MEANS FOR HYDROGEN

The present invention relates to a device for producing hydrogen from a liquid effluent fermentation substrate, and the use of the device and to a method for the production of hydrogen.

The present invention finds an industrial application in the production even hydrogen, but also in the manufacture of devices for such production.

The references between hook ([..]) refer to the list of references at the end of the section "examples".

Currently, the perspective of a hydrogen economy a lot of interest, both at scientific and industrial applications or policy. This due to the properties of hydrogen, non-carbon molecule with a high energy density. First the hydrogen is a reagent pçur industry, which currently consumes substantially all of the hydrogen produced in the world, half of which is used by the petroleum industry for petroleum refining, and 40% for the production of ammonia, but it is also increasingly perceived as an energy carrier, either by direct combustion, which is not the optimal method, either through its use in fuel cells for electrical production, or the conventional biofuel production, enable capturing carbon dioxide molecules.

The report by the Parliamentary evaluation of the scientific and technological choices French Parliament produces good perspectives to the introduction of the hydrogen as an energy carrier. The various application hydrogen Y are underlined, as well as the potential of decentralized production units for reducing transportation costs.

The production of hydrogen or dihydrogen is typically provided by catalytic reforming processes fossil hydrocarbon. The dihydrogen is for example generated from natural gas, or from other hydrocarbons with various degrees of efficiency. It retains many of these methods are disadvantageous in cost and pollution of the ecosystem.

The dihydrogen may also be removed from the water by methods of chemical reduction or thermolysis.

The dihydrogen can also be produced by biological processes, for example by means of algae. However, this method very little pollution compared to chemical extractions from hydrocarbons, currently not suitable for industrial production.

Beyond these considerations for producing, transporting hydrogen involves significant costs. Each year, in the case of France produces about 920,000 tons of hydrogen the majority of which is directly consumed at the production site. A significant portion (40%) is produced by steam reforming of methane, which is easier to ship from site to site, the use of methods for renewable production on site, from déôftets, would allow processes as hydrogen consumers by limiting the energy consumption and the production of greenhouse gasses.

Analyses life-cycle methods of producing hydrogen production methods show that enhanced bioremediation can potentially competing with thermochemical biomass conversion, for example by gasification, pyrolysis, hydrothermal liquefaction, with a positive environmental impact,

Production costs are to be paralleled with the true cost of hydrogen to purchase, comprising packaging and transportation of the product. The large centralized units for industrial production of catalytic reforming do not allow transporting hydrogen in a small amount to the local scale without observing an explosion transportation costs. The use of delivery systems to smaller scale, context aware energy and industrial premises would optimize costs and energy efficiency of the hydrogen distribution network.

There is thus in this context generally a real need of finding alternative means to those previously known particularly for solving the aforementioned problems and produce locally, at reduced cost and with an environmentally friendly method, hydrogen on industrial scale to make possible the use of hydrogen advantage, especially for clean energy production.

The present inventors have precisely this need by providing a device and method of use which provides a solution to many problems mentioned earlier.

It is a device, a method for the biological production of hydrogen by fermentation of biomass coupled to a separation membrane type.

The present invention particularly relates to a device particularly for producing hydrogen from a liquid effluent fermentation substrate, said device comprising a membrane reactor or "membrane module", the membrane reactor being connected directly or indirectly to an extraction means hydrogen, said membrane reactor comprising:

a shell defining an interior volume, said interior volume comprising two sealed compartments therebetween, namely an outlet manifold (26 of bis) and a compartment (24) effluent to be filled by the liquid effluent fermentation substrate,

hollow fibers each formed by a tubular wall surrounding a day of fiber, said tubular wall being liquid-tight and gas-permeable, hollow fibers are housed in said effluent compartment and arranged so that their day fiber communicates with said outlet manifold is sealed with respect to the effluent liquid when present in the effluent compartment, and

a gas outlet communicating between said outlet manifold and said extraction means hydrogen,

said compartment effluent, said hollow fibers, said manifold gas outlet, said gas outlet means and said hydrogen extraction being arranged so that the passage of a gas or a gas mixture from the liquid effluent when in the compartment effluent through the day of fiber hollow fibers into said means hydrogen extraction to move freely or by means of a sweep gas.

The present invention provides a biochemical process for the conversion of biomass into hydrogen, less energy-consuming that the thermochemical processes previously known, such as gasification, pyrolysis, and so forth, or other biochemical processes, for the production and removal of hydrogen or bio-hydrogen by fermentation. The greatest energy efficiency of the fermentation process is 67% of dark, in practice of 15 to 33%, value well above for example than the photolysis of water.

The device of the present invention is particularly useful for implementing a hydrogen production by fermentation dark. The term "fermentation-dark" in the sense of the present invention, the anaerobic fermentation is built up in by fermentation of organic substrates bio-hydrogen. It is a complex process by variously microorganism communities, involving a series of biochemical reactions using three similar steps of anaerobic digestion. The fermentation process dark differs from photo-fermenting in that it takes place in the absence of light.

Thus, unconstrained light delivery and under anaerobic conditions, the fermentation process dark corresponds to the degradation of organic compounds by an anaerobic microbial consortium, for producing hydrogen continuously, for example from fermentable waste, e.g. in fermentation self-regulating.

The present invention is for example advantageously the production and removal of hydrogen from organic wastes at a low temperature, for example 35 to 39 °c or over a wider range, e.g. 20 °c to 70 °c, from biomasses, including sewage sludge and industrial sludge. Agricultural biomass has for example been used for continuously feeding the biecéacteur membrane of the present invention with results that confirmed the very many advantages of the present invention.

Among the biochemical processes known for the production of hydrogen, only three allow direct hydrogen: photo-fermenting, the photolysis of water and fermentation dark. Dark fermentation is the only one of these processes not requiring light energy, thereby reducing the energy demand of providing continuous operation of the system. In addition, the fermentation dark is the process that results in the best productivities in hydrogen, up to 25 mmolH2/ L._mid the II_had/ hr against 2.5 mmolH₂/ L._mid iie binding_U/ H for the photolysis of water and 10 mmolH2/ L._mid the II_had/ H for photo-fermenting,

According to the invention, by "liquid effluent fermentation substrate", is meant any liquid effluent of natural origin or not may include in particular the nutrients that are required for a microbial consortium to enable it to produce hydrogen by a fermentation process; the nutrients that can be derived from a biomass, and, which can include a microbial consortium capable of producing hydrogen by the fermentation process.

The elements related to the effluent liquid fermentation substrate within the scope of the present invention are developed below, with reference to use of the device of the invention for the manufacture of hydrogen, use thereof, and to the method of the invention.

The substrate or biomass in turn constitutes the "feed" of the bacterial consortium. By "substrate" or "biomass" in the sense of the present invention, is meant any material liquid, solid, semi-solid, slurry, sludge, waste fermented targets usable by the bacterial consortium, present in the substrate itself and/or complemented, to generate a fermentation process for producing hydrogen. The substrate or biomass constitutes an effluent liquid fermentation substrate. It may be for example a sewage sludge, organic waste naturally occurring or not, animal, plant or human, of ultimate waste from e.g. agricultural technology, for example wine or wine, or food, organic waste fermentable little upgradable by other routes, so on one of the many advantages of the present invention appearing herein is that it can be advantageously practiced using locally available metabolizable organic waste, for local production of hydrogen.

For example, the biomass can be advantageously derived from the wine-making industry, irrespective the geographic origin. It may be for example of mires, marc, filter cakes, bound, marc exhausted, scrubbing water kits harvesting, sorting and wine making. The generation of organic waste, including byproducts such as grape marc, the mires, the dregs, cobs and dewatered sludge, the scrubbing water for wine, the scrubbing water tanks elutriating, is seen as one of the most important issues which must be global wine industry. The device, the method of the invention also provide a solution to this problem. Further, the advantage of this biomass, is that it does not require delivery of microbial inoculum, nor for supplementation, nor thermal pretreatment tending to suppress the activity of microorganisms in hydrogen consumption.

The device of the invention is therefore dedicated to the upgrading of biomass for hydrogen and optimizes the production and extraction of hydrogen from the biomass, regardless of the medium in which the biomass is grown or which the biomass is derived from.

According to the invention, the reactor membrane or membrane bioreactor (BRMs) comprises the essential elements mentioned. The membrane reactor components are described below and the examples of membrane reactors according to the invention are also described below and in the section "examples" below. By way of example, include a membrane reactor of the mark SEPAREL (registered trademark) and Liqui aparatus (registered trademark).

The reactor of the present invention is thus most preferably dark, c'est to say that it does not pass the lumen to the effluent. This applies at membrane reactor and also to additional reactors under which the effluent fermentation substrate is present, in static or flow.

According to the invention, the hollow fibers may advantageously be microfibers. According to the invention, the fiber diameter can be selected based on the diameter and/or length of the membrane reactor. For example, the fibers can have an internal diameter of 0.2 to 15 mm, for example 0.2 to 10 mm, for example 0.2 to 5 mm, for example of 0.4 to 0.5 mm, with, independently or respectively, for example an outer diameter of 0.3 to 20 mm, e.g. 0.3 to 15 mm, e.g. 0.3 to 10 mm, e.g. 0.3 to 5 mm, for example 0.7 to 0.9 mm. The length of the fibers depends of course of the length of the membrane reactor, these fibers preferably through the entire length of the membrane reactor. As purely illustrative, for an inside diameter of the envelope of 50 mm for example, the length of the fibers can be of 30 cm.

According to the invention, the hollow fibers may advantageously be made of polymeric material having a permeability to gas selectively or nonselectively. When the permeability is selective, it may preferentially allow passage of hydrogen in the day of fiber hollow fibers for its collection. When the permeability is non-selective, all the gasses, e.g. c0₂ and H₂ pass through the wall of the hollow fibers and find themselves in the day of fiber of such fibers.

It may be for example of fibers made of any material known to those skilled in the and for this gas permeability and impermeability to the effluent. According to the invention, it may be for example of a material usable as material ultrafiltration. It may be of a material selected from polytetrafluoroethylene (ePTFE), polysulphone, matrimid/polybenzamidazole, polymethylsilane, andc.

According to the invention, the walls of the fibers may include for example 0.05 to 0.15 pores of micrometer, for example 0.08 to 0.12 micron, for example about 0.1 micrometer, particularly for non-selective extraction of gas made from fermentation. According to the invention, the walls of nonporous or fibers may for example include 0.5 to 20 nm pores, for example 0.5 to 10 nm, for example 0.5 to 2 nm, for example 2 to 5 nm for a selective extraction of hydrogen.

In this case, generally, it may be of a fiber composed of a permeation membrane material.

According to the invention, the membrane reactor may comprise e.g. a number of fibers depending on the diameter of said reactor and/or the length and/or diameter of the fibers. Depending on the volume of effluent present or passing into the reactor and gas volume produced by the fermentation, the number of fibers is preferably sufficient to allow optimal recovery of the gas produced by the fermentation, and thus hydrogen. By way of example only, the volume of the hollow fibers may occupy about 5 to 30% of the reactor volume, e.g. 7 to 10%. This number is chosen in particular to optimize the contact area between the effluent and the gas in the day of fiber hollow fibers and allow an optimal flow of the effluent in the reactor for the production of gas, thus hydrogen. For example, for a reactor of about 5 cm in diameter, the number can be for example 200 to 500 fibers, for example of about 220 to 260 hollow fiber size as exemplified above.

According to the invention, said compartment membrane reactor effluent, said hollow fibers, said outlet manifold, said gas outlet means and said hydrogen extraction are arranged such that the passage of a gas or a gas mixture from the liquid effluent of fermentation substrate when in the compartment effluent through the day of fiber hollow fibers into said means hydrogen extraction to move freely or by means of a sweep gas.

For example, the fibers may be held together at the outlet manifold, in a waterproof boot through which they pass, leaving freely communicating the day of hollow fiber with the outlet manifold, while preserving sealing with respect to the compartment effluent. The waterproof boot can be for example made of resin or other materials to be sealingly passed through by the fibers without bruising.

Preferably, the membrane reactor is positioned vertically in the device of the present invention in use. Thus, according to the invention, when the device of the invention is used, a sky of gas is formed in the upper part of the membrane reactor, at the outlet manifold.

In the device of the present invention, the interior volume defined by the reactor shell membrane may comprise for example an inlet manifold, the compartment effluent being housed between said inlet manifold and said outlet header, said inlet header being sealingly separated from said effluent compartment, each hollow fiber being arranged to pass through said compartment to effluent and such that said headers communicate with each other via said day of fiber.

In the same way, according to the invention, the fibers can also be held together at the inlet header, in a waterproof boot through which they pass, leaving freely communicating the day of hollow fiber with the inlet manifold, while preserving sealing with respect to the compartment effluent. The waterproof boot can be for example made of a resin, for example epoxy resin or any other material capable of being sealingly passed through by the fibers without bruising.

According to the invention, the membrane reactor may include:

an effluent inlet fermentation substrate communicating between the exterior of said casing and said effluent compartment,

effluent output fermentation substrate communicating between said effluent compartment and the exterior of said shell,

said device may include a recirculation loop hydraulic effluent connected on one side to the effluent outlet and the other to the input effluent.

In the device of the present invention, the recirculation loop may be arranged to allow recirculation "direct" or "indirect" effluent fermentation substrate between said effluent outlet and said inlet effluent.

According to the invention, by recirculation "direct", is meant that the device may be provided with a hydraulic communication for circulating the effluent outside the reactor membrane, from the effluent outlet fermentation substrate, for example to improve the homogeneity between the effluent outlet and the effluent fermentation substrate to the input of said effluent compartment, e.g. for analysis parameters pH, pressure, fermentation activity, temperature of the effluent, and so on, for example to control the temperature of the effluent, for example for enriching the effluent in microbial consortium, before reinjection through inlet fermentation effluent. This circuit can be provided with elements necessary analyses and/or regulation of temperature and/or adding supplements mentioned, and, advantageously of the elements necessary for the control of these parameters to maintain such optimized hydrogen production.

According to the invention, by recirculation "indirect", is meant that the device may further include a reservoir for receiving the effluent fermentation substrate and arranged in fluid communication with the compartment to membrane reactor effluent. This hydraulic connection may be a communication with the outlet of the reactor effluent ung.part membrane and with the inlet of the reactor effluent membrane on the other hand. The reservoir may be designed to contain the effluent fermentation substrate prior to injection into the membrane reactor and receiving, continuously or discontinuously, the effluent from the reactor membrane. The reservoir may be for example a reactor jacket, for example provided with a stirring system, for instance also provided with heating and/or cooling, for instance also, provided with means for regulating the temperature and other parameters outlined herein, preferably automatic, for instance also means for adding nutrients useful at bacterial consortium for optimizing its activity of hydrogen production, for instance also for adding the bacterial consortium, in order to maintain optimal conditions of fermentation operated to produce hydrogen according to the present invention. Thus, these heating means, cooling, and other regulation, can be those known to the skilled person for the culture and/or maintenance of fermentation. This tank or reactor can also be fitted with a purging system the effluent prior to injection into the reactor membrane, degassing system may e.g. be a system by bubbling nitrogen into the reservoir. This tank or additional reactor can be used in particular for preparing the effluent fermentation substrate before entering the membrane reactor, for example for the homogenizing, for acclimation the bacterial consortium used, for the optional addition of nutrient compounds and/or bacterial consortium for initiating or renew the initial environment and depletion of microbial consortium, for adding water, for example when the effluent could not circulate otherwise, and more generally for control of the effluent and the regulation of its quality for the production of hydrogen in the reactor membrane. It can comprise means of substantially continuous stirring of the effluent, for its injection into the membrane reactor. These means can for example be in the form of blades mechanical mixing, the resistance of which is adapted to the viscosity of the effluent.

According to the invention, the circulation of the effluent between said effluent outlet and said inlet effluent can be achieved without any connection.

According to the invention, the compartments of the membrane reactor may be arranged successively from the bottom up according to the following order:

the inlet header,

- the compartment effluent, then

the outlet header.

In the device of the present invention, can also be advantageously providing means for injecting a sweep gas to the interior of the hollow fibers through said inlet header. The application of a sweep gas according to the invention unexpectedly increased by approximately 35% the yield of hydrogen. The inventors have found that the rate of extraction of gas products plays an important role on hydrogen production. According to the invention, the injection means of a sweep gas may include a gas reservoir to inject and means of control and regulation of the gas injected. It may also include means for collecting gas from the exploitation of the hydrogen from the gas produced by the fermentation, or recycled gas, and optionally filtering this recycled gas, to feed back the recycled gas as a sweep gas into the hollow fibers of the membrane reactor.

According to the invention, the membrane reactor may also be provided with means for heating and/or cooling of its contents, in particular of the effluent therein, and means for the control and regulation of the temperature of said effluent.

According to the invention, the extracting means hydrogen may advantageously include a means of separating the hydrogen gas from the reactor via the hollow fiber membrane, and a means of removing the separated hydrogen, this venting means preferably being adapted to be connected to a gas storage device. This means hydrogen separation may be conventional means known to the skilled person. It may be for example a means for removing carbon dioxide by adsorption by ethanolamides or by heat dissolution in a solution of carbonates, e.g. potassium. They can also be a selective adsorption on a molecular sieve bed ("PSAs" for "pression golf swing adsorber" or "TSAs" for "golf swing adsorber spa"),]. Or it can be a membrane technology, e.g. permeation [9]

According to the invention, the device may advantageously include means for heating and cooling means, and means for the control and regulation of the temperature, these means being able to regulate the temperature so as to be optimal, in particular for the activity of the microbial consortium in the presence, as defined below, for the production of hydrogen. These different means can be positioned at the membrane reactor and/or at the one or more reactor (e) (e) additional, if necessary.

According to the invention, the device may further include a storage container substrate or biomass, that can be added at additional reservoir, if desired, and be in fluid communication therewith. Such a container may be dedicated for example storage of the substrate or biomass, pending its use vja the additional reservoir. It may be a reactor for maintaining a storage temperature, as described in the present, or increasing the temperature for use of the substrate for the production of hydrogen according to the invention. A preparation of the fermentation can occur in such container, and an effluent fermentation substrate can be constructed for injected into the auxiliary tank, for regulating and acclimation for its injection into the membrane reactor.

The various elements of the device of the invention can be connected to each other, for the regulation of hydraulic communications, by a valve assembly, and, if appropriate control systems control the opening and closing of the valves.

The present invention also relates to the use of a device according to the invention for the manufacture of hydrogen, and a method for production of hydrogen which can advantageously be implemented with the device of the invention.

The effluent liquid fermentation substrate, which can be used as microbial inoculum, operable to implement the present invention, contains a set of microorganisms or "microbial consortium" to the origin of the phenomenon fermentation gas producer, whose hydrogen. The consortium already exists in the waste of natural origin or may be of or complemented artificially from a mixture of microorganisms known to the skilled person to produce hydrogen. For example, in the waste generated by the wine-making industry, as described below, the microbial consortium isolated predominantly comprises a microbial flora belonging to the genera The Enterobacter Klebsella KluyveraButtiauxella // / for the marc, e.g. grape marc and species A Clostridium saccharobutylicum, a Clostridium SPPs. (saccharoperbutyiacetonicum, beijerinckii, puniceum, diolis, roseum fed), streptococcal spp. (lutetiensis, infantarius, equinus, luteciae), Clostridium butyrlcum for the mires. The consortium can for example be used for hydrogen production by fermentation of agricultural biomass darkened, for example because it pre-exists in the substrate or biomass used, or because it is inoculated into said biomass to enrich it with active microorganisms for the production of hydrogen.

For carrying out the method of the invention, may be indeed emphasized that microbial consortium present in the biomass may contain hydrogenotrophic microorganisms that will consume the H2 for the synthesis of methane, e.g. hydrogenotrophic methanogens archaea, or acetate, for example bacteria homoacetogens. More strategies pretreatment of the consortium exist and can be used when carrying out the method of the invention to restrict the activity of these bacteria, for example by aeration, acid-base treatment, heat treatment or chemical. A thermodynamic point of view, the presence of H₂ in the reaction medium, inhibits the reaction for preparing the same. This leads thus to a manufacturing métabiques other pathways, for example for the production of other metabolites such as ethanol or lactate. It has been shown that it is possible to limit the negative impact related to a partial pressure of H₂ high on hydrogen production by including the use of inbred strains tolerant to hydrogen. The removal of hydrogen product of the reaction medium is an important parameter of the operation of the bioreactor to limit its consumption by the microorganisms and increase the efficiency thereof,. The device of the invention allows the extraction in a very advantageous manner, throughout the production of hydrogen in the membrane bioreactor, either freely, or, advantageously, according to the invention, by drive by means of a sweep gas, for example nitrogen from air or carbon dioxide.

According to the invention, the effluent fermentation substrate is liquid. If the biomass is too thick, it will add water, without requiring that such water is clean. It may be for example of water resulting from a process tank cleaning or river water or sewage treatment processes. This mixture of biomass with water may for example be made in a subsequent reactor as defined above, the resulting effluent is then injected into the membrane reactor for the production of hydrogen.

The fermentation temperature used may have a significant impact on the bacterial species active in the reaction medium. In pure culture, the performance of the bioreactor is thus bonded to the optimal conditions of the species or bacterial species used for the production of hydrogen, for example the mesophilic, the thermophilic, and the hyperthermophilic. In the case of using mixed cultures, this are the operating conditions which allow the emergence of bacterial species adapted, of the species present in the consortium overall. The use of a same inoculum will therefore in different temperature conditions, the development of different species, more or lesser performance for hydrogen production. Thus, depending on the microorganisms potentially present in the inoculum mixed, the optimum temperature of fermentation is adapted. The optimum temperature of the effluent during the implementation of the method of the invention can easily depending on the origin of the biomass or substrate, for example from an ID microorganisms in the presence and/or by starting from an effluent temperature of 35 to about 39 °c to and by finding by gradually varying temperature the temperature at which the hydrogen production is optimal. By way of example, when the biomass is original wine and/or when the microbial consortium produced therefrom or is similar thereto, the optimum temperature is about 37 °c + / - 2 °c. The temperature control can be performed, as described above, by means of temperature control, heating and/or cooling.

In the case of microorganisms, the pH of the effluent may also influence the fermentation process, and therefore the production of hydrogen. The pH of the effluent is preferably at optimum values for the fermentation. According to the invention, the pH of the effluent in the bin to effluent in the fermentation step can for example advantageously be maintained at a pH of from 5 to 7. by way of example, when the biomass is original wine and/or when the microbial consortium produced therefrom or is similar thereto, the optimum pH is about 5.5 + / - 0.2. Those skilled in the know determine, from culture media made on different boxes of kneaded or in suspension, for example in strict anaerobic conditions or not, and at different pH values determine the pH region most suitable for implementing the method of the invention.

According to the invention, in use of the apparatus or the implementation of the method, the produced hydrogen which is conserved in the day of fiber hollow fibers can be driven freely means hydrogen extraction by the thrust of gas produced in the day hollow fibers or, advantageously by means of a sweep gas, preferably neutral, c'est to say not chemically reactive under the operating conditions of the reactor with the hydrogen produced, e.g. nitrogen, or for example carbon dioxide, emanating from the PCV products. The flow of the sweep gas is selected to drive the produced gas having passed through the wall of said hollow fibers. The transmembrane pressure difference to allow transfer of gas products of said compartment to effluent to the day of the hollow fibers.

According to the invention, the sweep gas may be injected into the day of fiber hollow fibers at a pressure or a flow rate just sufficient (th) for pushing the gas, produced by fermentation in the membrane reactor, which is conserved in the day of fiber hollow fibers. This avoids produce too much gas to be treated for the removal of hydrogen from the reaction tank membrane according to the invention. The sweep gas flow is therefore adjusted in accordance with the gas production fermentation using the present invention.

The device, the method of the invention thus allow, by the separation of product gas from the effluent by the walls of the hollow fibers, and advantageously with further scanning by a sweep gas, avoiding the presence of H₂ in contact with the effluent or dissolved in the effluent, this gas inhibiting production of these gasses. The yield of hydrogen production is significantly improved by the present invention.

Further, by the method of present invention, unexpectedly, the inventors found also a production optimization organic acids, for example acetic acid, butyric acid, lactic acid, fumaric acid, succinic acid, as well as optimization of production of alcohols, for example ethanol, butanol, and the like, along with the hydrogen, the present invention also relates to the use of the device of the invention and the method for the manufacture of organic acids and/or alcohols. In the case where the production of organic acids and/or alcohols is the aim of carrying out the present invention, the elements relating to the extraction of hydrogen are not useful, and can be optionally employed or incidentally. The requirements for such production fermentation of organic are for example those developed in the present, geared for the production of these products, including with the scanning of the day fiber hollow fibers by a sweep gas.

The use and method of the present invention may comprise the following steps:

immersing hollow fibers such as defined herein in a liquid effluent fermentation substrate as defined in the present, the immersion is carried out in a compartment to effluent as defined herein,

fermenting the substrate to generate a gaseous mixture,

recovery of the gas mixture through the day of fiber hollow fibers,

removing the hydrogen from the reformate gas mixture through the day of fiber hollow fibers.

In other words, according to the invention, the manufacture of hydrogen according to the method or using the device of the present invention may include:

a step of filling the compartment to effluent by an effluent liquid fermentation substrate, so that the hollow fibers housed inside the compartment effluent is immersed in said effluent,

a step of fermenting the substrate inside the reactor membrane,

a step for collecting gas produced by the fermentation step.

According to the invention, the fermentation step and said collecting step can advantageously be continued leaving accumulate deposits of particles and/or microorganisms to the outer surface of the fibers and therebetween. The inventors have found unexpectedly that the accumulation of these deposits of particles and/or microorganisms would be seen immobilizing the microorganisms in the membrane reactor and supplying a benefit to the production of hydrogen from a stirred reactor in continuous operation.

According to the invention, when the device of the invention includes an additional reservoir in fluid communication with said compartment effluent, said use or method may advantageously include:

a step of filling said tank with said effluent liquid fermentation substrate,

a step for communicating with the compartment of the tank effluent, following said step of filling the compartment effluent with the effluent liquid fermentation substrate,

a step of continuously supplying the effluent compartment liquid effluent fermentation substrate from said reservoir.

Thus, according to the invention, the device, the method of the invention allows a continuous production of hydrogen, including, if desired a continuous circulation of the effluent fermentation substrate in the membrane reactor.

According to the invention, the internal temperature of said additional reservoir, when it is used to store the waste or biomass before use, is preferably controlled to be less than 10 °c, preferably less than 5 °c, for example about 4 °c. This reduces the need for engaging the fermentation process in said further tank, c'est to say before entering the membrane reactor or bioreactor.

According to the invention, the method or the use may also or in addition:

a step of stopping production of hydrogen, then

a step in which the membrane reactor is purged of liquid effluent fermentation substrate, and then

a step of cleaning the effluent compartment and hollow fiber, performed to maintain a live microbial film on the outer surface of the hollow fibers, and

a new step of filling the compartment to effluent by the effluent liquid fermentation substrate with the same hollow fibers and said bacterial film preserved.

According to the invention, the step of cleaning the effluent compartment, in particular hollow fibers, is cleaning which preferably do not damage the hollow fibers. Rinsing may be done with, for example with a pH which can be adjusted for preservation of the biofilm formed by the bacterial consortium on the fibers.

This water may contain a mild detergent, kind of RBS, preferably much diluted, or bleach with chlorine to 1 ppm.

Tests for fermenting substrate have been made thereby, without additional external microbial bacterial inoculum, using biofilm created on the surface of hollow fibers in contact with the effluent during previous fermentation tests, with equally satisfactory results than obtained with an initial sowing. The fermentation is quickly started and then has stabilized for a continuous production of hydrogen. Yield and productivity in hydrogen important were obtained, highly competitive with the tests using a freshly obtained from bacterial inoculum to the sewage treatment plant, revealing the interest of the membrane device of the present invention for stabilizing the bacterial consortium, and more widely microbial, hydrogen producing on the hollow fibers. In addition, these results demonstrate a simple large-scale implementation of the inventive method, by limiting the needs of reseeding of the reaction medium.

According to the invention, the use or method may advantageously include a step of degassing of the liquid effluent fermentation substrate prior to its introduction into the compartment effluent, in particular when the effluent or biomass is in an additional reactor as described above.

According to the invention, the use or method may advantageously comprise an acclimation time biomass and/or the effluent from fermentation substrate containing a bacterial consortium or microbial control in the device of the invention, in particular in the membrane reactor and/or, where appropriate in the additional reactor as described above, e.g. of 5 hours. These include an initial phase of acclimatization of microorganisms in the device of the invention, phase which is for example preferred at start of continuous fermentations in order to rapidly reach optimized production of hydrogen. Those skilled in the know easily determine the duration of the acclimation time which precedes somewhat production phase of hydrogen, e.g. by the fermentative activity in the effluent fermentation substrate. This phase may correspond for example to the initial phase of a reference test operating mode "a semi-batch process", once fermentation to its maximum production, e.g. after about 4 to 6 hours in the bioreactor, the fermentation reactor effluent containing fermentation substrate climatized can then be continuously supplied in substrate or biomass.

According to the invention, such a degassing and/or such a time has (have) acclimation advantageously allowed to further improve performance of the device, of the method of the invention, in particular by allowing a continuous and efficient production of hydrogen for more than 400 hours.

The inventors have thus developed a device and method economical and clean hydrogen production which will certainly accelerate the placement of a "hydrogen economy".

The device, the method of the present invention provide a very favorable energy gain from ultimate waste, little upgradable by other routes. The supply successful nutrient with waste agricultural biomass provides validation of industrial use.

The production of H₂ darkened by fermentation using the apparatus or method of the present invention, is less energy-consuming and can be performed under non-sterile conditions, using mixed organic waste inexpensive as substrates.

In the continued operation of the device of the invention has been successful and has allowed the passage to the operation of the membrane module, not only by interlocking with an additional reactor, but also alone as a membrane bioreactor.

Further, the use of the membrane bioreactor (BRMs) of the present invention alone ensures excellent perspective because it shows better stability of hydrogen production of prior devices, especially with continuously stirred tank reactor (RACs).

Further, the use of agricultural biomass used by the inventors of the present, whose annual production of 850,000 tons is French, would produce 20,000 000 M.³ hydrogen per year.

Also, the device, the method of the invention can easily and advantageously adapt to the flow of sludge from treatment plants and organic waste metabolizable available locally.

Still other advantages will appear to those skilled in the reading examples, illustrated by the figures, given by way of illustration.

Figure 1 schematically depicts a longitudinal sectional view of an exemplary membrane reactor usable in the device of the invention and/or for implementing the method of the invention.

Figure 2 schematically shows an overview of a first exemplary device of the present invention.

Figure 3 schematically depicts an overview of a second exemplary device of the present invention.

In this example, a membrane reactor, hereinafter "membrane module", hollow fiber usable for the extraction of gas from the effluent according to the invention.

This reactor is shown schematically in Figure 1 appended. It comprises:

a housing 20 defining an interior volume, said interior volume comprising two sealed compartments therebetween, namely an outlet manifold 26 bis and an effluent compartment 24, also called calender, to be filled by the liquid effluent fermentation substrate;

hollow fibers 28 each formed by a tubular wall surrounding a day of fiber, said tubular wall being liquid-tight and gas-permeable, hollow fibers are housed in said effluent compartment and arranged so that their day fiber communicates with said outlet manifold is sealed with respect to the effluent liquid when present in the effluent compartment, and

a gas outlet 30 communicating between said outlet manifold and said extraction means hydrogen.

an effluent inlet 34 fermentation substrate communicating between the exterior of said casing and said compartment 20 24 effluent, and

an effluent outlet 38 for fermenting substrate communicating between said effluent compartment 24 and the exterior of said shell.

The compartment effluent, the hollow fibers, the outlet manifold, the gas outlet and the means of extracting hydrogen are arranged such that the passage of a gas or a gas mixture from the liquid effluent when in the compartment effluent through the day of fiber hollow fibers into said means hydrogen extraction (not shown in Figure 1) to move freely or by means of a sweep gas.

The envelope 20, including effluent to the compartment 24, is made of a plastic material it may particularly include a calender metal, or formed of another material such as PVC.

The envelope 20 is opaque to allow dark fermentation effluent in the compartment 24.

The membrane module used is composed of hollow fiber 238 polytetrafluoroethylene (ePTFE), an inner diameter of 0.45 mm and an outer diameter of 0.87 mm and pores whose diameter is of the order of 0.1 MW.

These are arranged in a volume filling ratio of 8.5%. In one example embodiment the useful volume is 478 ml.

In the illustrated example, the interior volume of the shell 20 includes an inlet header 22 of bis. The effluent compartment 24 is housed between the inlet header 22 bis and the outlet header 26 of bis, the inlet header 22 of bis being sealingly separated from said effluent compartment 24.

The gas input manifold includes a gas inlet 32, especially the connection to a gas injection means.

Each hollow fiber 28 is arranged to pass through the compartment effluent 24 and so that the collectors 22 bis and 26 of bis intercommunicate via the day of fiber.

The hollow fibers are held together on one side by a first sleeve 22 at the inlet header 22 bis a sweep gas, on the other hand a second sleeve 26 at the outlet manifold 26 bis-gas. These two sleeves 22 and 26 is epoxy resin and ensure tightness at the fiber which pass through them, so as not to let the liquid effluent at the inlet manifolds and gas outlet, in communication while leaving free the day of hollow fiber with the input manifold 22 bis and the outlet header 26 bis-gas.

The stream of liquid that can be circulated in the compartment effluent 24 of this module, the module being delimited by the calender, from bottom to top, between the inlet 34 and its effluent outlet 38, allows the relative suspending solids potentially present in the effluent liquid fermentation substrate.

As may be seen in Figure 1, the input effluent the effluent outlet 34 and 38 are transverse, i.e. they open on the sides of the compartment effluent.

Here the input effluent 34 is near the sleeve 22 between the part for flue 22bis 24 of the inlet manifold. The distance between the inlet 34 and the sleeve effluent 22 defines a dead volume in the height at which arrives the effluent. There is little, if any, effluent flow in the dead volume. The dead volume can be used to leave the possibility of depositing particles of solids from the effluent or to enable development of a bacterial consortium also at that level.

In an alternative embodiment, not shown, instead of arriving transversely, the effluent outlet and/or the inlet (s) effluent arrives respectively from above and/or below.

For example, the effluent outlet arrives at the top of the calender, beside the gas outlet but without communicating with it. In this case, the effluent outlet comprises a conduit which passes through the outlet manifold, while being sealed with the interior of the outlet manifold. This line then passes through the sleeve between the outlet manifold and the compartment to effluent and runs into the effluent, either flush with the sleeve, or remote from it.

Similarly, the input of effluent can arrive at the bottom of the calender, beside the gas inlet but without communicating with it. In this case, the input effluent comprises a conduit which passes through the gas inlet manifold, while being sealed with the interior of the gas inlet manifold. This line then passes through the sleeve separating the gas inlet manifold and the compartment to effluent and runs into the effluent, either rastfe-EE is sleeve 26, or remotely therefrom, providing in the latter case a dead volume.

Placed in an upstanding position, a top gas forms in the upper chamber of the membrane reactor, above the liquid outlet 38. For example, the top gas has an average volume of 55 ml in the example embodiment where the working volume is 478 ml. A volume of liquid is thus present in the 423 ± 5 ml volume provided for the effluent in the calender of the membrane module of this example.

Heating the compartment to effluent to the fermentation temperature is achieved by water circulation of a thermostated bath (5a - PolystatBioblock vein) in a coil with a compact composed of a flexible tube surrounding the calender vinyl.

This module or membrane bioreactor (BRMs) according to the invention was tested in several devices according to the present invention. Examples are shown below.

The system used is defined so as to operate in a continuous mode. It is represented in Figure 2 appended.

It comprises a supply tank 10 double-walled glass walls ovenized Scilabware (trademark), which can contain a solution or fermentation medium substrate. It is used to power the membrane reactor of the example 1, or membrane bioreactor (BRMs) 2, at a controlled rate by a peristaltic pump 70 (Ismatec-IP, trademark).

A circulation of the environment is performed with a second peristaltic pump 61 in a recirculation loop 6 on which is formed the addition of the substrate solution or biomass 75 and a discharge rate equivalent of the effluent exiting excess, in order to regulate the volume of the medium within the BRMs 2. to maintain the conditions of anoxia strict in the reaction medium, a continuous degassing of the substrate or biomass is carried out by bubbling nitrogen (not shown) in the reservoir 10 of feed solution.

To avoid the fermentation of the substrate in the supply tank 10, a liquid coolant from a thermostatically controlled bath 56 (dewar tc40 - Huber and coupled to a Polystat 5a - Bioblock vein) circulates between the double walls of the supply tank 10. The thermostatically controlled bath can be connected to a heat control means 58, for maintaining the temperature of the puddle to 4 °c.

This allows the fermentation of the substrate starts essentially, if not exclusively, in the membrane bioreactor BRMs 2. the supply tank 10 sidearm in series to the recirculation loop 6 via a supply line 73. This supply pipe 73 joins therefore the pipe 60 of the recirculation loop 6, which joins the effluent outlet 38 to the input effluent 34, thereby allowing the arrival of new substrate to the effluent compartment 24.

The supply line includes a supply valve 72 and the peristaltic pump 70, which in each case one opening and the flow rate are monitored so as to continuously supply the membrane bioreactor effluent 2, substrate or biomass.

To control the pH within the effluent compartment 24, the pipe 60 of the recirculation loop is connected to a container filled with a basic solution 64, for example a sodium hydroxide solution (NaOH solution). A probe 62 pH measuring the pH value in the line 60. A peristaltic pump 66 controls the flow rate of basic solution in the recirculation loop 6, based on the measurements received by the probe 62 pH. This test can be automated for example.

A pressure sensor 68 measures the difference in pressure in the line 60 of the recirculation loop 6, between the effluent outlet 38 and the inlet effluent 34.

In this example, a sample valve 74 on the pipe 60 of the recirculation loop 6 optionally comprises samples for microbiological analysis. A sampling valve 74 of bis on channeling the flow outlet 75 optionally comprises samples of the effluent leaving, for example for measuring the sugar content.

The pipe 60 may include also the outlet stream 75 arranged to discharge the surplus liquid in the recirculation loop 6. example, an outlet free by overflow to vent the excess effluent fermentation substrate.

A heat transfer fluid or liquid, for example water, from another thermostatically controlled bath 52 (PolystatBioblock ITU - 5a

- marketed by that boundary) circulates in a coil around the calender 24 BRMs 2. the thermostatically controlled bath 52 is connected to a heat control means 54, for maintaining the temperature of the bath between 35 °c. 39 °c and the coil can be formed by a flexible tube made of vinyl. It can be arranged compactly around the support, in particular with its contiguous turns.

The means for injecting the sweep gas 8 is connected to the gas inlet 22 of the inlet manifold 32 of bis. This injection means 8 includes a source nitrogen 48, thus qtHune valve and a mass flowmeter 50 (to Brooks implement - model 5850e) coupled to a regulator (implement to Brooks - 0254 model) to control the flow of nitrogen.

The sweep gas, here nitrogen, flows from the inlet manifold to the outlet manifold 22 bis-substituted bis 26, via the days of fibers. It then flows out through the gas outlet 30, which is connected to a gas extraction means 4.

The number of traps to cold or cold traps is not limiting, and the extracting means 44 may comprise two cold traps, operating in particular between 2 and 3 °c and a separator gas/liquid membrane is installed in series with the line BRMs gas outlet 2.

The cold traps are immersed in a bath whose temperature is regulated e.g. by a cryostat tc40 brand HVS. The separator may be for example an elf 3s5sec25 marketed by single CTB Choffel. They are, preferably, arranged so as to condense the liquid prior to the arrival of gas flow towards the online analysis by PGCs to-DCT.

The extraction means 4 comprises a cold trap gas 44 and a cryostat 42 arranged to trap any condensates.

The extraction means 4 also optionally includes a means for separating hydrogen from other gas, especially carbon dioxide.

The extraction means 4 comprises a means of removing the separated hydrogen, especially a vent 12, adapted to be, in particular, connected to a gas storage device.

A gas chromatograph 46, for example a micro-chromatograph gas phase comprising two modules have sensors to thermal conductivity (also known as the GC-DCT, for "automated gas Chromatography - spa Fluoride Detector"), is connected upstream of the discharge means 12 and downstream of the separating means (not shown) enable analysis of the generated gas. Said micro-chromatograph may for example be of the Agilent's mark.

The makes and models, in particular apparatus, measuring devices and reservoir used in this example are exemplary non-limiting. They correspond to the templates used in the performed tests in the laboratory for this example.

According to another example, illustrated in Figure 3, the recirculation loop 106 is arranged so as to allow recirculation of the substrate between said indirect effluent outlet 38 and said inlet effluent 34.

In this example 3, the supply tank 10 and its cooling means (not shown in Figure 3), the membrane bioreactor (BRMs) 2 and its thermal regulation means 52, 54, and the gas injection means 8 are identical to those of the example 2. their references are therefore identical and their description will not be resumed in detail.

In the example 3, the recirculation loop 106 comprises an additional reactor 103 in hydraulic communication with the BRMs 2. this reactor is for receiving the effluent fermentation substrate.

This hydraulic connection may be a communication with the outlet of the effluent 38 BRMs 2 on the one hand, and a fluid communication with the effluent inlet 34 BRMs 2, on the other hand. This further reactor 103 may be designed to contain the effluent prior to injection into the BRMs 2 and receiving, continuously or discontinuously, the effluent from the BRMs 2.

This further reactor 103 is, in this example a double jacket reactor provided with a stirrer 101 with blades to agitate continuously the effluent, for its injection into the BRMs 2. the further reactor 103 thereby forms a continuously stirred tank reactor (RACs).

For example, the further reactor 103 may be a reactor of the mark Büchi of Ag. For example, for a 2 478 ml of BRMs, the supplemental reactor 103 may have an internal volume of 1.0 to 1.5 liter.

A heat transfer liquid from a thermostatically controlled bath 156 flows between the double walls of the additional reactor 103. The thermostatically controlled bath is connected to a heat control means 158, for maintaining the temperature of the bath between 35 °c and 39 °c.

The additional reactor 103 may comprise a temperature sensor 159, for example connected to a heat control means 158.

To control the pH within the additional reactor 103, and thus into the recirculation loop 106 and in the BRMs 2, the supplemental reactor 103 is connected to a container filled with a basic solution 164, for example a sodium hydroxide solution (NaOH solution). A probe 162 pH measuring the pH value in the pipe connecting the effluent outlet 38 BRMs 2 additional reactor at 103. A régulateurautomatique pH 166 actuates a valve connected between the container 164 and the further reactor 103.

As a pH regulator can be used, for example, the model ΔBL 7916 sold by Hanna's tools.

In order to improve the homogenization of the fermentation medium prior to injection into the reactor are, a circulation of the reaction medium, i.e. the effluent fermentation substrate, is performed with a peristaltic pump 161 in the recirculation loop 106, on which is formed the addition of substrate or biomass. This pump 161 is connected in fluid communication input to the additional reactor 103 and in output to the inlet 34 of the effluent BRMs 2. the valve 160 on the recirculation loop 106 to cut fluid communication. This recirculation valve 160 and 161 thus allow the peristaltic pump, by control respectively of the opening of the valve and the flow through the pump 161 of control the recirculation.

A pressure sensor 168 measures the pressure difference in the pipes of the recirculation loop 106, between the effluent outlet 38 and the inlet effluent 34.

In this example, two sampling valves 174 are arranged respectively on the additional reactor 103 and on the pipe connecting the pump 161 of the recirculation loop 106 to the input effluent 34, e.g. to measure the sugar content.

The recirculation loop 106 may further include a flow outlet arranged between the 2 and the further reactor BRMs 103, to drain any surplus liquid in the recirculation loop 106. For example, a valve (not shown) connected to a peristaltic pump 163 can control this discharge at a rate equivalent of the effluent or fermentation medium of excess substrate, in order to regulate the volume of the medium within the BRMs 2 and in the further reactor 103.

Another sample valve 174 of bis on the pipe allowing discharge of stream optionally comprises samples of the effluent leaving, for example for measuring the sugar content.

This further reactor 103 can also be fitted with a purging system for fermenting the effluent prior to injection into the substrate 2 BRMs, degassing system may e.g. be a system 147 and 150 by bubbling nitrogen into the reservoir. The system includes for example a source nitrogen 147, and a valve and a mass flowmeter 150 to control the flow of nitrogen.

The supply tank 10 supplies the BRMs 2 at a controlled rate by a peristaltic pump 170 (Ismatec-IP, trademark). This supply tank 10 can also be fitted with a degassing system (not shown) of the substrate or biomass before it is injected into the pipe 173.

The supply tank 10 is connected in series with the recirculation loop 106 via a supply line 173. This supply line 173 therefore joins the piping of the recirculation loop 106, in this example between the recirculation pump and the inlet 161 effluent 34, thereby allowing the arrival of substrate or biomass in the compartment effluent 24. la supply line includes a supply valve 172 and the peristaltic pump 170, which in each case one opening and flow rate are controlled so as to continuously supply the BRMs 2 in substrate or biomass.

The gas injection means 8 scan is connected over the gas inlet 32 of the inlet header 22 of bis in the same manner as for the example 2 and will not be further described.

A gas extraction means 104 is connected to the gas outlet 30 of the BRMs 2 and to a gas outlet of the additional reactor 103, via two circuits 141 and 143 corresponding gas. This allows also the recovery of hydrogen would be produced in the supplemental reactor 103.

This means gas extraction 104 includes a set of a cryostat cold traps 144 and 142 arranged to trap any condensable compounds ("condensate" once retained) within gas and corresponding 141 and 143.

The extracting means 104 also optionally includes one or separation means (not shown) of hydrogen other gas products, especially carbon dioxide.

The extracting means 104 comprises a means of removing the separated hydrogen, especially a vent 112, can be, in particular, connected to a gas storage device.

In this example, each of the gas flow emerges on this venting means 112, so that the hydrogen from the reactor and BRMs 2 additional 103 can be collected from a single discharge means 112.

Two-gas chromatographs to thermal conductivity detector 146 and 147 are connected upstream of the discharge means 112 and downstream of the separating means, for allowing ' analyze the gas products.

The makes and models, in particular apparatus, measuring devices, reactor and reservoir used in this example are exemplary non-limiting. They correspond to the templates used in the performed tests in the laboratory for this example.

Providing nutrient compounds to the reaction medium or effluent fermentation substrate has rendered it renew the initial medium (or aclimaté), and depletion of nutrients detrimental to the production of hydrogen. It has been carried out, either by adding spot directly on the recirculation loop, either by continuous addition of the feed solution of substrate or biomass.

The use of mires from the die or wine as real substrate or biomass feed microbial or bacterial consortium was produced with the membrane module of the example 1.

When using a medium rich in carbohydrates and nutrients (mixture model or mires), a microbial or bacterial growth is observed in the supply tank 10 in substrate at room temperature. Therefore, a reduction in the temperature of the tank was made at about 4 °c. In addition, thorough cleaning of the tank is performed currently in the case of contamination, by brushing with detergent and rinse with 70% ethanol. This intervention is being fermented, without interruption of power, through the use of a buffer volume of substrate or biomass.

In addition to the cleaning procedure after use, the cleaning of the membrane bioreactor is achieved by repeated flushes and emptying of the calender with bleach (1 ppm of active chlorine). This washing remove particulates present in the calender, but its purpose is not to destroy the biofilm is carried on the hollow fibers during fermentation, to maintain its potential activity. Reseeding the membrane reactor 2 can still be produced at the beginning of each test fermentation with activated sludge wastewater treatment plant acclimated, for example in, in this exemplele additional reactor 103. A fermentation test without application of inoculum outside has been performed, with supply power substrate containing nutrients, to observe the potential of this biofilm production for hydrogen production by fermentation dark.

A phase of acclimation of the microbial or bacterial consortium have shown to be advantageous before the continuous hydrogen production device of the invention. The acclimation of activated sludge is carried out in a stirred reactor in operation a semi-batch process (rSBA) with the optimized parameters by microorganisms in the presence function. The acclimation period is 5 hours, duration corresponding to the time required for the test performed on the biomass to reach its maximum gas production. And then, once the bearing for producing hydrogen reaches, the RAC is continually supplied by the effluent fermentation substrate from a storage, and a continual withdrawal of the effluent leaving the reactor is designed to compensate for the supply maintaining a volume of effluent in the (RACs) substantially constant. The mode of operation of this example is not shown. A recirculation loop is placed, on which the effluent leaving is removed and the substrate is added, thereby improving the homogenization of the reaction medium in the reactor membrane.

Nitrogen is used as a flushing gas with a flow rate of 10 ml/min for the extraction of the gas produced during fermentation.

A fermentation test-rac has has been performed in the RAC with feed rate to a substrate (SAR) of 0.62/ghexoseLmiiieu/H and a hydraulic residence time (TSHR) of 73 hours. This test has enabled the production of H2 period to 240 hours.

A reduction in the production of H2is observed after about 60 hr, followed by rising wide toward a bearing production relatively stable. However, a significant increase in the production CO2 is observed, being considerably higher than that of H₂ from 140 hours. The production profile has thus been divided into three phases for producing stable whose corresponding production parameters are given in table 1 below.

Table 1: yield values and in productivity by hydrogen, the mole ratio of H₂/ C0₂ and consumption of hexose operation for testing the rac rac has

* the first value corresponds to the average yield of hydrogen per mole hexose assumes an accumulation in the reactor; the second value corresponds to the average yield of hydrogen per mole hexose assuming a steady state.

The first phase of fermentation studied, between 7 H and 60 hr, present the hydrogen yield the highest obtained with 1.95 mole_H2/molhexose consumed with a productivity of 158 rnLfWLmiiieu/hr. This first phase of operation additionally has a ratio H2COEXTRUDED /2 large 1.13, superior to results observed during tests of a fermentation mode of operation in a semi-batch process stirred reactor (rSBA).

The second production phase identified is between 70 108 H and H fermentation after an important decrease of the efficiency and the productivity in hydrogen (46 - 54% et - 57% respectively) whose minimum value corresponds approximately to the TSHR applied to the system (73 hr) or a total renewal of the reaction medium.

Finally, a third phase, longer, will occur after 160 H with an efficiency and productivity similar to those of the second phase and a H₂/ C0₂ stabilized around 0.65, down more than 30% with respect to the two preceding phase.

The decrease in productivity in hydrogen between the two phases to be attributable as none nutrient except glucose is added to the medium at réactionnelà effluent fermentation substrate (not source of nitrogen or phosphorus) and that a part of the consortium or microbial is extracted from the (RACs) by controlling the volume therein. The substrate solution is stored in static air, an assumption that the reaction medium is in microaerobic condition can also be advanced; introducing oxygen into the bioreactor RACs can induce a change in the microbial or bacterial consortium, explaining the reduction in the production of H₂ the benefit of the CO-2. Finally a third hypothesis, relates the accumulation of metabolites of fermentation in the reaction medium, with concentrations approximately 10 times higher in ethanol, acetic acid and butyric acid with respect to an operation of the stirred reactor in semi-batch process. The metabolites in significant concentration limit thermodynamically production and therefore the coproduction of hydrogen.

In optics improve production results by limiting the drop in productivity by hydrogen and oxygen ratio₂/ C0₂ , a fermentation test-rac b has been performed by a modification of the configuration by using a global TRAb 12 hr, performed by the increase of the volume flow supply substrate. For this test the rac b, the concentration of the feed solution was reduced to 12/50 ghexoseghexose an L/L for the test-rac has, to obtain a DAS about 1/hr/ghexoseLmiiieu, therefore increased by 60% relative to the das ghexoseLmiiieu 0.62 // H of the rac-testing. The reduction of the TSHR should allow an improvement, notammentd ' increasing turnover of the reaction medium, by limiting the accumulation of metabolites and their potential negative impact on the production of hydrogen. The microaerobic, which can be generated by a small but continuous supply of dissolved oxygen in the substrate, creating conditions favorable to the emergence of metabolic pathways or competing. In order to avoid the feed of oxygen into the reaction medium, a continuous degassing is performed in the reservoir of the substrate.

To better ascertain the impact of the TSHR on the fermentation, an increase by step (12 hours, 18 hours and 23 hr) is performed as the fermentation test Rae-b receptor. The modification of the TSH is performed by modifying the substrate concentration of the feed solution and the volumetric feed, in order to maintain a steady SAR (about 1/hr/ghexoseLmiiieu). For keeping a recirculation rate fixed, the rate of recirculation is modified so as identical to that of the power supply.

The production profile obtained during testing fermentation globalStnent Rae-b is more stable than for the fermentation test-rac has, notwithstanding the modifications TSH performed. The production rate of H₂ and c0₂ remains close throughout the fermentation, without significant inversion, thereby confirming the positive impact operating conditions implemented.The following table 2 gives the results obtained for the use of the different TRAb used with the RAC. It should be noted that the results obtained for a TSH 73 hr, tested by the fermentation test-rac has, correspond to only the initial period of the test (of 7 to 60 hr) and therefore do not address the decrease production of H₂ .

Table 2: yield values and in productivity by hydrogen, the mole ratio of H₂/ C0₂ and DP hexose sugar consumption Err and function of hydraulic residence time (TSHR) for the RAC

On the three-TSH tested during testing fermentation Rae b, 12 hr TSH achieves the best production results of H2, with a yield of 0.93 moWmolhexose consumed and productivity of 47 ml _H2/hr / L miiieu. However, on the one hand relatively short h.gxose added is consumed by the fermentation (38%). The increase in the TSHR 18 hr to results from an important decrease of productivity (-43%) H₂ , with a slight decreased consumption of hexose 32%, leading to a drop in yield of hydrogen per mole hexose consumed (-30%). A new drop in the economic efficiency of 17% is observed when increasing the TSHR to 23 hr, the hexose (35%) consumption remains relatively low. The ratio H₂/ C0₂ is good (1.0) for TSH 12 hr, and remains large (0.88) for the TSH 18 and 23 hr, 160 H of operation. Compared to the result obtained on the initial phase of the test-rac has, the hydrogen yield is 52% lower with a TSH 12 wherein HET productivity by hydrogen is greatly reduced to 47 ml of 158 Lmineu ^ // hr. This is due to the rapid renewal of the reaction medium with a TSH-short, the bacterial population initially present in the bacterial inoculum is leached and did not develop sufficient time for consuming the entire substrate, as in the case of testing-rac b for which the reaction medium has not been renewed (TSH 73:00).

Microbiological analysis of the bacterial consortium present in the reaction medium of the RAC was made, for three samples:

with a TSH 73 hr (rac has) to 29 hr (initial phase) and 241 H fermentation (final phase is stable), and with a 12 hr TSH (RACs b) to 68 hr (table 3) fermentation. These results demonstrate that the majority of the bacterial consortium is composed of bacteria of the phylum Of Remediation, with more than 75% of the sequences analyzed. On the initial phase of fermentation, with a TSH 73 or 12 hr, the profile of the bacterial consortium is very similar, with the vast majority of bacteria of the genera A Clostridium Saccharomyces sensu stricto to 1 and 10 (about 91% sequences analyzed) with the common majority species Clostridium butyrlcum (29.2% and 52.8% respectively for the TSH 73 H and 12 hr).

Table 3: results obtained by sequencing of samples of biomasses fermentation tests rac-and rac-β - groups having an abundance>1% for at least one of the samples

However, other species of bacteria of genus Clostridium Difficile demonstrate in these two samples differ in their distribution, with a considerable portion of A Clostridium barati (26.9%) for TSH 73 H and a species indeterminate A Clostridium (// aminomutase thiosulfatireducenssuifidigenes) for TSH 12 hr. This variability can be explained by the evolution of the consortium as a function of time, in effect the test sample is taken from rac has 29 hr while the DFT Rae-b is taken to 68 hr. Wherein the species predominantly identified during tests rSBA is A Clostridium barati, and remain still important to 29 H fermentation, while less than half of the reaction medium was renewed. While in the case of the sample using a TSH 12 hr, removing a 68 H is conducted after more turnovers of the reaction medium, the bacteria present are therefore the more suited to continuous operation. The Species A Clostridium barati thus appears to be less well-matched to the continuous mode of operation and its population decreases as of renewal of the reaction medium, to become relatively minor (3.2% in the sample Rae-b to 68 hr). In contrast, the species Clostridium butyrlcum, where the population was less than 10% in operation rSBA, is evolving in the initial phase of the RAC-(29.2%) and becomes the predominant species after establishing the continuous operating mode in the sample of the RAC-b receptor (52.8%). Similar reasoning can be performed with regard to the occurrence of species A Clostridium aminomutase and (// aminomutase thiosulfatireducenssuifidigenes) and the disappearance of the unidentified species of A Clostridium SPs. (nº n°: lc020510).

The obtained sample 241 H fermentation during testing fermentation rac has, therefore after the relative increase in the production of c0₂ compared with that of H₂ , has a greater microbiological diversity, with the presence of the phylum from other species (Actinobacteria and Bacteroidetes) in significant quantity (>10%) and the significant reduction of frequent species in the sample collected 29 H fermentation: Clostridium butyrlcum (2%) and A Clostridium barati (0.6%). In contrast, other bacteria of the genus Clostridium Difficile have emerged with respect to the sample of the initial phase, in particular the species indeterminate A Clostridium (// aminomutase thiosulfatireducenssulfidigenes) whose representativeness increases of 0.3 to 12.0%. The emergence of other species of the phylum Of Remediation is also observed, with species of the genera Ruminiclostridium (16.4%), Sporolactobacillus (13.5%) and Ethanoligenens ( 10.9%).

Example 6: example of operating the membrane bioreactor (BRMs) according to the invention with a continuously stirred tank reactor (RACs) as in the previous example - operating mode/BRMs Rae

After identifying conditions permitting operation continuously effective in the RAC, in example 5, the addition of the membrane bioreactor (BRMs) system has been experienced to verify the improvement of the extraction of the product gas, and potentially increase the production of hydrogen.

In a first step, the membrane module has been added to the recirculation loop of the RAC, to couple these two devices, so as to correspond to the device of Figure 3, in example 3. in the same way as for the RAC (example 5), acclimation of activated sludge wastewater treatment plant is performed in the stirred reactor in operation a semi-batch process (rSBA), then once the bearing production H₂ reached, a portion of the sludge acclimated is transferred to the calender 24 of the membrane module 2. this test has improved and stabilize the operation of the membrane module in membrane bioreactor.

The membrane bioreactor (BRMs) or membrane module 2 of the present invention was coupled with an additional reactor 103 which is continuously stirred tank reactor (RACs) forming the device with coupling of Figure 3 (the rac/BRMs).

In this configuration with coupling, a recirculation of the effluent fermentation takes place between substrate and second portions of the system: the RAC 103 and the BRMs 2. both parts of the system are equipped with a gas feed scanning (from gas sources 48 and 147) with a flow rate of 10 ml/min and a gas outlet with on-line analysis (by the microchromatographes 146 and 147). In the case of BRMs 2, nitrogen is circulated within the hollow fibers 28, the reaction medium being located outside, in the calandre24. The samples of reaction medium taken by the valve 174 background of the RAC 103 correspond to the output of the RAC and hence to the input of the BRMs, whereas the samples taken in the effluent outlet by the valve 174 of bis correspond to the input of the RAC. As well as the test-rac has (example 5), this preliminary testing Rae/BRMs was performed without continuous degassing the supply reservoir substrate 10.

An initial volume of 1200 ml of activated sludge is used as microbial inoculum or to enable filling of the shell 24 of the BRMs 2 while maintaining in the RAC 103 a volume of liquid for pH control and regulation thereof. The volume of effluent for fermenting substratdans 24 during operation of the calender is about 440 ml, while an average volume of 670 ml is maintained in the RAC 103 over the test period. The addition of substrate is performed on the recirculation loop 106, prior to entry into the calender 24, thereby resulting in the increased production in this part of the system. The hydraulic residence time (TSHR) overall system is 99 hr, while the TSH on parts 103 and BRMs Rae 2 are shorter with respectively 12 H and 7 hr. The significant difference is explained by the use of a recirculation flow rate of the pump 161 is large relative to the feed rate of substrate of the pump 170, recirculation rate of 4.9.

This design enables production of H₂ in the BRMs, with a supply of reaction medium of relatively large volume in the RAC, in order to increase the inertia of the operation to avoid the leaching the bacterial consortium or microbial present in the reaction medium. A substrate feed rate (SAR) similar to that of the tests performed during testing of the RAC-b receptor (example 5) is used to power the system's BRMs (1.1 ghexose/L BRMs/hr) which in fact corresponds to a DAS reduced system-wide (Rae/BRMs) with 0.5/ghexoseLmiiieu/hr. The control of the pH of the reaction medium of the system is exclusively carried out by the addition of a NaOH solution to 3 mol/L in the RAC 103 by the automatic control system 164 - 166. A pH measurement is made directly at the output of the BRMs 2 by the probe 162 on the recirculation line 106 of the reaction medium. Hydrogen production being used first the BRMs 2, the value of pH control in the RAC 103 is modified to provide a pH of about 5.5 in outlet 24 of the calender.

Production patterns of gas (hr₂ and c0₂ ) have been studied for the RAC 103. After transfer of slurry in the membrane module, the production profile of H₂ in the RAC 103 is close to that of a test rSBA during the 7 h. Indeed, the consortium present in the RAC 103 continues to produce gas from the substrate introduced at the initial time (T-₀ ), but does not receive directly continuous substrate. After 13:30, an increase in the production of hydrogen is observed, which corresponds, approximately to the period of the utility of the reaction medium in the BRMs 2 from the beginning of the continuous substrate feeding. The remaining substrate is not consumed in the calender 24 arrives, in effect, in the RAC 103 and energizes the microbial or bacterial consortium.

103 continuously stirred tank reactor (RACs):

The production period between 20 H and 116 H is used for the calculation of production parameters (table 4).

Yield values and in productivity by hydrogen, the ratio

h-molar2COEXTRUDED /2 and consumption of hexose sugar for testing/BRMs Rae

This phase has a CO2 relatively constant, while that of H₂ , more varying, fluctuates between 0.8 ml_H2//Min and 0.2 ml. Lmiiieu ^ / Lmiiieu/minutes. Productivity by hydrogen during this period is low with 39 ml/L._{the m} the III_had/ hr), the bacterial consortium or microbial does indeed a residual portion of the substrate in the system (unconsumed portion in the BRMs 2). A hydrogen yield of 0.95 mole / ^ nriolhexose consumed is calculated on the production phase of the RAC, taking into account the quantity measurements of glucose added (exiting the BRMs 2) and residual (coming out of the RAC 103). This efficiency is relatively high because of adverse conditions of controlled pH to 6.8 from 29 H in the RAC 103. Consumption of 95% of hexose introduced into the RAC is measured. Similarly, the mole ratio of H₂/ C0₂ calculated on this phase is relatively good with 0.93.

The evolution of gas production in the RAC 103 can also be explained by the pH applied therein, be high (6.8) to obtain a pH of 5.5 outputted from the BRMs TSH means 2. the reaction medium into the BRMs 2 being of 7 hr, the impact of pH change in the RAC 103 is not immediately observed at the output of the BRMs 2, where the pH is measured; changes the regulation value have been carried out in several successive steps. A pH regulating 6.8 in the RAC 103 has finally allowed to reach pH 5.4 output calender 24. The operation at a pH of 6.8 is not impossible for the bacterial consortium production of H₂ , but this does not correspond to the optimum conditions. The TSH means the reaction medium into the RAC 103 being 12 hr, hydrogen producing bacteria are potentially impacted negatively by these fermentation conditions. In addition the pH being about 5 out of the BRMs 2, relative basic shock is subjected to bacterial arriving in the RAC 103. Although the use of a shock acid-base can permit selection sporeformers, continuous operation of the method appears inappropriate. This is a possible explanation of changes in the bacterial consortium, with a negative impact on certain hydrogen-producing bacteria.

2 membrane bioreactor (BRMs):

The production of gas generated in parallel by the BRMs 2 is generally greater. After the transfer of sludge acclimated rac 103 to the BRMs 2 and the beginning of the supply substrate to + 5 hr, a region of stable production is observed between 8 H and 36 hr. " Yield 1.70 moWmolhexose consumed is reached, much higher than the yield obtained over the initial period of the test the rac-b with TSH 12 hr (example 5). The DAS used herein, calculated on the part BRMs is 1.1/ghexoseLmiiieu/hr, close to that used for testing Rae b (1/hr/ghexoseLmiiieu). In either case, the microbial or bacterial consortium does not metabolize all the substrate; but consumption is greater with the BRMs, 53 - 58% against 29 - 38% with the RAC. In addition, with the coupled system/Rae BRMs, whereupon the part of the substrate is transferred to the rac 103 power consuming substantially all. The overall system efficiency is improved accordingly. The value of the productivity (149 rnLH2/ L._{the m}iiieu/hr) is three times greater than that for the period of the RAC-b with TSH 12 hr (47 rnL_H 2. Lmiiieu // Hr). The ratio H₂/ C0₂ (1.11) is high and comparable to those obtained in initial phases during the tests in the RAC (example 5).

After this first production phase stabilized, producing H₂ and CO2 decreases relatively sharply up to 68 hr (-55% et - 27% respectively), and then making up. This decrease production is accompanied by a reversal curves for production of H₂ and c0₂ , with a drop in the ratio H₂/ C0₂ of 1.11 to 0.68, producing c0₂ being less impacted than that of H₂ . These negative effects on hydrogen production appear more linked to accumulation of acids, in particular organic, in the reaction medium and at itself.

This build up would cause the occurrence of the fermenting secondary channels, non-producer H₂ , but c0₂ . Indeed, despite the changes in the value of pH control in the RAC 103, the pH at the output of the BRMs 2 remains less than 5.3 up to T₀ + 68 hr, corresponding to when the production begins to rise.

A third production phase stabilized builds in the BRMs 2 after hoisting the production between 90 H and 145 hr. The efficiency value obtained on this period is relatively low (0.79 moWmolhexose consumed) and can be explained by the footprint of metabolic pathways non-producer H₂ , but c0₂ , since the consumption of hexose sugar is it on the increase (+ 31% between phases 1 and 3). The microaerobic the reaction medium bonded to substrate feeding non-degassed may be responsible for this phenomenon. In contrast, productivity value H₂ is large, compared with those observed during testing in rac has (example 5), in the second and third phase stabilized (around 70 mLH2/ L._{the m}iiieu/hr). A ratio H2COEXTRUDED /2 of 0.58 over this area is calculated, this value is comparable to that seen at the end of the test-rac has (hr2COEXTRUDED /2 =0.65) after inversion of production curves of H2 and CO2, as for this test Rae/BRMs.

Biomass samples taken at 145 H output of the RAC 103 and BRMs 2 were analyzed by high throughput sequencing (table 5). A predominance of the phylum Of Remediation is again observed with more than 55% of the sequences analyzed, but with a significant presence of the phylum Bacteroidetes more than 30%. A large range of different bacterial species observed over that obtained previously in the RAC alone (example 5). The differences of fermentation conditions, in both parts of the system, the rac 103 and BRMs 2, can cause the development of different bacteria, then homogenized by recirculation of the reaction medium via the recirculation loop 106, explaining why the bacterial diversity of two media is close enough. Bacteria from the genera A Clostridium strictu sensu lato are poorly represented in the fermentative medium, with 16% and 8% in the RAC 103 in the BRMs 2, Clostridium butyrlcum is the species predominant in the medium (with 6.3% and 3.1% respectively).

Table 5'. results obtained by sequencing of samples to 145 H of fermentation in the RAC 103 and the BRMs 2 during testing/BRMs Rae - groups having an abundance>2% for at least one of the samples

The fermentation conditions being generally relatively close to those of RAC-has, these results may be related, with those obtained for the sample-rac has 241 hr, obtained after evolution of microbial or bacterial consortium to increased production of CO2. The distribution bacterial (the phylum) is similar, in this sample, although less diverse, with 75.6% of bacteria of the phylum Of Remediation and 10% of Bacteroidetes. The bacteria of the genus -Prevotella, are present at about 10% in the two samples of the test coupling/Rae BRMs and 241 in the sample to the test H of the rac has. Their development appears to thereby derive from fermentation conditions.

Despite the difficulties of regulating the pH, hydrogen production to be kept up on 145 hr within the BRMs. A production profile quite similar to continuous testing in Rae was observed, with a first phase of large production, followed by a deceleration, and then once more moderate production hydrogen, with an inversion of the production curves of H₂ and c0₂ , thereby to be attributable TSH important overall (99 hr) and the absence of tank ventilation supply substrate 10. In contrast, a improved outcomes for producing hydrogen on the initial part of the fermentation in the BRMs relative to the RACs in the test Rae b (example 5) with similar conditions (TSH 12 H and SAR of about 1 g/l_mid J.|_had/ hr). This shows the interest of BRMs for the production of H₂ . Further study relates to the use of BRMs as bioreactor alone for hydrogen generation according to the configuration of Figure 2, described in the example 2.

In this second configuration of the BRMs 2, once the sludge acclimated transferred therein, the RAC is separated from the recirculation loop. Thus, the BRMs alone is used for the production of hydrogen (fig. 2). The reaction medium is recirculated through the recirculation loop 6 where the pH is measured, the substrate and soda are added and the excess liquid is removed. The addition of soda which cannot be performed automatically, it is made by a peristaltic pump 66 whose throughput is manually changed as the test. The residence time of the reaction medium (TSHR) of about 6 H in the calender 24 thus differs in time, control of pH which is thus performed step by step.

The test BRMs BPA is realized with a global TRAb 46 hr, using the feed rates and recirculation used for testing rac has (example 5), but with a volume of reaction medium reduces the volume of the membrane module, 440 ml. A das ghexoseLmineu/0.9/hr was used for this test fermentation. The same holds for the fermentation test-rac-and Rac/BRMs (example 5 and 6), the supply tank 10 substrate has not been degassed during testing fermentation. The first few hours of this test, temporary increases in the pressure have been observed in the calender 24, followed by a rapid pressure drop. This phenomenon is due to an overpressure in the calender linked to an imbalance between the amount of liquid injected and the amount of liquid removed from the BRMs 2. to avoid such pressure fluctuations, a free extraction of the excess volume of the reaction medium has been layered from T₀ + 29 hr. This modification prevents rapid changes in production due to mounted and pressure drops in the calender, observed in particular to 16.5 H and 19.3 hr. It should be noted that any output bdües or liquid was observed in pathways gas downstream from the membrane module, despite the large pressures observed, which attests to the effectiveness of the 2 g/l BRMs for separation in this configuration.

Analyses of the evolution of the production of H₂ and c0₂ as a function of time for the fermentation test BRMs-were carried out. After transferring sludge acclimated in the shell 24, the production in the BRMs 2 is slowed during a period which approximately corresponds to the reaction medium in the TSHR calender 24 (7 hr) thereby to be attributable an adaptation period of the consortium at opératoirés conditions change (continuous feeding, reactor non-stirred, extracting the transmembrane gas products, andc...). After this period, a phase of stable production is observed yield H₂ of 0.77 mole_H2/nriolhexose (table 6) added, what is relatively low (- 23%) relative to the initial production in the BRMs during testing in coupling Rae (example 6)/BRMs. The concentration of hexose BRMs output 2 could not be analyzed accurately because the sampling has been performed on discharge from the BRMs but downstream of the feed point (valve 74), making it difficult readings taking into account the rate of glucose consumed, the approximation made giving the outliers. Productivity by hydrogen obtained on the initial phase is 97 ml of_H2Lmiiieu/hr /, which is 35% lower than for the test Rae (example 6)/BRMs. The ratio H₂/ C0₂ (1.01) is good and similar to test mode (example 5) Rae.

Table 6: yield values and in productivity by hydrogen, the mole ratio of H₂/ C0₂ for the fermentation test BRMs-operation BRMs

A phase stable fermentation, much longer, is observed between 72 169 H and H fermentation, yield H₂ of 0.64 mole_H2/molhexose added which is equivalent to the yield obtained on the second producing stage BRMs 2 during the test in the rac coupling/BRMs (0.62 molh₂ mole / hexose sugar added) (example 6). Similarly, productivities obtained for the two tests are equivalent with 91 and 81 ml_H2/Lmiiieu/hr respectively for testing/RACs BRMs and testing BRMs-. The ratio H2COEXTRUDED /2 is equivalent to those obtained after inversion productions H₂ and c0₂ in steady state operation when the tank ventilation in substrate is not used. The phenomenon of microaerobic membrane reactor seems to be the cause of a H₂/ C0₂ descending.

An important point to note with respect to the fermentation profile obtained in this test is the substantial stability of the rate of hydrogen production over the entire production period, about 1.5 ml_H2/Lmiiieu/Minutes. In contrast, as observed during previous tests and rac rac-/ BRMs (examples 5 and 6), the rate of generation of c0₂ increases during testing and exceeds that of H₂ , herein from 48 hr, which is faster than for the test-rac has, but 12 hours later than for the test in coupling/Rae BRMs. These findings may be correlated to the TSHR, upper for continuous operation (73 hr) with respect to operation in the membrane module (46 hr) and lower for the testing/Rae BRMs on the part BRMs (7 hr).

These production results validate the present invention regards the use of the membrane module since they show better stability of hydrogen production for the tests with the rac alone.

Placing a tank ventilation supply substrate 10 and testing different TRAb have additionally been made as a result of this study and allows further improvement of the yield and productivity of hydrogen in this mode of operation over longer periods of time.

Improving the operation: testing BRMs b

In to further improve the production results obtained using the present invention, the configuration of the BRMs has been modified by the use of a TSH 12 hr, achieved by increasing the flow rate of the peristaltic pump 70 substrate feed with reducing the concentration thereof of 12 to 50 ghexose/I, to obtain a substrate feed rate (SAR) equivalent. The reduction of the TSHR provides improvement of renewal of the reaction medium, by limiting the accumulation of metabolites produced by fermentation and their impact potentially negative ouinhibiteur on hydrogen production. Placing a recirculation rate 1 is also tested.

Unlike Rae (example 5), wherein the reduction of the residence time (TSHR) causes a washing the bacterial consortium unfixed, the vertical configuration of the BRMs 2, without agitation, enables materials in suspension settle to some extent in the lower portion of the shell 24, without being discharged at the effluent outlet 38, nor washed by the renewal of the medium. In addition, the bacterial growth on the wall of the membrane fibers 28 can effect a stable biofilm will not be evacuated by reducing possible to the TSHR. In order to avoid the feed of oxygen into the reaction medium, which is a source of microaerobic, create conditions favorable for the parasitics of metabolic pathways, a continuous degassing is carried out in the supply tank of substrate 10.

A fermentation test similar to that used for the fermentation test Rae b (example 5) is used here with the use of different successive TRAb (11 hr, 17 H and 20 hr) to assess their impact. The extensive study of this parameter is shown below, the example present consisting of an assessment of the membrane process compared to the RACs for efficient production of hydrogen. The analysis of the results was performed by establishing a production profile of H₂ and c0₂ as a function of time for the fermentation test BRMs-b receptor. Significant stability of gas production was observed, with a H₂/ C0₂ remaining close to 1, without inversion production curves, after more than 250 H of operation. A power outage has occurred in the wrong substrate 82 to about 10 hr hr during. Resuming the supply has been accompanied by a rapid recovery of the gas production. The results expressed for the fermentation zone with a TSH 17 H corresponding to the average of the results obtained before and after the stopping phase fermentation then resumed. The invention thus provides a stable and efficient means for the production of hydrogen.

A comparison of production results obtained with the tests fermentation Rae-b and BRMs-b is formed in the wiring board 7 below for TSHR about 12 H and 18 hr. A significant improvement of the hydrogen production is seen with the use of the BRMs, with an improvement of 93% of the hydrogen yield (moWmolhexose consumed) and a three-fold increase productivity H₂ to a TSH-about 12 hr. Similarly, the TSH about 18 hr allows a significant increase of the efficiency and the productivity by hydrogen with 146 - 168% and + + 399 - 417% respectively.

Yield values and in productivity by hydrogen, the mole ratio of H₂/ C0₂ and consumption of hexose sugar depending on the mode of operation RAC and BRMs for two-tsh (about 12 hr to about 18 hr).

A significant improvement of the ratio H (>19%)₂/ C0₂ is also observed. The substrate into the bioreactor is more consumed in the case of BRMs, with 59% against 35% for the RAC. This increase is directly related to the increase in productivity by hydrogen. The bacteria that evolved in part on the hollow fibers and the washing of the reaction medium being reduced by configuring the BRMs, a concentration greater bacteria may be assumed, for faster metabolization of the substrate introduced into the reaction medium.

Biomass samples taken at 68 hr (TSHR=12 hr) during testing fermentation in the rac-b and in BRMs-B have been analyzed by high throughput sequencing (table 8). A predominance of the phylum Of Remediation is again observed, but with significantly less in the presence BRMs (54.6%) than in the RAC (96.3%). A wider variety of the composition of the bacterial consortium is observed in the BRMs relative to RACs, with in particular the presence calibrate Bacteroidetes (11.4%) and Synergistetes (7.1%).

Results obtained by sequencing of samples to 68 H fermentation test RAC and BRMs TSH with a 12 hr groups having an abundance>1% for at least one of the samples

The bacteria A Clostridium strictu sensu lato remain present in significant quantity, with 30.5% in the BRMs, although this proportion is of 91.3% in the RAC. This observation used point on the importance of diversity into the reaction mixture to obtain an effective hydrogen production. The Species Clostridium butyrlcum is the predominant species in either configuration of the bioreactor, which confirms its important role for hydrogen production, however it is only 10.7% sequences analyzed in the sample of the BRMs, against 52.8% for the RAC. The Species A Clostridium barati, majority operation rSBA is herein has to about 3.5% in both configurations. This species appears to least adapted to the continuous mode of operation.

The production results obtained enable the invention as to the interest of using the BRMs relative to that of the RAC alone for efficient hydrogen production, both in efficiency in productivity. The optimization of the production process in operation BRMs is described experimentally below.

The interest of operating mode BRMs being established, the configuration used for the functional test BRMs b (example 7), was retained as an example for further study, with a 12 hr TSH.

The impact of different operating points, regarding recirculation of the medium, using a sequencing of the power supply and, TSH, as well as substrate feeding enhanced with nutrients or biomass, is tested. Finally, a fermentation test was performed without sludge acclimated (inoculum or external) and supply of eflluent fermentation substrate containing only sugars and nutrients; hydrogen production taking place by bio film established on the hollow fibers of the BRMs.

Reproducibility of results:

Three tests fermentation operation BRMs were performed to assess the reproducibility of results of gas production: BRMs b, BRMs-c and BRMs d-. These tests are then used as a reference point for the study of optimization. A TSH-about 12 H and a DAS of about 1 g/l/h are used. Production patterns of H2 for these three tests were analyzed by establishing curves showing the rate of H₂ as a function of time. An evolution relatively close to three profiles was observed, in particular for the tests BRMs-b and BRMs a-d, the test BRMs c having a slightly weaker production rate. This less production in the case of testing BRMs-c may be explained by a pH at the output of the weaker BRMs, about 4.5 to about 5.0 instead of fermentation tests BRMs-b and BRMs d-. A hydrogen yield means 1.0 ± 0.2 mole_H2/ mole_hrexose added is calculated, the fermentation test BRMs-c are predominantly responsible for the error of 20% on the value (table 9 below). Ultimately, the productivity is average 128 ± 31 ml of hydrogen_H2/Lmiiie_U/ hr, always much greater than the productivity of the fermentation test Rae b (47 rnL_H2//Hr Lmiiieu) (example 5). The ratio H₂/ C0₂ is 1.2 ± 0.1 and corresponds to a high value and reproducible on the different tests performed.

Table 9: yield values and in productivity by hydrogen, the mole ratio of H₂/ C0₂ and consumption of hexose sugar for three tests reference operation BRMs

(mole_H2/ mole

hexose sugar added)

Yield

(mole_H2/ mole

hexose sugar consumed,

(ml._H 2/L_{the m} i-|IEUs/hr)

hr₂ coextruded /₂

Hexose sugar consumption (%)

All of these results shows test repeatability fermentation made. A certain sensitivity is noted, presumably related to certain operating parameters such as pH, which regulation has not been automated to accomplish the test, and directly related to production performance being proffered in the bioreactor.

Operating with or without recirculation loop:

The set of test previously realized the fermentation was with a recirculation loop 6, reducing local at the TSH BRMs 2 6 hr instead of 12 hr; the residence time overall hydraulic being kept at 12 hr.

A test BRMs-E is performed without this recirculation loop 6, c'est to say that all of the effluent leaving the BRMs 2 is discharged from the system through the pipe 75. Profiles for producing gas for the fermentation test were performed. A production rate of H2 generally in the low range of the fermentation tests with recirculation loop is obtained. This slight decrease rate of H₂ hydrogen in the differential input low appears to be directly related to the recirculation, in effect, the medium is in this case less well homogenized, the bacteria is contacted with the substrate is less well performed, and a larger percentage of the hexose added is extracted from the calender 24, directly to the waste line 75 flows in the configuration with recirculation loop, the substrate, injected into the system by the pipe 73 traverses in theory twice the shell 24 and the recirculation loop 6 before being discharged, which improves contact with the bacteria. The flow rate of CO2 remains far less than that of H₂ over more than 84 H of fermentation with a flow rate difference significantly large, between 0.2 and 0.4 ml/L._mid iie binding_U//minute.

The table 10 below gives the results of producing H₂ obtained with this configuration, compared to the results obtained with the means recirculation loop 6. yield and productivity H₂ as well as the ratio H₂/ C0₂ , are in the average of the tests performed with the recirculation loop 6. without recirculation, the hexose sugar consumption is 55% against 60% with the loop. This validates the assumption of worse bringing the substrate into contact with the microorganisms, leading to a reduction in power consumption, and therefore also of the productivity. However, the hydrogen yield calculated per mole hexose consumed is very important, 1, 79 - 2, 85 mole_H2/ mole of hexose consumed, that seems to imply that the hand of substrate metabolized by the hydrogen-producing pathways is larger than for the other configuration, with recirculation of reaction medium. A second hypothesis may be that any homogenization of the medium is unfavorable to some non-hydrogen-producing microorganisms, and whose metabolism consumes a substantial portion of the substrate, thereby reducing the yield observed per mole hexose consumed when the recirculation loop is used.

Table 10: yield values and in productivity by hydrogen, the mole ratio of H₂/ C0₂ and consumption of hexose sugar depending on the mode of operation with/without recirculation loop

fermen - test cloth is function - one lamination efficiency (mole_H2mole / hexose sugar added) hexose sugar consumed) yield (mole_H2/ mole (ml._H 2. Lmiiieu // Hr) Productivity C0₂H₂/ Consump - hexose ignition

Analysis by high throughput sequencing was carried out on a sample taken from the effluent of the biomass BRMs to 68 H of fermentation for the tests with and without recirculation loop (BRMs-c and BRMs a-e, respectively). The main results obtained are presented in table 11 below. A larger percentage of the phylum Of Remediation is observed in the case of testing without recirculation loop, against 55% 72%, whose gap is mainly due to the family 1 Clostridiaceae. This family includes the bacteria of the genus A Clostridium strictu sensu lato, percentage of 54% without recirculation loop, against 31% with. The Species A Clostridium aminomutase is primarily responsible for this increase, representing 24.7% of the sequences identified in the microbial consortium the test without recirculation loop, while the latter was 1.1% with recirculation. This evolution shows that the fermentation conditions without recirculation loop appear more favorable to such bacteria could be responsible, in part, to the improvement of the hydrogen yield. The population of Sporolactobacillus falls when the recirculation loop is not used, the bacteria are generally lactic acid, non co-producer gas, supporting the hypothesis of inhibition of bacteria using the substrate for competing metabolic pathways.

Table 11 results obtained by sequencing of samples to 68 H fermentation test BRMs-c and BRMs-th groups having an abundance>2% for at least one of the samples

These results clearly show the interest of using the configuration of the bioreactor without recirculation loop, providing a better utilization of the substrate, in an amount less but more specifically to the production of hydrogen. This configuration is however less favorable to productivity H2 high, and limit the homogenization of the reaction medium, whose progress is more difficulty followed (pH value) since the pH measured at the outlet of the BRMs is the result of compensation by adding soda made 12 hr upstream, against 6 H in the configuration with recirculation loop. In a optical parametric optimization of the system, the homogenization and possibility of close tracking fermentation conditions have been preferred. For further study, the configuration with recirculation loop was used. In a optical hydrogen production, the use of a system without recirculation loop can however be envisaged.

Adding nutrients in the effluent fermentation substrate:

Fermentation in the tests using the present invention, a perpetual renewal of the reaction medium in the calender 24 occurs, up to five times during 68 H of fermentation with a TSH 12 hours, with the addition of substrate, containing only glucose. The nutrients are initially present in the sludge acclimated are readily consumed or removed to channeling the flow outlet 75. In order to renew the nutrient medium, required for bacterial metabolism, adding a solution containing a potassium dihydrogen phosphate, ammonium sulfate, iron sulfate, magnesium sulfate and nickel chloride, is made through the solution supply substrate. Two tests fermentation, BRMs-f and BRMs g-, were made by using the same conditions as in reference fermentation tests described above.

Production profiles obtained with adding nutrients have been made. The profiles of the two tests are quite similar, and have a production rate of H2 greater than those obtained without the use of nutrients in the first hours of operation 40. However, a drop in engine speed appears subsequently with a stabilization of the production at 1.7 ITs) LHRH2/ L._mid the II_had/ min to 50 hr. Further, at this period, an increase in the production of CO2 is observed, thereby more than the production of H₂ . A change in the metabolism appears to be initiated this reversal of the production, related to the use of nutrients in the solution supply substrate.

The table 12 below gives the results obtained with the two test means fermentation with nutrients (test BRMs a-f, BRMs g-) compared to those obtained without the use of nutrients (test BRMs b, BRMs-C., BRMs a-d). The test results of fermentation with nutrients are very close, with less variation than for those obtained with the tests without nutrients, however these utterances are modulating given a limited number of tests fermentation made. Adding nutrients may to some extent limit environmental changes in bound to the substrate, allowing better reproducibility of the results. The rated performance of the hexose added and producing H2 average 68 on the first hours of fermentation are included in l'écart subtype calculated on the values obtained without use of nutrients, although being located in the upper range. The major difference out of these results is the consumption subtotal in the calender added substrate 24 with the use of nutrients, 97% against 60%. Adding nutrients allows greater consumption of the substrate, but the substrate is not substantially more used for hydrogen production, as indicated by the yield H2 per mole hexose consumed: 1.15 against 1.81 without l'utilisation-of nutrient. The presence of nutrients in the reaction medium thus appears to allow the development of microorganisms of non-hydrogen and the consumers of the substrate, or activation of competing metabolic pathways, these two hypotheses being bonded. A ratio H2COEXTRUDED /2 of 1.05 is obtained with the presence of nutrients in the substrate solution. This ratio is good, but lower than that obtained for the test without nutrients, thereby denote a CO2 larger, coproduced by secondary metabolic pathways as the solvantogénèse.

Table 12: mean values of yield and productivity by hydrogen, the mole ratio of H₂/ C0₂ and consumption of hexose sugar fermentation tests with and without adding nutrients

These results demonstrate that the presence of nutrients selected in this study and their contents in the feed improves the reproducibility of results, but also promotes the emergence of unwanted secondary metabolic pathways. However, the production of hydrogen is long and the substrate is completely consumed, and the composition of the final reaction medium will provide advice on the definitive interest nutrients, since a production of metabolites of interest may prove a point positive additional hydrogen production, in a perspective of industrial production.

Hydraulic residence time (TSHR):

The TSH is a major parameter of the fermentation in steady state operation, several tests have therefore been made to analyze a fairly broad range TSH, ranging from 6 hr to 46 hr. After the fermentation test BRMs g-, using glucose as a substrate feed with nutrients, a reduction in the TSH 12 H to 6 H has been performed for 70 hr, and then a return at 12 hr TSH has been performed. A reduction in the yield of H₂ and c0₂ is observed as a result of the reduction of the TSHR. The production of H₂ is reduced by about half, while that of c0₂ about 20% decreases only insignificantly. The return to the TSHR to 12 H is followed by a rising of the producing c0₂ , but that of H₂ remains low. An imbalance in the bacterial consortium to a TSH 6 H can be responsible for the drop in the production of H₂ , irreversible after a return to the initial conditions.

The production results of H₂ , presented in table 13, show output drop means H₂ of 66% as a result of the reduction of the TSH 12 hr to 6 hr. Over the same period, productivity is reduced to 50 142 mLH2/ L._{the m} the III_had the ratio H / H and₂/ C0₂ drop 1.05 to 0.45. The nearly total consumption of hexose introduced into the reaction medium is observed down production of H₂ . Other secondary production pathways are therefore used for consuming the substrate, at the expense of the production of H₂ , which can be explained by variation in the bacterial consortium active bacteria. The return to a TSH 12 hr does not allow to lift the hydrogen production which is generally reduced again of 41% on average over the entire period, while production of c0₂ rises, explaining the decreased ratio H₂/ C0₂ to 0.24. A consumption rate of 100% is hexose conserved over this period, thereby confirms competing metabolic pathways in the bacterial consortium.

Table 13: yield values and in productivity by hydrogen, the mole ratio of H₂/ C0₂ for the fermentation test BRMs-G with

modification of the TSH 12 hr to 6 hr

This fermentation test showed that the use of a TSH 6 hr was not conducive to the production of hydrogen, relative to a TSH 12 hr. The next test is the test fermentation BRMs b, no additional nutrients, with TSH 11, 17 and 20 hr, already mentioned above (example 7). A DAS stability has been preserved as modifications of the TSHR. A slight decrease in the production output of H₂ and c0₂ is observed when increasing the TSHR to 17 H and then 20 hr. A break the wrong substrate feeding has occurred to about 10 hr hr during 82. Resuming the supply has been accompanied by a rapid recovery of the gas production. The results expressed for the fermentation zone with a TSH 17 H corresponding to the average of the results obtained before and after the stopping phase fermentation then resumed.

The results obtained with the test fermentation BRMs-above (example 7), using a TSH 46 H can also be used to study the impact of the TSHR on the performance of the membrane bioreactor. The table 14 below summarizes the results obtained with different tsh risk their fermentation made. An optimum TSH appears between 12 and results of these 17 hr, with a yield of 1.81 mole / ^ molhexose consumed and productivity up to 153 mLH2/ L._{the m} the III_had/ hr (testing BRMs b). The ratio H₂/ C0₂ is optimal, greater than 1, and consumption of about 60% hexose is observed. With the rise of the TSHR, a reduced efficiency H₂ per mole of hexose sugar added is observed, accompanied by a reduction in the consumption of hexose sugar, which significantly increases the efficiency H₂ per mole hexose consumed, evaluated at 2.77 mole / ^ molhexose consumed to a TSH 20 hr. It seems therefore that the producing bacteria of H₂ either the most promoted under these conditions, however a decrease in productivity is observed.

The use of a TSH 46 hr causes the decrease of yield to 0.77 mole_H2/molhexose added and productivity to 97 mLH2/ L._midiieu/hr.

Table 14: yield values and in productivity by hydrogen, the mole ratio of H₂/ C0₂ for the tests with different TRAb BRMs fermentation

Unlike RACs, wherein reduction of the TSHR causes a washing the bacterial consortium, the vertical configuration of the BRMs 2, without agitation, enables materials in suspension settle in the lower portion of the shell 24, without being discharged, nor washed by the renewal of the medium on the one hand, and on the other hand, the configuration BRMs allows a biofilm deposited on the outer wall of the hollow fibers. Therefore a TSH-reduced to 12 hr does not adversely affect the production of H2. The increase in the TSHR appears to restrict the activity of microorganisms consumers hexose and non-producers of H₂ , thereby limiting the consumption of substrate for unwanted metabolic pathways. However, the increase in the TSHR beyond 20 H is also a cause of a reduction in the productivity H₂ . A TSH 12 H is maintained for the rest of the study (use of a real substrate) because this utility was used previously with several operating conditions, thus allowing for more recoil for the study of the environmental impact as the use of a real substrate. In addition, for a same SAR, the use of a short-tsh reduces the concentration of the feed solution used, offering a wider range of biodegradable organic material can potentially be used as a substrate or biomass.

Supply mires:

Two tests were performed fermentation agricultural biomass fermentation as effluent from power substrate: mires of wine. The fermentation test BRMs-I has been carried out under conditions identical to the initial conditions are in place for testing fermentation BRMs-b receptor. Of mires from, of the vinification of grape Pfister on the cultivar Riesling and domain were used. These mires are diluted in water network of Strasbourg to obtain the proper concentration (12 ghexose/liter), and adjusted to pH 7.0 before being stored at 4 °c in the supply tank 10 in order to prevent bacterial growth.

A profile of gas production obtained with the use of these mires as substrate has been provided, to follow the flow of H₂ and c0₂ as a function of time. Three phases of fermentation are observed, with a first phase of acclimatization which lasts about the time required for two turnovers of the reaction medium, for good acclimation of the consortium to this feed solution complex and the potential introduction of novel bacteria have in these mires. A moderate production of H₂ , with about 1.5 ml/L._{the m} iII_U/ min is observed on the first phase, which is then increased to about 2.5 ml/Lmiiieu/nnin on phase 2 of stable production. A deviation of the production with decrease in the production of H₂ , accompanied by a slight increase of c0₂ is however observed from approximately 96 hr, intensifying the phase 3.

Table 15 below presents the production results obtained on these three phases of operation. A maximum yield of 1.66 mole_H2/molhexose consumed is observed in the first phase, linked to 98 ml of moderate productivity_H2/ L._{the m} iII_U/ hr. The hexose consumption greatly increases on the second production phase of H₂ passing from 51 to 83% hexose consumed, which is accompanied by an increase in productivity to 148 ml_H2/ L._midiieu/hr, located in the high range of results obtained with a model substrate. The yield H₂ per mole hexose added increases, while the yield of 50% H₂ per mole hexose consumed is reduced by 20%. This involves the consumption of a substantial portion of the substrate for the production of secondary metabolites. Finally, on the third phase, representing the deviation of the production, efficiency decreased from about 20% is observed relative to the phase ratio H 2.₂/ C0₂ final average is 0.64, is 35% lower than that of the production phase 2. a high consumption of the substrate is always observed (97%). Metabolic pathways not coproductrices H₂ appear to be activated by the supply of mires. Potentially, the development of secondary metabolic pathways is due to the introduction of non-hydrogen coproductrices, meet the continuous conditions and consuming the substrate instead hydrogen producing bacteria.

Table 15: yield values and in productivity by hydrogen, the mole ratio of H2COEXTRUDED /2 for the fermentation test BRMs-I with the use of supplied biomass as mires

A substantial portion of the substrate is not used for the production of H₂ , the use of a DAS lowest has been contemplated for optimizing results of hydrogen production by avoiding the cogeneration of unwanted metabolites. In addition, in order to limit the supply of microbial origin mires, in addition to treatment previously explained, the mires diluted were filtered (pore diameter of 0.45 MW), limiting the presence of suspended solids and microbes. This test fermentation, BRMs a-j, is made from mires different, from a vineyard Burgundy of cultivar Fermented Chardonnay domain Poncétys the duct ecophyto set points of the plane. A DAS initial 0.2 // hr gtiexoseLmiiieu is used in this test, increased to 0.4 and then 0.7 during the fermentation, corresponding to 20, 40 and 70% the SDA used previously, while retaining a TSH 12 hr.

A low production is observed data rate is less than 0.5mLH2/ L._{the m}iiieu/min during the first hours of fermentation test, but relatively large in view of the low SAR applied to the system. Production c0₂ very low is also observed, representing about half that of H₂ , when performed. With the increase of the DAS to 0.4/ghexoseLmiiieu/hr, a reversal curves of H₂ and c0₂ is however observed, accompanied by the expected increase in production output of H₂ , although weaker than hoped, limited to 0.5 the L/ml._mid iie binding_U/ minutes. Finally, the change of DAS to 0.7 ghexose/L._{the m}iiieu/hr to 149 hr causes a gradual increase in gas production for several tens of hours, which probably indicative of variations in the bacterial consortium, quite slow to implement, which may be associated with the growth of the microbial population.

The results show although the ratio H₂/ C0₂ very important, 2.22, obtained with the SAR of 0.2 grams_hrexose/L._{the m}iiieu/hr, contrast with those obtained of SAR higher (about 0.9) (Table 16 below). A nearly complete consumption of the substrate is observed 0.2/ghexoseLmiiieu/hr, while it decreases with increasing SAR;

the greater the amount of substrate that is supplied is large, more substrate remains unconsumed in the effluent leaving. The yield H₂ per mole of hexose sugar added is relatively low, maximum 0.2/ghexoseLmiiieu/H with 0.88moWmolhexose added and falls to 0.56 ^ / mole molhexose added to 0.4 // hr ghexoseLmiiieu. Despite the expectations, the yield per mole hexose consumed is not very high, with a maximum of 1.14 ^ / mole molhexose consumed with a SAR of 0.7 // hr ghexoseLmiiieu. Assuming the reduction SAR would concentrate the use of substrate for production pathways of H₂ thus is not to be retained. Division of the substrate occurs, although the channels coproductrices CO2 appear unfavorable to a DAS low. The addition of substrate being limited, productivity observed is low, with 23 ml_H2/Lmiiieu/H at 0.2/ghexoseLmiiieu/hr, and 68 mLH₂//H at 0.7 Lmilieughexose|Lmj/game/hr.

Table 16: yield values and in productivity by hydrogen, the mole ratio of H₂/ C0₂ for the fermentation test BRMs a-j

A sample of biomass harvested during this test fermentation, to 68 H fermentation was analyzed by high throughput sequencing. The results are compared with those obtained with a conventional feed (testing BRMs a-c) (table 17 below). In these two samples, the majority of sequences identified bacteria of the phylum Of Remediation (55 and 75% for the tests BRMs-c and BRMs a-j, respectively). However a relatively large changes between these two samples, the proportion of bacteria of the family 1 Clostridiaceae is thereby reduced from 18% to 30 sequences, while the families Lachnospiraceae and Bacillaceae In strongly increases in proportion for testing using mires as supplied biomass, with respectively 20 and 15% sequences analyzed, against 1 and 5% glucose supplied for testing. The bacterial species most abundant in the fermentation test supplied with mires is be Anaerobacillus (alkalidiazo-to-trophicus/alkalilacustris) with 15.4% sequences. This species was already present in the test supplied with glucose with 3.6% sequences, which thus does not account for either delivery of this bacterium via the mires, but rather the creation of conditions favorable to their development. In contrast, other bacterial species identified in quite large amount were not present in the test without additional mires, it is Vallitaleaguaymasensis (11.3%), (pavanii/andPseudomonas maltophilia) Stenotrophomonas (8.4%) and species of indeterminate The Streptococcus SPs. /Nostocoidalimicola (7.4%).

Table 17:

Results obtained by sequencing of samples to 68 H fermentation test BRMs-c and BRMs-th - groups having an abundance>2% for at least one of the samples

The use of a carbohydrate rich agricultural waste and other nutrients, such as mires, is attractive for biohydrogen production. This biomass was relatively efficient, although the provision of novel bacteria native to the agricultural biomass, appears to have modified the microbial consortium active and the general metabolism of the effluent fermentation substrate. The renewal of the microbial consortium, continuously, accompanied by a complete addition nutrients will nevertheless consider offering significant stability of the production process over the long term.

Operation from the biofilm BRMs

The biofilm created as fermentation tests on the fibers 28 BRMs 2 was used as a microbial consortium exclusive production of H2, therefore without external supply of microbial or bacterial inoculum. Placing the fermentation test differs therefore tests previously realized since the acclimation phase of activated sludge wastewater treatment plant, during the previous 5 hr up continuously, is not performed. The BRMs is directly brought into continuous feeding with a template substrate containing exclusively glucose, with added nutrients from of the T₀ fermentation broth with a hexose sugar/1, 9 das Lmj|game/H and TSH 12 hr.

The production of H₂ begins little by little on the twenty first hours of fermentation test and then settles 23 hours after the start of feed of the BRMs 2. a production rate of H₂ importantly observed, up to 3.5/rnLLmiiieu/minutes. An inversion of the production profile of H₂ and c0₂ takes place between 39 H and 60 hr. The production of gas settles to 80 H with a production rate of H₂ to 2 ml/L._mid the II_had/ min and c0₂ to 2.5 ml/L._{the m} iII_U/ minutes.

A very H₂ per mole hexose consumed is observed on the first phase of operation of 23 to 39 hr, with 2.24 mole_H2/molhexoseconsumed, with a consumption rate in large substrate, productivity and 204 ml./ hr / ^ Lmineu, maximum value of this study (table 18 below). On the second production phase, a slight decrease in these results is observed (generally 15%), but are still significant, particularly with regard to the productivity with 179 mLH₂/Lmiiieu/Hr. The decrease of the ratio H₂/ C0₂ is 27% between the two phases.

Yield and productivity H₂ important could be obtained, highly competitive with the tests using a microbial inoculum freshly taken from the sewage treatment plant, revealing the interest of BRMs for stabilization of the microbial consortium producer of H₂ on the hollow fibers. In addition, these results suggest additional simpler method of long-term, by limiting the needs of reseeding of the reaction medium.

Table 18: yield values and in productivity by hydrogen, the mole ratio of H₂/ C0₂ for the fermentation test BRMs K

In operation continuously BRMs has been successful and validated the invention, either in use of the membrane module coupled (example 6:

The rac/BRMs) or alone as a membrane bioreactor (example 7:

BRMS). The use of the membrane bioreactor (BRMs) alone offers excellent results because it shows better stability of hydrogen production tests (example 5) with RACs. Under the prevailing conditions tested by way of illustration only, the use of a TSH 12 H and degassing the solution supply substrate has improved the performance thereof by allowing efficient production of H₂ for more than 400 hours. After testing modes of operation with and without a sweep gas on the membrane module, the use of the pumping system developed on the semiconductor reactor batch process was tested for extraction of the gas produced by the fermentation showing that the configuration with sweep gas is most effective. A real substrate from agricultural biomass was used as a solution of continuous feeding of the BRMs with promising results. Finally a fermentation test has been made without delivery of inoculum outside, using the biofilm created on the hollow fibers during previous fermentation tests.

The online analysis of gas products is ensured by a micro gas chromatograph (PGCs) (Agilent's m200), comprising two modules have sensors to thermal conductivity (DCT) (46 fig. 2 - 146 and 147 in Figure 3). The table 19 gives the configuration of the PGCs used. The first module includes a column type zeolite molecular sieve 5 confectioneries permitting the separation of non-polar gas, while the second column contains a polymeric PoraPLOT U-permitting the separation of polar compounds, in particular the c0₂ in this study.

Table 19 : Description of the configuration of the modules of the PGCs to-DCT used gas analysis

A membrane gas/liquid, placed upstream of the PGCs to-DCT allows the arrival of all condensed phases on the micro chromatograph. A series of 3 analyses of 75 s is performed every 10 min during the fermentation.

Analyzing metabolites fermentation:

Specimen in a 15 ml tube by the valve of the reactor, the biomass sample is centrifuged at 4500 rpm, then the supernatant is transferred to a second 15 ml tube. The two tubes are stored at -20 °c until analysis.

Analysis by gas chromatography

The samples are thawed and then filtered (pore size: 0.22 MW) to remove any residual particles. An aliquot of solution of trifluoroacetic acid (a TFA) (3.42, 10'² the L/mole) is added to the sample to protonate the volatile fatty acids to be analyzed (ph=2). The analysis is performed by the GC FID (Agilent's technology 7890a CG). The chromatographic column used is of the type 4dB-FFAP (15 x m 0.1 mm, 0.1 pm) specific to the analysis of polar organic compounds: alcohols and volatile organic acids in our case.

Analysis by liquid chromatography

As well as for analysis by GC of FID, the samples are thawed and then filtered (pore diameter: 0.22 pm) to remove any residual particles. The chromatograph (Agilent's technology sets 1260) coupled to UV detection/visible is used with a chromatographic column Agilent's HI Plex hr, 7.7 χ 300 mm, 8 mW (and pre-column HI Plex hr, 50 mm X-7.7) specific to the separation of polar compounds such as alcohols and organic acids.

Analysis of carbohydrates

The total amount of carbohydrates present in the samples is evaluated by colorimetric assay. In glass test tubes, an aliquot (1 ml) of each sample is mixed with 2 ml of solution of anthrone (2 g/l) in hr2OS4 concentrate (95%). The tubes are then heated to 80 °c during 20 minutes, and then cooled and preserved in ice during 20 minutes. Each sample is then analyzed by UV-visible (spectrophotometer Varian model Cary 3) at a wavelength of 625 nm. A new calibration curve is performed with each batch of samples from a reference range: 0, 10, 20, 40, 60 and 80 ppm of glucose. The results obtained are given equivalent hexose sugar. This achieves a result independent of the nature of the sugars present in the sample: reducing or not, simple or complex. The analysis of the raw biomass yields the total concentration of carbohydrates, dissolved or not. Centrifuging the biomass allows separation of the supernatant, containing the metabolizable carbohydrate low molecular weight, and of the base, containing carbohydrates molar mass larger, non-soluble, the majority of which is not metabolizable without hydrolysis step. Quantification of carbohydrates dissolved is carried out by the analysis of the supernatant. A sample dilution of biomass in the terminals of the calibration range is performed with deionized water before analysis.

[1]A. Duigou, Miguet m, Y Amalric, "French of hydrogen markets on in-2008 - village bednet expectant pro Aspects", Int.. J Hydrog. The ENERGY, flight. 36, no. 15, e. 8822 - 8830, 2011.

[2] S. Giddey, sec. Greenish Badwal, A Kulkarni, "Review electrochemical ammoniac producing technology NDA 1003", Int.. J Hydrog. The ENERGY, flight. 38, no. 34, ρ. 14576 - 14594, ηον. 2013.

[3] Kalinowski L, j m. Lowincome, " hydrogen: vector of the transition energy? ", As parliamentary evaluation of scientific and technological choice, French, the assembled national parliamentary report no. 1672 and the Senate no. 253, 2013.

[4] F. Romagnoli, e. Blumberga, and i. Pilicka, " The Assessment Life Cycle Assessment biohydrogen producing in photosynthetic processes rather ", multinational Journal hydrogen the ENERGY, flight. 36, no. 13, ρ. 7866, 7871.2011.

[5]YΜ .Wong reported, and Wu Y T, and J. C. in the Juan, "industry has tendered hydrogen producing estimation using seed-sludge through DSD fermenting", the Renewable and Spheral révisions, flight. 34, ρ. 471,482, June 2014.

[6]MY. Azwar, Μ. A Hadziyannis, A Κ. Abdul Wahab, "reforestation biohydrogen producing bypass photobiological, fermenting NDA electrochemical processes rather: raise numerous", Finalize. Sustainer. The ENERGY Rev., flight. 31, ρ. 158 - 173, march 2014.

[7] Bourbonneux g, hydrogen production, petroleum refining, t3 transformation methods, (1998), 463 - 515.

[8] memento timepiece systems hydrogen, plug 3.2.1, AFH2.

[9] P. small, separation and liquefaction gas, gas separation techniques, engineering technology j3600 (1995).

[10] Pendyala B, R. S. Chaganti, J A Lalman, R. S. stayed, d. at Heath, LAU Reject bw K, "Pretreating externally mixed anaerobic enfant de la naissance differ source: Correlating tea hydrogen-hydrogenase activity and good microbial avec la of yield particularly vegetables", Int.. J Hydrog. The ENERGY, flight. 37, no. 17, ρ. 12175 - 12186, seven. 2012.

[11] Bw Hallenbeck, "the fermentative hydrogen producing: principes, the PROGRESS, NDA formulations", Int.. J Hydrog. The ENERGY, flight. 34, no. 17, ρ. 7379 - 7389, 2009.

[12] D Β .Kriz, L on Pitt, Μ .from Love, "producing Biohydrogen: NDA leads to verbs restrictions on Practical applica-activation", Int.. J Hydrog. The ENERGY, flight. 29, no. 2, ρ. 173 - 185, Feb. 2004.

[13] THE X. Wang's J, n Q NER, informally by Sheng Fang W, W Qian as Guo, "response of gaseous 1 Zend relationship has inhibiting DNA the nutrient limitation the experimental community hydrogen producing bypass hydrogen flame-producing fermented bacte-to-Yemeni Β 49", Int.. J Hydrog. The ENERGY, flight. 32, no. 6, ρ. 748D754, May 2007.

[14] V Clion, C Dickson, sec. Guard, Β. of Ernst, "vaccine potency analysis using biohydrogen producing bypass DSD fermenting", Canadian Journal Chemical Engineering from, 93 (2) (2015) 309 - 316