WO2014016785A1

WO2014016785A1 - System for producing electrical energy based on active elements such as solid oxide fuel

Info

Publication number: WO2014016785A1
Application number: PCT/IB2013/056079
Authority: WO
Inventors: Pierluigi LEONE; Giovanni PASQUALE; Massimo SANTARELLI; Orfeo Zerbinati
Original assignee: Acacia Cleantech S.R.L.
Priority date: 2012-07-27
Filing date: 2013-07-24
Publication date: 2014-01-30
Also published as: ITTO20120663A1; EP2878030A1

Abstract

System for producing electrical energy based on active elements, such as solid oxide fuel cells, wherein, after the redox reactions of a combustible substance, which is fed to an anode electrode of said cells, and of an oxidant, which is fed to a cathode electrode of said cells, a continuous flow of electrical energy is generated at the ends of said electrodes. Said system comprises an upper head (2) and a lower head (3), between which an electrical energy generation unit (4) is arranged, which houses said cells. Said upper and lower heads comprise a plurality of feeding and distribution ducts and chambers, for feeding and distributing a cathode oxidant flow and an anode fuel flow in the direction of said electrodes, between which a central duct (7) substantially extends through the whole generation unit, thus establishing a communication between the two heads and allowing the cathode and anode flows to flow from the upper head to the lower head, and vice versa.

Description

System for producing electrical energy based on active elements such as solid oxide fuel

The present invention relates to electrochemical systems for producing electrical energy based on active elements, such as solid oxide fuel cells (SOFC). In particular, the invention relates to an innovative solution for distributing reagents and for discharging reaction products from within a fuel cell stack.

Solid oxide fuel cells generate a continuous flow of electrical energy following redox reactions of a combustible substance, which is fed to an anode electrode, and an oxidant, which is fed to a cathode electrode (typically air). The ceramic electrolyte, typically yttria-stabilized zirconia or another ceramic oxide, transports oxygen ions from the cathode electrode to the anode electrode as a function of the operating temperature. This type of technology is currently deemed to be promising in many application fields (e.g. distributed power generation, domestic microgeneration, portable energy production devices, auxiliary units) because of its highly efficient conversion of fuel into electricity and the possibility of integration with other devices for generating additional electrical energy by exploiting the hot exhaust fumes (300-500 °C). The operating temperature of solid oxide fuel cells is normally within the range of 500-800 °C, depending on cell geometry and materials. The high temperature allows reducing the main cell overvoltages resulting from thermally activated physical phenomena. In addition, the high operating temperature allows integration with reforming reactors and ensures multi-fuel flexibility. Finally, it allows various power system configurations based on solid oxide fuel cells, such as systems for thermal energy and heat cogeneration, trigenerative systems for further production of refrigeration energy, hybrid systems for producing electrical energy with a cascade thermoelectrical system.

A typical configuration a single fuel cell utilizes a solid and dense electrolyte, which ensures transportation of oxygen ions; yttria oxide-stabilized zirconia oxide (YSZ) is most used because of its chemical-physical stability in reducing/oxidizing atmospheres and its optimal conduction of oxygen ions at temperatures above 650-700°C; a cathode material which allows reduction of an oxidant and which is chemically and physically stable with the other functional layers and with the feed gas atmospheres, such as, for example, strontium-doped lanthanum manganite (LSM); an anode material which allows oxidation of a fuel, which may contain hydrogen, carbon monoxide and gaseous hydrocarbons, and which is chemically and physically stable with the other functional layers and with the feed gas atmospheres, such as, for example, a nickel (Ni) and YSZ compound; a cell interconnection material, typically consisting of ferritic steel, which has good conduction properties in the oxide state, in contact with the surfaces of the electrodes, which ensures a good electrical connection between contiguous cells, and which promotes the flow of the reactant fluids towards the anode and the cathode.

The problem tackled by the present invention is to optimize both the feeding of the fuel and the discharge of the exhaust products to/from a generation unit including a plurality of cells organized according to a predefined scheme. The Applicant has also faced the problem of how to organize a generation unit including a plurality of cells. The present invention proposes an integrated structure for thermal and fluidic integration of the pre-reformer, the post-catalytic combustor and the reagent pre-heating sections with the stack.

One aspect of the present invention relates to a system for producing electrical energy based on active elements, such as solid oxide fuel cells, having the features set out in the appended claim 1.

Further features and advantages of the invention will be hereafter described with reference to an embodiment thereof as shown in the annexed drawings, wherein:

• Figure 1 shows a cross-section of the integrated system according to the present invention,

• Figures 2A and 2B show, respectively, an enlarged view of the upper and lower heads of the integrated system, FIG. 2A illustrating the cathode feed and exhaust flows and FIG. 2B illustrating the anode feed and exhaust flows,

• Figure 3 shows the packing order of the components of one layer of the multi-cell stack,

• Figure 4 is a partial sectional view of the plates making up the stack, highlighting the feeding channels, the communication holes and the distribution and exhaust channels, • Figures 5A and 5B show the circulation of the cathode flow on the feeding plate and on the distribution and discharge plate, respectively,

• Figures 6A and 6B show the circulation of the anode flow on the feeding plate and on the distribution and discharge plate, respectively,

• Figure 7 shows the details of one of the plates that make up the pre- reforming unit,

• Figure 8 shows the details of one of the plates that make up the post- combustion unit,

• Figures 9A, 9B and 9C show the circulation of the anode flow on the separator plate, on the feeding plate and on the distribution and discharge plate, respectively, in the alternative representation of the multi-cell layer, according to an alternative embodiment of the present invention.

With reference to the above-listed drawings, the system for producing electrical energy according to the present invention comprises an upper head 2 and a lower head 3, between which a generation unit 4 is arranged.

Said upper and lower heads comprise a plurality of feeding and distribution ducts and chambers, for distributing a cathode flow and an anode flow.

For the purposes of the present invention, the terms "anode flow" or "anode feed" refer to any combustible substance that contains methane (e.g. natural gas or biogas from anaerobic digestion) or another carbon-hydrogen-oxygen- based combustible substance (e.g. propane, LPG, methanol, ethanol, etc.) which can be converted into a synthesis gas mixed with water; the term "cathode flow" or "cathode feed" refer to an oxidant, e.g. air.

A first heat exchange unit 5 (or pre-reformer) is arranged between the upper head and the generation unit, and a second heat exchange unit 6 (or post- combustor) is arranged between the lower head and the generation unit. A central duct 7 substantially extends through the whole structure, thus establishing a communication between the two heads and allowing the cathode and anode flows to flow from the upper head to the lower head, and vice versa. Said duct, in fact, comprises a first cathode channel 71 for the cathode flow and an anode channel 72 for the anode flow, arranged within the cathode channel. The first heat exchange unit (pre-reformer) and the second heat exchange unit (post-combustor) are both conceived as extensions, respectively upwards and downwards, of the generation unit. The first unit comprises a heat exchanger that receives the cathode exhaust flow coming out of an exhaust duct of the generation unit and the anode fresh flow coming from the upper head. The second unit comprises a combustor also acting as an exchanger: in fact, it receives the anode exhaust flow coming from the respective exhaust duct of the generation unit, which burns with a stoichiometric percentage of the cathode flow coming from the central duct, thereby heating the remaining flow.

The upper head comprises a chamber 21 for distributing the anode feed flow and feeding the pre-reforming unit, a chamber 22 for receiving the cathode flow with an additional function as a fresh flow/cathode exhaust flow exchanger, and a chamber 23 for expelling the cathode exhaust flow. The lower head includes a chamber 31 for distributing the cathode flow and feeding the post-combustor, a chamber 32 for collecting and expelling the anode exhaust flow, and inlet 33 of the duct that admits the fresh anode flow mixture into the integrated system. Circulation of the cathode flow takes place as follows (see Figure 1). The cathode feed goes in through the inlet in communication with the exchanger arranged in the upper head. The cathode flow expands within chamber 22, where it is pre-heated by ducts 24, in which the cathode exhaust flows are directed towards exhaust chamber 23, corresponding to the external annular region of the upper layer. Once the exchange chamber has been totally filled, the cathode flow begins flowing towards the passage volume (internal region of the upper layer) through communication holes 25. From there, it is immediately made to flow into central tube/tie rod 7 through as many holes 26 in its side wall.

After having entered central duct 7, the cathode flow runs through it until it reaches the opposite end in the lower head, while at the same time being heated by surrounding generation unit 4 and heating the anode flow in the smaller-diameter internal duct 72. A foil turbulator may, if necessary, be secured to the anode flow duct in order to improve the thermal exchange. When the descent is over, the cathode feed exits through other holes 341 , which deliver it into an annular-section chamber 34 delimited at the top by a sealing ring; through further upper slots 35, it then arrives at the distribution chamber 31 , where it will be distributed to the post-combustion unit. Admission into the post- combustion plates occurs through smaller holes 36, for the percentage that must be burned with the anode exhaust flow, and bigger holes 37, for the remaining flow that must be heated. While flowing through the section where heat exchange takes place, the cathode flow travels towards a narrow annular slot between the central region of the plates and the central tube/tie rod, which will feed all the layers of the above stack. The cathode exhaust products of the reactions that have taken place on the cells are collected into elongated ducts 41 positioned on the outer perimeter of the stack plates. The exhaust flows are directed towards the pre-reforming unit and flow radially through it until they arrive at the ducts that will carry them towards the exhaust chamber. These ducts go through the anode feed flow distribution chamber and the exchanger chamber while thermally interacting therewith. Once they have arrived at the exhaust chamber 23, the cathode exhaust flows are removed from the integrated system through the outlet duct.

The path followed by the anode flow is as follows. The anode feed flow (see the corresponding arrows in Figure 1) enters the system perpendicularly to the lower head through a narrow duct 33. The flow then runs through a vaporization chamber 39, where the quantity of water contained in the flow is completely nebulized thanks to the heat yielded by the surrounding chamber for collecting and expelling the anode combustion fumes. The added water dose is such that no mixtures can form which might determine any "carbon deposition" phenomena during the reactions occurring in the generation unit. The small duct then goes into central duct 7 and runs throughout its length up to the top of the upper head. Within this section, thermal exchange phenomena occur due to the contemporaneous passage of the countercurrent cathode flow. At this point, the anode flow, which has also been slightly heated while crossing the pre- reforming unit, exits the system and flows through an external mixer 8 that will ensure an additional supply of water to the anode flow, in the event of shortage, during the operation of the device. Mixer 8 can also be used in order to supply water for the purpose of producing inerting steam in the event of a failure or overtemperature. The anode feed then flows back into the system through inlet 29 towards fuel distribution chamber 21. From there, the anode feed flow moves towards the underlying pre-reforming unit through passages 28 formed on the bottom base. The anode flow goes radially through the pre-reforming unit towards feeding ducts 42 of generation unit 4. During this transition, a heat exchange occurs with the cathode flows flowing in the opposite direction, from the periphery to the centre, towards the upper head. The anode exhaust flows produced inside generation unit 4 are conveyed into exhaust ducts that will carry them to the post-combustion unit. The fumes produced by the combustion of the anode exhaust flows, together with the incoming molar fraction of cathode feed, flow through cathode flow distribution chamber 31 in a duct that will carry them to exhaust collection chamber 32, from which they will then be expelled through an outlet duct.

Generation unit 4 according to the present invention comprises a plurality of solid oxide fuel cells that generate a continuous flow of electrical energy after the redox reactions of a combustible substance, which is fed to an anode electrode, and an oxidant, which is fed to a cathode electrode. Said cells are organized on stacked layers P; on each layer P there are preferably a plurality of cells, e.g. six, distributed in circular form, among which a parallel connection is established. The cells are closed on their sides by two groups of metal plates, which regulate the circulation of the different reactant flows and allow to establish a series electrical connection between the various layers. At the ends of the unit as a whole, which is made up of a plurality of stacked layers, there are two collectors for anode current 43 and for cathode current 44, respectively. Finally, the whole stack of layers is electrically insulated from the rest of the system by an insulating plate 45 placed on top of each collector. Preferably, the layers are oriented in a manner such that the cathode of each cell faces downwards.

The number of cells in each layer can be chosen at will; advantageously, however, any even number of cells will create a symmetric geometry that will facilitate the distribution of the flows.

Considering the cathode side first and proceeding from bottom to top, one can see a feeding plate AC, an exhaust distribution and collection plate DC, and a plate CC for holding metal meshes R, the function of which is to improve the expansion and homogeneity of the flow directed towards different cells C. FIG. 4 shows the cross-section of an assembled layer. The feeding plate has a series of conduits ACC, the depth of which equals the total thickness of the plate. As can be seen in FIG. 5A, the grid of conduits is also formed by secondary transverse ribs ACN, the depth of which is half that of the plate, which connect the main conduits to one another.

The conduits have a constant cross-section throughout their extension, and as a whole they are so organized as to exactly reproduce the shape of the cells. In particular, said conduits are arranged longitudinally relative to each other, and the outermost ones are slightly arched, so that the grid will take a particular shape that copies the shape of the above cell. In this grid of conduits, the incoming fresh flow is distributed evenly prior to reaching the distribution plate DC, shown in FIG. 5B. Passage occurs through a series of holes DCF, the number of which may vary, obtained in the exhaust distribution and collection plate and arranged axially at the intersections of the grid of the underlying feeding plate and possibly at the apex of main branches ACC. Between the two ends of the holes a slight pressure gradient is created, which imparts a slight acceleration to the flow going through them, thereby ensuring that the jet through the holes will reach the cell.

Distribution plate DC has the same geometry with longitudinal conduits: these are grooves DCC, the depth of which is half the plate thickness, interposed between the feeding holes and used for collecting and removing the exhaust flows produced. Between the cells and distribution plate DC there are metal meshes R, each positioned exactly under one cell of the layer and having the same surface area, arranged within housings provided in holding plate CC. Mesh R has such a weft and passage section as to much enlarge the diameter of the jet coming out of each hole, so that the overall distribution of the flow over the entire surface of the corresponding electrode will be as homogeneous as possible. Furthermore, the mesh acts both as a good heat exchanger, thus allowing the flow to reach adequate temperatures before coming into contact with the cell, and as a good electric conductor, thus generating low contact resistances on both the cell cathode and the exhaust distribution and collection plate.

The group of plates on the anode side is wholly similar to that on the cathode side, but the order is reversed. Analogy can also be observed in the assembly of the plates and in the distribution of the conduits on feeding plates AA and on the distribution and discharge plates DA, as shown in FIGS. 6A and 6B. Anode feeding plate AA, and hence conduits AAC formed therein, have a reduced thickness compared to the cathode side, due to the lower rate of the anode flow.

Between the anode compartment and the cathode compartment, three plates are inserted, i.e. two mesh holding plates CC and one insulating plate PI. The function of the mesh holding plates is to hold the meshes, whereas the function of the insulating plate, made of dielectric material, is to contain the cells, to ensure electrical insulation between contiguous cells, and to prevent gases from moving from one electrode to another, thereby avoiding mixtures that might cause the entire device to malfunction.

Finally, the cathode side of each layer is separated from the anode side of the next layer by means of a very thin metal separator plate PS, which constitutes the bottom of the two adjacent feeding plates. In a possible alternative configuration, the two anode and cathode feeding plates and the separator plate may be replaced with a single bipolar plate.

FIGS. 5A and 5B show in detail the layout of the upper surfaces of the two plates forming the pairs on the cathode sides of the cells of each layer. FIG. 5A particularly refers to cathode flow feeding plate AC, whereas FIG. 5B shows plate DC for distributing the fresh flow and collecting the exhaust product. The cathode flow comes from a narrow annular slot between central tube/ tie rod 7 and the plate itself, and is diffused in the grid of feeding conduits ACC after crossing an equally deep connection region also formed in the plate. Through the system of holes DCF, the flow then arrives into the upper distribution plate DC, from which it reaches the cathode and is involved in the reactions that allow producing electrical energy. The products obtained from the reactions settle into the collection and exhaust channels DCC, while a fresh flow comes up towards the cell along the distribution channels and replaces the previous flow in a sort of convective movement. The cathode exhaust flow is carried by dedicated channels towards apertures ASC on the side opposite to the inlet, on the outer edge of the plate, which apertures act as exhaust ducts.

As regards the anode side, the event dynamics is the same as the one described for the cathode side, the only difference consisting of the different locations of the inlets and outlets for the anode flow. FIGS. 6A and 6B show that the flow arrives at the feeding plate through circular ducts AAF on the outer edge of the plate, between which there are as many ducts ASA for collecting the anode exhaust flows. Each duct supplies the quantity of anode feed required for the operation of two cells. The anode flow is delivered into a wide initial channel, the depth of which is equal to the thickness of the plate and preferably equal to the depth of the conduits, which is then split into two parts by a tongue LA that separates the flows intended for the two adjacent cells. Likewise, in the exhaust distribution and collection plate DA each duct collects the exhaust flows of two cells. Finally, in this case as well the exhaust collection ducts are in a position opposite to the feeding ducts.

In this manner, the phase of feeding the cells belonging to each layer of unit 4, on both the anode and cathode sides, is achieved through a system of channels that ensures a homogeneous distribution of the flows on the cell surface. The circulation of the gases obtained during the feeding phase, on both the anode and cathode sides, allows to attain a substantially uniform temperature profile on the surface of each cell.

The phase of collecting and discharging the products generated by the reactions occurred on the cells of each layer, on both the anode and cathode sides, takes place separately from the feeding phase.

In an alternative representation of the multi-cell layer, the cells lying in the same plane operate in series. In particular, each cell is fed with an anode mixture composed partly of fresh fuel coming from the stack and, for the remaining part, of a percentage of the anode exhaust flows produced by the preceding cell. When the stack is operational, therefore, on each layer belonging thereto a recirculation is created wherein a part of the feed for each cell is supplied by the immediately preceding cell. The percentages at which the involved flows participate in the feed mixture depend on the operating parameters at which every single cell is to operate, i.e. fuel utilisation and steam-to-carbon ratio, and are determined by the dimensions of the various feeding ducts. For this representation it has been chosen to keep such parameters constant for all the cells lying in the same plane. FIGS. 9A, 9B and 9C show plates making up the anode side of this alternative representation; no modification is however required on the cathode side, compared to the plates shown in FIGS. 6A and 6B. In particular, FIG. 9A shows metal separator plate PS that separates the anode and cathode plates of two consecutive layers. The fresh anode flow is supplied through right-hand duct CD of each pair present on the outer edge of the plate. It travels along a short channel CS carved into the metal plate, to then enter the passage that will carry it to the underlying feeding plate, visible in FIG. 9B. At this point, the fresh fuel feed flows through a converging/diverging channel CDC, the depth of which is half that of the plate that accommodates it. At the ends of this channel a pressure difference is created, which accelerates the incoming flow towards the grid of conduits in front of it. While flowing in this channel, the fresh fuel gets into contact and mixes with a part of the anode exhaust flows produced by the cell that occupies the preceding position. The exhaust flows, which also contain a fuel percentage that did not participate in the reactions that took place in the preceding cell according to the preset Fuel Utilisation parameter, come from the exhaust distribution and collection plate shown in FIG. 9C. Once they have been collected by the appropriate channels, they move towards duct ASA, which is shaped like a truncated circular crown, and then diffuse into the corresponding channel carved in the feeding plate to a depth equal to half the plate's depth. From there they flow together into the converging/diverging channel, where they will mix with the fresh fuel flow and will travel towards the adjacent distribution region through two side ducts CLS carved to half depth. Thanks to the pressure difference generated by the converging/diverging channel, the anode exhaust flows are drawn in the proper direction, and there is no risk that they, or even any fresh fuel, might be sucked and jeopardize the operation of the stack. The dimensions of the two side ducts are determined on the basis of the part of the exhaust flow of the preceding cell to be reused. It is important that mixing occurs before the diffusion into the grid of conduits, in order to avoid creating any spatial unevenness in the feed composition due to the different nature of the two participating flows. Finally, it is necessary to prevent any excessive accumulation of water from occurring within the plate over time due to the recirculation established therein. For this reason, a part of the anode exhaust flows produced by each cell, containing a significant percentage of water, is expelled from the stack through small ducts CEPS situated beside the truncated circular crowns and put in communication therewith via short ducts as deep as the plate. This type of configuration allows obtaining, for each layer of the stack, overall values of fuel utilisation and steam-to-carbon ratio which are lower than those associated with each cell. In this alternative representation, each layer can house any number of cells because each cell is fed independently of the other ones, thus eliminating any plate geometry constraints.

On top of generation unit 4, preferably formed by multiple layers, a first heat exchange unit 5 (or pre-reforming) is arranged for pre-reforming the anode feed flow. This process involves the decomposition of the heavy hydrocarbons present in the fuel used in the system. In general, at the outlet of the pre- reformer there will be a mixture rich in methane, steam and, to a lesser extent, hydrogen, carbon monoxide and carbon dioxide. The quantity of water is such that no carbon deposits can be formed in the ducts that feed the combustible mixture into the cells or in the cells themselves.

With reference to FIG. 7, the pre-reforming unit consists of a stack of equal plates PR, the dimensions of which are the same as those of the feeding and distribution plates and of the cell holding plates of each layer, thus constituting an extension thereof in the vertical direction.

The number of plates composing this unit will be strongly dependent on the type of fuel supplied to the system and on the difficulty encountered in creating a mixture that can be directly fed to the cells. Modulation of this unit in terms of fuel feed can also be attained by selecting the type of catalyst to be deposited in the regions where the pre-reforming reactions will take place.

Since the pre-reforming process is very endothermal, the heat required for its activation is taken from the cathode exhaust flow coming from generation unit 4. For this reason, each plate of the pre-reforming unit is so designed as to include countercurrent heat exchangers S, each of which is composed of two channels SC for the passage of the cathode exhaust flow, between which there is an oval section SO where the pre-reforming reactions occur. Each element is obtained by carving the corresponding plate to a depth preferably equal to half its total thickness. The feed flow comes from small holes in the inner part of the plate and travels towards anode feed holes AAF in front, which will distribute it to the different underlying multi-cell layers. The path extends through a catalytic grid RC positioned within each section for the purpose of promoting the pre- reforming reactions. At the same time, the cathode exhaust flow coming from the unit enters through collection duct ASC and travels towards the centre of the plate in the direction opposite to that of the anode flow. After having completed its path, the exhaust flow will enter a duct that will carry it towards the cathode exhaust collection plate in the upper head, from where it will be definitively expelled from the integrated system. In order to maximize the heat exchange between the two flows, the channels for the cathode exhaust flow are strewn with many small cylinders, so as to increase the contact area.

At the lower end of generation unit 4 a second heat exchange unit 6 (post- combustor) is arranged, which is dedicated to the combustion of the anode exhaust flow with a small molar fraction taken from the cathode fresh flow coming from central tube/tie rod 7.

Unit 6 comprises a series of plates PC designed as shown in FIG. 8. As in the pre-reformer case, the size of the plates is the same as that of the plates that make up the generation unit, being in fact a downward extension thereof. The reaction mainly produces carbon dioxide and water, and its highly exothermal nature promotes the generation of a very large amount of heat, which is used for heating the air flow to be then fed into generation unit 4.

For a better exploitation of the heat produced by the reaction, combustion sections are alternated on the plate with channels for the passage of the flow to be heated, the depth of both elements being preferably half the plate thickness. The anode exhaust flow enters the combustion section through the exhaust duct ASA of generation unit 4 and propagates through a catalytic grid RC, added in order to promote the combustion reaction. The cathode flow coming from central duct 7 is split into two flows, the rates of which are determined by the different diameters of the holes. The molar fraction necessary for activating the combustion process flows into the smaller holes FP, whereas the flow that will be heated is conveyed into the bigger holes FG. The combustion products flow out into duct FE, which will carry them towards the exhaust flow discharge plate in the lower head. The flow to be heated moves in a radial direction with respect to the centre of symmetry, moving from the inlets on the outer edge of the plate to the inner end thereof, where it will then enter the narrow annular section leading to the different layers of generation unit 4. Advantageously, as for unit 5 or pre-reformer, the transit channels for the cathode feed are strewn with small cylinders in order to increase the thermal interaction with the adjacent combustion sections. The paths of the reactant fluids are so organized as to maximize the heat exchanges between the high-temperature exhaust flows, which have to be expelled from the system, and the low-temperature fresh flows, which, on the other hand, are directed towards generation unit 4, where the reactions take place.

The produced anode and cathode exhaust flows follow two separate paths for a more profitable exploitation of their calorific power.

The plates that make up the central units of the system described so far are not subjected to any welding. As shown in FIG. 1 , perfect adherence and compaction of the parts are ensured by the pressure exerted on the device by the upper and lower heads, which are connected to each other by means of the central duct, which also acts as a central tube/tie rod. Fastening is completed on the lower head by the application of contrast plates P1 and P2 pressed by Belleville washers MT; the whole assembly is tightened by means of a screw and a nut D. An additional seal T, made of heat-resistant rubber, is inserted within the contrast plates to prevent the cathode flow from leaking through the lower head.

The upper head is composed of three cylindrical chambers 21 , 22 and 23, wherein the uppermost one, to which the tube/tie rod is welded, is smaller than the plates of the central block, while the remaining ones are bigger. FIG. 2A shows the order in which the three blocks are assembled. Lower chamber 21 is used for distributing the anode flow towards the underlying pre-reforming unit. The incoming flow must travel along a first spiral section that slows down its expansion within the available volume, thereby allowing for heat exchange with the cathode exhaust ducts that cross the chamber towards the exhaust chamber. As an alternative to the spiral inlet, a circular grid with small holes may be inserted in the external region of the chamber to create a preliminary zone where the anode flow can be distributed evenly prior to expanding towards the central zone, where there are passages, also small in size, that communicate with the pre-reforming section. Upper chamber 22, instead, only contains the ducts coming from generation unit 4, which are formed by overlapping cylindrical and elliptical plates. In this case as well, it is possible to introduce a circular grid in the external region of the chamber to make the cathode flow homogeneous before it enters the heat exchange zone. Finally, last chamber 23 comprises two zones separated by a circular wall: the outermost zone, annular in shape, collects the cathode exhaust flows coming from unit 4 and expels them from the system; the innermost zone acts as a passage for the cathode flow before it enters the central tube/tie rod, thereby also allowing a last heat exchange to occur between the different moving flows. The lower head comprises two cylindrical chambers 31 and 32 which are larger than the plates of unit 4. With reference to FIG. 2B, distribution chamber 31 receives the cathode flow from the central tube/tie rod and delivers it to the post-combustion unit. Prior to reaching the actual distribution site, the cathode flow crosses a first annular-section volume 34 to prevent any friction between the distribution chamber and the tube due to a possible thermal expansion of the components of the lower head. This initial volume is separated from the remaining volume by a ribbed wall with flow passages 35 perfectly matching those on tube/tie rod 341. Furthermore, thanks to the presence of radial ribs 342, the wall ensures the necessary rigidity to the structure of the distribution chamber, which is subject to the load exerted through the Belleville washers by the nut screwed to the end portion of the tube/tie rod. A sealing ring is arranged over the annular section to ensure the separation between the fresh flow entering the distribution chamber and the hot flow coming out of the post- combustion unit, which will feed the cathode section of unit 4. Except for the ribbed wall having the same function as the one of the upper layer, lower chamber 32 has an empty volume which is completely dedicated to the collection and discharge of the anode fumes coming from the post-combustion unit.

Central duct 7, which can be observed in FIG.1 , has a dual function. Firstly, it plays a role as an internal tie rod, acting as a connection between the two heads. This task also involves the inner walls of the two heads, which, in addition to playing a containment role, also give rigidity to the whole system. Furthermore, the position of the tube allows the cathode flow flowing therein to undergo heat exchanges with both generation unit 4 and the anode flow passing through smaller duct 72 contained therein. Any non-linear sections, which may be implemented through bends 721 of the tube itself, will increase the time of permanence of the anode flow in the central section, resulting in a longer thermal interaction time. In an alternative embodiment of the device, it is also conceivable to modify the surface of the outer or inner wall of the duct by adding foil turbulators or finned components in order to maximize the exchange area.

Claims

1. System for producing electrical energy based on active elements, such as solid oxide fuel cells, wherein, after the redox reactions of a combustible substance, which is fed to an anode electrode of said cells, and of an oxidant, which is fed to a cathode electrode of said cells, a continuous flow of electrical energy is generated at the ends of said electrodes,

characterized in that it comprises

• an upper head (2) and a lower head (3), between which an electrical energy generation unit (4) is arranged, which houses said cells,

• said upper and lower heads comprise a plurality of feeding and distribution ducts and chambers, for feeding and distributing a cathode oxidant flow and an anode fuel flow in the direction of said electrodes,

• a first heat exchange unit (5) or "pre-reforming" is arranged between the upper head and the generation unit, and a second heat exchange unit (6) or "post-combustor" is arranged between the lower head and the generation unit,

• a central duct (7) substantially extending through the whole generation unit, thus establishing a communication between the two heads and allowing the cathode and anode flows to flow from the upper head to the lower head, and vice versa.

2. System according to claim 1 , wherein the anode fuel flow preferably contains methane (e.g. natural gas and biogas) and the cathode oxidant flow is air.

3. System according to claim 1 , wherein said combustible substance is a natural gas or biogas from anaerobic digestion or another carbon-hydrogen- oxygen-based combustible substance (e.g. propane, LPG, methanol, ethanol, etc.), which can be reformed into a synthesis gas for being directly fed to the cells.

4. System according to claim 1 , wherein the pre-reforming and post- combustion units are manufactured as heat exchangers, which interact in a horizontal direction and have dimensions that are similar to those of the generation unit, from which they extend as prolongations in the axial direction.

5. System according to claim 1 , wherein said pre-reforming unit comprises a plurality of stacked plates (PR), which are equal to one another and whose dimensions are the same as those of the feeding and distribution plates, thus constituting their extension along the vertical direction.

6. System according to claim 5, wherein the number of plates that make up said unit and the type of catalyst deposed inside the areas where the pre- reforming reactions take place depend of the fuel introduced into the system.

7. System according to claim 1 , wherein the packing of the integrated system is performed by means of the central duct (7), which, besides transporting the reactant fluids, acts as a tie rod, since it is hooked to the two heads and causes them to press against the central core of the system.

8. System according to claim 1 , wherein the two heads arranged at the ends of the integrated system produce the pressure that is necessary to pack and seal the plates of the pre-former (PR), of the post-combustor (PC) and of the generation unit (4).

9. System according to claim 1 , wherein the paths of the reactant fluids are so organized as to maximize the heat exchanges between the high-temperature exhaust flows, which have to be expelled from the system, and the low- temperature fresh flows, which, on the other hand, are directed towards the generation unit (4), where the reactions take place.

10. System according to claim 1 , wherein the anode and cathode exhaust products follow two separate paths, so as to exploit their calorific value in a more efficient manner.