed NH3 as competing with H2S for O2, and considered good NH3 destruction depend-
ent on faster oxidation rates than for H2S. Today that view is discounted by ASRL, whose research indi-
cates that the predominant route for NH3 oxidation is likely by reaction with SO2:

        2 NH3 + SO2 → N2 + H2S + 2 H2O

A minimum temperature of 2300-2350°F (1260-1290°C) has long been cited as necessary for complete
NH3 destruction. Certainly this is true for competing single zone furnaces which, in addition to their de-
pendence on high-intensity burners, often also require supplemental provisions for air/gas preheat, or
enrichment with natural gas or H2.

While similar temperatures are generally claimed necessary for zone 1 of a split-flow furnace, there is
field experience to the contrary, and burner intensity also becomes less critical.

With a split-flow furnace, such emphasis on temperature should arguably be avoided, as it only encour-
ages the operator to run oxidizing, with the aforementioned adverse impact of SO3 formation. Calculating
the split is usually the best way to not only maximize the temperature, but also ensure reducing condi-

Thermal dissociation of NH3 is favored by the higher temperatures inherent with O2 enrichment, in which
case a greater percentage of the amine acid gas can be routed to zone 1. Typical practice is to route ad-
ditional amine gas to zone 1 as necessary to limit, to the extent possible, the local temperature to 2500°F

The industry consensus is that the amount of NH3 that can be conventionally processed without generat-
ing NOx is limited to 30-35% of total SRU feed on a wet basis, ostensibly based on old Amoco correla-
tions. NO2 can oxidize SO2 to SO3:

        NO2 + SO2 → NO + SO3

However, such generalizations are arguably vague by virtue of neglecting such factors as temperature,
residence and whether split-flow.

Despite the industry trend toward increased hydrotreating activity, NH3 levels > 25% of total SRU feed
(molar wet basis) remain rare. However where such high levels of ammonia occur, current options for
handling the ammonia gas include conversion of the NH3-bearing stream to ammonium thiosulphate fertil-
izer, and purification of the NH3 for marketing or incineration.

                                                                                                               4 of 12
For units processing substantial NH3, Claus sulphur recovery efficiency is typically < 95% and secondary
treatment is necessary to meet emission limits. There are fundamentally two types of Claus tail gas treat-
ing units (TGUs) as reductive and oxidative.

The most prevalent are reductive units, which catalytically convert virtually all non-H2S sulphur com-
pounds and elemental sulphur vapor (Sx) to H2S for (1) subsequent absorption in a regenerable solution,
typically an alkanolamine, for recycle to the Claus unit, or (2) precipitation of elemental sulphur particles
in a redox process.

Oxidative TGUs thermally oxidize all sulphur compounds and Sx to SO2 for subsequent absorption in a
regenerable solution for recycle to the Claus unit.

WorleyParsons is presenting a new approach to handling gaseous NH3 streams potentially containing
minor concentrations of H2S. This new approach may find applications in many situations. For example,
purification facilities may have been installed to produce saleable NH3 where such markets no longer ex-
ist, or elimination of expensive purification steps may be desirable. Also, with bulk H2S removal from the
NH3 gas via established processes, it provides an economical means of increasing Claus sulphur recov-
ery capacity or processing greater quantities of NH3 than otherwise practical. Two variations are de-
scribed, depending on whether the associated TGU is reductive or oxidative.

                  Figure 1. WorleyParsons Conventional Ammonia Destruction Scheme

A New Approach - Reductive Tail Gas Units

In the reductive scheme (Figure 2), the NH3 gas is combusted sub-stoichiometrically to generate usable
heat and hydrogen (H2). As such, it replaces the conventional TGU feed heater or Reducing Gas Genera-
tor (RGG), which can be (1) an RGG burner, typically combusting natural gas stoichiometrically or sub-
stoichiometrically, or (2) an indirect heat exchanger, which will typically increase the reactor inlet tem-
perature to 550-650°F (290 - 340°C) in order to promote the aforementioned conversion reactions. Spe-
cifically, SO2 and Sx are converted by hydrogenation as follows:

        SO2 + 3 H2 → H2S + 4 H2O

        Sx + x H2 → x H2S

                                                                                                                5 of 12
While the Claus tail gas will supply a substantial share of the H2 needed, additional H2 must usually be
either imported or generated by air deficiency in the TGU RGG burner or Claus reaction furnace. While
NH3 combustion results in oxidation to N2 and water (H2O) according to reaction 4, thermal dissociation to
H2 will also occur to an appreciable extent, particularly at higher 2300°F, as follows:

        2 NH3 → N2 + 3 H2

When, according to established practice, NH3 is processed in the Claus reaction furnace, incomplete NH3
destruction resulting from failure to maintain optimum conditions of, for example, temperature, residence
time and/or SO2/NH3 ratio will tend to foul the downstream catalyst beds and heat exchanger tubes with
deposition of ammonium thiosulphate and sulphate. In the case of this process, however, NH3 destruction
need not be complete, as most of any such residual will simply be recycled to the sour water system via
the contact condenser blow down. Where downstream H2S absorption is by a recirculating regenerable
alkanolamine solution, as is usually the case, any minor NH3 not absorbed in the contact condenser will
be absorbed by the amine solution and ultimately concentrate in the amine regenerator reflux, which can
also be conveniently purged to the sour water system, consistent with common practice.

Among the benefits of this process, therefore, is that (1) reducing (air-deficient) combustion conditions
can be conveniently maintained without undue concern for potentially incomplete NH3 destruction, and (2)
the approach to stoichiometric air for complete NH3 combustion is automatically adjusted to maintain the
desired H2 residual downstream of the reactor, typically 2-3% H2 by volume.

A further useful consequence of combusting ammonia under reducing conditions is the avoidance, or at
least minimization, of byproduct nitrous oxides (NOx), which are undesirable pollutants. Furthermore, any
such compounds potentially formed will normally be reduced to N2 or NH3 in the hydrogenation reactor by
reaction with H2S, H2 or NH3. H2S, in particular, has proven to be an effective NO reducing agent, in part
due to its low dissociation temperature, and the abundance of H2S in the reduced Claus tail gas thus
plays a key part in NOx reduction in the hydrogenation reactor.

Consistent with established practice, carbonyl sulphide (COS) and carbon disulphide (CS2) in the Claus
tail gas are converted to H2S by hydrolysis, as follows:

        COS + H2O → H2S + CO2

        CS2 + 2 H2O → 2 H2S + CO2

Extraneous hydrocarbons often present in such NH3 acid gas streams will be partially oxidized to carbon
monoxide (CO) which, also consistent with established practice, can be substantially hydrolyzed to car-
bon dioxide (CO2) and H2 according to the so-called water gas shift reaction, as follows:

        CO + H2O → H2 + CO2

Obviously any H2S associated with the NH3 gas will increase the H2S load on the downstream TGU ab-
sorption medium, and that any SO2 surviving the NH3 combustion zone will be reduced to H2S in the hy-
drogenation reactor. Also, minor quantities of hydrogen cyanide (HCN), often found in association with

                                                                                                             6 of 12
NH3, will be readily hydrolyzed to NH3 and CO in the NH3 combustion zone, as they typically are in the
Claus reaction furnace, as follows:

        HCN + H2O → NH3 + CO

Where the heat release potentially exceeds process demand, a portion of the combustion gases are di-
verted through a waste heat boiler to achieve the desired net downstream reactor inlet temperature.
While fuel gas usage normally can be eliminated, it may be preferable to maintain at least a minimum
sub-stoichiometric natural gas fire which can be automatically increased in the event of NH3 gas curtail-
ment. This can also provide supplemental heat potentially necessary to ensure sufficient temperatures for
virtually complete thermal dissociation of un-oxidized NH3.

Key advantages of the process, in the case of a reductive TGU utilizing a regenerable amine for H2S ab-
sorption, are summarized as follows:

    Reduced Claus train size (or increased sulphur recovery capacity)
    Reduced natural gas usage in the tail gas treating unit
    Reduced CO2 load on the tail gas treating unit amine
    Optimization of the air/NH3 ratio independent of Claus stoichiometry

With regard to item the optimization the air/NH3 ratio independent of Claus stoichiometry, an inherent
problem with special furnace configurations which permit abnormally high NH3 rates relative to H2S is that
Claus stoichiometry limits the air/NH3 ratio so that residual tail gas H2 is often excessive, to the point of
resulting in unstable operation of the downstream incinerator due to sporadic secondary combustion
within the stack.

In some cases, process conditions may make one or both of the following options beneficial:

    NH3 Cracking Catalyst:
    When necessary to minimize NOx, maximize NH3 conversion and/or maximize H2 yield, a high-
    temperature NH3 cracking catalyst will permit partial oxidation at lower air/gas ratios. Such catalysts
    typically contain nickel, iron or both on a suitable substrate, and are well known.

    Tail Gas Recycle:
    Recycling a minor portion of the reduced tail gas from downstream of the contact condenser to the
    NH3 burner will directionally reduce NOx formation by a combination of temperature moderation and
    reduction by H2S.

A New Approach - Oxidative Tail Gas Units

In the alternative oxidative TGU scheme, a fuel is typically burned, with excess air, so that subsequent
combination of the flue and Claus tail gas streams will result in a net temperature sufficient for thermal
oxidation of all combustible sulphur compounds. In some cases the resultant SO2 can be discharged to
atmosphere, but environmental regulations will more typically require that the SO2 be recovered by an

                                                                                                                7 of 12
absorption medium. In the latter case the hot tail gas stream will typically be cooled in, for example, a
waste heat boiler prior to gas/liquid contact in the absorber.

In this case (Figure 3) the NH3 gas is combusted sub-stoichiometrically in zone 1, thus supplanting most
of the fuel otherwise required, usually a hydrocarbon gas. The ratio of air to NH3 gas is automatically ad-
justed to oxidize most of the NH3, while maintaining sufficiently reducing conditions to avoid, or at least
minimize, NOx formation.

A supplemental fuel, typically a hydrocarbon gas, is combusted with excess air in parallel zone 2 at a rate
necessary to achieve the desired temperature elevation of the Claus tail gas stream. When no supple-
mental heat is required, a minimum fire is preferably maintained to facilitate a prompt increase in the firing
rate in the event of NH3 gas curtailment.

Zone 1 and 2 combustion products are combined in zone 3, where residual H2, potential CO and potential
H2S from zone 1 are oxidized by excess O2 from zone 2.

Claus tail gas is thermally oxidized by combination with zone 3 effluent gas in zone 4, where supplemen-
tal air is also injected if necessary.

Process Description - RAC™ in a Reductive Tail Gas Unit

In conventional Reducing Gas Generator, the tail gas from the final condensers of the sulphur recovery
units enters the hydrogenation section through the Reducing Gas Generator. The reducing gas generator
has the dual purpose of heating the tail gas to a temperature that will permit the desired hydrogenation
and hydrolysis reactions to proceed in the reactor and to supply reducing gases, H2 and CO, to supple-
ment those present in the tail gas. These functions are carried out by the combustion of natural gas with
air supplied by the RGG Combustion Air Blower at substoichiometric conditions. Hot combustion products
are mixed with the tail gas, and the resulting stream flows to the Hydrogenation Reactor.

In the Hydrogenation Reactor, sulphur compounds are converted to H2S by the hydrogenation and hy-
drolysis reactions described above. These reactions are exothermic creating a temperature rise across
the catalyst bed.

In a typical reductive TGU (Figure 2), a NH3 gas stream, potentially containing minor but significant con-
centrations of H2S, mercaptans, HCN, hydrocarbons or other organic contaminants is combusted sub-
stoichiometrically to generate usable heat and H2. An auxiliary fuel stream such as natural gas, for exam-
ple, may be combusted to the extent necessary to satisfy downstream process demand for heat and H2,
The auxiliary fuel is combusted in zone 1 of the combustion chamber, and the NH3 gas is combusted in
downstream zone 2. Combustion air may be sub-substoichiometric, stoichiometric or super-stoichiometric
with respect to the auxiliary fuel, except that super-stoichiometric combustion is not advisable in the ab-
sence of downstream NH3 combustion.

The combined combustion gases are subsequently combined with the Claus tail gas to achieve a net
temperature of 550-650°F (290 - 340°C), following which the stream is passed through a fixed bed of
catalyst which has been impregnated with oxides of cobalt or nickel and molybdenum. The catalyst

                                                                                                                 8 of 12
serves to hydrogenate SO2 and Sx vapor to H2S, hydrolyze COS and CS2 to H2S, hydrolyze CO to CO2
and H2, and reduce NOx to NH3.

The approach to stoichiometric air for the combustion of the NH3, and potentially the auxiliary fuel, is lim-
ited as necessary to generate sufficient H2 for complete hydrogenation of SO2 and Sx to H2S. Consistent
with established practice, a typical residual H2 concentration of 1-3% on a dry, molar basis is maintained
downstream of the reactor.

Also consistent with established practice, the reactor effluent gas, nominally comprised of 1/3 water va-
por, is optionally cooled to 300-350°F (150-175°C) by indirect waste heat steam generation, then further
to 80-100°F (27-38°C) by direct contact with a recycle water stream in the contact condenser, which also
serves to condense most of the water vapor and absorb most of any NH3 which may be present. Surplus
recycle water is continuously purged to the plant's sour water collection system.

The cooled, relatively dry gas is then contacted with a suitable absorption medium for recovery of most of
the H2S. In the common case where an amine solution is used, for example, any NH3 not absorbed in the
contact condenser will be absorbed by the amine and ultimately concentrate in the regenerator reflux, a
slipstream of which can be purged to sour water if necessary to avoid excessive buildup.

Process Description - RAC™ in a Oxidative Tail Gas Unit

In an oxidative TGU (Figure 3), the NH3 is similarly combusted sub-stoichiometrically in zone 1, whereby
the air/gas ratio is automatically adjusted to achieve a residual H2 concentration which has been found
sufficient to ensure that NOx formation is minimized.

In this case supplemental fuel, typically a hydrocarbon gas, is combusted with excess air in zone 2 to the
extent necessary to achieve the desired downstream temperature of the Claus tail gas. When no supple-
mental heat is required, a minimum standby fire is preferably maintained so that the firing rate can be
promptly increased in the event of NH3 gas curtailment.

Combustion gases from zones 1 and 2 are combined in zone 3, where residual H2 and potential CO, H2S
and other miscellaneous combustibles are thermally oxidized by excess O2 from zone 2.

Zone 3 effluent gases are combined with the Claus tail gas stream in zone 4 to achieve the net average
temperature necessary for the desired oxidation of combustibles, which will typically include H2S, Sx,
COS, CS2 and CO. The temperature required for said thermal oxidation will typically be in the range of
800-1500°F (425-815°C), depending on prevailing environmental regulations, residence time and the na-
ture and concentration of key combustibles. If necessary, additional combustion air may be injected to
supplement residual O2 in the zone 3 effluent. A typical target would be 1-3% residual O2 on a molar wet
basis in the combined tail gas stream.

In some cases, environmental regulations will permit the discharge of the hot oxidized tail gas to the at-
mosphere. More commonly, the oxidized tail gas will be subsequently cooled by various conventional
means such as, for example, indirect heat transfer to generate waste heat steam, followed by SO2 recov-
ery using various established processes.

                                                                                                                9 of 12
Figure 2. WorleyParsons Ammonia Destruction in a Reductive Claus Tail Gas Treating Unit

                                                                                          10 of 12
Figure 3. WorleyParsons Ammonia Destruction in an Oxidative Claus Tail Gas Treating Unit

                                                                                           11 of 12

Ammonia Destruction in a Claus Tail Gas Treating Units is an alternative solution where the high ammo-
nia content needs to be processed beyond 30 to 35% of the total Claus feed on a wet basis. As described
in this paper, the cost of other alternatives will normally be substantially greater than this new invention.
Considering crude oil from several areas contains high level of nitrogen which results a high level of am-
monia content that needs to be processed. Two preferred methods are described and could be applied
where they are applicable.

Ammonia Destruction in a Claus Tail gas Treating has the following advantages.

1.    The sulfur recovery and tail gas unit together have the capabilities of ammonia destruction beyond
      30 to 35% for nay new or existing Claus unit.

2.    A proven and established process and do not require any pilot testing

3.    Flexibility of mode of operation with air only, on oxygen-enriched air, or on oxygen only, which in-
      volves only the Oxygen supply system.

4.    Provide plant flexibility for turndown, ability to handle different feeds, and changes in production

5.    Provide high reliability and easy maintenance

6.    Reduce Plot Space with fewer modifications, less impact on downstream equipment

7.    No Concerns for plugging downstream equipment

8.    No oxygen enrichment required for such high ammonia content

9.    Reduce Plot Space with fewer modifications, less impact on downstream equipment for revamps

10.   Reduce Capital and Operating Costs compare to alternatives


1.    Rameshni M., Cost effective Options to Expand SRU Capacity Using Oxygen, Sulphur Recovery
      Symposium, Brimstone Engineering Services, Inc., Banff, Canada, May 2002.

2.    Rameshni M., Challenges for SRU Expansion with Oxygen, Symposium, Brimstone Engineering
      Services, Inc., Vail, Colorado, USA, September 2002.

                                                                                                                12 of 12

To top