OA10872A - Conversion of hydrocarbons assisted by gliding electric arcs - Google Patents

Description

1

CONVERSION OF HYDROCARBONS ASSISTED BYGLIDING ELECTRIC ARCS IN THE PRESENCE OFWATER VAPOR AND/OR CARBON DIOXIDE 010872

5 RELATED APPLICATIONS

This application is based on a French patent application, National Registration No.9700364, fîled January 13, 1997.

BACKGROUND OF THE INVENTION 10

This invention concerne a hydrocarbon conversion process assisted by spécial gliding arc plasma in the presence of carbon dioxide (CO2) and/or water vapor. This process is illustrated bythe conversion of two model mixtures in an arc reactor equipped with a maturation post-plasmacompartment: a natural gas containing mainly methane and some ethane, propane and butanes, 15 - a "propane" containing some ethane and butanes.

Thersfore, the invention can be applied to any pure hydrocarbon, such as CH4, C2H,„ C3HB orC4H1? and to their mixtures.

In the presence of water vapor and/or of CO2, it is then possible to couvert, totally orpartially, ail these hydrocarbons basically into synthesis gas (consisting of a majority of hydrogen 20 H, and of carbon monoxide CO), but also into other valuable products, such as ethylene (C2H4), acety'.ene (C2H) andpropylene (C3H6), and ail without using traditional catalysts. The process isbasée mainly on steam reforming reactions, such as: CH.+ H3Ovap= CO + 3 H2 25 CH,+2 H2Ovap= 2 CO + 5 H2C5Hb+ 3 H2Ovap= 3 CO + 7 H2C4K10 + 4 H2Ovap = 4 CO + 9 H2 G) (2) (3) (4) reforming reactions with carbon dioxide, such as: CH++ CO2 = 2 CO + 2H2CJîs + 2C02 = 4C0 + 3H2CA + 3 CO2 = 6 CO + 4 H2CA3+4CO2 = 8CO + 5H2 (5) (6) (7) (8) cracking reactions, such as: 2 CH4 = C2H4 + 2 H22 CK4 = C2H2 + 3 H2CHtt = C2H4 + H2ca = CA + 2 h2ca = ca + h2CA.,=2C2H4+H2CAtf=2C2H2 + 3H2 (9) (10)(H)(12) (13) (14) (15) 010872 as well as single and inverse water shift: CO + H2O = CO2 + H2 (16) CO2 + H2 = CO + H2O (17)

Ail these reactions are performed in a medium highly activated by the presence of a5 spécial plasma produced by the gliding electric arcs. The activation of the medium is évident by the presence of rather unusual species (with respect to the traditional hydrocarbon conversionconditions) originating from the matter in which these arcs arc developed. Thus, électrons can becetected, as well as atoms, ions and/or molecular radicals such as: 02, H+, O+, 0/, O2', HO2,CH3, CH2, CH, C2 and many others. Most of these species can exist in their excited clectronic or 1 a vibrational States with a very long lifetime. They are also known as being extremely active càemically.

The production of synthesis gas starting from light saturated hydrocarbons is a very well-known and very important stage, especially for the upgrading of natural gases. The most usedprocess at the présent time, the catalytic steam reforming (or "steam reforming”) encounters 15 major problems. In principle, a high température (thermodynamic ratio) and a high pressure (forkinetic ratios) are sufficient for this process. However, in practice, despite the know-how for theproduction of "synthesis gas" according to the processes, the joint management of thecompositions, pressures and températures is délicate, even impossible without resorting tocacalysts. 2 û Then, in order to perform natural gas (mainly rich in methane) reforming with water vapor, usually a catalytic way is sought: presence of an active solid substance for températureswhich can be attained without too much difficulty. Therefore, the traditional steam reformingtecfcnologv uses fumaces in which several hundred fragile métal tubes (filled with a catalyst andhavmg a length which can reach several dozen meters) are located, heated with natural gas. This 2 5 technology is tied to the very strong drops in pressure and, especially, in heating energy. The température which the fumace pipes can withstand prevents also the réduction of CO2 content(awkward product originating from a parasite reaction at too low of a température).

Other problems are connected with catalyst poisoning (by sulfur and/or nitrogen), withcatalyst agïng, with the necessary excess of water vapor and/or with the formation of soot which 3C blocics the cubular System at a macroscopie scale and, most of ail, the microscopie pores of thecatalyst. These problems are observed particularly with steam reforming of hydrocarbons heavierthan methane; they are more fragile and, hence, more coking.

The conversion of hydrocarbons according to the endothermie reactions (1) through (15)requires a supply of energy (preferably "clean’’), without connection with any internai or extemal 3 5 combustion- The best way to promote these reactions would be to strike electric arcs dircctly in die medium io be converted, imposing a permanent distribution of energy in the largest volume tobe treaced. The transfer of energy of electric origin to the gas mixture would be made by directcransfer of the energy to the molécules. This would resuit in excitation, ionization anddissociation phenomena and also in part by Joule effect, considering the ionized mixture as a 5 010872 gaseous conductor. This is to say that the gaseous mixture, wiûch bas been made inlo aconductor after ionization, itself due to dielectric breakdown (hence, a preionization) betweenélectrodes brought to different potentials, would be considered as an electric résistance and, at thesame time, as a sort of electrolyte in gaseous phase: the plasma. 5 Plasma is defined as the fourth State of matter and, thcrefore, cannot in any case be taken as a criterion of similitude for previously known different processes. Wanting to claim theconcept of plasma or any type of reaction capable of developing at the plasma state, cornes towanting to claim ail the reactions developing at the liquid state.... There are onc thousand onetypes of plasma, and one thousand one ways to obtain these plasmas. By définition (simplified), 1 ü plasma is a gaseous medium in which the particles are in part ionized. Likewise, a part of the électrons is not associated with an atom, a molécule, an ion or a radical. Thus, even tlioughglobally, at a scale of a few microns, the medium is electrically neutral, two large families can bedefined, in a simplistic way: the heavy particles (radicals, atoms, molécules and ions) and theélectron cloud. 15 In most plasmas, the main macroscopie physical parameter - température - is the same for ail the components: this is thermodynamic equilibrium. These conditions can be very easilyobtained: it is sufïïcient to supply much energy, as in the case of plasma torches (plasmatrons, forsome), where the plasma is produced by a very high electric arc curent. There are also otherdevices capable of generating this state, such as, for example, induction or radiofrequency torches 2 C whereby the gaseous medium becomes résonant with an electric circuit. Such plasmas are called thermal plasmas by the experts. It is obvious that a thermal plasma will modify the chemistry of agas medium, simply by destroying ail the molécules, particularly the fragile ones, such as thehydrocarbons. The fragments found at the end of the process originate from partial rccombinationphenomena, often yielding too simple molécules. Such chemistry offers very poor prospects, 25 requires much energy and présents problems connected with the high température (such as therésistance of the materials).

Professional chemists indisputably prefer the idea of a plasma which does not respect theconditions of complété thermodynamic equilibrium. For example, it is sufficient to act on the freeélectrons by taking advantage of the fact that they are much lighter. It is also possible to act on 2- o the rotation or vibration properties of some molécules. In terms of energy, this cornes to breakingthe energy exchange equilibrium between the plasma and the surrounding medium (heat,electrical energy, radiation, etc.). This state is qualified as non-equilibrium. Such plasmas areeften called '"low température" plasmas, although the concept of température cannot be used:there are several methods whereby such plasmas may be generated: microwaves, électron beams, • 5 Came front, etc. However, the generators of these plasmas are rare on an industrial scale and areappropriate only for a very précisé application. This is the reason why, despite the great nuniberof patents, such plasmas are rarely used in chemistry.

Also, when a plasma is established or when its existence is ended, the equilibrium isbrokers These transitory States are actually non-equilibrium plasmas and last only a few 010872 milliseconds. One type of plasma takes advantage of this phenomenon, the gliding electric arcplasma, known under the name of "GlidArc", a relatively recent invention (1988) by H.LESUEUR et al. [”Low Température Plasma Génération Device through the Formation ofGliding Electric Discharges", BF 2,639,172]. Outside of the numerous géométrie possibilities of 5 a GlidArc plasma generator, and in a very global way, the parameters on which a chemist can actare: pressure, température, gas speed, current, electrical frequency and voltage. Such a number ofparameters exceeds the conventional reasoning capabilities of the man of the trade. For eachapplication, a real know-how and an inventive activity are necessary in order to obtain a resuit theobjectives of which are both the économie profitability and tire respect ofthe ecological L o principles. The approach allowed by the GlidArc enables the chemist to envision the distributionof a supply of energy directly in the gaseous mixture without, for example, resorting to catalysts.The chemist can also (to a certain extent) distribute directly the energy either in thermal form orin Chemical form. He can also intervene on the flux still loaded with active species leaving thegliding arc zone, to hâve these species reach with the load to be converted in a maturation post- 15 plasma zone.

Our bibliographie research conceming the last three décades yields few published and/orpatented results conceming the partially oxidizing conversion of saturated hydrocarbons assistedby plasma. This may be due to the problems connected with tire presence of oxygen originatingfrom dissociation of the H2O and/or CO2 molécules and attacking the traditional tungsten or 20 graphite électrodes of classic plasma devices. Nevertheless, we report these attempts to usedifferent sources of plasma. Systematically, both the approach and the reaction process aredifferent from ours. They hâve only point in common: the use of the word "plasma" or thepossibilitv of treating the same hydrocarbon molécules. K. KARL et al. [“Verfahren zur Herstellung eines wasserstoflreichen Gases aus 25 Kohlenwasserstoffen”, CH 378,296 (1957)] proposed hydrocarbon steam reforming under 66.7kPa - 0.3 MPa pressure, in a "silent" discharge characterized by a 0.3-0.5 MV/m very intenseelectric field. This source of plasma has been known for a century and is totally different fromthat ofthe invention. R J. HEASON presented, in 1964, his doctorate thesis conceming methane pyrolysis and 30 the reaction of CH4 with water vapor in an arc plasma (700 A, 20 V) in argon. These results arepublished only in manuscript form ["Investigation of methane and methane-steam reactions in anargon plasma", Dissertation, Ohio State Univ., Columbus], A "thermal" plasma and a deviceconsuming a great quantity of argon (2 moles Ar for 1 mole CH4) are involved. C.H. LEIGH and E.A. DANCY ["Study ofthe reforming of natural gas by a plasma arc", 3 5 Proc, cf the Int. Round Table on Study and Appl. of Transport Phenomena in Thermal Plasmas,contribution 1.5, Odeillo, 1975,11 pages] heated a mixture of CH4/CO2 ~ 1 in a jet of argonplasma produced by a traditional arc plasma torch. The jet température was approximatcly 10kK. The argon flux was of the same order of magnitude as that ofthe mixture to be treated. Theseresearchers observed a 11-74% conversion of carbon to H2, CO, C2H4 and C2H4 (without having

010872 ever detected C2H2 or H2O in the products?). No application was possible because of the highconsumption of electrical energy (70% of it passed in the plasma torch cooling water) and ofnoble gas.

Also P. CAPEZZUTO et al. ["The oxidation of methane with carbon dioxide, water vaporand oxygen in radio-frequency discharges at moderate pressures", 3rd Int. Symp. of PlasmaChemistry, Limoges, 1976, contribution G.5.11, 7 pages] studied partial oxidation of methaneplaced separately in mixture with CO2, either with O2 or with H2O, with the ratios CH4/oxidizer =1. The 35 MHz radioffequency (RF) plasma reactor needed an additional argon flux and couldonly work at low pressures of approximately 2.7 kPa. For a 3 to 36 l(n)/min total flow of enteringgas, the energy density varied from 1 to 12 kWh/m3. No industrial use was possible bccause ofthe high consumption of electrical energy and of noble gas (in addition to the complexity of theelectrical supply and the requirement to work under vacuum). The mechanical setup constraints,the low energy yield and the insufficient unit powers of the sources of RF plasma make the use ofthis method economically poorly suited for the transformation of major volumes of gas.

However, it is interesting to note that, in ail the cases, the authors observe an almost totalconversion of the methane and an appearance of the following products:

For the CH4/CO2 Systems, mostly H2, CO, C2H2, with presence of C2H4 (<5%) and ofC2H,(< 1%).

For the CH4/H2O System, the same as above but with a few traces of CO2. A patent by S. SANTEN et al. ["Thermal reforming of gaseous hydrocarbon" GB-A-2172,011] of 1986, daims the use of a plasma generator to heat reagents (a gaseous hydrocarbon,some water vapor and, possibly, some coal), completely or partially, up to a températureexceeding 1200°C. At such températures, these inventors expect favorable conditions to carry outtheir purely thermal process without the use of catalysts. The températures reached in the reactorand the thermal mode of the reforming (claimed and even emphasized in the title of the patent),thereiore indicate a treatment of hydrocarbons under thermodynamic equilibrium. The process isbased on a direct arc (two annular électrodes) or transferred arc plasma generator, which arc verytraditional devices known for almost a century. L. KERKER writes in a general manner on the tests on production of synthesis gas atHuis [“Herstellung von Reduktionsgas oder Synthesegas mit Lichtbogenplasmaverfahren”,Elektrowarme international B, Industrielle Elektrowarme, vol. 45(3-4), 155-61 (1987)]. Theillustrations indicate that a tubular reactor with traditional arc, with very high power (1 to 9 MW),is involved; it has been used at this plant since 1939 to produce acetylene. This time, the caseinvolved is a natural gas steam reforming project for the production of 99.9% pure hydrogen, at avery compétitive price with respect to electrolysis (although still more expensive than thehydrogen generated by the traditional steam reforming or partial oxidation methods).

Our team in Orléans has also been working since 1986 on the conversion of hydrocarbonsin thermal plasma reactors. These traditional torches with simple or transferred arc plasma make D 010072 it possible to obtain plasmas with relatively small volume, but at very liigh températures ( T > 10kK). Although these devices may be potential sources of active spccies, they are, nevertheless,poorly suited for Chemical applications requiring lower températures (in order not to completelydemolish the hydrocarbon molécules to soot) and, above ail, greater plasma volumes to be able to 5 act intimately on ail the fluid to be treated. The plasma torch technology, for examplc, wellestablished in the solid projection domain, has thus been found at the same time very costly andvsry difficult to implement for Chemical processes. However, we hâve obtained somcimprovements in the thermal plasma domain in the case of a transformation of methane withcarbon dioxide or elementaiy oxygen in a specifically controlled electric arc, see P. 1 o JORGENSEN et al., "Process for the Production of Reactive Gases Rich in Hydrogen and in

Carbon Oxide in an Electric Post-Arc, BF 2,593,493, (1986). The structure of the device placedin operation at the time unfortunately did not allow using water vapor as reagent or to workwithout consuming the argon necessary as plasmagenic gas of a first pilot arc. Later we usedahnost the same arc with higher current (20 - 150 A) to study the oxidation of ethylene, see K. 15 MEGUERNES et al., "Oxidation of ethane C2H6 by CO2 or O2 in an electric arc". J. High Temp.Chem. Process, vol. 1(3), p. 71-76 (1992), without much improvement in the consumption ofelectric energy or of plasmagenic argon.

Summary of the Invention 2 C II is in order to correct these problems that we studied reforming of pure methane by carbon dioxide in an electro-reactor which had just been invented by our team. It consîsted ofthree électrodes between which gliding discharges develop; the plasma medium thus obtainedwas very much out of thermodynamic equilibrium and contained numerous excited spccies whichmade it highly reactive. This plasma device has been mentioned above under the name of 25 GlidArc. Our first tests on the production of synthesis gas starting from a CH4 + CO, mixtureinjeeted into this new type of plasma (without any cooling or plasmagenic argon) were reportedby H. LESUEUR et al., "Production of synthesis gas (CO + H2) starting from the oxidation ofCK, by CO2 in a gliding discharge electro-reactor", Physics Colloquium, Supplément to theJournal of Physics, vol. 51 (18), p. C5-49 - C5-58 (1990). We later made a more systematic 30 conrpariscn of methane reforming with carbon dioxide in a transferred arc and in the GlidArc to show the great superiority of the gliding arc reactor, see K. MEGUERNES et al., "Oxidation ofCH, by CO2 in an electric arc and in a coid discharge", 1 lth In. Symp. on Plasma Chem.,Loushborough (England), 1993, vol. 2, p. 710-715. Lastly, a complété article on the conversionof CH4 by CO2 was published by H. LESUEUR et al., "Electrically assisted partial oxidation of 35 methane", Int. J, Hydrogen Energy, vol. 19(20), p. 139-144 (1994).

This (pure) methane reforming by (pure) CO2 has shown a very interesting way to upgrade certain gases with high contents of carbon dioxide. However, the products leaving ourreactor had an H/CO molar ratio between 0.5 and 0.8, ahnost in agreement with reaction (5).Therefore, diis gas composition was totally unsuitable for the Fischer-Tropsch technology 7 010872 (synthesis of hydrocarbon synthetic fuels, "syncrude") or similar technology for the production ofmethanol. The two processes require synthesis gas with an H2/CO ratio near 2:1.

We discovered also that, after a few improvements, the same GlidArc devicc is wellsuiied for a supply of pure water vapor as the only plasmagenic medium. The overhcating tests of 5 the water vapor by means of this device were performed at laboratory scale and at atmosphericpressure. The improved GlidArc was supplied with very wet water vapor at 105’C. Nodecerioraâon of the plasma generator supplied with water vapor was observed after scvcral longexperiments. The water vapor thus overheated at atmospheric pressure and Chemical 1 y activatedby the presence of H, O, OH and other metastable species may be of interest for drying or for lû Chemical transformations, see P. CZERNICHOWSKI and A. CZERN1CHOWSKI, "Glidingelectric arcs to overheat water vapor", 9th University-Industry Colloquium "Electrical techniquesand qualiry of drying", Bordeaux-Talence, 1994, p. Bl-l-Bl-7.

It is at this stage that we thought that traditional steam reforming of pure méthane can beimproved in the presence of gliding electric arcs which contribute to the reaction medium an 15 easijy coctrollable enthalpy and some highly reactive species. These particular arcs may then playthe -oie o; a catalyst in homogeneous phase, see A. CZERNICHOWSKI et al., "Assistancedeviice and process by means of plasma in the non-catalytic steam cracking of hydrocarbon andhalegenated organic compounds", BF 2,724,808 (1994).

The previously mentioned methane steam reforming endothermie reaction (1) requires, in 20 order te be fully executed under standard conditions (298 K, 1 atm), an energy équivalent to 206kJ per tracsformed CH„ mole, or else at 0.64 kWh per 1 m3(n) of the CO + 2 H2 mixtureprodmeed. When the reaction is barely started under standard conditions (the CH„ transformationratios being only 0.005%), it is necessary, according to Thermodynamics, to heat the reagents tohigher températures, which requires not only to supply the reaction enthalpy, but also leads to 25 reheahrg adl the mixture. Our calculations indicated that a minimum cost, 0.933 kWh per l m3(n),for the CH*. CO- equimolar mixture is situated at the température of 950 K, where a 75%transdemarion of initial CH4 is attained. At this stage, the H2/CO rnolar ratio is too higli (4.98) forsome applications of this synthesis gas. In order to increase the transformation ratio of methane toapproximately 97%, it would be necessary to heat ail the reaction mixture to approximately 1200 30 K at the theoretical cost of 0.986 kWh per 1 m3(n) of the CO/H2 mixture, but the excess hydrogenstill oers ac an HVCO level equal to 3.04.

The pure CH4 décomposition in the presence of overheated water vapor in a simpleGlidArc reactor, without maturation, has actually yielded large quantifies (in terms of percentagesby voiiume) of H- (up to 68%) and CO (up to 14.8%), while the percentagc of C2H2 by volume 35 (max. I. :% and C2H4 (max. ~ 0.34%) were low. In ail cases, we had H2/CO molar ratiosexceedmg the value of 4 and even reaching 5.8! I: is possible to convert CO into CO, or, inversely, CO2 into CO via almost athermicréactions (16) and (17), called "shift". This makes it possible to préparé mixtures with the desiredcompoâton. of synthesis gas for a particular application. Nevertheless, in practice in the industry, η 010872 these reactions require a separate reactor, and thc presence of catalysts and thcy are accompaniedby ail the problems due to the complexity, poisoning and aging of the catalytic load, etc.

In order to explain the phenomenon observed of too high an H2/CO ratio in our puremethane steam cracking tests assisted by the GlidArc plasma, we performed a sériés of tests, see 5 A. CZERNICHOWSKI and K. MEGUERNES. "Electrically assisted water shift reaction", 12thInt Symp. on Plasma Chem., Minneapolis, Minnesota, 1995, vol. 2, p. 729-33. By injecting amixture of caibon monoxide with water vapor in a GlidArc reactor, wc did observe reaction (16),without the least presence of traditional catalyst. Therefore, it is the plasma itself which catalyzedthis shift, converting CO into H2. 10 The objective of the process and the plasma assistance device to steam reforming, to the reforming with CO2 or to simultaneous reforming with an H2O/CO2 mixture of hydrocarbons isthe production of gases rich in CO and H2, containing also high ratios of C2H2, C2H4 and C3H6,without formation of soot or coke. The process makes it also possible to upgrade the CO2 byconverting it into CO in the presence of hydrocarbons. 15 This mixture of valuable products is obtained in a reactor /1/ with electric gliding arcs /4/ which strike directly into an endothermie reaction medium consisting of hydrocarbons mixedwith H2O and/or CO2. The reactor is equipped with a diaphragm /19/ with a convergent/divergenthole /20/ to reinforce the agitation of the arcs with the load to be converted and, at the saine time,to hâve the conversion of the load progress after prolonged contact with catalytic species derived 20 front the plasma.

Brief Description of the Drawings

Figure 1 is a schematic of the reactor used in the inventive method.

Figure 2 is a schematic diagram of the entire reactor System used in the inventive method. 25

Detaüed Description of the Invention

Therefore, we had a new idea, which is the subject of this invention, to applysimultaneously H2O and CO2 to a mixture (with variable composition, as needed) in order toobtain simultaneously during one single operation in the GlidArc reactor, a conversion of certain 30 hydrocarbons by steam reforming (reactions 1 through 4), reforming with carbon dioxide (réactions 5 through 8) and an inverse shift of part of the hydrogen (reaction 17). The purpose ofthis is to obtnin a synthesis gas with a désirable H2/CO molar ratio per further use of thissynthesâs gas, for example through a Fischer-Tropsch process. This objective has been achievedand, fînrthermore, we hâve been suiprised by the appearance of other conversion products of the 25 foad: CjHj and C3H8 at quite high contents. These unsaturated products can then contribute an addational value to this hydrocarbon conversion process assisted by gliding electric arcs.

Another new idea, which is another feature of this invention, is to divide the old GlidArcreactor into two compartments. By adding a partition, in the form of a diaphragm, we create inthis macmer a gliding arc compartment with reinforcement of the rccirculation of the reagents, * 010872 and another maturation compartment where the reactions gcneratcd in the arc zone can becompleted. The two parts of the reactor communicate through a very large ho le allowing thereagents and the active species to penetrate the maturation post-plasma zone.

Several types of GlidArc reactors may be used. That sketched in Fig. 1 is a small size5 device (laboratoiy scale) used to illustrate the invention. Of course, it is only a non-restrictive example of execution of a future industrial-size reactor. The small gliding arc reactor /1/ uses sixstainless Steel 0.8 mm thick profiled sheet électrodes /21 (only two of the six électrodessymmetrically arranged around the axis of the flow of the fluid to be treated are shown in Fig, 1).Each one of the électrodes is 14 cm long and 25 mm wide. The électrodes delimit a nipple-shaped 1 o space /3/ in which the gliding electric arcs /4/ can develop. This reactor contains a 1.8 mmdiameter nozzle /5/ blowing the fluid /6/ to be converted into space /7/ between the électrodesarranged so that the fluid circulâtes along the central part of these électrodes exposed to the arcs.Thus, the roots /8/ of the arcs, cracking and pre-ionizing the gas at site /9/ where the distancebetween the électrodes is the least, glide on these électrodes, then disappear at site /10/ near the 15 end of the électrodes, to reappear at the initial site. The process is sequential and tire life of an arc/4/ is observed to be between 1 and 20 ms, depending on the linear specd of the fluid in zones /7/,/9/, /3/ and /10/ between électrodes /2/. The gliding arcs /4/ hâve variable characteristics startingfrom site/9/ where they are started, up to their extinction /10/, specifically with dissipations ofenergy which grown in time. The reactor is closed by means of a lid /11/ holding tire electrically 20 insulated électrodes with high voltage connections /12/. The entire structure is sealcd; it withstands a partial vacuum (in the order of 7 kPa) as well as a 12 bars overpressure at the time ofcombustion of methane-rich mixtures. Lastly, an orifice /13/ is provided as the outlet of theproducts of the treatment. The reactor (with 80 mm inside diameter and 1.5 liters capacity) isequipped with a closed stainless steel double wall /14/, as an envelope. These double walls /14/, 25 insulated by means of minerai wadding /15,/ are used to recycle the energy released in the reactor,injecîing it into the incoming fluid (6a). The heat losses of the reactor can be further limitcd bymeans of aresistor/16/ wound around the reactor and carrying an electric current. Othcr fluidscan be added separately through intake (6b), to form a mixture which is then injected by nozzle/5/. Some faoles or take-ofïs /17/ are used, for example, to branch a pressure gauge, run a 30 thermocouple wire or a sampler of the fluid entering the reactor. Through a fast ( > 10 m/s), almost punctifonn injection of the fluid between électrodes /2/, a recirculation phenomenon /18/of the reageats in the gliding arcs zone is already produced. To reinforce this recirculation, weadd a ceramic diaphragm /19/ provided with a wide axial hole /20/, thus dividing the reactor intwo parts: a compartment of arcs /21/, approximately 2/3 of the total volume of the reactor, and a 35 "maturation" compartment /22/, equal to approximately 1/3 of the total volume of the reactor.Hole /20/, with convergent/divergent shape (18 mm in diameter in the narrow part) allows thereagersîs (paitially used up) to pass, as well as the long-living active species originating from theexcitation of the gases by the gliding arcs. Therefore, in the maturation zone, the conversion islikely to be eaded in an environment in which the température is much lower. The fluid, once in 10 01 0 8 7 2 this post-plasma zone, cannot any longer retum to the arc zone. The bright zone ofthe glidingj electric arcs can be observed through a 12 mm diameter porthole /23/, in order to make certain of | the proper operation ofthe reactor. Very important information can be drawn from the émission l · spectrum of this zone! The conversion of the hydrocarbons can be sufficiently advanced at the 5 time of passage through a single GlidArc reactor. Otherwise, the products partially converted in a ; ” reactor can be treated in several reactors thus described and placed in sériés (not shown).

Spécial care must then be emphasized at the time of installation of diaphragm/19/in theshape of a convergent/divergent hole. These new means create a new maturation reaction zone inwhich very active and metastable species (thus having catalytic properties) make it possible to 1 o reform hydrocarbons resulting from violent reactions in the plasma zone, can be deactivated on other molécules and thus cause the conversion of the reagents to progress even faster. Physics provide us with information on such atomic and molecular species as H*, OH*, O2*, CO2*, CO*, ; H2*, H3* (and many others) which hâve a sufficiently long life to travel long distances in the gas • flux, even at atmospheric pressure. This phenomenon is very important for the conversion of 15 hydrocarbons known for their fragility. In fact, the action of a non-thermal (or out of equilibrium) ' plasma, such as the GlidArc plasma coupled with the maturation post-plasma zone enables us to completely prevent coking of the hydrocarbon load. Long hours of operation of the reactor thusbuflt and perfect transparency of the porthole (ail this in the presence of hydrocarbons as fragileas propane and butanes) are the best proof of "soft" transformations which can be executed in a 2 o GlidArc reactor with said post-plasma compartment.

The reactor is supplied by controlled flows (by mass flowmeters) of gas taken frombottles (or other sources) and/or of the water vapor produced by a generator. The supply of thereactor with an initially liquid substance at ambient température (for example, a heavierhydrocarbon or water) can also be carried out by using a dosing pump. The constant flow of this 25 liquid, controlled by a valve and a flowmeter, is thus evaporated in an oven, to be then injectedbetween the double walls and, lastly, into the reactor, whether or not previously mixed withanother fluid of the process.

Chemical analyses are performed, using traditional gas chromatographie methods. We usethree chromatographs, each assigned to the spécifie dry gases: CO, CO2 and CH4 for the first, 30 hydrogen alone for the second, and ail the hydrocarbons for the third. The flow of the water vaporin the products is quantified by trapping a known volume of exiting gases.

The gliding arcs inside the reactor are supplied by a spécial high voltage system ensuringat the same time preionization of the medium and then transfer of the electrical energy to theplasma. The electric power of the reactor used varies between 0.57 and 1.09 kW under 0.1 or 0.2 35 A for a flow rate of fluids to be treated from 0.57 to 1.23 m3(n)/hr; the energy supply with respectto the ïoad is 0.47 to 1.23 kWh/m3(n). Nothing nevertheless prevents using more power, higherflow raies and/or greater energy for industrial operations.

Refonning of a naturel gas (NG) or of a "propane" will be better understood with the helpof Fig. 2. The reactor used is that shown in Fig. 1. Fig. 2 is a schematic représentation of the 11 010872 apparatus as a whole. In this figure, the GlidArc reactor /1/ is supplied by a spécial high voltagepower generator /24/. It is operated directly with, as plasmagenic gas, a NG takcn from the citysupply network /25/ (or with the "propane" /26/ from a pressurized cylinder), mixed with carbondioxide /27/, water vapor (or liquid water) /28/ or with the CO2/H2O mixture. The gas flow rates 5 are controlled by mass flowmeters /29/. The gas mixture entering (dry) can be sampled for chromatographie analysis through a take-off /17/. The flow rate of the water vapor is also knownafter calibration of a dosing pump connected to the entry /28/. The thermocouple /29a/ makes itpossible to measure the température of the fluid at the entry of the injection nozzle while theprobes /30/ and /31/ indicate the températures in the two compartments of the reactor. A pressure 10 gauge /32/ gives at any time the pressure inside the reactor: this pressure is kept slightly higherthan atmospheric pressure. The products leaving the reactor are cooled in a heat exchanger in theair /33/. After leaving the exchanger, the gases are directed to a direction invertor tap /34/ whichssnds them either to analysis /35/ or to évacuation stack l36aJ. At the time of our tests, we collectand weigh the water leaving the reactor, by condensation /37/ and absorption /38/, as well as the 15 dry gaseous product for chromatographie analyses. To this effect, the wet gas is conveyed tooutlet /36a/, then when we estimate that the reactor is operating in stable condition (pressure,températures, gas flow rates, water vapor flow rate, electrical power), tap /34/ is reversed and it issent to analysis /35/. The water is stored in the greatly cooled flask /37/ and in an absorbingmateriaL Tap /39/ being first closed and taps /40 and /40a/ open, the diy gas runs through a bulb ; c or a spherical flask /41/ then through gas meter /42/ and leaves the experimental device through/36b/ for the évacuation stack. The température of the gas at the outlet of the meter /42/ ismeasured by a thermometer /43/. At the time of each test, also the atmospheric pressure ismeasured with a barometer, in order to bring our balances of volumes to normal conditions (n).

Numerous feasibility tests of the reforming process of natural gas or "propane" were 2 5 penormed in the new reactor with the maturation compartment (we are presenting only the mostsignificant tests). The composition (% by volume) of the NG originating from the citydistribution network was not changing much: CH4 from 89.7 to 91.9; C2H6 from 6.6 to 6.S; C3Hgfrom 1.1 to 1.2, C4Hi0 from 0.25 to 0.29 (mixture of n- and iso-butane); O2 from 0.17 to 0.34; andN2 from 1.2 to 1.8. Besides, we were analyzing carefully this NG at the time of each test in order 5 : to establish an exact balance of matter. The composition (% by volume) of the "propane" contained in a bottle was: CH4 0.1; C2H61.0; C3Hg 96.7; C3H60.3; and C4HI01.9 (also a mixtureof n- and iso-butane).

Table 1 summarizes examples G1 through G5 of natural gas steam reforming. Table 2summarizes examples G11 and G12 of NG reforming with CO2 alone. Table 3 summarizes z 5 examples G21 through G23 of NG reforming simultaneously with an H2O/CO2 mixture. Lastly,Table - illustrâtes our tests PI through P3 with the "propane" in the simultaneous presence ofwater vapor and carbon dioxide. Ail our experiments were performed at a slightly higher thanamiospheric pressure. rr—v-.x-rr-v··-......T’·’’ 010872

Each table is divided horizontally in three parts. The first part indicates the nature andquantity of the fluids injected in the reactor and the spécifie energy injected in the plasma (theactual electric energy of the GlidArc compared to the normal hourly flow rate of ail the enteringreagents), as well as the température of the fluid entering the reactor, that inside the plasma 5 compartment (but not in contact with the gliding arcs) and that inside the maturationcompartment.

The second part of each table indicates the volumes (in normal liters) of dry productsfrom the process leaving the reactor after the injection of 1 kWh of electric energy in the GlidArcplasma under expérimentation conditions. Thus, these values indicate a real energy cost (in 1 o electricity) of the process at laboratory scale. This section indicates also the energy cost of a unit mass of CO (other products considered "at no cost") or of a unit volume of synthesis gas (otherproducts aiso considered "at no cost") having a given H/CO ratio.

The third part of each Table indicates other results of calculations based on theexperimental data: the global rate of conversion of carbon of NG origin (or from "propane") and 15 possïbly of CO, origin, the conversion rates of the different hydrocarbons présent in the NG (or in the "propane"), as well as the specificities pertaining to conversion of carbon présent in the NG(or in the "propane") and possibly of CO2 to useful products.

We add again the absence of coke, soot, tar or other pyrolytic compounds in our products(withün the îimits not exceeding 0.5% expressed as mass of converted carbon). 20 010872 13

Table 1

Example G1 G2 G3 G4 G5 5 Incoming flow rate l(n)/h NG 424 424 424 424 424 water vapor 473 606 785 803 458 Spécifie energy, kWh/m3(n) 1.13 1.02 0.90 0.47 1.21 Température (deg C) entry 1 220 215 215 200 250 reaction 630 590 560 490 680 10 maturation 310 300 290 300 380 Outgoing l(n)/kWh c2h4 5.8 5.0 4.5 4.9 7.0 c2h2 14.0 10.0 8.7 12.8 9.6 c3h6 0.4 0.3 0.3 0.3 0.5 15 CO 67.5 65.8 64.4 65.6 60.1 CO2 5.0 5.9 8.5 7.9 6.2 H2 262 248 240 248 272 HVCO, mol/mol 3.9 3.8 3.8 3.9 4.5 Energy cost CO,kWh/kg 11.9 12.2 12.6 12.6 13.3 20 H2+CO, kWh/m3(n) 3.0 3.2 3.3 3.2 3.0 Carbon conversion (%) 25.1 23.5 23.6 13.4 23.6

Conversion of hydrocarbons présent in the NG (%) ch4 23 22 22 13 20 25 c2h6 35 31 32 15 41 Specificities pertaining to C3Hs 42 34 37 21 47 carbon conversion (%) to c2h4 10 10 9 9 14 c2h2 25 20 18 24 19 30 c3h6 1 1 1 1 2 CO 59 64 64 59 59 CO2 4 6 9 7 6 35 010872

Table 2

Example Gll G12 5 Incoming flow rate, l(n)/h NG 328 328 CO2 438 438 Spécifie energy kWh/m3(n) 0.75 1.42 Température (deg C) entry 140 165 reaction 290 380 10 maturation 160 180 Exit l(n)/kWh C2H4 3.9 2.8 C2H2 13.9 7.1 c3h6 0.2 0.2 15 CO 205 151 H2O 38 19 h2 173 124 Hj/CO, mol/mol 0.84 0.82 Energy cost CO,kWh/kg 3.9 5.3 20 H2+CO,kWh/m3(n) 2.6 3.6 Carbon conversion (%) of NG origin 8.3 11.4 of CO2 origin Conversion of hydrocarbons présent 9.0 11.8 25 intheNG(%): CH4 17 24 CA 25 35 C3H8 Specindties regarding carbon 25 33 conversion (%) to .... C2H4 3 3 30 C2H2 11 8 c3h6 0.2 0.3 CO 85 88 35 15

Table 3 010872

Example G21 G22 G23 5 Incoming flow rate, l(n)/h NG 495 484 446 CO2 52 138 138 Water Vapor 332 254 177 Spécifie energy kWh/m3(n) 1.23 1.07 1.33 Température (deg C) entry 240 230 230 10 reaction 665 660 675 maturation 390 395 405 Exit C2H4 6.8 6.4 6.2 c2h2 13.5 17.2 6.7 15 c3h6 0.5 0.5 0.6 CO 80.9 95.3 88.0 H2 268 245 218 H/CO, mol/mol 3.3 2.6 2.5 Energy cost CO.kWh/kg 9.9 8.4 9.1 20 H2+CO.kWh/m3(n) 2.9 2.9 3.3 Caifaon conversion (%) of NG origin 21.6 17.0 15.6 O2 of CO2 origin 1-1 3.5 3.2

Conversion of hydrocarbons présent intheNG(%): CH4 20 18 17 25 C2H6 41 36 37 C3Hg 47 40 42 Specificities regarding carbon conversion (%) to C2H4 11 9 11 30 C2H2 22 24 11 CA 1 1 1 CO 66 66 76 15

Table 4

Exampfie PI P2 P3 1b 010872 5 Incoming flow rate, l(n)/h "propane" 343 221 512 0.77 215 510 330 223 221 226 1.33 220 610 325 207 180 309 1.36 220 625 350 CO2 water vapor Spécifie energy kWh/m3(n) Température (deg C) entry reaction maturation Εχίζ l(n)/kWh C2H4 19.5 21.3 23.5 10 c2h2 31.0 19.1 21.5 5.2 5.9 6.5 CO 105 120 123 H2 227 203 222 EL./CO, mol/mol 2.2 1.7 1.8 15 Energy cost CO. kWh/kg 7.6 6.8 6.5 H2+CO,kWh/m3(n) 2.9 3.1 2.9 Carbon conversion (%) of "propane" origin 13.4 20.1 2.7 of CO2 origin Specificities regarding carbon 2.5 3.9 4.1 20 conversion (%) to C2H4 16 18 24 C2H2 26 16 22 c3h6 5 6 7 CO 44 50 46 25 The comparison of our recent results from NG steam reforming (shown in Table 1 ) with the preceding results taken from experiments performed on pure methane in tlie GlidArc reactorwithout maturation compartment (see A. CZERNICHOWSKI et al., 1994, Tables 2 and 4,experiments M4 and M10) clearly indicates the superiority of the new device (described above).Table 5 illustrâtes these différences for similar conditions (respectively G3 and G4), conceming 30 the ELO/hydrocarbon ratio and the energy supply to the load to be converted:

Table 5

Example G3 G4 M4 M10 Spécifie energy, kWh/m3(n) 0.90 0.47 0.94 0.42 H2O/hydrocaxbon at entry (mol/mol) 1.85 1.89 1.89 1.71 Température (deg C), reaction 560 490 345 220 maturation 310 300 non non 17 010872

Exit (mol/mol) C2H4/C2H2 0.52 0.38 0.28 0.22 (Cft+CA+CjHÜ/CO 0.21 0.27 0.11 0.18 h2/co 3.8 3.9 4.0 4.3

Hence, we now obtain many more unsaturated hydrocarbons. At the same timc, for asimilar H7CO ratio, the C2H4/C2H2 ratio is higher. These results witness reinforcement of therecirculation in the GlidArc compartment shortened by installation of the diaphragm. Thus, thehydrocarbon load can be in doser and more prolonged contact with the gliding arc zone; that iswhere much acetylene is created. At the same time, we observe partial hydrogénation of theacetylene to ethylene, which occurs outside the arcs in the maturation compartment. In anenvironment in which the température is sufficient to ensure very rapid partial hydrogénationkinetics, part of the acetylene is converted to ethylene, a product even more sought for itsmultiple applications.

We point out that, for the first time, we hâve performed steam reforming of ethane,propane and butanes présent in the natural gas used as reagent. On the basis of our comparativeChemical analyses and our exact balances of material entering and leaving the GlidArc reactor(see Table 1), we détermine that the conversion of the ethane and of the propane is much higherthan that of the methane. Furthermore, the conversion of the propane is greater than that of theethane. The ratio of these hydrocarbons in the incoming gas is CH4:C2H6:C3H8 ~ 79:6:1. Theirmean conversion is (in relative scale) in inverse proportion to CH4:C2H6:C3H8 ~ 1:1.5:1.8. Thisindicates thaï, thanks to this steam reforming process of hydrocarbon loads containingincreasingly heavy hydrocarbons, their conversion is attained with increasing ease and with thesame spécifie energy applied to the incoming load. The steam reforming process assisted bygliding arcs could then be applied, whatever the natural gas (or other mixture of hydrocarbons) tobe converted.

We note that the global conversion rate is limited in ail the experiments presented here inorder to better study the individual conversion phenomena of each component of the NG or of the"propane". This conversion can obviously be much greater, for example, following an increase ofthe spécifie energy injected in the reagents.

The other comparison of our results of conversion of the NG containing CO2 (shown inTable 2) with our previous results conceming experiments on the mixture of pure methane withsome CO2 brought into a GlidArc reactor without maturation compartment (see H. LESUEUR etal., 1994, Table 1) confirms the superiority of the device now described. For example, under theprevious "B" conditions (spécifie energy equal to 0.94 kWh/m3(n) and the CO,/CH4 molar ratio =1.13), the energy cost of the CO produced is similar, but the H2/CO ratio obtained is better,exceeding the 0.8 value, while the previous ratio was 0.6. We emphasize also that, for the firsttime, we hâve performed reforming with carbon dioxide of ethane, propane and butanes présentin theNG (used as reagent). According to our analyses and exact balances (see Table 2), we 1t. 010872 observe that the conversion of the ethane and of the propane is more pronounced than that of themethane. Their average conversion is, on a relative scale, in CH4:C2H6:C3Hg ratio of~ 1:1.5:1.5,despite a very high excess of methane in the NG studied. Tliis indicates again that the reformingprocess with CO2 of hydrocarbons heavier than methane would be easier. 5 The reforming process with CO2, assisted by gliding arcs, could then be applied with any natural gas (or other mixture of hydrocarbons) to be converted. We are thinking, for example, ofthe different biogases or of certain gas resources with mixtures of hydrocarbons and carbondioxide. These gases can thus be upgraded without costly séparation of CO2. Moreover, havingavailable a "clean" energy source (solar, hydraulic, nuclear, etc.), we could thus recycle the 1 o carbon dioxide, which is a formidable contemporary problem.

We are demonstrating for the fîrst tirne the feasibility of a ncw hydrocarbon conversion process assisted by gliding arc plasma in the simultaneous presence of carbon dioxide and watervapor. This process is illustrated in Tables 3 and 4 by the conversion of two model mixtures ofhydrocarbons in a new reactor provided with a post-plasma maturation compartment. In the 15 simultaneous presence of water vapor and CO2, we can thus convert ail hydrocarbons such asCH4, QH* C3Hg and/or C4H10 into synthesis gas and partially also into other valuable products: QHj and C3H6, without using traditional catalysts. In particular, in the Asia-Pacificcountries and Pakistan there exists great amounts of CO2 in the natural gas. Huge gas fields arereported in Indonesia having CO2 contents upwards of 70 v% (Exxon Natuna, for example). 20 Eields in Pakistan range from 6 to 80 v% CO2. Removing this CO2 is not only expensive but alsoprésents a disposai problem. While reinjection into an aquifer is a possibility, it is also expensiveand an adéquate aquifer must be located nearby. This invention uniquely enables large CO2contents to remain in the natural gas and yet produce synthesis gas suitable for synfuel orpetrocbcmical production. The ability of this invention to convert high CO2 natural gas into 25 synthesis gas to produce valuable end products promises to open new routes to reduce globalcarbon émissions. A wide range of ratios of two oxidizers can be used. Although our examples are given forE20/CO2 values between 1.0 and 6.4, the fact of being able to use only one oxidizer makes itpossible to widen this ratio for values between 0 and oo. Hence, ail the H2O/CO2/hydrocarbon 30 nmxtures can be converted in the GlidArc reactors without prior séparation of components.Accordmg to necessity, we can then obtain a synthesis gas with an H2/CO ratio near 2 for thesynthesis of synthetic oil or of methanol, or of a synthesis gas very rich in hydrogen for tiresynthesis of ammonia, or yet of a gas very rich in CO Per "oxo" synthèses..., these examples notbenng restrictive. 3 5 We note the complété absence of soot, cokes or other undesirable products from the comversiom of heavy hydrocarbons, such as the butanes présent in non-negligible quantity at thetùme of ouïr tests. On the contrary, the increasing fragility of increasingly heavy hydrocarbons is a"phas" for our process, from the point of view of the energy cost for the production of CO andalscr of other valuable unsaturated products. In some cases, this cost is reduced by half by passing 19 010872 from methane-rich gas to propane-rich gas. This is a strong point of our process when comparedwith the traditional processes confronted with the problem of déposition of cokes and tars,especially in the presence of heavier hydrocarbons than methane.

Lastly, we point out the presence of non-negligible quantities (but at adjustable content) 5 of unsaturated hydrocarbons C2H4, C2H2 and C3H6 in our products from conversion assisted by

GlidArc plasma. They contribute an additional value as final commercial product (acetylene) oras raw material for other organic synthèses. Mixed with synthesis gas, they also facilitate theconstruction of hydrocarbon chains at the time of the Fischer-Tropsch synthesis (informationfrom recent scientific work by Professor A. LAPEDUS of the Organic Chemistry Institute of 10 Moscow). Thus, formed simultaneously with the CO and H2 during the conversion of hydrocarbons in the GlidArc, these unsaturated molécules can contribute to the direct applicationof an improved synthesis of liquid hydrocarbons.

On a more technical level, it must be pointed out how surprisingly easy is the operationof the reactor and of its assembly, without détérioration of électrodes, electrode holders, 15 diaphragm or wall of the reactor or of the maturation compartment, ail submitted to the action ofthe incoming reagents and of the outgoing products. This is explained by the moderatetempérature of the assembly ( < 680 deg C) and by a very short contact time between the roots ofthe arcs with the électrodes, even if made of Steel and even if not cooled. We did not encounterany problems in the implémentation of the plasmagenic gases chosen: the mixtures of 2 0 hydrocarbons with water vapor and/or CO2.

Our experiments hâve demonstrated the feasibility of the new process of production of gases rich in hydrogen and carbon monoxide, containing also very large quantities of C2H4, C2H2and CjHg.

The process consists of manufacturing these gases by means of gliding electric arcs which 25 strike directly in the hydrocarbon mixed with water vapor and/or with carbon dioxide in any proportions. This causes the oxidation and/or partial cracking of these hydrocarbons, avoiding thedisacvantages of the existing processes. The reagents, partially converted in a gliding arccompartment, then penetrate another maturation compartment which is separated from the directreaction zone by a diaphragm with a large hole. There, in the presence of the still active species 30 produced in the arcs and transported by the gas leaving the arc zone, the gas undergoes an additional conversion at a much lower température than that présent in the direct reaction zone.

The subject of this invention then is a process which allows the partial cracking andoxidarion of the hydrocarbons in the active presence of water vapor and/or carbon dioxide,without any need for other reagents or catalysts and without the formation of soot, coke or tar 35 with the proper operation of the reactor. The tests clearly demonstrate the ease of reforming withsteam, or carbon dioxide or simultaneous reforming with an H2O/CO2 mixture accompanied bynon-catalytîe hydrocarbon cracking.

The invention makes it also possible to transfer directly electrical energy under highvoltage and relatively low current to an endothermie reaction medium. These electrical ί i 1

J i 10 15 20 010872 conditions, combined with high speed of the plasmagenic medium in the arc zone, cause a strongelectric and also thermodynamic non-equilibria. The material injected into this non-cquilibriumplasma zone created in the GlidArc device then reacts in non-thermal manner.

No difficulty was noted at the time of the experiments and the extrapolation for largevolumes is easy. Despite a non-optimized reactor and only one pass of the reagents through theGlidArc compartment, a large part of the initial molécules is converted into synthesis gas and intounsaturated hydrocarbons. This conversion is greatly improved by the almost punctiforminjection of the reagents into the arc zone by using a fine nozzle and also by means of adiaphragm with a convergent/divergent hole placed axially and reinforcing recirculation of thereagents in this direct reaction zone.

Other positive points can also be claimed for a future practical application:transformation of hydrocarbons and possibly of CO2 into products with much greatervalue (H2, CO, unsaturated hydrocarbons), the only reagent necessary is water and/or CO2,the absence of any catalyst, the very compact equipment which can be installed at sites with restricted surface area (for example on offshore oil platforms for the conversion of associated gases).the method does not dépend on the Chemical composition of the mixture of hydrocarbons,the GlidArc reactor has no Chemical inertia and can respond very quickly to controlsignais, the incoming and outgoing products, after condensation of the water vapor, do not carryany foreign ballast increasing their volume, which makes the conversion operationseaaer. 4·

Claims (1)

1. 21 010872 WHATIS CLAIMEDIS: 1) A Hydrocarbon conversion process comprising reforming with an oxidizcr wherein amixture of one or more hydrocarbons with an oxidizer is submitted to a gliding arc plasma to startand maintain endothermie Chemical conversion reactions of said mixture into synthesis gas, saidsynthesis gas being a mixture ofhydrogen H2 and of carbon monoxide CO, and wherein saidsynthesis gas further comprises on or more unsaturated hydrocarbons. 2) The process of claim 1, wherein said oxidizer comprises water vapor H2O. 3) The process of claim 1, wherein said oxidizer comprises carbon dioxide CO2. 4) The process of claim 1, wherein said oxidizer comprises carbon dioxide CO2 and watervapor H2O. 5) The process of claim 1, wherein said unsaturated hydrocarbons comprise one or moremembers of the group consisting of acetylene C2H2, ethylene C2H4 and propylene C3H6. 6) The process of claim 1, wherein said mixture contains oxidizer in volumétrieoxidizer/hydrocarbon ratio equal to at least 0.7. 7) The process of claim 6, wherein the CO2 reacts with said one or more hydrocarbons incontact with said gliding arc and is converted into carbon monoxide. 8) The process of claim 7, wherein said conversion is obtained at a pressure between 7kPa and 12 bars and wherein the température of the gas outside said gliding arc is less than orequal to 680°C. 9) The process of claim 8, wherein said synthesis gas further comprises one or moremembers of the group consisting of ethylene, acetylene and propylene, and wherein said synthesisgas has a H7C0 ratio of between 0.8 mol/mol and 4.5 mol/mol, and an unsaturated hydrocarbons/CO ratio greater than 0.06 mol/mol. 10) The process of claim 9, wherein said conversion is obtained without décompositionof the hydrocarbon load into soot, coke or tars in relative quantifies exceeding 0.5% expressed inmass of converted carbon. 22 010872 11) A device for hydrocarbon conversion comprising a gliding arc structure for creating aplasma, said gliding arc structure placed in an arc compartment, a maturation compartmcntseparated from said arc compartment by means of a diaphragm, said diaphragm having a holetherethrough such that gases are allowed to pass between said arc compartment and said 5 maturation compartment directly through said hole in order to reinforce recirculation in the arccompartment. 22 psgss CARBON RESOURCES, LTDpar procuration