Electrolysis cell system and method for operating an electrolysis cell system

Abstract

Method for operating an electrolysis cell system with at least one electrolysis cell in which a product gas containing hydrogen (H2) is produced from a feed gas containing water (H2O) by means of electrolysis using electrical energy, comprising the steps of: providing an electrolysis cell system with an inlet for the feed gas and an outlet for the product gas; providing a measuring device for recording a measured value (UN,in) representing the partial pressure of oxygen in the feed gas; providing a measuring device for recording a measured value (UN,out) representing the partial pressure of oxygen in the product gas and / or providing a measuring device for recording a measured value (UN,diff) representing the difference between the partial pressure of oxygen in the feed gas and the partial pressure of oxygen in the product gas (16); feeding an electrolysis current into the electrolysis cell system;Determining an actual feed conversion value (FCist); determining a feed conversion control deviation (FCdelta) between the actual feed conversion value (FCist) and a predefined target feed conversion value (FCsoll); generating a control signal (S) depending on the feed conversion control deviation (FCdelta); adjusting one or more process parameters of the electrolysis cell system depending on the control signal (S).

Description

[0001] The invention relates to a method for monitoring a conversion rate of an electrolysis cell system in which a product gas comprising hydrogen (H2) is produced from a feed gas comprising water (H2O) by means of electrolysis using electrical energy.

[0002] The method according to the invention is particularly suitable for use in solid oxide electrolyzer cell (SOEC) systems and in reversible solid oxide cell (rSOC) systems in electrolysis operation.

[0003] A solid oxide electrolyzer (SOEC) comprises at least one electrolysis cell, which uses electrical energy to split water (H₂O) into its components, hydrogen and oxygen. The design and function of a solid oxide electrolyzer are similar to those of a solid oxide fuel cell (SOFC), as they are based on the same technology. A key difference is that electrolysis uses water vapor as the feed gas, while fuel cells use oxygen and fuel gas (e.g., hydrogen) as their feed gases.

[0004] A solid oxide electrolyzer consists of several components, each fulfilling different functions. These components generally include an anode support layer, an anode, an electrolyte layer, a cathode, and a cathode support layer. The anode support layer forms the base of the cell and provides mechanical support. It is often made of a porous material such as nickel oxide (NiO). The anode often consists of a mixture of nickel and electrolyte materials. Water, usually in the form of water vapor, is supplied to the anode and diffuses through it. At the anode, the electrooxidation of the water vapor takes place, generating oxygen ions and electrons. The electrolyte layer consists of a solid electrolyte material, such as yttrium-stabilized zirconium dioxide (YSZ). This layer facilitates the transport of oxygen ions (O₂⁻) from the anode to the cathode.The cathode consists of cathode materials such as lanthanum-manganese-cobaltite (LaMnO₃) or lanthanum-ferrite (LaFeO₃). At the cathode, oxygen ions are extracted from the electrolyte and combined with electrons to produce oxygen. The cathode support layer provides structure and stability for the cell. It is structured similarly to the anode support layer.

[0005] A solid oxide electrolyzer utilizes high-temperature operation (typically 700–1000°C) to activate the electrolyte and facilitate ion flow. The water splitting process proceeds as follows: Water vapor (H₂O(g)) is directed to the anode. There, the water vapor decomposes into oxygen ions (O₂⁻) and protons (H⁺). The oxygen ions diffuse through the electrolyte layer to the cathode. The electrolyte layer allows the transport of oxygen ions (O₂⁻) from the anode to the cathode while blocking the flow of electrons. This is achieved by an electrical potential difference between the anode and cathode. At the cathode, the oxygen ions (O₂⁻) combine with electrons supplied by the external circuitry to produce oxygen (O₂).

[0006] Solid oxide electrolysis cells are efficient due to their high operating temperatures and can produce clean hydrogen. They are used in hydrogen production, energy storage, and other industrial processes.

[0007] To operate water electrolysis in a solid oxide electrolysis cell with maximum efficiency, process parameters must be monitored and adjusted. Of particular importance is the so-called "feed conversion rate" (FC) (also known as "feed conversion" or "current-to-reactant ratio" (CR)) of a cell unit (stack or stack module). Feed conversion refers to the ratio between the electrical current input and the chemical reactants (starting materials) used in the electrolysis cell to produce products such as hydrogen and oxygen. Feed conversion values ​​typically range between 0.1 and 0.9. Feed conversion is a measure of the electrochemical conversion of the gas components contributing to electrolysis (so-called oxygen donors) in the feed gas within the stack relative to the amount available at the inlet.It can therefore also be interpreted as a relative change in the oxygen partial pressure between the inlet and outlet of an electrolysis cell or a stack of fluidically communicating electrolysis cells (stack modules) at the stack inlet.

[0008] It has been shown that exceeding a certain feed conversion value, for example 0.9, can lead to soot formation in individual cells of the electrolysis system. Furthermore, exceeding a feed conversion value of 0.9 poses a risk of feed gas depletion in the electrolysis cell or stack. Soot formation and feed gas depletion can each lead to total system failure, or failure of individual cells or a cell stack, during subsequent operation. Therefore, feed conversion serves as a control variable, among other things, for regulating electrical current and throughput.

[0009] To prevent soot formation or depletion of the electrolysis cells in the cell system, feed conversion is monitored during the electrolysis process. Monitoring feed conversion using conventional methods requires a significant amount of sensory equipment. Both the quantity and composition of the feedstock supplied to the electrolysis cells must be precisely known. Furthermore, the temperature and electrolysis current for the electrolysis cell or stack must be measured accurately. This necessitates a complex measurement system and the evaluation of numerous, inherently error-prone measurements, which can lead to measurement inaccuracies of + / - 10% in determining the feed conversion. Regular calibration of the measuring instruments is also required to maintain these inaccuracies over time.

[0010] The invention is based on the objective of providing a method and a system of the type mentioned above that allows improved monitoring of feed conversion.

[0011] The problem is solved by a method according to claim 1. Advantageous embodiments are specified in the dependent claims.

[0012] A key concept of the present invention is the monitoring of the feed conversion of the electrolysis cell system in order to influence process parameters of the electrolysis process on this basis.

[0013] Feed conversion can be interpreted as the relative change in oxygen partial pressure between stack inlet and outlet.

[0014] Lambda sensors are commonly used to determine the partial pressure of oxygen in a gas mixture at high temperatures. A lambda sensor can be considered a miniaturized high-temperature solid oxide fuel cell operating at idle. A lambda sensor measures the Nernst voltage (OCV = open-circuit voltage) that develops between the fuel gas (measurement gas) and a reference gas (pure oxygen or air).

[0015] The challenge now lies in the selection, arrangement, and interconnection of multiple lambda sensors, as well as the correlation of the measured values, in order to directly convert measured Nernst voltages to the feed conversion of a stack. It can be shown that (in steady-state operation) it is insufficient to determine the Nernst voltage only at a single point before or after the stack; the change in Nernst voltages across the stack must also be known. This can be achieved using a lambda sensor ("differential lambda sensor") positioned between the stack inlet and outlet, exposed to feed gas on one side and product gas on the other. Alternatively, the Nernst voltage difference can also be measured using two lambda sensors before and after the stack, measuring against a common reference gas. FC = 1 − exp 2 F U N , in − U N , out RT 1 + K T ⋅ p p ⊖ ⋅ exp − 2 FU N , out RT F - Faraday constant, T - temperature, R - Boltzmann constant, K(T) - lookup table

[0016] This analytical equation, derived from certain simplified model assumptions, is already capable of predicting feed conversion over a wide range of different gas mixtures, temperatures, and pressures with an accuracy of +10.9 / -5.2% at FC = 60% to +3.1 / -2.5% at FC = 90%. The increasing accuracy for higher (and therefore potentially more critical) FC values ​​is noteworthy. The model's uncertainty, which is also reflected in the systematic deviation from the target value, can be further reduced by using an empirical approach based on multidimensional quadratic polynomial regression instead of the analytical equation. This approach demonstrates that pressure measurement can be neglected, as it does not significantly reduce accuracy. This eliminates the need for corresponding instrumentation.Between FC = 60% and 90%, the relationship can be approximated as follows: . FC = − 35.551 = − 35.551 + 2.9882 ⋅ U N . in V + 49.915 ⋅ U N . in V + 46.927 ⋅ U N . out V − 46.927 ⋅ U N . diff V + 0.020335 ⋅ T ° C + 0.020335 ⋅ T ° C − 7.1475 ⋅ U N . in V 2 − 16.913 ⋅ U N . in V 2 + 10.598 ⋅ U N . in V ⋅ U N . out V + 30.128 ⋅ u N . in V ⋅ U N . diff V − 0.0024074 ⋅ U N . in V ⋅ T ° C − 0.014378 ⋅ U N . in V ⋅ T ° C − 20.363 ⋅ U N . out V 2 − 20.363 ⋅ U N . diff V 2 − 0.011971 ⋅ U N . out V ⋅ T ° C + 0.011971 ⋅ U N . diff V ⋅ T ° C − 2.9674 ⋅ 10 − 6 ⋅ T ° C 2 − 2.9674 ⋅ 10 − 6 ⋅ T ° C 2

[0017] For the correct correlation of the values, it has proven crucial that the temperature at each lambda sensor must also be determined. The resulting accuracy of the fit ranges from ±3.5 percentage points at FC=60% to ±1.5 percentage points at FC=90%. The stated values ​​initially represent only the accuracy of the model. Additional uncertainties due to the limited accuracy of the input variables to be measured (Nernst voltages, temperature) must still be taken into account. Both sensor configurations (one lambda sensor before and one after the stack, or one lambda sensor before the stack and a differential lambda sensor between the stack input and output) are similarly sensitive to measurement deviations. The total uncertainty using standard measurement techniques is 6.0 or 7.8 percentage points at FC=83.3%, which is significantly lower than the values ​​of the conventional FC measurement method.More precise instruments make it possible to reduce this overall uncertainty to less than 3 percentage points (2.7 or 2.4 percentage points).

[0018] The operating method according to the invention offers the following advantages: Precise determination of the feed conversion rate compared to conventional methods at the same or lower costs. Less instrumentation is required. Online measurement and monitoring of feed conversion is possible. No special plant operation is necessary for feed conversion determination. Measuring instruments such as lambda probes can be externally connected to the feed gas and product gas lines of the electrolysis cell system. This allows for quick instrument replacement.

[0019] In state-of-the-art methods, feed conversion is calculated, for example, using the following equation: FC = jA eff m cells 2 F n ˙ H 2 O , in + 2 F n ˙ CO 2 , in

[0020] To calculate the feed conversion according to this equation, five parameters must be known: two geometric parameters (cell count & cross-sectional area), and three measured values ​​for steam, CO2 and electricity.

[0021] A disadvantage of this approach is that feed conversion is determined as a balance. This balance typically comprises a module with several hydraulically parallel stack towers. However, each stack tower has different characteristics regarding hydraulic and electrical flow. As a result, the feed conversion (FC) varies between the stack towers. In the worst case, the FC becomes too high in one tower, leading to soot formation (Co-SOEC) or membrane degradation (SOEC). However, locally high and low feed conversion rates can balance each other out in the feed conversion balance, so this critical condition goes undetected. To avoid critical conditions, the feed conversion rate must therefore be reduced across the board to prevent any tower from exceeding critical values, which reduces the efficiency of the SOEC. To refine the balance, high-temperature flow measurements would be necessary.Only a few measurement principles are available for this purpose and they are difficult to integrate into a SOEC, as straight inlet sections are required, but are not provided by design.

[0022] In contrast, the feed conversion balancing area can be limited to a single stack tower using the method according to the invention, since measuring devices for detecting oxygen partial pressures, for example using lambda probes at high temperatures (such as 600°C to 1000°C), are available. By limiting the balancing area to individual stacks, the variations between the stack towers can be detected. Furthermore, the feed conversion rate can be increased using the method according to the invention. The system can be operated in such a way that no stack tower experiences an excessively high feed conversion rate.

[0023] If the cost analysis were carried out analogously to the current state of the art, there would be a cost advantage. Two probes and two temperature measurements are required. With available measurement technology, costs of approximately €4,500 are to be expected, with the potential to reduce them to below €1,000. It should be noted that at these costs, tower-specific measurements are no longer possible. With one measurement for each tower, the costs increase proportionally depending on the number of towers, but are still more favorable than the loss of efficiency. REFERENCE MARK LIST

[0024] 10 Electrolysis cell 11 Oxygen electrode 12 Hydrogen electrode 13 Air / purge medium 14 Oxygen-enriched air / purge medium 15 Feed gas 16 Product gas 17 Catalytic reactor / reformer 18 Hydrocarbons (C m H n ) 19 Feed gas 20 Power source for electrolysis current 21 Measuring device 22 Measuring device 23 Measuring device QR: Heat dissipated during external reforming; QE: Heat consumed during internal reforming

Claims

1. Method for operating an electrolysis cell system with at least one electrolysis cell (10) in which a product gas (16) containing hydrogen (H2) is produced from a feed gas (15) containing water (H2O) by means of electrolysis using electrical energy, comprising the steps of: providing an electrolysis cell system with a feed line for the supply of feed gas (15) and a discharge for the discharge of product gas (16); providing a measuring device (22) for recording a measured value representing the partial pressure of oxygen in the feed gas (15) (U). N,in ); Providing a measuring device (21) for recording a measured value representing the oxygen partial pressure in the product gas (16) (U N,out ) and / or providing a measuring device (23) for recording a measured value representing the difference between the oxygen partial pressure in the feed gas (15) and the oxygen partial pressure in the product gas (16) (U N,diff), Feeding an electrolysis current into the electrolysis cell system; determining a feed conversion actual value (FC) ist ); Determining a feed conversion rule difference (FC) delta ) between the feed conversion actual value (FC ist ) and a predefinable target feed conversion value (FC) soll ); Generating a control signal (S) depending on the feed conversion control difference (FC) delta ); Setting one or more process parameters of the electrolysis cell system depending on the control signal (S).

2. Method according to claim 1, characterized by the fact that A solid oxide electrolysis cell system (SOEC), in particular a solid oxide cell (rSOC) that can be operated in reverse mode, is used as the electrolysis cell system.

3. Method according to one of claims 1 or 2, characterized by the fact that the electrolysis cell system comprises at least one stack containing several electrolysis cells (10).

4. Method according to claim 3, characterized by the fact that the electrolysis cell system comprises several stacks, several of which form a stack group which are fed with the feed gas (15) from a common feed line and wherein the stacks of the stack group each feed the product gas (16) produced in the stack into a common outlet.

5. Method according to claim 4, characterized by the fact that the measuring device (21) is arranged and designed such that the measured value representing the oxygen partial pressure in the product gas (16) (U) N,out ) is detected at the output of each stack of the stack group, namely upstream before the product gas (16) of the respective stack is fed into the common outlet.

6. Method according to claim 4, characterized by the fact that the measuring device (23) is arranged and designed such that the measured value (U) representing the difference between the oxygen partial pressure in the feed gas (15) and the oxygen partial pressure in the product gas (16) N,diff) each between the feed gas inlet of the stack and the outlet of the stack, namely upstream before the product gas (16) of the respective stack is fed into the common outlet.

7. Method according to any of the preceding claims, characterized by the fact that The electrolysis current is a process parameter that can be set depending on the control signal (S).

8. Method according to any of the preceding claims, characterized by the fact that the measuring device (21) comprises a lambda probe, wherein the measured value (U) acquired by means of the lambda probe N,out ) the Nernst tension, which represents the oxygen partial pressure in the product gas (16).

9. Method according to claim 8, characterized by the fact that The recorded Nernst voltage is analyzed using a Fourier analysis or a Fourier filter at a predefinable modulation frequency.

10. Electrolysis cell system that is set up and configured to carry out a method according to any of the preceding claims.

Images