7.2.1 Background and basics Developments in environmental legislation in various countries have made it difficult to commercialize residues and heavy oils with high viscosity and high levels of contaminants (fractions with a high concentration of polynuclear aromatic hydrocarbons, asphaltenes, sulphur, nitrogen, and metals). This has created the right conditions for the development of hydroconversion processes suitable for transforming these residues into more valuable products, or at least making them more environmentally friendly. The main objectives of residue upgrading processes are: pretreatment of feedstock for other conversion units, increasing distillates and the production of fuels with a low sulphur content (Le Page et al., 1992; Dickenson et al., 1997). Traditionally, this type of process has been classified into two categories: technologies based on reducing the carbon content (carbon rejection), and hydrogen addition technologies. The first class comprises typically thermal processes, already developed in the first half of the past century, such as coking, deasphalting, and visbreaking. These processes, while having the advantage of flexibility and low cost, produce distillates of modest quality and residues of low value (fuel oil, coke, asphalt; see Chapter 7.1). The second class comprises more recent catalytic processes for the hydroconversion of residues, which increase the H/C ratio without simultaneously producing fractions with a higher carbon content; these have the further advantage that the contaminant content is also reduced. Chapter 1.1 describes the place occupied by these processes in the principal refinery schemes. It should be remembered that conventional catalytic cracking processes may also be partially fed with residues (O’Connor et al., 1991). The hydroconversion processes available today differ mainly in terms of the typology of reactors, operating conditions and the catalytic systems used. Reactions and catalysts Hydroconversion processes for residues, both thermal and catalytic, are characterized by the fact that they generally operate under severe conditions, especially as concerns temperature and pressure. Space velocities, generally low, are obviously linked to the optimization of operating conditions, in accordance with the properties of the feedstocks and the products we wish to obtain. Hydrodesulphurization (HDS), hydrodenitrogenation (HDN) and hydrodemetalization (HDM) reactions are catalytic, and of similar type to those described in other parts of this volume (see Chapters 3.1 and 6.2), with the difference that in this case the molecules are more complex molecules and have a higher molecular weight. Sulphur is present in residues mainly in the form of sulphides (sulphur links in asphaltenes) or thiophene compounds. The former are susceptible to thermal decomposition due to the low energy of the C-S bond, whereas the decomposition of thiophenes, still catalytic, may occur with or without a preliminary dehydrogenation stage. This stage is always present in the conversion of nitrogen compounds (see Chapter 3.1). Hydrodemetalization is usually thermocatalytic, preceded by the partial hydrogenation and decomposition of asphaltenes and porphyrins, with the production of heavy metal sulphides, typically of vanadium and nickel. The life-span and performance of the catalysts used in hydrodemetalization are strongly conditioned by phenomena of inhibition and the physical obstruction of the pores. Hydrogenation reactions, generally catalytic, are aimed at conversion, but also involve the olefins and free radicals which form due to thermal cracking, 309 VOLUME II / REFINING AND PETROCHEMICALS 7.2 The hydroconversion of residues
Transcript
7.2.1 Background and basics
Developments in environmental legislation in variouscountries have made it difficult to commercializeresidues and heavy oils with high viscosity and highlevels of contaminants (fractions with a highconcentration of polynuclear aromatic hydrocarbons,asphaltenes, sulphur, nitrogen, and metals). This hascreated the right conditions for the development ofhydroconversion processes suitable for transformingthese residues into more valuable products, or at leastmaking them more environmentally friendly.
The main objectives of residue upgradingprocesses are: pretreatment of feedstock for otherconversion units, increasing distillates and theproduction of fuels with a low sulphur content (LePage et al., 1992; Dickenson et al., 1997).
Traditionally, this type of process has beenclassified into two categories: technologies based onreducing the carbon content (carbon rejection), andhydrogen addition technologies. The first classcomprises typically thermal processes, alreadydeveloped in the first half of the past century, such ascoking, deasphalting, and visbreaking. Theseprocesses, while having the advantage of flexibilityand low cost, produce distillates of modest quality andresidues of low value (fuel oil, coke, asphalt; seeChapter 7.1). The second class comprises more recentcatalytic processes for the hydroconversion ofresidues, which increase the H/C ratio withoutsimultaneously producing fractions with a highercarbon content; these have the further advantage thatthe contaminant content is also reduced.
Chapter 1.1 describes the place occupied by theseprocesses in the principal refinery schemes. It shouldbe remembered that conventional catalytic crackingprocesses may also be partially fed with residues(O’Connor et al., 1991). The hydroconversion
processes available today differ mainly in terms of thetypology of reactors, operating conditions and thecatalytic systems used.
Reactions and catalystsHydroconversion processes for residues, both
thermal and catalytic, are characterized by the fact thatthey generally operate under severe conditions,especially as concerns temperature and pressure.Space velocities, generally low, are obviously linked tothe optimization of operating conditions, inaccordance with the properties of the feedstocks andthe products we wish to obtain.
Hydrodesulphurization (HDS), hydrodenitrogenation(HDN) and hydrodemetalization (HDM) reactions arecatalytic, and of similar type to those described in otherparts of this volume (see Chapters 3.1 and 6.2), with thedifference that in this case the molecules are morecomplex molecules and have a higher molecular weight.Sulphur is present in residues mainly in the form ofsulphides (sulphur links in asphaltenes) or thiophenecompounds. The former are susceptible to thermaldecomposition due to the low energy of the C-S bond,whereas the decomposition of thiophenes, still catalytic,may occur with or without a preliminarydehydrogenation stage. This stage is always present inthe conversion of nitrogen compounds (see Chapter 3.1).
Hydrodemetalization is usually thermocatalytic,preceded by the partial hydrogenation anddecomposition of asphaltenes and porphyrins, with theproduction of heavy metal sulphides, typically ofvanadium and nickel. The life-span and performanceof the catalysts used in hydrodemetalization arestrongly conditioned by phenomena of inhibition andthe physical obstruction of the pores.
Hydrogenation reactions, generally catalytic, areaimed at conversion, but also involve the olefins andfree radicals which form due to thermal cracking,
309VOLUME II / REFINING AND PETROCHEMICALS
7.2
The hydroconversionof residues
significantly reducing polycondensation reactions (andthus the formation of coke) which, however, are alwayspresent. Also important are reactions which transferhydrogen between the various hydrocarbon structuresin solution (aromatics, cycloaliphatics, paraffins), andbetween these and free radicals.
Cracking and condensation reactions, by contrast,are mainly thermal, and lead to the formation of bothlight products and polyaromatic precursors of coke.Examples of reactions can be found in other parts ofthis volume (see Chapters 3.1 and 6.2 and Section7.2.2), and in the literature (Leprince, 2001).
The catalytic reactions mentioned above areexothermic, whereas thermal cracking reactions areendothermic. The overall hydroconversion process,however, is exothermic.
The complex chemical and physical properties ofresidues, and the simultaneous presence of thermaland catalytic reactions, make the kinetic analysis of
the chemical processes involved extremely complex.Thermal cracking and polycondensation reactionsbecome significant above 400-420°C, especially withfeedstocks rich in asphaltenes; under these conditions,high hydrogen pressures are needed to stabilize thefree radicals and hydrogenate the aromatic rings.
The physical shape of the catalyst, in terms ofparticle size and structure, plays an important role bothwithin individual processes, and in the different types ofprocesses. The main distinction is that betweensupported catalysts (similar to those used, for example,in hydrotreating processes) with a diameter of between0.8 and 3 mm, and catalysts used in the form ofdispersed microparticles. Supported catalysts are usedin fixed bed, moving bed and ebullated bed processes.
Very often, at least two different types of catalyst areused one after the other in a single process withdifferent functions: a typical sequence consists of one ormore HDM catalysts, placed to ‘guard’ the efficiency of
subsequent catalytic processes. Often these are precededby a bed of protection material with large porositywhich acts as a trap or filter for the solid particlespresent in the feedstock: iron compounds, NaCl,carbonaceous or inorganic particles of various types,sediments, etc. Table 1 shows some of the properties ofa typical catalytic material for industrial use.
The porosity of HDM catalysts must be sufficientlyhigh, in terms of pore dimensions and adsorption
capacity, in order to allow large molecules to enter andthen trap the metal sulphides which form. Furthermore,the support (for example alumina) must have as low anacidity level as possible. Considering the mechanismsdescribed, these catalysts can work at highertemperatures than HDS catalysts. The latter must alsohave the ability to encourage HDN reactions,hydrogenation and hydrocracking; for this purpose, theNi-Mo system may be preferred over the Co-Mo system.
Catalysts in the dispersed phase, used in slurryprocesses (see below), consist of very small particlesinjected in a fluid phase or which form in situ from aprecursor; the catalyst is entrained in the effluentwhich exits the reactor unconverted, and can berecycled with it.
One example is molybdenum naphthenate, which,due to thermal interaction in the reactor, turns intosulphide particles associated with a hydrocarbonmatrix (about 1 mµ in diameter). Despite the highsurface area, activity is generally lower than that ofsupported catalysts. The use of these catalysts atrelatively high temperatures causes fewer problemswith agglomeration and loss of pressure; they alsohave a good ability to block free radicals, thusreducing polycondensation reactions.
311VOLUME II / REFINING AND PETROCHEMICALS
THE HYDROCONVERSION OF RESIDUES
slurry
ebullated bed
fixed bed
resid FCC
metals (ppm)1 10 100 1,000 10,000
40
30
20
10
0
asph
alte
ne (
wt%
)
Fig. 1. Typical operating windows for residue conversiontechnologies, including catalytic cracking (FCC) for the sake of comparison.
MM
hydrogen make-uphydrogen recycle
oil
water
sour gas (H2S)
water to sourwater stripper
light liquidproduct
to distillation unit
cold lowpressureseparator
cold highpressureseparator
hot high pressureseparator
quench H2
feedaccumulator
filter
residue
R-01 R-02 R-03 R-04
hot low pressureaccumulator
H2
Fig. 2. Diagram of the RHU (Shell) fixed bed process. The HDM and HDS phases take place in reactors R-01/R-02 and R-03/R-04 respectively.
ProcessesThe various processes are differentiated on the
basis of the type of reactors, their number andsequence, as shown in Table 2 (Leprince, 2001;Anonymous, 2002; Meyers, 2004). Fig. 1 gives an ideaof the positioning considered optimal for the varioustechnologies, as a function of the main contaminants(asphaltenes and metals) contained in the feedstock.
Fixed bed processes (the most common at the end ofthe twentieth century), though sound, do not appear ableto adequately treat feedstocks with high contents ofasphaltenes, metals and other heteroatoms, due mainlyto problems relating to the deactivation of the catalysts.
Technologies of the ebullated bed type performwell even with relatively heavy feedstocks. Slurryprocesses, which operate with dispersed catalysts,ensure good feedstock upgrading performances, inaddition to considerable flexibility.
Fixed bed and moving bed processesFixed bed processes are traditionally characterized
by the presence of HDM stages and partial cracking,plus the HDS stage, as in the example shown in Fig. 2(Giavarini, 1999).
The reactors, of trickle flow type, are normallylarge due to the low space velocities and the largequantities of catalysts needed. The temperature, whichincreases from top to bottom due to the exothermicnature of the reactions, is controlled by adding quenchgas. The temperatures do not normally exceed 400-420°C, and pressures may be up to 160 bar. Space velocities, usually low, must be such as toensure sufficient wetting of the catalyst.
In moving bed processes, which are relatively lesscommon, the feedstock and hydrogen may circulatein equicurrent, as in the simplified diagram shown inFig. 3, and in countercurrent. In these processes, too,the second stage reactors are generally of a fixed bedtype. The catalyst moves towards the bottom onlyduring operations to extract the depleted catalyst(from the bottom). The slight expansion of the catalyst, in the form of pellets, caused by theflow of the feedstock, creates some problems withfriction and mechanical erosion. However, thisdisadvantage and that represented by more complexoperating procedures is countered by the advantageof longer working cycles with respect to fixed beds.
Ebullated bed processesProcesses of this type are characterized by the fact
that the circulation of the feedstock and the hydrogenfrom bottom to top keeps the catalyst in suspension(Fig. 4); a recirculation pump for liquid productsregulates the expansion of the bed. This pump may beplaced inside or outside the reactor (Parkash, 2003).
These reactors are large, being up to 30 m tall andwith diameters of up to 5 m; the volume must besuitably increased to take account of the expansion ofthe catalytic bed (about 30-50%). In this case, too,some friction between the particles of catalyst isunavoidable. The flow is agitated, rather than of pistontype as in fixed bed reactors; the temperature profile is
312 ENCYCLOPAEDIA OF HYDROCARBONS
DEEP CONVERSION OF RESIDUES
products toseparation
fixedbed
HDSreactors
freshcatalyst
moving bedreactor
spentcatalyst
feedhydrogen
catalystaddition
max liquidlevel
expandedlevel
settledcatalyst level
distributorgrid plate
recyclingoil
gas/liquidproduct
to separators
leveldetectors
catalystwithdrawal
make-up H2and feed oil
Fig. 3. Section of reactors for a moving bed process.
Fig. 4. Diagram showing how an ebullated bed reactor works.
isothermal. This, alongside the modest andcontrollable loss of pressure, represents one of theadvantages of this type of process.
The catalyst may be added and removed eithercontinuously or intermittently, thus avoiding variationsover time in the yield and quality of products, typical offixed bed processes. However, the consumption ofcatalysts is higher than for fixed beds, since the catalystremoved is found in varying stages of saturation anddeactivation. Operating procedures are also morecomplex. With high conversion rates, furthermore, thequality of the residue is generally fairly poor.
Fig. 5 shows a diagram of the H-Oil process(Axens), which uses two ebullated bed reactors(Leprince, 2001). Since hold-up times are lower thanfor fixed beds, the temperatures (390-450°C) andabove all pressures (90-240 bar) are higher; spacevelocities may range from 0.1-0.9 h�1.
Slurry processesThese processes, to an even greater extent than
ebullated bed processes, are suitable for treatingresidues with a high impurity content. The unitconsists of one or more reactors (which may be offixed bed type) with the feedstock, hydrogen andcatalyst (in the dispersed phase) circulating frombottom to top.
The catalyst generally consists of finely dispersedmetal sulphides (of iron and/or molybdenum),generated in situ by the decomposition of a precursor.The catalyst does not promote cracking, which is ofexclusively thermal type, whilst it activates thedesulphurization, free radical quenching andhydrogenation reactions. The velocity of the liquidsand gas must be sufficiently high to keep the catalystdispersed; for the rest, conditions are similar to thoseof the preceding class of processes.
The processes under examination are often stillin the pilot or pre-industrial stage (see againTab. 2). A recent process, developed by ENI andnamed EST (ENI Slurry Technology), stands outfor the fact that it combines the HDM/HDN/HDSstage using a MoS2 catalyst with a deasphaltingand asphaltene recycling operation, and theoptional recycling of deasphalted oil. This processis also characterized by high conversion rates andhigh quality products, without the simultaneousproduction of fuel oil.
References
Dickenson R.L. et al. (1997) Refiner options for convertingand utilizing heavy fuel oil, «Hydrocarbon Processing»,February, 57.
Giavarini C. (1999) Guida allo studio dei processi petroliferie petrolchimici, Roma, Siderea.
Le Page J.F. et al. (1992) Resid and heavy oil processing,Paris, Technip.
Leprince P. (edited by) (2001) Conversion processes, Paris,Technip.
Meyers R.A. (editor in chief) (2004) Handbook of petroleumrefining processes, New York, McGraw-Hill.
O’Connor P. et al. (1991) Improve resid processing,«Hydrocarbon Processing», November, 76-84.
Parkash S. (2003) Refining processes handbook, Amsterdam,Elsevier.
Dipartimento di Ingegneria Chimica, dei Materiali,delle Materie Prime e Metallurgia
Università degli Studi di Roma ‘La Sapienza’Roma, Italy
313VOLUME II / REFINING AND PETROCHEMICALS
THE HYDROCONVERSION OF RESIDUES
H2purge
H2make-up
catalystaddition
catalystwithdrawal
fuel gas
residuevacuum residue recycle
fractionation
vacuumresidue feed
naphtha
gas oil
vacuumdistillate(to FCC)
Fig. 5.Simplifieddiagram of the H-Oil(Axens)ebullatedbed process.
7.2.2 LC-Fining process
The LC-Fining residuum hydroconversion processwas developed to specifically target hydrocrackingof the most difficult, heavy, lower-valuehydrocarbon streams (petroleum residua, heavy oilsfrom tar sands, shale oils, solvent refined coalextracts, etc.) at residue conversion levels of up to80% and higher.
When coupled with an integrated, fixed-bed,wide-cut hydrotreater/hydrocracker, the LC-Finingprocess produces high quality finished productswithout significant quantities of undesirableby-products.
Residuum upgrading process requirementsshould include the ability to: a) handle high heatsof reaction via efficient use of reactor volume; b) handle extraneous material withoutplugging; c) provide uniform distribution ofreactants and efficient contacting; d) operate overextended periods without shutdown. The nature of the LC-Fining process, makes itideally suited for the conversion of residues tolighter, more valuable products. The process canbe tailored for the feedstocks, the degree ofconversion, and the product qualities required,especially for the production of high-qualityresidual fuel oil with a low sulphur content andgood pipeline stability, or high quality syntheticcrude oils.
The process is based on technology initiallydeveloped and commercially demonstrated byCities Service, and subsequently improved andrefined by ABB Lummus Global, BP (formerlyAmoco Oil Company) and ChevronTexacoCorporation. With this process, heavy oil feeds –including gas oils, petroleum atmospheric andvacuum residue, coal liquids, asphalt, bitumenfrom tar sands, and shale oil – are hydrogenatedand converted into a wide spectrum of lighter,more valuable products such as naphtha, light andmiddle distillates, and atmospheric and vacuum gasoils. Products can be used as fuel oil, syntheticcrude, or feedstock for a coker, visbreaker, solventdeasphalter, or residuum catalytic cracker.Operating conditions and catalyst type and activitycan be varied to achieve the desired hydrocarbon
Process chemistryResidua and heavy oils differ from distillates in
that they contain asphaltenes and other highmolecular weight, highly polynuclear aromaticstructures and coke-forming constituents includingcarbon residue and organometallic compounds(with nickel and vanadium). Asphaltenes have amarked effect on the chemistry of hydrocrackingand result in the deposition of carbon and cokeprecursors on the catalyst. Such depositionsseriously affect the catalyst activity due, primarily,to the condensation and polymerization reactionsduring residue conversions (especially at deeperconversion levels).
Residuum hydrocracking is accomplished atrelatively high temperatures and high pressures inthe presence of hydrogen, and of a catalyst capableto hydrogenate the products and preventpolymerization of the free radicals as crackingreactions proceed.
The catalyst consists of a combination of metalsthat promote hydrogenation (e.g., cobalt andmolybdenum, or nickel and molybdenum)deposited on an alumina base.
The two most important reactions that takeplace in residuum hydrocracking, i.e. thermalcracking to lighter products and catalytic removalof feed contaminants, generally require operatingtemperatures between 385 and 450°C, hydrogenpartial pressures from 7.50 to 15.00 mPa and spacevelocities ranging from 0.1 to 0.8 vol h�1.
The chemical reactions in residhydroprocessing are sub-divided into three mainclasses: cracking of large molecules followed bysaturation (capping of free radicals); elimination of heteroatoms (S, N, O); demetallization (removalof vanadium, nickel and trace metals).Typicalchemical reactions during hydroprocessing areshown in Fig. 1.
Traditionally, the compounds present in theresiduum are generally classified as oils, resins,and asphaltenes. The asphaltenes are the high-molecular-weight material that typically contain a
314 ENCYCLOPAEDIA OF HYDROCARBONS
DEEP CONVERSION OF RESIDUES
large concentration of sulphur, nitrogen, metals,and highly condensed polynuclear aromatics.
Nitrogen removal is generally much moredifficult than sulphur removal. Some nitrogencompounds in the cracking reactions are merelyconverted to lower boiling range nitrogencompounds rather than being converted to NH3.
The highest concentration of metals (V and Ni)mostly reside in the asphaltene fraction with somein the resin fraction. The oil fraction tends to benearly free of metals. Metals are removed as metalsulphides. Unlike sulphur and nitrogen, which are converted and “escape” as H2S and NH3, thevanadium and nickel removed are absorbedon the catalyst. These metals are known to plugthe catalyst pores, and this pore blockage results incatalyst deactivation.
There are many factors that affect the sedimentformation rate and, consequently, the reactoroperability and resid conversion limits, including:a) asphaltene content; b) CCR reactivity; c)thermal severity (conversion); d) catalyst type andactivity; e) hydrogen partial pressure; f ) type andquantity of diluents; g) resin content; h) reactortemperature. Of these, the first six have the mostinfluence. Many pilot plant tests showed thatsediment (i.e. coke) precursor formation is directlyproportional to the asphaltene content of the feedand inversely proportional to the CCR reactivity.
The nature and origin of the coke precursors areoften not precisely known. However, a mechanismof their formation in processes that involve thermal
cracking in addition to hydrocracking andhydrogenation, such as the LC-Fining process, hasbeen generally postulated as described by thefollowing reaction chemistry:• Thermal cracking – formation of free radicals:
Rx�Ry�� Rx��Ry�
• Reaction of free radicals to form olefins orasphaltenes:
R�CH2�CH2��� R�CH��CH2�H�
R�CH2�CH2��Rx��� R�CH2�CH2�Rx
• Ermination or recapping of free radicals byhydrogenation:
Ry��H��� R y�H
At elevated temperatures, thermal crackingreactions generate free radical species due to therupture of carbon-carbon bonds. The free radicalsmay react with hydrogen in the presence of thecatalyst to form stable products. This reactionpredominates in the LC-Fining process where highhydrogen partial pressures are always maintained.If proper conditions are not maintained, the freeradicals may also combine with other free radicalsto form higher molecular-weight free radicals. Thischain reaction can continue until very highmolecular-weight, insoluble species (cokeprecursors/sediments) are produced. As thetemperature is increased to obtain higherconversions, the rate of generation of free radicals,and consequently coke precursors, can increase,creating phase separation and potential instabilityin the reactor if it is allowed to exceed thesolubility limit.
Control of coke precursors (organic sediments)can be accomplished in three ways: their formationis minimized or eliminated by using extremely highhydrogen partial pressures or very active catalyst;the coke precursors are maintained in solution byadding aromatic diluents; the coke precursors areremoved from the system.
The catalyst and aromatic diluents used in theprocess have an excellent ability to control theformation of these coke precursors. The continuousphysical removal of coke precursors (via filtration,centrifugation, etc.) from the reactor loop, can beaccomplished by bottoms recycle and removal ofthe coke precursors from the recycle stream, anapproach pilot tested and patented by Lummus forHigh Conversion LC-Fining.
Catalysts and reaction kineticsA series of catalysts are available for use in
LC-Fining units. The first-generation catalysts in
315VOLUME II / REFINING AND PETROCHEMICALS
THE HYDROCONVERSION OF RESIDUES
CH3
S
hydrodesulphurization
� 2H2 � H2S
N
C
CH3
C3H7
C6H13
C3H7 C H
hydrodenitrogenation
� 7H2 � NH3
hydrocracking
Cn C C C Cm � H2 Cn C C � Cm
� 7H2 �C2H5
�
C2H5
C2H5
�
C2H5
Fig. 1. Typical chemical reactionsin resid hydroconversion.
commercial use had adequate HDM(Hydrodemetallization)-HDS(Hydrodesulphurization) activity with acceptablesediment levels. These were less expensive thanmore recently developed, enhanced contaminantremoval/sediment control catalysts. Newgeneration catalysts are needed to produce lowsulphur fuel oils (from vacuum bottoms) of 1 wt%sulphur or less, with minimum sediment levels(�0.15 wt%) for pipeline stability. The otherrequirement of a good LC-Fining catalyst is tomaintain improved reactor operability/stability athigh temperature/deep resid conversions.
The residuum hydroprocessing catalysts aresmall (8 mm to 16 mm in size), extruded,cylindrical pellets made from an aluminium base.The pellets are impregnated with active metals(Co, Ni, Mo, W and other proprietary materials)that have good hydrogenation, demetallization,desulphurization and sediment control activity.Catalyst manufacturing processes are tailored tomanipulate physical and mechanical propertiessuch as: a)size (length and diameter); b) attritionresistance; c) crush strength; d) pore sizedistribution; e) pore volume; f ) effective surfacearea. Catalytic performance is affected by thecomplicated nature of the “active site” anddispersion and distribution of activators andpromoters.
Pore size control and distribution are keyfactors in the behaviour and formulation ofresiduum conversion catalysts. The pore sizes needto be sufficiently large to allow the diffusion of thelarge asphaltene molecules that require upgrading.Unfortunately, as the pore diameter increases, thesurface area and the hydrogenation activitydecrease. The diffusion of large molecules isreduced further because of pore mouth pluggingdue to carbon laydown and metal sulphidebuild-up from vanadium and nickel atoms that are removed from the resid feed. Metal sulphides are formed from the oxidativestate of the catalyst in the reactor environment(presulphiding reactions with sulphur in heavy oils, etc.).
Catalysts are also optimized for specificfunctions – such as metals removal, sulphurremoval, carbon residue reduction, and highconversion – while maintaining a clean product lowin organic sediments. The catalyst systemdeveloped by BP for their LC-Fining unit at TexasCity utilizes a proprietary demetallization catalystin the first reactor, and a high activity NI/Modesulphurization catalyst in the second and thirdreactors.
One of the key features of the LC-Finingprocess is the use of counter-current catalystaddition to optimize catalyst usage. Fresh catalyst is added to the third reactor, then itis reused by withdrawing it and adding it to thesecond reactor. The catalyst can then be useda third time, by withdrawing it from the second reactor and adding that material to thefirst reactor. Catalyst cascading results in higheroverall kinetics rate constants and, therefore,better overall catalyst utilization based on theconcentration of metals in the spent catalystdischarged from the first stage. This mode ofaddition/withdrawal has the added benefitof exposing the most highly converted resid to themost active catalyst. This reduces the sedimentformation in the last reactor and thus allowsreactor operability/conversion limits to beextended.
The rate at which a catalyst deactivates duringresidual oil hydrocracking is a complex functionof many parameters that can be categorized intothree distinct classes: the first consists of physicaland chemical properties of the residual feedstockto be processed; the second is concerned with thenature of the catalyst itself; the third is the effectof the operating conditions (temperature, spacevelocity, hydrogen partial pressure, etc.) used toobtain the desired levels of conversion anddesulphurization.
The most significant causes of catalystdeactivation are metals and carbon laydown.Concurrent with the desulphurization of residua isa demetallization reaction. Products from theresidual hydrocracking have nickel and vanadiumcontents markedly lower than the feed to the unit.The metals accumulate on the catalyst, causingdeactivation. It has been proposed that theorganometallics simply block the outer physicalsurface of the catalyst.
Carbon laydown on catalyst is influenced byfeedstock characteristics and conversion severity.Carbon accumulation is high in all operatingscenarios, ranging from slightly under 10 wt% forlow temperature HDS of atmospheric residuum, toover 40 wt% for high conversion of vacuumbottoms. The residuum hydrocracking is,apparently, diffusion controlled. It has been foundthat 8 mm catalyst performs better than 16 mmcatalyst.
Kinetics rate constants provide the fundamentalbasis for scaling-up from pilot plant to commercialconditions. As for hydrocarbon conversion, designfrom pilot plant data uses the back-mix-reactors-inseries model where kinetics is described in terms
316 ENCYCLOPAEDIA OF HYDROCARBONS
DEEP CONVERSION OF RESIDUES
of simple pseudo-first-order expressions; the rateconstant is given by the following expression :
100 1/NkHC�N ��1111��1�LHSV100�c
where kHC is the first-order hydrocarbon conversionrate constant, LHSV is the Liquid Hourly SpaceVelocity, C is the residue conversion, and N is thenumber of ebullated bed reactors in series.
As for residue hydrodesulphurization, kineticsis considerably more complex than that related tohydrodesulphurization of model organic sulphurcompounds or, for that matter, narrow-boilingpetroleum fractions (see Chapter 3.1). In publishedstudies of the kinetics of residuahydrodesulphurization, one of three approaches hasgenerally been taken:• The reactions can be described in terms of
simple first-order expressions.• The reactions can be described by use of two
simultaneous first-order expressions – oneexpression for easy-to-remove sulphur and aseparate expression for difficult-to-removesulphur.
• The reactions can be described using a pseudo-second-order expression.Referring to the latter approach and two back-
mix reactors in series model, it is possible to obtainan expression for the second-orderhydrodesulphurization rate constant kS. A similarexpression is used to evaluate the second-orderCCR reduction rate constant.
As for demetallization, the rate constant isgiven by a simple pseudo-first-order expression:
kV o kNi�2WDSV((Cf /Cp)0.5�1)
where kV and kNi are the first-order vanadium andnickel removal rate constants, WDSV is the DailyWeight Space Velocity (m3/g/day), Cf and Cp arethe concentrations of reacting species in the feedand product.
Process descriptionA simplified process flow diagram of an
LC-Fining unit with a close-coupled, integrated, fixedbed hydrotreater/hydrocracker is shown in Fig. 2.
Oil feed and hydrogen are heated separately,combined, and then passed into the LC-Fining
317VOLUME II / REFINING AND PETROCHEMICALS
THE HYDROCONVERSION OF RESIDUES
feed
H2heater
expa
nded
bed
rea
ctor
fixe
dbe
d re
acto
r
HP/HTseparator
HP/HTseparator
LP/HTseparator
LP/MTseparator
LP/LTseparator
washwater
make-upH2
H2S
amine
fuel gas
products
residue
LP/MTseparator
LP/LTseparator
oilheater
recycleheater
distillaterecycle
surge drumrecyclestripper
stea
m
Fig. 2. Process flow sketch : LC-Fining unit withintegrated distillate hydrotreating. HP, High Pressure;HT, High Temperature; MT, Medium Temperature; LP, Low Pressure; LT, Low Temperature.
hydrocracking reactor (Fig. 3). Fresh feed andhydrogen enter the reactor at the bottom and passup through a catalyst bed where, under the effectsof time, temperature and hydrogen pressure, andaided by catalysts, the feed oil is cracked andhydrogenated to produce lighter, higher-qualityproducts. A portion of the liquid product from thelarge pan at the top of the reactor is recycled,through the central downcomer, by means of apump mounted in the bottom head of the reactor.This flow provides the velocity necessary to keepthe catalyst bed in a state of motion somewhatexpanded over its settled level (i.e. ebullated).Moreover, this flow aids in maintaining nearisothermal reactor temperature. As a matter of fact,the hydrocracking reactions are exothermic,resulting in a temperature rise from inlet to outletdepending upon the reaction operating severity.However, because of the mixing effect of theinternal recycle, the bed operates essentiallyisothermically. The catalyst level is monitored andcontrolled by radioactive density detectors, wherethe source is contained inside the reactor and thedetectors are mounted outside. Temperature ismonitored by internal couples and skin couples.The performance of the ebullated bed is
continuously monitored and controlled with thedensity detectors and temperature measurementsthat verify proper distribution of gas and liquidthroughout the catalyst bed. Temperature deviationsoutside the normal expected ranges that mightsuggest maldistribution will cause the DistributedControl System (DCS) to activate alarms or initiateautomatic shutdown on the heaters, hydrogen feed,and/or reactor section, as required.
Catalyst is added and withdrawn likewise tomaintain an equilibrium catalyst activity withoutthe need for unit shutdown. The reactors can bestaged in series, where the product from the firstreactor passes to a second reactor and, if necessary,to a third reactor. After the final reactor, theproduct goes to a high pressure/high temperatureseparator.
Vapour, effluent from the separator, is let downin pressure before heat exchange, removal ofcondensates, and purification. Handling the recyclegas at low pressure offers considerable savings ininvestment over a conventional high pressurerecycle gas purification system.
After stripping, the recycle liquid is pumpedthrough the coke precursor removal step (a physicalmeans of separation such as centrifugation, etc.)where very small quantities of insoluble heavyhydrocarbons or carbonaceous solids are removed.The clean liquid recycle then passes to the suctiondrum of the feed pump. The net product from thetop of the recycle stripper goes to fractionation; netheavy oil product is directed from the stripperbottoms pump discharge to vacuum fractionation.
The reactor effluent vapour, along withdistillate recovered from the heavy oil stripperoverhead, any virgin atmospheric gas oil recoveredin the prefractionator upstream of the LC-Finingunit, and vacuum gas oil recovered in the vacuumfractionator, are all charged to a “wide-cut”, close-coupled, integrated, fixed bedhydrotreater/hydrocracker located immediatelydownstream from the last ebullated bed reactor.The inlet temperature to the first bed is controlledby adjusting the amount of heat extracted from thereactor vapour stream and the temperature of thedistillate liquid is controlled by a combination ofhydrogen and liquid quench. The effluent from the hydrotreating reactors isseparated into a vapour and heavy distillate liquidstream, with the liquid stream routed to thehydrotreated distillate fractionator. The vapourstream is amine treated, purified through a PressureSwing Adsorption (PSA) or membrane unit,recompressed and recirculated to the LC-Finingreactors.
318 ENCYCLOPAEDIA OF HYDROCARBONS
DEEP CONVERSION OF RESIDUES
density detectorradiation sourcewell
catalyst addition line
densitydetectors
skinthermocouples
thermowellnozzle
effluent
catalystwithdrawal
line
normal bedlevel
feed
recycle pump
Fig. 3. LC-Fining reactor.
The high conversion (>80%) LC-Fining processdiffers from the basic process in that bottomsrecycle is practiced. The recycle liquid is let downin pressure and passes to the recycle stripper whereit is fractionated to the proper boiling range forreturn to the reactor. In this way, the concentrationof bottoms in the reactor, and therefore the reactionproducts slate can be controlled.
Technology featuresSeveral advances in the LC-Fining resid
hydroconversion technology have significantlyreduced the capital investment and have furtherextended the conversion limits and processinglimitations. These include: a) H2 purificationsystems; b) integrated fixed bedhydrotreater/hydrocracker; c) interreactorseparator/stripper; d ) modified liquid recycle pan;e) reactor bottom head feed distributor; f ) improved reactor distributor design.
Hydrogen purification systemsIn early designs, a lean oil system was used to
purify the recycle gas and the maximum purityachievable was 82%. In 1984, Lummus developedand patented a low pressure H2 purification system,which has been utilized in all commercialoperating units since. With low pressure H2purification, the gas exiting the last reactor isimmediately let down and cooled at low pressureand then purified in a PSA unit. This permits highhydrogen treat gas purities, generally exceeding 97vol%. As a result, treat gas circulation rates werereduced by 50 to 60% and reaction system designpressure by 10%, while still satisfying thehydrogen partial pressure requirements. Thischange, in conjunction with the replacement of acertain amount of high pressure equipment withlow pressure equipment, significantly reduced theunit investment cost. However, with this lowpressure H2 purification scheme the recycle gashad to be recompressed from low pressure back toreactor operating pressure, requiring an increase of25 to 30% in the overall power consumption.
In 1998, the use of membranes was evaluatedfor purification of the recycle gas, and similar treatgas purities were achieved with membranes as witha PSA system. Membranes allow the same reducedreaction system pressure and lower treat gascirculation rates as with a PSA unit, but with theadded benefit that the purified recycle gas isavailable at higher pressures. Consequently, therecycle gas can be recompressed in a single stageof compression versus the two or three stages ofcompression required with a PSA system, resulting
in a 20% reduction in the overall unit powerconsumption. In addition, based on current highpressure equipment pricing, the unit investment isslightly less with membranes than with a PSA unit.
By using high purity recycle gas, it is possibleto achieve the desired hydrogen partial pressurewith much lower hydrogen gas rates. The low gasrates have two primary benefits: they reduce unitinvestment and velocity and, therefore, the gashold-up in the reactor. This provides for moreliquid residence time per unit reactor volume,thereby reducing the reactor volume. The use oflow treat gas rates is utilized in all commerciallyoperating LC-Fining units, which work witha total-hydrogen-to-chemical-hydrogenratio of 2.5 to 3.
Integrated fixed bed hydrotreater/hydrocrackerSeveral designs incorporated a close-coupled,
integrated, fixed-bed hydrotreater/hydrocrackerimmediately downstream of the ebullated bedreactors. In this design, the vapour stream from theebullated bed reactors, plus the distillate recoveredfrom the heavy oil stripper overhead and thestraight run atmospheric and Vacuum Gas Oils(VGO), are fed to a fixed bedhydrotreater/hydrocracker operating at essentiallythe same pressure level.
Excess hydrogen contained in the ebullated bedreactor effluent vapour is used to hydrotreat thedistillate fractions. Additional hydrogen, equivalentonly to the chemical hydrogen consumed in thefixed bed reactor, is introduced as quench to thesecond and third catalyst beds.
The ebullated bed reactor effluent vapour isfirst contacted with VGO in a wash tower in orderto remove any potential resid entrainment andentrainment of catalyst fines from the fixed bedreactor. By incorporating the fixed bedhydrotreater within the LC-Fining reaction system,the HP system service count is greatly reduced. Inaddition, since excess hydrogen in the LC-Finingreactor effluent vapour is used to hydrotreat thestraight run and LC-Fining distillate fractions, theneed for additional recycle gas compression iseliminated. As a result, the investment issignificantly reduced compared to a stand-alonehydrotreater/hydrocracker.
Interreactor separator/stripperRecently proposed process design
configurations incorporate the use of aninterreactor separator/stripper, which permitshigher liquid capacities to be achieved for a givenreactor cross-section. Gas superficial velocities
319VOLUME II / REFINING AND PETROCHEMICALS
THE HYDROCONVERSION OF RESIDUES
through the downstream reactors are reduced byseparating the vapour between reaction stages androuting it to the final reactor effluent separator.This design provides for parallel flow of gas toeach reaction stage while maintaining the benefitsof series flow liquid operations. In a conventionallydesigned unit, the effluent vapour from theupstream reactor is combined with additionaltreat/quench gas and all the vapour is directed tothe downstream reactors.
Liquid recycle pan and bottom head feeddistributor
A modified, two-stage liquid recycle pan designhas increased the conversion (about 4%) in thereactor and minimized upsets associated with therecycle pump bed expansion system. Subsequently,a new pan design has been developed permittingoperation at still higher capacities and treat gasrates.
In 1985-86, a bottom head feed distributor wasadded to the LC-Fining reactors. It allows for bettermixing of the feed oil, gas and recycle oil, whileproviding better distribution of oil and gas to thecap and riser assembly.
Improved reactor distributor designThe primary reactor grid distributor is a
bubble cap type with slotted risers for distributionof vapour and liquid to each cap. Each risercontains a seat and ball to prevent back-migrationof catalyst below the grid into the reactor bottomhead should the recycle pump stop. The seatcontains a small v-notch to permit oil to bedrained from the reactor.
Amoco Oil Company performed substantialcold flow modelling of the distributor grid thatled to the installation of catalyst slides. This wasfound to reduce instances of localized catalystsettling near the wall, maintaining a cleanerreactor environment and increasing the run lengthbetween turnarounds.
Process flexibilityThe LC-Fining unit has great inherent flexibility to
meet variations in feed quality/throughput, productquality and reaction operating severities (temperature,space velocity, conversion, etc.). This flexibility is adirect result of the ebullated catalyst bed reactor system. In an ebullated bed unit, if the metalsor sulphur content of the feed increases, the productquality is maintained by increasing catalystconsumption.
The world’s first ebullated bed residual upgrader,operated by Cities Service Oil Company, utilized this
flexibility to process atmospheric bottoms, FCC(Fluid Catalytic Cracking) heavy cycle oil, propanedeasphalter bottoms, and vacuum bottoms.
Sufficient operating flexibility is also normallyforeseen in the design to enable the unit to operatein the future with vacuum bottom recycle whichprovides for future options to increase eitherconversion or unit throughput.
A wide range of heavy oils have been processedin LC-Fining units. For example, the BP unithandles many of the poorest quality vacuumresidua in the world, including Mexican,Venezuelan and Middle Eastern. Feed typically isunder 5°API and has more than 4 wt% sulphur andmore than 400 ppm metals. LC-Fining unit productyields for processing Arabian Heavy VacuumBottoms to conversion levels of 40%, 65%, and80% are listed in Table 1. All of these conversionscan be achieved in the same plant.
The plant and product properties are estimatedfrom generalized correlations that were derivedfrom extensive pilot plant and commercial data.Typical product properties for a 65 vol%conversion case are shown in Table 2.
Operating variablesThe main operating variables in the ebullated
bed residuum hydroconversion process are thefollowing: a) feed quality; b) reactor temperature;c) hydrogen partial pressure; d) space velocity; e) treat gas to oil ratio; f ) catalyst addition rate.Typical ranges of operating parameters are reportedin Table 3.
Feed QualityThe quality of feed (gravity, S, N, CCR,
metals, viscosity, C5 and C7, asphaltenes, ashcontent, trace metals, Na, Ca, Si, Fe, distillation,etc.) has a decisive effect on the choice ofoperating variables and the final processingobjectives. Variations in contaminant levelsinfluence the chemical hydrogen consumption andheat release.
Higher contents of asphaltenes, metals, CCRlead to a reduction in the catalytic activity. Higher viscosity/gravity feeds significantly affectthe bed hydrodynamics resulting in higher gashold-up.
Vacuum residua with varying paraffinic,aromatic, naphthenic and unsaturatedhydrocarbons contents have significantly differentcracking/HDS/HDN(Hydrodenitrogenation)/HDM/HCCR(Hydro Conradson Carbon Residue) reaction rate constants. Asphaltenes aredifficult to crack and saturate and may
320 ENCYCLOPAEDIA OF HYDROCARBONS
DEEP CONVERSION OF RESIDUES
polymerize on the catalyst surface ascarbonaceous deposits.
Heavy metals, particularly nickel andvanadium, in addition to alkalis (such as sodium)
and alkaline earths (such as calcium andmagnesium) are strongly attracted to the catalystand irreversibly destroy its activity. These metals are found predominately in
321VOLUME II / REFINING AND PETROCHEMICALS
THE HYDROCONVERSION OF RESIDUES
Table 1. Typical LC-Fining unit product yields
Crude source Arabian Heavy vacuum bottoms + cat cracker HCO
Hydrogen consumption S m3/m3 Fresh Feed 168 155 221 283
Table 2. LC-Fining unit product properties
Arabian Heavy 65% vol. conversion
Boiling range, °C C5-165 165-370 370-550 550+
Wt% on feed 5.27 26.50 33.71 28.25
Vol.% on feed 7.60 31.50 36.95 28.00
Gravity, °API 61.2 31.2 19.0 4.6
Sulphur, wt% 0.01 0.11 0.53 1.6
Nitrogen, wt% 0.02 0.08 0.19 0.45
Aniline point, °C 50 73
Cetane index 41
Conradson Carbon, wt% 26.3
Metals: vanadium, wppm 48nickel, wppm 26
Viscosity, cSt at 25°C 4.699°C 1.2 7.8
150°C 3.1 70177°C 30
C7, Asphaltenes, wt% 9.3
asphaltenes; however, they are also present inporphyrins. Porphyrins are large, condensed-ringmolecules that boil at a lower temperature thanasphaltenes.
Iron, silica and calcium carried in with the feedcan be troublesome trace metal contaminants.These may be chemically combined with the heavyhydrocarbon molecules. In either case, these result not only in deactivationof the catalyst through pore mouth plugging, butalso plug the catalyst interstices causing excessivepressure drop.
Reactor temperatureAt a given space velocity, the ebullated bed
reactor temperature influences the hydroconversionand contaminant removal rate constants. However,an increase in temperature also accelerates thesecondary undesirable reactions of carbonlaydown, catalyst deactivation, build-up ofimpurities on active sites, formation of hydrogendeficient coke precursors, pore mouth plugging,sintering, etc.
Hydrogen partial pressureHydrogen partial pressure is of fundamental
importance for the successful operation of the LC-Fining reactors. It has an important effect oncatalytic activity and fouling rate. An increase in hydrogen partial pressuresuppresses the coke fouling of catalyst andimproves the reaction rates (first or second-orderwith respect to H2 partial pressure) for the catalyticand non-catalytic reactions. For a given operation, hydrogen partial pressure can be
increased by: increasing total system pressure,increasing make-up hydrogen purity andcirculation rate, increasing recycle gas purity and circulation rate.
Space velocityThe severity of reactor operation is established
not only by the reaction temperature, but also bythe Liquid Hourly Space Velocity (LHSV) whichestablishes the residuum conversion targets: ahigher LHSV (feed rate) would require highertemperature for the same resid conversion andproduct quality.
Treat gas to oil ratioHigher hydrogen partial pressures are required
for the completion of hydroconversion reactions.Higher treat gas to oil ratio improves thehydrogen partial pressures. However, optimum balance is required to reduce thecompression costs and excessive gas hold-up athigher gas rates. Typically, the hydrogencirculation rates through the reactors are muchhigher than the stoichiometric (chemical)hydrogen requirements. In order to reduce thetotal make-up hydrogen requirements, themajority of the excess gas is purified and recycledwith make-up being equal to chemical, solubilityand mechanical losses. Typically, the excess overchemical treat gas to oil ratios vary between 2 to4 depending on the purity of make-up and recyclehydrogen.
Catalyst addition rateEquilibrium catalyst activity influences two
basic requirements. These are resid conversion tolighter material and hydrogenation severity.The hydrogenation reactions (heteroatom removal, demetallization and CCR reduction) are promoted by Co/Mo, Ni/Mo, Ni/Wcomponents. As the ongoing hydrogenationreactions proceed, the activity of the catalystdeclines while metals and carbon build-up on thecatalyst. Thus, in the LC-Fining reactors, thedesired equilibrium catalyst activity is maintainedby adding fresh catalyst and removing spentcatalyst. The catalyst addition rate is an importantvariable to achieve desired product quality goals.
Bibliography
Gupta A. (2002) in: Proceedings of the LC-Fining symposium,Milazzo (Italy), Agip research center at CRS.
322 ENCYCLOPAEDIA OF HYDROCARBONS
DEEP CONVERSION OF RESIDUES
Table 3. Typical ranges of operating parametersand variables
Gupta A. (2003) Chevron Lummus Global ebullated bed bottomof the barrel hydroconversion (LC-Fining process), in:Meyers R.A. (editor in chief) Handbook of petroleumrefining processes, New York, McGraw-Hill, Chapter 8.7.
Gupta A., Baldassarri M. (2004) Recent advances inebullating bed residue hydrocracking (LC-Fining) process,in: Proceedings of the Advanced refining technologiesworkshop, Napa (CA).
Gupta A. et al. (1991) Update on LC-Fining technology, in:Proceedings of the Petroleum Society of CIM and AOSTRA,Alberta.
