The present invention relates to a process for the selective hydrodesulfurization of naphtha streams containing sulfur and olefins. An olefinic naphtha stream is hy- drodesulfurized at a high gas to oil ratio, resulting in effective hydrodesulfurization and maintenance of octane values.

The requirements to sulfur levels in gasoline have continually been increased, recently to below 10 ppmw. In general, this will require deep desulfurization of olefinic naphthas. Deep desulfurization of naphtha requires improved technology to reduce sulfur levels without the severe loss of octane number that accompanies the undesirable saturation of olefins.

Hydrodesulfurization is a hydrotreating process for the removal of feed sulfur by con version to hydrogen sulfide. Conversion is typically achieved by reaction of the feed with hydrogen over non-noble metal sulfided supported and unsupported catalysts, es- pecially those of Co/Mo and Ni/Mo. Severe temperatures and pressures may be re- quired to meet product quality specifications by conventional means.

Olefinic cracked naphthas and coker naphthas typically contain more than about 20 weight percent olefins. At least a portion of the olefins are hydrogenated during con ventional hydrodesulfurization. Since olefins are relatively high octane number compo- nents, it is desirable to retain the olefins rather than to hydrogenate them to saturated compounds. Conventional hydrodesulfurization catalysts have both hydrogenation and desulfurization activity. Hydrodesulfurization of cracked naphthas using conventional naphtha desulfurization catalysts under conventional conditions required for sulfur re moval, results in a significant loss of olefins through hydrogenation. This results in a lower grade fuel product that needs additional refining, such as isomerization, catalytic reforming, blending, etc., to produce higher octane fuels. This, of course, adds signifi cantly to production costs.

Selective hydrodesulfurization involves removing sulfur while minimizing hydrogenation of olefins and octane reduction by various techniques, such as selective catalysts, sep aration of feedstocks, with individual treatments of fractions at specific process condi- tions, or both.

In regular hydrodesulfurization processes, the gas to oil ratio (GOR) is typically kept below 500 Nm ³ /m ³ since it has been believed that higher GOR will push the reaction towards a higher hydrogenation of the olefins. In addition, there has been little motiva- tion to increase GOR, as a higher GOR will be related with additional cost due to a re- quirement for excess hydrogen circulation in the process, and an elevated consumption of hydrogen by reactions forming products without increased value has also been as- sumed. The typical pressure for such processes have been around 20-25 barg, in an expectation of lower catalyst deactivation compared to lower pressures. In addition, 20- 25 barg is also a common hydrogen supply pressure.

It has however, now surprisingly been discovered that GOR above 500 Nm ³ /m ³ in com bination with moderate absolute pressures, contrary to expectations will have the effect of enabling a higher desulfurization with reduced loss of octane numbers. The selectiv ity balance between hydrodesulfurization and the hydrogenation of olefins is a possible definition of the overall quality of a FCC naphtha post-treat process. In US 7,629,289 the best performing catalyst was able to obtain a selectivity of 6.5, and processes ob taining more than 60-80% level of hydrodesulfurization are typically performing at or below such a selectivity.

The present invention is a process enabling an improved selectivity towards hy drodesulfurization over hydrogenation of olefins, by increasing the GOR and limiting the pressure, which is commercially attractive, since the value of naphtha is highly re- lated to the octane number. Traditionally the octane number has been maintained by providing process modifications increasing the complexity of processes or by develop- ment of complex specific catalysts.

The gas to oil ratio shall in accordance with the terminology of the skilled person of re- finery technology in the following be construed to mean the ratio between hydrogen containing gas and naphtha feedstock, as determined by the individual flows of the streams at the point where the hydrogen containing gas and the feedstock are mixed.

In the present text the term GOR is used as an abbreviation for the gas to oil ratio. The two terms shall be construed as fully equivalent. The unit for GOR is given as Nm ³ /m ³ . The numerator of the unit (Nm ³ ) shall be understood as“normal” m ³ , i.e. the amount of gas taken up this volume at 0°C and 1 atmosphere and the denominator of the unit (m ³ ) shall be understood as the volumetric flow of oil at standard conditions, typically at

60°F and 1 atmosphere. The pressure and temperature shall in accordance with the terminology of the skilled person of refinery technology in the following be construed as the pressure and tem- perature respectively at the inlet of a reactor.

The hydrogen partial pressure shall be construed as the partial pressure of hydrogen in the treat gas.

The space velocity shall in accordance with the terminology of the skilled person of re- finery technology in the following be construed as the LHSV (liquid hourly space veloc- ity) over a single catalytically active material unless otherwise indicated.

The initial boiling point (IBP), the final boiling point (FBP) and the temperatures corre- sponding to recovered amounts of sample, shall be understood in accordance with the ASTM D86 standard. T ₅ , T10, T ₅ o and T95 boiling points shall accordingly be understood as the distillation temperatures where 5vol%, 10vol%, 50vol% and 95vol% respectively have been recovered.

The research octane number (RON) shall be understood as the octane number meas- ured in accordance with ASTM D2699.

Olefins shall in accordance with the IUPAC definition and the language of the skilled person be understood as acyclic and cyclic hydrocarbons having one or more carbon- carbon double bonds.

Di-olefins shall similarly be understood as acyclic and cyclic hydrocarbons having two or more carbon-carbon double bonds.

The severity of reaction conditions shall be understood as the extent to which a given reaction will take place. Hydrodesulfurization severity, shall be understood as being increased if one or more physical or chemical conditions are changed in a way having the consequence that the degree of hydrodesulfurization is increased.

Where concentrations are stated in vol% or ppmv this shall be understood as volume/volume % and volume/volume parts per million.

Where concentrations are stated in wt% or ppmw this shall be understood as weight/weight % and weight/weight parts per million.

Where an amount of sulfur is specified, this shall be construed as the wt% of atomic sulfur, relative to the total stream. Where an amount of organic sulfur is specified, this shall be construed as the wt% of atomic sulfur in organic molecules, relative to the total stream.

The term conversion shall be construed as the net conversion, as calculated from the inlet concentration of a species and the outlet concentration of a species relative to the inlet concentration of the species.

The terms extent of hydrodesulfurization, HDS conversion or %HDS shall be consid- ered equivalent, unless stated otherwise, and shall be construed as the net conversion of organic sulfur to inorganic sulfur, as calculated from the wt% of atomic sulfur in or ganic molecules (e.g. excluding H2S) in the inlet stream and the outlet stream.

The terms extent of hydrogenation of olefins, olefin saturation or %OSAT shall be con- sidered equivalent, unless stated otherwise, and shall be construed as the net conver- sion of olefins, as calculated from the wt% of olefinic molecules in the inlet stream and the outlet stream.

The term selectivity shall be construed as the ratio between the extent of hydrodesulfu- rization and the extent of hydrogenation of olefins, e.g. %HDS/%OSAT.

A broad aspect of the present disclosure relates to a process for hydrodesulfurizing an olefinic naphtha feedstock while retaining a substantial amount of the olefins, which feedstock has a T95 boiling point below 250°C and contains at least 50 ppmw of organi cally bound sulfur and from 5% to 60% olefins, said process comprising hydrodesulfu- rizing the feedstock in a sulfur removal stage in the presence of a gas comprising hy drogen and a hydrodesulfurization catalyst, at hydrodesulfurization reaction conditions including a temperature from 200°C to 350°C, a pressure from 2 barg or 5 barg to 10 barg, 15 barg, 25 barg or 35 barg, and gas to oil ratio from 500 Nm ³ /m ³ , 600 Nm ³ /m ³ , 700 Nm ³ /m ³ or 750 Nm ³ /m ³ to 900 Nm ³ /m ³ or 1000 Nm ³ /m ³ , to convert at least 60 % of the organically bound sulfur to hydrogen sulfide and to produce a desulfurized product stream, with the associated benefit of such a process providing a lower octane loss at all severities above 60% HDS, compared to a process with similar conversion of or ganic sulfur with a lower gas to oil ratio, as measured by the selectivity slope, (%HDS- %OSAT)/(%OSAT ^* (100-%HDS)), while avoiding excessive increase of equipment size by limiting gas to oil ratio.

In a further embodiment the process severity is configured for converting at least 70%, 80% or 90% of the organically bound sulfur to hydrogen sulfide, with the associated benefit of providing a process suitable for current sulfur regulation with minimal octane loss.

In a further embodiment the gas to oil ratio and the pressure is configured for the selec- tivity slope, (%HDS-%OSAT)/(%OSAT ^* (100-%HDS)), to be above 0.55 or 0.7, with the associated benefit of obtaining desirable process configurations under the guidance of the parameter selectivity slope.

In a further embodiment less than 30% or 50% of the sulfur in the feedstock directed to the sulfur removal stage is found in mercaptans. Such a feedstock would be in need of a severe but specific hydrodesulfurization. The guidance of the selectivity slope is es- pecially suitable for such a more difficult feedstock, typically be found in FCC products, not having undergone significant hydrodesulfurization.

In a further embodiment the liquid hourly space velocity (LHSV) is from 1 .1 hr ¹ to 3 hr ¹ , with the associated benefit of a low LHSV being a possibility to employ the increased selectivity by an increase in process severity, without sacrificing olefins.

In a further embodiment the process further comprises the steps of

b) separating the feedstock in at least a heavy naphtha stream and a light naphtha stream according to boiling point

c) directing said heavy naphtha stream as the feedstock of said hydrodesulfurizing step, providing a desulfurized product stream

d) optionally directing the light naphtha stream as the feedstock to a further sulfur re- moval stage, providing a light desulfurized naphtha stream

e) combining said desulfurized product stream and either said light naphtha stream or said light desulfurized naphtha stream to form a final product stream,

with the associated benefit of such a process having a lower octane loss compared to a similar process without separation of the feedstock.

In a further embodiment said hydrodesulfurization catalyst comprises 0.5% or 1% to 5% cobalt and/or nickel and 3% to 20% molybdenum and/or tungsten, on a refractory support, with the associated benefit of such a catalyst being cost effective for hy drodesulfurization.

In a further embodiment said hydrodesulfurization catalyst comprises 0.5% or 1% to 5% cobalt and 3% to 20% molybdenum with the associated benefit of such a catalyst being cost effective for hydrodesulfurization and having limited activity in olefin satura- tion.

In a further embodiment said refractory support comprises alumina, silicaspinel or silica-alumina, with the associated benefit of such a support being highly robust. Alumina and silica shall be construed as materials of synthetic or natural origin being dominated by the oxides of aluminum and silicium. Alumina-silica shall be construed as a mixture, in any ratio, on any level down to atomic level of these oxides. Spinel shall be con- strued as an oxidic material comprising magnesium and aluminum in a common crystal structure.

In a further embodiment said step (c) comprises the substeps

(x) directing said heavy naphtha stream as the feedstock of a first hydrodesulfurizing step, providing a desulfurized heavy product stream

(y) optionally separating the desulfurized heavy product stream in a at least a desulfu- rized heavy naphtha stream and a gas stream,

(z) further desulfurizing the heavy desulfurized naphtha product stream, providing the desulfurized product stream

wherein the conditions and catalytically active material of steps (x) and (z) may be simi- lar or different with the associated benefit tailoring the catalytically active material of steps (x) and (z) to the relevant requirements for conversion of sulfur, and with the as- sociated benefit of removing hydrogen sulfide which may interfere with the hydrodesulfurization of step (z).

In a further embodiment said step (x) converts at least 75%, 80% or 85% of the organically bound sulfur to H2S, with the associated benefit of the high GOR and/or H20R of the process allowing such a severe HDS step, while avoiding excessive saturation of olefins.

In a further embodiment said step (y) is present and involves the steps (p) separating the desulfurized heavy product stream in at least a desulfurized heavy naphtha stream, a desulfurized intermediate naphtha stream and a gas stream, and one or both of the steps

(q) further desulfurizing the intermediate desulfurized naphtha product stream, provid- ing the intermediate desulfurized product stream and (r) combining two or more of the intermediate desulfurized product stream, the heavy desulfurized product stream, the light naphtha stream and the light desulfurized naph- tha stream to form a final product stream, with the associated benefit of providing even more possibility to fine tune the materials and conditions of the process.

In a further embodiment the process for hydrodesulfurizing the olefinic naphtha feed stock retains at least 20%, 40%, 60% or 80% of the olefins in the olefinic naphtha feed stock with the associated benefit of such a process providing a hydrocarbon being use ful as a component in a high octane low sulfur gasoline.

In a further embodiment the process further comprises a step of selective diolefin hy- drogenation prior to said hydrodesulfurizing step, with the associated benefit of reduc- ing the risk of polymerization of diolefins in the process and of reacting olefins and mer- captans to convert low-boiling mercaptans to higher boiling sulfides. The reaction be tween olefins and mercaptans has the effect of providing a light naphtha fraction corn- prising olefins and little or no sulfur and a heavy naphtha fraction comprising few ole- fins and the majority of sulfur. Such two fractions may be separated and treated individ ually.

In a further embodiment, the selective diolefin hydrogenation reaction conditions in volves a temperature from 80°C, 90°C, 100°C or 150°C to 200°C, a pressure from 2 barg or 5 barg to 40 barg or 50 barg, and gas to oil ratio from 2 Nm ³ /m ³ , 5 Nm ³ /m ³ or 10 Nm ³ /m ³ to 20 Nm ³ /m ³ or 25 Nm ³ /m ³ to convert at least 80% or 90 % of the diolefins to alkanes or mono-olefins or by reaction with mercaptans to sulfides, with the associ- ated benefit of such conditions being effective in hydrogenation of diolefins, with mini- mal mono-olefin saturation, and thus minimal RON loss. In addition, the conditions are effective in formation of sulfides from mercaptans and olefins, which has the potential effect of providing a light naphtha fraction comprising olefins and little or no sulfur and a heavy naphtha fraction comprising few olefins and the majority of sulfur. This difference in characteristics between light naphtha fraction and heavy naphtha fraction may be employed in specific treatment of the two fractions.

In a further embodiment the selective diolefin hydrogenation reaction conditions involve a temperature from 80°C, 90°C or 100°C to 200°C, a pressure of 5 barg to 40 barg or 50 barg, and a gas to oil ratio of 250 Nm ³ /m ³ to 2500 Nm ³ /m ³ to convert at least 80% or 90 % of the diolefins to alkanes or mono-olefins or by reaction with mercaptans to sul- fides, with the associated benefit of such a process not requiring separate hydrogen addition in the diolefin hydrogenation and hydrodesulfurizing steps.

The rate of a chemical process is controlled by chemical kinetics. Typically, reaction rates increase with increased temperature, increased reactant concentration, de creased product concentration and decreased space velocities (i.e. increased resi- dence times), but the relations may be more complex than expected, due to the nature of reaction mechanisms on the microscopic level. Especially in refinery processes, in- creasing the factors which increase reaction rates will be called increased severity of the process.

Hydrogenation processes are often employed in the conversion of hydrocarbons, e.g. for the removal of sulfur by hydrodesulfurization (HDS). The severity of hydrogenation is typically increased by increasing temperature, hydrogen partial pressure, the gas to oil ratio (GOR) or decreasing the space velocity.

A common intermediate product in refineries is naphtha withdrawn from a fluid catalytic cracker, which is suitable for use as gasoline. The amount of sulfur in this FCC naphtha is typically too high to be included in final gasoline product, and the sulfur is often re- duced by hydrotreatment, but at the same time it is desired that the amount of olefins is maintained, as removal of these would lead to a reduced octane number of the final gasoline product. As desulfurization as well as olefin saturation are hydrogenation pro- cesses the immediate expectation is that increasing the hydrogenation severity to ob tain a high extent of HDS will be associated with a high sacrifice of octane number due to olefin saturation. A further aspect of FCC naphtha post-treat is that the presence of di-olefins is undesired, as diolefins, which may be present in a concentration from 0.1 %, 0.5% or 1 % to around 5%, may polymerize and form solid products which will block the reactor.

The strategy for balancing a high diolefin saturation, a high HDS activity and a low ole- fin saturation has often been based on specific process conditions in combination with the choice of selective catalysts. For the diolefin-hydrogenation a nickel-molybdenum catalyst operating at low GOR and low temperature has been preferred, since the less severe conditions will not result in high hydrogenation of mono-olefins. EP 0 725 126 propose to split the FCC naphtha to be desulfurized in a light and a heavy naphtha stream, and treat these differently - e.g. by only hydrotreating the heavy naphtha stream, which will have the highest amount of sulfur, or by hydrotreating the heavy naphtha stream in two steps with or without intermediate separation. Often the first hy- drotreatment step is carried out in the presence of a cobalt/molybdenum catalyst, which is more active in HDS than in olefin saturation.

Recent environmental standards require the sulfur content to be as low as 10 ppm in gasoline. To obtain this for a feed with 1000 ppm sulfur as much as 99% HDS will be required. It is well known that this may be obtained by increasing the severity of the HDS process by increasing the temperature or hydrogen partial pressure. This in- crease in temperature or hydrogen partial pressure will however have the drawback of also increasing the olefin saturation, such that the octane number and thus the gaso line value is reduced.

Similarly, the decrease of space velocity may also result in increased HDS, but also in this situation a sacrifice of octane number is observed.

According to the prior art, the GOR for HDS of FCC naphtha has typically been 300 Nm ³ /m ³ to 500 Nm ³ /m ³ , but studies of the effect of varying GOR have not been made.

Increasing GOR has however been considered an increase of hydrotreatment severity, and therefore a common expectation has been that increased GOR would result in in- creased rates of other hydrogenation processes. The experiments in the present docu ment, evidence that a process in which GOR is above 500 Nm ³ /m ³ results in increased HDS without increasing olefin saturation; on the contrary a reduction in olefin saturation is observed. This surprising experimental observation may be implemented in a novel and inventive process, involving operation of a HDS reactor at a combination of moder ately increased GOR, such as above 500 Nm ³ /m ³ , 600 Nm ³ /m ³ or 700 Nm ³ /m ³ and lower pressures than practiced in the prior art. Beneficially an upper limit, such as 1000 Nm ³ /m ³ may be imposed on the GOR, as this will limit equipment size.

The analysis of experimental data has identified that it is possible to reach previously unrealized values of selectivity above 10 and even 20, by reducing the pressure in the process, especially in combination with high GOR. Without being bound by theory it is believed that the effect is due to a combination of low pressure limiting olefin saturation as well as mercaptan recombination and the high GOR limiting catalyst deactivation.

However, increasing the GOR is associated with drawbacks, such as excessive volume of process equipment as well as excessive requirements for e.g. compressor capacity, and therefore it is preferred to operate with a GOR below 1000 Nm ³ /m ³ . In addition, the realization of high selectivities at moderate conversion and thus low se verity opens a previously unrealized opportunity for increasing the window of operation by choosing a low space velocity, such as 1 .1 hr ¹ to 3 hr ¹ .

It has further been identified that an appropriate evaluation of the balance between high HDS conversion and low olefin saturation may beneficially be made by plotting se lectivity (%HDS/%OSAT) vs. (100-%HDS), which in the interval of severity defined by %HDS being 60% to 99%. These experimental data were observed to have a linear correlation with an asymptotic selectivity of 1 at high severity with 100% HDS, and therefore to follow the correlation (%HDS/%OSAT) = ai ^* (100-%HDS)+1 , where a, is a constant corresponding to combination (i) of conditions, such as feedstock, GOR, pres sure and catalyst, but varying severity. This surprising realization lead to the identifica tion of the parameter“selectivity slope” ai=(%HDS-%OSAT)/(%OSAT ^* (100-%HDS)), which due to the linear nature of the correlation may be used to characterize a single set of conditions as highly selective, irrespectively of the chosen severity (within the se- verity range defined by %HDS between 60% and 99%).

According to the present disclosure it is therefore recommended to evaluate process selectivity by the parameter“selectivity slope” and to optimize the selectivity slope by varying GOR and pressure, with increasing GOR and decreasing pressure leading to increased selectivity slope.

It has been surprisingly identified that a previously unrealized potential for a good pro- cess performance, as shown by a steep selectivity slope above 0.55 or 0.7, may be ob tained for difficult feedstocks of which less than e.g. 30% or 50% of the sulfur is in the form of mercaptans. Such a moderate content of mercaptans would typically be found in FCC products, not having undergone significant hydrodesulfurization.

Reducing sulfur content while having low or no reduction of octane number, by a high GOR and a low pressure has the benefit that complex process layouts may be avoided or that it is made possible to obtain very low sulfur levels in combination with satisfac- tory octane numbers, which would otherwise be hard to obtain. It may however also be found beneficial to combine a process with a high GOR and low pressure with the ex- isting process designs, such as an initial hydrogenation of diolefins, a separation of heavy and light naphtha streams, and treatment of one or both of these streams, in one or more steps. Some or all of the process steps involving hydrodesulfurisation may be carried out at increased GOR and low pressure in accordance with the present disclo- sure.

The hydrogenation of diolefins is preferably carried out at moderate conditions. The reason is that the hydrogenation of the first double bond in diolefins is readily carried out at low temperature, and by limiting the temperature the second double bond may be protected. Therefore, the GOR are kept very low, typically below 25 Nm ³ /m ³ , 10 Nm ³ /m ³ or even 5 Nm ³ /m ³ , but also temperature is kept low, e.g. around 100°C-200°C. The GOR must however be sufficient for the desired saturation of diolefins present.

The present disclosure also include combination of the aspects and embodiments listed above.

Figures:

Figure 1 shows a simple process, implementing the present disclosure.

Figure 2 shows an implementation of the present disclosure in a process involving pre treatment and separation.

Figure 3 shows experimental data presented as olefin saturation vs. hydrodesulfuriza tion.

Figure 4 shows experimental data presented as selectivity vs. 100% - hydrodesulfuriza tion, together with linear fits of the experimental data.

102 Hydrocarbon feedstock

104 Stream of hydrogen containing gas

106 Combined feedstock

108 Material catalytically active in hydrodesulfurization

1 10 Desulfurized naphtha stream

202 Di-olefinic hydrocarbon feedstock

204 Hydrogen containing gas

206 Di-olefinic feedstock reaction mixture

208 Material catalytically active in diolefin saturation

210 Intermediate product

212 Separator

214 Light naphtha stream

216 Heavy naphtha stream

218 Hydrogen containing gas

220 Heavy naphtha reaction mixture

222 First material catalytically active in hydrodesulfurization

222 Material catalytically active in hydrodesulfurization

224 Partly desulfurized heavy naphtha

226 Further catalytically active material

228 Desulfurized heavy naphtha 230 Desulfurized naphtha product

Figure 1 shows a process for removing organically bound sulfur from hydrocarbons.

The process involves combining a hydrocarbon feedstock 102 containing organically bound sulfur and olefins with a stream of hydrogen containing gas 104 such that the ra- tio of hydrogen containing gas to feedstock is at least 750 Nm ³ /m ³ . The combined feed- stock 106 is directed to contact a material catalytically active in hydrodesulfurization 108, such as 1 % cobalt and 3% molybdenum, on an alumina support, at a temperature around 250°C. A desulfurized naphtha stream 1 10 is withdrawn from the catalytically active material.

In a further embodiment the catalytically active material may have a different composi tion such as 1 % to 5% cobalt and 3% to 20% molybdenum or tungsten, on a refractory support, which may be alumina, silica, spinel or silica-alumina.

In a further embodiment the hydrogen containing gas may comprise significant amounts of other gases, e.g. more than 25%, 50% or even 75% nitrogen, methane, ethane or mixtures hereof.

Figure 2 shows a process for removing organically bound sulfur from hydrocarbons comprising di-olefins. The process involves combining a di-olefinic hydrocarbon feed stock 202 containing organically bound sulfur, olefins and diolefins with a stream of hy- drogen containing gas 204 such that the ratio of hydrogen containing gas to feedstock is around 5-10 Nm ³ /m ³ providing a di-olefinic feedstock reaction mixture 206. The di- olefinic feedstock reaction mixture 206 is directed to contact a material catalytically ac- tive in diolefin saturation 208, such as 2% nickel or cobalt and 7% molybdenum or tungsten, on an alumina support, at a temperature around 100-200°C, to provide an in- termediate product 210 comprising less than 0.1 % or 0.3% di-olefins. Under such mild conditions, it is considered that the lighter sulfur components of the di-olefinic hydrocar- bon feedstock do not react to release organic sulfur as H2S, but instead they may un dergo recombination reactions with olefins to form heavier sulfides. The intermediate product 210 is directed to a separator 212, from which a light naphtha stream 214 and a heavy naphtha stream 216 are withdrawn. The heavy naphtha stream 216 is com- bined with a stream of hydrogen containing gas 218 such that the ratio of hydrogen containing gas to feedstock in the resulting heavy naphtha reaction mixture 220 is at least 750 Nm ³ /m ³ and directed to contact a first material catalytically active in hy- drodesulfurization 222, such as 1% cobalt and 3% molybdenum, on an alumina sup- port, at a temperature around 250°C, providing a partly desulfurized heavy naphtha 224. The partly desulfurized heavy naphtha 224 may optionally be directed to a further catalytically active material 226 such as 12% nickel on an alumina support, typically op- erating at a temperature higher than the first material catalytically active in hydrodesul- furization 222, such as 300°C to 360°C, providing a desulfurized heavy naphtha 228. The desulfurized heavy naphtha 228 is then combined with the light naphtha stream 214 to provide a desulfurized naphtha product 230. In Figure 1 and Figure 2 the tem- perature control of the reactions are not shown, but since the FIDS reactions are exo thermic, it is typical to add cold hydrogen containing gas or cold recycled product to maintain a low temperature increase. If the GOR is increased the requirement for using product recycle may be reduced, as more quench gas will be available.

In a further embodiment the light naphtha may also be desulfurized by contact with a material catalytically active in hydrotreatment, but typically at less severe conditions than the heavy stream(s).

In a further embodiment the partly desulfurized heavy naphtha may be directed to a separator to provide the heavy sulfurized naphtha fraction contacting the third catalyti cally active material and an intermediate naphtha fraction which may either be treated by contact with a further catalytically active material or be combined into the desulfu- rized naphtha product.

Examples:

Two feedstocks of commercial, heavy catalyst cracked naphtha boiling between 60 and 200°C were directed to hydrodesulfurization in an isothermal downflow pilot plant reac tor. The feedstocks are characterized in Table 1 and Table 2. The hydrodesulfurization conditions in the reactor are further specified below.

The reactor effluent was cooled to ca. -5°C to condense the treated naphtha product, which was separated from a remaining gas phase comprising H2S and unreacted H2, and subsequently stripped using N2 to remove any dissolved H2S from the product. The catalyst used was a hydrodesulfurization catalyst comprising 1 .1 wt % Co and 3.2 wt % Mo on alumina support. The catalyst was a 1/20 inch trilobe size in Example 1 and a 1/10 inch quadlobe size in the remaining examples.

The experimental results are listed in Table 3 to Table 7, and depicted in Figure 3 and Figure 4.

Figure 3 plots %OSAT vs. %HDS. The attractive region of parameters of high %HDS and low %OSAT is proximate to the lower right corner.

Figure 4 plots selectivity (%OSAT/%HDS) vs. 100-%HDS. Here the attractive region of parameters corresponds to the steepest line, especially to the top left. For each set of experiments a linear fit is made, with intercept forced to 1 , corresponding to a selectiv- ity of 1 for maximum severity. For each individual experiment the corresponding value of selectivity slope between the experimental point and the point ((%OSAT/%HDS) , (100-%HDS))=(0,1 ) is calculated as (%HDS-%OSAT)/(%OSAT ^* (100-%HDS)) and in cluded in the tables reporting the experiments.

Example 1 :

In Example 1 Feedstock 1 was treated under a GOR level of 500 Nm ³ /m ³ , with 100% hydrogen treat gas. The severity of hydrodesulfurization was controlled by varying the temperature from 200 to 280 °C and the gas to feedstock ratio (GOR) of 250 to 1400 Nm ³ /m ³ , with an inlet pressure of 20 barg. The liquid hourly space velocity (LHSV) was 2.5 1/hr (v/v/hr). Experimental results are shown in Table 3, and in Figures 3 and 4 us- ing the symbol‘x’.

Example 2:

In Example 2 Feedstock 2 was treated under a GOR level of 1200 Nm ³ /m ³ with 100% hydrogen treat gas with an inlet pressure of 20 barg. The severity of hydrodesulfuriza- tion was controlled by varying the temperature from 220 to 265 °C. The liquid hourly space velocity (LHSV) was 2.5 1/hr (v/v/hr). Experimental results are shown in Table 4, and in Figures 3 and 4 using the closed circle symbol '·'.

Example 3:

In Example 3 Feedstock 2 was treated under a GOR level of 1200 Nm ³ /m ³ with a treat gas mixture of H ₂ and CH ₄ with a total inlet pressure of 20 barg. The severity of hy- drodesulfurization was controlled by varying the H ₂ concentration in the treat gas from 42% to 75%. The temperature was 235 °C. The liquid hourly space velocity (LHSV) was 2.5 1/hr (v/v/hr). Experimental results are shown in Table 5, and in Figures 3 and 4 using the closed triangle symbol‘A’.

Example 4:

In Example 4 Feedstock 2 was treated under a GOR level of 1200 Nm ³ /m ³ with a 100% hydrogen treat gas with an inlet pressure of 8.3 barg. The severity of hydrodesulfuriza- tion was controlled by varying the temperature from 220 to 265 °C. The liquid hourly space velocity (LHSV) was 2.5 1/hr (v/v/hr). Experimental results are shown in Table 6, and in Figures 3 and 4 using the open square symbol‘n’.

Example 5

In Example 5 Feedstock 1 was treated under a GOR level varying from 250 Nm ³ /m ³ to 1200 Nm ³ /m ³ with a 100% hydrogen treat gas with an inlet pressure of 20 barg and a temperature of 230°C. The severity of hydrodesulfurization was not varied further. The liquid hourly space velocity (LHSV) was 2.5 1/hr (v/v/hr). Experimental results are shown in Table 7, and in Figures 3 and 4 using the open circle symbol‘o’.

The results of Example 4 and Example 5 in combination indicate that it will be possible to obtain attractive high selectivity conditions, while avoiding excessive equipment cost, by operating at GOR below 1000 Nm ³ /m ³ and at low pressures.

Analyzing the experimental results by the direct inspection using Figure 3 shows that Example 4 at low pressure and high GOR operate in the more desirable range of high %HDS and low %OSAT, while Example 1 according to the prior art at low GOR and high pressure is the least desirable. Examples 2 and 3 are similar to each other with a position between the other two experiments.

Figure 4 shows a transformed representation of the experimental results, by plotting selectivity (%HDS/%OSAT) vs. %HDS for the experiments where %HDS is above 60. The parameters of the fitted lines are shown in Table 8, for a two-parameter fit of slope and intercept and for a linear fit with intercept forced to 1 . In Figure 4 the lines with in tercept 1 is shown to benefit the comparison of lines.

The low conversion experiments (two experiments of Example 1 ) were omitted as they deviated from the linear trend, with a selectivity slope far below the other experiments of Example 1 . Example 3 indicate that keeping the absolute pressure, while reducing partial pressure has an effect upon selectivity similar to changing severity by changing temperature. A comparison of Example 3 and 4 indicate that for conditions with the same partial pres- sure of hydrogen (42% hydrogen at 20 barg vs. 100% hydrogen at 8.3 barg), selectivity slope is higher when the absolute pressure is lower.

Table 8 shows that for Examples 4 according to the present invention the slope is close to 1 , and much higher than for Examples 1 ,2 and 3. This documents that operation at high GOR and low pressure provides an optimal parameter space in which the desired selectivity for %HDS over %OSAT is possible, and furthermore that this optimal param- eter space is conveniently identified by evaluating the slope of selectivity assuming an asymptotic selectivity of 1 at 100 %HDS. The assumption of an asymptotic selectivity also has the convenience that a measure of the quality of conditions may be estimated from a single experiment and calculated as (%HDS-%OSAT)/(%OSAT ^* (100-%HDS)). From Tables 3 to 7 it is seen that the selectivity slope varies little with severity within similar experiments. The high intercept value for Example 3, is considered to an arte fact due to statistical uncertainty, and as shown in Table 5 the selectivity slope values for the two experiments are consistent, confirming the appropriateness of using the se lectivity slope parameter. It is seen that for moderate GOR, only experiments with low absolute pressure have values above 0.55.

Table 1

Feedstock

Property Method of Analysis

Sulfur ASTM D 4294 250 ppmw

SG 60/60° F ASTM D 4052 0.7605

Olefin ASTM D 6839 35 w%

RON ASTM D 2699 89.8

Boiling point ASTM D 7213 SimDist

IBP 37 °C

5% 62 °C

10% 71 °C

50% 117 °C

95% 173 °C

FBP 201 °C

Table 2

Feedstock

Property Method of Analysis

Sulfur ASTM D 4294 249 ppmw

SG 60/60° F ASTM D 4052 0.7517

Olefin ASTM D 6839 35 w%

RON ASTM D 2699 90.7

Boiling point ASTM D 6729 DHA

IBP -3.4 °C

10 wt% 49 °C

20 wt% 69 °C

50 wt% 115 °C

90 wt% 166 °C

FBP 189 °C

Table 3

Temperature Pressure % Olefin Selectivity

GOR % H2 % HDS

[°C] (barg) Saturation slope

200 20.0 502 100 41 4 0.14

210 20.0 502 100 61 7 0.20

230 20.0 502 100 88 18 0.32

240 20.0 502 100 92 24 0.37

250 20.0 502 100 95 34 0.41

260 20.0 502 100 97 47 0.36

280 20.0 502 100 99 75 0.38

Table 4

Temperature Pressure % Olefin Selectivity

GOR % H2 % HDS

[°C] (barg) Saturation slope

220 20.0 1200 100 75 5 0.52

235 20.0 1204 100 90 12 0.62

235 20.0 1200 100 92 16 0.58

265 20.0 1200 100 98 39 0.61

Table 5

Temperature Pressure % Olefin Selectivity

GOR % H2 % HDS

[°C] (barg) Saturation slope

235 20.0 1200 42 83 7 0.60

235 20.0 1200 63 88 10 0.67

Table 6

Temperature Pressure % Olefin Selectivity

GOR % H2 % HDS

[°C] (barg) Saturation slope

220 8.3 1200 100 69 2 1.00

235 8.3 1200 100 86 6 0.98

245 8.3 1200 100 90 8 1.04

265 8.3 1203 100 95 16 1.00 Table 7

Temperature Pressure % Olefin Selectivity

GOR % H2 % HDS

[°C] (barg) Saturation slope

230 20.0 252 100 82 18 0.19

230 20.0 502 100 88 18 0.32

230 20.0 752 100 91 18 0.47

230 20.0 904 100 92 18 0.51

230 20.0 1104 100 93 18 0.60

230 20.0 1405 100 94 17 0.83

Table 8

Slope Intercept Slope w. intercepts

Example 1 0.320 1.2 0.345

Example 2 0.497 1.7 0.534

Example 3 0.417 4.1 0.622

Example 4 0.991 1.1 0.997