FR3009204A1 - METHOD FOR CAPTURING HEAVY METAL CONTENT IN WET GAS WITH DILUTION OF WET GAS TO CONTROL RELATIVE GAS MOISTURE - Google Patents

Abstract

The method for capturing at least one heavy metal, chosen from mercury and arsenic, contained in a wet gas comprising water vapor, the wet gas having a relative humidity RH greater than 75%, comprises the following steps: a) the wet gas is mixed (1000) with a dilution gas so as to obtain a gas mixture having a relative humidity HM lower than the relative humidity RH, b) the gas mixture is brought into contact with a mass of capturing said heavy metal (2000) to obtain a heavy metal depleted gas mixture.

Description

The present invention relates to the field of the treatment of gaseous effluents containing heavy metals, in particular effluents of petroleum origin and their derivatives such as natural gas and synthesis gas. More specifically, the invention relates to the capture of heavy metals, and in particular mercury or arsenic, present in a wet gaseous effluent, using a process which, at first, makes it possible to reduce the hygrometry gas and then, in a second step, to purify the effluent thus partially dehumidified by a heavy metal capture technology. Mercury is a metal contaminant found in gaseous or liquid hydrocarbons produced in many parts of the world such as the Gulf of Niger, South America, North Africa or the Asia-Pacific region. The removal of mercury from hydrocarbons is desired at the industrial level for several reasons. On the one hand, the presence of mercury in these hydrocarbons poses risks to operators working in contact with these products because mercury is toxic. In elemental form, mercury is volatile and presents a serious risk of inhalation neurotoxicity. In organic form, mercury has similar risks of skin contact neurotoxicity. On the other hand, the presence of mercury in hydrocarbons is detrimental to conventional processing operations used to recover these hydrocarbons. Conventionally, the hydrocarbons are subjected to catalytic reactions such as the selective hydrogenation of olefins produced by steam cracking or catalytic cracking of liquid hydrocarbons. However, the catalysts used generally comprising noble metals such as platinum and palladium can be deactivated by mercury. Indeed, mercury induces sintering of catalysts by amalgam nanoparticles noble metals. The reduction of the specific surface of the catalysts leads to a very significant loss of their catalytic activity. Finally, the elimination of mercury is important because its presence can lead to corrosion problems of cryogenic heat exchangers based on aluminum, which can lead to serious industrial consequences.

Among other things for these reasons, it is desired to eliminate or at least reduce the mercury concentration in the gaseous hydrocarbon effluents. Industrially, the elimination of mercury from the gaseous effluents is achieved by a circulation of the effluent to be treated through mercury guard beds (or demercurization unit) filled with adsorbent materials, otherwise called capture masses. The impurity to be removed, here mercury, is then retained irreversibly, preferably by chemisorption, within or on the surface of the capture mass and the effluent discharged from the capture mass bed is then purified. Mercury uptake can be achieved by reacting mercury with an elemental sulfur phase in a capture mass. Indeed, the elemental sulfur, S, reacts irreversibly with the elemental mercury, Hg °, as follows: 10 Hg ° (g / 1) + S (s) HgS (s) (1) By "Hg ° (g / 1) "is understood to mean that the mercury is dissolved in a gaseous (g) or liquid (I) fluid phase. In contrast, "(s)" refers to the solid phases consisting of the active phase of the uptake mass and the product of the reaction. The formed product, HgS, called cinnabar or metacinabrium, is a chemically and solid inert mineral phase over a wide temperature range. The mercury is thus trapped in the capture mass and the effluent to be treated is purified. Other active phases may be used such as metal sulphides such as, for example, copper sulphide (CuS). Typically, the capture masses are obtained by active phase impregnation methods on porous supports of the activated carbon or alumina type or by co-granulation of the active phase with a binder such as for example oxide aluminum. However, these capture masses may have malfunctions when the gaseous effluent to be treated is wet. On the one hand, depending on the active phase chosen, it may be leached by the presence of liquid water or liquid hydrocarbons and be entrained in the stream to be treated, or the binder may react with water to form a dense and compact phase that will block the flow of gas. On the other hand, the presence of water in the form of steam at high hygrometry levels in the gas to be treated may lead to the appearance of capillary condensation phenomena on the porous supports used. This phenomenon results in the appearance, at a given temperature, of liquid water at pressures lower than the saturation vapor pressure of water (Po).

For a cylindrical model pore, the Kelvin equation (Equation 2) allows to determine the critical pore radius (Rc) below which the pores will be filled with liquid water.

Rc = -27Vmcos9 / RT / log10 (P / P0) (2) where P is the gas pressure, T is the gas temperature, R is the ideal gas constant (R = 8.314 JK - 'mol-1) , Vm is the molar volume of water, y is the gas / water surface tension and 0 is the water / solid contact angle. P / Po corresponds to the definition of the relative humidity of the effluent. Smaller pores, especially micropores (d <2nm) are therefore much more sensitive than mesopores (2 <d <50 nm) or macropores (d> 50nm) to the phenomenon of capillary condensation (d corresponding to the diameter of pores). The relative humidity of the natural gas, also called humidity rate or hygrometry rate or degree of hygrometry, corresponds to the ratio of the partial pressure of water vapor contained in the gas to the saturated vapor pressure, also called vapor pressure, at the same temperature. It is therefore a measure of the ratio between the water vapor content of the air and its maximum capacity to contain it under these conditions of pressure and temperature.

The capillary condensation mechanism can also take place with hydrocarbon vapors or volatile organic compounds. The presence of capillary condensation strongly impacts the operation of the collection mass since it causes the appearance of a high resistance to transfer of material into the bed and prevents the mercury from accessing the entire active phase. Very often, the performances of the guard bed are then strongly altered. A decrease in the performance of sulfur-type capture masses on activated carbon has thus been observed in the nocturnal operating range of the guard bed. This dysfunction is attributed to the drop in temperature of the reactor during the night which causes the appearance of capillary condensation in the bed. Similarly, it has been shown that for a model gas (mercury in nitrogen), having a relative humidity of 10%, the performance of the sulfur-based capture mass deposited on activated carbon decreases by 25%.

It is therefore important to place the mercury guard bed at an appropriate place on the process chain to operate the unit optimally. In general, on a natural gas exploitation, the gas is extracted from the geological medium on production wells. At the output of production wells, natural gas is loaded with water and liquid hydrocarbons (condensates). A three-phase flow composed of gas, liquid and sludge is thus obtained, which is channeled into a unit for trapping sludge, commonly called "slug catcher". Generally, the "slug catcher" is in the form of pipes arranged according to a studied slope which makes it possible to control the flow of the mixture and to deposit the sludge.

At the output of this unit, the mixture is sent to a separator called primary. Three phases are thus obtained: water, condensates and gas. The gas from the primary separator is sent to a coalescer which returns to the primary separator the liquid entrainment. At the exit of the coalescer, the gas is theoretically just saturation, but in practice there is often the presence of more or less significant liquid entrainment, usually in the form of droplets suspended in the gas, depending coalescer performance installed. The gas is then generally deacidified by an acid gas removal unit, generally an amine treatment, which selectively removes H 2 S and O 2. The gas thus treated then passes to a drying unit which makes it possible to lower the rate of hygrometry up to values of a few ppm. The drying of the natural gas can be conducted exclusively by contact with a glycol solution or by circulation in a bed of adsorbents. The mercury guard bed is very often placed in the process chain after these dryers. Since the gas is significantly dried, the guard bed can operate under favorable conditions, in particular by avoiding problems of capillary condensation in the mercury adsorbent material. However, these drying units often use glycol compounds in which the mercury can be solubilized. WO2005 / 047438 shows in particular that the concentration of mercury in the glycol can reach high values, of the order of 2.9 ppm. During the regeneration step, the glycol solution is heated to temperatures close to T = 200 ° C and a portion of the mercury is then released into the atmosphere. Another solution proposed in US Pat. No. 5,223,145 consists of a process that eliminates both mercury and water. In this scheme, a first bed of molecular sieve is used to remove water and mercury concomitantly. As the purification efficiency decreases, the flow is directed to a second regenerated molecular sieve bed. The first bed is then regenerated by passing a hot dry gas at the inlet of the column to generate a wet gas polluted with mercury at the end of this column. This regeneration gas must then itself be treated with a guard bed to remove the mercury. This solution is particularly unsuitable since on the one hand it merely transfers the problem from one effluent to another and on the other hand it requires a) to regenerate the adsorbers frequently at a temperature which is difficult to reach on a treatment plant. gas and b) to use a second non-regenerable mercury removal system, thereby increasing the associated investment costs. Another solution proposed in US5120515 is to have in a single column desiccants for removing water and aluminas impregnated with an active phase for capturing mercury. This solution is also not economically viable since it is necessary to periodically regenerate desiccants which leads to additional economic costs. Moreover, it is mentioned in this patent that this regeneration step results in mercury releases that must be reprocessed by another unit of demercurization. In addition, the positioning of the mercury guard bed downstream of the drying device causes the mercury contamination of the entire process chain upstream of the guard bed. Even if the mercury guard bed is placed between the deacidification and dehydration units, the entire process chain upstream of the guard bed, in particular the amine treatment unit, is polluted with mercury. It is possible to place the mercury guard bed as far upstream as possible on the process chain, for example after the coalescer. This positioning nevertheless leads to several disadvantages. If the coalescer is not very effective or even damaged, continuous liquid (water and condensate) drives can feed the demercurization unit. On the other hand, even at saturation, heat losses in the line between the coalescer and the demercurization unit can generate condensation. Especially if the line is long, if it is raining or if it is cold, the line is not insulated the demercurization unit will receive more liquids.

Finally, there is a phenomenon of retrograde condensation, thermodynamic particularity of natural gas, generating condensation during the isothermal pressure drop of natural gas. In practice, the pressure losses in the line, but also in the demercurization unit, can generate a partial condensation of water and hydrocarbon condensates. The most direct solution is then to lower the hygrometry rate of the wet gas by applying overheating of the natural gas through for example the presence of a superheater. However, given the flows commonly encountered in natural gas units, this solution proves to be very expensive energy and consequently financially. The object of this invention is to propose an optimized wet gas treatment method which makes it possible to place the heavy metal guard bed downstream of a gas-liquid separator, for example a coalescer, and, preferably, upstream of the acid gas removal step while ensuring optimal operation of the heavy metal collection masses by performing a partial dehumidification of the wet gas from the gas-liquid separator.

The Applicant has discovered that, cleverly, the dilution of the wet gas by a dilution gas, preferably with a dilution gas which increases the saturation vapor pressure, makes it possible to respond to the problem previously described by ensuring sufficient relative humidity low to avoid the problems of condensation and, a fortiori, of capillary condensation on the masses of the guard bed of heavy metal. In general, the invention relates to a process for capturing at least one heavy metal, chosen from mercury or arsenic, contained in a wet gas comprising water vapor, the wet gas having a higher relative humidity RH. at 75%, in which the following steps are carried out: a) the moist gas is mixed with a dilution gas so as to obtain a gas mixture having a relative humidity HM lower than the relative humidity HR, b) it is necessary to contact the gas mixture with a heavy metal capture mass to obtain a heavy metal depleted gas mixture. According to the invention, the dilution gas can be selected from 002, H2S and a gas having a relative humidity of less than 50%. In step a), the flow rate of the dilution gas can be determined so as to obtain a gas mixture having a relative humidity HM lower than the relative humidity RH. The moist gas can be mixed with a dilution gas to obtain a gas mixture having a relative humidity HM of less than 90%, preferably less than 80%, or less than 75%. To capture the mercury, the capture mass may comprise an active phase selected from at least one metal sulphide based on a metal selected from the group consisting of copper (Cu), chromium (Cr), manganese (Mn) , iron (Fe), cobalt (Co) and nickel (Ni). The capture mass may comprise an active phase composed of elemental sulfur. When the gas contains H2S, the capture mass may comprise an active phase composed of at least one metal oxide precursor of a metal selected from copper (Cu), chromium (Cr), manganese ( Mn), iron (Fe), cobalt (Co) and nickel (Ni). To capture arsenic, the capture mass may comprise an active phase composed of at least one metal oxide of a metal selected from copper (Cu) and lead (Pb). The active phase may be distributed on a porous support, the porous support being chosen from the group consisting of aluminas, phosphorus aluminas, silica-aluminas, silicas, clays, activated carbons, zeolites, titanium, zirconium oxides, and their mixtures. After step b), the following steps can be carried out: c) at least a portion of the H 2 S and CO 2 contained in the heavy metal depleted gas mixture are removed so as to obtain a mixture of depleted gases in H25 and in CO2 and a flow rich in H25 and 002, then d) is removed at least a portion of the water vapor contained in the H2S-depleted gas mixture, to CO2 to obtain a gas depleted in water.

In the case where CO2 is chosen as the dilution gas, after step b) and before step c), a membrane separator can be used to separate at least part of the CO2 content. in the heavy metal depleted gas mixture so as to obtain a CO2-rich stream and at least a portion of said CO2-rich stream can be recycled in step a) as a dilution gas. At least a portion of said H2S and CO2 rich stream can be recycled in step a) as a diluent gas. A portion of the water-depleted gas obtained in step d) can be recycled in step a) as a diluent gas.

The wet gas may be selected from one of the following gases: natural gas, shale gas, coal gas, synthesis gas, combustion fumes, gaseous hydrocarbon effluent, chlorine alkali plant vents , rare earth production plant vents. For the case of a natural gas, before carrying out step a), the following steps can be carried out: a natural gas is extracted from an underground deposit, and then the natural gas is introduced into an elimination device sludge, then the natural gas is introduced into a liquid gas separation device, then the natural gas is sent to step a), said liquid gas separation device may comprise a coalescer to separate the condensed water from the natural gas . Other features and advantages of the invention will be better understood and will be apparent from the following description with reference to the drawings, in which: FIG. 1 shows the method of lowering the partial pressure of steam of water of a wet gas according to the invention, Figures 2 to 4 show three embodiments of the method according to the invention, integrated in a process for producing a natural gas. The invention involves diluting a stream of wet gas to be treated on a heavy metal guard. This dilution is done by mixing with a dilution gas, which is preferably dry. The gas mixture is then found under conditions close to the initial conditions but with a partial pressure of water vapor substantially away from the saturation vapor pressure. The water vapor pressure in the stream downstream of the gas dilution step according to the invention can be regulated by at least two parameters: First, the amount of dilution gas used to dilute the flow of wet gas. Secondly, the nature of the dilution gas. Indeed, depending on the nature of the dilution gas chosen, the saturation vapor pressure may be modified. In the case of using dilution gases such as CO2 or H2S, the saturation vapor pressure increases. The judicious choice of the dilution gas saves the amount of diluent to be used for an identical result. Referring to Figure 1, the wet gas arrives through the conduit (100). The gas flowing through the duct (100) can be at a pressure of between 2 and 10 MPa, preferably at a pressure of between 5 and 9 MPa, and at a temperature of between 20 and 80 ° C., preferably at a temperature between 25 and 70 ° C. The wet gas comprises water vapor, for example at a relative humidity of between 30% and 100%. The wet gas also comprises at least one heavy metal, for example mercury and / or arsenic, in proportions ranging between 10 nanograms and 1 gram of mercury per Nm3 of gas. The gas flowing in the duct (100) is mixed in the equipment (1000) with a dilution gas arriving via the duct (200). This results in a diluted wet gas discharged through the conduit (101). For example, the equipment (1000) may be a capacity supplied by the ducts (100) and (200) and provided with an outlet duct (101) for the diluted effluent or a simple union of the ducts (100) and (200) resulting in a common duct (101), for example in a Y configuration. Mixing the dilution gas with the humid gas makes it possible to obtain a gas mixture having a relative humidity lower than the relative humidity of the gas wet. According to the invention, the quantity, that is to say the flow rate of gas arriving via the conduit (200) can be determined in such a way that the gas mixture obtained in the conduit (101) has a lower relative humidity at 90%, preferably less than 80%, very preferably less than 75% so as to prevent capillary condensation in the guard bed (2000) containing the heavy metal capture mass. This hygrometry threshold depends on the porous characteristics of the heavy metal capture mass. For example, the molar flow ratio between (100) and (200) is thus typically between 2 and 10, preferably between 4 and 10. The previously mentioned ratio also depends on the nature of the gas. Preferably, the dilution gas has a relative humidity of water of between 0 and 50%, preferably between 0 and 30%, an excellent value of relative humidity of the dilution gas being between 0 and 20%. Preferably, a dilution gas is chosen, for example CO2 or H2S, which has an effect of increasing the saturation vapor pressure of the water when it is mixed with the initial gas with respect to the vapor pressure. saturating in water of the initial gas. The pressure of the dilution gas arriving via the duct (200) may be between the value P of the pressure of the humid gas arriving via the duct (100) and the value P + 0.5 MPa. The gas mixture is sent through the conduit (101) into the heavy metal guard bed (2000). The depleted mercury gas is evacuated from the heavy metal guard bed (2000) through the conduit (102). The treated gas (102) is sent to the unit (3000) whose function is to separate the dilution gas from the wet gas. The operation of the unit (3000) depends on the nature of the dilution gas. As a result of the treatment by the unit (3000): an undiluted heavy metal depleted gas effluent discharged through the conduit (103), a flow of dilution gas which is recycled via the conduit (200), where appropriate excess of water separated from the diluent gas discharged via the pipe (300) and, where appropriate, an effluent (201) composed mainly of recycle gas in a proportion corresponding to the diluent gas that would already be originally contained in the wet gas to be treated (100).

According to a first embodiment, it is possible to use a dilution gas that is condensable. If the dilution gas is a condensable gas, the principle of the unit (3000) is to condense at least a portion of the dilution gas, optionally to dry it and then return it via the conduit (200) to be mixed with the wet gas to be treated (100). The condensation step can be done, for example by lowering the temperature or increasing the pressure. According to a second embodiment, the unit (3000) can implement a membrane having the ability to selectively separate the dilution gas from natural gas. The separated dilution gas is raised in pressure, for example by compression, and then recycled (200). This embodiment may use 002 or H2S as the diluent gas. Preferably, these gases are already present in the gas to be treated and the dilution gas is produced by means of a step of at least partial recycling of one or more of these gases around the capture mass.

According to a third embodiment, it is possible to choose the dilution gas already contained in the wet gas, in particular in natural gas, for example CO2 or H2S. In this case, the unit (3000) can be composed of an absorber and a regenerator between which a liquid absorbent having the property of preferentially absorbing the diluent gas flows. This absorbent is preferably an aqueous solution of an organic compound having at least one amine function. The dilution gas obtained at the regenerator is recycled via the conduit (200). According to a fourth embodiment, it is possible to choose a portion of the wet gas depleted of heavy metal and depleted of water as a dilution gas. In this case, the unit (3000) may be composed of an element having the ability to deplete the wet gas in water. This element can be, for example an absorber and a regenerator between which circulates a liquid absorbent, preferably glycol, having the property of absorbing water, or a solid adsorbing water. Part of the depleted heavy metal gas depleted of water is recycled through the conduit (200).

The guard bed (2000) is composed of capture masses that have the capacity to adsorb the heavy metal. It is possible to use any type of capture masses known to those skilled in the art. The capture mass used may be chosen from those known to those skilled in the art. The capture mass comprises a compound, commonly called the active phase, which reacts with the heavy metal so as to capture the heavy metal on the capture mass. The capture mass is preferably in the form of a bed composed of granules. The active phase of the capture mass may include, in particular for capturing mercury, metals which, in their sulfur form, react with mercury. The metal sulphide or sulphides contained in the capture mass according to the invention are based on a metal chosen from the group consisting of copper (Cu), chromium (Cr), manganese (Mn), iron (Fe ), cobalt (Co) and nickel (Ni). Preferably, the metal (s) of the metal sulphide (s) are chosen from the group consisting of copper (Cu), manganese (Mn), iron (Fe) and nickel (Ni). Very preferably, if a single metal sulfide is present, the copper sulfide is selected. The active phase used may also be elemental sulfur, for example as described in patent document FR 2 529 802.

The mercury guard bed being arranged upstream of the deacidification units (for example an amine unit), the gas to be treated contains H2S. It is therefore also possible to use the metal oxide precursor, the metal being selected from the group consisting of copper (Cu), chromium (Cr), manganese (Mn), iron (Fe), cobalt (Co) ) and nickel (Ni), preferably copper oxide, which concomitantly eliminates H2S and mercury being in a first step sulphurized by H2S in metal sulphide and then reacting with mercury . The active phase of the capture mass can comprise, in particular for capturing arsenic, and in particular in its gaseous form AsH3, metals which, in their oxide form, react with arsenic. The metal oxide (s) contained in the capture mass according to the invention are based on a metal chosen from the group consisting of copper (Cu) and lead (Pb). Preferably, if only one metal oxide is present, the choice is made on the copper oxide (II) also called cupric copper oxide (Cu0). Very preferably, if only one metal oxide is present, the choice is made on lead oxide (II) (PbO).

Preferably, the capture mass may consist of an active phase, as described above, distributed over a porous support. The porous support may be chosen preferably from aluminas, phosphorus aluminas, silica-aluminas, silicas, clays, activated carbons, zeolites, titanium oxides, zirconium oxides, and mixtures of these. this.

A capture mass containing a support and copper sulphide is described, for example, in US Pat. No. 4,094,777. The capture mass can be obtained by any preparation route known to a person skilled in the art, for example impregnation or coagulation. -granulation. The treatment of the treated gas with the capture mass (5000) is preferably carried out by injecting the effluent to be treated into a reactor containing the capture mass in the form of a fixed bed. In the equipment (2000), the contacting of the effluent to be treated with the capture mass can be carried out at a temperature between -50 ° C and 115 ° C, preferably between 0 ° C and 110 ° C, and more preferably between 20 ° C and 100 ° C. In addition, it can be carried out at an absolute pressure of between 0.01 MPa (0.1 bar) and 20 MPa (200 bar), preferably between 0.1 MPa (1 bar) and 15 MPa (150 bar), and more preferably between 0.1 MPa (1 bar) and 12 MPa (120 bar).

In addition, this step of contacting the effluent to be treated with the capture mass can be carried out with a V.V.H. between 0.1 h-1 and 50000 h-1. "V.V.H. The hourly Volumic Velocity of the gaseous effluent in the capture mass, that is to say the volume of the gaseous effluent per volume of reactor and per hour. For an effluent to be treated gas, V.V.H. may be preferably between 50 h-1 and 500 h-1. The contact with the capture mass advantageously makes it possible to capture the heavy metals, in particular mercury, contained in the effluent to be treated, and to obtain an effluent having a content of heavy metals, in particular mercury, reduced in relation to the content of the initial effluent, or even completely eliminate heavy metals from the effluent.

Advantageously, the reduction in the total content by weight of heavy metal between the gaseous effluent before treatment and the effluent obtained after treatment with the capture mass can represent at least 90%, preferably at least 95%, and more preferably at least 99%.

FIG. 2 diagrammatically represents an example of a process for producing natural gas from the extraction to the deacidification operation, in which the method according to the invention is implemented, according to the second embodiment mentioned above. . The references of Figure 2 identical to the references of Figure 1 designate the same elements.

The natural gas extracted from the underground deposit (1) is sent via the conduit (2) into a device (3) for trapping sludge, commonly called "slug catcher". The sludge is evacuated from the "slug catcher" by the conduit (4). The natural gas from the device (3) is introduced through the conduit (5) into a primary liquid gas separator (6) which makes it possible to eliminate a liquid flow comprising water and hydrocarbons via the conduit (7).

The wet gas discharged from the separator (6) is then introduced via the conduit (110) into the liquid gas separation equipment (4000). The function of this equipment (4000) is to separate the wet gas from the droplets of liquid resulting from the saturation of the condensables at the temperature of the system. This equipment (4000) is for example a coalescer. EP2473250A2 describes an embodiment of a coalescer. The coalescer (4000) may be composed of an enclosure provided with internal elements such as baffles and or fiber mattresses that promote the condensation of the liquid droplets contained in the gas. The elimination of the droplets of liquid is preferable because it avoids ending up in a system where there is a saturated gas which leads in addition to condensed liquids which will then settle on the mercury guard bed. The condensates are evacuated from the coalescer (4000) through the conduit (111), while the depleted gas is evacuated through the conduit (100). The gas flowing in the conduit (100) is then mixed in the equipment (1000) with a dilution gas arriving via the conduit (200). The dilution gas may be 002, or a dried gas having a composition similar to that of the gas flowing through the conduit (100). The gas mixture discharged from the equipment (1000) through the conduit (101) is introduced into the mercury guard bed (2000) to produce a mercury depleted gas discharged through the conduit (102).

The mercury-depleted gas is introduced via the conduit (102) into the unit (3000) comprising a reactor equipped with a membrane. The membrane is chosen to be more permeable to the dilution gas than to natural gas. The permeate, that is to say the flow having passed through the membrane is discharged through the conduit (200), then compressed (not shown) and recycled to the unit (1000). For example, a pressure difference between the two faces of the membrane can be applied to migrate the CO2 through the membrane. For example, the pressure difference may be between 0.2 and 3 MPa, preferably between 0.5 and 2.5 MPa. A membrane made of a material selected from cellulose acetate, polymides, polysulfones and polaramides can be used. Optionally, a dilution gas refill is performed via the conduit (202). This embodiment can use CO2 or H2S as a dilution gas. Preferably, these gases are already present in the gas to be treated and the dilution gas is produced by performing a step of recycling at least a portion of the dilution gas around the capture mass. The retentate, that is to say the fluid that has not passed through the membrane is removed from the unit (3000) and sent through the conduit (103) in the deacidification unit (8). The pressure of the flow flowing in the duct (200) can be increased by using a compressor.

In the unit (8), the gas is contacted with a liquid solution of amine which absorbs the H2S and optionally the remaining CO2 contained in the gas. The gas depleted of acidic compounds is introduced via line (9) into the dehydration unit (10). In the unit (10), the natural gas can be contacted with a liquid solution of glycol that absorbs the water contained in the gas. Alternatively, in the unit (10), the natural gas can be contacted with solid masses adsorbing water. The gas discharged from the unit (10) via the conduit (11) can be marketed. FIG. 3 schematically represents an example of a process for producing natural gas from the extraction to the deacidification operation, in which the method according to the invention is implemented, according to the third embodiment mentioned above. . The references in FIG. 3 identical to the references in FIG. 2 denote the same elements. With reference to FIG. 3, the natural gas from the underground deposit (1) is freed from production sludge in the separator (3), then liquid phases in the primary separator (3) and droplets of liquid in the coalescer ( 4000). Then, the water-depleted natural gas (100) is mixed with the dilution gas flowing through the conduit (200) into the equipment (1000). The gas mixture obtained is introduced into the mercury storage bed (2000), and then the mercury-depleted gas is introduced into the deacidification unit (8). In the unit (8), the gas is brought into contact with a liquid solution of amine, arriving via the conduit (81), which absorbs the H2S and optionally the CO2 contained in the gas. The absorbent solution is then introduced via the conduit (82) into a regenerator (80) which is maintained at temperature by the reboiler (83). In the regenerator, the acidic compounds, in particular CO2 and H2S, are released and are discharged at the top of the regenerator. According to the third embodiment of the invention, these acidic compounds obtained at the top of the regenerator are discharged through the conduit (200) to be recycled in the equipment (1000). The regenerated absorbent solution obtained at the bottom of the regenerator (80) is introduced via the conduit (81) into the unit (8).

The gas depleted of acidic compounds is removed from the unit (8) via the conduit (9) and is introduced into the dehydration unit (10). The gas discharged from the unit (10) via the conduit (11) can be marketed.

FIG. 4 schematically represents an example of a process for producing natural gas from the extraction to the deacidification operation, in which the method according to the invention is implemented, according to the fourth embodiment mentioned above. . The references of FIG. 4 identical to the references of FIG. 2 denote the same elements. With reference to FIG. 4, natural gas from the underground deposit (1) is freed from production sludge in the separator 3, then liquid phases in the primary separator (3) and droplets of liquid in the coalescer (4000). . Then the water depleted natural gas (100) is mixed with the dilution gas flowing through the conduit (200) into the equipment (1000). The gas mixture obtained is introduced into the mercury storage bed (2000), and then the mercury-depleted gas is introduced into the deacidification unit (8). In the unit (8), the gas is brought into contact with a liquid solution of amine which absorbs the H2S and optionally the CO2 contained in the gas. The acid-depleted gas is introduced through line (9) into the dehydration unit (10). In the unit (10), the natural gas can be contacted with a liquid solution of glycol that absorbs the water contained in the gas. Alternatively, in the unit (10), the natural gas can be contacted with solid masses adsorbing water. The gas, which can be described as dry natural gas, discharged from the unit (10) through the conduit (11) has a reduced water content compatible with the specifications established for marketing the natural gas. For example, the water content in the natural gas flowing in the duct (11) may be between 0 and 200 ppm. According to the fourth embodiment of the invention, a portion of the dry natural gas is withdrawn and the withdrawn portion is sent via the conduit (200) into the mixing unit (1000) to produce a gas mixture (101) having a relative humidity lowered relative to the gas arriving through the conduit (100) in the unit (1000). The process according to the invention is particularly well suited to the treatment of natural gas. Nevertheless, the process according to the invention can be applied to other types of gaseous effluents such as combustion fumes, synthesis gas or else a gaseous hydrocarbon effluent, for example a shale gas, a gas coal, a gas oil cut or a refinery fuel oil, chlorine alkali plant vents, rare earth production plant vents. The gaseous effluent contains water in variable proportion which depends in particular on its origin and the previous treatments that it has been able to undergo. Typically the relative humidity of the fumes is between 30% and 100%. The combustion fumes are produced in particular by the combustion of hydrocarbons, biogas, coal in a boiler or by a combustion gas turbine, for example for the purpose of producing electricity. These fumes have a temperature of between 20 and 60 ° C., a pressure of between 1 and 5 bar (1 bar = 0.1 MPa) and can comprise between 50 and 80% of nitrogen, between 5 and 40% of carbon, between 1 and 20% oxygen, and some impurities such as SOx and NOx, if they have not been removed upstream by a deacidification process. The synthesis gas contains carbon monoxide CO, hydrogen H 2 (generally in a H 2 / CO ratio equal to 2), water vapor (generally at saturation at the temperature at which washing is performed) and carbon dioxide CO2 (of the order of ten percent). The pressure is generally between 20 and 30 bar, but can reach up to 70 bar. It contains, in addition, sulfur impurities (H2S, COS, etc.), nitrogen (NH3, HCN) and halogenated impurities. Depending on the positioning of the demercurization step in the purification chain of the synthesis gas, its relative humidity is between 5% and 100%. The examples presented below make it possible to highlight the advantages of the invention. Example 1 (Context) In this example, two commercial mesoporous solids A and B used for demercurization and characterized by the porous volume distributions shown in Table 1 are considered. Table 1 - Porosity Type Proportion of Solids A and B studied% Volume porous Solid A Solid B Micropores (d <2nm) 0 0 Mesopores (2 <d <50 nm) 49 60 Macropores (d> 50 nm) 51 40 d corresponding to the pore diameter of the solid, measured by a mercury porosimetry, for example the method described in the document [Rouquerol F .; Rouerai J .; Singh K. Adsorption by powers & porous solids: Principle, methodology and 5 applications, Academic Press, 1999] According to the hypothesis of cylindrical pores, one can then apply the Kelvin law (equation 2) to determine, according to the degree of hygrometry, the proportion of the total pore volume filled with liquid water. It is placed under the conditions of a perfectly wetting fluid on the surface (0 = 0), the surface tension is equal to that of the air / water system 10 (y = 67.9 mNm -1 at T = 50 ° C. ) and the temperature T = 50 ° C. The saturation vapor pressure of the water Po is determined by the law of Antoine (PO = 105.20389-1733.926 / (T-39.485)). The result of this calculation is shown in Table 2. Table 2 - Percentage of pore volume filled as a function of the degree of hygrometry (3% of the total pore volume filled in the hygrometry range) Degree of hygrometry Solid A Solid B 0 <W / Po <25 0 0 25 <W / W <50 0.02 0.01 50 <W / W <70 1.55 1.27 70 <W / W <80 6.52 8.01 80 <W / W <90 30.72 42.40 90 <W / Po < 99.9 57.62 45.55 It is clearly observed that the majority of the pores are not affected below hygrometry levels of about 80%, after which there is a sharp increase in the proportion of the total pore volume filled with liquid water. up to almost 100% for hygrometry rates close to saturation According to the assumption of a homogeneous distribution of the active phase on the surface of the support, it is possible to calculate the amount of active phase rendered inactive. because it is inaccessible due to the filling of the pores as a function of the hygrometry rate (3% of active phase inaccessible in s the hygrometry range Degree of hygrometry Solid A Solid B 0 <P / Po <25 0 0 25 <P / P <50 0 1.10 50 <P / P <70 7.06 5.08 70 <P / P <80 21.29 23.03 80 < P / Po <90 46.17 58.75 90 <P / Po <99.9 25.48 12.04 In accordance with these data, a dynamic mercury uptake test conducted on 1 g of the capture mass A showed that the purification efficiency was 100% at room temperature. after one hour for a hygrometry rate of 25%; 100% after one hour for a hygrometry rate of 50%; 90% after one hour for a hygrometry rate of 75%; 25% after one hour for a hygrometry rate of 90% and 0% after one hour for a hygrometry rate of 100%. In agreement with these data, a dynamic mercury uptake test conducted on 1 g of the capture mass B showed that the purification efficiency was 100% after one hour for a humidity level of 25%; 100% after one hour for a hygrometry rate of 50%; 90% after one hour for a hygrometry rate of 75%; 15% after one hour for a hygrometry rate of 90% and 0% after one hour for a hygrometry rate of 100%.

The operation of the capture masses will therefore only be moderately impaired until humidity levels of 80%. A fortiori, in order to have a safety margin, it is recommended to maintain the relative hydrometry of the gas to be treated in the demercurization step at a relative humidity of less than 75%. Example 2 (not in accordance with the This example is given for comparative purposes. It aims to evaluate the amount of energy required to reach 75% relative humidity for the gas saturated with water under the conditions of Example 1 by simply heating the gas flow.

The following initial conditions are used: the wet gas is a natural gas saturated with water, produced at 70 bar and 30 ° C for a dry flow rate of 140 000 Nm3 / h. Its dry composition is given in Table 3 below: Table 3 - Composition of dry gas% mol CO2 10.7 02 0.8 N2 0.9 Cl 77.2 C2 6.6 C3 2.2 iC4 0.2 nC4 0.5 iC5 0.2 nC5 0.3 C6 0.2 C7 0.1 C8 0.1 When this gas is saturated with water, its water vapor content is 880 ppm mol. The results presented below result from a simulation under the Aspen Hysys v7.2 software. The thermodynamic model selected is SRK.

The temperature at which a relative humidity of 75% is reached is 35.5 ° C. The thermal power required to pass the gas flow from 30 ° C to 35.5 ° C is 500 kWth. Then the heated gas is treated on a mercury guard bed.

Example 3 (According to the Invention) This example according to the invention proposes to lower the water content of a wet gas below 75% according to what is set forth in Examples 1 and 2 in order to prevent any phenomenon solid mass of capillary condensation in the mercury capture masses as described in Example 1. The method of FIG. 2 is implemented. The results presented below result from a simulation under the Aspen Hysys v7.2 software. The thermodynamic model selected is SRK (Soave-Redlich-Kwong). The following initial conditions are used: the wet gas (100) is a natural gas saturated with water, produced at 70 bar and 30 ° C for a dry flow rate of 140 000 Nm3 / h. Its dry composition is given in Table 3 of Example 2. The molar flow of wet gas is 6252 kmol / h. When this gas is saturated with water, its water vapor content is 880 ppm mol. The diluent gas flowing through the conduit (200) is carbon dioxide. it is brought to a molar flow of 1100 kmol / h. Mixing the diluent gas (200) with the wet gas (100) produces a resulting gas (101) having a relative humidity of 74.74%. The ratio between the molar flow of the gas flowing in the duct (100) and the molar flow of the gas flowing in the duct (200) is 5.7. The gas mixture (101) is then introduced into the capture mass to be depleted of mercury. Example 4 (according to the invention) This example according to the invention proposes to lower the water content of a wet gas below 75% according to what is set forth in Examples 1 and 2 in order to prevent any massive phenomena of capillary condensation in the mercury capture masses as described in Example 1. The method of FIG. 4 is implemented. The results presented below result from a simulation under the Aspen Hysys v7 software. 2. The thermodynamic model selected is SRK (Soave-Redlich-Kwong).

The following initial conditions are used: the wet gas (100) is a natural gas saturated with water, produced at 70 bar and 30 ° C for a dry flow rate of 140 000 Nm3 / h. Its dry composition is given in Table 3 of Example 2. The molar flow of wet gas is 6252 kmol / h.

When this gas is saturated with water, its water vapor content is 880 ppm mol. The dilution gas (200) corresponds to a fraction of the exported gas obtained after the gas treatment units (8) and (9). Thus, the dilution gas (200) has the same composition as the gas (100) given in Table 3, except that this gas is dry.

A molar flow of 2110 kmol / h is withdrawn and sent via the conduit (200) to the mixing unit (1000). The ratio between the molar flow of the gas flowing in the duct (100) and the molar flow of the gas flowing in the duct (200) is 3. The mixture of the dilution gas (200) with the humid gas (100) produces a resulting gas (101) having a relative humidity of 74.75%.

Conclusions on Examples 3 and 4: The comparison of these two examples shows the comparative advantage of using a diluent gas which will have the effect of increasing the saturating vapor pressure as in Example 3 rather than a gas which has only a diluting effect as in Example 4. This reduces by nearly half the flow of dilution gas to be used to lower the relative humidity. Example 5 (according to the invention) This example according to the invention proposes to lower the water content of a wet gas below 75% according to what is set out in Examples 1 and 2 in order to prevent any phenomena Massive capillary condensation in the mercury capture masses as described in Example 1. The method of FIG. 2 is implemented. The results presented below result from a simulation under the Aspen 30 Hysys v7.2 software. . The thermodynamic model selected is SRK (Soave-Redlich-Kwong). The following initial conditions are used: the wet gas (100) is a natural gas saturated with water, produced at 70 bar and 30 ° C for a dry flow rate of 140 000 Nm3 / h. Its dry composition is given in Table 3 of Example 2. The molar flow of wet gas is 6252 kmol / h. When this gas is saturated with water, its water vapor content is 880 ppm mol.

The dilution gas (200) corresponds to a partial recycle step of the CO2 contained in the gas initially. This recycling step is done by means of a membrane module (3000) having the ability to preferentially separate CO2 and a compressor intended to compensate for the pressure drop on the CO2 produced here estimated at 20 bar.

The membrane module (3000) makes it possible to separate 54.6% of the CO2 contained in the flow (102). The CO2 / CH4 ratio in the stream (200) is adjusted to 4 to reflect the non-ideality of the operation of the membrane. The action of this module (3000) has the effect of stabilizing the concentration of CO2 in the flow (101) at 20.4 mol% against 10.7 mol% initially and 69.2 mol% in CH4 against 77.2 mol. % mol initially. Mixing the diluent gas (200) with the wet gas (100) produces a resulting gas (101) having the following composition given in Table 4: Table 4 - Composition of the gas mixture Compound% mol CO2 20.42 02 0 , 7 N2 0.7 Cl 77.2 C2 5.7 C3 1.9 iC4 0.2 nC4 0.4 iC5 0.2 nC5 0.3 C6 0.2 C7 0.1 C8 0.1 H2O 0.1 Relative humidity of the resulting gas (101) is 74.35. The ratio between the molar flow of the gas flowing in the duct (100) and the molar flow of the gas flowing in the duct (200) is 6 with 6252 and 1021 kmol / h, respectively.

The power of the 75% efficiency compressor to raise the pressure of the recycled gas by 20 bar is 230 kW electric.

Claims (16)

CLAIMS1) A method for capturing at least one heavy metal, selected from mercury and arsenic, contained in a wet gas comprising water vapor, the wet gas having a relative humidity RH greater than 75%, in which one carries out the following steps: a) the wet gas is mixed (1000) with a dilution gas so as to obtain a gas mixture having a relative humidity HM lower than the relative humidity HR, b) contacting (2000) the mixing gas with a capture mass of said heavy metal to obtain a mixture of heavy metal depleted gas. 2) Process according to claim 1, wherein the dilution gas is selected from CO2, H2S and a natural gas having a relative humidity of less than 50%. 3) Method according to one of claims 1 and 2, wherein in step a), the flow rate of the dilution gas is determined so as to obtain a gas mixture having a relative humidity HM less than the RH relative humidity . 4) Method according to one of claims 1 to 3, wherein the wet gas is mixed with a dilution gas so as to obtain a gas mixture having a relative humidity HM of less than 90%, preferably less than 80%, see less than 75%. 5) Method according to one of claims 1 to 4, wherein the capture mass comprises an active phase selected from at least one metal sulfide based on a metal selected from the group consisting of copper (Cu), chromium (Cr), manganese (Mn), iron (Fe), cobalt (Co) and nickel (Ni). 6) Method according to one of claims 1 to 4, wherein the capture mass comprises an active phase composed of elemental sulfur. 7) Method according to one of claims 1 to 4, wherein the capture mass comprises an active phase composed of at least one metal oxide of a metal selected from copper (Cu) and lead (Pb). 8) Method according to one of claims 1 to 4, wherein when the wet gas contains H2S, the capture mass comprises at least one metal oxide precursor of a metal selected from copper (Cu), the chromium (Cr), manganese (Mn), iron (Fe), cobalt (Co) and nickel (Ni). 9) Method according to one of claims 5 to 8, wherein the active phase is distributed on a porous support, the porous support being selected from the group consisting of aluminas, phosphorus aluminas, silica-aluminas, silicas, clays, activated carbons, zeolites, titanium oxides, zirconium oxides, and mixtures thereof. 10) Method according to one of claims 1 to 9, wherein after step b), the following steps are performed: c) at least a portion of the H2S and CO2 contained in the depleted gas mixture is removed; of heavy metal, so as to obtain a mixture of H2S and CO2 depleted gas and a flow rich in H2S and CO2, then d) remove at least a portion of the water vapor contained in the depleted gas mixture in H2S, in CO2 so as to obtain a gas depleted in water. 11) Method according to one of the preceding claims, wherein the dilution gas is CO2, and wherein after step b) and before step c), is implemented a membrane separator to effect a separation of at least a portion of the CO2 contained in the heavy metal depleted gas mixture so as to obtain a CO2-rich stream and in that at least a portion of said CO2-rich stream is recycled to step a) as a as dilution gas. The method of claim 10, wherein at least a portion of said H2S and CO2 rich stream is recycled to step a) as a diluent gas. 13) The method of claim 10, wherein in step a) is recycled a portion of said water-depleted gas as a diluent gas. 14) Method according to one of the preceding claims, wherein the wet gas is selected from one of the following gases: a natural gas, a shale gas, a coal gas, a synthesis gas, combustion fumes, gaseous hydrocarbon effluent, chlorine alkali plant vents, rare earth production plant vents. 15) Method according to one of the preceding claim, wherein the wet gas is composed of a natural gas and wherein before performing step a), the following steps are carried out: - a natural gas is extracted from an underground deposit, then - the natural gas is introduced into a sludge removal device, then - the natural gas is introduced into a liquid gas separation device, then the natural gas is sent to step a). The method of claim 15, wherein said liquid gas separation device comprises a coalescer for separating the condensed water from the natural gas.