ELECTROLYSIS SYSTEM WITH CONTROLLED THERMAL PROFILE

Abstract

This invention relates to a system comprising one or more electrolysis cell(s) and at least one power electronic unit that supplies the cell(s) with a fluctuating voltage, and to a method for operating one or more electrolysis cell(s), comprising providing one or more voltage fluctuations to the electrolysis cell(s) by at least one power electronic unit, enabling the provision of a low-cost electrolysis system which simultaneously allows for fast-response dynamic operation, improved electrolysis efficiency, increased lifetime and high impurity tolerance.

Claims

1. A system for operating one or more electrolysis cell(s), comprising: one or more electrolysis cell(s); and at least one power electronic unit, wherein the power electronic unit(s) provide(s) one or more voltage fluctuations to the electrolysis cell(s), wherein the voltage fluctuation(s) are configured such that near-thermoneutral operation at part load is enabled by matching the integral Joule heat production with the integral reaction heat consumption inside said cell(s).

2. The system according to claim 1, wherein the one or more electrolysis cell(s) are configured to operate above 120° C.

3. The system according to claim 1, wherein the one or more electrolysis cell(s) are selected from solid oxide electrolysis/fuel cells (SOEC/SOFC), molten carbonate electrolysis/fuel cells (MCEC/MCFC), high temperature and pressure alkaline electrolysis/fuel cells, and ceramic electrolyte proton conducting electrolysis/fuel cells (PCEC/PCFC).

4. The system according to claim 1, further comprising at least one PID system controlling the voltage fluctuation(s) based on measurements of the inlet and outlet temperature of fluids sent to and from a stack.

5. The system according to claim 1, wherein the voltage fluctuation(s) are configured to effect desorption or dissolution of side reaction compounds adsorbed, precipitated or otherwise formed in the electrodes of the cell(s).

6. The system according to claim 1, wherein a duration of each voltage fluctuation(s) is in the range of from 1 μs to 1000 s.

7. The system according to claim 1, wherein the power electronic unit comprises a DC power supply with a pulse width modulation (PWM) motor controller, a bi-directional power supply, or a power supply in combination with an e-load.

8. The system according to claim 1, wherein the range of the voltage fluctuation(s) is between 0.2 V and 2.0 V.

9. The system according to claim 8, wherein the range of the voltage fluctuation(s) is between 0.5 V and 1.9 V.

10. The system according to claim 1, wherein the one or more electrolysis cell(s) perform electrolysis of H.sub.2O and/or CO.sub.2.

11. A method for operating one or more electrolysis cell(s), comprising: providing one or more voltage fluctuations to the electrolysis cell(s) by at least one power electronic unit, wherein the voltage fluctuation(s) are configured such that near-thermoneutral operation at part load is enabled by matching the integral Joule heat production with the integral reaction heat consumption inside said cell(s).

12. The method according to claim 11, wherein for a fraction of the time of the voltage fluctuation, the current in the cell(s) is reversed such that the cell(s) operate in fuel cell mode.

13. The method according to claim 11, wherein the one or more electrolysis cell(s) perform electrolysis of at least CO.sub.2.

14. The method according to claim 11 wherein the one or more electrolysis cell(s) are selected from solid oxide electrolysis/fuel cells (SOEC/SOFC), molten carbonate electrolysis/fuel cells (MCEC/MCFC), high temperature and pressure alkaline electrolysis/fuel cells, and ceramic electrolyte proton conducting electrolysis/fuel cells (PCEC/PCFC).

15. The method of claim 11 comprising controlling the voltage fluctuation(s) based on measurements of the inlet and outlet temperature of fluids sent to and from a stack.

16. The method of claim 11 wherein the voltage fluctuation(s) are configured to effect desorption or dissolution of side reaction compounds adsorbed, precipitated or otherwise formed in the electrodes of the cell(s).

17. The method of claim 11 wherein a duration of each voltage fluctuation(s) is in the range of from 1 μs to 1000 s.

18. The method of claim 11 wherein the range of the voltage fluctuation(s) is between 0.2 V and 2.0 V.

19. The method of claim 18 wherein the range of the voltage fluctuation(s) is between 0.5 V and 1.9 V.

20. The method of claim 11 wherein the one or more electrolysis cell(s) are operated above 120° C.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0025] FIG. 1 shows the simulated OCV, temperature and current density, for conventional CO.sub.2 electrolysis operating profile with 5x oxygen overblow (i.e. with a cell footprint of 100 cm.sup.2, 5% CO at gas inlet and 28% CO at gas outlet, CO.sub.2 as buffer gas, a SRU voltage of 1.228 V, and a total current of 50 A).

[0026] FIG. 2A schematically illustrates an electrolysis cell.

[0027] FIG. 2B illustrates an electrolysis system comprising two cells stacked in series.

[0028] FIG. 3A and 3B show examples of a steam electrolysis using voltages with sine wave-shaped fluctuations. Corresponding average voltages are shown as horizontal lines.

[0029] FIG. 4 shows an example of a steam electrolysis using fluctuating voltage operation at E.sub.tn 80% of the time and at OCV 20% of the time (thick line). The average voltage is shown as the thin line. E.sub.tn and OCV are shown with dotted lines.

[0030] FIG. 5 shows an example of a CO.sub.2 electrolysis using square wave-shaped fluctuating voltage (thick line) with average cell voltages of 1.18 V and 1.32 V (thin line), respectively. E.sub.tn and OCV are shown with dotted lines.

[0031] FIG. 6 is a diagram showing OCV, temperature and current density for the square wave-shaped fluctuating voltage with average cell voltages of 1.114 V. The temperature profile is shown for 5x oxygen overblow (a cell footprint of 100 cm.sup.2, 5% CO at gas inlet and 28% CO at gas outlet, CO.sub.2 as buffer gas, a cell voltage of 1.3 V for 74% of the time and 0.7 V for 26% of the time, and the total current being 53 A).

[0032] FIG. 7 is a diagram showing the OCV, temperature and current density under the conditions of FIG. 6, except for the omission of oxygen overblow.

[0033] FIG. 8 is a diagram showing the OCV, temperature and current density calculated for Example 3.

[0034] FIG. 9 shows the cell area specific resistance, current and voltage as function of time during AC/DC H.sub.2O electrolysis.

[0035] FIG. 10 shows the comparison of the area specific resistances for DC and AC/DC test in H.sub.2O electrolysis.

[0036] FIG. 11 depicts the OCV, temperature and current density profiles calculated for Example 4.

[0037] FIG. 12 shows the comparison of the area specific resistances for DC and AC/DC test in CO.sub.2 electrolysis and gas cleaning effects.

DETAILED DESCRIPTION OF THE INVENTION

[0038] For a more complete understanding of the present invention, reference is now made to the following description of the illustrative embodiments thereof:

[0039] In a first embodiment, the present invention relates to a system for operating one or more electrolysis cell(s), comprising: one or more electrolysis cell(s); and at least one power electronic unit, wherein the power electronic unit(s) provide(s) one or more voltage fluctuations to the electrolysis cell(s), wherein the voltage fluctuation(s) are configured such that near-thermoneutral operation at part load is enabled by matching the integral Joule heat production with the integral reaction heat consumption inside said cell(s).

[0040] In practice, actual electrolysis systems typically operate at conditions that are neither fully isothermal nor fully adiabatic. The term “near-thermoneutral operation”, as used herein, denotes electrolysis operation where the absolute value of the difference between the integrated Joule heat production and the integrated reaction heat consumption (both integrated over a period of more than 3600 seconds) is less than the absolute value of the integrated heat consumption or the absolute value of the integrated heat production, or both. In preferred embodiments, “near-thermoneutral operation” is understood as electrothermal balanced operation, which uses electric (Joule) heat to balance the required reaction heat and can be distinguished from conventional thermal balanced operation where the thermal capacity of excess air flow is used limit temperature variations in electrolysis cells and stacks.

[0041] A schematic representation of an exemplary electrolysis cell constructed for electrolyzing different types of reactant materials into desired reaction products is shown in FIG. 2a. The cell 10 comprises a first electrode 11 and a second electrode 13, as well as an electrolyte 12 provided between the first electrode 11 and the second electrode 13. During electrolysis operation, electrical power is supplied to the cell 10 (i.e. across the first electrode 11 and the second electrode 13) through a power electronic unit 16 and a reactant 14 passed over the first electrode 11 is separated into ions 15 of a second reaction product 15a and a mixture 15b of a first reaction product and unreacted reactant. The ions 15 pass though the electrolyte 12 and the second reaction product 15a is formed at the second electrode 13. According to the present invention, the power electronic unit 16 provides one or more voltage fluctuations to the electrolysis cell(s). In a preferred embodiment, the system according to the present invention further comprises at least one proportional-integral-derivative (PID) system 17 controlling the voltage fluctuation(s) based on measurements of the inlet and outlet temperature of the fluid (gas or liquid) sent to and from the electrolysis cell(s) or the stack or by measurements of the temperature directly in the cell compartments (not shown), wherein each of the temperature measurements may be conducted by a temperature detection means 18. The PID controller is configured to continuously calculate an error value as the difference between a desired temperature setpoint and the measured temperatures, enabling the connected power electronic unit to apply a correction of the voltage fluctuations based on proportional, integral, and derivative terms.

[0042] The reactant materials are not particularly limited. In preferred embodiments, the one or more electrolysis cell(s) perform electrolysis of H.sub.2O, CO.sub.2, or co-electrolysis of H.sub.2O and CO.sub.2.

[0043] If the cell 10 is a solid oxide electrolysis cell (SOEC) designed to electrolyze water, the first electrode 11 represents the hydrogen electrode, the second electrode 13 may be referred to as the oxygen electrode, and the reactant 14 will be high temperature steam. A catalyst in the hydrogen electrode facilitates separation of steam 14 into oxygen ions 15 and a mixture of hydrogen gas and unreacted steam 15b. As the oxygen ions 15 pass though the electrolyte 10, oxygen gas 15a will be formed at a catalyst in the oxygen electrode 13. However, it is understood that FIG. 2a serves to schematically illustrate an exemplary electrolysis system, and the present invention is not limited to such a configuration (or a SOEC) and may include further layers (e.g. membranes or diaphragms) and components as long as the electrolysis cell enables electrolytic operation. In embodiments, the electrolytic cell may be constructed from fuel cells which operate in electrolytic mode (reversible fuel cells). Furthermore, the electrolysis cell is not limited to a flat stack configuration but may also incorporate other designs (including cylindrical configurations, for example). FIG. 2b illustrates an electrolysis stack 20 that is constructed from two electrolysis cells 10. Depending on electrolysis production requirements, additional cells could also be added to the stack 20.

[0044] Unless otherwise noted, the term “electrolysis cell(s)”, as used herein, will be understood to encompass both a single cell, such as the cell 10 shown in FIG. 2a, and a stack of two or more cells, such as the stack 20 shown in FIG. 2b.

[0045] It is noted that details pertaining to materials and construction techniques for electrolysis cells are well-known to the skilled artisan and will not be described herein.

[0046] In general, a distinction may be made between low-temperature and high-temperature electrolysis. In high-temperature electrolysis, the one or more electrolysis cell(s) are configured to operate above 120° C., such as 200° C. to 1100° C., or 650° C. to 1000° C., for example. While not being limited thereto, the one or more electrolysis cell(s) for high-temperature electrolysis are preferably selected from solid oxide electrolysis/fuel cells (SOEC/SOFC), molten carbonate electrolysis/fuel cells (MCEC/MCFC), high temperature and pressure alkaline electrolysis/fuel cells, and ceramic electrolyte proton conducting electrolysis/fuel cells (PCEC/PCFC).

[0047] From the above description, it will be understood that the term “electrolysis cell(s)”, as used herein, also encompasses fuel cells, i.e. cells which operate in fuel cell mode exclusively. Accordingly, the systems and methods of the present invention may also be used to extend the lifetime of fuel cells and fuel cell stacks.

[0048] In electrolytic mode, the electrolysis reaction is generally endothermic, i.e. the reaction heat is negative. The Joule heat due to the necessary overpotential and current is positive in both fuel cell and electrolysis mode. In high-temperature electrolysis, near-thermoneutral operation at an operating voltage below E.sub.tn is desirable for optimal performance. For operating voltages between OCV and E.sub.tn, near-thermoneutral operation requires heat addition during the electrolysis process. In such a system, heat addition is further necessary to reduce tensile stress at the interconnect/cell interface, which potentially leads to delamination and loss of contact, poor performance and degradation. Conventionally, heat is supplied through the use of a heated sweep gas or active heating devices, for example. In contrast, in the present invention, the Joule heat is balanced with the reaction heat (plus heat loss to surroundings) by supplying the electrolysis cell(s) with one or more voltage fluctuations via one or more power electronic unit(s). Thus, the electrolysis system can be operated near-thermoneutrally with no need for external heating sources.

[0049] By adjusting the voltage variation the Joule heat can be set to balance or slightly exceed the reaction heat. This enables near-thermoneutral operation, or slightly exothermal operation with an average (integral) SRU voltage between OCV and E.sub.tn while having the temperature of the outlet gas to be the same or slightly higher than the temperature of the inlet gas. Therefore the operation with the voltage variation enables improved control of the thermal profile in the SRU. Optimizing the average (integral) SRU voltage and the thermal profile can be used to increase the carbon activity in a carbonaceous gas inside the electrolysis stack without risking detrimental carbon formation in the electrodes. For CO.sub.2 electrolysis this translates into higher outlet gas CO concentration which again translates into decreased expenses for gas separation.

[0050] The term “voltage fluctuation”, as used herein, denotes a predetermined variation of the cell voltage, which may be applied in the form of a periodical voltage variation which recurs in predefined intervals. From the viewpoint of reducing mechanical tension, the duration of each voltage fluctuation(s) is preferably set to a range of from 1 μs to 100 s. By operating the electrolysis system accordingly, the duration of each fluctuation is so short that the temperature change in the fluid (e.g. gas) and the cells and stack is negligible. In this way, accumulation of mechanical tension at the weak interfaces in the stack can be avoided, making it possible to achieve the increased lifetimes enabled by reversible operation. In further preferred embodiments, the frequency of the voltage fluctuation(s) is in the range of from 10 mHz to 100 kHz.

[0051] In preferred embodiments, the range of the voltage fluctuation(s) is between 0.2 V and 2.0 V, especially preferably between 0.5 V and 1.9 V.

[0052] In a preferred embodiment of the present invention, the power electronic unit comprises a DC power supply with a pulse width modulation (PWM) motor controller, a bi-directional power supply, or a power supply combined with an electronic load (e-load).

[0053] In a preferred embodiment, the voltage fluctuation(s) is/are configured to effect volatilization, desorption or dissolution of side reaction compounds adsorbed, precipitated or otherwise formed in the electrodes of the cell(s), e.g. by increasing the oxidation state (oxidation) or decreasing the oxidation state (reduction) of said side reaction compounds, which leads to degradation decrease, more stable cell voltage and extended lifetime of the cells. While not being limited thereto as long as their formation is reversible, such side reaction compounds may be undesired intermediates or originate from impurities in the reactant (e.g., hydroxides formed by alkaline earth metals, hydrocarbons, sulphur-based compounds, formaldehyde, formic acid ammonia, halogenated compounds) or from electrolysis cell materials (e.g., Si-based impurities from glass components). Desorption or dissolution of side reaction compounds may be achieved by periodical changes to the cell voltage so that the electrochemical cell switches between electrolysis and fuel cell mode operation, for example.

[0054] Ni-migration is known to be one of the main degradation mechanisms in Ni/YSZ-electrodes used for conventional DC electrolysis, observed both in H.sub.2O and CO.sub.2 electrolysis. Impurities such as sulphur strongly bind to the Ni-surface, which is known to accelerate Ni-particle coarsening. Hence, desorption of sulphur (and other) impurities may impede Ni-migration.

[0055] Without being bound to theory, it is assumed that the reduction in the degradation rate may to a certain extent related to desorption of impurities adsorbed at the electrochemically active sites during cathodic polarization of Ni/YSZ electrodes. As an example thereof, the formation of SiO.sub.2 at the active sites during cathodic polarization of the Ni/YSZ electrode may be mentioned. This is expected to occur via the reaction Si(OH).sub.4(g)->H.sub.2O(g)+SiO.sub.2(l) as described earlier by A. Hauch et al., J Electrochem Soc. 2007; 154(7):A619-A26. During a short anodic polarization, H.sub.2O may be formed and SiO.sub.2 may be desorbed. The shape of the voltage fluctuation can be in principle of any type. However, voltage fluctuations comprising sine-wave shaped and/or square-wave shaped voltage fluctuation profiles are preferable. A mix between the sine-shaped and the square-shaped voltage fluctuations is especially preferred to minimize the peak voltage and to minimize erroneous operation conditions related to induction phenomena. Calculations of near-thermoneutral operation conditions using sine-wave shaped and/or square-wave shaped voltage fluctuation profiles will be described by means of examples below.

[0056] In a second embodiment, the present invention relates to a method for operating one or more electrolysis cell(s), comprising: providing one or more voltage fluctuations to the electrolysis cell(s) by at least one power electronic unit, wherein the voltage fluctuation(s) are configured such that near-thermoneutral operation at part load is enabled by matching the integral Joule heat production with the integral reaction heat consumption inside said cell(s).

[0057] In a preferred embodiment, the current in the cell(s) is reversed for a fraction of the time of the voltage fluctuation, so that the cell(s) operate in fuel cell mode. Advantageously, this process may reduce damages to the electrode microstructure and/or effect desorption or dissolution of side reaction compounds adsorbed, precipitated or otherwise formed in the electrodes of the cell(s). Furthermore, during the fraction of time where the current is reversed, not all the products from the integral electrochemical reaction which still reside inside the cell(s) are converted back to reactants. Therefore there is no need to change the fluid (e.g. gas) composition, as opposed to conventional (DC voltage) operation.

[0058] In solid oxide fuel cells, Cr poisoning is known to limit the SOFC air-electrode lifetime, so that alumina coated steel pipes are typically required for air supply. Cr posioning is expected to proceed by reaction of gaseous CrO.sub.2(OH).sub.2 with solid strontium oxide (SrO) present at the reaction sites in strontium-rich air electrodes, resulting in the formation of SrCrO.sub.4 and H.sub.2O. By applying voltage fluctuations according to the present invention during fuel cell operation, desorption of SrCrO.sub.4 can be effected.

[0059] It will be understood that the preferred features of the first embodiment may be freely combined with the second embodiment in any combination, except for combinations where at least some of the features are mutually exclusive.

EXAMPLES

Example 1: Sine-Wave Shaped Voltage Fluctuations

[0060] In general, the voltage fluctuation provided to the electrolytic cell(s) can be regarded as a DC voltage superimposed by a smaller AC voltage (AC/DC voltage).

[0061] A DC voltage U.sub.1 superimposed with an AC sine voltage can be written as

[00001]$\begin{array}{cc}U\left(t\right)=U_0\sin \left(\omega t\right)+U_1& \left(\mathrm{Eq}.1\right)\end{array}$

[0062] The reaction heat can be written as

[00002]$\begin{array}{cc}\stackrel{.}{Q_r}=\frac{U-E}{R}K& \left(\mathrm{Eq}.2\right)\end{array}$

[0063] Herein, U represents the cell voltage, E is the Nernst voltage, R the cell resistance, and K is the Coulomb specific reaction heat

[00003]$\left(K=\frac{T\Delta S}{nF}\right).$

For electrolysis, the reaction heat is negative.

[0064] The Joule heat can be written as:

[00004]$\begin{array}{cc}{\stackrel{.}{Q}}_J={\left(\frac{U-E}{R}\right)}^2& \left(\mathrm{Eq}.3\right)\end{array}$

[0065] For thermo-neutral operation, it is required that:

[00005]$\begin{array}{cc}\Downarrow \underset{0}{\overset{T}{\int }}{\stackrel{.Math.}{Q}}_r+{\stackrel{.Math.}{Q}}_J\mathrm{dt}=0\Updownarrow \underset{0}{\overset{T}{\int }}\left(\frac{U_0\sin \left(\omega t\right)+U_1-E}{R}\right).Math.K+\frac{{\left(U_0\sin \left(\omega t\right)-U_1-E\right)}^2}{R}\mathrm{dt}=0\Updownarrow \underset{0}{\overset{T}{\int }}\left(U_0\sin \left(\omega t\right)+U_1-E\right).Math.K+U_0^2{\sin }^2\left(\omega t\right)+{\left(U_1-E\right)}^2+2\left(U_1-E\right)U_0\sin \left(\omega t\right)\mathrm{dt}=0\Updownarrow {\left[\begin{array}{c}-\frac{U_{0}K}{\omega }\cos \left(\omega t\right)+\hspace{1em}\left(U_1-E\right)\mathrm{Kt}+\\ \begin{array}{c}{U}_0^2\left(\frac{t}{2}-\frac{1}{4}\sin \left(2\omega t\right)\right)+\\ {\left(U_1-E\right)}^{2}t-2\frac{\left(U_1-E\right)}{\omega }\cos \left(\omega t\right)\end{array}\end{array}\right]}_0^T=0\Updownarrow \left(U_1-E\right)\mathrm{KT}+\frac{U_0^2}{2}T+{\left(U_1-E\right)}^{2}T=0\Updownarrow \left(U_1-E\right)=\frac{-K\pm \sqrt{K^2-2U_0^2}}{2}& \left(\mathrm{Eq}.4\right)\end{array}$

[0066] FIGS. 3A and 3B illustrate exemplary sine-shaped fluctuations operated at 50 Hz. The graphs in FIG. 3A refer to the + sign, and the in FIG. 3B graphs refer to the − sign in the last line of Equation 4. Corresponding average voltages are shown as horizontal lines.

Example 2: Square-Wave Shaped Voltage Fluctuations

[0067] If square-shaped voltage variations are used, the following expression applies for near-thermoneutral operation:

[00006]$\begin{array}{cc}\Downarrow T_1\left({\stackrel{.Math.}{Q}}_{r,1}+{\stackrel{.Math.}{Q}}_{J,1}\right)+T_2\left({\stackrel{.Math.}{Q}}_{r,2}+{\stackrel{.Math.}{Q}}_{J,2}\right)=0\Updownarrow T_1\left(\frac{\left(U_1-E\right)}{R}K+\frac{{\left(U_1-E\right)}^2}{R}\right)+T_2\left(\frac{\left(U_2-E\right)}{R}K+\frac{{\left(U_2-E\right)}^2}{R}\right)=0\Updownarrow T_{1}K\Delta U_1+T_1\Delta U_1^2+T_{2}K\Delta U_2+T_2\Delta U_2^2=0\Updownarrow t_{1}K\Delta U_1+t_1\Delta U_1^2+\left(1-t_1\right)K\Delta U_2+\left(1-t_1\right)\Delta U_2^2=0\Updownarrow \Updownarrow \Delta U_2=\frac{\begin{array}{c}-K\left(1-t_1\right)\pm \\ \sqrt{{\left(K\left(1-t_1\right)\right)}^2-4\left(1-t_1\right)\left(\Delta U_1^2+K\Delta U_1\right)t_1}\end{array}}{2\left(1-t_1\right)}& \left(\mathrm{Eq}.5\right)\end{array}$

[0068] Herein, ΔU.sub.1=(U.sub.1−E) and ΔU.sub.2=(U.sub.2−E). In Equation 5, it is assumed that the voltages only switch between U.sub.1 and U.sub.2 such that the total period T=T.sub.1+T.sub.2.

[0069] The average voltage U.sub.av is given as:

[00007]$\begin{array}{cc}{U}_{\mathrm{av}}=T_1.Math.U_1+T_2.Math.U_{2}E=1V\mathrm{and}KK=-0.5\mathrm{for}{\mathrm{CO}}_2{U}_1\mathrm{and}{U}_0\mathrm{can}\mathrm{be}\mathrm{obtained}\mathrm{from}& \left(1\right)\end{array}$

Equation 4 for the sine function. Using the same assumptions for E and K , a relation from Equation 5 between U.sub.1, U.sub.2 and T.sub.1 for the square shaped voltage variation is obtained.

[0070] The voltage variation can be any shape. The sine-wave shape and square-wave shape are provided for mathematical simplicity. The important aspect is the integrated Joule heat balances the integrated reaction heat. Smooth curves are preferred over square-wave shaped curves to minimize stray-inductance and eddy currents in the SRU.

[0071] FIG. 4 shows an example with a square-shaped voltage fluctuation for steam (H.sub.2O) electrolysis. Herein, the voltage fluctuates between OCV and E.sub.tn. At these voltages, the cell is operated near-thermoneutrally.

[0072] FIG. 5 shows an example with a square-shaped voltage fluctuation during thermo-neutral CO.sub.2 electrolysis. Herein, the square-wave shaped cell voltage fluctuation causes a switch between electrolysis mode and fuel cell mode when applying a cell voltage below OCV, which promotes desorption or dissolution of side reaction compounds formed in the cell electrodes. Moreover, the heat generated during the operation in fuel cell mode (i.e. the 20% of the time where the cell voltage is below OCV) balances the heat consumed during the operation in electrolysis mode (i.e. the 80% of the time where the cell voltage is above OCV). The average cell voltages are 1.18 V and 1.32 V (thin line), respectively. The two examples shown in FIGS. 4 and 5 are both operated at 1 kHz.

[0073] In a further experiment shown in FIG. 6, OCV/temperature/current density profiles were calculated in line with FIG. 1 (i.e. using a cell footprint of 100 cm.sup.2, 5% CO at gas inlet and 27.8% CO at gas outlet, CO.sub.2 as buffer gas, 5x oxygen overblow) with the exception that voltage fluctuations according to FIG. 4 were applied (1.3 V for 74% of the fluctuation period and 0.7 V for 26% of the fluctuation period; Integral current=53 A). FIG. 6 demonstrates that, in contrast to conventional electrolysis operation (cf. FIG. 1), a flat temperature profile is achieved inside the stack and current density changes between gas inlet and gas outlet are kept at a minimum.

[0074] FIG. 7 shows the OCV, temperature and current density under the conditions of FIG. 6, except for the omission of oxygen overblow. In comparison to FIG. 6, a larger temperature variation is observed. However, the current density can be maintained on a similarly stable level.

[0075] It is shown that the present invention enables near-thermoneutral operation of electrolysis cells and stacks by controlling the size and shape of voltage fluctuations. By reducing thermo-mechanical stress in the stack, the lifetime of electrolysis cells and stacks may be further improved.

[0076] Accordingly, it is possible to provide a low-cost electrolysis system which simultaneously enables fast-response dynamic operation, improved electrolysis efficiency, increased lifetime and high impurity tolerance and improved conditions for CO production.

Example 3: H.SUB.2./H.SUB.2.O Electrolysis Tests

[0077] An electrolysis system for H.sub.2/H.sub.2O electrolysis was set up as follows: An asymmetric square-shaped wave function was set at a function generator (Wavetek Model 145). The thus produced signal was amplified by a bipolar power supply (Kepco BOP 20-20D) and transferred to a test setup in accordance to C. Graves et al., Nature Materials 2015, 14, 239-244. An oscilloscope (Philips PM 3384) was used to monitor the cell voltage as well as the signal from the function generator. In AC/DC mode, the electrolysis cell was operated at 30 Hz fluctuating between 1.27 V (˜90% of the time) and 0.8 V (˜10% of the time), i.e. 90% duty, giving an average voltage of 1.22 V. For the sake of comparison, the electrolysis cell was also operated in DC mode, with the operating conditions for the H.sub.2O electrolysis tests being shown in Table 1 below. For both electrolysis tests, the gas flow to the negative electrode was 24 l/h of H.sub.2: H.sub.2O (with a ratio of 0.5:0.5), and the inlet temperature was 700° C.

TABLE-US-00001 TABLE 1 Average Cell Average Fuel side Inlet Air flow Current Cell Electrolysis flow rate H.sub.2O:H.sub.2 rate Density Voltage mode [l .Math. h.sup.−1] ratio [l .Math. h.sup.−1] [A .Math. cm.sup.−2] [V] DC H.sub.2O 24 0.5:0.5 140 ~−0.73 1.27 AC/DC H.sub.2O 24 0.5:0.5 140 ~−0.55 1.22

[0078] For the AC/DC test, with the current density being approximately −0.55 A/cm.sup.2, the H.sub.2O utilization was 30% and the air overblow factor was 16. The Nernst voltage vs. air was calculated to be 941 mV and 965 mV at the inlet and outlet, respectively. The calculated outlet temperature and gas composition was 701° C., and 65% H.sub.2+35% H.sub.2O. The profiles for temperature, Nernst voltage and current density from inlet to outlet are presented in FIG. 8 (for a co-flow configuration). FIG. 9 shows the cell area specific resistance, current and voltage as function of time during AC/DC H.sub.2O electrolysis.

[0079] The evolution of the cell area specific resistance for the AC/DC and DC H.sub.2O electrolysis tests are compared in FIG. 10, which shows that the application of voltage fluctuations according to the present invention effectively prevents the increase of cell resistance during cell operation. It is noted that the measured area specific resistance for the AC/DC test is initially higher than for the DC test. Impedance spectra recorded before and after the AC/DC test indicated that this effect was based on a higher positive electrode resistance when compared to the positive electrode resistance the resistance of the reference cell.

Example 4: CO/CO.SUB.2 .Electrolysis Tests

[0080] Three cells were tested in CO:CO.sub.2 electrolysis mode, using the test setup described in Example 3.

[0081] All tested cells were multi-layer tapecasted cells (MTC) cells, having a CGO (cerium gadolinium oxide) barrier layer, an LSC (lanthanum strontium cobaltite)-CGO oxygen electrode and an LSC contact layer.

[0082] The cells were tested under the following conditions:

[0083] 1) DC test without gas cleaning: total flow of 18 L/h with inlet CO:CO.sub.2 ratio of 0.13:0.87; average cell current density of 0.31 A/cm.sup.2, inlet temperature of ˜695° C., 13.3% fuel utilization (outlet CO:CO.sub.2 ratio 0.24:0.76), a total fuel flow rate of 18 L/h, and an air flow rate of 50 L/h to the oxygen electrode.

[0084] 2) DC test with gas cleaning: inlet flow of 10.5 L/h with re-circulation of the outlet gas, incl. 20% CO, a total fuel flow rate of 21 L/h and an inlet CO:CO.sub.2 ratio of 0.1:0.9, average cell current density of 0.31 A/cm.sup.2, inlet temperature of ˜695° C., 11% fuel utilization (outlet CO:CO.sub.2 ratio 0.20:0.80), and an air flow rate of 50 L/h to the oxygen electrode.

[0085] 3) AC/DC test: total flow of 16 L/h with CO:CO.sub.2 ratio of 0.11:0.89, average cell current density of ˜0.3 A/cm.sup.2, inlet temperature of ˜695° C., ˜14.7% fuel utilization (outlet CO:CO.sub.2 ratio ˜0.24:0.76), a total fuel flow rate of 16 L/h and an air flow rate of 140 L/h to the oxygen electrode. The gas cleaner was switched off after 250 h testing. The cell was operated at 30 Hz fluctuating between 1.30 V (˜60% time) and 0.75 V (˜40% time), i.e. a 60% duty giving an average voltage of 1.09 V.

[0086] The calculated temperature, current density and Nernst voltage profile for the AC/DC test from fuel and air inlet to outlet is shown in FIG. 11. A counter-flow test setup was used for the actual cell test whereas a co-flow configuration was used for the presented calculation, however, the temperature profile is relatively flat in both cases. The Nernst voltage (OCV) at the inlet and outlet was 877 mV and 913 mV, respectively. A slight temperature increase from 695° C. to 696° C. is measured from inlet to outlet.

[0087] A comparison of the cell area specific resistance profiles for the two DC tests and the AC/DC test, respectively, is presented in FIG. 12. The degradation rates shown in FIG. 12, which have been calculated as the time-dependent change in area specific resistance (in mΩ.Math.cm.sup.2.Math.kh.sup.−1), demonstrate that in the case of operating with cleaned gases, the AC/DC operation reduces the degradation rate by a factor of approximately 3.7, compared to conventional DC operation. The results indicate that increase of area specific cell resistance during cell operation may also be effectively supressed in CO:CO.sub.2 electrolysis cells, which enables an increased cell lifetime, higher efficiency and stable cell operation.

[0088] Once given the above disclosure, many other features, modifications, and improvements will become apparent to the skilled artisan.

REFERENCE NUMERALS

[0089] 10: electrolysis cell [0090] 11: first electrode [0091] 12: electrolyte [0092] 13: second electrode [0093] 14: reactant [0094] 15: ions [0095] 15a: second reaction product [0096] 15b: mixture of first reaction product and unreacted reactant [0097] 16: power electronic unit [0098] 17: optional PID system [0099] 18: temperature detection means [0100] 10: electrolysis cell stack