3.3 Sour oil and gas management - Treccani

3.3.1 Introduction

Petroleum streams containing hydrogen sulphide, H2S,in amounts greater than a few ppm are referred to as‘sour’, a term whose origin goes back to the earliestdays of the oil and gas industry, when the oil wastasted to verify whether or not it contained thiscompound. Hydrogen sulphide is a colourless, highlytoxic and dense gas (density 1.39 g/l at 25°C and 1bar; boiling point �60°C) which is readily soluble inhydrocarbons. It is characterized by a foul odour,detectable at 1 ppm, but tends to rapidly saturate thehuman olfactory system, making it particularlypernicious. The 8 h exposure limit is 10 ppm, andexposure to 600 ppm rapidly provokes respiratorydisturbance and death.

Not surprisingly, sour petroleum resources havealways posed special problems to operators. If in thepast, a highly sour resource (containing several mole%or more of H2S in the associated gas) might have beenconsidered unattractive for development relative toother assets, today operators are developing fieldscontaining as much as 30% H2S. The interest for theseresources reflects several intersecting trends, includingthe diminishing availability of easy-to-produceconventional oil resources, increased natural gas (NG)prices in North America, and the opening of Russia,Central Asia and the Middle East to internationaloperators. Their sour petroleum resources, in fact, havehistorically characterized these geographical regionsand international oil and gas companies must be readyto face the challenges of highly sour production inexchange for the opportunity to operate in these richpetroleum provinces.

Hence, the development of sour oil and gasresources is rapidly emerging as a major industrial andtechnological theme. The problems posed by sourreservoirs include, in the first place, health and safety

environment (HSE) issues, which accompany alloperations, from the exploration phase through to fieldabandonment. In fact, the high toxicity of H2S requiresspecial operating procedures to ensure worker safetyduring drilling and in production operations. The usualby-product of processing sour oil and gas is elementalsulphur, a bright yellow solid that melts at 120°C,which also must be disposed of correctly, so as toavoid impact on the area surrounding the site ofproduction.

The presence of H2S, moreover, impacts negativelyupon the economic value of an asset, with the need toemploy special materials for all fluid streamscontaining H2S and humidity; the unit cost ofcorrosion-resistant metal alloys is as much as ten timesgreater than that of the carbon steel that would beemployed in a sweet oil development. The processesused for NG sweetening also add considerably tocapital and operational expenditure; for a remote gasfield, the economic limit for gas sweetening usingpresent technologies may be reached for NGcontaining c. 15 mol% H2S (see below).

The conventional technology train for thetreatment of a sour gas stream is shown in Fig. 1 A.This consists of a sweetening step, where the H2S isremoved from the gas, followed by the SulphurRecovery Unit (SRU). Central to the SRU is theClaus plant, in which the H2S is converted toelemental sulphur. The gaseous effluent from theClaus plant still contains 2-5% of the total sulphur,which must be further processed in a TGU (TailGas Unit) to meet emissions standards for SO2.Present technologies can be pushed to recover asmuch as 99.9% of the total sulphur separated fromthe NG. Where high total volumes of H2S and/orhigh concentrations must be treated, new solutionsare being sought in order to reduce developmentcosts (Fig. 1 B).

237VOLUME III / NEW DEVELOPMENTS: ENERGY, TRANSPORT, SUSTAINABILITY

3.3

Sour oil and gas management

One of the most important developments, whichoriginated in North America and is only nowbeginning to be applied internationally, employsre-injection of the separated acid gas stream into thereservoir or a suitable, sufficiently deep saline aquifer.In this way, the costs associated with the Claus plantand the TGU are eliminated and the production ofelemental sulphur is avoided. At Karachaganak, Eni asjoint operator has adopted a variant on this process:the H2S removed from that part of the associated gasstream utilized for energy requirements and for gasexport is mixed with the remaining raw gas andre-injected into the upper part of the reservoir. The c. 500 bar injection pressure constitutes the worldrecord for this type of operation.

Another major focus of industrial R&D regards thedevelopment of new, low-cost processes capable ofeffecting a partial, or bulk, removal of H2S fromhighly sour gas streams. Such processes would beinstalled upstream of a traditional sweetening plant,reducing the volume of H2S entering the latter, andappear particularly promising where they can becombined with acid gas re-injection (see Fig. 1Bagain). Many view the development of bulk separationtechnology as a prerequisite for the economicdevelopment of highly sour gas reservoirs, and themajor oil companies are active in this area.

The sulphur that emerges from the SRU unit isanother source of problems for the operator,particularly in remote geographical locations. Thoughconstituting one of the principal feedstocks of modern

society (annual consumption is about 60 million tons),there is a chronic glut of sulphur on the world market.Sulphur produced in remote locations is stranded, andmust be stored; this is done by forming large, million-ton blocks of solid, elemental sulphur. The prospect oflong-term sulphur storage is driving a search for newstorage technologies, or permanent disposal solutions,capable of reducing the long-term environmental andeconomic liabilities associated with block storage.Interest continues also in the development of newcommercial uses for sulphur (see below).

A further word must be said about smaller volumeoperations (H2S 1-10 t/d) such as those found in Italyand elsewhere, where H2S is removed and converted tosulphur via a redox process. The consolidated redoxtechnology has been used for a long time, but the moretons of H2S removed, the higher the costs of thetechnology. It is difficult to operate optimally,produces a foul odour, and supplies sulphur of lowquality that must be disposed of rather than sold.Research for improved processing options, therefore,continues in this area as well.

3.3.2 Origin of H2S in oil and gasaccumulations

The ability to predict the geological H2S risk duringexploration operations would help both to prioritizepotential exploration targets better and to ensure thesafety of drilling personnel. Today, this is possible only

238 ENCYCLOPAEDIA OF HYDROCARBONS

NEW UPSTREAM TECHNOLOGIES

A

B

sour naturalgas

(containingH2S, CO2)

sweeteningunit

natural gasto sales

emissions toatmosphere

elemental sulphur(to sales or storage)

Clausunit

tail gasunit

H2S

CO2

H2S

SO2, CO2

partiallysweetenednatural gas sweetening

unit

natural gasto sales

emissions toatmosphere

elemental sulphur(to sales or storage)

acid gas stream to re-injection

Clausunit

tail gasunit

bulksweetening

process

H2S

CO2

H2S

SO2,CO2

sour natural gas(containingH2S, CO2)

Fig. 1. Conventionalprocessing scheme for large volumes of sourgas (A) and a scheme of an innovative process for bulk sweetening with re-injection of acid gas (B).

to a limited extent, and in areas where considerableknowledge of the subsurface has already been gainedthrough petroleum exploration and development.Reliable prediction of the geological H2S risk requiresdetailed knowledge of the mechanisms governing H2Sgeneration and its migration and accumulation inpetroleum-filled structures.

A number of chemical and biological pathwaysthat can lead to the formation of H2S have beenidentified (Table 1). Of these, the most widespread andimportant are believed to be Biological SulphateReduction (BSR) and Thermal Sulphate Reduction(TSR) (Machel, 2001). BSR is a process mediated bySulphate Reducing Bacteria (SRB) which are activeunder strictly anaerobic conditions. Thesemicroorganisms metabolize hydrocarbons byemploying sulphate ions present in the brine as theelectron acceptor (oxidant) in a process which is faston the geological timescale and most frequently iscontrolled by the limited availability or diffusion rateof sulphate or another nutrient. The conditionsfavourable for BSR are believed to include:temperature between 40 and 80°C, the presence ofdissolved sulphate ions, and low-to-moderate brinesalinity. BSR of crude oil is believed to involve theoxidation of linear saturated hydrocarbons andalkylbenzenes. The H2S produced inhibits BSRactivity above approximately 5 mol% of H2S in the

hydrocarbon phase, which defines the upper limit onthe concentration of H2S that can be produced by thisprocess; because of the existence of pathways leadingto the spontaneous removal of H2S (see below), rarelydoes BSR lead to H2S concentrations in excess of1-2%.

While the considerations above refer to thepetroleum quality in a reservoir prior to itsdevelopment; BSR can, under favourable conditions,take place on the shorter timescale of oil productionoperations. In fact, operators in the North Sea and theGulf of Mexico are all too familiar with thephenomenon of reservoir souring, in which sulphateions in the injected seawater are used by SRB,probably introduced together with the seawater, tometabolize the water-soluble hydrocarbons. This canlead, in extreme cases, to the generation of severalthousand ppm H2S in the produced fluids, creatingtremendous problems for the operator. In fact, seriouscorrosion problems and expensive platform retrofittingcan be the result if the reservoir development plan hasbeen based on the expectation of a sweet petroleumstream. A similar problem can be encountered whenproducing reservoirs containing sulphate-rich brinesfrom reservoirs too hot for BSR. Once the producedfluids cool to below 80°C, SRB microbial activitybecomes a risk and can generate troublesome levels ofH2S in the surface facilities and transport lines.


SOUR OIL AND GAS MANAGEMENT

Table 1. Main sources of H2S in petroleum accumulations

Origin of H2S H2S generation mechanism Main characteristics

Biological sulphatereduction (BSR)

Petroleum�CaSO4(s)�� CaCO3(s)�H2S�H2O�contaminatedpetroleum and bitumen

Maximum T: c. 80°C, inhibitedby high levels of salinityRarely produces concentrationsof H2S in petroleum above a fewmole percent

Degradation of organicsulphur compounds

Hydrolysis of organic sulphur compounds

Generation of H2S limitedby the sulphur contentof petroleum of less than a fewmole percent

Reaction with elementalsulphur 4S�CH4(aq)�2H2O�� CO2(g)�4H2S(g)

Elemental sulphur is presentalmost exclusively as tracesin petroleum reservoirs

Volcanic seepageThe H2S, generated in the deep subsurface, migrates into thereservoir along deep fractures or faults.

Restricted to areasof volcanic activity; rarelyassociated with hydrocarbonaccumulations

Thermal sulphatereduction (TSR)

Petroleum�CaSO4(s)�� CaCO3(s)�H2S�H2O�contaminatedpetroleum and bitumen

Minimum T: 120-140°CRequires the presence ofanhydritesMay generate percentagesof H2S up to 95 %

Present understanding of these phenomena isincomplete, and research efforts continue toinvestigate the factors which govern the reservoir andsurface souring processes (Burger et al., 2005).

As noted above, TSR is the dominant mechanismresponsible for the generation of high concentrationsof H2S in petroleum reservoirs, and has beenimplicated in the formation of NG reservoirscontaining above 95 mol% H2S. The presently limitedunderstanding of this process can be attributed both tothe complexity of the reactions involved and to thedifficulty of reproducing the high temperature,pressure conditions and long time-scale of TSRprocesses in the laboratory. Nevertheless, through acombination of laboratory and field case studies, thegeneral picture of this process has been developed(Machel, 2001; Cross et al., 2004).

TSR involves the reaction of hydrocarboncompounds with sulphate minerals (anhydrites), suchas evaporitic beds or other mixed lithologies, in thepresence of brine. These reactions proceed on ageological timescale and modification of thepetroleum accumulation may be only partial. Theminimum temperature at which this process becomesimportant is estimated by various authors to liebetween 120 and 140°C. Petroleum quality, thesulphate dissolution rate and brine availability arebelieved to influence the rate of TSR, while pH, thepresence of H2S and other minerals are thought tocatalyze the process. Deposition of the solid CaCO3

by-product of reaction may inhibit complete reactionof the available sulphate minerals. The chemicaltransformations generally invoked are indicated belowfor methane and higher saturated hydrocarbons.

In the case of methane, the following reactionsoccur:

CaSO4(s)��

��Ca2�(aq)�SO2�

4(aq)

CH4(aq)�SO2�4(aq)�H�

(aq)��

��HCO�3(aq)�

�H2S(g)�H2O

Ca2�(aq)�HCO�

3(aq)��

��CaCO3(s)�H�(aq)

In the case of ethane and higher alkanes, however,the reactions are as follows:

CaSO4(s)��

��Ca2�(aq)�SO2�

4(aq)

C2H6(aq)�2SO2�4(aq)�2H�

(aq)��

��2HCO�3(aq)�

�S �H2S(g)�2H2O

Ca2�(aq)�HCO�

3(aq)��

��CaCO3(s)�H��(aq)

4S �CH4(aq)�2H2O��CO2(g)�4H2S(g)

As indicated for the higher alkanes, elementalsulphur is a product of the TSR reaction. Although itcan further react with CH4 to generate H2S, it can alsobe chemically incorporated into other components of

the petroleum, as further discussed below, althoughsmall but significant amounts may remain in thereservoir as elemental sulphur. Other routes to sulphurproduction include the following:

CaSO4�3H2S�� CaO �3H2O �1/2S8

TSR processes take place under conditions oftemperature in which thermal maturation occurs, andcontribute to the change in the oil quality. Thesaturated hydrocarbon compounds present inpetroleum are preferentially consumed by TSRreactions, with methane being the least reactive ofthese. TSR consequently results in a relative increasein methane and in the aromatic components of thecrude at the lower range of temperature for TSR.Nevertheless, the correlation of TSR with API gravityis weak because benzothiophenes and heavier aromaticsulphur compounds may be formed as the result of thereaction of the elemental sulphur by-product withhydrocarbons, offsetting the consumption of heavierhydrocarbons. The increase in the percentage ofmethane may favour precipitation of the asphalteniccomponent, with negative consequences for reservoirquality. In advanced stages and at a highertemperature, TSR favours the formation of dry naturalgas, with an H2S content of 95 mol% or even higher.Under these conditions significant amounts ofelemental sulphur may also be present in the gasphase.

The amount of H2S found within a reservoir isdetermined not only by the amount of H2S generatedby TSR reactions, but also by the presence of pathwayscapable of removing it from the oil during migrationor within the reservoir. The presence of iron or othertransition metal ions in the brines or mineral phaseswith which the H2S comes into contact can scavenge aconsiderable amount of H2S through the formation ofmetal sulphides (e.g. pyrite). Common iron mineralsinclude siderite (FeCO3) and magnetite (Fe3O4), andthe clay phase chlorite; other clay phases mayincorporate iron ions as well (Aagard et al., 2001).Thus, in general, clay-rich porous media tend to beless prone to high H2S accumulations. For the samereason, carbonate rocks containing anhydrites are thebest candidates for high H2S accumulations. Anotherpathway for H2S removal, dissolution into the brinephase, can be important where the aquifer ishydrodynamically active.

The distribution of H2S within a reservoir can bequite heterogeneous and cases of extreme, systematicvariability across a structure are known (Al-Eid et al.,2001). Several mechanisms can bring about thisphenomenon. One is the migration of H2S into areservoir successive to hydrocarbon charging, in whichcase the slow homogenization of the H2S may be



incomplete. A heterogeneous distribution of H2S alsocan be produced where there are hydraulically isolatedcompartments within the reservoir, or where clays andother H2S scavenging minerals are unevenlydistributed within the structure.

Sulphur isotopic analyses on H2S, organic sulphurcomponents of the oil and reservoir mineral phases(e.g. anhydrites and pyrite), when used in conjunctionwith other information, constitute a particularlypowerful tool for determining the likely origin of theH2S within a structure. The isotopic parameter d34S isdefined according to:

(34S/32S)sampled34S ��11124111�1� �1.000(34S/32S)standard

where d34S is assigned a value of zero in the standard,which is troilite found in the Diablo Canyon meteorite.In fact, where TSR operates, the value of d34S for H2Sshould be very similar to that of the anhydrite source,and in fact can be used to identify that source. Inseveral cases, sulphur isotopic data have been used,together with other geological information, to infer theminimum temperature for onset of TSR in thepetroleum system (Manzano et al., 1997). Whereas thed34S parameter of the organic sulphur compounds inan oil not subjected to TSR will have the same value asthe source rock, sulphur incorporated chemically intothe petroleum following TSR should tend toward thesame d34S value as the H2S. The BSR pathway for H2Sgeneration, on the other hand, is characterized by alarge isotopic fractionation between sulphate andsulphide.

3.3.3 H2S removal in smallvolume plants

Desulphurizing hydrocarbons to remove H2S in excessof the required specifications is one of the mostexpensive aspects of the treatment of acid gas streams.Being a highly toxic gas, H2S must not only beremoved from the hydrocarbon stream withefficiencies close to 100%, but also turned into a lesshazardous species, usually elemental sulphur, withadditional costs (Gas […], 1994). In large treatmentplants which remove dozens of tons of H2S a day ormore, Claus processes are normally used. For smallervolumes the Claus process is uneconomic, and H2Stransformation makes use of other, generally lesseffective, conversion systems. In these instances, thetechnological options available are essentially of twotypes: non-regenerated sorbents and redox processes(Kensell and Leppin, 1996; Quinlan, 1996). One of themost effective parameters for the selection of theoptimal process is the plant’s capacity in terms of

removed sulphur. Fig. 2 shows an estimate of the costof treating an NG stream as plant capacity varies. Inaddition to showing the rapid rise in the cost oftreating the gas as the H2S content increases, thediagram highlights the need to use differentdesulphurization technologies depending on thequantity of H2S to be removed.

To remove a few dozen to several hundred kg ofH2S per day from acid gaseous hydrocarbon streams,non-regenerated chemical sorbents are generally used(Foral and Al-Ubaidi, 1995). The acid gas comes intocontact with solid or liquid chemicals which react withthe H2S to produce a spent stream which must bedisposed of. The sorbent may form the packing in anabsorption column, as in the Sulfatreat and IronSponge technologies, or, for liquid reagents, a washingcolumn may be used; alternatively, the reagent, usuallya triazine or less commonly an amine, may be injecteddirectly into the acid gas stream inside the gaspipeline. Injection into the pipeline requires small andsimple equipment (a liquid/gas mixer at the entranceto the pipeline and a separator at the exit) but, due tothe cost of the chemicals, it is not viable for volumesabove a few dozen kg of sulphur per day. In general,the use of non-regenerated solid or liquid sorbentsmakes it possible to reach the required qualityspecifications for gas with low capital costs. However,operating costs are high, up to or over 15 euro per kgof sulphur produced. These costs are consideredacceptable if the quantity of sulphur to be removed issmall, but become unsustainable when tons of sulphurper day must be removed from the gas streams.

Regenerable reagents have been developed sincethe 1940s, but large numbers of desulphurizationprocesses based on regenerable reagents have beenavailable to the oil industry only since the 1970s. Mostof these processes are based on reversible oxidereduction in the liquid phase; they are therefore known



trea

tmen

t cos

t ($/

m3 )

10�3

10�2

10�1

sulphur production (t/d)

adsorbents

Claus technology

redox technologies

0.1 1 10 100

Fig. 2. Desulphurization costs for a natural gas production facility (capacity = 106 Sm3/d).

collectively as ‘liquid phase redox processes’. Redoxprocesses (Connock, 1996) can be used for productionvolumes lying in the range of a few hundred kg ofsulphur per day up to the lower economic limit ofamine/Claus processes, about several dozen tons ofsulphur per day.

Fig. 3 shows the typical scheme of a redox process.In the absorption stage, the acid gas comes intocontact with the absorbent solution; the H2S reactschemically with the solution and the sweetened gasexits the reactor. In most cases, the solution contains areagent which oxidizes the H2S to sulphur; in others,oxidation occurs at a later stage. After the oxidation ofthe absorbed H2S, the partially spent circulatingsolution is sent to the regeneration stage, where itsoxidization capacity is regenerated with air throughoxidation reactions which are generally catalyzed. Aseparation stage, either before or after regeneration,recovers the sulphur produced in the form of aconcentrated suspension. After the replacement of thecirculating solution to re-establish the originalcomposition and volume, it is returned to theabsorption reactor for a new cycle.

One of the first redox processes developed was theStretford process (Keene, 1989). Based on theabsorption of H2S in an alkaline solution followed byoxidation to sulphur with the redox pair(VO2

�)/(V4O22�), the Stretford process covered a

significant portion of the redox desulphurizationmarket for many years. For environmental and safetyreasons, the presence of vanadium in the redoxsolution led to the abandonment of the construction ofnew facilities in favour of safer and moreenvironmentally-acceptable processes based, forexample, on the redox pair Fe2+/Fe3+. Liquid phaseredox processes using solutions of iron salts areprevalent in the desulphurization capacity rangebetween 0.5 and 20 t/d of sulphur. Hundreds of plantshave been built with the LO-CAT (Merichem GTP,

ARI, Wheelabrator, US Filter and others) and Sulferox(Dow Chemical) technologies (Mamrosh and Allen,1994, Nalg, 1995).

Despite the commercial success of these plants,redox desulphurization processes continue to presenthigh costs and various operating problems. Highcapital expenditure is required for the materialsnecessary to resist chemical attack by the oxidizingsolutions; the reactors must therefore often be built ofstainless steel. Operating costs are high mainly due tothe costs of the chemicals, which contribute todesulphurization costs to the tune of several hundredeuro, and in some cases over a thousand euro, per tonof sulphur produced. Furthermore, operating problemsencountered include the tendency to form stablefoams, to deposit sulphur in the transport lines, and theproduction of sulphur which does not meetcommercial specifications, and must be disposed of,for example, by landfilling. These limitations becomeeven more significant for high-pressure applications oroffshore platforms.

The response to the problems presented bycommercial redox processes has been thedevelopment of new processes which aim to simplifyredox chemistry, reduce operating costs and makeplant management less problematic (Quinlan et al.,1997). Some of these new processes use organicsolvents to avoid the problems associated with thepresence of sulphur dispersed in aqueous streams.Examples include the Crystasulf process developedby Radian International and Hysulf processdeveloped by Marathon Oil. In other processes, likeThiopaq, originally developed by Paques, and later incollaboration with Shell Global Solution, a basicaqueous solution is used to sweeten the acid gas. Thisprevents elemental sulphur from being formed in theabsorption reactor; instead, sulphur forms during theregeneration phase, presenting less serious operatingproblems, especially for pressurized applications.The Thiopaq process belongs to the family of redoxprocesses which use biological reactors (Anders andWebb, 1995). In Thiopaq plants, the regeneration ofthe absorbent solution occurs in a bioreactor wherethe sulphide deriving from the absorption of the H2Sinto the basic solution is oxidized by aerobicmicro-organisms to sulphur and in part to sulphateminerals. The alkalinity used for absorption, whichwould otherwise render the process uneconomic, isthus regenerated. The Desulfeen process, developedby EniTecnologie (Gianna et al., 2004) also employsa biological reactor, in this case for the regenerationof the absorption solution. In the Desulfeen process,based on the redox pair Fe2+/Fe3+, the circulatingsolution, acidified with sulphuric acid, is regeneratedin a bioreactor containing Thiobacillus ferrooxidans



absorptionand oxidation

of H2S

sweet gas

sourgas

solutionmake-up

chemicals

redoxsolution

regeneration

air

sulphurseparation

sulphur

Fig. 3. Diagram of the workings of redox processes.

bacteria. This choice, made possible by using strainsof Thiobacillus ferrooxidans able to reoxidize ironunder acid pH conditions, entails a series ofoperating and economic benefits: less foam, betterseparability of sulphur, the low cost of the chemicalsand higher quality of the sulphur produced.

As an alternative to liquid phase redox processes,some suppliers are developing technologies based onthermal processes adapted to small volumes, whichmake use of the fastest H2S oxidation kineticsobtainable at relatively high temperatures. An exampleis the direct oxidation process offered by Sulfatreat, inwhich the acid gas is sweetened by oxidizing the H2Swith air, mixed with the gas in stoichiometricquantities in a catalytic reactor. The percentages ofH2S removed (around 90%) are far from thoseobtainable with non-regenerable sorbents or liquidphase redox processes, many of which allow acidhydrocarbon streams to be desulphurized to H2Sconcentrations below 1 ppm vol. However, theadvantages are significant: the absence of circulatingsolutions, the multi-year life of the catalyst and theproduction of extremely pure liquid sulphur.

Many new low-volume desulphurizationtechnologies have been developed to thepilot/demonstration plant scale, and in some caseshave been used in industrial plants. However, despitethe availability of numerous new processes (thosementioned are only a few of those discussed in thespecialized literature), the commercial scene continuesto be dominated by conventional processes. Thissituation may change in the near future; under pressurefrom environmental regulations on sulphur emissions,a growing fraction of acid hydrocarbon streams in thepetroleum industry are destined to be desulphurized,creating a significant increase in the demand fordesulphurization capacity in the interval suited toredox processes. It is likely that in a cost-consciousmarket such as that for gas treatment, the increase indemand may pave the way for the emergence of moreinnovative processes.

3.3.4 High volume sour gastreatment processes

Sour petroleum streams and large total H2Svolumes to process constitute special challengesfor oil and gas operators, on account of the highcapital and operating costs of the consolidatedtechnologies, and the need to manage the largevolumes of elemental sulphur produced,particularly in remote locations. In Canada, theMiddle East and Central Asia facilities areoperating which treat sour gas streams containing

as much as 30% H2S, and total amounts as high asseveral thousand tonnes of H2S per day.

Sour gas sweetening

The pipeline specification for NG generallyrequires an H2S level of 5 ppm or less. The upgrading,or ‘sweetening’, of a sour natural gas is accomplishedusing processes which selectively remove the H2Sfrom the NG, with high hydrocarbon recovery. Theacid gas thus removed, which may also containsignificant amounts of CO2, COS, CS2 andmercaptans, is sent to a SRU (e.g. Claus plant) or,increasingly, to re-injection (see Section 3.3.5).

The main processes used for NG sweetening arebased on selective absorption using solvents,although adsorption using molecular sieves is alsoused for some intensive purification processes. Thechief characteristics of the consolidated processesare discussed briefly in this section (for a moredetailed treatment, see Chapter 3.2). These and otherprocesses are the object of industrial R&Dactivities, pursued with the aim of reducing the costof producing highly sour (H2S 10-30 mol% orhigher) petroleum resources. As mentioned above(see Section 3.3.1), most efforts are currentlyfocussed on the bulk sweetening of highly sour NGupstream of the final sweetening plant, and envisionre-injection into the reservoir or another deepgeological structure, rather than Claus processing, ofthe acid gas separated.

Consolidated sweetening processes

Absorption processesThe selectivity of solvent-based sweetening

processes for sour natural gas relies on either achemical or physical affinity. Chemical solvents reactwith the acid gas components to form loosely-bondedchemical complexes. On heating at reduced pressure,these complexes dissociate and release the acid gasfrom the solvent. Physical solvents absorb acid gasesby dissolution. After absorption, the acid gases may bereleased from the solvent by reducing the pressure,thermal stripping, or a combination of both. Recently,blends of physical and chemical solvents have beendeveloped which combine some of the advantages ofboth; these are also commonly regenerated by thermalstripping. In each case, the regenerated solvent isrecycled to the absorber, while the acid gas streamflows to the SRU. The choice of solvent is based on thegas composition, expected sweet gas specifications,requirements of the SRU, etc. For a list of the majorsolvent families and the blends presently in use for NGsweetening, see Chapter 3.2.



The most widely used chemical solvents forthe removal of acid gases from NG streams arealkanolamines (referred to generally as aminesolvents), employed as aqueous solutions. Thesechemical solvent processes are particularlyapplicable when acid gas partial pressures are lowand/or low levels of acid gas are desired in theresidue gas. Because of the low hydrocarbonsolubility in the aqueous solution, these processesare particularly effective for treating gases rich inheavier hydrocarbons. Some alkanolamines canbe used to selectively remove H2S in the presenceof CO2.

The basic chemical processes involved in H2Scapture by amines are illustrated below:

[1] 2RR�R�N�H2S��(RR�R�NH)2S

[2] (RR�R�NH)2S �H2S��2RR�R�NH� HS�

where R denotes an alkanol group and R� and R� canbe either alkyl or alkanol groups, hydrogen, or amixture of the two depending on whether the amine isprimary, secondary or tertiary. Alkanolamines alsoreact with carbon dioxide in two ways:• formation of carbonate and bicarbonate

[3] 2RR�R�N � CO2�H2O��(RR�R�NH)2CO3

[4] (RR�R�NH)2CO3�CO2��H2O��

��2RR�R�NH�CO3H�

• formation of carbamate

[5] 2R�R�NH �CO2��

��R�R�NCOO�R�R�NH2�

While reactions [1] and [2] are fast, [3] and [4] areslow.

Amine chemical structure has an importantimpact on reactivity with H2S and CO2, and can beused to enhance selectivity for H2S. Stericallyhindered amines are the most basic and have thefastest reaction rates with H2S, but are less prone tocarbamate formation (carbamate formation isentirely excluded with tertiary amines). Benefits ofselective H2S removal include the need for reducedsolution flow rates, higher H2S concentrations in theseparated acid gas and smaller amine regenerationunits. In many industrial applications, removal ofH2S from the gas stream is the only objective of theprocess, such as in streams with low ratios of H2S toCO2 (as in Claus plants for the treatment of tail gas),and in NG with a CO2 content at or below the salesgas specification.

The basic process for amine scrubbing involvescontacting the gas (at elevated pressure) with thesolvent counter-currently in a packed or trayedabsorption column (Fig. 4). A significant portion ofthe co-absorbed hydrocarbons is recovered byreducing the pressure of the rich amine in a flash

drum. Finally, the chemical solvent is regenerated,after heat exchange and filtration, in a distillationcolumn which operates at, or near, 1 bar.

Processes using physical solvents are based onphysical absorption of the acid gases and employ aprocess scheme resembling that for the aminesolvents (see Fig. 4 again), with the addition of aseries of flash units between the absorption andregeneration columns. Physical solvent processes arefavoured where the partial pressure of the acid gas inthe feed is high (greater than c. 345 kPa), heavyhydrocarbon content is low and selective removal ofH2S is desired. Some follow-up treatment orblending is generally required to achieve pipelinespecifications for H2S. The interest in theseprocesses derives from the low energy requirementsfor regeneration, which is achieved by multi-stageflashing at low pressures, by stripping with inert gasat low temperature or by heating and stripping withsteam or solvent vapours.

Physical solvents generally can remove COS,CS2, and mercaptans as well as H2S while, in certaininstances, simultaneously dehydrating the gas. Theabsorption step is operated at, or lower than,ambient temperature in order to enhance thesolubility of the acid gases. The solvents arerelatively non-corrosive, so carbon steel can be usedin the plants. However, the physical solvents alsoabsorb heavy hydrocarbons from the gas stream,resulting in high hydrocarbon content in the acid gasstream as well as possibly significant hydrocarbonlosses. On the main physical absorption processes,see Chapter 3.2.

One of the most recent developments in naturalgas sweetening technology is the use of hybridsolvents which employ a mixture of chemical andphysical solvents in order to combine the favourablecharacteristics of each (e.g. selectivity for CO2 ormercaptans), and may be tailored to the needs of thespecific application. As an example, the mixtures ofsulpholane with diisopropanolamine (DIPA) ormethyldiethanolamine (MDEA) employ the physicalsolvent sulpholane in an aqueous medium, and makeit possible to remove H2S, CO2, COS, CS2,mercaptans and polysulphides. DIPA is used wherecomplete removal of H2S, CO2 and COS is desiredwhile MDEA is preferred for the selective removalof H2S in the presence of CO2, with partial removalof COS. The main advantages reside in the lowenergy requirements, high acid gas loadings and lowcorrosivity of the solvent. The disadvantages arethose typical of physical and chemical solvents: co-absorption of heavy hydrocarbons in the NG streamand the need to regenerate the chemical solvent forgas streams containing CO2.



Adsorption processesAdsorption processes employing 13X and 5A

molecular sieves have been used for a variety ofapplications where very high gas purity is required.This technology provides simultaneous water and acidgas removal down to very low water contents (0.1 ppmvol). Large-pore molecular sieves such as 13X alsoprovide separation of mercaptans. In the presence ofCO2, however, 13X molecular sieves tend to catalyzethe formation of COS by reaction between H2S andCO2. Commercial units with a capacity up to 6·106

Sm3/d (standard m3 per day) are in operation. Theseunits may have two or three adsorber beds, and themolecular sieves are regenerated thermally. Traces ofglycol, glycol degradation products or oil can poisonthe molecular sieve.

Bulk sweetening processesThe movement of the industry towards re-injection

of acid gases as a means of reducing the volume ofelemental sulphur produced in petroleum operationshas important implications for NG sweeteningprocesses. Where moderate H2S volumes are treated, itmay be convenient to employ conventional sweeteningprocesses with the treated gas exiting at pipelinespecification and the separated acid gas stream sent tore-injection. For very large volume operations,however, the industry is exploring the development ofpartial, or bulk sweetening processes to be installedahead of a final, conventional sweetening process(Fig. 5). The underlying concept of this research is thatit may be possible to develop bulk sweeteningprocesses capable of removing a large part (50-95%)of the H2S at a much lower unit cost than theconventional processes.

This technology may be the key to theeconomic development of highly sour gas

resources in the Middle East and Central Asia,which would otherwise have treatment costsapproaching the commercial value of thesweetened NG. To illustrate this point, Table 2reports an estimate of the total costs for sweeteningan NG (c. 7 million m3/d) containing 19 mol% of



sweetgas

optional equipment

sourgas

acid gas

abso

rber

stri

pper

outletseparator

leancooler

rich/leanexchanger

flashgas

flashtank

reboiler

condenser

refluxseparator

reclaimer

inlet separator

Fig. 4. Simplified process scheme for acid gas removal by amine scrubbing.

Table 2. Costs per unit of gas treatedwith conventional technology

(amine sweetening, Claus, TGU, sulphur storage)for an associated gas containing 19% of H2S

dollars/MBtudollars/t

of sulphur

Materials 0.001 0.118

Facilities 0.19 24.82

Sulphur storage 0.06 8.00

Total,gross variable costs

0.25 32.93

Credit-by-products(steam)

0.11 14.40

Total,net variable costs

0.14 18.53

Fixed costs 0.10 12.59

Total costs 0.24 31.12

Depreciation (10%/y) 0.22 29.35

Production costs 0.46 60.47

Profits (15%/y) 0.34 44.03

Recovered productioncosts

0.80 104.50

H2S, assuming the use of conventional technology(amine sweetening, Claus processing, tail gascleanup and sulphur storage) and a remotelocation. The treatment cost, c. 0.80 dollars/MBtugas treated, is equal to, or greater than, the marketvalue of the NG itself for many remote locations.Thus we can say that today, an approximate cut-offfor the economic viability of a remote sour gasreservoir is around 15% H2S or less. A similarconclusion has been reached by Lallemand andMinkkinen (2002). Although the future marketvalue for remote gas is hard to predict, this analysis

serves to underscore the economic burden oftraditional sour gas processing on the developmentof such resources, and to identify the economic‘prize’ that would be associated with abreakthrough in the treatment of such resources.

Bulk sweetening processes also represent asolution for incrementing the total NG throughputof an existing treatment plant without the need toadd extra SRU capacity. The first example of alarge bulk sweetening process, based on theMorphysorb process at the Kwoen gas plant inNortheastern British Columbia, Canada was built



recycleflashdrum

sour naturalfeed gas solvent pump

upgraded sour gas

acid gas to injection well

filter,separator

recycle-gascompression,

stage 2

recycle-gascompression,

stage 1acid-gas

compression,stages 2-4

acid-gascompression,

stages 1absorbers

acid-gasflashdrum

acid-gasflashdrum

recycleflashdrum

Fig. 5. Schematic of the Kwoen bulk gas sweetening process (Palla et al., 2004b).

Table 3. Bulk gas sweetening at the Kwoen plant

Vapour Acid gas feed Treated acid gas Acid gas

Flow rate (106 Sm3/d) 8.49 7.53 0.96

Pressure (MPa) 7.45 7.38 6.97*

Temperature (°C) 17 13 49

Composition (% mol)

CO2 8.60 7.21 19.60

H2S 13.54 5.33 78.71

CH4 77.26 86.81 1.47

C2H6 0.21 0.23 0.09

C3H8 0.02 0.02 0.02

COS 0.02 0.02 0.05

CH3SH 0.01 0.00 0.04

N2 0.34 0.38 0.00

H2O 0.01 0.00 0.04

* After compression

for precisely this purpose (Palla et al., 2004a,2004b). This process employs a physical solvent,comprised of a mixture of N-formylmorpholine andN-acetylmorpholine, which displays strongselectivity for H2S over CO2 and lowerhydrocarbon losses to acid gas than other physicalsolvents. It also possesses good chemical andthermal stability and is environmentallycompatible. The Kwoen gas plant initiatedoperation in 2002 and is designed for bulk H2Sremoval and disposal of more than 875 t/d ofsulphur through the injection of the separated andliquefied acid gas into a depleted gas reservoir. Theplant reduces the H2S content of the NG streamfrom 13.5% to 5.3% mol, which allows thecentralized gas gathering and upgrading plant torealize its full gas production capacity without theneed for additional sulphur recovery capacity. Table 3shows the properties and compositions of the gasstreams entering and exiting the Kwoen gas plant.The process, illustrated in Fig. 5, was realized witha series of simple flash-regeneration stages in placeof a regeneration column so as to minimize energyrequirements and capital cost. The gas is absorbedby a Morphysorb solution in two parallel packedcolumns. The acid gas final flash drums operate at0.45 MPa and 0.18 MPa respectively. The acid gasthat flows from the final flash drum is compressedto a final discharge pressure of 7.6 MPa. Thecompressor aftercooler liquefies the acid gas priorto its entering a 14 km, 6-inch (15.24 cm) diameterpipeline which carries it to the injection well.

Again in this sector, several innovative physicalprocesses are being studied as a basis for the bulkseparation of H2S from highly sour NG streams(Table 4). The general characteristics desired insuch a system, where the acid gas is destined forre-injection, can be identified as follows: the bulk

removal process should produce the acid gas atelevated pressure, and preferably cold and dry, inorder to reduce the cost of compressing it to theformation pressure; it should also require minimalenergy input for gas regeneration, as surplus heatenergy from the Claus unit is not available in thisscenario. However, compared with a commercialgas development project, in a bulk sweeteningprocess methane recovery efficiency will be less ofa performance constraint. This is true in remotelocations (such as Tengiz, Kashagan), where someof the acid gas by-product of oil production isdestined for re-injection.

The Sprex processA scheme for bulk H2S removal has been

proposed by the Institut Français du Pétrole (IFP),and Total (Lallemand and Minkkinen, 2002). Theoverall process incorporates an H2S bulk removalstep upstream of a chemical solvent treatment(using activated MDEA) and is intended for use inconjunction with acid gas re-injection into adisposal reservoir. Bulk removal is accomplishedin the Sprex contactor through a low temperaturedistillation of H2S from methane carried out atelevated pressure (c. 100 bar). Mercaptans are notremoved, and as there are significant losses ofhigher hydrocarbons with the H2S stream, the mostinteresting applications are for the treatment oflean natural gases (C3� �1.0%). One of the mainadvantages of this process is that the H2S-richstream from the bulk separation step is obtained asa liquid, which reduces the cost of compression forre-injection. All water contained in the inlet rawgas will be dissolved in this liquid stream, whichwill also contain some of the incominghydrocarbons. A simplified Sprex process flowdiagram is depicted in Fig. 6.

The dried gas leaving the Sprex contactor iscooled through a gas-gas heat exchange, gas-condensate heat exchange, and then through arefrigerated chiller before being passed to a LowTemperature Separator (LTS) drum at linepressure. Fluid composition and the LTStemperature largely determine H2S removalperformance. The higher the acid gas content ofthe inlet gas, the higher the pre-extractionperformance. Table 5 reports the material balancecalculated for an NG stream containing 34% molH2S with the LTS operating at �30°C. A pilotplant-scale test of the project is being readied forstart-up near Pau, France at the time of writing.The economic assessment reported in Table 6shows that, for highly sour gas streams, the Sprexprocess can significantly reduce CAPEX (CAPital



Table 4. Proposed processes for the bulk sweeteningof acid gas

Process Physical basis

Sprex Cryogenic distillation of H2S

Membraneseparation

Use of selective membranes for theseparation of H2S

Absorptionby hydrocarbons

Absorption of H2S with a physicalsolvent

MorphysorbAbsorption of H2S with a physicalsolvent

Separation withhydrates

Selective formation of H2S hydrates and their separation from the natural gas

EXpenditure) and power consumption compared toconventional sweetening processes.

H2S absorption by field hydrocarbons Eni is currently exploring a simple and inexpensive

process for bulk separation, in which the removal ofH2S is performed by absorption with hydrocarbonfractions available at the well site (e.g. virgin naphthaor stabilized crude oil). This application is aimed atremote oil production operations with large volumesof associated gas, where there is very limited marketvalue for the light hydrocarbons.

Process simulation studies suggest the possibilityof achieving methane recoveries as high as 95% froman NG containing 14% H2S and 4% CO2, with aproduct gas containing 0.4% mol H2S. The chemicalstability of the hydrocarbon solvent offers thepossibility of effecting desorption at high temperature,producing the H2S-rich disposal stream at highpressure (10-15 bar), which reduces the compressionratio required for acid gas re-injection. It should bepossible to further enhance methane recovery, thoughat the expense of an increase in the treatment schemecomplexity (e.g. by employing intermediate desorption



H2S

sour gas

acidgas

fuel gas

re-injection well

pipeline gasSprex

refrigerationElf Activated MDEA

dehydrationunit

cryogenic column

Fig. 6. The Sprex process for bulk H2S removal (Lallemand and Minkinnen, 2002).

Table 5. Predicted performance of the Sprex process(Lallemand and Minkkinen, 2002)

Composition of acid gas(pressure, 70 bar)

% vol

H2S 35.0

CO2 7.5

Methane 56.5

Ethane 0.6

Propane and heavier hydrocarbons 0.4

Water Saturated

Specifications of treated gas

H2S content (ppm vol) 4

CO2 content (ppm vol) 2

Waste gas

Re-injection pressure (bar) 150

Table 6. Economic comparisonof the Sprex process and the MDEA sweetening process for a natural gas

with an H2S content of 35% (Lallemand and Minkkinen, 2002)

CAPEX(base 100 for conventional

treatment)

WithoutSprex

WithSprex

Sprex unit 0 23

Compression MDEA and acid gas + facilities

100 57

Total for process units 100 80

Energy requirements (MW)

Electricity (motors, pumpsand compressors)

52 29

Heat energy(low pressure steam)

46 34

and a partial recycle of gas to the absorption column).The overall process is similar to that for other physicalsolvents. A pilot-scale plant for testing this process hasbeen in operation near Milazzo, Italy since early 2005.

The loss of heavier hydrocarbons with the H2Sstream favours use of the process in remote locationswhere the market value of these products is the lowest.A comparison of the CAPEX and OPEX (OPeratingEXpenditure) costs estimated for a conventional plant(for example amine, Claus, TGU) and the sameprocess with an upstream bulk separation step (withthe acid gas recovered re-injected into the reservoir)shows that this process offers considerable potentialeconomic advantages. The composition of the gastreated is the same as that in Table 3 whereas ageographically and logistically remote site wereassumed in the cost estimate exercise.

Separation of H2S via hydrate formation Another physical property, based on the different

tendency of H2S and NG to form clathrate hydrates,has also been proposed as the basis for effecting bulksweetening of sour gas. In fact, hydrate formation withH2S is significantly favoured thermodynamicallycompared to methane (CO2 is intermediate betweenthe two); the onset for H2S hydrate formation is only3 bar at 10°C, whereas methane forms hydrates at thistemperature only above 70 bar. In a multi-componentsystem the gas concentrations in the gas and solidphases tend to be enriched or reduced in thecomponents present in the starting mixture.

A thermodynamic study showed that the mostrelevant variable for the yield of the sweeteningprocess (defined as the ratio between Py and Qz, beingQ the molar amount of feed gas, P the molar amountof equilibrium gas, z the molar concentration ofCH4 in the feed gas and y the molar concentrationof CH4 in the equilibrium gas) is the amount of water.Predicted equilibrium compositions of a humid sour

gas stream as a function of T, P and water content arereported in Table 7.

A schematic flow diagram of a process for gassweetening via hydrate formation is shown in Fig. 7.The process involves the following steps: a)pressurization, if required, of the feed gas mixtureusing an adiabatic compressor to the set pressure(higher than hydrate formation pressure); b)precooling of the pressurized feed gas to a giventemperature by heat exchange with the purified NGstream from the hydrate formation unit; c) cooling ofthe feed gas to the set temperature (lower than hydrateformation temperature) using an ethylene glycolsolution stream; d ) hydrate formation in a highpressure unit kept at constant temperature with asecond ethylene glycol solution stream; e) hydratedissociation in a dissociation unit by heating with astream of hot water (seawater or produced water) of ahighly-concentrated H2S gas stream; f ) the waternecessary for hydrate formation is supplied by arecycling stream from the hydrate dissociation unit.

Potential benefits of hydrate separation reside inthe possibility of treating associated gas containing ahigh level of H2S without the need for special solventsor substrates (solid catalyst, membranes, etc.). Theonly requirements are water, pressure (20-50 bar) andcooling/heating systems (0-50°C); the energy neededfor sustaining the process (heating/cooling cycles) canbe obtained by combusting part of the sweetenednatural gas. The process should also be capable ofproviding the acid gas stream at moderately elevatedpressure to facilitate re-injection.

Gas permeation: membrane processesIn membrane separation processes a high pressure

gaseous mixture is passed over a thin layer of densematerial (the membrane) where the dissolution of eachcomponent into the membrane is based on solubilityand partial pressure. Diffusion across the membrane



Table 7. Yields of the hydrate sweetening process for a binary mixture with an H2S/CH4 ratioof (15/85)% mol

Yield (% mol)

10 MPa, 10°C 4 MPa, 4°C 4 MPa, 1°C

Water (% wt)

Methanerecovered

H2S (%)in treated gas

Methanerecovered


Methanerecovered


10 99.29 13.91 99.49 13.74 99.51 13.72

50 90.12 6.85 91.50 5.81 91.54 5.63

70 67.36 2.90 69.26 2.15 69.05 2.00

85 1.29 1.00 5.24 0.72 4.59 0.68

produces two streams: the permeate, enriched in themost permeable gases, and the retentate or product,which is that part of the feed that has not crossed themembrane (Fig. 8 A).

Increasing the partial pressure of thecomponent to be removed from the feed gas (byincreasing its molar fraction or the absolutepressure of the feed gas, or by decreasing its molarfraction or the absolute pressure in the permeateside) reduces the membrane area required forseparation. Commercial membranes for acid gasseparation are composed of either cellulose acetateor polyimide polymer. The spiral wound membranegeometry, which consists of a stack of flat sheets ofmembrane alternated with porous spacers wrappedaround a perforated collecting pipe (Fig. 8 B), ispreferred as it allows for the highest partialpressures. The feed gas flows in the axial direction,parallel to the central pipe, while the permeatestream flows radially towards it. The membranescan operate at pressures up to 130 bar andtemperatures from 30-90°C. The membranemodules are usually arranged inside pressurevessels containing 1-8 modules.

Gas permeation is increasingly being appliedby upstream operators for CO2 removal from NG,with capacities up to 7 ·106 Sm3/d being installedand volumes up to 14 ·106 Sm3/d underdevelopment or construction. However, nocommercial applications for H2S removal havebeen set up and operational experience is limited togas streams with a low H2S content.

The separation of H2S from NG by permeationusing current membrane materials is characterizedby a trade-off between residual H2S in the treatedNG and the loss of hydrocarbons with the acid gaspermeate. This has blocked commercialdevelopment of membrane separation in traditionalgas processing schemes. For bulk separationprocesses, however, this trade-off may be lesssignificant, and industry efforts are presently

underway to evaluate the potential benefits of thistechnology. Membrane processes, in fact, possesssome important intrinsic advantages: they areenvironmentally safe, have lower energyrequirements and are simple to operate, as there areno moving parts, adjustments or solvents. Thesecharacteristics make the application of membranesystems particularly attractive for remote orinaccessible locations. The modular nature ofmembrane separation allows incremental additionsto capacity.

H2S treatment

Today, H2S treatment processes all lead to thegeneration of elemental sulphur. Large scale sulphurrecovery operations (�20 t/d of S) treat H2S asfurnished from the sweetening plant and convert it toelemental sulphur through the Claus process, whichaccomplishes the following overall transformation:

H2S �1/2O2�� H2O �S

This reaction is carried out in three or moreseparate stages in a large and complex plant.

As the regulations on emissions from oil and gasoperations have tightened, processes capturing asmuch as 99.9% of the total sulphur present in NG havebeen developed. Large scale H2S processing plantsrepresent a significant capital expense and, togetherwith the tail gas units needed to clean the off-gas tospecifications, can triple the cost of a gas-sweeteningunit. For this reason, efforts continue to improve theeconomic as well as the environmental performance ofthese processes.

This section examines the Claus process and itsmajor variants, which continue to dominate sulphurrecovery, and the processes employed for tail gasclean-up (see also Chapter 3.2). Finally, alternativeprocesses and major modifications of the Clausprocess which are presently the object of R&Dare discussed.



rawgas

1 2 3

4

5

6

H2S rich gas

purifiedgas ethylene

glycolethyleneglycol

water hotwater

Fig. 7. Process for gas sweetening via selected formation of H2S hydrates.

The Claus process and its modificationsThe best-known and most widespread process

for the treatment of acid gases is the Clausprocess, which converts H2S into liquid elementalsulphur. Claus plants treat gas streams containing15-80% mol H2S, up to several % molhydrocarbons, with the remainder comprised ofCO2, water vapour and other trace compoundssuch as COS and CS2.

The Claus process, first conceived in 1883,comprises two steps, the first of which is theoxidation of one third of the H2S to SO2 in afurnace, and the second is the reaction of SO2 withthe remaining H2S (the Claus reaction) to formelemental sulphur and H2O (GPSA, 1998):

H2S �3/2O2�� SO2�H2O

2H2S �SO2��

��(3/x)Sx�2H2O

This reaction is endothermic at high temperatureand exothermic at lower temperature (�10 Kcal/molSO2 at 800°C and �25 Kcal/mol SO2 at 100°C) onaccount of association equilibria involving sulphur in

the vapour phase, which can be represented as follows(the stoichiometry coefficients are omitted):

S8��

��S7��

��S6��

��S5��

��S4��

��S3��

��S2

These equilibria are shifted to the left at lowtemperature and to the right at higher temperature(above 800°C, only S2 is present). The equilibriumconversion to sulphur in the Claus reaction as afunction of temperature is reported in Fig. 9, whichis characterized by a minimum around 600°C.

In practice, the Claus reaction is carried out inseveral separate stages. In the first, at atemperature between 980 and 1,370°C and in theabsence of catalyst, one-third of the H2S togetherwith any hydrocarbons and other combustiblesubstances are burned in air in a furnace. In theseconditions, the oxidation of H2S to SO2 takesplace, followed by the slower reaction between H2Se SO2 to yield sulphur and water. This stagegenerates the major part of the heat of reaction,which is recovered for various uses within the plant(e.g. for regeneration of the solvent in thesweetening unit); conversion to sulphur is about70%. Two or more catalytic stages are generallyinstalled downstream of the furnace to boostsulphur recovery. These generally operate below370°C and provide a maximum sulphur recovery of97-98%. Higher recoveries are achieved andsulphur (SO2) emissions requirements are satisfiedonly through the use of expensive tail gasprocessing units (TGU).

The main advantages of the Claus processinclude the production of high quality sulphur, thepossibility to convert COS, destroy NH3 and HCN,and the production of large amounts of heat in theform of steam. On the other hand, the Clausprocess places strict requirements on the incomingacid gas characteristics (a minimum of 15% mol H2Sin the gas stream, limited hydrocarbon content, etc.).The push to achieve ever-greater levels of sulphurrecovery has led to an increase in the numberof catalytic conversion stages and to the addition oftail gas treatment units of increasing sophistication,increasing the cost and complexity of the Clausprocess.

There are two principal variants of the Clausprocess: straight-through and split flow. The first isemployed where the concentration of H2S isgreater than 50%, and oxidizes only part of theH2S in the incoming stream. The second isgenerally employed for lower H2S concentrations,and splits the incoming acid gas stream into twoparts (1/3 and 2/3), the smaller of which is fullyoxidized before being combined with thenon-oxidized portion.



A

B

highpressurefeed gas

highpermeabilitycomponents

product

selectivelayer

poroussupport

permeate

feed in

concentrateout

permeate-sidespacer

membranepermeateout

roll toassemble

feed-sidespacer

permeate flow(after passage

through membrane)

Fig. 8. Membrane-based separation of gas mixtures (A) and the spiral wound geometry (B) commonly used for NG treating.

A schematic diagram of the Claus process inthe straight-through configuration, comprised ofthree catalytic stages is reported in Fig. 10. Theprincipal phases of the process are:• Combustion (in a deficiency of air) of c. one third

of the H2S present in the acid gas stream, of thehydrocarbons and any other combustiblesubstances. The furnace operates at 20-100 kPagreater than the atmospheric pressure.

• Cooling of the combustion products with therecovery of heat and production of steam (10-35bar, 185-245°C). The mean temperature of theeffluent gas is generally 315-370°C.

• Condensation and recovery of sulphur from thefurnace effluent and the effluent from the catalyticconverters, with production of low-pressure steam(3.4-4.8 bar). The effluent gas temperature fromthe first condensers is c. 177°C, and 127-149°Cfrom the final condenser.

• The gas exiting from the condensers must bereheated to above the dew point of sulphur to avoidcondensation in the successive converter.

• The catalytic converters (generally employing anallumina catalyst) promote the reaction betweenH2S and SO2 at the low temperatures downstreamof the furnace.A new variant on the Claus process employs

oxygen-enriched air. The reduced volume of the airemployed leads to more economic plant dimensions forlarge-scale operations, and can also be used to enhanceexisting plant capacity. Oxygen enrichment is generallykept below 28% to avoid the need for special materials,but some plants are operating with 60% oxygen-enriched air. Though the application of this variant ismost convenient in refinery operations which canjustify the cost of an oxygen plant or have access toother sources of oxygen, even stand-alone upstreamoperations may have favourable economics for verylarge-scale acid gas plants (e.g. 2,000 t/d of S).

Tail gas treatmentThe steadily increasing sulphur content of

produced oil and NG and growing environmentalpressure have pushed oil and gas producers andrefiners to adopt technologies offering very highlevels of sulphur recovery. Typical sulphurrecoveries for a two-stage Claus plant are usually90-96%, increasing to 95-98% with the addition ofa third stage. Other tail gas unit (TGU),technologies are required to achieve recoveriesof 99% or higher, of which two general typesexist: H2S recycle and subdewpoint (see alsoChapter 3.2).

H2S recycling processesIn these processes the oxidized sulphur compounds

exiting from the Claus plant are reconverted to H2Sand recycled. The general layout of one of these, theShell Claus Offgas Treating (SCOT) process, is shownin Fig. 11.

The typical composition of the tail gas entering thetail gas treatment plant (TGU) is as follows: COS,200-5,000 ppm vol; CS2, 200-5,000 ppm vol;H2S, 0.3-5% vol; SO2, 0.15-0.75% vol; Svap, saturatedat the operating T and P.

The SCOT process is comprised of two mainsections: the reactor, where hydrolysis andhydrogenation reactions are carried out (see below),and the absorption section.

The hydrolysis reactions are as follows:

COS �H2O�� CO2�H2S

CS2�2H2O�� CO2�2H2S

CO �H2O�� CO2�H2

The following hydrogenation reactions also takeplace:

SO2�3H2�� 2H2O �H2S

Sn�nH2�� nH2S



conv

ersi

on o

f H

2S to

S (

%)

40

60

80

100

temperature (°F)

temperature (°C)

acid gas from Wellhead Facilitiesacid gas from refinerydata of Gamson and Elkins

0

0 200

catalyticrange

flamerange

400 600 800 1,000 1,200 1,400 1,600

400 800 1,200 1,600 2,000 2,400 2,800

Fig. 9. H2S conversion to elemental sulphur as a function of temperature for: an acid gas containing 3.5% molhydrocarbons (in red); a refinery acid gas stream (c. 7% mol hydrocarbons,1% mol mercaptans) (in blue); pure H2S (in green) (GPSA, 1998).

The gases exiting the Claus plant are heated to c.300°C, mixed with a reducing gas containing H2, andthen passed over a fixed bed reactor containing aCo/Mo catalyst where the oxidized sulphur species areconverted to H2S. The gas leaving the reactor at320-340°C is cooled in two steps to 40°C. The cooledgases are sent to an absorption column where achemical solvent (e.g. MDEA) is used to capture H2Sand CO2. The off-gas typically contains 150 ppm ofH2S and can be sent to the incinerator. The acid gasesare recovered from the rich solvent in a strippingcolumn and recycled to the Claus unit.

Subdewpoint processes These processes operate at temperatures below the

dewpoint of elemental sulphur, thereby ensuring ahigh conversion of H2S and SO2 to sulphur. TheSulfreen process is a typical example of a lowtemperature, dry bed Claus process. It consists of twoor more reactors, a blower for regeneration, a gasheater and a separator. The catalytic beds arecomprised of an activated allumina, which actssimultaneously as adsorbent and catalyst. At thereaction temperature (120-140°C) H2S and SO2 arealmost completely converted to sulphur. This remains

adsorbed onto the catalytic bed, which must beregenerated periodically at 300°C using gas from theClaus furnace. Two or more reactors are thereforerequired for continuous operation. A Claus plant withthree catalytic beds and a Sulfreen unit can achieve99.5% sulphur recovery; in a further refinement, COS,CS2 and H2S can be converted to sulphur in ahydrolysis unit installed ahead of the Sulfreen reactor.

The highest total sulphur recoveries are presentlyobtained with the Clauspol process, in which the Clausreaction proceeds to completion in a non-volatileorganic solvent. This is accomplished by sending thetail gas from the Claus unit to the bottom of a filledreactor where it flows upwards against acountercurrent solvent flow. The H2S and SO2 areabsorbed by the solvent which contains a catalyst. Thesulphur produced by the Claus reaction is onlypartially soluble in the solvent and is recovered as aliquid phase at the bottom of the reactor. The newestversion of the Clauspol process maintains the solventbelow the sulphur saturation limit, to reduce sulphur inthe off-gas to less than 50 ppm vol. The ‘desaturationloop’ is comprised of three heat exchangers and aseparator. In this configuration a combinedClaus/Clauspol process can achieve conversions of



converterno.1

converter2

converter3

reheater1

waste heatboiler

reactionfurnace

high pressuresteam

lowpressuresteam

lowpressuresteam

lowpressuresteam

boilerfeed water

boilerfeed water

boilerfeed water

boilerfeed water

boilerfeed water

acidgas

knockout

drum

air blower

sulphur pit

sulphur pump

liquidsulphur

air

reheater2

reheater3

condenser1

condenser2

condenser3

condenser4

heatedboiler

feed water

tailgas

no. no.

no. no. no.

no. no. no. no.

Fig. 10. Example of three-stage Claus plant: straight-through operating with acid gas-fueled inline burners for reheating (GPSA, 1998).

over 99.9% by utilizing a high performance TiO2

catalyst in the Claus process to ensure conversion ofCO2 and CS2.

Innovative processes for H2S treatmentThe increasing cost and complexity of the

existing Claus and TGU processes haveencouraged a continued search for new H2Streatment and emissions control technologies andprocesses. Of particular interest are catalyticpartial oxidation processes in short contact timereactors, which offer the possibility of achievingsignificant reductions in the dimensions of thereactors and reducing the total number of stagescompared to the Claus process. Secondly, effortscontinue on the development of treatmentprocesses capable of recovering significantamounts of H2 from H2S.

Short contact-time Claus reactorThe sulphur partial oxidation catalysis process

currently under study by ConocoPhillips (Allisonet al., 2003) carries out the transformation of H2Sto sulphur in the presence of air in a compactreactor. The results achieved thus far suggest thatthe current level of performance of the Clausprocess can be achieved with the elimination ofone or more catalytic beds. The reduction innumber and size of the stages may lead to capitalexpenditure savings up of to 50% compared to aconventional Claus process. The reactortemperature lies in the range from 700-1,500°Cand the catalyst used, covered by a patent, is amixture of Pt/Rh catalyst containing a lanthanidemetal supported on zirconia or a-allumina.Laboratory tests have been carried out with gas

streams containing from 20 to 100% H2S.Residence time in the catalytic bed is in the orderof 10-50 ms, and reactor dimensions are 10% orless of those of a Claus furnace. Operatingconditions and the results of laboratory tests arereported in Table 8.

The sulphur yield (75%) can be compared withthat of a conventional Claus furnace (55-65%).Laboratory testing indicates that the process canstand both significant (greater than two-fold)variations in the gas flow rate and ample variationin the ratio air/H2S (from 1.7-2.7) withoutsignificant impact on the H2S conversion rate andyield of sulphur.

A pilot scale plant (1 t/d of sulphur produced)comprised of two identical reactors has beenrealized at the ConocoPhillips Ponca City refinery(Oklahoma, USA), which has provided a gas



reheater

gascooler

off gas toincinerator

recycleto Clausplant

steam

air

fuelgas

steam

Claustail gas

condensate tosour water stripper

reac

tor

quen

ch c

olum

n

abso

rber

rege

nera

tor

Fig. 11. Basic design of the SCOT process.

Table 8. Partial catalytic oxidation of sulphur:conditions and results of laboratory tests

(Allison et al., 2003).GHSV�Gas Hourly Space Velocity

H2S conversion (%) 83

Selectivity for sulphur (%) 90

Sulphur yield (%) 75

SO2 yield (%) 8

H2 yield (%) 10

air/H2S 2.1

GHSV (h�1) 265,000

stream containing 80-88% H2S, 1-2%hydrocarbons and ammonia, with the balance madeup by CO2. The focus of current studies is toincrease the yield of sulphur to over 94%, whichwould make it possible to eliminate the need forpost-reactor catalytic stages entirely.

Thermal dissociation of H2S to sulphur and H2

Hydrogen production from H2S, favouredthermodynamically, is certainly among the mostattractive hypothetical alternatives to the Clausprocess currently under study. Notwithstandingnumerous efforts in this area, it has not yet beenpossible to identify a process that captures thispotential. Studies conducted by Alberta SulphurResearch (ASR) during the 1990s identified thethermodynamic limitations to thetransformation of H2S to H2 and sulphur via athermal-cracking route (Clark et al., 2003).Among the principal problems identified werethe inability to find a favourable temperaturefor the catalytic Claus converters; above 325°C,the H2 produced reacts with SO2 in the presenceof the catalyst, whereas the conversion of H2Swas insufficient at lower temperatures. Anotherlimitation identified regards the difficulty ofseparating H2 from the effluent stream. All theevaluations carried out indicate that thermaldissociation of H2S to sulphur and hydrogen viathe process studied is not economicallyattractive.

ASR is presently exploring partial catalyticoxidation in a short contact time reactor as a meansfor achieving the non-equilibrium conversion ofH2S to H2 and sulphur. The thermodynamics for

three alternative pathways for the reaction of H2Swith oxygen are reported below:

[1] H2S �1/2O2�� H2O �1/8S8

(�208.7 kJ/mol)

[2] H2S �3/2O2�� H2O �SO2

(�518.2 kJ/mol)

[3] 2H2S �1/2O2�� H2�H2O�S2

(�72.1 kJ/mol)

Assuming that oxidation of H2S proceeds via aradical pathway, the reaction pathway reportedbelow has been suggested as a possible route toH2 and S2, provided that the competing oxidation toSO2 can be avoided (Clark et al., 2004). This resultwould require catalytic conditions with theefficient removal of heat.

Results for the uncatalyzed conversion of H2S inthe presence of O2 over a range of temperatures andcontact times are reported in Table 9. Small amountsof H2 are detected above 600°C, the yield increasing atshorter contact times. These results suggest that thereaction is within the kinetic regime and that H2 is notthe result of thermal dissociation at this temperature.On the other hand, the significant amounts of SO2

formed are contrary to the predictions of the reactionscheme reported below and the equilibrium predictionat 700°C.

H2S �O2�� HS ��HO2

H2S ��HO2�� HS �[H2O2]

[H2O2]�� 2�OH

2H2S �2�OH�� 2�HS �2H2O

2�HS�� H2�S2



Table 9. Oxidation of H2S in the gas phase in a quartz reactor* (Clark et al., 2004)

T (°C)**Contact time

(ms)

Yield (mol/100 mol of reagent) Conversion (%)

H2 SO2 H2S O2

200 180 0.0 0.1 1.2 2.7

400 127 0.1 0.3 21.6 43.4

600 98 5.9 6.7 41.9 100

700*** 88 6.4 6.5 43 100

44 7.4 6.0 46.8 100

24 7.8 5.4 48.3 100

* Feed: 79.9% of H2S, 20.3% of O2** Temperature of external furnace*** Results of the thermodynamic calculation at 700°C: yield of H2, 0.50 mol; yeld of SO2, 1.17 mol; H2S conversion, 47.8%; O2 conversion, 100%

To depress the undesired reactions leading toSO2 formation, the possibility of using a- andg-allumina catalysts was examined (Table 10). Theseun-doped catalysts significantly reduced SO2

formation, with g-allumina proving particularlyeffective in catalyzing the partial oxidation of H2Sto H2 at 400°C, while limiting SO2 formation to avery low level (0.25% conversion to SO2). Theselectivity improved at still lower external furnacetemperatures, with no SO2 detectable in theproduct stream. H2 selectivity was also examinedas a function of the H2S/O2 ratio in the feedbetween 200 and 600°C. Below 500°C, amaximum in the selectivity for H2S was found foran H2S/O2 ratio of 4, which agrees with thestoichiometric requirement for the formal reaction[3] shown above. Fig. 12 summarizes theobservations on the factors controlling H2

selectivity in the partial oxidation of H2S andrelates these to the presumed chemical stepsunderlying the process.

The Gas Technology Institute (GTI) hasexplored high temperature, non-catalytic routes forthe decomposition of H2S to sulphur and H2

(Slimane et al., 2002). As reported in Fig. 13, H2Sdecomposition to H2 and sulphur is disfavouredthermodynamically until a very high temperature,sulphur conversion reaching 50% only above1,350°C. In the process studied by GTI, the energyrequired for the transformation is furnished by theoxidation of part of the H2S. The partial oxidationand decomposition of the H2S take place within a

porous medium where temperatures in the order of1,400°C are reached. The process is complex, withthe reaction front advancing progressively withinthe reactor. Periodic inversion of the direction offlow, as shown in Fig. 14, has been proposed as ameans to confine the reaction. Published reports onlaboratory-scale tests claim H2 yields of up tonearly 20% (Fig. 15). The major part of the H2

contained in the H2S should be converted to H2O(not reported), while the production of SO2 shouldbe comparable to that obtained with the Clausprocess. How to recover the H2 generated from thehigh temperature reaction gas remains an openproblem.

3.3.5 Sour gas re-injection

Acid gas re-injection is attracting much attentionas an environmentally-sound and cost-effectiveapproach that can avoid the cost of traditional H2Sprocessing and the problems of handling theelemental sulphur product, particularly for verysour NG streams. In this process, the acid gasesseparated are compressed and injected into thedisposal reservoir through a special well, in amanner similar to the disposal of produced water(Fig. 16). The disposal zone can be either ahydrocarbon reservoir (the producing reservoiritself or a nearby, depleted one) or a saline aquifer.In operations where a part of the sour associatedgas is destined for re-injection, this can be enriched



Table 10. Catalytic oxidation of H2S over a- and g-allumina* (Clark et al., 2004)

T (°C)**Contact time

(ms)

Yield (mol/100 mol of reagent) Conversionof H2S (%)

Selectivityfor H2 (%)H2 SO2

a-allumina

300 6.3 0.5 0.1 51.6 1.1

400 8.7 4.2 0.2 55.1 9.6

500 7.6 5.8 1.2 56.6 12.9

g-allumina

200 18.5 7.1 0.0 60.5 14.6

300 15.3 7.8 0.0 61.9 15.7

400*** 13.0 9.2 0.2 62.2 18.6

* Feed: 79.9% of H2S, 20.3% of O2** Temperature of external furnace*** After 4 h: yield of H2, 10.7 mol; yield of SO2, 0.02 mol; conversion of H2S, 64.6%; selectivity for H2, 20.8%

with H2S separated from the sales gas. One of thepotential benefits of injecting H2S and acid/sourgas into the reservoir is that the excellent solubilitycharacteristics of H2S favour oil recovery (Thibeauet. al., 2003). The downside of such a miscibleflood scheme, of course, is the risk of earlybreakthrough of highly sour gas to the producerwells, and the inevitable build-up of acid gaswithin the reservoir which will lead, late in theproduction life, to acid gas recycling.

Acid gas re-injection technology was developedin North America where over 60 operations areactive at the present time, approximately two-thirdslocated in Alberta, Canada. Overall, it is estimatedthat a total of 750,000 t of H2S and 635,000 t ofCO2 are disposed of each year in Western Canadaalone. One of the largest facilities, the Kwoenproject (see Section 3.3.4), re-injects over 860 t/dof H2S.

To date, acid gas re-injection processes inNorth America have employed relatively shallowaquifers or depleted gas reservoirs requiringrelatively low pressures for injection (�100 bar;Wichert and Royan, 1997). One of the major issuesfor the extension of this process to other regions ofthe world is the much higher injection pressuresthat will be required (Miwa et al., 2002). One ofthe most important raw gas injection plants yetbuilt is in the oil-processing centre of the sour (c. 3-4% of H2S) giant Karachaganak condensatefield, in the north-west of Kazakhstan. This facilitydisposes of 6.4 billion cubic metres of sour gas peryear at pressures of c. 500 bar by re-injection intothe reservoir. A similar solution is under evaluationfor the Tengiz and Kashagan fields, which contain



H2

sele

ctiv

ity

increasing H2S/O2 ratio

less O2 in feed

increasingbranch

decreasingbranch

vs

Fig. 12. Controlling factors for H2 selectivity in partial oxidation of H2S (Clark et al., 2004).

prod

uct g

as c

ompo

siti

on(%

vol

)

0

20

40

60

80

100

temperature (°C)700 900 1,100 1,300 1,500

H2S

H2

S2

Fig. 13. Thermal equilibrium for H2S decomposition to H2 and sulphur versus temperature (Slimane et al., 2002).

periodicallyswitched flow

products (S, H2, etc.)to separation

H2S

air

H2S

air

thermal insulation

self-sustaining superadiabaticreaction zone

tem

pera

ture

Fig. 14. Conceptual scheme for a superadiabatic reactor for H2S decomposition (Slimane et al., 2002).

H2S[H2S]0

H2[H2S]0

S2 [H2S]0H

2S c

onve

rsio

n (%

)

0

20

40

60

80

100

equivalence ratio1 2 3 4 5 6

Fig. 15. Experimental results on H2S conversion; equivalence ratio is defined as the molar ratio of O2supplied to the O2 that is stoichiometrically required to burn all the H2S (Slimane et al., 2002).

up to 20% mol of H2S in the associated gas andhave even higher static pressures. Work byequipment manufacturers is underway to ensurethat the compressor capacity required for suchinjection schemes will be available to thedevelopment projects.

Properties of acid gas mixturesThe acid gas stream recovered from NG and

sent to the compressor for re-injection consistsmainly of H2S, CO2 and a small amount ofhydrocarbons. Where conventional chemicalsolvents are used for separation, the gas mixturecoming from the regeneration column will also besaturated with water; for example, the watercontent is c. 65 g/Sm3 at 35-50 kPa and 50°C. Thephysical properties of this mixture which arerelevant to the design of the compression andinjection schemes include: a) vapour/liquid phasebehaviour; b) water content at saturation in thevapour, liquid and dense phases; c) conditions forhydrate formation; d ) the density of the vapour,liquid and dense phases; e) heat capacity over abroad range of pressure and temperature. Whilethese properties are available for pure H2S andCO2, equations of state or other means must beused to estimate the properties of their mixtures,particularly under high-pressure conditions, forwhich there is little operational experience.

Two important concerns for acid gas systemsare the increased potential for corrosion and thepropensity to form hydrates at elevated pressures.The corrosion risk can be controlled by avoidingthe formation of a free water phase in the system,while hydrate formation can be avoided byoperating so that the water pressure is maintainedbelow the thermodynamic threshold. Depending onthe acid gas composition and injection conditions,chilling, glycol dehydration, desiccants and/or

interstage cooling during acid gas compression canbe used to remove water.

Fig. 17 reports the saturation water content versuspressure for different compositions of H2S and CO2 at49°C (Caroll and Maddocks, 1999). The water contentdecreases with increasing pressure and below about4 MPa is insensitive to the acid gas composition. Formixtures containing 50% or more H2S, a jump inwater content is observed in correspondence with theformation of a liquid H2S phase, which has a greaterwater-holding capacity than the gas phase. For 1/1 and3/1 H2S/CO2 mixtures, there is a range of pressuresover which three phases exist (note that CO2 does notliquefy at this temperature).

Hydrates are solid clathrate complexes of waterincorporating guest molecules, which form at elevatedpressure and at temperatures greater than 0°C. As seenabove (v. Section 3.3.4), H2S is a strong hydrateformer and can form hydrates at pressures as low as3 bar. Hydrate formation must be avoided at all pointsin the re-injection system, from the compressor to thewell bore. Fig. 18 reports the pressure and temperatureconditions for hydrate formation of H2S and other purecompounds (the hydrates are stable to the left of andbelow the lines drawn on the graph). For acid gasinjection, the compression cycles must be designed toavoid crossing both the phase envelope and the hydrateregion, as shown in Fig. 19. Hydrocarbon contentsgreater than 5% will decrease the ability of the acidgas mixture to hold water at high pressures.Hydrocarbons also reduce the bubble pointtemperature, meaning that a lower temperature will berequired to convert the acid gas mixture to a liquidphase at a given pressure.

Acid gas compression and injection facilitiesThe ultimate pressure to which the acid gas

must be compressed for injection depends on therate of injection, the length and size of the



raw sour gasproduction

gas plant

salesgas acid gas

injection

Fig. 16. Schematic illustrating process of managing acid gas production via re-injection(Royan and Wichert, 1995).

injection line, the diameter of the well tubing, thepressure of the reservoir, the permeability of thezone and the depth of the zone. Provided that theacid gas mixture contains little hydrocarbon,liquefaction can be achieved under moderateconditions (e.g. 6 MPa at 20°C for a methanecontent below 3%). The density of the acid gasmixture under these conditions will lie in the rangebetween 0.6 and 0.8 g/l. The hydrostatic head inthe well bore, due to the density of the acid gas inthe liquid and dense states, will contribute todetermining the injection pressure into thereservoir.

A wellhead injection pressure of 6 to 9 MPacan be achieved by a four-stage compressor, with acompression ratio between 2.2 and 2.9 per stage.Although large-scale compressors are available forsour gases, there is still a need to demonstrate thepotential for using large centrifugal multistage acidgas compressors at discharge pressures of8-10 MPa, followed by pumping to higher pressures.

The total power requirement to compress acidgas from 0.1 MPa to about 9.6 MPa is about7.94 kW for 1,000 m3/d. Were it possible to recoverthe acid gas from the sweetening unit at a higherpressure, there would be considerable savings inenergy requirements; providing the acid gas streamat 1.2 MPa would, in fact, halve the number ofcompression stages. Meriting particular interest,therefore, are those processes for bulk separationin which the solvent is regenerated at moderatepressure.

One of the main concerns for re-injectionoperations is that the compression section will besensitive to any changes in the composition of theacid gas coming from the sour NG treating section.

For every application, the capital costs, life cycleoperating costs, and life cycle energy efficiencymust be considered to select the optimum design.

Acid gas behaviour in the reservoirThe possibility of employing acid gas injection

to manage produced H2S depends, of course, onthe availability of a suitable reservoir or salineaquifer to receive it. A disposal zone must provideadequate injectivity and capacity and guaranteecontainment of the acid gas over a time-scale thatcan be measured in hundreds or thousands ofyears. Major properties that must be evaluated to



H2S

75/25

50/50

25/75

CO2

wat

er c

onte

nt (

g/S

m3 )

0

4

8

6

2

10

12

14

16

18

20

22

24

pressure (MPa)0 1 2 3 4 5 6 7 8 9 1110

Fig. 17. Water content of acid gas at 49°C (dashed areas denote three-phase regions) (Caroll and Maddocks, 1999).

pres

sure

(M

Pa)

0.1

0.5

1

10

5

50

100

temperature (°C)0 5 10

isobutanepropane

methane

carbon dioxide

ethane

hydrogen sulfide

quadruple point

15 20 25 30 35

Fig. 18. Thermodynamic conditions for hydrate formation with pure compounds.

0

10

9

8

7

6

5

4

3

2

1

�120 �80 �40 0 40 80 160120

pres

sure

(M

Pa)

temperature (°C)

Fig. 19. Four-stage acid gas compression showing pressure rise and temperature variations for a mixture containing 49% of H2S, 49% of CO2and 2% of CH4. The phase envelope for the liquid drop-out phase is shown by the blue line, while the phase boundary for the formation of hydrates is shown by the broken line.

ensure disposal zone suitability include: a)extension of the cap rock, and cap rockimpermeability and resistance to acid gas overtime; b) isolation of the disposal reservoir fromother producing reservoirs or aquifers; c) distanceof migration of the acid gas over a long time scale;d) response of the disposal zone minerals andfluids to the mixture injected.

The well materials and cement used forcompletion and abandonment must providesuitable, long-term resistance to the acid gasmixture in order to ensure against leakage upwardsfrom the well path to higher formations or to thesurface. This aspect of acid gas re-injection meritsfurther study, particularly as regards cementintegrity around the well bore. Standard Class Gwell cements have been shown to be unsuitable foruse in injection wells, while resin-based blends andlatex blends have been found to have acceptableresistance, at least in one specific case (Whatley,2000).

Where the injection zone is a depleted oil orgas reservoir that previously contained asignificant amount of H2S, the uncertaintiesassociated with the cap rock (areal extension,sealing capability and compatibility with H2S) andthe response of reservoir minerals to H2S injectionmay be small. In general, the final pressure of thedisposal reservoir is kept below the originalpressure in order to reduce the risk of migration ofthe injected gas out of the target formation. On theother hand, where a saline aquifer is the geologicalformation destined for injection, a full suite ofgeological and geochemical studies may beappropriate, including studies on the long-termmigration of the acid gas from the point ofinjection (Whatley, 2000).

The response of minerals and fluids present inthe disposal zone to the acid gas is an issue thatshould be evaluated with experimental studies.Formation of insoluble phases (mineral precipitatesor hydrocarbon solid drop-out) in the near wellbore region, resulting from a mixture of the acidgas components with the oil or brine phase, couldcompromise well injectivity. On the other hand,mineral phase dissolution in the acid gas mixturecould structurally compromise the well bore,leading to possible collapse, or open up pathwaysfor migration of the acid gas upwards and out ofthe target zone.

Although many studies are generally run priorto, and during, an injection project to test materialsand rock behaviour, another key issue that mustalways be considered is storage site monitoring.Proper actions must be programmed not only over

the injection period (tens of years), but also on alonger time scale.

Case study: KwoenAt the Kwoen site in British Columbia (see Section

3.3.4), acid gas from the final flash drum iscompressed to 1,350 psi (�9 MPa) and liquefied in acompressor aftercooler. Liquid pumps operating at2,125 psi (�14 MPa) send the acid gas through a 14 km, 6-in (15.24 cm) diameter pipeline to theinjection site where the acid gas liquid flows down theinjection well tubing to a depth of 2,630 m. Theinjection reservoir (a depleted gas field) is a naturallyfractured carbonate, tight matrix; the reservoirpressure at the start of injection is 600 psi (�4 MPa).Currently, one injection well disposing of 860 t/d ofsulphur is in operation, but there are plans to add asecond injection well later.

3.3.6 Sulphur storage and disposal

The preceding paragraphs described the technologiesused by the oil industry to desulphurize petroleumproducts, leading to the production of large amounts ofH2S. Despite growing interest in re-injection intounderground formations, most H2S is transformed intosulphur, a non-hazardous substance and an essentialraw material for industry. The production of sulphurfrom H2S has increased over the years, paralleling thegrowth of demand for energy; this has gradually madethe petroleum industry one of the main sources ofsupply for this element on the world markets.

Production of sulphur from petroleumWorld sulphur consumption, in elemental form or

as an alloy, reached an historic peak of 63 million tonsin 2003 (Stone, 2003). For decades, the main problemlinked to the supply of this raw material was to ensurethe necessary transportation logistics from productionsites to the industries using it. Technological advancesand concerns about the environment and safety haveled to the modification over the years of transportationsystems for elemental sulphur. Questionable practicesfrom an environmental point of view, such as thetransportation of large blocks of solid material, havebeen eliminated in favour of safer systems such astransportation in tablet or granular form or as liquidsulphur.

During some periods of low demand, thelogistical issue was complicated by the concomitantneed to temporarily store large quantities ofelemental sulphur. During the 1980s, millions oftons of elemental sulphur from the desulphurizationof hydrocarbons were stored, awaiting the right



market conditions for sale. However, theaccumulation of unsold products was a temporaryphenomenon, due to the capacity of the system tobalance supply and demand in a relatively shorttime. Up until the late 1990s, in fact, a large part ofsulphur production came from mining. When thedemand for sulphur lagged behind production andthe surplus became substantial, the drop in pricesforced part of the sulphur originating from minesout of the market; some sulphur mines closed andthe balance was re-established. This regulationmechanism between demand and supply workedefficiently for many years, up to the point whenmining production practically disappeared,definitively removing the only tool for rebalancingthe market available to operators.

Currently, over 90% of worldwide elementalsulphur production derives, as an unavoidableby-product, from the oil industry’s hydrocarbondesulphurization plants. The market for industrialby-products, by their very nature, cannot usedemand and supply regulation mechanismseffectively, and the accumulation of excessproduction is often the only way of preventing thecollapse of the market. Recently, due to theavailability of increasingly large quantities ofsulphur from the oil industry, there has been astructural imbalance in worldwide sulphurproduction, with an excess in the order of2-3 million tons per year (Kitto, 2002).

It is extremely hard to predict if and whengrowth in demand will be able to absorb excessproduction and rebalance the market. Sulphurproduction continues to increase, driven by thegrowth in energy requirements and the tendency toprogressively lower the maximum permittedconcentration of sulphur in fuels. On the demandside, factors such as the growth in worldpopulation, the rapid development of emergingcountries, especially China and India, theeconomic recovery of the ex-Soviet countries andother, equally complex, phenomena will beimportant for rebalancing the relations betweenproduction and consumption.

While awaiting the recovery of world demandfor sulphur, some oil companies are forced to storeenormous quantities of sulphur on their productionsites. This has created a new management issue,which arises more from the awareness that thesulphur will probably remain in its elemental formfor a long time than from the amounts of unsoldsulphur. This is an unusual situation, since sulphurhas always been managed as an intermediatedestined for transformation into something else ata later point. The need to store sulphur for long

periods poses completely new problems, the mostpressing of which is to ensure that theenvironmental impact of large amounts ofelemental sulphur concentrated in relatively smallareas is kept under control.

There are two possible solutions. The firstconsiders sulphur a valuable material which shouldbe stored until the market allows it to be sold. Thisline of action involves developing technologies forthe construction of sulphur storage facilities, safefor people and the environment, which allow it tobe recovered when required by the market. Thesecond solution considers excess sulphur as wasteto be disposed of; this premise underlies more orless definitive disposal technologies, deriving inpart from traditional technologies and systemsapplied to industrial waste.

Sulphur storage facilities and environmental issuesThe sulphur industry uses storage facilities of

all types and sizes. However, when large quantitiesof sulphur are to be stored for a long time, the onlytechnology used is the construction of enormousparallelepipeds of solid material. The storage ofliquid or solid sulphur in other physical forms(granules, tablets, flakes) is part of the logisticalchain needed to transport the material to market; ingeneral, the time required is kept as short aspossible. For long-term storage, needed for surplusproduction, the only way of storing the material isin the form of blocks, each containing hundreds ofthousands of tons of sulphur (Rutland, 1998).During the 1980s in Canada, the leading producerof this element, up to twenty million tons werestored at one time in this form.

The technology used to build storage facilitiesis based on the solidification of liquid sulphurfrom the recovery plant, usually a Claus plant, inblocks several hundred metres long and between 10and 20 m tall (greater heights would lead to therisk of collapse, as the material’s mechanicalresistance would be exceeded). From production tothe storage facility, the sulphur is transported in aliquid state through pipelines, which are oftenthermally insulated and, if necessary, heated. Toavoid the sulphur solidifying in the transport lines,the distance between the sulphur productionfacility and the storage site must be as short aspossible: usually the storage area is no more than afew dozen kilometres away.

The surface of the area used for storage shouldbe as large as possible. This is because sulphur isnot a good heat conductor and solidifies slowly; itmust therefore be poured in thin layers to give eachlayer the time needed for it to cool completely



before the next layer is poured. The preparation ofthe area destined for storage generally involvescovering the ground with a layer to isolate it fromthe sulphur. A layer of asphalt may be used, but thissolution is expensive; on most occasions low-costsolutions are used, such as a sheet of plastic,generally high density polyethylene or PVC, placedon the ground, which is suitably prepared beforethe construction of the block. The constructiontechnique involves enclosing the area destined tohost the sulphur block with mobile metal sheetingset around the whole perimeter. Liquid sulphur isthen poured into the area limited by the panelsfrom a distribution tower equipped with mobilearms through which the liquid is poured over theconfined space to a depth of about 10-15 cm, thenit is allowed to cool. To obtain a uniform layer, themobile arms are moved during pouring to differentareas of the block, so as to cover the whole surface.

As it solidifies, the sulphur contracts and takeson a residual porosity which, depending on thethickness of the layer and the cooling conditions, isin the order of a few percent. After solidification,the layer of sulphur is allowed to cool further topermit the transposition from the allotropic b form,which the sulphur takes on immediately aftersolidification, to the a form, thermodynamicallystable below 90oC. During this transition the solidcontracts further, with a 5.6% reduction in volume,and acquires its definitive density. The emptyspaces produced by contraction may be partiallyfilled as later layers are poured, improving thecompactness of the material. However, the sulphurblock inevitably fractures diffusely following thecontraction in volume. After the entire volumeenclosed by the metal panels has been filled, thepanels are dismantled and moved to a higher levelto allow further layers of sulphur to be poured, andso forth until the whole block has been formed. Toprolong cooling times, two blocks may be builtsimultaneously, pouring the sulphur alternately intoone or the other. It is normal practice to contain thesides with metal panels to improve the block’sstability.

The accumulation of large amounts ofelemental sulphur in small areas causes concernsabout potential risks for the environment and forpeople. The first effect is the visual impact on thelandscape, especially significant if the site is neartowns; there is also the risk of significant sulphurloss to the surrounding environment. Sulphur lossmay occur by various mechanisms; a smallquantity of H2S and SO2 is present in the liquidsulphur sent for storage, and these gases may bereleased both during solidification and

subsequently by diffusion from the mass ofsolidified sulphur. Special degasificationtreatments for the liquid sulphur lower the presenceof these toxic gases to a few tens of ppm before itis sent for storage, thus significantly reducing theproblem. A second mechanism of sulphur leakagefrom storage facilities into the environment is theformation of sulphur powder on exposed surfacesdue to erosion (Fig. 20) and evaporation. Althoughnot quantified, these phenomena may generatelarge amounts of sulphur powder, which can betransported by the wind over long distances. Theelemental sulphur released into the environment inthe dispersed form is relatively rapidlyincorporated into natural biological cycles byoxidation, mainly biological, to sulphuric acid.Oxidation is the product of the metabolic activityof thiobacilli, extremophile autotrophic bacteriacapable of using the oxidation of sulphur as asource of metabolic energy. This natural process inthe sulphur cycle is usually beneficial for theenvironment, but may represent a risk when thegeneration of sulphuric acid produced by thebacteria exceeds the capacity of the receptorsystem to block the acidity produced. If thisoccurs, the pH tends to decrease, sometimesreaching levels far more acidic than natural ones.In these cases, it is necessary to intervene bysupplying basic substances such as lime,calcareous rocks, etc. to return the system’s pH toits original values.

The production of sulphuric acid in excess ofthe capacity of the soil to absorb it is whatgenerally occurs around sulphur storage facilities,on the blocks and in their immediate vicinity. Theblock itself and the sulphur powders released fromit are an abundant source of sulphur for theubiquitous thiobacilli, whose activity leads to therapid acidification of the meteoric waters cominginto contact with the sulphur. This is probably themost significant environmental issue. Given its lowpH, the surface water from the storage site must becollected and neutralized before it can be releasedinto water courses. In addition to its costs, thisphenomenon leads in the short term to amanagement problem and in the long term to aresponsibility for the owner of the sulphur. For thelarge blocks of sulphur at some industrial facilities,the cost of treating surface water may be extremelyhigh (in the order of millions of dollars per year).

Although there is no reliable data on the actualimpact on the environment in the vicinity ofsulphur storage facilities, the pressure exerted bypublic authorities in favour of environmentally-friendly technologies is leading to the



abandonment of current storage practices.Recently, the demand for new sulphur storagefacilities has increased, driven by some petroleumproduction sites destined to generate largequantities of sulphur remote from potentialmarkets. It is easy to predict that the increase incapacity must be met with innovative technologiesable to guarantee a high level of safety for theenvironment and for people. The problem iscomplex and expensive, and is perceived as apotential obstacle to production. If the technologiestraditionally used to store hazardous industrialwaste were applied to sulphur storage, the costswould be far higher than those of current systems.It is thus imperative to identify innovative low-costsulphur storage systems which are environmentallysafe and wholly acceptable to the population andpublic authorities.

New solutions for sulphur storageThe solutions under study for the construction

of environmentally safer sulphur storage facilitiesare based on the peculiarity of the environmentalrisk associated with the presence of large quantitiesof elemental sulphur, almost entirely due to thebiological process of sulphur oxidation bythiobacilli. One possible solution is being tested atFort McMurray-Athabasca (Alberta, Canada),where enormous quantities of sulphur are produced

from oil sands by the Athabasca Oil Sands Project(AOSP). The technology tested by Alberta SulphurResearch (Clark, 1998), in collaboration withSyncrude, involves burying the blocks of sulphurat a depth which will ensure a low and constanttemperature throughout the year. If, as in someparts of northern Canada, the mean temperature isbelow that needed for the survival andreproduction of thiobacilli (�5oC), the oxidation ofsulphur is in practice completely inhibited,eliminating the risk of the acidification of watersand soils. In more temperate zones, however, thissolution cannot be applied.

A different solution has been developed by Eni.The approach adopted involves exploiting a naturalproperty of the thiobacilli in order to reduce the riskof acidifying surface waters at sulphur storagefacilities. It has been discovered that these bacteria,though able to oxidize sulphur under all theconditions present in different natural habitats, arehighly sensitive to environments which are bothacidic and have high ionic strength; in practice, inenvironments with high ionic strength, the thiobacillistop growing as soon as the pH reaches acidic values.The solution proposed (Crescenzi et al., 2005) isbased on the inhibiting action exerted by high ionicstrength environments on thiobacilli, and involvespassivizing the surface of the sulphur blocks bytreating them with non-organic salts to inhibit



Fig. 20. Erosion of the side of a sulphur block(by courtesy of the Author).

biological oxidation. Subsequently, the blocks arecovered with a layer of soil that, in addition toprotecting the block from erosion, also have thefunction of absorbing any emissions of H2S and SO2.Compared to the burial of the blocks describedabove, this approach is applicable everywhere andprobably less expensive, since it makes it possible toinhibit the activity of thiobacilli at all temperatures,and does not require the use of pits.

Underground disposalTo solve the problem of surplus production

definitively, as an alternative to surface storagefacilities, disposing of sulphur in undergroundstructures has been suggested, as in the disposal ofsome other industrial waste products. Two possibletechnologies have been proposed for sulphur disposal.The first uses injection under hydraulic fracturingconditions and involves the preparation, starting withthe liquid sulphur from the Claus unit, of a sulphurand water slurry which is then pumped under pressureinto the geological structure destined to receive it(Cobianco et al., 2005). Although not yet developedon an industrial scale, this solution makes use ofknowledge and equipment which are already knownand widely used in the oil industry, for example in thedisposal of drilling debris. Preliminary estimatessuggest that this technology is economicallyadvantageous when compared with the costs andenvironmental liabilities (in other words the costs ofthe environmental rehabilitation of an industrial site)associated with the construction and maintenance overseveral decades of sulphur storage facilities. Thesulphur thus disposed of would only be partiallyrecoverable, however, so that this solution is aimedprimarily at disposing of production when very longstorage times are predicted. In these cases, theenvironmental liabilities associated with open-airstorage may be heavier than the commercial value ofthe sulphur. The other disposal route proposed forsulphur makes use of an underground cavern createdin a salt formation (Pickren, 2003). This cavern mayalready exist, or be created by water leaching inaccordance with conventional mining practices. In thelatter case, however, large amounts of water areneeded, many times the volume of the cavern to becreated; this water, saturated in salt, must then bedisposed of. The sulphur is pumped into theunderground caverns in liquid form or as a waterslurry. Despite its higher costs, the latter option isconsidered preferable since the sulphur, when injectedinto underground caverns above its melting point, mayremain liquid for a considerable time (several years),generating greater risks than storage in the form ofslurry. By using the Frash-type recovery techniques

adopted in mines, the sulphur disposed of in saltformations can technically be recovered in case ofmarket demand. There are no known industrialapplications.

It is possible that sulphur will continue to beproduced in quantities higher than industrial capacity.The R&D efforts being made by oil companies leadone to hope that in the near future solutions suited todifferent production contexts will become availableand that, when necessary, sulphur will be stored forconsiderable periods without damaging theenvironment or endangering the general public.

3.3.7 The sulphur marketand new uses

The closure of the large sulphur mines over the pastdecade has led worldwide elemental sulphurproduction to practically coincide with that derivingfrom the desulphurization of hydrocarbons. The oilindustry, which has in practice become the mainsource of this form of sulphur, has few optionsavailable for controlling the amounts which it is forcedto produce. Without recourse to the large-scalere-injection of H2S, the production of sulphur byupstream petroleum activities is destined to increase,as is that from downstream activities, given that allforecasts indicate greater recourse in the future to sourreservoirs to meet increasing demand for energy.

Generally speaking, the sulphur produced by theoil industry represents an economic resource whichpartially offsets production costs. However, it becomesa problem when the price of sulphur and the distancefrom end users makes transportation uneconomic,forcing companies to maintain large stores of unsoldmaterial. This aspect of the management of sulphurproduced by the desulphurization of hydrocarbons isthe object of much attention from oil companies. Asecond aspect of ‘enforced’ sulphur production relatesto the effects of the relationship between demand andsupply on the world market. Forecasts predict thatgrowing enforced production capacity will consolidatea situation of glut in the sulphur market. Increasing theworld demand for sulphur has thus become anobjective, pursued by creating new opportunities intraditional markets or promoting the development ofnew sectors capable of absorbing at least a part of theexcess sulphur.

World sulphur production

Sulphur is an essential raw material for a series oftransformation industries which consume more than60 million tons per year worldwide to produce



essential commodities for economic development suchas fertilizers, paints, non-ferrous metals, etc.

This important raw material is available on themarkets in different forms. Elemental sulphur is themost important commercially, which is now derivedalmost entirely from the oil industry as a by-product offossil fuel production (see above). Another majorsource of sulphur is pyrites, iron sulphide-basedminerals. In the worldwide production of sulphuricacid, pyrites contribute about 8% of sulphurconsumption, especially in developing countries. Theremainder of the industrial demand for sulphur is metby other sources, largely the production of non-ferrousmetals. Table 11 shows the percentage contribution ofeach source of supply, from the total of about 64million tons of sulphur consumed by industry in 2001.

Sulphur consumptions

Sulphur is mainly used to make fertilizers. Theavailability of sulphur is therefore an essentialcondition for the development of industrial agricultureand for meeting the food requirements of populations.Other industries essential for economic developmentalso depend heavily on the availability of this rawmaterial, and for many years sulphur consumption wasone of the most reliable indicators for measuring thedegree of development of industrialized countries.This correlation does not hold true today, but sulphurcontinues to be a vital material for the development ofemerging countries undergoing strong economicgrowth, such as China and India. Fig. 21 shows thedifferentiation of sulphur consumption in the formswhich are commercially important worldwide.Observing consumption patterns during the modernperiod, it is easy to see that it tends to grow in parallelwith the increase in gross domestic product,confirming the central and irreplaceable role ofsulphur in industrial development. However, for someyears, despite its abundance and low cost,technological changes have tended to minimizeconsumption wherever possible. The reasons for thislie in a greater sensitivity to environmental issues,leading to the lowering of emissions of sulphur oxides

into the atmosphere, due to their fundamental role inthe generation of acid rain, and to a drastic decrease inwaste production. The industrial use of sulphur, inparticular as sulphuric acid, almost always createslarge amounts of industrial waste; in the production ofphosphate fertilizers, for example, each kg of sulphurused generates 3 to 5 kg of phosphogypsum to bedisposed of. When it is economically viable, therefore,efforts are made to adopt technologies which keep theuse of sulphur to a minimum. Hitherto, this trend hasnot altered the speed of growth of world sulphurconsumption but in the future, concern forenvironmental issues may lead to a significantdecrease in the use of this material in some productionsectors, contributing to even greater imbalances in theratio of largely enforced production to consumption.

Non-conventional uses of sulphur

Even during periods of low availability, farfrom today’s glut, sulphur has always been amaterial available at low cost and with a highdegree of purity. The potential for finding new usesfor elemental sulphur has received constantattention from operators, leading to thedevelopment of applications in the most diversefields, such as lighting, electrical accumulators,slow-release fertilizers, adhesives and foams.However, the amounts needed for these uses arerelatively modest compared to those required for



sulphur

in all forms100

17 uses not basedon sulphuric acid

10 elemental sulphur

90 other sulphur compounds

51 phosphoric acid14 fertilizers27 chemicals (TiO2, HF, etc.)

8 mineral acid leaching

83 uses basedon sulphuric acid

Fig. 21. Consumption of all types of sulphur in the different sectors of use.

Table 11. Distribution of world sulphur productionamong different sources as recorded in 2001

Sources of sulphur in all forms %

Elemental sulphur recovered from thedesulphurization of fuels

64

Sulphur from pyrites 8

Elemental sulphur from mines 2

Other sources (for example, productionof non-ferrous metals)

26

conventional uses. During the periods of abundantsupply which have occurred since the 1970s,interest in developing new uses for sulphur hasincreased in those industrial sectors which aretheoretically able to generate a consumption ofmillions of tons per year, in other words so as toabsorb a significant proportion of worldproduction. This possibility is basically linked toits use in two sectors: in agriculture to supplysulphur to crops and in construction materials,such as sulphur asphalt for the construction ofroads and sulphur cement for the production ofconcrete. These non-conventional uses will beexamined separately below; finally, an interestingnew proposal for using sulphur in the energyproduction sector is noted.

The use of sulphur as a fertilizerThe most promising sector in terms of potential

market size is that of sulphur-based fertilizers tomeet the demands for agricultural use (Boswelland Friendsen, 1993). Sulphur, which is animportant component of plants, is removed fromfields in significant quantities with every harvest.If it is not replaced, its absence limits yields andlowers the quality of the produce (Messick andFan, 1999). In fact, the lack of sulphur is not acommon condition in agricultural soils. Sulphurhas always entered the soil in abundance togetherwith phosphate and nitrogen fertilizers, whichcontain sulphates, and in meteoric waters in theform of sulphuric acid deriving from the sulphuroxides present in the air. Modern cultivationpractices tend to use concentrated fertilizerswithout sulphates, and regulations on atmosphericpollution have drastically decreased emissions ofSO2 and consequently the presence of sulphuricacid in atmospheric precipitation. The result is agradual depletion of sulphur in cultivated lands,which promises to give rise to massive worlddemand for elemental sulphur in the future.Although recorded on numerous occasions in manyparts of the world (Graziano, 1995), the sulphurdeficit has not yet stimulated a significant marketfor elemental sulphur in the fertilizer sector. Thereasons for this are not clear; it depends largely onthe reluctance of farmers to introduce newcultivation practices, but lack of investment by thefertilizer industry may also play a part.

Use of sulphur in the construction industryWhilst awaiting the expected entry of elemental

sulphur into common agricultural fertilizationpractices, the construction materials sector, inwhich sulphur is present as sulphur asphalt and

sulphur cement, remains the only marketpotentially able to increase the demand for sulphursufficiently to reduce the gap between productionand consumption. Sulphur asphalt and sulphurcement are binding materials which have beenknown for decades (Petrossi et al., 1972; McBee,1981). Sulphur asphalt is a substitute for theasphalt used for road surfacing. Sulphur cement isused to make concrete in place of hydraulicPortland cements. The commercial interest in theseuses for sulphur has experienced ups and downsover the years, depending on fluctuations insulphur prices, but R&D activities have beenconstant, creating products of ever-increasingquality. Both sulphur asphalt and sulphur cementhave been developed commercially and occupymarket niches which, however, have not yetreached the size required to significantly increaseindustrial demand for elemental sulphur. Theimpression shared by operators is that thelimitations, which in the past have hindered thelarge-scale development of the sulphur market inconstruction materials, have not yet been overcomeand, even under conditions of production surplus,prevent the growth in demand for these products.

Sulphur asphaltSulphur was first used to improve the properties

of asphalt during the 19th century, but only duringthe 1970s did the availability of low-cost sulphurand the high cost of bitumen lead to an effectivedrive to develop a market for sulphur asphalt(Rennie, 1977). Sulphur asphalt is generallyprepared by mixing hot liquid sulphur and bitumen.In the blends, sulphur exists in three different forms:one part chemically bonded to the hydrocarbonmolecules, one part dissolved in the bitumen, andthe remainder in the form of small crystallinesulphur particles dispersed in the bituminous mass.The part which reacts is generally small, since thereaction between sulphur and hydrocarbons is slowat the blending temperature. The solubility ofsulphur in bitumen depends on its composition andtemperature; for a typical commercial bitumen at140oC one can expect a solubility of around 20%, sothat as a general rule, the crystalline part accountsfor 20% in weight of the mass of the sulphurpresent. The sulphur present in the blend modifiesthe original properties of the bitumen (Bencowitzand Boe, 1938; Akili, 1984). For example, viscositydecreases as sulphur content increases, as does theresistance to penetration. The properties of thebituminous concretes formulated with sulphurasphalt, however, do not change significantly up tosulphur concentrations of 15% in volume (around



30% in weight). With higher concentrations, greaterrigidity and a more limited application temperatureinterval are observed. For these reasons, in practice,the highest possible concentration is 35% in volumein the absence of plasticizers.

Since its first introduction over a century ago,sulphur asphalt has been tested numerous times forthe construction of road surfaces. Hundreds ofkilometres of road were built with increasinglyperfected materials during the 1970s and 1980s,especially in North America (Prince, 1978); theresults have rarely been negative. In the greatmajority of cases, the quality and durability havebeen deemed similar to, or better than,conventional roads. The economic competitivenessof sulphur asphalt is usually also guaranteed duringperiods of availability of low-cost sulphur.However, there is one characteristic which hindersthe large-scale use of sulphur asphalt. In field tests,the emission of unacceptable amounts of sulphurgases, consisting mainly of H2S and SO2, duringthe application of sulphur asphalt has beenrecorded on numerous occasions. Based on thepublished evidence (Prince, 1978) collected byexperts in this sector, it appears that significantamounts of these gases may be produced duringthe heating and application while hot of the sulphurasphalt, due in part to the reaction of sulphur withhydrocarbons. Only a careful control of thetemperature of the materials to keep the sulphurasphalt below the limit temperature of 150oC cancontain this phenomenon.

Understandably, the danger of toxic emissionsrepresents a significant obstacle to the use of a newbitumen. In construction sites, it is difficult tocontrol the temperature during operations toprepare and apply the road surface with greataccuracy. The products available today on themarkets are better from this point of view than

those in the past; it is therefore probable thatperfecting the materials will make it possible toprepare formulations which are thermally morestable, thus overcoming this significant limitationon the development of the sulphur asphalt market.

Sulphur cementThe binding properties of liquid sulphur have

been known from the earliest times. In the modernera, systematic attempts to use sulphur to formulateconcrete date from the beginning of the 20thcentury; however, these were abandoned due to thelow resistance and durability of the products. Theturning point came in the early 1970s (Sullivanet al., 1975), when additives capable of overcomingthe structural defects produced by the solidificationprocess were combined with the sulphur; during thetransition from the liquid to the solid state this led toa loss of over 10% in volume. Over the years, anincredible number of products, both organic andmineral, have been used to this end: olefins,dicyclopentadiene, styrene, phosphorus, blackcarbon, vegetable oils, crude oils are only some ofthe additives tested to improve sulphur’s bindingproperties. The sulphur cement currently availableon the market possesses excellent properties,without doubt better than those of Portland cement(Sulphur […], 1994). The concretes formulated withsulphur cement, as well as having better mechanicalproperties than those of hydraulic cements, also havethe chemical inertia of sulphur, making sulphurcement highly competitive for applications inaggressive or corrosive environments, e.g. inchemical plants, in marine environments or wheretoxic and hazardous waste products are renderedinert (Darnell, 1996). Table 12 shows some of themechanical properties of a typical sulphur concrete,compared with those of the hydraulic equivalentmade with Portland cement.



Table 12. Comparison between the properties of sulphur concrete and conventional concrete

Sulphur concrete Portland concrete

Resistance to compression (MPa) 45-70 20-35

Resistance to bending (MPa) 6-10 3-4

Modulus of rupture (MPa) 9-13 535

Thermal expansion (°C�1) 14-15 12

Elastic modulus (103�MPa) 30-40 30

Humidity absorption % �0.10 0.3-3.0

Chemical resistance high low

The consumption of sulphur cement hasincreased since the 1980s, when the firstcommercially competitive products appeared on themarket. At the end of the 1990s, the volume of theworld market had reached 35,000 t/yr. The high costand greater complexity of its use compared toPortland cement represent the main limitationshindering its development in a wider market. Thecost of producing sulphur concrete is in fact two orthree times that of an identical volume of hydraulicconcrete. This high cost is in part linked to thehigher cost of the raw material and in part to greatermanufacturing costs, given the need to operate attemperatures above the melting point of sulphur(119oC). Despite its excellent properties, therefore,it is unlikely that sulphur cement will replacePortland cement for common uses. In difficultenvironments, however, its better qualities ofresistance and durability seem sufficient to allow theprogressive affirmation of this material.

Use of sulphur as a fuelSulphur’s calorific value is 2,160 kcal/kg; although

this is far lower than that of fossil fuels, it is highenough to hypothesize a thermal use to produce energy,if it is impossible to sell. This way of monetizingsulphur has been proposed by Alberta SulphurResearch. The thermal monetization of sulphur isnecessarily accompanied by the production of twice themass of SO2, to be disposed of by injecting it intounderground reservoirs in liquid form. This proposalhas aroused the interest of some oil companies, whichhave examined its technical feasibility and economicand environmental sustainability. No pilot-scale testingactivities are currently known. One important doubtconcerns the behaviour of the SO2 in the rockformation destined to receive it, since contact withwater generates a strongly acidic solution. Aninteresting hypothesis is the possibility of injecting theSO2 into a sour reservoir; in this case the Clausreaction between SO2 and H2S inside the reservoirwould lead to the formation of sulphur. Despite doubtsregarding the actual potential for energy monetization,it is probable that, as long as the sulphur glut lasts, thisscheme to produce energy from an unusable residuewithout emitting greenhouse gases will attract theattention of operators.

References

Aagard P. et al. (2001) H2S-controlling reactions in clastichydrocarbon reservoirs from the Norwegian Shelf and USGulf Coast, in: Proceedings of the 10th Internationalsymposium on water-rock interaction, Villasimius (Italy),10-15 July, v.I, 129-132.

Akili W. (1984) Modification of sand-asphalt by the additionof sulphur, in: Sulphur ’84. Proceedings of the internationalconference, Calgary (Canada), 3-6 June, 617-628.

Al-Eid M. et al. (2001) Investigation of H2S migration inthe Marjan complex, «SPE Reservoir Evaluation &Engineering», 4, 509-515.

Allison J.D. et al. (2003) Partial oxidation of hydrogen sulfideto elemental sulfur: a Claus alternative, in: Sulphur 2003.Proceedings of the international conference, Banff (Canada),2-5 November.

Anders B., Webb E.C. (1995) Treatment of H2S containinggases: a review of microbiological alternatives, «Enzymeand Microbial Technology», 17, 2-10.

Bencowitz I., Boe E.S. (1938) Effect of sulphur upon someproperties of asphalts, in: Proceedings of the AmericanSociety for Testing Materials annual meeting, 39, Part II,539.

Boswell C.C., Friendsen D.K. (1993) Elemental sulfurfertilizers and their use on crops and pastures, «NutrientCycling in Agroecosystems», 35, 127-149.

Burger E.D. et al. (2005) A mechanistic model to evaluatereservoir souring in the Ekofisk Field, in: Proceedings ofthe International symposium on oilfield chemistry, Houston(TX), 2-4 February, SPE 93297.

Caroll J.J., Maddocks J.R. (1999) Design considerations foracid gas injection, in: Proceedings of the Laurance Reidgas conditioning conference, Norman (OK), February.

Clark P.D. (1998) Improved strategies for blocking andunderground storage of elemental sulphur, «ASRL QuarterlyBulletin», 35, 16-37.

Clark P.D. et al. (2003) High temperature cracking of H2S tohydrogen in ceramic reactor systems, «ASRL QuarterlyBulletin»,126, 3-61.

Clark P.D. et al. (2004) Production of H2 from catalytic partialoxidation of H2S in a short-contact-time reactor, «CatalysisCommunications», 5, 743-747.

Cobianco S. et al. (2005) Permanent disposal of sulphur: anenvironmental and economically sound strategy for thefuture?, in: Proceedings of the 7th Offshore Mediterraneanconference, Ravenna (Italy), 16-18 March.

Connock L. (1996) A better understanding of LRSR processes,«Sulphur», January-February, 42-52.

Crescenzi F. et al. (2005) Italian Patent EA6453 to Eni,EniTecnologie.

Cross M.M. et al. (2004) Thermochemical Sulphate Reduction(TSR): experimental determination of reaction kinetics andimplications of the observed reaction rates for petroleumreservoirs, «Organic Geochemistry», 35, 393-404.

Darnell G.R. (1996) Sulphur polymer cement a final wasteform for radioactive and hazardous wastes, AmericanSociety for Testing Materials special technical publication,STP 1240, 272-281.

Foral A.J., Al-Ubaidi B.H. (1995) Evaluation of H2Sscavenger technologies, Gas Research Institute, ReportGRI-94/0197.

Gas processes ’94. A special report (1994), «HydrocarbonProcessing», 73, 67-116.

Gianna R. et al. (2004) Patent WO2004056454 toEniTecnologie.

GPSA (Gas Processors Suppliers Association) (1998) Sulfurrecovery, in: Engineering data book, Tulsa (OK), GPSA,2v.; v.II, section 21.



Graziano P.L. (1995) Preventing iron chlorosis in fruit treesusing elemental sulphur, «Sulphur in Agriculture», 19,63-66.

Keene D.E. (1989) British Gas Stretford Process: the basis ofgood process design and operations, in: Proceedings of theGas Research Institute liquid redox sulphur recoveryconference, Austin (TX), 7-9 May, 72-115.

Kensell W., Leppin D. (1996) Review of the H2S directoxidation processes, in: Proceedings of the 7th Gas ResearchInstitute sulphur recovery conference, Austin (TX), 24-27September 1995, 81-100.

Kitto M. (2002) The sulphur market: developments and trends,«Sulphur», July, 18-23.

Lallemand F., Minkkinen A. (2002) Processes combine toassist acid-gas handling, reinjection, «Oil & Gas Journal»,21 January.

McBee W.C. (1981) Modified sulphur concrete technology,in: Proceedings of Sulphur ’81: an international conference,Calgary (Canada), 367-388.

Machel H.G. (2001) Bacterial and thermochemical sulfatereduction in diagenetic settings-old and new insights,«Sedimentary Geology», 140, 143-175.

Mamrosh D.L., Allen M.C. (1994) Status of sulferoxtechnology and recent developments, in: Proceedings ofthe 6th Gas Research Institute sulphur recovery conference,Austin (TX), 15-17 May, 51-65.

Manzano B.K. et al. (1997) The influence of thermochemicalsulphate reduction on hydrocarbon composition in Niskureservoirs, Brazeau river area, Alberta, Canada, «OrganicGeochemistry», 27, 507-521.

Messick D.L., Fan M.X. (1999) Plant nutrient sulphur gainsglobal momentum, «Sulphur», 260, 26-33.

Miwa M. et al. (2002) Sour gas injection project, in:Proceedings of the Abu Dhabi International petroleumexhibition and conference, Abu Dhabi, 13-16 October, SPE78547.

Nalg G.J. (1995) ARI-LO-CAT/ARI LO-CAT II. Applicationand process technology update, in: Proceedings of the 7th

Gas Research Institute sulphur recovery conference, Austin(TX), 24-27 September, 29-37.

Palla N. et al. (2004a) Morphysorb process proves feasiblein first commercial plant, «Oil & Gas Journal», 5 July,54-59.

Palla N. et al. (2004b) Duke Energy installs first Morphysorbunit at Kwoen gas plant, «Oil & Gas Journal», 12 July,52-57.

Petrossi U. et al. (1972) Reactions and technological propertiesof sulphur treated asphalt, «Industrial and Engineering

Chemistry. Product Research and Development», 11, 214-219.

PICKREN R.A. (2003) US Patent 6582025 to CrescentTechnology.

Prince S.M. (1978) Construction and performance of sulphur-asphalt road in Texas, «Advances in Chemistry Series»,165, 161-171.

Quinlan M.P. (1996) Evaluation of selected emerging sulphurrecovery processes, «GasTips», 3, 26-35.

Quinlan M.P. et al. (1997) Evaluation of selected emergingoffshore sulphur recovery processes, in: Proceedings of theGas Processors Association annual convention, San Antonio(TX), 10-12 March, 183-190.

Rennie W.J. (1977) Sulphur asphalts. The Pronk 51/A emulsionbinder system, Calgary (Canada), SUDIC.

Royan T., Wichert E. (1995) Sulphur disposal by acid gasinjection, in: Proceedings of the Acid gas Canadian gasprocessors association meeting, Calgary (Canada), 20September.

Rutland C. (1998) The right way. The old practice of pouringsulphur to poorly prepared or unprepared ground can haveserious and costly future ramifications reports, «Sulphur»,256, 41-44.

Slimane R.B. et al. (2002) Production of hydrogen bysuperadiabatic decomposition of hydrogen sulfide, in:Proceedings of the US Department of Energy Hydrogenprogram review meeting, Golden (CO), 6-10 May,NREL/CP-610-32405.

Stone K. (2003) Sulphur. Mineral and metal commodityreviews, in: Canadian minerals yearbook, 52, 1-6.

Sullivan T.A. et al. (1975) Sulphur in coating and structuralmaterials, «Advances in Chemistry Series», 140, 55-74.

Sulphur polymer cement concrete (1994), Washington (D.C.),The Sulphur Institute, 58.

Thibeau S. et al. (2003) Simulation of sour gas injection intolow permeability oil reservoirs, in: Proceedings of theSociety of Petroleum Engineers annual technical conferenceand exhibition, Denver (CO), 5-8 October, SPE 84362.

Whatley L. (2000) Acid-gas injection proves economic forWest Texas gas plant, «Oil & Gas Journal», 98, 58-61.

Wichert E., Royan T. (1997) Acid gas injection eliminatessulfur recovery expense, «Oil & Gas Journal», 95, 63-72.

Thomas Lockhart Francesco Crescenzi

EniTecnologieSan Donato Milanese, Milano, Italy



Date post:	13-Nov-2021
Category:	Documents
Upload:	others
View:	1 times
Download:	0 times

3.3 Sour oil and gas management - Treccani

Documents