WO2004079849A1

WO2004079849A1 - Multi voltage tap redox flow battery composed of stacked cell modules of adjustable cell area

Info

Publication number: WO2004079849A1
Application number: PCT/IT2003/000129
Authority: WO
Inventors: Andrea Zocchi; Krisada Kampanatsanyakorn; Placido Maria Spaziante
Original assignee: Squirrel Holdings Ltd.
Priority date: 2003-03-04
Filing date: 2003-03-04
Publication date: 2004-09-16
Also published as: CN1765027A; WO2004079849A8; AU2003219521A1

Abstract

A redox flow battery includes a plurality of cells electrically connected in series between two end terminals of the battery composed of elements all of a certain active area stackable in a filter-press arrangement. Positive and negative electrolyte solutions flow in parallel and/or in cascade, respectively through positive electrolyte flow compartments and through negative electrolyte flow compartments of the cells. The cells are filter-press stackable multi-compartment cell modules comprising monopolar electrodes, each module composed of a certain even integer number n+1 of flow compartments alternately for said positive electrolyte solution and for said negative electrolyte solution separated by an odd integer number n of perm-ionic membrane separators, and alternately housing a positive and a negative electrode. The redox flow battery has a plurality of intermediate voltage taps in the form of externally protruding tabs of said electrodes. Cell area partializing switches are controlled for connecting in common a selectable number of either positive or negative electrodes composing a cell module.

Description

MULTI VOLTAGE TAP REDOX FLOW BATTERY COMPOSED OF STACKED CELL MODULES OF ADJUSTABLE CELL AREA

TECHNICAL FIELD OF THE INVENTION

5 The present invention relates to systems for storing and/or transforming energy, based on redox flow batteries.

BACKGROUND OF THE INVENTION

Storage batteries, and in particular redox flow batteries, are often employed in the exploitation of renewable energy sources, in load leveling, and in the generation 10 and distribution networks of electrical energy. The use of storage batteries is necessary in "stand-alone" photovoltaic (solar) panels systems not connected to any power distribution grid. Redox flow batteries offer many advantages for these types of application compared to other types of storage batteries.

Among redox flow batteries, all vanadium batteries, i.e. batteries that employ a 15 vanadium-vanadium redox couple in the negative electrolyte as well as in the positive electrolyte, are particularly advantageous.

Performances of a storage plant employing vanadium redox flow batteries are reported and analyzed in the article: "Evaluation of control maintaining electrical power quality by use of rechargeable battery system", by Daiichi Kaisuda and 20 Tetsuo Sasaki IEEE 2000.

There is a wealth of literature on redox flow batteries and in particular about vanadium redox flow batteries. Therefore, a detailed description of the peculiarities and advantages of such batteries in respect to other types of batteries does not seem necessary in order to fully describe the present invention.

25 Among the many advantages of redox flow batteries, it is worth remarking though their suitability to being charged at different charging voltages. To accomplish this, intermediate voltage taps along the chain of elementary cells in electrical series that constitute the battery, may be used. Depending on the voltage of the available source, most appropriate taps are selected for coupling to the recharging voltage source an appropriate number of cells. This is possible because, unlike other types of storage batteries, redox flow battery systems store energy in the electrolyte solutions containing the redox couples (briefly electrolytes) that circulate through the cells and that are stored in separated tanks. The battery represents exclusively the electrochemical device where electrical energy transforms into chemical energy and vice versa, and the electrodes of the cells do not undergo any chemical transformation during charge and discharge processes.

In renewable energy systems, there are conditions, generally of a variable nature, that may affect the transformation process and the eventual energy storage.

In case of aeolian generators there is the problem of providing for constant characteristics of the electrical energy that is supplied to electrical loads. In case a DC generator (dynamo) is used, the generated voltage varies with the rotation speed and each aeolian generator is often provided with mechanical devices to increase the useful range of wind conditions. In case an alternator is used to generate an AC voltage, speed variations cause frequency variations of the generated AC voltage and rectifiers DC-DC converters and inverters may be necessary. In case it is necessary to store electrical energy in batteries, a battery charger must be coupled to the alternator.

Similar problems are encountered also in hydroelectric power plants.

When interconnection to the local mains is contemplated, the electrical power produced on site must have the same voltage and frequency as the distribution network. This applies for example in all those applications where electrical power produced from renewable energy sources satisfies only partially the local energy demand and the difference is made up by absorbing power from the electrical utility network or, during periods of favorable weather, the electrical power exceeds the demand and the excess is fed into the distribution network.

In the use of redox batteries for energy storage in these plants, interconnection to the distribution network may increase considerably the exploitation of natural renewable energy sources, allowing the generation of electrical power even under sub-optimal conditions that would not produce the standard electrical voltage and frequency characteristics required by local electrical loads or permit uploading of excess power on the distribution network (in order to gain energy credits).

It is clear that the design of these power plants exploiting renewable energy sources of unpredictable characteristics implies the identification of ranges of useful conditions. On that basis, rectifiers, DC-DC converters, transformers, inverters, mechanical transmission ratio converters, and the like are required for allowing exploitation of renewable energy sources for longer periods and at levels economically convenient in respect to the investment. As already said, the use of storage batteries is a necessary condition to enhance exploitation.

In many cases the cost of these ancillary devices and accessories may surpass the cost of the generator and/or of the eventual storage batteries. Moreover, a low efficiency figure of these devices may severely lower the overall efficiency figure of the whole renewable energy source plant.

Generally, electrical distribution networks, and as a consequence electrical machinery and devices, operate with an AC voltage because it is relatively easy to modify using simple static machines such as electrical transformers.

This has also imposed the establishment of a standard (50 or 60 Hz) mains frequency (AC) and all electrical machines and devices of common household use, are normally designed, and/or operated at this fixed mains frequency.

By contrast, batteries typically store and deliver DC electrical power.

Interfacing problems between these two systems are evident and are commonly overcome by employing battery chargers on one side and inverters on the other side. These ancillary devices significantly lower the overall efficiency of the energy transformation processes (battery charging and discharging processes).

One important area of technological advancements is the solution of technical problems tied to the interfacing of electrical systems of generation and/or distribution of AC power and/or of storage and successive release of electrical power using a redox flow battery to efficiently store energy in the battery independently of the electrical characteristics of the source in terms of voltage and/or frequency.

Prior PCT patent publication No. WO 03/007464, discloses a hybrid inductor-less or transformer- less inverter system based on a redox flow battery provided with a number of intermediate voltage taps established at intervals along the stack of elementary cells composing the battery of elementary cells in electrical series between the two end terminals of the battery.

The output AC waveform is reconstructed by sequentially and cyclically switching the intermediate voltage taps on the output line at the desired frequency.

Prior PCT patent application No. PCT/IT02/00653 discloses a method and relative structure for efficiently charging a redox flow battery provided with an array of intermediate voltage taps established along the stack of elementary cells that compose the battery, in electrical series between the two end terminals of the battery, from any DC or AC electrical source by functionally switching the DC source or the output of a rectified AC waveform to an appropriate intermediate voltage tap in order to charge the elementary cells included in the circuit at appropriate charging conditions.

The document also discloses a complete hybrid system of battery charger and inverter based on the same multi voltage tap redox flow battery, capable of transforming electrical energy from any DC or AC source of any frequency into electrical energy deliverable to an electrical load at a specified AC voltage and frequency. Such a system is ideally suited for plants exploiting renewable energy sources.

Prior PCT patent publication No. WO 99/39397, discloses a redox flow battery composed of a plurality of elementary cells in electrical series, in the form of a bipolar filter-press electrolyzer, implementing a cascaded flow of the positive and negative electrolyte solutions containing the redox couples from the first cell at one end of the stack to the last cell at the other end of the stack, thus preventing critical conditions of bypass (parasitic) currents and associated pitting and corrosion problems on surfaces of electrically conducting elements (electrodes).

The cascaded flow of the positive and negative electrolyte solution (electrolyte) is realized by an appropriate coordination of through holes and slots, customarily created in the frame portions of the stackable elements that are normally made of molded plastic material (essentially of a not conductive material) that, upon assembling of the various elements that compose the stack in a so called filter- press arrangement, create the internal ducts of the battery for the circulation of the electrolyte solutions.

On the other side, it is well known to redox flow battery operators that the voltage tends to vary basically as a function of two variables, namely: the state of charge of the electrolyte solutions that are circulated in the cells of the battery and of the electrical load (current flowing through the battery).

While the variation of the battery voltage due to the state of charge of the positive and negative electrolyte solutions is relatively small and may be easily controlled, variation of the voltage due to the electrical load may be rather large.

Variations of voltage due to the variation of the state of charge of the electrolytes may be readily corrected in an automatic manner by modulating accordingly the pumping rate of the electrolytes through the battery from the reservoirs of charged electrolyte solutions to the reservoirs in which spent electrolyte solutions are recovered during a discharge phase of the battery system. Variations of the battery voltage (or most significantly of the internal voltage drops) due to electrical current variations, during charge and discharge phases, depend on the fact that the battery voltage varies with the varying of the electrical load or of the charging current because of internal resistance parameters of the redox flow cells that compose the battery.

By examining the problem at elementary cell level, at null load the cell voltage is equal to the open circuit voltage which is function of the state of charge of the electrolytes in the cell compartments. At the full rated load the voltage is much lower than the open circuit voltage because of the not negligible internal resistance of the cell. In case of an all-vanadium redox flow cell, by defining for example as the full rated load the load at which the cell voltage is equal to 80% of its open circuit voltage, (which with a state of charge of about 50% would correspond to about 1.2 N (during a discharge phase) and to 1.5 N (during a charging phase)), therefore, during a discharge process, the cell voltage will vary from 1.35 N, at null load, to 1.2 N at full rated load. This represents a voltage variation of about 11 %.

Therefore, the full range of variation of the cell voltage is of 0.3 N (from a minimum of 1.2 V during a discharge phase to a maximum of 1.5 V during a charge phase) that is a variation approximately comprised between 20 and 25 %.

For many applications such a relatively large range of voltage variation may not be acceptable. For instance in the case of telecommunication switching stations, a hybrid energy transformation and storage system, as the one disclosed in said prior patent application PCT/IT02/00653 based on redox flow battery, could cause problems because of an excessively large range of variation of the battery voltage.

A related parameter is the so-called voltage efficiency of the battery that is defined as: ηv^Vd/Vc, where Vd is the discharge battery voltage and Vc is the charge battery voltage. This figure of merit accounts for the irreversibility of the ohmic losses in the battery that is a function of the current density over the effective cell area (i. e. the electrode and/or membrane area). In applications sensitive to a large voltage swing the alternative would be to revert to the use of electronic DC/DC converters to provide for a substantial constancy of the battery voltage.

Of course the other important figure of merit of a redox flow cell as well as of any other type of storage battery, is the Faraday (current) efficiency that is defined as: rjF = Qd/Qc, where Qd is the electrical charge delivered to a load during discharging and Qc is the electrical charge received by the battery during charging. In redox flow batteries it may be as high as 99.8 %.

The overall energy storage efficiency ηj? is the product of the above specified efficiencies:

η_E = η_vη

Given that filter-press type bipolar redox flow batteries may often conveniently placed and operated as a vertically stacked assembly, that is with horizontally laying elements (electrodes, membranes), any upgrading of the maximum current deliverable by the battery before a limiting drop of the battery voltage is experienced, requires an increase of the cell area and therefore of the "footprint" of a vertically stacked battery. This may be a problem in existing installations because of an impossibility of so enlarging the footprint of the battery.

OBJECT AND SUMMARY OF THIS INVENTION

We have now found an effective solution to the problems caused by an excessively large range of variation of the voltage at the terminals and/or at intermediate voltage taps of a redox flow battery.

The novel structure of this invention significantly reduces the range of variation of the battery voltage while preserving the advantages of a filter-press stack architecture of the battery as well as the possibility of implementing a cascaded flow of the positive and negative electrolyte solutions containing the relative redox couples through the respective positive electrode and negative electrode compartments of the elementary cells that compose the battery.

Basically, according to this invention, at least some of the elementary cells that are electrically connected in series between the two end terminals of the battery and composed of elements that are stackable in a filter-press arrangement have different areas from one another. The difference of area of one elementary cell to another along the stack of cells is determined by making at least certain cells in the form of multi-compartment monopolar cell modules, having a certain number of flow compartments of opposite polarity, containing a positive or a negative electrode, respectively.

The monopolar electrodes of the multi-compartment monopolar cell modules are selectively connectable, via external switches, into the electrical circuit of the battery, according to needs. In this way, the effective (working) cell area of any monopolar module of the battery may be incremented or decremented by selecting or deselecting certain monopolar electrodes of each polarity, that are eventually connected in common into the electrical circuit of the battery.

Substantially, all the stackable electrode elements, all the permionic membrane elements and optionally even any bipolar electrode element composing the battery stack have the same area and a geometrically similar frame portion of non- conductive material providing for the sealability of the cell compartments according to a filter-press stack assembling mode.

The frame portions of the distinct stackable elements composing the battery are provided with coordinately matching through holes, borings and/or slots defining the internal ducting either for flowing the positive and the negative electrolyte solutions in cascade or in parallel, respectively through positive electrolyte flow compartments and through negative electrolyte flow compartments of the elementary cells or of the multi-compartment monopolar cell modules.

At least a terminal electrode of every multi compartment monopolar cell module may advantageously coincide with an intermediate voltage tap of the battery stack, preferably each monopolar electrode of each multi- compartment monopolar cell module coinciding with an externally connectable intermediate voltage tap, thus providing for other adaptivity possibilities for best matching the active cell area to the current flowing through the battery, in order to meet the limits of variation of the battery voltage.

The possibility of switching any one intermediate voltage tap of the battery to the node of a source of electrical energy (for charging the battery) and/or to an output node for powering an electrical load (in discharging energy from the battery), besides permitting waveform discretization as disclosed in the prior patent application PCT/IT02/00653, permits one also to maintain the current density in the selected elementary cells at values appropriate to maintain the internal voltage drop through the selected cells within acceptable limits.

For a certain current flowing through a certain number of monopolar cell modules of the battery, or even through a single monopolar cell module, accounting for a certain battery voltage by incrementing or decrementing the actual working cell area will proportionally make the internal resistance smaller or larger.

Selection of the appropriate group of elementary cells, or of the number of monopolar electrodes connected, of each monopolar cell module, when charging and/or discharging the battery, will be made dependent on the respective (generally variable and different) power levels, determined by the electrical characteristics of the source providing the charge current and/or by the electrical load powered by the battery.

By virtue of this flexibility of adapting the operating conditions of the single elementary cells of the battery to the external electrical requisites, in any situation of use of the battery, the battery architecture of this invention is particularly effective for implementing the hybrid systems disclosed in the cited prior PCT patent publication No. WO 03/007464and in the cited prior patent application PCT/IT02/00653. In fact, discretization of an AC waveform implies that different groups of elementary cells of the battery that are sequentially and cyclically selected for connection into an external circuit (for charging and/or for delivering current to an electrical load) are subject to phase currents of different levels, as clearly illustrated in the above identified documents. Under such peculiar conditions of operation, the possibility offered by the present invention to "modulate" the effective cell size of the cells in electrical series that form the multi intermediate voltage tap battery according to the different load conditions (current) that exist among the different phase switchings of de-construction (when charging from an AC source) and of re-construction (when powering an AC load) of an AC waveform, permits an outstanding flexibility for optimizing or best adapting the characteristics of the battery system and sensibly restricting the range of variation of the battery voltage.

BRIEF DESCRIPTION OF THE DRAWINGS

The characterizing aspects and features of the redox flow battery of this invention will be further clarified in the following description of several embodiments and by referring to the attached drawings, wherein:

Figure 1 is a basic electrical diagram of a four compartment monopolar cell module according to this invention; Figure 2 is a functional electrical diagram of the module of Figure 1 showing the switches of partialization (sizing) of the effective cell area;

Figure 3 is a hydraulic diagram of the module of Figures 1 and 2, showing the electrolytes flow ducting;

Figure 4 is a functional electrical diagram of a battery of four monopolar cell modules of modifiable effective area, in electrical series;

Figure 5 is a perspective sectional view of a multi-compartment monopolar cell module;

Figure 6 is an enlarged functional diagram of a battery stack composed of four monopolar cell modules, the cell area of which may be incremented by 100% and by an additional 50% of a certain basic (minimum) cell area, employing exclusively monopolar electrodes;

Figure 7 is a plot of voltage-current characteristics of a monopolar cell module for different conditions of partialization of the effective cell area; Figure 8 is a plot of voltage-current characteristics of a battery composed of four stacked monopolar cell modules for different configurations of partialization of the effective cell area;

Figure 9 is a functional diagram of a battery stack equivalent to the one of Figure 5 but employing monopolar electrodes and bipolar electrodes;- Figures 10 and 11 are respectfully a top view and a bottom view of a stackable monopolar electrode element of the module shown in Figure 5;

Figures from 12 to 18 are cross sections of the frame portion of the stackable element of Figures 10 and 11, showing the various features thereof

DESCRIPTION OF EMBODIMENTS OF THE INVENTION

As depicted, by way of example, in Figures 1, 2, 3 and 4, a stackable monopolar cell module MM is divided by a certain odd integer number n of membranes 1 (in the above-mentioned figures n=3) into n+1 flow compartments, alternately for the positive electrolyte solution as shown in Figure 3, and for the negative electrolyte solution, and into which are alternately installed positive electrodes 2 and negative electrodes 3.

By electrically connecting into the electrical circuit of the battery one or more electrodes of the same polarity of the monopolar cell module, the effective cell area is effectively partialized. In the example shown in the figures, the effective cell area of a monopolar cell module may be doubled, by closing the switch al of Figure 2 or the switches al, bl, cl and dl of Figure 4, or tripled, by closing also the switch a2 of Figure 2 or the switches a2, b2, c2, d2 of Figure 4.

Depending on the arrangement of interconnections that is implemented, the effective (working) cell area may be incremented/decremented (partialized) by an integer factor up to a maximum factor corresponding to an odd number n, that is to the number of perm-ionic membrane separators of the multi-compartment monopolar cell module.

A simplified three-dimensional cross sectional view of a multi-compartment monopolar cell module is shown in Figure 5.

In the embodiment shown in Figure 5, the module MM includes n=5 membranes 1 that define n+l=6 flow compartments each housing alternately a positive electrode 2 or a negative electrode 3. Apart from the positive end electrode 2 and the negative end electrode 3, all the electrodes there between have a "two-face" structure.

Generally, each electrode, indicated as a whole with 2 or 3 in the Figure, includes a pervious or impervious base plate, 2a and 3 a, of an electrically conductive material, chemically resistant to the electrolyte solutions, that may be either a metallic material, a conductive cermet material, glassy carbon, graphite or a conductive aggregate of a binder (typically a moldable plastic binder) and of particles and/or fibers of said conductive materials. The base plate, 2a and 3a, has primarily the function of current collector of substantial rigidity and mechanical strength to constitute, in a "tongue" or "tab" portion (not visible in the section of Figure 5) projecting out of the outer perimeter of the molded plastic frames 4 of the stackable electrode elements, electrical terminals suitable to be ordinarily connected by common fixtures with copper wirings in the battery circuit.

Most preferably, as indicated in Figure 5, each electrode has an active or primarily working portion generally in the form of a porous three-dimensional conductive structure, 2b, 3b, through open pores of which the respective electrolyte solution may readily flow or circulate. For example, the working three- dimensional porous portion, 2b, 3 b, of each electrode may be a mat or felt of carbon fibers in electrical contact with the current collecting (or distributing) base plate 2a, 3 a, and that may extend to bear at a plurality of points onto the surface of the perm-ionic membrane separator 1. The membrane 1 may in practice be sandwiched between the porous mats of the positive and negative electrodes of the two flow compartments separated by the membrane, upon tightening of the filter- press assembly, exploiting the ability of the carbon felt to be moderately compressed in an elastic fashion.

Ordinary die-stamped gaskets of a chemically resistant elastomer (not shown in the figures) may be placed over at least one of the counter opposed surfaces of two adjacent plastic frames 4 of the stackable electrode elements for providing a leak proof hydraulic sealing of the flow compartments according to common filter-press assembling practices.

The flow of the two electrolyte solutions of opposite polarity, each containing the respective redox couple, through the respective electrode (flow) compartments of the monopolar cell module occurs through respective inlet and outlet manifolds 5 and 6, communicating through port holes 7 and 8, with the respective flow compartments.

In the cross section of Figure 5, the arrows schematically indicate the flow path of the negative electrolyte solution (negative electrolyte). The inlet manifold 5, the outlet manifold 6, the inlet port holes 7 and the outlet port holes 8 may all be defined by appropriately aligned holes and borings or slots made in the plastic frames 4 of the stackable monopolar electrode elements, according to common practices in the art.

Of course, similar inlet and outlet manifolds and respective inlet and outlet port holes are also present for operatively flowing the positive electrolyte in the other flow compartments of the multi-compartment monopolar cell module.

In the embodiment shown in Figure 5, the module includes two end plates 9 and 10 that render the stackable module completely "self contained", usable as a discrete component (unit or elementary cell) of a battery stack assembled to achieve a specifically required nominal battery voltage (that is a multiple of the monopolar or elementary cell voltage).

As shown in the Figure 5, the end plates 9 and 10 as well as all the frames 4 that composed the module MM, may be provided with through holes 11 and 12 in areas external to the outer perimeter of the hydraulic seal, through which tie rods may pass for tightening the battery stack in a filter-press mode, according to common practices in the art.

The frame 4 as well as the end plates 9 and 10 may all be made of molded plastic, for example of polyethylene, polypropylene and equivalent moldable non- conductive and chemically resistant materials.

For example, a battery stack with a nominal voltage equal to four times the unit cell voltage, may be realized by stacking together four multi-compartment monopolar cell modules of the type illustrated in Figure 5.

A stack could have a functional scheme as depicted in Figure 6.

In Figure 6, to facilitate the reading of the diagram, the monopolar electrodes or more precisely the porous mat portion thereof are drawn with different shadings for distinguishing negative electrodes from positive electrodes, as also indicated by the relative symbol of the respective electrolyte flow compartment.

The negative and positive electrodes may be assumed to have the same composition though they may even be made of different materials and/or with a different structure and/or morphology, talcing into consideration the different half- cell reactions that occur on a positive electrode (oxidation) and on a negative electrode (reduction).

In Figure 6 are schematically shown the switches al, a2, bl, b2, cl, c2 that permits to adapt the effective cell area to the current level, during discharging as well as during recharging of the battery, consequent to the sample structure depicted in this figure.

By closing only the switches al, bl, cl, and dl, the cell area may be doubled.

By closing only a2, b2, c2, and d2 the cell area of the relative monopolar cell module MM(i) may be increased by 25%.

By closing all the switches, the cell area may be increased by one and a half the basic area, which in the examples so far illustrated in the figures corresponds to twice the projected area of an electrode/membrane/electrode assembly.

The effects that may be implemented by the partializing possibilities provided by a battery architecture of this invention will now be analyzed for a sample embodiment. Let Vo be the open circuit voltage of a cell. Let I be the current, and let A be the working or effective area of a cell. Then the voltage across the cell during charging is

V_c = Vo + kl/A,

where k is a constant which is determined by the properties of the cell.

In the case of multi-compartment cell module battery , for example according to the electrical scheme embodiment circuit of Figure 4, the effective cell area is given by

A = (n + l)Ao,

where Ao is the basic cell (minimum) area, and n is the number of area partialization switches that are closed. The discharge voltage across the cell is

V_d = Vo - kl/A = Vo - kl/(n + l)Ao.

Let the state of charge of the electrolyte solutions be 50%; then V₀ = 1.35 V. Let the design voltage efficiency Vd/V_c be 0.85 when n = 0. Then l /Ao = 0.11 V, V_c = 1.46 V and Vd = 1.24 V. For a given basic (minimum) cell area Ao there exists a current Io at which these cell voltages are verified.

In these calculations it is assumed that the ohmic resistance of the electrode structures (electronic current carrying structures as distinct from ionic current carrying media) can be neglected compared with the electrical resistance of the electrolyte solution and of the membrane.

In the batteries used for the tests: Ao = O.lmxO.lm = 0.01m² and I₀ = 4A; therefore k = 2.75xlO^"4Ωm².

When the current I drawn from the battery varies, the discharge voltage Vd of each cell varies in a substantially linear fashion as shown in Figure 7.

The equations of the three plots shown are:

Vd = Vo - (l o/Ao)(I Io) when both switches are open (n = 0),

Vd = Vo - (kIo/2Ao)(LTo) when at is closed and a₂ is open (n - 1),

Vd = No - (kI₀/3Ao)(I/Io) when both switches are closed (n = 2).

The design discharge voltage 1.24 V occurs at I = I₀ when n = 0, at I = 2I₀ when n = 1, and at 1=31₀ when n = 2.

By automatically configuring the cell area partializing switches, for example by monitoring the battery voltage, comparing it with specified thresholds, it is possible to keep the discharge voltage between specified limits of variation, namely: V_d(min) = 1.20 V and Vd(max) = 1.28 V when the current absorbed by the load varies between a minimum value: I(min) = 0.67 I₀ and a maximum value: I(max) = 4.09 I₀.

Let n_a, n_b, n₀, and nd be the number of cell area partializing switches of the respective multi-compartment monopolar cell modules MMa, MMb, MMc, and MMd that compose the battery of Figure4, that are closed at a certain time. Then the discharge voltage of the battery Vd is given by

V_d = V_d(a) + Vd(b) + V_d(c) + V_d(d)

= 4Vo - [l/(n_a + 1) + l/(n_b + 1) +l/(n_c + 1) +l/(n_d + l)](kI₀/Ao)(I/Io).

This voltage depends on the combination of the four parameters n_a, n_b, n_c, and n , but not on their permutation.

Table 1 shows all the possible switch configurations for the considered sample embodiment of Figure 4 and the battery voltage that they produce when the current is I₀.

The open circuit battery voltage is 4Vo = 5.4V, and kI₀/Ao = 0.1 IV. Therefore, each cell module has voltage efficiency 85%, but the overall voltage efficiency for the four module battery rises above 85% as the cell area partializing (incrementing) switches are progressively closed.

Table 1.

n_a nb n_c, n_d Battery Voltage Voltage Efficiency

0 0 0 0 4.960 0.85

1 0 0 0 5.015 0.87

1 1 0 0 5.070 0.88

1 1 1 0 5.125 0.90

1 1 1 1 5.180 0.92

2 0 0 0 5.033 0.87

2 1 0 0 5.088 0.89

2 1 1 0 5.143 0.91

2 1 1 1 5.198 0.93

2 2 0 0 5.107 0.90

2 2 1 0 5.162 0.92

2 2 1 1 5.217 0.93

2 2 2 0 5.180 0.92

2 2 2 1 5.235 0.94

2 2 2 2 5.253 0.95

Of course, plots of the discharge battery voltage Vd versus current (expressed as a ratio I/lo) can be drawn for each of the above configurations, as depicted in Figure 8. An appropriate configuration of the cell area partializing switches allows to keep the discharge voltage between specified limits: Vd(min) = 4.93 V and Vd(max) = 4.99 V, while the current varies between the values: I(min) = 0.92 Io and I(max) = 3.20 Io.

Unlike the case of traditional bipolar batteries, wherein the elementary cells have the same membrane and electrode area, and a constant battery voltage cannot be maintained while the load on the battery varies, a battery made according to this invention, in the form of a stack of multi-compartment monopolar cell modules with cell area partializing switches, can provide a substantially constant battery voltage a substantial constancy of the battery voltage may be provided notwithstanding variations of the current absorbed by the load between certain minimum and maximum values.

Though the total voltage of the battery stack corresponds to the sum of the voltages of the individual monopolar cell modules that compose the battery, the operating voltage of each monopolar cell module may be adjusted in relatively small steps by varying its effective area independently of the other cells. By virtue of the fact that any individual cell of the battery stack, connected in series, may have its area partializing switches configured in a way such as to produce a cell voltage different from another individual cell of the stack, the resolution of the voltage control that can be performed is particularly fine and the resulting battery voltage can be kept within relatively narrow limits over a wide range of currents.

There is no practical limit to the number of membranes and of compartments of each monopolar cell module. Therefore, it is possible for a properly designed battery stack to support a large current even though the area of the stackable elements (footprint) is small.

The modules in a stack may have different numbers of flow (electrode) compartments and of area partializing switches. This is a great advantage that may be exploited in designing battery stacks destined to applications that contemplate a process of AC waveform discretization/reconstruction, according to the systems disclosed in said prior applications WO 03/007464 and PCT/IT02/00653.

For example, the monopolar cell modules (normally at one end of the stack) that support the relatively large phase currents in coincidence with the peaks of a discretized AC waveform being reconstructed may have a proportionately larger number of selectable elements than the cells at the other end of the battery stack that support relatively smaller phase currents, some of which may even be ordinary bipolar cells of fixed area.

Although a battery stack according to this invention may be composed by stacking four modules of the self contained type described in Figure 5 as schematically depicted in Figure 6, a battery stack may be realized using stackable bipolar electrode elements as joining elements between adjacently stacked monopolar cell modules.

This alternative embodiment in terms of structure of the battery stack is schematically depicted in Figure 9, for the case of a battery stack composed of five monopolar cell modules of partializable cell area.

The bipolar electrodes are recognizable by the different shadings over opposite faces thereof that indicate the polarity of the porous three-dimensional electrode mats in electrical content with the base plate (current collector) that in this case must of course be solid (without pores) and impervious to any flow there through by the electrolyte solutions of opposite polarity flowing in contact and through the pores of the respective electrode mats on the opposite faces of the current collecting base plate.

Of course, also the position of the cell area partializing switches is adapted to the use of bipolar electrodes for structurally and electrically joining two adjacent multi-compartment monopolar cell modules.

The use of bipolar electrodes according to the embodiment of Figure 9, favors a reduction of the total height of stack by eliminating the presence of juxtaposed pairs of plastic end plates, 9 and 10, present in the stack assembly of Figure 6.

A typical stackable element for composing the multi-compartment monopolar cell module of the invention may be constructed as in the example shown in Figures 10 to 18.

The conductive base plate 2a (or 3 a) may be in the form of a glassy carbon plate, preferably having a metallic core which may be a zinc expanded metal plate or aluminum or another highly conductive metal or alloy.

The base plate should have a good electrical conductivity (enhanced by a metal core sheet) and be perfectly resistant to chemical attack by the electrolyte solution in contact therewith (normally a strong acid). In the example shown, the elements have a square area and the base plate 3a has a substantial portion of its perimeter (at least two opposite sides) permanently embedded in a plastic frame 4 molded thereon. At least one and preferably two tongue portions 3t of the base plate 3a extend outside the frame portion 4 molded over the square perimeter of the base plate to constitute electrically connectable terminals of the monopolar electrode.

The molded frame portion 4 has four corner holes 13 for accommodating tie rods of the filter press assembly.

Moreover, the plastic frame portion 4 has a plurality of through holes 14 distributed along the fours sides thereof, and located in an outer portion of the molded plastic frame 4, clear of the perimeter of the embedded conductive base plate 3 a.

Depending on whether the stackable monopolar electrode element is a positive electrode or a negative electrode, the corresponding electrolyte solution is circulated in the compartment of the monopolar cell module containing the electrode by producing in the molded plastic frame a plurality of inlet slots or ports 15, communicating with inlet manifolds defined by the through hole 14 from the same side (inlet side) of the plastic frame and outlet slots or ports 16 communicating with the outlet manifolds formed by the through holes 14 disposed on the opposite side of the plastic frame.

The stackable element, shown from one side in Figure 10 and the reverse side in Figure 11, is by way of example an element containing a negative electrode 3 a. Of course, an adjacently stackable electrode element will be contain a positive electrode 2a and the inlet and outlet ports will place the respective flow compartment in connection with the inlet manifold constituted by the through holes 14 present in the lower side of the molded frame and with the outlet manifold constituted by the through hole 14 present on the upper side of the molded plastic frame.

In both cases, the fact that the inlet and outlet manifolds are constituted by several through holes through which the electrolyte solution flows in parallel, and the fact that consequently the port holes for flowing part of the electrolyte through the cell compartment housing the relative monopolar electrode, provides for an even distributions of the electrolyte in the electrode compartment and through the porous carbon felt active electrode placed in electrical contact over the entire surface of the conductive base plate 3 a.

As shown in the previous Figures 10 and 11, the molded plastic frame may be provided with matching recesses and protrusions 17, 18 of differentiated shape, to facilitate the stacking of one element over the other in a perfect alignment and for preventing orientation errors.

Appropriate gaskets may be interposed between the frames.

The perm-ionic membrane separator is normally placed between the gasketed surfaces of adjacent frames. Alternatively and preferably, the membrane, cut to size, may have its perimetral portion retained into the annular groove 18, purposely cut over a face of each electrode frame, using a retainer ring (not shown in the drawings) for pushing the membrane into the groove 18 upon tightening together the stacked elements. The set of Figures 12-18 are sectional views of the frame portion of the stackable element of Figures 10 and 11 all of these respective plane sections are identified by the capital letters in the figures.

Claims

C L A IM S

1. A redox flow battery including a plurality of cells electrically connected in series between two end terminals of the battery and composed of elements all of a certain active area and stackable in a filter-press arrangement, with positive and negative electrolyte solutions flowing in parallel and/or in cascade, respectively through positive electrolyte flow compartments and through negative electrolyte flow compartments of said cells and a plurality of externally connectable intermediate voltage taps between said two end terminals, characterized in that at least one of said cells of said stacked filter-press battery assembly has a cell area that is an integer number multiple of said certain active area defined by the dimensions of said stackable elements.

2. The redox flow battery of claim 1, wherein said cells of different areas are filter-press stackable multi-compartment monopolar cell modules, each composed of a certain even integer number n+1 of flow compartments alternately for said positive electrolyte solution and for said negative electrolyte solution separated by an odd integer number n of perm-ionic membrane separators, and alternately housing a positive and a negative electrode.

3. The redox flow battery of claim 2, characterized by comprising a plurality of intermediate voltage taps, one of said taps coinciding with at least an end electrode of one of said multi-compartment monopolar cell modules.

4. The redox flow battery of claim 2, characterized in that each of said multi-compartment monopolar cell modules employs at least a two-face composite electrode of a first polarity, including a conductive base plate, over opposite faces of which are disposed porous fluid permeable three-dimensional electrode mats contacting over at least a plurality of points said perm-ionic membrane separating the flow compartment containing said two-face electrode from a flow compartment containing an electrode of a second or opposite polarity.

5. The redox flow battery according to claim 2, characterized in that back- to-back juxtaposed end electrodes of two of said filter-press stackable multi- compartment monopolar cell modules of said battery assembly constitute a fluid impervious bipolar electrode, electrically joining a monopolar cell module to the other monopolar cell module in said electrical series of the battery.

6. The redox flow battery according to claim 5, wherein said fluid impervious bipolar electrode constitutes one of said intermediate voltage taps of the battery.

7. A redox flow battery according to claim 2, characterized in that each monopolar cell module has cell area partializing switches for connecting in common a selectable number of either positive or negative electrodes composing the cell module.

8. The redox flow battery of claim 2, characterized in that all the electrodes of each multi-compartment monopolar cell module have a tab portion extending outside the hydraulic seal perimeter of the flowing electrolyte solutions of the stackable elements of said filter-press stack assembly for electrical connection of the electrode.

9. The redox flow battery of any of the precedent claims characterized in that it is composed of a plurality of filter-press assembled stackable electrode elements, each composed of a conductive base plate, a perimetral portion of which is embedded in a molded frame portion of non conductive material and having one or more tab portions protruding out of the outer perimeter of said molded frame portion with the function of a connectable outer terminal of the electrode base plate.

10. A method of operating a redox flow battery in the form of a filter-press stack of cells including at least one cell having a cell area that is a multiple of a certain active area defined by the dimension of stackable elements of said filter- press assembly, characterized in that the method includes varying the effective cell area of said cell or cells of multiple area as a function of the current flowing through the battery to limit the variation of the battery voltage.

11. The method according to claim 10, characterized in that the effective cell area is automatically incremented or decremented as a function of the monitored level of the current through the battery by closing or opening cell area partialization switches, each connecting either a positive or a negative electrode of said certain active area in common with other electrodes of the same polarity belonging to said cell of multiple area.