NO872183L - SELECTIVE PROCESSING OF GASES CONTAINING OLEFINES BY THE MEHRA PROCESS. - Google Patents

Description

The present invention relates to the recovery of olefins from pyrolysed hydrocarbon gases and particularly relates to the recovery of olefins from split hydrocarbon gases by absorption with a preferred physical solvent.

Olefins have a wide range of petrochemical applications. Ethylene is a fundamental building block within the petrochemical industry. The most important single application is the conversion to polyethylenes of low and high density, which are used in packaging, within communications, construction, the automotive industry, the manufacture of household appliances and within many other industries. Other main uses include oxidation to ethylene oxide and chlorination to ethylene dichloride.

Olefins are generally produced by thermal or catalytic cracking of gaseous or liquid hydrocarbons to produce split gases. Three general processes for separating or concentrating the components of split gases for the recovery of moderate and high purity ethylene have been available for many years. One involves the use of a solid adsorbent such as charcoal or silica gel. The other two use fractional distillation in two variants, the first is a simple low-temperature fractionation process and the second includes absorption in liquid that has a low vapor pressure, thereby avoiding very low temperatures in the fractionation system.

Low temperature fractionation, an improvement on the staged liquefaction process, has been used since the mid-1960s in the majority of ethylene plants in operation, as described in "Low Temperature Fractionation of Light Hydrocarbons" by A.W. Pratt and N.L. Foskett, Transactions of the American Institute of Chemical Engineers, Volume 42, 1946, page 149.

In a typical modern olefin plant where low-temperature fractionation is used for ethylene extraction as illustrated in Figure 1, the dominant feedstock consists of ethane, propane, butanes, naphtha, gas oil, condensate, and other hydrocarbons derived from crude oil cracking. These raw materials are pretreated and split using conventional steam splitting devices. The split gases leave the split furnaces at 815 to 1200°C. These gases are quenched and cooled down to approx. 27 to 49" C at pressures lower than 210 kpa. The raw materials above can also be split catalytically under suitable operating conditions.

Depending on the feedstock and the degree and selectivity of the fission, the fissured gases include hydrogen, methane, carbon monoxide, carbon dioxide, acetylene, ethylene, ethane, methylacetylene, propadiene, propylene, propane, butadienes, butenes, butanes, C^ compounds, C^-Cg non-aromatic compounds, benzene, toluene, xylene, ethylbenzene, styrene, Cg to 200"C gasoline, 200+°C fuel oil and water.

These gases are compressed in multi-stage compression units to pressures of the order of 2800 to 4g00 kpa. During compression, some heavier hydrocarbons and water are separated. Depending on the composition, the separated hydrocarbons are stabilized and can be used as part of the raw material for the fractionation part. The uncondensed gases are further dried after compression using molecular sieves or activated aluminum oxide to a water dew point of less than -100°C. The cracked gases can be dried at an intermediate pressure level corresponding to the intermediate pressure of the multi-stage cracked gas compressor.

The dry gases are then cooled in a series of steps to cryogenic temperatures of -129° C using a complicated series of equipment where extensive heat exchange and cooling systems for ethylene and propylene are used. The purpose of the cooling is to separate ethylene and heavier hydrocarbons from the methane and hydrogen present in the split gases. The remaining flow of methane and hydrogen is further separated into hydrogen-rich and methane-rich streams inside the cryogenic cooling plant. Part of the separated hydrogen stream is preferably further purified using conventional pressure swing adsorption techniques before it is used for hydrogenation of the acetylene to desired products, while the methane-rich stream is used as fuel gas for the steam cracking furnaces.

The speared liquid streams, containing ethylene and heavier hydrocarbons, are further fractionated in a low-to-high-pressure demethanizer (1100 to 3200 kpa). The low-thickness demethanizer uses ethylene refrigerant for the overhead condenser, while the high-pressure demethanizer uses low-level propylene refrigerant, sometimes in conjunction with an expansion unit, for such condensation.

The specification for methane in the bottom product from the demethanisation device is very strict, as methane represents a contaminant when it is present in the ethylene product. Correspondingly, it is highly desirable to recover most of the ethylene present in the raw materials of the demethanization device in order to reduce losses and thereby minimize the need for recycling top product from the demethanization device to the split gas compressors for further recovery of this desired product.

Olefin plants are known to be energy intensive. For a plant with an ethylene capacity of 640 tonnes per years using modern available technology for 50/50 condensate and ethane/propane feedstock, the input energy needs of the gas compressor can vary between 37,000 and 40,000 horsepower (PS), while the corresponding needs for the propylene cooling system can vary from 18,000 to 22,000 PS, and the needs of the ethylene cooling system can vary from 4000 to 8000 PS. Consequently, depending on the technology used and the desired product state, the total compression energy can vary between 60,000 and 70,000 PS, this requires large capital investments and represents a significant operating cost blinded by energy consumption. In summary, depending on the desired product, available raw material, degree and selectivity of cleavage and separation technology used, the specific energy consumption for an olefin plant can vary between 3100 and 7500 kcal/kg ethylene product. This represents approx. 25 .456 to 62.4$ of the total heating value of ethylene of 12017 kcal/kg.

It is important to note that the cracking system for olefin production appears to have been improved and refined to a high efficiency operating condition. Although the steam cracking device requires significant amounts of fuel, as the cracking process takes place at very high temperatures, most of the energy used in the cracking furnaces is recovered through the extensive use of waste heat recovery equipment. Of the energy released by thermal decomposition of the hot gases, e.g. only approx. 3-556, and the tpa takes place through the chimney. Most of the energy thus recovered is used for compression of the split gases and provides cooling for the cooling part. Therefore, the predominant energy consumption in an olefin plant is associated with the cooling and the fractionation section to separate different components of the split gas stream.

According to a recent article, there are twenty-three companies operating thirty-two olefin plants in the United States. This article is entitled, "New Capacity Forces Ethylene Producers to Aim for Lower Costs, Flexibility", by Ted Wett, and was printed in the September 12, 1985 issue of the Oil and Gas Journal, page 39. The equivalent total installed production capacity of these thirty-two plants or the factories were approx. 17 million tonnes of ethylene in June 1985. Worldwide, the total ethylene plant capacity was approx. 49 million tonnes per year in 37 countries. The raw materials used in the facilities or factories were at least one of ethane, propane, butane, refinery streams, naphtha, gas oil, natural gas liquids (NGL), liquefied petroleum gas (LPG), etc.

As an example of the most modern ethylene plants, Europe's newest ethylene crackers started production around October 1985 at a plant at Mossmoran, Scotland, with six crackers running on ethane as feedstock as part of a $1.6 trillion joint venture. It is the first cracking plant known to use a gas turbine to drive the process gas compressor, and the first to use the turbine exhaust gas as preheated combustion air for the furnaces. As the second part of the medal in terms of ethylene capacity, cracking plants with a capacity of approx. 600,000 tonnes per year in England and Germany as a result of the competitive effect.

Problems of imbalance between capacity and demand have plagued the olefin production industry for many years. Nevertheless, new facilities of ever-increasing size and complexity incorporating the latest technologies have been built. In most cases, however, plants are built because of cheap sources of raw materials or because the incorporation of the latest technologies may enable a plant to compete favorably with older plants and capture part of the existing market and/or its expected expansion. As the newer facilities come into operation, the older facilities may have to be closed, reduce their capacity, or reduce production costs.

Among the changes that can be made to effect cost reduction are changes to the splitting facilities to provide flexibility for processing a wide variety of raw materials depending on market conditions. Another cost-reducing step is to revise existing facilities to optimize raw material and energy requirements, possibly including cogeneration systems with different units. A third step is to make full use of computer control 1 to maximize plant operating efficiency. A fourth step is to replace outdated facilities within a factory with more efficient facilities that use improved technology.

The next step for improving the efficiency of 01 ef inproduction is believed to be the replacement of low-temperature facilities for the extraction and separation of olefins from split gases. Although the industry has quietly continued to use low-temperature fractionation for many years after discarding solvent absorption systems such as the Kniel process, it pays a relatively high price for this because low-temperature fractionation requires extreme dehydration, and this is achieved by applying energy -intensive ethylene and propylene cooling systems.

Around approx. 1948 was, according to page 651 of "Petroleum Refinery Engineering" by W.L. Nelson, Third Edition, McGraw-Hill Book Company, Inc., New York, N.Y., 1949, 830 pages, ethylene is fundamentally important in the manufacture of many chemicals, such as alcohol, ethyl ether, styrene, ethylene glycol, and tetraethyl lead. A total daily plant capacity of 1,500 tonnes per day was built or was in operation in 1948, this does not include plants that produced ethylene for hydration to alcohol. The plants were mainly of the thermal decomposition type, operating mainly on ethane and propane, where the split gases were fed to an absorption split column which produced fuel gas as the top product and a rich solvent which was fed to a deethanisation column as the first component of a fractional coupling portion, as shown in Figure 1 of U.S. Pat. patent no. 2,573,341 which is figure 2 in the drawings of the present invention, as representative of previously known absorption technology.

The absorption process is discussed for an absorption-type recovery plant in an article in Chemical Engineering Progress, by Ludwig Kniel and W.H. Slager, Vol. 43, No. 7, pages 335-342, July 1947, which uses the same process in Figure 2 shown on page 652 of the book, "Petroleum Refinery Engineering", and in Figure 1 of U.S. Patent No. 2,573,341.

This ethylene plant of Monsanto Chemical Co. at Texas City, Texas, was mainly used for the production of ethylene and was primarily fueled by ethane and propane. Typical final ethylene yields were 75 wt-56 from ethane, 48 wt-56 from propane, or 25-32 wt-56 from gas oil. In case of a conversion per passage of approx. 4556 when splitting propane, the yields were approx. 16.7 wt-56 ethylene and 15.8 wt-56 propylene, in one review. The ethylene was separated together with the ethane and heavier components by means of a low-temperature absorption, an aromatic distillate is produced in the process containing more than 50 wt-56 benzene and toluene and significant amounts of naphthenes, among which cyclopentane and cyclohexane have been identified.

Typical analyzes of three ethylene-containing streams were reported in this paper: (a) a typical coke oven gas, (b) a refinery off-gas, and (c) the effluent from a pyrolysis unit fed with propane and operated to produce a maximum amount of ethylene. For these three raw materials, the ethylene concentration was 4-27 mol-56, diluents lighter than ethylene amounted to 94-17 mol-56, thus covering most of the commercial gases from which ethylene or ethylene + propylene can be extracted economically.

In an article in Petroleum Refiner by Ludwig Kniel, vol. 27, no. 11, November 1948, the design of a fractionating absorption device, which is essentially the same apparatus as the absorption splitter discussed in the previous article, is described for the separation of methane from ethylene in a pyrolysis gas obtained from the splitting of propane. It had been found in plant operation that the effect of such fractionating absorption devices exceeded the requirements that were expected in the design, both in terms of recovery and purity. The reason for this was thought to be either due to the type of lean oil used which contained significant proportions of aromatic compounds, particularly benzene, or to be the result of extra recirculation that took place between the plates where the intercoolers were located.

An article in Hydrocarbon Processing & Petroleum Refiner, by Joe. J. Weatherby, April 1962, Vol. 41, No. 4, described a low temperature absorption plant which recovered 40% propane from 295,000 Nm<3>/day of wet natural gas. This installation used a fractionator-absorber having an evaporator and no intercoolers. This column not only absorbed hydrocarbons from the gas stream, but fractionated the methane and ethane so that a de-ethanized product was obtained. A portion of the Cy+ fraction of the product was pumped over the top of this column as an absorption sink. Ethylene glycol was injected into the inlet gas stream upstream of the heat exchanger and inlet gas cooler where it was cooled to -7°C with ammonia as refrigerant. A gas exchanger and inlet gas cooler removed about half of the heat from the absorption outside of the absorption device, but they correspondingly reduced the average effective absorption device temperature, and the oil content of the tail gas leaving the top of the column was in equilibrium with the oil at the minimum temperature of

-7°C, thereby minimizing oil loss.

Numerous processes are known in the solvent absorption technique for olefin recovery from cracked gases, refinery gases and synthetic gases containing these unsaturated compounds. Some processes use an aromatic absorption oil as a solvent inside an absorption-splitting column that has an evaporator. Such processes are described in one or more of U.S. Pat. patentene nr. 2,187,631, 2,282,549, 2,308,856, 2,325,379, 2,357,028, 2,433,286, 2,455,803, 2,570,066, 2,573,341, 2,588,323, 2,708,580, 2,849,371, 3,055,183, 3,082,271, 3,349,145, 3,686,344, 4,072,604 og 4,479,812.

U.S. Patent No. 2,573,341, Ludwig Kniel, which was issued based on the U.S. application no. 717,264, submitted on 7 December 1946, relates to a process for the extraction of olefinic hydrocarbons and especially ethylene of high purity from coke oven gas, refinery off-gas and pyrolysis gas, which have an ethylene content of respectively 4.0, 5.0 and 27.0 mol-#, which are the raw materials of a rectifier-absorption device, also known as an absorption-splitter. The column or tower has an evaporator at the bottom and two intercoolers to remove the heat of extraction. The top product is fuel gas, and the bottom product is fed to a de-ethanisation column, which has an evaporator at the bottom and a condenser for the top product from which part of the condensate is returned to the de-ethanisation column as return flow. The lean oil volume is not more than 4.2 kg per kg raw material, an amount which ensures retention of 99 mol-Sé of the ethylene entering the rectification-absorption tower.

Absorption, exemplified by the process in Figure 1 of the U.S. patent no. 573,341, was stated on page 651 of Nelson's book to be the process mainly used for ethylene recovery around 1948. In the mid-1950s, a number of absorption plants were in operation that used refrigerated belly oils such as propane, butane, or light aromatic fractions.

Around the mid-1960s, according to page 85 of "Manufacturing Ethylene", by S.B. Zdonik, E.D. Green and L.P. Hallee, Petroleum Publishing Co., 1970, a few absorption plants in operation, but the number was decreasing, while all the newly built plants used some form of low-temperature fractionation to separate ethylene from light hydrocarbons. For plant capacities of approx. 45,000 tonnes per year produced ethylene, refrigerated absorption extraction plants were economically comparable to low-temperature fractionation extraction plants.

As the size of ethylene plants increased, the large heating and cooling loads on the solvent in absorption-type plants caused the plants to become more uneconomical in terms of operating and facility costs, even though the low-temperature fractionation plant used a more complicated cooling system due to the extremely low temperatures used.

It therefore seems as if a classic competitive battle was fought during the twenty years from the end of the Second World War to approx. 1965, where the winner was 1 of the temperature fractionation plant and the loser was the solvent absorption plant. An overview of the reasons for this decision on the market is in order.

Referring to Figure 1 of the U.S. patent No. 2,573,341 as the dominant absorption process according to Nelson, there appear to be at least seven reasons why plants using the Kniel absorption process, for the recovery and separation of olefins from thermally cracked hydrocarbon gases, did not win the competition. Firstly, it should be noted that this Kniel process could not be applied to the separation of methane from hydrogen, whereas such a separation was required for the hydrogenation of acetylene to ethylene. Hydrogen was therefore used as fuel, which is an economically disadvantageous process.

Secondly, the process seems to have been aimed at continuous production of aromatic fractions and retention thereof in the aromatic system, to the extent necessary to maintain a supply of absorbent oil. This situation indicates that the losses of lean absorption oil were significant and that such losses went to fuel use. The cracking furnaces appear to have been designed for the production of the aromatic absorbent oil rather than for producing ethylene.

Third, it is stated in the U.S. patent no. 2,573,341 that it is also essential that methane is eliminated by means of the absorption-splitting column, without loss of ethylene in the overhead product stream, and that larger quantities of methane than expected are not allowed in the bottom product materials from the absorption-splitting column because such larger amounts could destroy the operation of the de-aeration unit and the ethylene fractionator by causing an inability to condense the return stream in these towers at the temperatures necessary to achieve the desired concentrations in the overhead fractions from each of these the towers. According to an Amendment filed on January 8, 1951 during the consideration of the U.S. application no. 717,264, "It was found that when designing such a plant there was a tendency to either lose ethylene top product in the fuel gas pipe or to accumulate methane in the bottom products from the absorption facility, which contaminated the ethylene product". The solution that appears to be chosen and stated in the requirements of the U.S. patent no. 2,573,341, as shown in Figure 2 of the patent, was to add the demethanization column, but the demethanization column is redundant if the absorption-split column is properly designed.

Fourthly, approx. 9656 of the methane rejected, this indicates that a content of 45é methane was present in the ethylene product, a figure which is confirmed by the ethylene purity of 97$. However, the demands for purity have steadily increased, and it may have been an increasing economic burden to meet these demands that the Kniel-type absorption process was unable to meet.

Fifth, it appears that the absorption-splitter column was installed mainly to reduce cracking requirements, by preliminary removal of hydrogen and methane in the raffiery off-gas feedstocks, rather than for recovery and separation of cracking gases. Furthermore, the lean oil fraction used in the absorption-splitting column consisted of 54.0% benzene and toluene and 17.6% pentanes and lighter hydrocarbons, which is considerably lighter than the usual lean oil iron.

Sixth, the description also indicates that the process was limited to ethylene as a product, and was not flexible enough to provide additional desirable by-products, such as propylene, which was sent to the heaters for cracking, and acetylene, which was totally ignored in the process and also recycled . Furthermore, other unreacted paraffins and olefin fractions heavier than ethylene were recycled to the spar furnaces to make the aromatic distillate used as absorbent oil, although a market existed for the propylene and undoubtedly for the butadiene at this time. This constituted another uneconomic feature of the process.

As the seventh reason, the degrees of freedom for operating the Kniel process were limited by the use of only temperatures and lean oil flow rates. However, pressure was also available as a control factor. That its application was neglected may have led to some of the problems with the process.

Three Kniel process absorption plants were built by the Lummus Company on February 1, 1950, according to an affidavit of Ludwig Kniel filed in pending application No. 717,264. If one were confronted with the high losses of absorbent oil in one of these absorption plants, one could have chosen any of several improvement methods.

One such method would be to use a higher molecular weight oil, such as the heavy fractions produced in the oil-gas separators 1 figure 1 of the U.S. Patent No. 2,573,341 or the heavy fractions rejected by the redistillation tower, but it was known that fewer absorber molecules would then be available for absorption, resulting in a lower loading capacity. Lower load capacities mean a larger solvent supply, larger pumps, more backflow, etc. This approach was apparently rejected.

Another method would be to regulate the loss situation when producing absorbent oil in the plant. It is e.g. known that low partial pressure and high vapor ratio maximize olefin production and that high partial pressure and low vapor ratio plus long residence time maximize production of aromatic compounds. Accordingly, it would have been a relatively simple matter to make adjustments in the operation of the cracking furnaces to produce a desired amount of aromatic compounds to replace the losses of absorbent oil. This procedure was apparently chosen. However, the process would then be locked into the production of aromatic compounds with concomitantly decreasing ethylene production and an apparent inability to compete with low-temperature fractionation processes.

Consequently, the Kniel absorption process appears to have been plagued with the following problems: A. uneconomical burning of hydrogen, propylene and acetylene; B. extensive loss of aromatic absorbent oil and replacement with distillate produced in the process, thereby reducing ethylene production; C. an apparent need for an additional column (the demethanizer) for the separation of methane from ethylene to be able to produce ethylene of sufficiently high purity; and D. the inability to financially meet the growing after- the question of ethylene purity and recovery. The original patents and applications relating to the Mehra process have used preferred physical solvents for the extraction of hydrocarbon gas liquids from natural gas reservoirs in two versions: extractive pressure relief and extractive splitting.

The extractive depressurization embodiment of the Mehra process involves extracting natural gas streams with a preferred physical solvent, depressurizing the rich solvent, and compressing, cooling, and condensing the desired C 2+ hydrocarbons, as described in U.S. Pat. patents 4,421,535, 4,511,381, 4,526,594 and 4,578,094 and in the U.S. applications no. 758,351 and 759,327. The condensed hydrocarbons are then selectively demethanized to preserve selected C2+, C3+, or C4+ hydrocarbons, and the removed Cj_, C1+ C2, or C1+C2+C3 hydrocarbons are recycled to the extraction step. The extractive pressure relief embodiment is described on pages 7 and 8 of the October 14, 1985 issue of Gas Processors Report, P.O.Box 33002, Tulsa, Ok 74153, USA.

The extractive splitting operation of the Mehra process, as pitched in the U.S. applications Nos. 784,566, 808,463, 828,988 and 828,996, involve contacting a raw gas stream with a preferred physical solvent in an extractive splitting column including an upper extraction section and a lower splitting section. The gas enters the column between the extraction section and moves upwards where it contacts lean, preferably physical, solvent which, after entering the extraction section at the top of the column, flows downward and countercurrent to the upward gas stream. The contact takes place over mass transfer surfaces, such as column packing or distillation plates. The solvent leaving the bottom of the extraction part is rich in Ci and heavier hydrocarbons.

This C^+ -rich solvent enters the split portion of the column and flows downward, where it contacts the upwardly split vapors from the evaporator at the bottom of the column. The split vapors mainly consist of unwanted components, such as methane, if the desired purpose is to extract ethane and heavier hydrocarbons, or methane and ethane, if the desired purpose is to extract propane and heavier hydrocarbons, etc., depending on the desired extraction purposes.

Returning to opportunities for cost reductions for olefin plants, it seems that a likely place to achieve this is the cryogenic cooling section in which hydrogen, methane and ethylene fractions are separated. A system that can effect such separation using a less energy-intensive method can be useful in many existing olefin plants.

Specifically, the demethanization column is a particularly high intensive energy consumer in a low temperature fractionation plant because it requires extremely low temperatures. An improved process that can eliminate the demethanization column is therefore required.

A third area that invites improvement is the production of high purity ethylene, at ethylene recovery levels better than 98% and at ethylene purities higher than 99.9%, in a more economical way by reducing energy consumption at a reasonable cost.

It is consequently an object of the present invention to provide a process for the separation and extraction of hydrogen to a selected degree from a gaseous stream containing olefins.

It is further an object to provide a method for separating and extracting methane to a selected degree from a gaseous stream containing olefins.

It is further an object to provide a method for separating olefin-containing gases into hydrogen, methane and Cg=+ hydrocarbon products which meet desired specifications.

It is another object to provide a process for the separation and recovery of hydrogen, methane and C2=+ hydrocarbon gases from olefinic gas streams while causing low solvent losses and reduced energy consumption compared to previously known processes.

It is also an object of the present invention to provide an extraction plant for the separation and extraction of hydrogen, methane and Cg=+ hydrocarbon gases from thermally split hydrocarbon gases and to thereby be able to retrofit an existing olefin production plant.

It is another object to provide such an extraction plant and a method for using this as part of a new olefin production plant.

According to these objects and principles of the present invention, it has surprisingly been discovered that a hydrogen-rich stream, a methane-rich stream and a Cg=+ hydrocarbon stream, which are extremely pure with respect to methane and from which extremely pure ethylene can later be recovered, can be separated and recovered from a stream of compressed, mecaptan-free, dry hydrocarbon gases, which include a significant proportion of unsaturated hydrocarbons, by passing the hydrocarbon gas stream through two extraction columns in series, countercurrent to streams of lean preferred physical solvent, and by adding at least one solvent-rich bottom fraction stream to at least one distillation column which regenerates the rich solvent to form lean solvent and produces at least one of the three products At least one extraction column is an extractive splitting column.

Each extractive splitting column contains a splitting part, which has a device for heat supply, and an extraction part above this. Either one or both of the extractive splitting columns can additionally be equipped with a rectification part which is placed above the extraction part and which includes a partial condenser which receives the top fraction flow from the top of the column, an accumulator, and a return pipe to the top of the column. Each extractive splitting column is part of a solvent loop. The same solvent can be used in both loops.

The two extraction columns and the one or two distillation columns can be arranged according to any of three designs. In two of the three embodiments, the three products are recovered as, or from, top fraction streams from three of the four columns. In all three of these embodiments, the stream of compressed, mercaptan-free, dried hydrocarbon gas is fed to the middle section of the first extraction column, which is the ethylene extraction column (EEC).

The arrangements of these designs, the most important contents of the top fractionation and bottom fraction streams, the column designations and the solvent stream and one or two loops are reproduced in Table I. It appears that the arrangements of the extraction columns in these designs can be briefly described as EEC /MEC, EEC/MPC and EEC/MEC/FV. In the EEC/MEC design, the top fraction ES stream is partially recycled and partially fed to the MEC column, which is the methane extraction column (MEG), and the bottom fraction stream is fed to the first distillation column, which is the crude product column (RPC). It regenerates lean solvent for the first solvent loop and produces the product stream of Cg=+ hydrocarbons. The second extraction column produces the hydrogen-rich gas product stream and its overhead fraction and a solvent-rich bottom fraction stream which is fed to the second distillation column, designated the methane product column (MPC). It produces the methane-rich gas product stream and its top fraction and regenerated solvent for the second solvent loop as the bottom fraction stream.

In the IEEC/MPC embodiment, the ethylene extraction column (EEC) extracts most of the methane, as well as the ethylene and other C2+ hydrocarbons (both saturated and unsaturated) in the free solvent stream, producing the hydrogen-rich gas product stream as its overhead fraction. The rich solvent stream is fed to the first distillation column (SRC) which regenerates lean solvent for the first solvent loop and produces C]_+ hydrocarbons as overhead fraction stream for feeding to the middle part of the second extraction column, which is the methane product column (MPC). This column produces the methane-rich gas product stream as an overhead fraction and a rich solvent stream containing C 2+ hydrocarbons. This rich solvent stream is fed to the second distillation column, which is the crude product column (RPC), it produces the third product stream of C2=+ hydrocarbons and regenerates the lean solvent for the second solvent loop.

In the EEC/MEC/FV design, the ethylene extraction column (EEC), which is an extractive splitting column, extracts the ethylene and heavier hydrocarbons in its rich solvent well fraction stream which is fed to the distillation or crude product column (RPC) where lean solvent is regenerated and C2=+ hydrocarbon product is produced as peak fractional flow. The lean solvent stream is fed to the top of the second extraction column (MEC), which generally has no splitter, but may use one if methane and hydrogen specifications require it, while the top fraction EEC stream is fed to the bottom. The second extraction column (MEC) can operate at a higher pressure than the EEC, but the operating pressure of the MEC must be less than the critical pressure of the gas-liquid mixture in the MEC. The MEC top product stream is hydrogen-rich gas product. The bottom fraction stream of rich solvent is depressurized, producing a methane-rich gas product stream and lean solvent which is fed to the top of the ethylene extraction column.

The EEC/MEC design produces one product from the extraction columns and two products from the distillation columns. The EEC/MPC seeding produces two products from the extraction columns and one product from the distillation columns. The EEC/MEC/F V version produces one product from the extraction columns, one product from the pressure relief vessel, and the third product from the single distillation column.

The cracked gases do not need to be compressed to a typical pressure of 3600 kpa, but instead need only be compressed to 2200-3200 kpa to extract ethylene and heavier hydrocarbons from the typical cracked gases or refinery gases.

The method according to the present invention uses the Mehra process technology concept for separating the components of a compressed, mecaptan-free and dehydrated cracked gas stream from an olefin plant by effectively using a preferred physical solvent for effective washing out of the gas stream and preferably removing selected hydrocarbons from this. Additional solvent information is provided in U.S. Pat. application no. 808,463.

For the purposes of the present invention and to illustrate the improvements compared to prior art, the process begins at 1 mark the intake of the split gas compressor which represents the first connection point, the production of ethylene, i.e. the splitting process, is outside the scope of the present invention. The ethylene-plus component stream entering the de-ethanization column, as the main product stream according to the invention, represents the second connection point, as downstream fractionation into individual products is also outside the scope of the present invention. Other products of the method according to the invention are a hydrogen-rich gas stream and a methane-rich gas stream. These three products, i.e. ethylene-plus, methane-rich and hydrogen-rich stream, are then treated in the same way as in prior art.

Accordingly, the method according to the present invention relates to the separation of ethylene and heavier hydrocarbons from methane and hydrogen gases which are present in the split gases from an olefin plant, refinery off-gas streams, coke oven gas streams or synthesis gas streams.

As the Mehra process concept can be applied to any gas stream containing olefins derived from any raw material and as it relates to separation processes that are currently energy intensive, the Mehra process is potentially suitable for retrofitting any ethylene plant in the world over, which today has an equivalent installed ethylene capacity of 49 million tonnes per year.

In the olefin separation schemes of Figures 1 and 3, the three steps common and identical to a typical olefin plant and the process of the present invention are cracking, waste heat recovery and deethanization, plus any subsequent steps in downstream equipment, such as ethylene/ethane fractionators, depropanizers, propylene/ propane fraction ionizers, debutanization devices, etc.

The consumption of energy for feed gas compression is reduced by this replacement of a low temperature fractionation plant with a Mehra process system, because the Mehra process extraction process does not require high pressures for the separation of the components. Similarly, acid gas treatment and dehydration can take place at lower pressures. Dehydration needs are also significantly reduced because extreme cryogenic temperatures are never used in the Mehra process for separation of a desired ethylene and heavier hydrocarbons from the split gas streams. Accordingly, simple glycol dehydration may be sufficient for use in the Mehra process concept.

As the Mehra process extraction unit does not require extreme cryogenic temperatures, there is no need for low levels of cooling, as provided by the ethylene and propylene cooling systems in a conventional low temperature fractionation plant. The extraction unit is also able to separate hydrogen from methane, thereby producing two separate gas streams as products. The C2=+ product from the Mehra process unit meets the specifications for the lighter components, ie methane and hydrogen, of the ethylene product and is therefore suitable for downstream processing in a conventional deethanizer.

The split gases leaving the waste heat recovery unit are compressed in a multi-stage processor to a desired pressure in the range from 1100 to 3200 kpa. The compressed gases can optionally be cooled to a temperature of no less than -29°C before extraction. The split gases enter an ethylene extraction column (EEC), which can only consist of an extraction section and a split section with side and bottom evaporators. This column can additionally use a top-fraction partial condenser for the generation of return flow to a rectification section to achieve a very high recovery, of the order of 98 to 99%, of the ethylene present in the cracked gases when the feedstock and the degree of cracking result in large amounts of methane in the split gases, thereby minimizing the loss of valuable ethylene with the overhead fraction stream.

From there, an EEC can consist of a rectification part with an upper condenser at the very top, a middle part for the extraction of ethylene and heavier hydrocarbons with a preferred physical solvent, and a lower splitting part with side and bottom evaporators. ~ However, a simple vapor extraction column, consisting of an extraction section at the top and a split section with suitable side and bottom evaporators, may be equally suitable for desired recoveries, naturally depending on the economic optimization of the capital and energy consumption parameters.

The purpose of the extraction section is to recover ethylene and heavier hydrocarbons from the cracked gas stream which enters at the bottom of the extraction section and flows upwards through a transfer medium, such as packing or fractionation trays or alternatively by using "HIGEE" trays, while the contacts the downflowing preferred physical solvent fed to the EEC at a solvent flow rate selectively controlled in the range of 0.1 to 70 m<3> of solvent per 1000 m<3>gas stream, and depending on the composition and flow rate of the gas stream. "HIGEE" is a trademark of Imperial Chemical Industries. During the extraction process, some of the contained methane in the split gases will also be recovered at the same time with the help of the physical solvent.

The solvent stream leaving the bottom of the extraction section of the EEC is split for unwanted methane by effectively using additional mass transfer medium in the split section of the column. The split vapors are preferably generated by heating the rich solvent stream in side and/or bottom evaporators, but they can be from an external source of compatible gas stream. The source of heat energy can be an external source or waste heat recovered by the lean solvent loop. The solvent-containing ethylene and heavier hydrocarbons leaving the bottom of the split portion of the EEC do not contain more than allowable amounts of undesirable components, such as methane, according to the specifications of the ethylene product.

In the process of splitting unwanted components from the rich solvent and extracting the split vapors in conjunction with the split gas stream in the extraction section, the vapors leaving the top of the extraction section carry with them some of the desired components, such as ethyne, whose recovery is desired to be of the order of 98 to 99%. Depending on the composition of the methane and hydrogen components of the cracked gas stream and the relative economics of recovering ethylene below desired levels by varying the solvent flow rate to the extraction section of the EEC, a rectification section may be appropriate to provide to effectively perform the desired rectification, resulting in improved recovery of the ethylene component from the cracked gas stream.

Accordingly, the top fraction vapors are partially condensed, so that adequate amounts of reflux are generated for the rectification section. Furthermore, the rectification part also recovers any physical solvent that may have been carried away with the methane and hydrogen streams.

The solvent containing only the desired components of ethylene and heavier hydrocarbons can be heated further if this is economically desirable before processing in the crude product column (RPC). In this column, the hydrocarbons extracted from the cracked gas stream are separated from the physical solvent. The rectification section is operated so that solvent losses with the hydrocarbon product from the top of the column are minimized. The operating conditions at the top of the column are preferably such that the top fraction can be condensed using available condensation media, such as air or cooling water. At operating pressures at the bottom of the crude product column, the temperature of the bottom fractions is generally lower than the equivalent boiling temperature for pure component solvent. In general, a small amount of heavier hydrocarbons (less than 2 mol%) may remain in the lean solvent to keep the size of the RPC within economically advantageous limits.

The top product from RPC is further processed in the de-ethanizer and downstream equipment by conventional steps. The split solvent from the bottom of the RPC is recycled to the top of the extraction section of the EEC after cooling the lean solvent to the desired temperature via a heat recovery loop and lean solvent cooler, similar to, but not limited to, the arrangement shown in Figure 4.

Any liquids formed and separated at the intermediate pressure levels of the cracked gas compressor can be stabilized for further processing in the fractionation section of a typical olefin plant. However, depending on the composition and quantity of the separated stream, it may be beneficial to stabilize through EEC. This flow therefore enters the EEC at a suitable place in the splitting section, preferably between the splitting and extraction sections.

As a Mehra process plant does not require high pressures for extraction of the ethylene plus components it may be economically advantageous to provide a secondary but parallel extraction column operated on a common solvent loop with the first column and operated at an intermediate pressure level in accordance with the desired intermediate pressure for the cracked gas compressor. The obvious advantage of this would be to selectively extract C3+ or C4+ hydrocarbons from the cracked gas to further reduce the compression requirements of the cracked gas compressor. It would be preferable, but not necessary, to use an operating pressure level for the secondary parallel extraction column that is somewhat higher than the operating pressure of the crude product column (RPC). The rich solvent stream from the bottom of the parallel extraction column is such that it meets the requirements regarding undesirable components, such as methane, applicable to the EEC, so that it can be combined with the rich solvent stream from the EEC before processing in the RPC .

As an alternative, in some cases it may be more economically advantageous to recycle a small vapor stream from the top of the accumulator for RPC to the appropriate intermediate stage of the cracked gas compressor, thereby relaxing the strict specification for methane at the bottom of the EEC. The purpose in any situation must always be to maintain financial advantage and flexibility.

The uncondensed vapors from the top of the EEC include mainly hydrogen and methane. In two embodiments described, this stream is then fed to the methane extraction column (MEC) where methane is selectively extracted from the hydrogen-methane stream with a preferred physical solvent in solvent loop No. 2. The column is operated under strict specifications in terms of hydrogen content in the methane stream exiting the column through the bottom of the MEC , while most of the methane found in the raw material for MEC is extracted as a bottom product. Similarly, the operating purpose of the MSC is to maintain a low methane content in the hydrogen stream leaving the top of the column. Depending on the specifications of the hydrogen-rich gas stream, a rectification section can advantageously be used on top of the MEC. The hydrogen product that leaves the MEC can be processed further before use inside the olefin plant, e.g. for the hydrogenation of acetylene to ethylene.

The rich solvent containing methane and residual amounts of recovered ethylene are separated from the solvent in loop no. 2 by fractionation of the methane in the methane product column (MPC). The solvent that is thereby split is recycled to the top of the MEC for further extraction of methane. The solvent in loop no. 2 is treated in a similar way to the solvent in loop no. 1 by heat recovery and cooling to the desired temperature for extraction in the MEC.

It is possible to use two different preferred physical solvents in the two different solvent circulation vanes. Consequently, the solvent used in loop no. 2 will be relatively more selective towards methane than hydrogen, compared to the corresponding properties of the solvent used in solvent loop no. 1. It is also satisfactory if both solvent loops use the same physical solvent. However, it is important to note that the operating conditions for flows, temperature and pressure may be different as appropriate for the operation.

Another advantage of the method according to the present invention is that the olefin plant becomes more flexible when it comes to the possibility of processing different raw materials. In most designs, it has always been a serious challenge how much flexibility should be built in, so that the plant is kept as economical as possible at any given time. A significant remaining restriction is the limitations regarding the cooling systems and the cryogenic cooling parts. By applying the Mehra process, it became relatively easy to provide additional raw material flexibility. Due to the reduction of energy utilization in the separation section, the existing facilities can further be made more cost-effective and/or can be expanded at a nominal cost.

Under normal operating conditions for hydrocarbon distillation systems, alpha or the relative volatility of methane over ethylene is approx. 5.0. However, as stated in a paper entitled "High C02- High H2S Removal with SELEXOL Solvent" by John Sweny, presented at the 59th Annual Gas Processors Association Cinvention, 17-19 March 1980, the relative volatility of methane over ethylene 7.3. Consequently, in the presence of dimethyl ether of polyethylene glycol (DMPEG), the relative volatility of methane over ethylene is significantly improved (46%) compared to a normal fractionation system, indicating that DMPEG is selective towards the recovery of ethylene from a stream containing methane and ethylene. Since the relative behavior of methane and ethylene changes in the presence of a preferred physical solvent such as DMPEG, the contact process between the gas stream and the physical solvent is selective adsorption, or better defined as extraction rather than absorption.

The preferred physical solvents are defined for the purposes of the present invention as solvents having a minimal relative volatility of methane over ethylene of at least 5.5 (thereby defining the improved selectivity for ethylene over methane) and in addition a solubility of at least 7.0 normal cubic meters of gaseous hydrocarbons per m<3> of the solvent (thereby defining its hydrocarbon loading capacity), or alternatively, a preferential factor of at least 38. The preferential factor for selecting a physical solvent for use in the Mehra process concept of the present invention is defined as a product of relative volatility for methane over ethylene multiplied by the solubility of ethylene in physical solvents, expressed as normal cubic meters of ethylene per m<3>solvent (Nm<3>/m<3>). However, the ideal preferred physical solvent should have an ethylene over methane selectivity of at least 10.0 and at the same time have a hydrocarbon loading capacity of at least 21.2 Nm<3>/m<3>. When an ethylene purity of at least 99.5% and ethylene recovery of 99% is required, a preferential factor of at least 49 is highly preferred. In order to achieve an ethylene purity of at least 99.95% and an ethylene recovery of 99.5%, a preferential factor of 70 is highly preferred.

In addition, the preferred physical solvent is selected from the group consisting of dialkyl ethers of polyalkylene glycol, N-methylpyrrolidone, dimethylformamide, propylene carbonate, sulfolane, glycol triacetate and Cg to C^q aromatic compounds having methyl, ethyl or propylaliphatic groups which specifically constitute a subgroup of mestylene, n-propylbenzene, n-butylbenzene, o-xylene, m-xylene, p-xylene and mixtures thereof and aromatic streams rich in mixed xylenes and other Cg-C^g aromatic compds.

However, the method according to the present invention is also capable of using two different solvents. Equipping the first extraction column with a rectification section is particularly desirable if the solvents are different.

The process according to the present invention provides an ethylene product of high purity at high recovery levels, and also provides useful clean product streams of hydrogen and methane. The method is also believed to be characterized by extremely low solvent losses, reduced maintenance requirements, simplified equipment and lower capital costs, elimination of freezing, increased operating time, increased flexibility, and the possibility of using a wide range of raw materials. Furthermore, the process is also capable of producing propylene, with the same properties and advantages. Figure 1 is a schematic sketch of a typical olefin plant in which cracked gases are extracted and partially separated by the low-temperature fractionation process. Figure 2 is a schematic sketch of an older olefin plant showing the Kniel process according to U.S. Pat. patent No. 2,573,341 for recovery and separation of split gases by solvent absorption. Figure 3 is a simplified, schematic sketch showing the method according to Figure 1 with the Mehra process, for recovery and separation of cracked and/or refinery gases by solvent extraction, substituted for the cryogenic cooling part and the demethanization device in Figure 1. Figure 4 is a schematic flow diagram for the process for recovery and separation of split gases into a stream of hydrogen-rich gas, a stream of methane-rich gas, and a deethanizer feed stream using the Mehra process. This diagram fits into the block labeled "Mehra Process Extraction" in Figure 3 so that its complete C2=+ product is fed to the deethanizer column. Figure 5 is a schematic flow diagram showing another embodiment of the Mehra process in which an ethylene extraction column (EEC) extracts all the hydrocarbons of its solvent-rich stream to form a hydrogen-rich gas stream as the first product, a methane product column (MPC) extracts the ethylene and heavier hydrocarbons to its solvent-rich stream, so that a methane-rich gas stream is formed as the second product, and a crude product column (RPC) receives the MPC bottoms fraction, so that it produ-

a C2=+ hydrocarbon product stream is seen as the third product.

Figure 6 is a schematic flow diagram showing a third embodiment of the Mehra process in which the ethylene extraction column (EEC) separates the hydrogen and methane from the rich solvent containing ethylene and heavier hydrocarbons. The upper H2/CH4 stream is again extracted with lean solvent to recover an iterium-rich gas stream and produce a methane-rich bottoms stream which is depressurized to recover a CE^-rich gas stream.

With respect to Figures 3, 4, 5 and 6, it should be emphasized that pipes are often indicated when streams are identified in the following and that streams are intended, if not stated, when materials are mentioned. Furthermore, it must be emphasized that flow control valves, temperature control devices, pumps etc. is installed and operated in conventional relation to the main components of the equipment shown in the drawings and discussed in the following with reference to the continuous operating process according to the present invention. All these valves, devices and pumps, as well as heat exchangers, accumulators, condensers etc., are included in the term "auxiliary equipment". The term "absorption device" is conventionally used for a gas/solvent absorption device, but when the term is used in the method according to the present invention with a preferred physical solvent, the Device is assumed to be an "extractor".

It should also be emphasized that when "C2=+-hydrocarbons" are mentioned, the designation indicates ethylene and heavier hydrocarbons, both saturated and unsaturated and both alkyl and aromatic compounds. Similarly, "C±+" denotes methane and heavier hydrocarbons, both saturated and unsaturated and both alkyl and aromatic compounds.

The process shown schematically in the flow diagram of Figure 4 includes a rectification-extractive splitting column designated as the ethylene extraction column (EEC) 20 and a crude product column (RPC) 30 in the first solvent loop and a methane extraction column (MEC) 60 and a methane product column (MPC) 70 in the second solvent loop.

The split or refinery inlet gas stream 11 enters the gas compressor 12 which is equipped with aftercoolers and systems for removing mercaptans and dehydrating the gas. After the heavier fractions, the acidic impurities and the water have been eliminated from the pyrolysis products, the actual recovery of ethylene and propylene can take place. When such an elimination pretreatment is incorporated into the compression cycle, the selection and execution of the final product separation system is simplified to a considerable extent.

The compressed, cooled, mercaptan-free and dehydrated gas passes through pipe 13, which selectively includes an optional gas cooler 15, to the middle section of the EEC 20. Gases then flow upwards through the extraction section 23, while also flowing past the lean solvent moving downwards from tube 49, and then enters the rectification section 25, while flowing past the downward flowing return stream from tube 35, and finally exits the column through tube 24 as top fraction flow.

The combined reflux and solvent passes through the extraction section 23 in the splitting section 21 while encountering upward vapors produced from the bottom fraction liquid, after it has been heated by recirculation through tube 26 and evaporator 27, and from the intermediate liquid, after it has been heated by recirculation through pipe 28 and side evaporator 29. A portion of the upward vapors is extracted by means of the combined reflux/solvent liquid. The remaining vapors rise to the extraction section and mix with the incoming gas stream from pipe 13. The hot bottom liquid from the column is discharged through pipe 22a, passes through lean/rich solvent cross-exchanger 33, and is then fed through pipe 22b to the middle section of RPC 30.

Column 30 includes a splitting part 31 and a rectification part 35. The upward gases pass through rectification part 35, at the same time as they encounter downwardly flowing lean solvent from return pipe 54, and leave the system as top fraction stream through pipe 34, to be condensed in upper condenser 51, pass through pipe 52 and three into accumulator 53 from which a part is recycled to the top of RPC 30 by means of reflux pump 55 through pipe 54, while another part is pumped with the help of product pump 57 through pipe 56, so that ethylene and heavier hydrocarbons are produced which is supplied to the de-ethanisation device and other downstream equipment. Another part of uncondensed gases can be selectively recycled, if it is economically advantageous to do so, through pipe 59 to compressor for split gas 12.

The returning hydrocarbons in tube 54 of loop no. 1 containing traces of solvent, move downwards, in countercurrent to the upward gas flow from tube 22, through rectification part 35 and enter splitting part 31 at the same time as they encounter upward vapors which is located therein which is produced by passing the bottom fraction liquid through recycling pipe 36 and evaporator 37. The purpose of the rectification part 35 is to reduce solvent loss with the C2=+ hydrocarbon product. The heated bottoms fraction material at the bottom of RPC 30 is discharged through tube 32 of the return section of solvent loop #1, passing through lean/rich solvent cross-exchanger 33, tube 41, evaporator 27, side evaporator 29, cooler for lean solvent 43, pipe 45, pipe 41, lean solvent pump 47 and supply pipe 49 to the top of extraction section 23 of EEC 20.

Top fraction section 24 passes through upper condenser 31 and pipe 32 and enters accumulator 34 from which liquid is recycled through pipe 35 by reflux pump 36 to the top of rectifier section 25 in EEC 20. The first solvent loop and its columns and auxiliaries are thus fully described .

Uncondensed gases in accumulator 34 are led through supply pipe 33 to the middle part of methane extraction column (MEC) 60, between splitting part 61 and extraction part 65. The gases from pipe 33 pass up and through extraction part 65, at the same time they meet lean solvent from pipe 89, and leave the column as a hydrogen-rich gas product stream 64. The descending lean solvent from pipe 89 passes through extraction section 65, at the same time effectively absorbing materials from the gas stream entering through pipe 33, at the same time encountering upward vapors and absorbing methane from these and at the same time it loses hydrogen to the vapors. The liquid bottom fraction material in the column is heated by recirculation through tube 66 and evaporator 67. The bottom fraction material in the column is discharged through tubes 62a and 62b, and lean/rich solvent cross-exchanger 73 and supplied to the middle part of MPC 70.

Column 70 includes a splitting part 71 and a rectification part 75. The gases heated in exchanger 73 pass upwards through rectification part 75 at the same time as they meet downwardly returning material from pipe 94. The top fraction stream from the column leaves it via pipe 74, upper condenser 91 and pipe 92, and enters accumulator 93. From this accumulator, liquid material is recycled through pipe 94 and reflux pump 95 to the top of the rectification part 75 of MPC 70. Steam from accumulator 93 is discharged as a methane-rich gas stream through pipe 99.

The falling reflux liquid takes up solvent in loop No. 2, while giving off methane and any C2=+ hydrocarbons during its passage through the rectifier section 75, and then enters the splitter section 71 where it encounters upward vapors from the heated bottom fraction material in the column. Heating of this material is achieved by recirculating it through pipe 76 and evaporator 77. The heated bottom fraction material, now lean solvent, is discharged through pipe 72, passing through lean/rich solvent cross exchanger 73, pipe 81, evaporator 67 , lean solvent cooler 83, pipe 85, lean solvent pump 87 and feed pipe 89, to the top of the extraction section 65 of the MEC 60. The second solvent loop and its columns and associated equipment are thus fully described.

In summary, the purpose of EEC 20 is to selectively extract all hydrocarbon components of inlet gas stream 13 by effectively utilizing the selectivity of the preferred physical solvent in stream 49. The purpose of RPC 30 is to remove all hydrocarbon components from the inlet gas stream that are present in rich solvent stream 22b and produce the lean solvent stream in tube 32 plus the porous stream of C2 = + hydrocarbon components present in the overhead fraction stream 34.

The purpose of MEC 60 is to isolate most of the hydrogen as a top fraction stream in tube 64 and isolate most of the methane as part of the rich solvent stream in tube 62a. The purpose of MPC 70 is to isolate substantially all of the remaining methane in methane-rich gas stream 99, which includes small amounts of C2 and C3 hydrocarbons, and to regenerate the lean solvent in stream 72.

Figure 4 shows a system in which two completely different preferred physical solvents can be used, if desired, in two solvent loops. The first, or primary loop, is aimed at separating the H2 and CH4 in the top product stream from the C2=+ hydrocarbons in the rich solvent stream, and then recovering the C2=+ hydrocarbons and regenerating the solvent. The second or secondary loop is directed towards the recovery of a hydrogen-rich gas stream and then towards the recovery of a methane-rich gas stream while simultaneously regenerating the solvent. The two loops are connected by the peak fraction flow from the first loop which constitutes the feed gas flow for the second loop.

When two different preferred physical solvents are used in the two different solvent circulation ice loops, the solvent used in loop no. 2 of figure 4 will be relatively more selective for methane over hydrogen, compared with corresponding properties for the solvent used in solvent loop no. 1. such first solvent is preferably an aromatic distillate, such as mesitylene or C8-C8 monocyclic aromatic compound which has at least one alkyl side chain and which has a lower boiling point and a higher preferential factor than the solvent in the second loop such as e.g. may be dimethyl ether of polyethylene glycol (DMPEG).

If the same preferred physical solvent is used in both solvent loops, C8-C8 monocyclic aromatic solvents are preferred. However, it is important to note that the operating conditions in terms of currents, temperatures and pressures may differ depending on what is required in the operation.

Each loop passes through two columns, one column is an extractive splitting column and the other column is a distillation column. Each extractive splitting column has at least one splitting part and/or a rectification part and/or an extraction part. Each of the four columns is also equipped with an evaporator for heating the well fraction material. The first column of each loop is an extractive splitter column, and control of its performance is accomplished in part by controlling the temperatures and flow rates of the lean solvent streams fed to the upper portions of the columns. Preferably, the first column of the first part is additionally equipped with a return flow which is supplied to the column at a supply point at a distance above the supply point of the lean solvent stream, thereby defining a rectification part therebetween. The second extractive splitting column is preferably not equipped with a return flow, but it can have this if relatively pure hydrogen is desired. The second column of each loop is a product column, the first is for the isolation of crude C2=+ hydrocarbons as an overhead fraction stream, and the second is for the isolation of impure methane as an overhead fraction stream.

Figure 5 illustrates an extractive splitting process with two columns for a gas stream containing olefins, in which e.g. initial separation of hydrogen from all the hydrocarbon components in the first column takes place. This process uses ethylene extraction column (EEC) 120, solvent regenerator column (SRC) 130, methane product column (MPC) 140, and crude product column (RPC) 150.

Compressed, mecraptan-free, cooled and dry gas enters column 120, just above the midpoint, through tube 121. Liquid at the bottom of the column circulates through tube 125 and vaporizer 126 to be heated. Bottom fraction in column 120 leaves this through pipe 123 and solvent exchanger 127. Top fraction from column 120 leaves this through pipe 124 as a hydrogen-rich gas product.

The heated rich solvent in exchanger 127 passes through tube 131 and enters column 130, slightly above its midpoint. Liquid at the bottom of column 130 circulates through tube 135 and evaporator 136 to be heated. Bottom fraction from column 130 leaves this through pipe 133 and pump 137 and enters heat exchanger 127 and passes through pipe 128, evaporator 126, and solvent cooler 129 and enters the top of column 120 as lean solvent.

An overhead fraction stream leaves column 130 through tube 134, is cooled in reflux condenser 134a, enters accumulator 138, and is separated into uncondensed and condensed hydrocarbons. The latter leaves the column through pipe 139a, is pumped by means of return flow pump 139 to the pressure of column 130, and enters the top of column 130 as return flow. The uncondensed hydrocarbons leave the accumulator 138 through pipe 141 and enter the column 140, slightly above its midpoint. Liquid at the bottom of column 140 circulates through pipe 145 and evaporator 146 to be heated. Bottom fraction leaves column 140 through tube 143 and enters rich/lean solvent exchanger 147 for heating therein. An overhead product gas stream exits the top of column 140 through pipe 144.

Heated solvent leaves exchanger 147 through tube 151 and enters column 150, slightly above its midpoint. Liquid at the bottom of column 150 is circulated through pipe 155 and evaporator 156 to be heated. Bottom fraction leaves column 150 through tube 153, is cooled in exchanger 147, passes through tube 148 and evaporator 146, and is further cooled in solvent cooler 149 before entering the top of column 140. A top fraction stream leaves the top of column 150 through tube 154 and is partially condensed in condenser 161 before it enters reflux accumulator 162. Condensed liquid in accumulator 162 leaves this through pipe 163 and is pumped by means of reflux pump 167 to the top of column 150. Uncondensed hydrocarbons in reflux accumulator 162 leave this through pipe 164 and become C2=+ hydrocarbon product.

Figure 6 illustrates a separation and recovery process for a gas containing olefins that has been compressed, cooled, freed from mercaptans and dried. The process uses two extraction columns which can be arranged as separate columns or built together as a single tall column. These are first or bottom fraction ethylene extraction column (EEC) 170 and a second or top column (MEC) 190. The process also uses a pressure relief vessel (FV) 200 and a crude product column (RPC) 210.

The inlet gas stream in tube 171 enters EEC 170, slightly above its midpoint, and passes upward to meet descending lean solvent from tube 178. Liquid at the bottom of column 170 circulates through tube 175, and vaporizer 176, and tube 177 to be heated and is returned to column 170. Bottom fraction in column 170 leaves this through pipe 173 and enters and is heated in rich/lean solvent exchanger 181. A top fraction stream in pipe 179 leaves the top of column 170, passes through any compressor 185 and tube 191, and enters MEC 190 near its bottom. The evaporator can, however, initially be used to achieve a further splitting part below the supply point which is slightly below the middle of the column.

Bottom fraction leaves the column 190 through pipe 193 and enters the pressure relief vessel 200, where the bottom fraction is separated into a methane-rich gas product stream which leaves the vessel 200 through the top fraction pipe 209, and a bottom fraction stream which leaves the vessel 200 through pipe 203 and solvent pump 205 before entering the top of column 170 through pipe 178. An overhead fraction stream leaves the top of column 190 through pipe 199 and passes through a possible power recovery turbine 187 and leaves the plant as a hydrogen-rich gas product stream in pipe 197.

Heated, rich solvent exits exchanger 181 through tube 211 and enters RPC 210, somewhat below its midpoint. Liquid in the bottom fraction thereof is circulated through tube 215, evaporator 216, and tube 217 to be heated and returned to column 210. Bottom fraction leaves column 210 through tube 213 and is cooled in exchanger 181. Lean solvent leaves exchanger 181 through tube 182, cooled in evaporator 176, passes through tube 183, further cooled in solvent cooler 184, and enters the top of MEC 190 through tube 198. An overhead fraction stream leaves RPC 210 through tube 219, is condensed by condenser 221, and enters return flow accumulator 223, through pipe 222 where it is separated into liquid and vapor parts. The liquid part leaves the device through pipe 225 and is pumped by means of return flow pump 226 through pipe 218 and enters the top of the column 2*10. The steam part leaves the device through pipe 224 and becomes C2 = + -hydrocarbon product which is suitably condensed.

Depending on the relative concentrations of Etø and CH4 compared to ethylene and heavier hydrocarbon components, the solvent requirements for EEC 170 may be significantly lower than those for MEC 190. It may therefore be economically desirable to divert a significant portion of the solvent in pipe 178 via pipe 206 to lean solvent pipe 183. By doing this, the control size for the EEC 170 and part of the solvent loop, including the units 176, 181, 210, 216, 221, 223 and 226 and their associated piping, can be significantly reduced.

EXAMPLE

An existing olefins plant, as shown in Figure 1, is being investigated for improvement by replacing the cooling section and the demethanizer column with a Mehra process extraction system, as shown in Figure 3. Portions of the propylene cooling system are to be preserved for use with the Mehra process - the extraction units. The ethylene cooling system can be closed, thereby further reducing the load on the propylene cooling system. Replacement of the molecular sieve or dehydration unit in Figure 1 with the glycol dehydration unit in Figure 3 shall be assessed on the basis of cost-effectiveness, but for the purposes of the present example it is assumed that the molecular sieve units remain in operation. The Mehra process extraction system, using o-xylene as one of the preferred physical solvents, is as shown in Figure 6.

Using a commercially available and reliable computer implementation, the specific process of Figure 6 was simulated. The operating conditions were only partially optimized and do not reflect a final performance. In this simulation, a stream of thermally cracked olefinic gases, which were compressed, cooled, free of mercaptans and dehydrated, was assumed to enter EEC 170 through pipe 171 at 8608.82 kg-mol/h at 32°C and 2758 kpa . A stream of lean solvent was also assumed to enter EEC 170 through pipe 178 and to flow at a rate of 15548.00 kg-mol/h at -13.20C and 3103 kpa. Three products were calculated as recovered: 1949.03 kg-mol/t of hydrogen-rich gas product in pipe 199, 1734.00 kg-mol/t of methane-rich gas product in pipe 209 and 4932.30 kg-mol/t of ethylene and heavier hydrocarbon product in tube 224. The compositions, other flow rates, temperatures and pressures are given in Table II. The composition of the cracked feedstock in pipe 171 is typically a thermally cracked 70/30 ethane/propane mixture on a liquid volume basis.

Product Purities

The composition of the hydrogen-rich gas product in pipe 199 should be noted because its hydrogen content is 93.4% on a kg-mole basis. Furthermore, its solvent content is only 0.0017% on a kg-mole basis and its ethylene content is essentially 0.

The composition of the methane-rich gas product in pipe 209 is also noteworthy because its methane content is 89.5% on a ln kg-mole basis. In addition, its solvent content is 0.04% on a kg-mole basis and its ethylene content is 0.043% on a kg-mole basis.

The composition of the stream of ethylene and heavier hydrocarbons in pipe 224 is 56.3% ethylene, 22.3% ethane and 0.030% methane on a kg-mole basis. Its o-xylene content is 0.0032% on a kg-mole basis.

Recycling conditions

Assuming that all compounds other than hydrogen, methane, ethylene, benzene, and toluene were to be recycled from the Mehra process system and returned to the heaters for further cracking, the recycle ratio would be 1.8, compared to the recycle ratio of 1.0 in the Kniel absorption process according to U.S. Patent No. 2,573,341. It is believed that this advantageous recycle ratio is the result of selecting ethylene-favoring reaction conditions, as compared to aromatics-favoring reaction conditions, which were clearly required by the Kniel process.

Significance of calculations

The results shown in table II are assumed to be very significant because the commercially available simulation package that was used in the present example is routinely used for the design of facilities. The results shown in Table II also reveal that the Mehra process fulfills what the prior art does not. Such previously unattainable results demonstrate that the present embodiment of the Mehra process represents a discovery of great importance to the ethylene production technique.

One can e.g. compare the losses of solvent which were a serious drawback of the Kniel absorption process with the losses shown in Table II. Streams 199, 209 and 224, all streams leaving the system, show a total loss of o-xylene amounting to 0.894 kg-mol/h. The total amount of o-xylene 1 stream 198 is 15427.96 kg-mol/h. This loss, expressed as a percentage of the total o-xylene stream, is just 0.06% on a kg-mole basis, a phenomenal feat in solvent absorption technology. If mesitylene were to be used as a preferred physical solvent as in this example, the expected solvent losses would be significantly lower than those of o-xylene.

Another possible comparison is on a methane rejection basis by reference to the 1947 article in Chemical Engineering Progress which states that the methane retained in the grease oil represented 4% of the ethylene in the cracked gas, in which the volumes of methane and ethylene were approximately equal. In contrast to this, the methane in streams 173 and 224 is equal to 1.50 kg-mol/h which is 0.05% of the ethylene in stream 171, in which the volume of methane is 2/3 of the volume of ethylene.

Purity of etvlen product

The crude hydrocarbon product is further fractionated downstream into ethylene and ethane. To do this while achieving sufficiently high purity in the top fraction ethylene stream requires a high level of efficiency in the ethylene fractionator column.

As can be seen from table II, at approx. 72 mol-% ethylene i t i 1 f r e 1 s s s stream to an ethyl enfractionator column Containing only ethylene and ethane, as 1 stream 224, an ethylene product is produced, with a purity of 99.95 mol-% at a recovery level of 99.94%. This effect cannot now be achieved with the Kniel process even when a demethanization column is used, such a column is also clearly not required with the Mehra process.

Claims (20)

1. Process for the production of a hydrogen-rich product stream, a methane-rich product stream, and a product stream of ethylene and heavier (C2=+) hydrocarbons, characterized by the extraction of a mass of compressed, mecaptan-free, dry gas containing olefins with a stream of a lean, preferably physical, solvent at a selected solvent flow rate. 2. Method according to claim 1, characterized in that the olefin-containing gas is selected from the group consisting of thermally cracked hydrocarbon gas, catalytically cracked hydrocarbon gas, refinery off-gas, coke oven gas and synthesis gas. 3. Method according to claim 2, characterized in that the preferred physical solvent is selective for ethylene and heavier hydrocarbon components of said C2=+ hydrocarbons, so that the relative volatility of methane over ethylene is at least 5.5, and the solubility of ethane in said solvent. is at least 7.0 normal m<3>ethylene per m<3>solvent, or the preferential factor is at least 39. 4. Method according to claim 3, characterized in that the preferential factor is at least 49. 5. Method according to claim 4, characterized in that the preferential factor is at least 70. 6. Process according to claim 3, characterized in that the preferred physical solvent is selected from the group consisting of dialkyl ethers of poly-alkylene glycol, N-methylpyrrolidone, dimethylformamide, propylene carbonate, sulfolane, glycol triacetate and Cg to C^q aromatic compounds having methyl, ethyl or propylaliphatic groups which specifically constitute a subgroup of mesitylene, n-propylbenzene, n-butylbenzene, o-xylene, m-xylene, p-xylene and mixtures thereof and aromatic streams rich in mixed xylenes and other Cg-C^g aromatic compounds. 7. Method according to claim 6, characterized in that: A. said selected solvent flow rate is within the range from 0.1 to 70 m<3> of solvent per 1000 normal m<3> of said gas stream and in accordance with the composition and flow rate of the gas stream; B. the olefin-containing gas stream is supplied to at least one extractive splitting column which has a device for heat supply "and to which a stream of the lean, feeds the drawn physical solvent is supplied, whereby said gas and said solvent pass countercurrently inside it, so that at least one top fraction stream and at least one bottom fraction stream are produced; and C. at least one bottom fraction stream is supplied to at least one distillation column in which the lean solvent stream is regenerated. 8. Process according to claim 7, characterized in that: A. the olefin-containing gas stream is supplied to the first of two extractive splitting columns placed in series, the gas stream passes in countercurrent in each column to a stream of the lean preferred physical solvent, thereby producing a first top fraction stream, a first bottom fraction stream, a second top fraction stream and a second bottom fraction stream, at least one of said top fraction streams being one of said products; and B. feeding the first and second bottom fraction ion streams to two distillation columns which regenerate said streams of said lean preferred physical solvent and produce at least one of said products. 9. Method according to claim 8, characterized in that each extraction column contains a splitting part, which has a device for heat supply and an extraction part above this. 10. Method according to claim 9, characterized in that the aforementioned device for heat supply includes a bottom evaporator and a side evaporator. 11. Method according to claim 10, characterized in that at least one of said extractive splitting columns additionally includes a rectification part to which a return flow is supplied, where the rectification part is placed above the extraction part. 12. Method according to claim 8, characterized in that each lean solvent stream flows as part of a solvent loop, where the first extractive splitting column contains part of a first solvent loop and the second extractive splitting column contains part of a second solvent loop. 13. Method according to claim 12, characterized in that the solvent in the first solvent loop and the solvent in the second solvent loop are the same preferred physical solvent. 14. Method according to claim 12, characterized in that the solvent in the first solvent loop and the solvent in the second solvent loop are different physical solvents. 15. Method according to claim 8, characterized in that: A. the first top fraction stream from the first extraction column is supplied to the second extractive splitting column which produces said hydrogen-rich product stream; B. the largest bottoms fraction stream is fed to the first of said distillation columns which regenerates the solvent and produces said stream of ethylene and heavier hydrocarbons; and C. the second bottoms fraction stream is fed to the second of said distillation columns which produces the methane-rich product stream as the top fraction stream. 16. Method according to claim 8, characterized in that: A. the first peak fraction stream is the aforementioned hydrogen-rich stream; B. the first bottom fraction stream is fed to the first of said distillation columns which regenerates the first stream of the lean preferred physical solvent and produces a top fraction stream of methane and heavier hydrocarbons which is fed to the second extractive splitting column; C. the second extractive splitting column produces said methane-rich product stream and said second bottoms fraction stream; and D. the second bottom fraction stream is fed to the second distillation column which regenerates the solvent stream and produces said stream of ethylene and lighter hydrocarbons. 17. Method according to claim 7, characterized in that said top fraction stream is fed to an extraction column to which a second stream of lean preferred physical solvent is fed, whereby the hydrogen-rich stream is produced a bottom fraction stream which is reduced in pressure to produce the said methane-rich product stream and a another stream of the preferred physical solvent which is fed to the first extractive splitting column. 18. Improved process according to claim 17, characterized in that the first extractive splitting column is a rectifying-extractive-splitting column having a partial overhead condenser for said overhead fraction stream and reflux means for returning condensed liquid to the column at a feed point above the feed point of the preferred the physical solvent, thereby defining a rectification part above the extraction part. 19. Improved process according to claim 18, characterized in that the contact in said rectifying-extracting-splitting column takes place at less than 3200 kpa. 20. Improved process according to claim 19, characterized in that the bottom fraction stream is a rich solvent stream which is heated by means of lean/rich solvent cross-exchange with said stream of lean preferred physical solvent.