FIELD OF THE INVENTION

The current invention concerns liquefied petroleum gas (LPG) . More in particular it concerns a process for converting a LPG stream that is contaminated with

oxygenates into a purified stream with a reduced

oxygenate content. More in particular it concerns a process for converting an LPG stream that is contaminated with an oxygenate such as dimethylether (DME) into a purified LPG stream with an oxygenate content that is reduced to 5 percent by weight (wt%) or less, preferably 50 ppmw or less, more preferably 10 ppmw or less.

BACKGROUND

Liquefied petroleum gas, also called LPG, GPL, LP Gas, liquid petroleum gas or autogas, is a flammable mixture of hydrocarbon gases, primarily propane (C3) and butane (C4) . Since LPG has such a simple chemical structure, it is among the cleanest of any alternative fuels. It is used as a fuel in heating appliances and vehicles. The boiling point of LPG is less than 25 °C, even less than 1 °C. Typically, the boiling point ranges from —42 °C to 0 °C depending on its mixture percentage of C3 and C4. Standard product specifications require LPG to be delivered as a single-phase pressurized liquid product. Both for propane, butane and mixtures thereof,

specifications are known on the content of other

hydrocarbons and other components. This includes

specifications on the total sulphur content (e.g., 15 ppmw maximum), hydrogen sulphide content (e.g., none) and total oxygenate content (e.g., 5 wt%, preferably 50 ppmw, more preferably 10 ppmw maximum) . It even contains a specification on olefinically unsaturated compounds (0.5 wt% maximum) . This is, for instance, known from the standard specification "Liquefied Petroleum Gas", put on the internet at

http://www.tasweeq.com.qa/EN/Documents/LPG.pdf by the Qatar International Petroleum Marketing Company.

LPG is typically prepared by refining petroleum or "wet" natural gas, and is almost entirely derived from fossil fuel sources. It is typically manufactured during the refining of petroleum (crude oil) , or extracted from petroleum or natural gas streams as they emerge from the ground. Typically, LPG is produced from a hydrocarbon fraction comprising at least 80 mol% C3 and/or C4, the remainder being lighter components such as ethane, and heavier components such as C5's. It may also contain a (small) amount of olefinically unsaturated compounds (ethene, C2=; propene, C3=, butenes C4=, and higher olefins) . This hydrocarbon fraction is referred to as an LPG stream. Typically, this fraction has an upper boiling temperature below 25°C at ambient pressure. In the final step the LPG stream is typically brought on-spec using a de-ethanizer and the like. Cryogenics may also be used, in particular for hydrocarbon feeds based on natural gas.

Oxygenates are used for Enhanced Oil Recovery

(hereafter "EOR") and/or Enhanced Gas Recovery (hereafter "EGR") . For instance, from US Patent Application

Publication 2012/037363, a system for producing oil and/or gas from an underground formation is known, including a well above the formation; a mechanism to inject an enhanced oil recovery formulation into the formation, wherein the enhanced oil recovery formulation includes dimethyl ether (DME) .

A downside of this EOR and/or EGR use is that the produced hydrocarbons are contaminated with oxygenates, DME in particular. When these contaminated fossil fuel sources are used, the LPG stream derived therefrom will contain anywhere from 20 to 80 mol%, more typically from 30 to 70 mol%, still more typically from 40 to 60 mol% DME. The total oxygenate specification is therefore greatly exceeded. This DME-contaminated LPG stream can be purified before the final on-spec production step, e.g., by distillation or by washing. However, this additional distillation or washing step adds costs. This adversely affects the economics of the LPG. Alternatively, the DME- contaminated LPG stream can be used as low quality fuel. Either way, DME, which is a valuable product is lost without any benefits. It is therefore desirable to improve the purity of the DME-contaminated LPG stream without increasing costs while at the same time

efficiently recovering the value of the DME.

SUMMARY OF INVENTION

Accordingly, the current invention provides a process for converting a contaminated LPG stream comprising C3 and/or C4 and further comprising from 80 to 20 mol% of DME, into a purified LPG stream with an oxygenate content of 10 ppmw or less, wherein the DME-contaminated LPG stream is fed to an oxygenate-to-olefin (OTO) process at conditions whereby the DME reacts to form olefins, and wherein the purified LPG stream is obtained.

The invention also relates to a system for producing oil and/or gas from an underground formation, including a well located in the formation; a mechanism to inject an enhanced oil recovery formulation into the formation, and an oxygenate-to-olefin reactor. DETAILED DESCRIPTION

EOR and/or EGR with DME as enhanced recovery

formulation is known from US2012/037363 mentioned

hereinbefore and incorporated herein in its entirety by reference. Thus, an enhanced recovery formulation

comprising DME is injected into an underground formation, and oil and/or gas are recovered from the production wells .

The recovery of oil and/or gas from the underground formation may be accomplished by any known method.

Suitable methods include subsea production, surface production, primary, secondary, or tertiary production. The selection of the method used to recover the oil and/or gas from the underground formation is not

critical .

For instance, the miscible enhanced oil recovery formulation may be injected into a single conduit in a single well, allowing dimethyl ether formulation to soak, and then pumping out at least a portion of the dimethyl ether formulation with gas and/or liquids. Another suitable method is injecting the miscible enhanced oil recovery formulation into a first injection well, and pumping out at least a portion of the miscible enhanced oil recovery formulation with gas and/or liquids through a second production well. The selection of the method used to inject at least a portion of the miscible enhanced oil recovery formulation and/or other liquids and/or gases is not critical.

The enhanced oil recovery formulation may comprise further components. Such further components include one or more of hydrogen sulfide, carbon dioxide, octane, pentane, LPG, C2-C6 aliphatic hydrocarbons, nitrogen, diesel, mineral spirits, naphtha solvent, asphalt solvent, kerosene, acetone, xylene, trichloroethane, and mixtures thereof. In addition, water in gas or liquid form, air, nitrogen, and mixtures thereof may be used.

Typically, the crude oil and or gas so produced is refined and separated in various fractions.

Conventionally, fractional distillation is used, for instance using single or multiple distillation steps or treatments. DME has a boiling point of -24 °C. It typically ends up in the light distillation fraction, which comprises at least 80 mol% C3 and/or C4, and which is used for the preparation of LPG. This fraction is referred to as the LPG fraction.

By fractional distillation and the like, the DME used in the EOR and/or EGR will concentrate in the LPG fraction with a content of from 20 to 80 mol%, typically from 30 to 70 mol%, more typically from 40 to 60 mol%, conveniently about 50 mol%. Thus the oxygenate specification is greatly exceeded and separation of DME on the one hand and the C3/C4 fraction on the other hand is required for use.

In the process of the current invention, the DME- contaminated LPG fraction is fed to an OTO reactor.

Oxygenate-to-olefin (OTO) processes are used for producing light olefins: ethylene (C2=) , propylene (C3=) and mixtures thereof. These light olefins are essential building blocks for the modern petrochemical and chemical industries. The search for alternative materials for light olefin production has led to the use of oxygenates such as alcohols and, more particularly, to the use of methanol (MeOH) , and or ethanol, or their derivatives such as dimethyl ether (DME), for example. Common

oxygenate feedstocks are methanol (MeOH) and/or

dimethylether (DME) , in which case the process is

referred to as a methanol-to-olefin process. In the OTO process, the oxygenate is converted in an oxygenate-to- olefin reactor to ethylene and propylene using a suitable conversion catalyst. Molecular sieves such as microporous crystalline zeolite and non-zeolitic catalysts,

particularly silicoaluminophosphates (SAPO) , are known to promote the conversion of oxygenates to hydrocarbon mixtures, particularly hydrocarbon mixtures composed largely of light olefins. The conversion of alcohols is for instance described in US 3,894,107 . It is generally known that the process can be optimized to produce a major fraction of C2 - C3 olefins. However, also paraffins, aromatics, naphthenes and higher olefins can be formed by a combination of hydrogen transfer, alkylation, cyclisation and

oligomerisation . Prior process proposals have included a separation section to recover ethylene and propylene from the reaction effluent.

As indicated, such processing typically produces or results in a reaction effluent containing a range of olefin reaction products as well as unreacted oxygenates, oxygenate derivatives, and other trace oxygenates.

Typical or common OTO processing schemes include an oxygenate absorber whereby methanol and or water

(preferably at a temperature in the range of 25 to 40°C) is used to absorb any remaining oxygenates, e.g.,

methanol and DME, from the reaction effluent, while cooling down the product reactor effluent. This

oxygenate-containing circulated water is subsequently stripped in an oxygenate stripper to recover methanol and DME, with such recovered materials ultimately recycled to the oxygenate conversion reactor. The oxygenate

conversion product stream resulting from the oxygenate absorber is typically passed to one or more compressors. Moreover, it is treated, for instance between the 4th and 5th stage of compression, in a C02 removal zone wherein the dewatered oxygenate conversion product stream is contacted with caustic solution to remove carbon dioxide.

Preferably the reaction effluent is first

fractionated. It is common to send the reaction effluent first to a quench tower. Here typically most of the water and residual oxygenates may be removed. Subsequently it is common to submit the first effluent stream to one or more compressors or compression stages; to an oxygenate stripper, and/or to a C02 removal zone. Finally, it is common to send the first effluent stream to a drier.

In the process of the current invention, these steps are preferably included, more preferably in the described order. For instance, an oxygenate stripper and/or C02 removal zone may be part of a "compression train". In this case, after C02 removal the product gas is further compressed. Next it is common to fractionate the reaction effluent stream into a first product fraction comprising C2 and C3 hydrocarbons and a second fraction comprising the C4+ olefins. Also a series of distillation columns may be used. For instance, the reaction effluent stream, or the first fraction if a split has already been made, may be sent to a de-ethanizer, typically operating at 22- 26 bar a, wherein C2 hydrocarbons are isolated from the top. The CI- component therein may be compressed and further purified, if needed. The bottom stream after the de-ethanizer may then be sent to a de-propanizer, typically operating at 12-21 bars, wherein C3

hydrocarbons are isolated from the top. The overhead of the de-ethanizer and of the de-propanizer may be sent to a C2 splitter, where the ethylene and the ethane are split, respectively to a C3- splitter, where the

propylene and the propane are split. The C4 saturates (butanes) may be removed from the bottom stream of the de-propanizer. This process may include recycle streams, for instance with respect to the product of the oxygenate stripper and/or part of the bottom stream of the de- propanizer .

The saturated components (primarily C3 and C4) of a DME-contaminated LPG fraction are substantially inert when fed into the OTO reactor. Subsequently, the C3 and C4 saturates may be separated from the C3 and C4 olefins contained in the DME-contaminated LPG fraction, if any, and/or produced in the OTO reactor. The olefinically unsaturated components of the LPG fraction, if any, may be removed from the OTO product together with the formed olefins. This process thus may result in LPG that satisfies the oxygenates specification and the olefins specification at the same time. Accordingly, the DME- contaminated LPG fraction can be a valuable stream of light olefins which may be used as building blocks in modern petrochemical and chemical industries. Remaining olefins in the LPG fraction after treatment in the OTO reactor, if any, may be hydrogenated to meet the olefin specification .

The LPG fraction that is contaminated with DME may be used as single feed in an OTO reaction, or mixed with another oxygenate feed. For instance, it may be used as supplement feed in an existing OTO process that runs on methanol or a methanol/olefin feed. If used as supplement feed, then the ratio is selected such as to achieve the highest efficiency in the OTO reaction.

In this respect, the use of a DME-contaminated LPG fraction in an existing OTO plant brings additional benefits. Thus, in a common OTO process steam is

typically used as reaction diluent. A reaction diluent is typically used in a molar ratio, diluent/oxygenate, of about 1 or greater, wherein ethers such as DME are considered to represent 2 oxygenate molecules. Preferably the diluent/oxygenate ratio is no more than 10,

preferably no more than 5, more preferably no more than 2.5. By introducing an LPG fraction, the consumption of steam may be significantly reduced. With sufficient C3 or C4 saturates in the DME-contaminated LPG fraction, steam may be entirely replaced. In other words, ideally the saturated hydrocarbons in the LPG fraction are used as diluent before any steam is introduced. If additional diluent is required (e.g., when the content of saturated hydrocarbons in the DME-contaminated LPG fraction is low, with a corresponding diluent/oxygenate ratio of less than 1) then steam may be introduced. Thus, steam is

introduced, if any, in an amount to achieve a total diluent/oxygenate ratio of at least 1, but preferably no more than 10.

The OTO process is described in more detail herein below. In this process an oxygenate feed is provided to a first reaction zone. The oxygenate feed in the current invention comprises DME that is part of the contaminated LPG fraction as main or supplementary feed. Reference herein to the oxygenate feed is to a single feed

comprising the DME-contaminated LPG fraction, together with other oxygenates such as MeOH and/or with steam, if any, or to two or more sub feeds each comprising one or more of the compounds of the oxygenate feed, which combined form the oxygenate feed.

In the first reaction zone, the oxygenate feed is contacted with a suitable catalyst. This may, for

instance, be a zeolite-comprising catalyst. Catalysts suitable for converting the oxygenate feed comprise one or more molecular sieves. Such molecular sieve- comprising catalysts typically also include binder materials, matrix material and optionally fillers.

Suitable matrix materials include clays, such as kaolin. Suitable binder materials include silica, alumina, silica-alumina, titania and zirconia, wherein silica is preferred due to its low acidity.

Molecular sieves preferably have a molecular

framework of one, preferably two or more corner-sharing tetrahedral units, more preferably, two or more [Si04],

[A104] and/or [P04] tetrahedral units. These silicon, aluminum and/or phosphorus based molecular sieves and metal containing silicon, aluminum and/or phosphorus based molecular sieves have been described in detail in numerous publications including for example, US

4,567,029. In a preferred embodiment, the molecular sieves have 8-, 10- or 12-ring structures and an average pore size in the range of from about 3 A to 15 A.

Suitable molecular sieves are silicoaluminophosphates

(SAPO), such as SAPO-17, -18, 34, -35, -44, but also SAPO-5, -8, -11, -20, -31, -36, 37, -40, -41, -42, -47 and -56; aluminophosphates (A1PO) and metal substituted

(silico) aluminophosphates (MeAlPO) , wherein the Me in MeAlPO refers to a substituted metal atom, including metal selected from one of Group IA, IIA, IB, IIIB, IVB, VB, VIB, VIIB, VIIIB and Lanthanides of the Periodic Table of Elements. Preferably Me is selected from one of the group consisting of Co, Cr, Cu, Fe, Ga, Ge, Mg, Mn, Ni, Sn, Ti, Zn and Zr.

Alternatively, the conversion of the oxygenate feed may be accomplished by the use of an aluminosilicate- comprising catalyst, in particular a zeolite-comprising catalyst. Suitable catalysts include those containing a zeolite of the ZSM group, in particular of the MFI type, such as ZSM-5, the MTT type, such as ZSM-23, the TON type, such as ZSM-22, the MEL type, such as ZSM-11, and the FER type. Other suitable zeolites are for example zeolites of the STF-type, such as SSZ-35, the SFF type, such as SSZ-44 and the EU-2 type, such as ZSM-48.

Aluminosilicate-comprising catalyst, and in

particular zeolite-comprising catalyst are preferred when an olefinic co-feed is fed to the oxygenate conversion zone together with oxygenate, for increased production of ethylene and propylene.

Preferred catalysts for OTO processes comprise SAPO, MEL and/or MFI type molecular sieves, whereby the latter two are zeolite molecular sieves. More preferred catalyst comprise SAPO-34, ZSM-11 and/or ZSM-5 type molecular sieves. A preferred MFI-type zeolite for the OTO catalyst has a silica-to-alumina ratio, SAR, of at least 60, preferably at least 80. More preferred MFI-type zeolite has a silica-to-alumina ratio, SAR, in the range of 60 to 150, preferably in the range of 80 to 100.

The catalyst may further comprise phosphorus as such or in a compound, i.e. phosphorus other than any

phosphorus included in the framework of the molecular sieve. It is preferred that a MEL or MFI-type zeolite comprising catalyst additionally comprises phosphorus.

The oxygenate feed is contacted with the catalyst at a temperature in the range of from 350 to 1000 °C, preferably of from 450 to 650°C, more preferably of from 530 to 620°C, even more preferably of from 580 to 610°C; and a pressure in the range of from 0.1 kPa (1 mbar) to 5 MPa (50 bar), preferably of from 100 kPa (1 bar) to 1.5 MPa (15 bar), more preferably of from 100 kPa (1 bar) to 300 kPa (3 bar) . Reference herein to pressures is to absolute pressures.

Obviously, it is also possible to use several

reactors in series, wherein the effluent of a first reactor is used as feed in a subsequent reactor.

The current invention also relates to a system for producing oil and/or gas from an underground formation, including a well above the formation; a mechanism to inject an enhanced oil recovery formulation into the formation, and an oxygenate-to-olefin reactor. Preferably, this is a system primarily for enhanced oil recovery, given the beneficial use of DME for EOR. More preferably, the system further includes a refinery wherein produced hydrocarbon from the system is

fractionated. This is of particular interest if the product is refined using fractional distillation to obtain an LPG fraction wherein the DME that was used for EOR is concentrated.

ILLUSTRATIVE EMBODIMENTS

In the process of the current invention, a system for producing oil and/or gas from an underground formation may be used, which system includes a well above the formation; a mechanism to inject an enhanced oil recovery formulation into the formation, and a production well.

A dimethyl ether formulation is injected into the formation. The crude oil so produced is subjected to conventional fractional distillation processes. This results in an LPG fraction containing about 50 mol% DME. EMBODIMENT 1

The LPG fraction containing DME is introduced as such into the OTO reactor. The feed is reacted in a fluidized bed reactor at a weight hourly space velocity (WHSV) of 5-40 h-1 at 600°C in the presence of ZSM-5 as catalyst. The reaction effluent is quenched and condensed. Any residual oxygenate is stripped and recycled. The product is then split into a light olefin fraction or fractions (combination of C2=/C3= or separate ethylene and

propylene fractions), and a paraffin fraction. Depending on the residual olefins level, the product is then subjected to a final hydrogenation process. The paraffin fraction is now an upgraded LPG fraction with an

oxygenate content below the 10 ppmw. It also has a reduced olefins content, thus meeting the specification requirements in this regard too.

EMBODIMENT 2

The LPG fraction containing DME is mixed 1:1 volume ratio with a MeOH feed. In other words, it is integrated in an existing OTO plant. However, no steam is used as would have been the case in a conventional process.

The feed is fed to an OTO reactor where the feed is reacted in the same manner as described in embodiment 1. In this case the amount of olefins produced will be higher, due to the higher content of oxygenates in the feed. The LPG so produced meets the specification

requirements with respect to the oxygenate content and the olefin content. As compared to a conventional OTO process, this embodiment has the additional advantage of a significantly reduced steam consumption.

Those of skill in the art will appreciate that many modifications and variations are possible in terms of the disclosed embodiments of the invention, configurations, materials and methods without departing from their spirit and scope. Indeed, integration with existing production systems and/or OTO reactors has additional capex and opex benefits. Accordingly, the scope of the claims appended hereafter and their functional equivalents should not be limited by particular embodiments described and

illustrated herein, as these are merely exemplary in nature .