METHOD AND SYSTEMS FOR DETERMINING A MEASUREMENT ERROR IN A HYDROGEN CONCENTRATION MEASUREMENT

Description

[0001] The invention relates to a computer-implemented method for generating a model for determining a measurement error in a concentration measurement of hydrogen contained in a gas mixture, wherein the gas mixture consists of hydrogen and a non-hydrogen carrier gas mixture.

[0002] Furthermore, the invention relates to a computer program product or a computer program comprising instructions which, when the computer program is executed by a computer, cause the computer to execute the aforementioned method.

[0003] Furthermore, the invention relates to a computer-readable data carrier on which the aforementioned computer program product or computer program is stored, and a data carrier signal that transmits the aforementioned computer program product or computer program.

[0004] Furthermore, the invention relates to a method for determining a measurement error in a concentration measurement of hydrogen contained in a gas mixture, wherein the gas mixture consists of hydrogen and a non-hydrogen carrier gas mixture.

[0005] Furthermore, the invention relates to a system for measuring hydrogen concentration comprising a gas analysis device and a computing device associated with the gas analysis device.

[0006] The use of hydrogen (especially from renewable sources) is becoming increasingly attractive for the energy sector. Its application as an additive to fuel gas, such as natural gas, plays a particularly important role. This hydrogen-fuel gas mixture can then be combusted in a turbine to generate electricity and heat. Controlling these turbines requires rapid hydrogen concentration measurement to maintain a stable operating point.

[0007] One possible approach to measuring hydrogen concentration is partial pressure measurement. Due to hydrogen's high thermal conductivity, partial pressure measurement is often performed using thermal conductivity sensors. For example, gas analyzers and field devices operating on the thermal conductivity principle are known for this purpose. Such devices can operate continuously and are primarily used for the quantitative determination of H₂ or He in binary or quasi-binary gas mixtures.

[0008] Other application examples include chlor-alkali electrolysis (0...10% H2 in Cl2); metallurgy (steel production and processing); H2 measurement in the LNG (Liquefied Natural Gas) process; ammonia synthesis; and artificial fertilizer production.

[0009] In the simplest case, the measurement takes place with a gas analyzer operating on the thermal conductivity principle in a binary mixture, e.g. H₂ (hydrogen) in N₂ (nitrogen) or H₂ in CH₄ (methane), etc. Any further admixture to the carrier gas impairs the accuracy of the measurement.

[0010] US patent application US 2011-158854 A1 discloses a method for improving the accuracy of hydrogen concentration determination, based on the understanding that ambient temperature and oxygen concentration in a sample gas negatively affect the accuracy of the hydrogen concentration measurement. However, due to the presence of oxygen, this method is unsuitable for use in a fuel gas application.

[0011] For use in a fuel gas application (e.g., natural gas-hydrogen mixture), it is therefore necessary to adapt the measurement to the existing gas matrix. Since the carrier gas is not isolated but only present in the presence of H₂, this is hardly possible. The analyzer (the gas analysis instrument) must therefore be prepared for all natural gas sources without requiring complex (and sometimes impossible) on-site adaptations.

[0012] The object of the present invention can therefore be seen as providing methods and systems that enable the aforementioned preparation of the analyzer.

[0013] The problem is solved according to the invention using a computer-implemented method mentioned above, in that data on at least two different non-hydrogen-containing carrier gas mixture compositions are provided. The non-hydrogen-containing carrier gas mixture is a natural gas mixture.

[0014] The data can be in the form of a file on a storage medium, such as a portable non-volatile storage device. The data can also be provided on a server in a network, such as on the internet, in a cloud, or similar.

[0015] Based on data concerning at least two different non-hydrogen carrier gas mixture compositions, different gas mixture compositions are created (e.g., automatically, computer-aided) or generated (e.g., in the form of tables on a computer). The non-hydrogen carrier gas mixture composition and / or the hydrogen content or hydrogen concentration differ between these different gas mixture compositions.

[0016] In one embodiment, it can be provided that, when generating the different gas mixture compositions, hydrogen is added to each non-hydrogen-containing carrier gas mixture composition, from, for example, approximately 0.0 vol.% to, for example, approximately 100.0 vol.%. In this way, a multitude of different gas mixture compositions can be generated based on a single carrier gas mixture composition by changing the ratio between the carrier gas mixture and the hydrogen in the gas mixture. This can be achieved, for example, by successively increasing the hydrogen concentration (and correspondingly reducing the concentration of the carrier gas mixture) in the gas mixture, e.g., in 1% increments. It is understood that smaller increments, e.g., of 0.1%, or larger increments, e.g., of 2%, can also be chosen.

[0017] Subsequently, (expected) hydrogen concentration measurements are determined for the generated gas mixture compositions. Based on these hydrogen concentration measurements, measurement errors are identified. This can be done, for example, by comparing the determined hydrogen concentration measurements with the (actual) hydrogen concentrations known from the generation of the gas mixture compositions. This establishes a correlation between the carrier gas mixture compositions and hydrogen concentration measurements on the one hand, and the (expected) measurement errors on the other.

[0018] In a further step, a model is provided (based on the established relationship) that maps data on carrier gas mixture compositions and hydrogen concentration measurements to measurement errors. In other words, the model contains relationships between the carrier gas mixture compositions and hydrogen concentration measurements and the (expected) measurement errors that enable the aforementioned mapping.

[0019] The model can be based on linear regression, gradient-boosted trees, or neural networks. Linear regression is preferably used to establish the relationships between the carrier-gas mixture compositions and hydrogen concentration measurements on the one hand, and the measurement errors on the other, thus generating the model.

[0020] The model is therefore generated or trained using data on gas mixture compositions, which is intended for use with a gas analysis device operating on the thermal conductivity measurement principle, and outputs a corresponding measurement error, with which the measured value is subject, for the measured value and a provided carrier gas mixture composition.

[0021] It should be noted here that the different gas mixture compositions can be generated not only virtually, as described above. It may be possible, for example for process control, to generate the gas mixture compositions in parallel in a laboratory, for instance using a mixing device capable of mixing a carrier gas mixture with hydrogen. In this case, the hydrogen concentration values ​​can be measured with a gas analyzer operating on the thermal conductivity principle, and the measurement errors detected during the measurement can be compared with the measurement errors determined using the model.

[0022] In one embodiment, it may therefore be provided that the hydrogen concentration measurements are based on a measurement.

[0023] Alternatively or additionally, in one embodiment it may be provided that the hydrogen concentration measurements are calculated.

[0024] It may be useful if the data includes a characteristic curve of a gas analysis device operating on a thermal conductivity measurement principle, whereby thermal conductivities of the gas mixture compositions are calculated, and the hydrogen concentration measurements are calculated from the (calculated) thermal conductivities based on the characteristic curve.

[0025] A computing device suitable for carrying out the aforementioned procedure is also disclosed. For example, the computing device may have a first interface with which it can receive data on at least two different non-hydrogen-containing carrier gas mixture compositions. The computing device may furthermore include a computing unit, e.g., a processor, and a memory operationally connected to the computing unit, e.g., volatile or non-volatile memory. The memory is, for example, designed to store or buffer the data, with the computing unit being configured to generate the different gas mixture compositions based on the data and to determine hydrogen concentration measurements for the different gas mixture compositions, e.g., using a characteristic curve optionally present in the data.Furthermore, the computing unit is configured to determine measurement errors based on the hydrogen concentration measurements, e.g., by comparing them with the actual hydrogen content specified during generation. The computing unit may also have a second interface through which it provides the model. Alternatively or additionally, the computing unit may be configured to provide the model via the second interface and store a copy of the model in its memory.

[0026] In one embodiment, the data may include information on the concentration of two, three, four, five or more components contained in the carrier gas mixture, preferably for each carrier gas mixture composition.

[0027] In one embodiment, it may be provided that each component is selected from the group: methane, carbon dioxide, nitrogen, ethane, propane.

[0028] In one embodiment, the carrier gas mixture may be a synthetic gas mixture, e.g. a two-component gas mixture (e.g. CO 2 -CH 4 mixture).

[0029] In one embodiment, it may be provided that a hydrogen concentration measurement value is determined for each gas mixture composition, and a measurement error is determined based on the hydrogen concentration measurement value and the hydrogen content in the gas mixture composition.

[0030] In one embodiment, data may be provided for four, twelve, 200, or more different non-hydrogen carrier gas mixture compositions. Such data may, for example, be based on or represent the NIST AGA8 table.

[0031] In one embodiment, it may be provided that a proportion of the data, preferably 10% to 20%, in particular 15% of the data, is used to validate the model.

[0032] Furthermore, the problem is solved according to the invention by the aforementioned method for determining a measurement error in that data on the composition of the non-hydrogen carrier gas mixture are provided, the hydrogen concentration in the gas mixture is measured with a gas analysis device operating on a thermal conductivity measurement principle in order to obtain a measured value, and a model generated as described above is applied to the measured value and to the data on the composition of the non-hydrogen carrier gas mixture in order to determine a measurement error for the measured value (absolute in vol.%).

[0033] It is understood that the model can either have been created (i.e., before the procedure is carried out) or be created (i.e., during the execution of the procedure).

[0034] Furthermore, the invention is solved according to the invention with a measuring system mentioned above in that the gas analysis device is configured to To measure the hydrogen concentration in a gas mixture consisting of hydrogen and a non-hydrogen carrier gas mixture using a thermal conductivity measurement principle in order to obtain a measured value, wherein the computing device includes a model generated as described above and is configured to obtain data on the composition of the non-hydrogen carrier gas mixture, and to apply the model to the measured value and to the data in order to determine a measurement error for the measured value (absolute in vol.-%).

[0035] In one embodiment, the computing device may be configured to correct the measured value taking into account the measurement error, and preferably to provide the corrected measured value.

[0036] In one embodiment, the computing device may be configured to transfer the measurement error to the gas analysis device so that it is automatically taken into account in future measurements.

[0037] In one embodiment, the gas analysis device may be configured to measure the concentration of at least one component contained in the non-hydrogen carrier gas mixture and to transmit the measured concentration value to the computing device or to make it available to the computing device.

[0038] In one embodiment, the gas analysis device may be configured to measure the concentration of at least two, three, four, five, or more components contained in the non-hydrogen-containing carrier gas mixture and to transmit or make available the measured concentration values ​​to the computing device.

[0039] In one embodiment, the gas analysis device may include a gas analysis device operating on a thermal conductivity measurement principle to measure hydrogen concentration, and a gas chromatograph device to measure the concentration of the at least one component contained in the non-hydrogen carrier gas mixture.

[0040] Further features, properties and advantages of the present invention will become apparent from the following description with reference to the accompanying figures. These schematically illustrate: FIG 1 a computing facility, FIG 2 a flowchart of a computer-implemented process, FIG 3 a flowchart of a procedure for determining a measurement error in a hydrogen concentration measurement, and FIG 4 a process control system.

[0041] In the exemplary embodiments and figures, identical or similarly acting elements may each be provided with the same reference numerals.

[0042] First, the focus will be on FIG 1 and FIG 2 Reference made to. FIG 1 Figure 1 shows a computing device 1 that is suitable for executing a computer-implemented procedure for generating a model 2, which is represented by a flowchart in FIG 2 This illustrates the point.

[0043] The one in FIG 1The computing device 1 shown can, for example, have a first interface 3 with which it can receive data 4 about at least two different non-hydrogen-containing carrier gas mixture compositions. The computing device 1 can also have a computing unit 5, e.g., a processor, and a memory 6, e.g., volatile or non-volatile memory, operationally connected to the computing unit 5. The memory 6 is, for example, designed to store or temporarily store the data 4, wherein the computing unit 5 is configured to generate different gas mixture compositions based on the data 4 and to determine hydrogen concentration measurements for the different gas mixture compositions, e.g., based on a characteristic curve optionally present in the data 4. Furthermore, the computing unit 5 is configured to use the hydrogen concentration measurements, e.g.,Measurement errors are calculated by comparing the measured values ​​with the actual hydrogen content specified during generation. This generates a model 2 which, based on determined relationships between the carrier gas mixture compositions and hydrogen concentration measurements on the one hand, and the (expected) measurement errors on the other, can map data on carrier gas mixture compositions and hydrogen concentration measurements to measurement errors corresponding to the hydrogen concentration measurements.

[0044] Computing unit 1 may also have a second interface 7 through which it provides model 2. Alternatively or additionally, computing unit 1 may be configured to provide model 2 via the second interface 7 and to store a copy of model 2 in its memory 6.

[0045] FIG 2The computer-implemented process is illustrated by means of a flowchart. In step S1, the data 4 are provided via at least two different non-hydrogen-containing carrier gas mixture compositions. The data can be in the form of a file on a storage medium, for example, a portable non-volatile storage medium. The data 4 can be provided, for example, on a server (not shown) in a network, such as on the internet, in a cloud, or similar. For this purpose, the computing unit 1 of the FIG 1 have a suitable connection interface.

[0046] In step S2, different gas mixture compositions are created or generated based on data 2. In principle, two parameters can be varied simultaneously or individually. First, the type of carrier gas mixture can be varied. Second, the hydrogen content in the gas mixture can be varied.

[0047] When generating different gas mixture compositions, hydrogen can be added to any non-hydrogen carrier gas mixture composition in amounts ranging from, for example, approximately 0.0 vol% to, for example, approximately 100.0 vol%. In this way, a variety of different gas mixture compositions can be generated based on a single carrier gas mixture composition by changing the ratio between the carrier gas mixture and the hydrogen content. This can be achieved, for example, by successively increasing the hydrogen concentration (and correspondingly reducing the concentration of the carrier gas mixture) in the gas mixture, e.g., in 1% increments.

[0048] In step S3, (expected) hydrogen concentration measurements are determined for the generated gas mixture compositions. Measurement errors are then calculated based on these measurements. Preferably, a hydrogen concentration measurement is determined for each generated gas mixture composition, and the corresponding measurement error is calculated based on this value. This can be done, for example, by comparing the determined hydrogen concentration measurements with the (actual) hydrogen concentrations known from the generation of the gas mixture compositions. This generates model 2, which can map data 4 on carrier gas mixture compositions and hydrogen concentration measurements to measurement errors.

[0049] In a further step, S4, the Model 2 will be made available.

[0050] FIG 1This reveals that memory 6 contains a computer program 8 with instructions for executing the procedure described above.

[0051] Hydrogen concentration measurements can be based on a measurement. Alternatively or additionally, the hydrogen concentration measurements can be calculated. For this purpose, a characteristic curve of a gas analyzer operating on a thermal conductivity measurement principle can be included in the data. In this case, the thermal conductivities of the gas mixture compositions can be calculated, and the hydrogen concentration measurements can be derived from the (calculated) thermal conductivities using the characteristic curve.

[0052] Preferably, the data includes information on the concentration of two, three, four, five, or more components contained in the carrier gas mixture. This information is preferably available for each carrier gas mixture composition. This increases the robustness of model 2.

[0053] The component can be selected from the following group: methane, carbon dioxide, nitrogen, ethane, propane.

[0054] All carrier gas mixtures are natural gas mixtures.

[0055] The compositions of various natural gases can be represented, for example, in the following TABLE 1: component min. [Vol.- %] typ. [Vol.-%] max. [Vol. - %] Thermal conductivity [W / m*K] Thermal conductivity [norm.] CH 4 60 88 95 0.0341 1 C 2 H 6 1 4 14 0.02029 0.6 C3H8 0.1 1.5 5 0.01682 0.49 CO2 0.1 3 11 0.01652 0.48 N 2 0.3 3.5 18 0.026 0.76

[0056] This includes the typical five components (methane, ethane, propane, carbon dioxide, nitrogen) of typical natural gases and their thermal conductivities, as well as the thermal conductivities normalized to CH4.

[0057] Four specific examples EG1, EG2, EG3, EG4 of possible natural gas compositions are given in TABLE 2: EG1 EG2 EG3 EG4 CH 4 [Vol.-%] 60.565 70.226 88.106 95.419 CO2 [Vol.-%] 3.943 10.58 2.71 0.174 N2 [Vol.-%] 17.582 4.348 3.172 0.25 C 2 H 6 [Vol.-%] 13.898 10.092 0.635 4.02 C 3 H 8 [Vol.-%] 2.953 3.413 4.425 0.088

[0058] The data used to generate Model 2 can also include more than four natural gas compositions, e.g., a large number of them, 200 or more.

[0059] For example, 404 gas mixture compositions can be generated from natural gas compositions EG1 to EG4 by gradually adding hydrogen to each of the natural gas compositions EG1 to EG4 in 1% increments. It is understood that, for example, to generate more data points, smaller increments, such as 0.1%, can also be chosen.

[0060] Subsequently, a hydrogen concentration measurement can be determined for each of the 404 gas mixture compositions. Then, based on the hydrogen concentration measurement and the hydrogen content in the gas mixture composition, a measurement error can be determined.

[0061] FIG 3shows a flowchart of a procedure for determining a measurement error in a hydrogen concentration measurement in a hydrogen-containing fuel gas mixture.

[0062] In step S01, data on the composition of the non-hydrogen carrier gas mixture and the model 2 generated as described above are provided.

[0063] In step S02, a gas analysis device operating on a thermal conductivity measurement principle is used to measure the hydrogen concentration in the gas mixture in order to obtain a measured value.

[0064] In step S03, model 2 is applied to the measured value and to the data on the composition of the non-hydrogen carrier gas mixture in order to determine a measurement error for the measured value (absolute in vol.-%).

[0065] The procedure may include a further step S04 in which the determined measurement error is used to recalibrate the gas analysis device.

[0066] FIG 4 Figure 100 shows a process control system in which a method for determining a measurement error in a hydrogen concentration measurement in a hydrogen-containing fuel gas mixture, for example the method of FIG 3 can be implemented.

[0067] The process control system 100 is designed to supply and operate a gas turbine 101 with hydrogen-containing fuel gas 102.

[0068] The fuel gas 102 is prepared in a mixing device 103. A carrier gas 104, for example natural gas or a synthetic gas, and hydrogen 105 are supplied to the mixing device 103, so that the mixing device 103 can mix hydrogen 105 with the carrier gas 104 to produce the fuel gas mixture 102, which is supplied to the turbine 101.

[0069] A preferred flow direction of the gases is illustrated by arrows 106.

[0070] Furthermore, the process control system 100 includes a gas analysis unit, which in this case comprises a gas analysis device 107, a computing unit 108 associated with the gas analysis device 107 and a gas chromatograph 109.

[0071] The gas analyzer 107 has a measuring inlet 1070 through which a sample gas, for example the fuel gas 102, is introduced into the gas analyzer 107. The gas analyzer 107 also has a measuring outlet 1071 to discharge the gas after the measurement.

[0072] The gas analysis device 107 also includes a measuring chamber 1072 that operates on the thermal conductivity principle. The measuring principle of the measuring chamber 1072 is based on the different thermal conductivities of gases. The heating of a heated measuring resistor (not shown) surrounded by gas is determined by the thermal conductivity of the gas.

[0073] In one operating mode, the measuring chamber 1072 can have a sensor (not shown) equipped with a micromechanically manufactured silicon chip, the measuring diaphragm of which can be provided with thin-film resistors. The resistors contained in the diaphragm are regulated to a constant temperature. For this purpose, a current can be used that assumes a specific value depending on the thermal conductivity of the measuring gas. This "raw value" can be further processed electronically and used to calculate the gas concentration. The sensor is characterized in particular by a short T90 time. Preferably, the sensor is located in a thermostatically controlled stainless steel housing to suppress the influence of the ambient temperature. Furthermore, it can be advantageous to only indirectly supply the sensor with fluid to avoid flow influences.This can be achieved, for example, by placing the sensor in a bore to the side of the flow channel.

[0074] In one operating mode, the measuring chamber 1072 may be provided with four measuring resistors connected to form a Wheatstone bridge (not shown). Two of these resistors are surrounded by the sample gas, while the other two are surrounded by the reference gas. A constant DC voltage heats the resistors above the temperature of the measuring block 1072. If the sample gas and reference gas have different thermal conductivities, the resistors heat up unevenly due to the heat dissipation. A change in the composition of the sample gas therefore also results in a change in the resistance values. The electrical equilibrium of the measuring bridge is disturbed, and a voltage is generated across the bridge diagonal. This voltage is a measure of the concentration of the measured component, in this case, hydrogen (H₂).

[0075] The gas analyzer 107 may have corresponding interfaces for power supply and / or control. For power supply, the gas analyzer 107 may have a power connection point 1073. In addition to the power connection point 1073, the gas analyzer 107 may include a bus system, preferably a fieldbus, in particular a CAN bus 1074 (Controller Area Network), and an analog / digital output 1075.

[0076] The computing unit 108 associated with the gas analysis device 107 can additionally be designed as a supply and control unit for the gas analysis device 107 and be coupled to the aforementioned connections of the gas analysis device 107 (see FIG. 4 ).

[0077] The gas analysis device 107 is designed to measure the hydrogen concentration in the fuel gas 102 in order to obtain a measurement value.

[0078] The computing device 108 comprises a model 2 generated as described above and is configured to obtain data 40 on the composition of the carrier gas 104 and to apply the model 2 to the measured value and to the data 40 in order to determine a measurement error (an absolute error in vol.-%) for the measured value.

[0079] The data 40 concerning the composition of the carrier gas 104 can be based on measurements performed on the carrier gas 104 by the gas chromatograph 109. The gas chromatograph 109 is configured to measure the concentration of at least one component contained in the carrier gas 104. For example, the gas chromatograph 109 is configured to measure the concentrations of methane, ethane, propane, carbon dioxide, and nitrogen in the carrier gas 104.

[0080] The gas chromatograph 109 can transmit the measured values ​​to a control unit 110 of the process control system 100. The control unit 110 is designed to support the control of the processes taking place in the process control system 100.

[0081] The control unit 110 can transmit the data 40 to the computing unit 108 or make it available to the computing unit 108 in some other way.

[0082] Furthermore, the computing unit 108 can be configured to correct the measured values ​​by the gas analyzer 107, taking into account the relevant measurement errors, and to provide the corrected measured values ​​to the control unit 110, so that the control unit 110 can, for example, display the corrected measured values.

[0083] Furthermore, the computing unit 108 can be configured to control the gas analyzer 107 according to the calculated measurement error, so that the gas analyzer 107 automatically takes the measurement error into account in the next measurement and displays a corrected measurement value or transmits it to the computing unit 108. For this purpose, the computing unit 108 can transmit the calculated measurement error to the gas analyzer 107, for example, via the fieldbus 1074.

[0084] The gas chromatograph instrument 109 can perform the measurement, for example, at regular time intervals, for example with a sampling rate of 180 seconds.

[0085] The gas analyzer 107 allows for much more frequent measurements than the gas chromatograph 109. A sampling rate of 5 seconds is preferred.

[0086] Although the invention has been illustrated and described in detail by means of exemplary embodiments, the invention is not limited by the disclosed examples. Variations thereof can be derived by a person skilled in the art without departing from the scope of protection of the invention as defined by the following claims. In particular, the features disclosed in connection with the systems described herein can be usefully employed to further develop the methods described herein, and vica versa.

Claims

1. Computer implemented method for generating a model (2) for determining a measurement error in a measurement of concentration of hydrogen (105) that is contained in a gas mixture (102), wherein the gas mixture (102) comprises hydrogen (105) and a non-hydrogenous carrier gas mixture (104), wherein - data (4, 40) is provided regarding at least two different non-hydrogenous carrier gas mixture compositions, - different gas mixture compositions are generated based on the data (4, 40), wherein in the case of different gas mixture compositions the non-hydrogenous carrier gas mixture composition and / or the hydrogen content are / is different, - for the gas mixture compositions, hydrogen concentration measured values are determined, and measurement errors are determined on the basis of the hydrogen concentration measured values, - a model (2) is provided that maps data regarding carrier gas mixture compositions and hydrogen concentration measured values to measurement errors, characterised in that the non-hydrogenous carrier gas mixture is a natural gas mixture (104).

2. Method according to claim 1, wherein the data (4, 40) comprises information regarding concentration of two, three, four, five or more components that are contained in the carrier gas mixture (104).

3. Method according to claim 2, wherein each component is selected from the group: methane, carbon dioxide, nitrogen, ethane, propane.

4. Method according to one of claims 1 to 3, wherein the gas mixture (102) comprises hydrogen (105) and two or more non-hydrogenous natural gas mixtures (104).

5. Method according to one of claims 1 to 4, wherein, when generating the different gas mixture compositions, hydrogen is added to each non-hydrogenous carrier gas mixture composition from 0.0% by volume to 100.0% by volume.

6. Method according to one of claims 1 to 5, wherein a hydrogen concentration measured value is determined for each gas mixture composition, and a measurement error is determined on the basis of the hydrogen concentration measured value and the hydrogen content in the gas mixture composition.

7. Method according to one of claims 1 to 6, wherein the data (4, 40) is provided regarding four, twelve, 200 or more different non-hydrogenous carrier gas mixture compositions.

8. Method according to one of claims 1 to 7, wherein the hydrogen concentration measured values are based on a measurement.

9. Method according to one of claims 1 to 8, wherein the hydrogen concentration measured values are calculated.

10. Method according to claim 9, wherein the data (4) comprises a characteristic curve of a gas analysis device (107) that operates according to a thermal conductivity measuring principle, wherein thermal conductivities of the gas mixture compositions are calculated, and the hydrogen concentration measured values are calculated from the thermal conductivities on the basis of the characteristic curve.

11. Computer program product comprising commands that, when the computer program is being executed by a computer, cause this computer to implement a method according to one of claims 1 to 10.

12. Computer readable data carrier on which the computer program product according to claim 11 is stored.

13. Data carrier signal that transmits the computer program product according to claim 11.

14. Method for determining a measurement error in a measurement of concentration of hydrogen (105) that is contained in a gas mixture (102), wherein the gas mixture (102) comprises hydrogen (105) and a non-hydrogenous carrier gas mixture (104), wherein the non-hydrogenous carrier gas mixture is a natural gas mixture (104), wherein - data (40) is provided regarding the composition of the non-hydrogenous carrier gas mixture (104), - hydrogen concentration in the gas mixture is measured using a gas analysis device (107), which operates according to a thermal conductivity measuring principle, in order to obtain a measured value, - a model (2) that is generated according to one of claims 1 to 10 is applied to the measured value and to the data (40) regarding the composition of the non-hydrogenous carrier gas mixture (104) in order to determine a measurement error for the measured value.

15. Measuring system comprising a gas analysis facility (107, 109) and a computing facility (108) that is allocated to the gas analysis facility (107, 109), wherein the gas analysis facility (107, 109) is configured so as to - measure hydrogen concentration in a gas mixture (102) comprising hydrogen (105) and a non-hydrogenous carrier gas mixture (104), wherein the non-hydrogenous carrier gas mixture is a natural gas mixture (104), according to a thermal conductivity measuring principle in order to obtain a measured value, wherein the computing facility (108) comprises a model (2) that is generated according to one of claims 1 to 10 and is configured so as to - obtain data (40) regarding the composition of the non-hydrogenous carrier gas mixture (104), and - apply the model (2) to the measured value and to the data (40) in order to determine a measurement error for the measured value.