Method and arrangement for leakage detection for a flow battery

Abstract

To improve the detection of leakages in flow batteries (1), a method according to the invention comprises the steps of flowing, at a first half-cell pressure (p1), a first test fluid (101) through a first group of half-cells (41); filling, at a second half-cell pressure (p2), a second test fluid (102) into a second group of half-cells (42); measuring, in an outlet line (14) of the first group of half-cells (41), a concentration of the second test fluid (102) in the first test fluid (101); deducing the presence of a leakage path from the first group of half-cells (41) to the second group of half-cells (42) when the measured concentration of the second test fluid (102) in the first test fluid (101) exceeds a predetermined threshold value.

Description

[0001] Method and arrangement for leakage detection for a flow battery

[0002] Field of the invention

[0003] The present invention relates to a method as well as to an arrangement for detecting a leakage path between a first and a second group of fluidly connected half-cells of a flow battery, preferably a flow battery using dissimilar electroactive species, most preferably a redox flow battery, the first and the second group of half-cells separated from one another at least partially by a membrane of the flow battery.

[0004] Background

[0005] A (redox) flow battery is an electrochemical device for storing energy on an electrochemical basis, usually comprising electrolyte tanks for storing positive and negative electrolytes and pumps and lines for circulating the electrolytes through one or more cell stacks having a number of individual cells. The individual cells of the cell stacks are each formed by a positive half-cell and a negative half-cell arranged side by side (hereafter, groups of positive and negative half-cells are also referred to as first and second groups of half-cells), the positive and negative half-cells of an individual cell being separated by a membrane, typically a semi-permeable, ion-exchange membrane. A membrane of a flow battery is, for example, a cation and / or anion exchange membrane, e.g. based on a sulphonated fluoropolymer such as polytetrafluoroethylene (PTFE). Flow batteries and their components, i.e. , membranes, flow paths through a battery created by said groups of half-cells, electrodes, etc., are covered at length in the prior art, cf. , e.g., WO 2007 / 052744 A1 or WO 2023 / 110799 A1, both documents providing extensive implementation details.

[0006] The half-cells in the positive or first group of half-cells contain a positive electrode in a frame through which the positive electrolyte flows. The half-cells in the negative or second group of half-cells contain a negative electrode in a frame through which the negative electrolyte flows. The positive electrolyte and the negative electrolyte are circulated separately through the half-cells. The positive and negative electrodes are usually made of porous felts of graphite through which said electrolytes can flow. Electrode plates, for example bipolar plates, are arranged between individual neighboring single cells of the cell stack as current collectors, which are usually made of a composite material comprising carbon and plastic. Current collectors are located on the electrode plates on the axially outer sides of the axially outer individual cells of the cell stack, via which an electrical contact is routed to the outside in order to be able to tap an electrical voltage across the entire cell stack (discharging the redox flowbattery) or to be able to apply an electrical voltage to the cell stack (charging the redox flow battery). The cell stack is terminated on the outer axial sides by an end plate, which holds the cell stack together.

[0007] An electrolyte of a redox flow battery is a liquid and essentially comprises an electrochemical redox pair consisting of a first redox element and a second redox element, each in the form of a redox-active element or ion or combinations of redox-active elements and / or ions with different electrical charges (oxidation numbers). There are a large number of redox-active elements or ions or combinations of redox-active elements and / or ions that can be used as redox pairs in a redox flow battery, or generally in flow batteries. The redox elements of a redox couple are usually dissolved in the electrolyte liquid. The electrolyte liquid is usually an aqueous acid, such as aqueous sulphuric acid. Redox flow batteries with a wide variety of combinations of redox pairs are known. Some non-exhaustive examples of known combinations of redox pairs are V2+ / V3+ vs. VO2+ / VO2+ (in a vanadium redox flow battery), V2+ / V3+ vs. Br- / CIBr2, Br2 / Br- vs. S / S2-, Br- / Br2 vs. Zn2+ / Zn, Ce4+ / Ce3+ vs. V2+ / V3+, Fe3+ / Fe2+ vs. Br2 / Br-, Fe3+ / Fe2+ vs. Cr3+ / Cr2+, Mn2+ / Mn3+ vs. Br2 / Br-, Fe3+ / Fe2+ vs. Ti2+ / TiO2+ and others. The redox elements can be contained in the electrolyte in different chemical compounds, for example in the form of sulphates, such as vanadium sulphate, or chlorides, such as vanadium chloride.

[0008] In a vanadium-based redox flow battery, the positive electrolyte in the charged state consists of a redox pair in the form of vanadium with the oxidation number +4 (also referred to as V(IV) or V4+) and vanadium with the oxidation number +5 (also referred to as V(V) or V5+). In the charged state, the negative electrolyte consists of a redox pair in the form of vanadium with an oxidation number of +2 (also referred to as VII or V2+) and vanadium with an oxidation number of +3 (also referred to as VIII or V3+), which means that the negative electrolyte has a more negative electrochemical potential than the positive electrolyte.

[0009] When operating a redox flow battery, electrical energy is released to a consumer or absorbed from an energy source. In the example of a vanadium-based redox flow battery, the known chemical reaction (redox reaction) 2H++ VC + V2+<- VO2++ V3++ H2O takes place during charging I discharging in a single cell of the redox flow battery. During the charging process, the reaction that takes place is caused by an external current source or voltage source through which an electrical voltage is applied to an individual cell (or the entire cell stack). To operate the redox flow battery, the electrolytes must be circulated through the cell stack, specifically through the individual cells of the cell stack, as is well known from the prior art, for example from WO 2018 / 087270 A1 or from WO 2014 / 131702 A1.While redox flow batteries are an established and well-understood technology, also a series of, to this date still unsolved, technical problems relating to redox flow batteries are reported. A particularly problematic topic concerns the electrolyte circuits through which the electrolyte liquids discussed above flow. In particular, redox flow battery stacks can leak, either early due to manufacturing errors or later in their lifespan, due to materials degradation or due to other causes. Leakages in flow batteries are frequently categorized as external and internal leakages, external leakages being understood as an electrolyte liquid leaving the battery system and contaminating its environment, and internal leakages meaning that an electrolyte may reach an area inside a battery system where it is not supposed to be, potentially causing harm to the battery system itself. For example, leakage can start with the failure of an internal stack seal and eventually become visible externally. Contact between liquid electrolyte liquid in a redox flow battery and an existing current collector must be fundamentally prevented since the acidic (e.g., in the case of a vanadium redox battery) or basic electrolyte liquid may cause corrosion of the usually metallic current collector. Aside from electrolyte contamination issues, small external leaks also present a potential hazard to maintenance workers (both chemically and electrically because the electrolyte may have an electrical potential different to surrounding conducting elements) and can lead to corrosion of surrounding equipment (e.g., battery rack, housing, neighboring cell stacks). Various approaches have been suggested in the art for monitoring for and for providing early warning of external leaks in such systems, see, e.g., JP 2012099416, or US2008050646 or WO 2022 / 248633 A1.

[0010] With regards to internal leakages, especially flow batteries with dissimilar electrolytes (electrolytes that do not have the same electroactive species) suffer from permanent capacity loss, if one electrolyte mixes with the other. Typical examples are electrolytes that may have different inorganic redox couples (e.g., Fe-Cr flow batteries), organic-inorganic redox couples, or all organic redox batteries (where the organic molecules differ in posolyte and negolyte). Once such a mixing is detected, it is important to identify where the leakage is occurring, so that specifically an affected component may be repaired or replaced. If the leakage cannot be located, all flow stacks may have to be replaced, which is wasteful and uneconomic. Hence, a detection method for detecting internal leakage paths must be sensitive, also to small leakages, because flow batteries are required to operate for many years with only low decrease in capacity. Further, a detection method must also be suitable for use in energy storage systems in the field, where batteries within the energy storage system contain multiple flow stacks.

[0011] For example, an acceptable rate of capacity loss may be <10% over 10 years. Therefore, if both electrolytes have similar charge densities and react together in a 1:1 ratio, then the acceptable mixing would be <10% of the electrolyte volume from one electrolyte into the otherover 10 years. Considering a system with 10000 L of electrolyte in both tanks, this would be equivalent to 1000 L mixing in 10 years or 0.2 ml min-1. This rate of mixing is much lower than a typical electrolyte flow rate through a stack. For a typical 28-cell stack, flow rates usually are above 1 L min-1. Therefore, the maximum acceptable crossover rate is <0.02% of the total flow rate. Thus, many known techniques that could be used to detect larger leaks (temperature rise, change in flow rate / pressure, change in state-of-charge, change in color or physical properties) would have to be of extremely high precision, which can usually not be achieved with reasonable outlay or effort, rendering known methods impractical and / or uneconomic for the use of internal leakage detection and internal leakage prediction in a flow battery.

[0012] Problem to be solved by the invention

[0013] Hence, it is an object of the invention to further improve the detection of leakages in flow batteries.

[0014] Summary

[0015] This object, for the method mentioned at the outset, is achieved by additionally carrying out the steps of flowing, at a first half-cell pressure, a first test fluid through the first group of halfcells; filling, at a second half-cell pressure, a second test fluid different from the first test fluid into the second group of half-cells, the second half-cell pressure exceeding the first half-cell pressure to allow for a leakage flow of the second test fluid into the first group of half-cells; measuring, in an outlet line of the first group of half-cells, a concentration of the second test fluid in the first test fluid; and deducing the presence of a leakage path from the first group of fluidly connected half-cells to the second group of fluidly connected half-cells when the measured concentration of the second test fluid in the first test fluid exceeds a predetermined threshold value.

[0016] In contrast to methods known from the prior art, such as monitoring changes in an open-circuit voltage of electrolyte leaving a flow battery stack in standard operation with equal (typically atmospheric) pressures in the respective half-cells, the usage of an increased pressure in the second group of half-cells allows to significantly improve both measurement resolution and measurement accuracy, while requiring almost no additional outlay. The approach according to the invention will force the second test fluid through any leakage path, e.g., (potentially very small) pinholes in the membrane, decrease in membrane permselectivity, defectiveseals, damaged bipolar plates, etc., to enter the opposing electrolyte. All of these types of leakage paths can be detected accurately by means of the invention.

[0017] As explained previously, detecting leakages in redox flow batteries is critical for safety, performance, and economic reasons. Leakages can expose individuals and the environment to toxic or corrosive electrolytes, potentially causing health hazards, environmental contamination, and even fire risks. Early, efficient and precise leakage detection is crucial to ensure the safe, efficient, and long-lasting operation of flow batteries. The invention allows for improvements in all of these aspects.

[0018] Specifically, the second half-cell pressure may exceed the first half-cell pressure by at least 1mbar, or by at least 10mbar, or by at least 50mbar, or by at least 100mbar, or by at least 500mbar, or by at least lOOOmbar. However, also other pressure ratios or other pressure differences may be applied, should this turn out to be beneficial for a specific setup in which a leakage is to be detected. There are several options to arrive at a pressure value representing a half-cell pressure of a group of half-cells. For instance, a single pressure value may be taken into account, e.g., by computing a (lumped) average pressure of a group of half cells, such that the first half-cell pressure represents an average pressure through the first group of half cells and the second half-cell pressure represents an average pressure through the second group of half cells. In another preferred embodiment, the first half-cell pressure considered in the invention may be selected as a pressure of the first test fluid acting in an inlet of the first group of fluidly connected half-cells, i.e. an inlet allowing to connect the group of halfcells with an external reservoir etc., and / or the second half-cell pressure may in the same fashion be chosen as a pressure of the second test fluid acting in an inlet of the first group of fluidly connected half-cells. However, the invention allows for flexibility in this regard. Specifically, e.g., also a first geometrical pressure-point different from the inlet may be selected in the fluid path of the first group of half cells as a point where a pressure is monitored as a first half cell pressure, and a second geometrical pressure-point in the second group of half cells, also different from the inlet of the second group of half cells, may be selected as a point to monitor the pressure. In such a case it may be abstained from calculating an average pressure. In this case, in a particularly advantageous embodiment, the second geometrical pressure-point may be chosen as the point within the second group of half cells with the shortest distance to the first geometrical pressure-point.

[0019] The predetermined threshold value, whose exceedance may be used as an indicator for the presence of a leakage, may be provided in the form of a an absolute concentration value, or in the form of a ratio between a concentration measurable in an uncontaminated first test fluid, preferably at an inlet of the first group of fluidly connected half-cells, and a concentration measurable in a contaminated first test fluid, preferably at an outlet of the first group offluidly connected half-cells, the ratio preferably being selected as a ratio of 0.5 or as a ratio of 0.1 or as a ratio of 0.01 or as a ratio of 0.005 or as a ratio of 0.001. With regards to absolute concentration values, 0.01 moles per liter or 0.05 moles per liter or 0.1 moles per liter or 1 mole per liter may be considered as thresholds. However, for liquid contaminants, mass concentration (e.g., g / L, ppm, ppb), weight / weight percentage, volume / volume percentage, and number concentration may as well be employed to express thresholds, each offering distinct advantages depending on the context. Gaseous contaminants often utilize volume / volume percentage (e.g., ppm, ppb) or mass concentration (e.g., mg / m3). Furthermore, specialized units like activity (e.g., Bq, Ci) may be used for radioactive contamination, while surface and linear contamination may be quantified using mass per unit area or length, respectively.

[0020] By means of the invention, not only the mere presence of a leakage path can be determined, however. In a preferred fashion, it is also conceivable to determine the specifics of a detected leakage path, e.g., its diameter and / or length and / or size and / or location etc., depending on the level of the concentration of the second test fluid. As elaborated earlier, information about leakages and leakage paths in a flow battery, such as their location, may be used in various ways. For example, in preferred embodiments, a specific flow stack may be replaced after finding that it is affected by leakage, or an electric load of a flow battery may be adapted, or a flow rate of an electrolyte may be adjusted etc.

[0021] As will be explained in detail later, the invention allows to achieve a second half-cell pressure higher than a first half-cell pressure in two ways. On the one hand, it is possible to flow both the first test fluid as well as the second test fluid through the respective group of half-cells, i.e., from an inlet of the group of half-cells to an outlet of the group of half-cells, i.e. through the flow battery. On the other hand, it is also possible to flow second test fluid into the second group of half-cells until a desired pressure level is achieved in the second group of halfcells, and then close the outlet of the second group of half-cells and also stop the further supply into the second group of half-cell, keeping the pressure close to the desired value. However, in both scenarios, the second group of half-cells is filled with the second test fluid, as outlined above.

[0022] As mentioned above, the invention also suggests an arrangement to carry out the method outlined above. To that end, the arrangement mentioned at the outset may be equipped with a first flowing means, preferably a first pump, to flow a first test fluid through the first group of half-cells at a first half-cell pressure; a second flowing means, preferably a second pump, to fill a second test fluid different from the first test fluid into the second group of half-cells at a second half-cell pressure, the second half-cell pressure exceeding the first half-cell pressure to allow for a leakage flow of the second test fluid into the first group of half-cells; detection means to measure a concentration of the second test fluid in the first test fluid in an outletline downstream the first group of half-cells and to indicate the presence of an internal leakage path when the measured concentration of the second test fluid in the first test fluid exceeds a predetermined threshold value.

[0023] Further preferred embodiments are reflected in the dependent claims, which will be explained in detail in the following.

[0024] Detailed description

[0025] The present invention is described in greater detail below with reference to Figures 1 to 4, which show schematic and non-limiting advantageous embodiments of the invention by way of example. The specific examples described herein are only used to explain the content of the present invention and are not intended to limit the present embodiment. In the figures:

[0026] Fig.1 shows a schematic view of a redox flow battery system whose redox flow battery comprises a single stack of redox flow cells compressed together into a series stack,

[0027] Fig. 2 shows a cross-sectional view of several cells near the lower end of the series stack,

[0028] Figs. 3a and 3b show schematic diagrams of exemplary leak detection circuits suitable for use in the present invention,

[0029] Fig. 4 shows a block diagram representing possible steps in a particularly advantageous procedure in which the present invention may be used.

[0030] With reference to Figs.1 and 2, the known construction of a common redox flow battery 1 according to the prior art will be explained. However, the invention is by no means restricted to redox flow batteries but may also be used to detect leakage paths in general flow batteries, particularly preferably in flow batteries using dissimilar electroactive species. As is well known, redox flow batteries imply that all electroactive species are dissolved and thus liquid. However, the invention may as well be used in hybrid flow batteries, in which at least one solid or gaseous electroactive species is used, as long as the flow battery has a membrane to define at least two half-cells.

[0031] A cell stack 2 of a redox flow battery 1 comprises a plurality of cells 4. Each cell is formed from a positive half-cell 42 and a negative half-cell 41, i.e., positive half-cells 42 and negative half-cells 41 are arranged alternately in the cell stack 2, the negative half-cells 41 forming afirst group of fluidly connected half-cells 41, and the positive half-cells 42 forming a second group of half-cells 42. A semi-permeable membrane 6, typically an ion exchange membrane (cation and / or anion exchange membrane, e.g., Nation®) is arranged between a positive halfcell 42 and a negative half-cell 41 in a cell 4. An electrode plate 7, for example a bipolar plate, is arranged between two adjacent cells 4. A positive electrode 422 is arranged in the frames 401 of the positive half-cells 42, and negative electrodes 412 are arranged in each of the frames 401 of the negative half-cells 41.

[0032] The positive electrodes 422 and the negative electrodes 412 are usually designed as mats made of carbon or graphite fibers. Via recesses 80 in the frames 401 of the positive half-cells 41 and negative half-cells 42, or cells 4, electrolytically differentially charged electrolyte fluids are pumped during normal operation through the cells 4 by means of pumps 71, 72, wherein the positive electrode 422 in a cell 4 or the respective positive half-cell 42 is perfused by the positive electrolyte fluid, and the negative electrode 412 of the negative half-cell 41 is perfused by the negative electrolyte fluid. In some types of redox flow batteries 1, such as a vanadium redox flow battery or a vanadium polyhalide battery, the two electrolyte liquids are chemically largely similar or have only a different oxidation state in the half-cells (e.g., V2+and V3+, VO2+and VC ). As mentioned above, the invention is not restricted to redox flow batteries, and hence also not to vanadium based redox flow batteries, such that also other types of flow batteries may be used within the scope of the invention, e.g. flow batteries building on electrolytes as iron-chromium, polysulfide-bromine, or uranium.

[0033] Fig.1 further shows tanks 91, 92, in which the electrolyte fluids for operation are stored. In normal operation, i.e. , in the course of power generation or storage, said electrolyte liquids are circulated at pressures p1, p2 close to atmospheric pressure between the negative halfcells 41 or positive half-cells 42 and the negative or positive tanks 91, 92, respectively, using said pumps 71, 72. While a pressure difference is needed between stack inlet and stack outlet to cause electrolyte flow through the stacks, in general, during operation, the inlet pressure is only about 0.01 - 1.5 bar higher than the outlet pressure, with the outlet pressure typically around atmospheric pressure (depending, e.g., on the relative position of stacks to tanks).

[0034] The negative or positive tanks 91, 92 may be spatially separate containers, but may also be formed, for example, as two compartments of a common container, separated by a suitable partition. The cell stack 2 is completed at the two axial ends by a negative end plate 60 and a positive end plate 61 , made, for example, of plastic. The negative end plate 60 and the positive end plate 61 are clamped by clamping means 50 consisting of passing bolts 51, nuts 52, washers 53 and springs 54, thus compressing the frames 401 of the negative half-cells 41 and positive half-cells 42 of the cell stack 2. An electrical connection 19 can be provided oneach of the negative end plate 60 and positive end plate 61 , via which the current collectors 3 in the interior of the redox flow battery 1 can be connected to an external circuit on both sides of the redox flow battery 1.

[0035] Further, the electrolyte fluid connections are provided for the supply and discharge of the electrolyte fluids in the exemplary embodiment at the end plates 60. A positive inlet 921 serves to supply the positive half-cells with electrolyte liquid (i.e., positive electrolyte liquid in normal operation), and a positive outlet 922 serves to return the electrolyte liquid to the positive tank 92 after passing the positive half-cells 42. Similarly, a negative inlet 911 serves to supply the negative half-cells 41 with electrolyte liquid (i.e., negative electrolyte liquid in normal operation), and a negative outlet 912 serves to return the electrolyte liquid to the negative tank 91 after passing the negative half-cells 41. It is to be mentioned that the polarity of the groups of half-cells 41, 42 may also be changed, e.g., by reversing the direction of a voltage during charging, or by interchanging the tanks 91, 92, such the first group of half-cells 41 can act as a group of positive half-cells and the second group of half-cells 42 can act as a group of negative half-cells.

[0036] In order to prevent a possible deformation of the, for example elastic frames of the cells 4 by said contact pressure, spacers 8 may be provided between the negative end plate 60 and positive end plate 61 in order to ensure a constant distance 8' between the negative end plate 60 and positive end plate 61.

[0037] As mentioned at the outset, internal leakages, in particular leakages through the membranes 6, constitute a severe problem in many practical applications. As elaborated earlier, known approaches for detecting internal leakages between negative half-cells 41 and positive halfcells 42 suffer from a series of drawbacks, and in particular do not provide the measurement accuracy needed for accurate long-term leakage prediction. To overcome these drawbacks, the present invention provides an improved method for detecting internal leakages, which mainly builds on the idea of using an appropriately selected pressure drop between the mentioned group of half-cells 41, 42 during a testing scenario outside of regular operation of a flow battery, preferably also using different fluids to simplify the detection of leakages.

[0038] Specifically, the method according to the invention comprises the steps of flowing, at a first half-cell pressure p1, a first test fluid 101 through the first group of half-cells 41; filling, at a second half-cell pressure p2, a second test fluid 102 different from the first test fluid into the second group of half-cells 42, the second half-cell pressure p2 exceeding the first half-cell pressure p1 to allow for a leakage flow of the second test fluid 102 into the first group of halfcells 41; measuring, in an outlet line 14 of the first group of half-cells 41, a concentration of the second test fluid 102 in the first test fluid 101; deducing the presence of a leakage pathfrom the first group of fluidly connected half-cells 41 to the second group of fluidly connected half-cells 42 when the measured concentration of the second test fluid 102 in the first test fluid 101 exceeds a predetermined threshold value.

[0039] In contrast to standard methods known from the prior art, such as monitoring changes in an open-circuit voltage of electrolyte leaving a flow battery stack in standard operation with equal (typically atmospheric) pressures in the respective half-cells 41, 42, the usage of an increased pressure in the second group of half-cells 42 allows to significantly improve both measurement resolution and measurement accuracy, while requiring almost no additional outlay. The approach according to the invention will force the second test fluid 102 through any leakage path (e.g., (potentially very small) pinholes in the membrane, decrease in membrane perm selectivity, defective seals, damaged bipolar plates, etc.) to enter the opposing electrolyte.

[0040] Moreover, the method according to the invention allows to fine-tune the sensitivity of the leakage detection by modifying the pressure difference between the two groups of half-cells. In case higher sensitivity is required, a larger pressure difference between the two flowing paths, i.e. , between the first and second group of half-cells 41, 42, may be applied. In case a smaller pressure difference suffices to achieve a desired measurement accuracy, also smaller pressures may be used, causing less strain to the flow battery. Specifically, it may be provided that the second half-cell pressure p2 exceeds the first half-cell pressure p1 by at least a factor of 1.1 , or by at least a factor of 1.25, or by at least a factor of 1.5, or by at least a factor of 2, or by at least a factor of 5.

[0041] In the scenario according to Fig.3a, the pressures p1, p2 are assumed as lumped parameters, i.e. a single pressure value of pressure p1 represents the fluid pressure inside the first group of half cells 41. As explained earlier, especially an average pressure may be used for this purpose. In many practical applications, the assumption of a lumped pressure is reasonable. In this scenario, the invention makes sure that an average pressure in the second group of half-cells 42 exceeds an average pressure in the first group of half-cells 41. However, as explained earlier, also pressures present at specific spots in the groups of half-cells 41, 42 may be considered, most preferably the pressures present at the inlets 911, 921 of the groups of half-cells 41 , 42. In this scenario, the invention ensures that the inlet pressure of the second group of half-cells 42 exceeds the inlet pressure of the first group of half-cells 41.

[0042] The invention thus provides great flexibility, allowing to adapt the described method to different flow batteries. Another aspect allowing for flexible adaptation of the invention is the predetermined threshold value, which may either be provided in form of an absoluteconcentration value, or in form of a ratio between a concentration measurable in an uncontaminated first test fluid 101 , preferably at an inlet 911 of the first group of fluidly connected half-cells 41 , and a concentration measurable in a contaminated first test fluid 101 , preferably at an outlet of the first group of fluidly connected half-cells 41 (“concentration^ : concentra-tionout”). Such a ratio may be selected as a ratio of 0.5 or as a ratio of 0.1 or as a ratio of 0.01 etc. Moreover, also a change in other properties dependent on concentration change may be tracked, such as changes in density, or in pH-value, or in opacity, or in transparency etc. Spectroscopic methods may be used to indirectly follow concentration changes (i.e. , at least contamination of one electrolyte with the opposing one).

[0043] To further elucidate the concept of the invention, Fig. 3a depicts, in a highly simplified manner, a cell stack 2 comprising a first electrolyte inlet line 911 , a second electrolyte inlet line 912, and respective outlet lines 914 and 913, for flowing the test fluids 101, 102 through the cell stack 2. Further, respective flowing paths through the groups of half-cells 41, 42 are shown schematically, separated by dashed lines and respective membranes 6, with different pressures p1, p2 being present in the different groups of half-cells 41, 42. Possible leakage flows are indicated by small arrows perpendicular to the membranes 6.

[0044] Moreover, detection means 95 are shown in Fig. 3a, in the outlet 914 downstream the first group of half-cells 41 , for measuring the concentration of the second test fluid 102 in the first test fluid 101. Already by visual inspection of the simplified diagram of Fig.3a, it becomes apparent that an increased pressure p2 increases leakages through said membranes 6, making measuring easier. Said detection means 95 may be digital, and may hence be implemented with digital components such as a CPU, a microprocessor, or an FPGA. In case of a digital implementation, the detection means 95 may further be programmed to indicate the presence of an internal leakage path when the concentration of the second test fluid in the first test fluid exceeds said predetermined threshold value. To that end, the detection means 95 may comprise a display via which such an indication may be output graphically. Alternatively, the detection means 95 may be analog, e.g., be a volume to trap the amount of leaked second test fluid 101, whose volume may then be inspected visually. The measurement principle applied to measure a concentration of the second test fluid 102 in the first test fluid 101 may be visual observation, conductivity measurement of a sample of fluid in the outlet of the first fluid, optical transmission (suitable wavelength, photodetector), capacitance, inductance, or impedance measurement of the first test fluid 101 in the outlet, or radioactive emission detection (second test fluid). Moreover, instead of triggering an alarm in case conductive fluids being present, non-conductive gas may be trapped in-line in a gas trap, thus cutting off a voltage measurement within the gas trap, hence using said non-conductive gas leakage detection. In case of different fluids, changes of measured voltages due to dilution could betracked. Also, sensors that may be submersed in-line in a test fluid could be used, the sensor generating changing measurement signals in case the composition of the test fluid changes due to leakage.

[0045] A particularly advantageous aspect of the invention is that it encompasses two alternative variants, utilizing different fluids for the firstand second test fluids 101, 102, which variants can be employed to implement the inventive concept. In the first variant, a gas is used as the second test fluid 102, and an electrolyte liquid is used as the first test fluid 101, preferably an electrolyte liquid as mentioned above (vanadium, iron-chromium, polysulfide-bromine, uranium). This gas is preferably unreactive with the first test fluid 101 , for which it may be selected from a group comprising helium, nitrogen, neon, krypton, argon, xenon, and radon. In a particularly preferred fashion, a non-battery-electrolyte may be used as the first test fluid 101 and / or as the second test fluid. A non-battery-electrolyte may be understood as, e.g., an unreactive gas, or generally an electrolyte that is not suitable for the regular use (energy conversion) in a flow battery.

[0046] In said second variant, an electrolyte liquid selected from a group comprising iron-chromium electrolyte, vanadium electrolyte, polysulfide-bromine electrolyte, and uranium electrolyte is used as the second test fluid 102, and a liquid different from the first test fluid, preferably a liquid selected from a group comprising deionized water, acid and base is used as the first test fluid, hence using exclusively liquids for detecting a leakage. In case of the second variant, further options for measuring a contamination of the first test fluid 101 with the second test fluid 102 can be considered. Specifically, the concentration of the second test fluid 102 within the first test fluid 101 may be quantified by measuring an optical absorption, a pH value, or a redox potential of the first test fluid 101 at the outlet of the first group of half-cells 41 , and compare it to a comparing value of an uncontaminated first test fluid 101.

[0047] A further configuration of a leakage detection circuit, suitable for implementation in the present invention, is illustrated in Fig.3b. Fig.3b illustrates an arrangement for detecting internal leakage paths between a first group of fluidly connected half-cells 41 and a second group of fluidly connected half-cells 42 of a flow battery 1, comprising, in comparison to Fig. 3a, a series of additional components.

[0048] Specifically, the detection means 95 arranged in the outlet line 912 is implemented by means of a gas trap, which allows for a particularly effective and simple measurement in case the first variant of the invention is used (based on gas as the second test fluid 102). As shown in Fig.3b, the gas trap may be formed by a simple bend in the outlet line. However, as is known to one of ordinary skill in the field of flow batteries, a gas trap may be implemented in a variety of other forms as well. As the first test fluid 101 in the form of an electrolyte liquid iscirculated through the stack at a slow pace, any gas that crosses over from the gas-filled half-cells 42 will accumulate in the gas trap, where it can be measured to indicate whether a leakage path through the membranes 6 is present. It is desirable that the entirety of the second test fluid 102 that has leaked into the first group of half-cells 41 is collected in the gas trap.

[0049] Further, the embodiment shown in Fig.3b shows valves 96, 97 in the outlet lines 912, 922 that can be closed. By means of valve 97, a particularly advantageous embodiment can be realized. Specifically, the second test fluid 102 can first be flowed through the second group of half-cells 42 until the second group of half-cells is free of any electrolyte liquid that may have still been held in half-cells 42 from a prior regular operation of the battery 1. Thereafter, the outlet line of the second group of half-cells can be closed by means of said valve 97, while an inflow of the second test fluid 102 may be continued into the second group of halfcells 42 until a predetermined pressure ratio p2:p1 between the second half-cell pressure p2 and the first half-cell pressure p1 is reached. To that end, the pressure in the inlet lines may be measured and the pumps mentioned before may be controlled based on the measured pressures p1, p2, as is indicated by the dashed lines connecting the pressure measuring points with the pumps. To operate the pumps 71, 72, a suitable control law may be implemented, e.g. a Pl-controller or a Sliding-Mode-controller or a Model-Predictive-controller, that maps a deviation between a desired pressure p2 and a measured pressure onto an according operation state of the pumps 71, 72, e.g., onto a desired rotating speed of pump 72. In this fashion, any desired pressure p1, p2 and any desired pressure ratio can be realized with highest precision. Of course, also the outlet valves 96, 97 may be opened and closed depending on measurements of the pressures p1, p2.

[0050] Moreover, measuring the pressures p1, p2 allows for a further, particularly advantageous embodiment of the invention, in which, in addition to detecting a concentration in an outlet, a pressure drop in the second group of half-cells 42 is monitored while the valve 97 in the outlet is closed, representing a further indicator of a leakage between the said groups of halfcells 41, 42, as apparently, without any leakage at all, the pressure inside a closed volume (as e.g. the second group of half-cells 42) would remain constant. In case a semi-permeable membrane is used that does not seal entirely against gas, usually slow and small pressure drops take place due to small amounts of gas passing through the semi-permeable membrane. However, also in this scenario, damage to the membrane or aging of the membrane or changes in membrane properties that cause leakages of electrolytes also cause the passing rate of a gas to go up, which may be detected in the scope of the invention. In some cases where a gas is used as a second test fluid 102, it frequently turns out to be advantageous to reduce or even stop pumping and hence stop or at least reduce the flow of the firsttest fluid 101 on the fluid-side of half-cells. In this fashion, it can be ensured that even only small amounts of crossing gas can be trapped in a gas trap, without being continuously depleted without continuous gas flux. Moreover, measuring a pressure drop on a gas side, i.e. a pressure drop of a gaseous second test fluid 102, frequently turns out to be a particularly sensitive variant to detect a leakage, in certain scenario significantly more sensitive than determining a cross-contamination.

[0051] As can further be seen in the embodiment depicted in Fig. 3b, reservoirs are provided for supplying the first test fluid 101 and the second test liquid 102. In a particularly preferred configuration, the reservoir and / or the external gas cylinder can be designed to retrieve the first test fluid 101 after a leakage test has been conducted, especially in case the test liquids 101, 102 are non-reactive gases or non-reactive liquids.

[0052] As stated previously, the invention provides significant flexibility, allowing to employ the concept of the invention also in more elaborate testing scenarios, e.g., to test a series of stacks. As illustrated in Fig. 4, one approach to that end may encompass six principal steps, using the first variant of the invention described above. These steps may be repeated for several stacks 2, allowing for stack individual testing.

[0053] In a first step I, gas may be pumped into the inlet line 912, thereby displacing the electrolyte into the aforementioned tank. In a second step II, once the second group of half-cells 42 is free of electrolyte, the line may be closed (e.g., by means of a valve 97). In a third step III, the gas pressure may be increased, as described before, e.g., to a pressure 50 mbar, or 100 mbar, or 500 mbar, or 1 bar or more higher than the pressure p1 in the opposing group of half-cells 41. The opposing electrolyte liquid meanwhile may be pumped at a low flow rate (sufficient to cause the agitation of any gas entering the electrolyte, allowing it to be collected in the gas trap). In a fourth step IV, a measurement according to the considerations laid out above may be carried out, inspecting whether any critical leakage is present. After that, in a fifth step V, suitable repairs or replacements may be made. Eventually, in a sixth step VI, electrolyte liquid may again be pumped through the second group of half-cells 42, to displace said test gas and to resume regular operation. The energy storage is then fit to resume normal operation, with reduced capacity fade rate.

[0054] This process allows for a significant extension of a battery’s lifetime. Consequently, stacks in a battery may be tested individually in sequential testing steps. Such a testing procedure may be conducted automatically or manually, and, in a highly advantageous fashion, also during a service operation.

Claims

Claims1. A method for detecting a leakage path between a first group of fluidly connected halfcells (41) and a second group of fluidly connected half-cells (42) of a flow battery (1), preferably of a flow battery using dissimilar electroactive species, most preferably of a redox flow battery, the first and the second group of half-cells (41 , 42) separated from one another at least partially by a membrane of the flow battery (1), the method comprising the steps of:- flowing a first test fluid (101) at a first half-cell pressure (p1) through the first group of half-cells (41);- filling a second test fluid (102) different from the first test fluid (101) at a second halfcell pressure (p2) into the second group of half-cells (42), the second half-cell pressure (p2) exceeding the first half-cell pressure (p1) to allow for a leakage flow of the second test fluid (102) into the first group of half-cells (41);- measuring, in an outlet line (14) of the first group of half-cells (41), a concentration of the second test fluid (102) in the first test fluid (101);- deducing the presence of a leakage path from the first group of fluidly connected halfcells (41) to the second group of fluidly connected half-cells (42) when the measured concentration of the second test fluid (102) in the first test fluid (101) exceeds a predetermined threshold value.

2. The method according to claim 1, wherein the second half-cell pressure (p2) exceeds the first half-cell pressure (p1) by at least 1mbar, or by at least 10mbar, or by at least 50mbar, or by at least 100mbar, or by at least 500mbar, or by at least lOOOmbar.

3. The method according to claim 1 or 2, wherein the second test fluid (102) is flowed through the second group of half-cells (42) while keeping the second group of half-cells (42) filled at said second half-cell pressure (p2).

4. The method according to any one of the previous claims, wherein an electrolyte liquid is used as the first test fluid (101) and wherein a gas is used as the second test fluid (102), preferably a gas unreactive with the first test fluid (101) being used as the second test fluid (102).

5. The method according to claim 4, wherein the gas is selected from a group comprising helium, nitrogen, neon, krypton, argon, xenon, carbon dioxide, pressurized air and radon.

6. The method according to claim 4 or 5, wherein at least a part, preferably all, of the second test fluid (102) that leaked into the first group of half-cells (41) is collected in a gas trap in said outlet line (912) of the first group of half-cells (41), the gas trap preferably comprising a bend in the outlet line.

7. The method according to any one of the previous claims, wherein said concentration of the second test fluid (102) in the first test fluid (101) is measured by means of visual observation, or by means of determination of a conductivity across the gas trap according to claim 6, or by means of optical transmission, preferably using light sources of suitable wavelength and a photodetector to inspect the gas trap according to claim 6, or by means of determining a capacitance or an inductance or an impedance of the first test fluid (101) in said outlet, or by detecting radioactive emission emitted by a the second test fluid (102) when in the first test fluid (101).

8. The method according to any one of claims 1 to 7, wherein said second test fluid (102) is flowed through the second group of half-cells (42) until the second group of half-cells (42) is free of previously held electrolyte, whereafter an outlet line of the second group of half-cells (42) is closed, preferably by closing an outlet valve in said outlet line of the second group of half-cells (42), whereafter said second test fluid (102) is flowed into the second group of halfcells (42) until a predetermined pressure ratio between the second half-cell pressure (p2) and the first half-cell pressure (p1) is reached, which pressure ratio is preferably maintained thereafter, preferably by controlling the second half-cell pressure (p2).

9. The method according to any one of claims 1 to 3, wherein an electrolyte liquid selected from a group comprising iron-chromium electrolyte, vanadium electrolyte, polysulfide-bromine electrolyte, and uranium electrolyte is used as the second test fluid (102), and wherein a liquid different from the second test fluid is used as the first test fluid (101), and wherein preferably a liquid selected from a group comprising deionized water, acid and base is used as the first test fluid (101).

10. The method according to claim 9, wherein said concentration of the second test fluid (102) in the first test fluid (101) is measured by determining an optical absorption ora pH-value or a redox potential of the first test fluid (101) in the outlet of the first group of half-cells (41).

11. The method according to any one of the previous claims, wherein a non-battery-electro-lyte is used as the first test fluid (101) and / or as the second test fluid (102).

12. The method according to any one of the previous claims, wherein the first half-cell pressure (p1) is a pressure of the first test fluid (101) acting in an inlet (911) of the first group of fluidly connected half-cells (41) and / or wherein the second half-cell pressure (p2) is a pressureof the second test fluid (102) acting in an inlet (922) of the first group of fluidly connected halfcells (42).

13. The method according to any one of to any one of the previous claims, wherein the predetermined threshold value is provided in the form of a an absolute concentration value, or in the form of a ratio between a concentration measurable in an uncontaminated first test fluid (101), preferably at an inlet of the first group of fluidly connected half-cells (41), and a concentration measurable in a contaminated first test fluid (101), preferably at an outlet of the first group of fluidly connected half-cells (41), the ratio preferably being selected as a ratio of 0.5 or as a ratio of 0.1 or as a ratio of 0.01 or as a ratio of 0.005 or as a ratio of 0.001.

14. An arrangement for detecting internal leakage paths between a first group of fluidly connected half-cells (41) and a second group of fluidly connected half-cells (42) of a flow battery (1), preferably of a flow battery using dissimilar electroactive species, most preferably of a redox flow battery, the first and the second group of half-cells (41 , 42) separated from one another at least partially by a membrane of the flow battery (1), the arrangement comprising:- a first flowing means, preferably a first pump (71), to flow a first test fluid (101) through the first group of half-cells (41) at a first half-cell pressure (p1);- a second flowing means, preferably a second pump (72), to fill a second test fluid (102) different from the first test fluid (101) into the second group of half-cells (42) at a second half-cell pressure (p2), the second half-cell pressure (p2) exceeding the first halfcell pressure (p1) to allow for a leakage flow of the second test fluid (102) into the first group of half-cells (41);- detection means (95) to measure a concentration of the second test fluid (102) in the first test fluid (101) in an outlet line (14) downstream the first group of half-cells (41) and to indicate the presence of an internal leakage path when the measured concentration of the second test fluid (102) in the first test fluid (101) exceeds a predetermined threshold value.

15. The arrangement according to claim 14, wherein an additional reservoir (91) and / or an external gas cylinder are provided for supplying said test fluids (101, 102), the reservoir and / or the external gas cylinder preferably being designed to retrieve the first test fluid (101 , 102) after a leakage test is completed and / or wherein a separate contamination-reservoir is provided to accept the first test fluid (101) contaminated by the second test fluid (102).

Images