Solid oxide electrolysis cell system and method of operating a solid oxide electrolysis cell system

Abstract

A method for operating a solid oxide electrolysis cell (SOEC) system at part load, the SOEC system including multiple branches electrically connected in parallel, each branch including at least one SOEC stack, includes determining a thermally neutral target voltage, where operation is endothermic below the thermally neutral target voltage and exothermic above the thermally neutral target voltage, and performing pulse-width modulated current control by cycling an on phase and an off phase for each branch such that the SOEC system operates at an average operating power equal to a selected percentage of the operating power at the thermally neutral target voltage. During the on phase, all of the SOEC stacks in the branch operate at the thermally neutral target voltage, and during the off phase, all of the SOEC stacks in the branch operate at 0% power. Each branch is configured to operate independently of the other branches.

Description

[Technical field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to and the benefit of U.S. Provisional Patent Application No. 63 / 060,854, filed August 4, 2020, the entire disclosure of which is incorporated herein by reference. [Background technology]

[0002] The present disclosure relates to electrolysis cell systems for producing fuels, primarily hydrogen, from electricity. In particular, the present disclosure relates to solid oxide electrolysis cell systems and methods of operating solid oxide electrolysis cell systems for part load and rapid response capabilities.

[0003] An electrolysis cell is a device capable of converting electrical input energy through an electrochemical reaction into chemical energy stored in a fuel, such as a hydrocarbon fuel. In general, an electrolysis cell comprises an anode, an electrolyte layer, and a cathode. The electrolyte layer serves to transport ions between the anode and the cathode and drives reactions within the anode and cathode that produce a chemical fuel.

[0004] Electrolysis cells are often characterized by a type of electrolyte layer used to transport specific ions. For example, one type of electrolysis cell is the solid oxide electrolysis cell (SOEC), which incorporates a solid ceramic electrolyte to transport negatively charged oxygen ions from the cathode to the anode.

[0005] At the cathode, an oxygen-containing reactant (often CO vapor) dissociates due to an applied electric current to form hydrogen or CO, along with oxygen ions that transport charge through the electrolyte to the anode. H2O+2e → H2+O 2- (1) CO2+2e→CO+O 2- (2)

[0006] The oxygen ions combine at the anode and react to form gaseous oxygen. O2 - →1 / 2O2+2e (1)

[0007] The solid ceramic electrolyte of the SOEC is a solid, non-porous metal oxide. For example, the solid ceramic electrolyte of the SOEC can include Y2O3-stabilized ZrO2 (YSZ). The cathode of the SOEC can include, for example, a metal / YSZ cermet. In some examples, the cathode of the SOEC can include a porous dual-phase nickel and yttria-stabilized zirconia (Ni / YSZ) cermet. The anode of the SOEC can be, for example, Sr-doped LaMnO3.

[0008] SOEC operating temperatures typically range from about 650°C to about 1000°C, at which ionic conduction by oxygen ions occurs. Referring to FIG. 1, a non-limiting example of the relationship between SOEC operating current (horizontal axis) and electrochemical heat generation (vertical axis) is shown, for example, at 650°C, 700°C, and 750°C. It is desirable for the SOEC to operate at a target operating point that operates in a thermally neutral state (i.e., a voltage below which the SOEC stack operation is endothermic and above which the SOEC stack operation is exothermic). The thermally neutral state is temperature dependent. For example, if the SOEC operating temperature is 750°C, the target operating point is approximately 1.285V / cell. Most electrolytic stack technologies operate only in their respective exothermic regions based on their reaction kinetics. High performance SOEC stacks offer the potential to operate either exothermic or endothermic depending on the conditions.

[0009] In the example of FIG. 1, point d is the target operating point, although it is understood that the target operating point may vary between systems and applications. The shaded zone identifies an example of a preferred operating range. This preferred operating range reflects the need to provide heat to or remove heat from the stack as it deviates from thermal neutrality. The higher the heat load (exothermic or endothermic), the more complex and expensive the resulting thermal management strategy, and the higher the stress applied to the SOEC stack. Depending on the design, the SOEC stack can be designed to operate at temperatures up to 0.2 W / cm2 However, the preferred operating range is 0.1 W / cm 2 and a more preferred operating range is less than 0.05 W / cm 2 At point d, the stack is at 750° C., −1 A / cm 2 700°C and is operating in a thermally neutral state. When operating at point d, the SOEC system only needs to preheat the inlet gas and can do so using the heat of the process exhaust. If the operating power is reduced, for example, to point b, the SOEC stack operates outside the preferred operating range (shaded zone) and the SOEC stack operation is endothermic. The SOEC stack will either cool towards 700°C (due to poor overall performance at low temperatures) and then reach thermally neutral again, or the SOEC stack will need to connect an auxiliary heater to maintain the stack at the operating temperature. If the SOEC stack is allowed to cool, the SOEC system will need to accommodate a different operating temperature.

[0010] If the operating power is increased again, for example to point a, the increase quickly propels the SOEC stack operation into a severely exothermic state. This tends to bring the stack back up to the target operating temperature, which is a relatively slow process. Considering a typical SOEC stack, a transition upwards of 50° C. requires a time span of minutes. Meanwhile, local overheating and increased exhaust temperature can start to adversely affect the SOEC stack and system operation and its useful life. The curve is not symmetrical. The overheating for a given increase in operating power is much greater than the cooling associated with the same magnitude decrease in operating power. In addition, the thermal conditions affect the cell voltage, and a change from an endothermic to an exothermic operating state can produce a large change in stack voltage that can exceed 30%. Therefore, the associated power electronics must be capable of handling a wider range of voltages, which tends to both increase the cost of the power electronics and decrease their efficiency.

[0011] If the SOEC load is reduced to 50% load and the SOEC stack is cooled to 650° C. (point c), it is difficult to increase the operating power beyond about 60% without waiting for the stack and process to heat up again, which is a relatively slow process: in a typical SOEC stack, a 100° C. increase in temperature can take tens of minutes.

[0012] The system turndown ratio is defined as the ratio of the system peak power to its lowest sustainable operating point. The operating turndown of a solid oxide electrolysis cell system is typically adapted by changing the process conditions rather than the stack conditions, in particular by changing the reactant flows or concentrations and / or by shifting the operating temperature. Conventional solutions for operating in or passing through the endothermic operating range present in high performance SOEC systems include passing through the endothermic region quickly to avoid subcooling, which essentially makes sustained part-load operation impossible. Another conventional solution involves adding supplemental heat to the process. If properly designed, these heaters also help to maintain the temperature during endothermic operation, since some heating is required for the initial heating up to the operating temperature. Yet another conventional solution involves allowing the stacks to cool, stopping the endothermic heat, and reducing their efficiency to a point where a new equilibrium is established at a lower operating temperature.

[0013] A need exists for a method of operating a solid oxide electrolysis cell system at high turndown without significantly driving balance of plant requirements in terms of additional heaters, ability to operate over a wide temperature range, etc. In particular, a need exists for a method of operating a solid oxide electrolysis cell system for part load and rapid response capabilities. Summary of the Invention

[0014] In a particular embodiment, a solid oxide electrolysis cell system includes a plurality of branches electrically connected in parallel, each branch including at least one solid oxide electrolysis cell stack, each solid oxide electrolysis cell stack including a plurality of solid oxide electrolysis cells. A method of operating a solid oxide electrolysis cell system at a part load includes determining a thermally neutral target voltage for a given operating temperature, where operation of the solid oxide electrolysis cell system is endothermic below the thermally neutral target voltage and operation of the solid oxide electrolysis cell system is exothermic above the thermally neutral target voltage, and performing pulse width modulated current control by cycling an on phase and an off phase for each of the branches during an operating cycle of the solid oxide electrolysis cell system, such that the solid oxide electrolysis cell system operates at an average operating power equal to a selected percentage of the operating power at the thermally neutral target voltage of the operating cycle of the solid oxide electrolysis cell system. During the on phase, all of the solid oxide electrolysis cell stacks in a given branch operate at the thermally neutral target voltage. In the off-phase, all of the solid oxide electrolysis cell stacks in a given branch are unloaded to their open circuit voltages and operate at 0% of their rated power. Each branch is configured to operate in either the on-phase or the off-phase, independent of the other branches.

[0015] In some aspects of the method or system, all of the branches are in the off phase during at least one period of the operating cycle.

[0016] In some aspects of the method or system, all of the branches are in the on-phase during at least one period of the actuation cycle.

[0017] In some aspects of the method or system, at least one branch is in an on-phase while at least one branch is in an off-phase during at least one period of the operating cycle. This minimizes the pulsation of the total output power at any one time. The more stacks that are electrically connected in parallel and individually switched, the less pulsation is imposed due to on-phase and off-phase transitions.

[0018] In some aspects of the method or system, switching between the on and off phases occurs during consecutive periods within an operating cycle, as opposed to during a fixed period of time.

[0019] In some aspects of the method or system, in an operating cycle, the first branch is in an on phase for the duration of the first period while the second branch is in an off phase for the duration of the first period, the second branch is switched to the on phase at the start of the second period and remains in the on phase for the duration of the second period, and the first branch remains in the on phase for the duration of the second period, or is switched to the off phase at the start of the second period and remains in the off phase for the duration of the second period.

[0020] In some aspects of the method or system, performing pulse width modulation current control includes selecting a pulse width modulation frequency and duty cycle such that the prevailing thermal conditions are thermal such that a transition period between an on phase and an off phase or between an off phase and an on phase occupies less than 10 percent of the duration of an operating cycle.

[0021] In some aspects of the method or system, the pulse width modulation frequency and duty cycle are selected such that each branch switches from the on phase to the off phase before cell starvation due to reactant depletion occurs.

[0022] In some aspects of the method or system, the pulse width modulation frequency and duty cycle are selected by a programmed controller to select the prevailing thermal condition to be thermally neutral or to target a slightly net endothermic or slightly net exothermic condition, such that the temperature of the solid oxide electrolysis cell stack is constant or ramped in a controlled manner between target operating temperatures that reflect system demand.

[0023] In some aspects of the method or system, the pulse width modulation time constant is τ e (the electrical time constant related to the voltage response to a change in current, typically on the order of a few milliseconds) and τ c (the chemical time constants associated with reactant supply and reactant depletion at the electrochemical locations in the SOEC stack, typically on the order of seconds).

[0024] In some aspects of the method or system, the prevailing thermal conditions avoid severe endothermic and severe heat generating regions of the overall stack operating window. In some aspects of the method or system, the operating current density of each branch is 50 mW / cm at a thermally neutral target voltage during the on-phase. 2 The current density may be within the range.

[0025] In some aspects of the method or system, the pulse width modulation time constant is selected such that heat generation during a complete pulse width modulation cycle is dominated by the on-time and off-time states, and not dominated by states during the transition between on and off or the transition between off and on.

[0026] In some aspects of the method or system, the on-phase is selected to be at or near thermal neutral. A near thermal neutral state may be selected if the system control desires to shift the operating temperature toward a new operating point (e.g., in preparation for a sustained idle period where cooling may be desired, or in preparation for recovery from a sustained idle period where heating may be desired).

[0027] In some aspects of the method or system, the on-phase is selected to be at or near thermal neutral. If the off-time state is not completely thermally neutral due to excessive heat loss or heat generation from the balance of the process, a near thermally neutral state may be selected, in which case the on-phase will be selected to provide the heat absorption or generation necessary to balance the overall thermal state to control the SOEC stack temperature.

[0028] In some aspects of the method or system, the pulse width modulation time constant is selected such that product and reactant concentrations near the electrochemically active region are governed by the average conditions during the entire pulse width modulation cycle, rather than the conditions during the on and off times of the actuation cycle.

[0029] In some aspects of the method or system, the pulse width modulation time constant is selected such that the thermal conditions and resulting temperatures during a complete pulse width modulation cycle are dominated by average conditions rather than by conditions during the on and off times of an actuation cycle.

[0030] In some aspects of the method or system, the pulse width modulation frequency and duty cycle are determined by a pulse width modulation time constant τ pwm is selected to define a thermal neutrality factor (Tf) that simultaneously allows partial power operation of the overall system while maintaining a near constant capacity factor and a substantially constant temperature due to the different magnitudes of the respective process time constants while operating at substantially thermal neutrality (or a small but controlled deviation from thermal neutrality).

[0031] In some aspects of the method or system, the resulting actuation strategy is τ as in a conventional controlled SOEC system. t (typically on the order of minutes) is limited by the thermal time constant associated with the thermal mass of the SOEC stack and the temperature response to heat flow, but by τ e This allows the SOEC stack to operate at thermal conditions (temperatures) that retain the ability to ramp to full power in times on the order of 100 Hz (eg, on the order of milliseconds).

[0032] In some aspects of the method or system, the resulting actuation strategy is τ as in a conventional controlled SOEC system. t (typically on the order of minutes) is limited by the thermal time constant associated with the thermal mass of the SOEC stack and the temperature response to heat flow, but by τ flowThis allows the SOEC stack to operate in thermal conditions that maintain operation at full power while retaining the ability to ramp to full power in times on the order of a few milliseconds (eg, seconds).

[0033] In some embodiments of the method or system, the resulting operating strategy allows the SOEC stack to operate under part-load conditions in a manner that allows for power turndown in increments of 10:1 while maintaining full operating temperature and operating in a thermally neutral manner without the need for substantial external heating.

[0034] In some embodiments of the method or system, the resulting actuation strategy is that two or more electrically parallel branches of the stack are switched according to a pulse width modulation sequence in such a manner that at least one branch is in the on-phase while at least one other branch is in the off-phase. For example, a 50% power demand will have only half the stacks in the on-phase at any particular time, but all stacks will experience a 50% duty cycle over the pulse width modulation repeat period.

[0035] In some embodiments of the method or system, the resulting actuation strategy is that two or more electrically parallel branches are switched according to a pulse width modulation sequence in such a manner that a weaker branch may be less loaded than a more powerful branch for the overall benefit of system life and efficiency.

[0036] In some aspects of the method or system, the total power cycle is a first predetermined percentage of rated power and each of the branches operates at 100% for a predetermined percentage of the time, the first predetermined percentage of rated power equal to the predetermined percentage of time during which each of the branches operates at 100%.

[0037] In some aspects of the method or system, the maximum instantaneous power step is a second predetermined percentage of the rated power, where the second predetermined percentage of the rated power is calculated by dividing 100% by the number of branches of the solid oxide electrolysis cell system.

[0038] In some aspects of the method or system, at least one of the branches operates at 100% of its rated power in a first on-phase and at 0% of its rated power in the next on-phase immediately following the first on-phase.

[0039] In some aspects of the method or system, at least one of the branches is unloaded and operated at 0% of rated power at regular intervals in the operating cycle of the solid oxide electrolysis cell system.

[0040] In some aspects of the method or system, at least one of the branches is unloaded at irregular intervals in an operating cycle of the solid oxide electrolysis cell system.

[0041] In some aspects of the method or system, for two consecutive cycles in an operating cycle of the solid oxide electrolysis cell system, the branches in the first subset are the same.

[0042] In some aspects of the method or system, for two consecutive cycles in an operating cycle of the solid oxide electrolysis cell system, the branches in the first subset are different.

[0043] In some aspects of the method or system, each of the plurality of solid oxide electrolysis cells comprises an anode, a cathode, and a solid ceramic electrolyte.

[0044] In some aspects of the method or system, the solid ceramic electrolyte comprises a non-porous metal oxide.

[0045] Those skilled in the art will appreciate that the above aspects are not mutually exclusive and can be combined.

[0046] These and other advantageous features will become apparent to one who considers the disclosure and drawings. [Brief description of the drawings]

[0047] [Figure 1] For example, FIG. 1 shows the relationship between operating current (horizontal axis) and electrochemical heat production (vertical axis) for each of 650° C., 700° C., and 750° C. [Diagram 2] An example of independent on / off switching of the SOEC stacks is shown in an SOEC system operating at 95% net power. [Diagram 3] An example of independent on / off switching of the SOEC stacks is shown in an SOEC system operating at 53% net power. [Figure 4] An example of independent on / off switching of the SOEC stacks is shown in an SOEC system operating at 12% net power. [Diagram 5] 1 illustrates experimental test results of a sub-scale solid oxide electrolysis stack operating in a furnace according to an exemplary embodiment. [Figure 6] Experimental voltage and temperature response results of a sub-scale solid oxide electrolysis stack stepped gradually through part load conditions from 10% to 70% load according to an exemplary embodiment are compared. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0048] Generally referring to the figures, disclosed herein is a method of operating an SOEC system for partial load and rapid response capabilities. As discussed above, the SOEC system includes at least one solid oxide electrolysis cell. Preferably, multiple solid oxide electrolysis cells can be stacked and interleaved with interconnecting plates that distribute gas to the electrode / electrolyte interface and act as current collectors to form a solid oxide electrolysis cell stack. The SOEC system can include multiple SOEC stacks forming at least some electrically parallel branches, for example, 40 SOEC stacks can be connected to 20 electrically parallel branches. In accordance with the operating method, each SOEC branch can be loaded from open circuit to full load while operating in a thermally neutral or near thermally neutral state, independent of the other SOEC branches. Each SOEC branch includes one or more SOEC stacks. The operating method also provides the ability for the SOEC system to operate at partial load, but under thermal conditions that allow for rapid (approximately instantaneous) transition to full load.

[0049] A typical part-duty cycle for an SOEC stack can be described as follows: Phase 1 - The SOEC stack is operated at zero current, which is thermally neutral, at the nominal operating temperature. The reactant flows through the stack are set according to the nominal power demand. - Phase 2 - The SOEC stack switches to full current in thermally neutral state. The SOEC stack voltage response is nearly instantaneous, with the time spent in partial load (endothermic) state being, for example, a few milliseconds. The heat absorbed during this transition time is negligible compared to the thermal mass of the SOEC stack. Negligible temperature change occurs. Phase 3 - The SOEC stack is now at full current in a thermally neutral state and the flow may be too low to support this current indefinitely. However, there is enough gas adjacent to the active area of ​​the solid oxide electrolysis cell to support some operating time. Depending on the design, this may be on the order of tenths of a second, for example. Phase 4 - The SOEC stack is switched to open circuit before cell starvation due to reactant depletion occurs. - Phase 5 - The stack is now held at open circuit voltage (OCV) for a period proportional to the desired load target and proportional to the time the SOEC stack has been at full load. For example, a target of 40% of full load operating point with an on-time of 1 / 10 seconds would have an off-time of 1.5 / 10 seconds. During this time, gas flow continues uninterrupted, clearing products from the active area of ​​the SOEC / SOEC stack and providing new reactants for the next cycle. The SOEC stack is in a thermally neutral state. The cycle then repeats from Phase 1.

[0050] Referring to Figure 1, a non-limiting example of the relationship between SOEC operating current (horizontal axis) and electrochemical heat production (vertical axis) is shown, for example, at 650°C, 700°C, and 750°C. In the example of Figure 1, point d is the target operating point, although it is understood that the target operating point may vary between systems and applications. At point d, the stack is at 750°C, -1 A / cm 2 At point d, the SOEC system only needs to preheat the inlet gas, which can be done using the heat of the process exhaust. If the operating power is reduced, for example, to point b, the SOEC stack is operating outside the preferred operating range (shaded zone) and the SOEC stack operation is endothermic. For example, the target operating point may be the thermally neutral 50 mW / cm depending on the stack and system design. 2The temperature may be within, above, or below the preferred operating range. A stack operating outside the preferred operating range may be described as heavily exothermic or heavily endothermic. A stack operating within the preferred operating point, but not at the thermally neutral target voltage, may be described as slightly exothermic or slightly endothermic. Upon power unloading from point d, the SOEC stack will either cool towards 700°C (due to poor overall performance at low temperatures) and then reach thermal neutrality again, or the SOEC stack will need to connect an auxiliary heater to maintain the stack at the operating temperature. If the SOEC stack is allowed to cool, the SOEC system will need to accommodate a different operating temperature.

[0051] For example, if the operating power is increased again to point a, the increase quickly propels the SOEC stack operation into a severely exothermic state. This tends to bring the stack back up to the target operating temperature, which is a relatively slow process associated with the thermal time constant of the SOEC stack. Before thermal equilibrium is re-established, local overheating and increased exhaust temperature can begin to adversely affect the SOEC stack and system operation and its useful life. The curve is not symmetrical. The overheating for a given increase in operating power is much greater than the cooling associated with the same magnitude decrease in operating power. In addition, the thermal conditions affect the cell voltage, and a change from an endothermic to an exothermic operating state can produce large changes in the stack that can exceed 30%. The associated power electronics must therefore be capable of handling a wider range of voltages, which tends to both increase the cost of the power electronics and decrease their efficiency.

[0052] If the SOEC load is reduced to 50% load and the SOEC stack is cooled to 650° C. (point c), it is difficult to increase the operating power beyond about 60% without waiting for the stack and process to heat up again, which is due to the thermal time constant τ t This is a relatively slow process, as shown by

[0053] A method of operating an SOEC system at part load (hereinafter "operation method") is described herein. The SOEC system includes multiple solid oxide electrolysis cell stacks connected in parallel. Each SOEC stack includes multiple solid oxide electrolysis cells. The operation method includes determining, for a given operating temperature, a thermally neutral target voltage below which operation of the SOEC system is endothermic below the thermally neutral target voltage and above which operation of the SOEC system is exothermic, and performing pulse width modulated current control by cycling the on and off phases of each SOEC stack such that during an operating cycle of the SOEC system, the SOEC system operates at an average operating power equal to a selected percentage of the operating power at the thermally neutral target voltage. As used herein, an "operation cycle" refers to a period during which the electrolysis cell stacks are operating in accordance with an embodiment of the present invention, as opposed to operating at a steady voltage or being powered off for an extended period of time. An electrolysis cell stack is considered to be operating at "100% power" or "full power" when it operates at a thermally neutral target voltage at a given operating temperature. This may also be referred to as "rated power" at a given operating temperature. During the on-phase, the electrolysis cell stack operates at a thermally neutral target voltage. During the off-phase, the electrolysis cell stack is unloaded and operates at 0% power. The solid oxide electrolysis cell stack may be divided into multiple electrically parallel branches. In any electrical branch in the on-phase, the solid oxide electrolysis cell stack in the branch operates at 100% power. In the off-phase, the solid oxide electrolysis cell stack in the branch is unloaded to an open circuit voltage. Each of the solid oxide electrolysis cell stack electrical branches is configured to operate at 100% power or be unloaded, independent of the other solid oxide electrolysis cell stack electrical branches. The SOEC system may include a controller programmed to perform steps of an operating method.

[0054] The pulse width and load of the SOEC stack according to the actuation method are controlled such that during the on-pulse (mode 1), the SOEC system operates in a near thermally neutral state. During the off-pulse (mode 2), the stack is unloaded, which is essentially a thermally neutral state. By switching between modes 1 and 2, the stack is kept in a near thermally neutral state while operating at partial load, which would otherwise be heavily endothermic. The actuation method takes advantage of different time constants associated with 1) the voltage response (cell voltage vs. current), which has a response time in the order of ms; 2) the electrochemical reaction response (flow vs. current) as a function of capacity factor, which has a response time in the order of sub-seconds; and 3) the thermal response (temperature vs. current), which has a response time in the order of multiple seconds. These time constants are physical characteristics of the materials and design of the SOEC stack and system.

[0055] At least three characteristic time constants can be identified within SOEC stacks and systems. τ e : the electrical time constant related to the voltage response to a change in current, typically in milliseconds; τ c : the chemical time constants associated with reactant supply and reactant depletion at the electrochemical locations in the SOEC stack, typically on the order of seconds, and τ t : Thermal time constant related to the thermal mass of the SOEC stack and its temperature response to heat flow, typically on the order of minutes.

[0056] The two dynamic time constants that are relevant can also be considered as follows: τ pwm : A controlled time constant associated with the pulse width modulated current modulation strategy described above. τ flow : A controlled time constant related to the ability of the overall system to ramp flow rate, for example, in response to a required step change in operating power.

[0057] As long as the operating frequency is slow enough and operation is dominated by on-times and off-times (not transition times), the stack will be nominally thermally neutral. As used herein, the term "prevailing thermal state" refers to the thermal state of the cell for more than 90 percent of the duration of the operating cycle. Thus, if the prevailing thermal state of an electrolysis cell stack is thermally neutral, the stack will operate in a thermally neutral state for more than 90% of the duration of the operating cycle. In particular, as long as the operating frequency is fast enough that the inlet gas does not pass all the way through the SOEC stack and passes through the cell active area without reacting, the effective capacity factor will be set by the nominal flow rate. These impose constraints on both the maximum and minimum operating frequencies. Given the several orders of magnitude difference in the time constants of these two processes, there is a large margin for finding operable frequencies that meet both requirements. Neither process imposes significant thermal transients on the stack or the system, since neither process is significantly faster than thermal heating.

[0058] For example, Figure 5 shows experimental test results of a sub-scale solid oxide electrolysis stack operating in a furnace according to an exemplary embodiment. This is an ideal environment for fast transient response, with furnace heat supporting endothermic stack operation. Figure 5 shows the same operating conditions first operated in pulse width modulation mode (the subject of this disclosure) and then repeated with a typical steady load strategy. The operating conditions in each case start at no load, step to 30% load, dwell for 30 minutes, then step to full load, dwell for 30 minutes, followed by unloading.

[0059] Keeping in mind that lower voltage reflects higher efficiency, it should be observed that, first, at 30% load, the pulse width modulation test is operating at a higher efficiency. This is because the operating temperature is maintained and therefore it can operate at peak efficiency. In contrast, steady state operation at 30% load is operating at about 5% less efficiency because the stack is cooling by operating endothermically. This is only possible if the test is performed in a furnace that acts to maintain the temperature around the endothermic stack. In a real system, it would be more difficult to maintain the operating temperature of the endothermic stack and the resulting cooling and efficiency loss would likely be greater.

[0060] Further, with reference to FIG. 5, secondly, it should be noted that for a step change to full load, there is a small voltage recovery in the pulse width modulation case, whereas in the conventional case, the recovery is much larger. This reflects the lower operating temperature reached during part load operation and the associated inefficiency when stepping to full load. In a real generation system, not only will the part load temperature be more likely to be lower, but the available drive voltage may not be high enough to drive the cold stack to peak current, and as a result, endothermic operation will continue to increase the transition time, potentially significantly.

[0061] 6 compares experimental voltage and temperature response results of a sub-scale solid oxide electrolysis stack stepped gradually through part load conditions from 10% to 70% load in accordance with an exemplary embodiment. The black data points are the average in-stack temperature (squares) and average cell voltage (circles) when operating in pulse width modulation mode, and the white data points are the same results when operating in conventional steady state mode.

[0062] It is noteworthy that in pulse width modulation mode, the stack temperature is constant regardless of the set point, whereas in conventional operation, the stack cools gradually with a minimum temperature occurring at about 40% load. This sub-scale test was performed in an electric furnace with a constant temperature set point; in a real system, the stack temperature would be expected to drop more significantly. It is also noteworthy that the efficiency of the stack at all part power operating points is higher, with a nearly perfect linear response. Higher efficiency is beneficial for obvious reasons, e.g., reducing the energy cost per volume of electrolytic gas. Linearity is beneficial for improving the predictability and controllability of the resulting overall system.

[0063] Furthermore, by avoiding endothermic operation, the stack (and system) is able to respond better to transient conditions compared to conventional operation, and therefore has better load following capabilities. However, the added benefit of easier temperature and electrical control also improves the system when the stack is controlled with a pulse width modulated control strategy. Furthermore, it is known that thermal transients are a degradation mechanism for solid oxide electrolysis stacks, and by operating with pulse width modulated control, the majority of thermal transients that would otherwise be induced can be avoided.

[0064] By allowing operation at part load while maintaining thermal neutrality in the stack, the system does not need to be designed to support severe endothermic operation or large fluctuations in operating temperature. This provides a significant simplification in the design of the system. Additionally, by maintaining the SOEC stack near thermal neutral, the system can respond much more sensitively to changes in power, even under part load conditions.

[0065] With reference to Figures 2-4, the method of operation of a relatively small SOEC system including five branches electrically connected in parallel is described in further detail. Each of the branches includes one SOEC stack and is identified as stack 1, stack 2, stack 3, stack 4, or stack 5. However, it is understood that each branch may include any number of SOEC stacks, for example, each branch may include 20-40 SOEC stacks. Examples of switching in the method of operation are shown at 95% net power (Figure 2), 53% net power (Figure 3), and 12% net power (Figure 4). The net power percentages represent the average operating power divided by the rated power of the system (i.e., operating power at thermal neutral voltage). Operating cycles for the first 100 cycles are shown, with each number on the x-axis identifying a cycle number. Each shade of the bar graph represents a particular SOEC stack operating at full current (mode 1). Thus, at cycle numbers where no shading is shown for a particular SOEC stack, it is understood that that particular SOEC stack is unloaded (Mode 2). The dashed line shows the total instantaneous power, while the dotted line shows the time averaged power. With five SOEC stacks, the maximum instantaneous power step (calculated by dividing 100% of the SOEC system by the number of SOEC stacks) is set to 20% of rated power, which sets the requirements for all electrical systems (e.g., capacitor banks) that may be required. As the number of stacks increases, the amount of smoothing required decreases.

[0066] Referring to FIG. 2, in the first cycle at 95% net power, the first SOEC stack is off (no shading on stack 1) and the other four SOEC stacks are operating at 100% power (shading on stacks 2-5), resulting in a total power that is less than 95% of rated power. In the second cycle, all of the SOEC stacks are operating at 100% power (shading on stacks 1-5), resulting in a total power that is greater than 95% of rated power, and this condition persists through the fourth cycle. In the fifth cycle, the third SOEC stack is off (no shading on stack 3) and the other four SOEC stacks are operating at 100% power (shading on stacks 1, 2, 4, and 5), resulting in a total power that is again less than 95% of rated power. As seen in Figure 2, each SOEC stack operates in thermal neutrality or near thermal neutrality, independent of the other SOEC stacks, and can be loaded from open circuit to full load while maintaining 95% net power. In some embodiments, the total power cycle is a first predetermined percentage of rated power, and each of the solid oxide electrolysis cell stacks operates in an on phase for a predetermined percentage of the time, the first predetermined rated power percentage being equal to the predetermined percentage of time during which each of the solid oxide electrolysis cell stacks operates at 100% power. In Figure 2, each of the five stacks is on 95% of the time (i.e., a predetermined percentage of time) of the cycle, and the total cycle power is similarly 95% of rated power (i.e., the first predetermined percentage of rated power).

[0067] As shown in Figure 2, each individual stack (or electrical branch) has an on-time that, when averaged over an entire operating cycle, is equal to the desired output power level (when expressed as a percentage of rated power). As shown in Figure 2, the timing of each stack (or electrical branch) is offset with respect to each other branch so that the sum of the stacks in their on-cycles is not fluctuated by more than one stack (or electrical branch) at a time. For example, over the course of the 100 step cycles shown, there is no time step where fewer than four stacks are on simultaneously, even though each of the stacks is off exactly five times.

[0068] In some examples, at least one of the SOEC stacks is unloaded at regular, repeating intervals in the operating cycle of the solid oxide electrolysis cell system (e.g., a regular pattern such as on for 5 on phases, off for 1 off phase, on for 5 on phases, off for 1 off phase, etc.). In some examples, at least one of the solid oxide electrolysis cell stacks is unloaded at irregular, non-repeating intervals in the operating cycle of the solid oxide electrolysis cell system (e.g., an irregular pattern such as on for 5 on phases, off for 2 off phases, on for 3 on phases, off for 4 off phases, etc.).

[0069] Referring to FIG. 3, an example of a switching in operation method to maintain 53% net output power is shown. In the first cycle, the fourth and fifth SOEC stacks are 100% on (stacks 4 and 5 are shaded), while the first through third SOEC stacks are off (stacks 1 through 3 are not shaded). The SOEC system is operating at 53% output power, and there is no cycle in which all five SOEC stacks are operating. Instead, in any given cycle, two to three SOEC stacks are 100% on. No more than four stacks and less than two stacks are on in any particular cycle. Each stack (or electrical branch) operates at the same 53% overall duty when evaluated over a sufficiently long period of time (shown and achieved here over 100 cycle steps). The target power level need not be a discrete value related to the number of stacks. As the desired resolution increases, the overall length of the switching cycle may also increase, but any part-load condition of the SOEC system in a suitable cycle is achievable. In 100 steps (as shown in FIG. 3), the output power can be modulated to nearly 1%. In FIG. 3, each SOEC stack is on for 53% of the overall cycle. There are relatively few positions where any particular stack is on for two consecutive cycles, and 53% is only slightly more than 50%, reflecting that all stacks are on exactly 50% of the time.

[0070] Referring to FIG. 4, an example of switching in the operating method to maintain 12% output power is shown. Each SOEC stack is turned on approximately once every 8 cycles. The SOEC stacks each switch between open circuit ("off"), which is essentially thermally neutral, and full power ("on"), which is configured to be thermally neutral, so that the operating point does not result in SOEC stack cooling other than waste heat loss from the process. This waste heat loss can be balanced by selecting an on-phase that is slightly exothermic, assuming that the heat loss is small, which should be typical for an efficient electrolysis system.

[0071] One constraint on turndown is the minimum switching frequency coupled to the period it takes for reactant molecules to flow from the inlet to the outlet of the electrochemically active region. In the example of FIG. 4, if the off period is 8 cycles, the switching frequency must be selected so that the 8 cycles do not result in excess reactant bypass flow through the unpowered SOEC stack. Perhaps this may not be as big a constraint since the system flow rate is adjusted for a lower power condition (12% net power). If the flow rate is 8 times lower, the flow rate period will also be 8 times longer, and therefore no change in switching frequency is required. If the flow rate is not 8 times lower, the switching period will not affect the overall capacity factor since there is excess reactant flow even during the on period.

[0072] 2-4, or any other net power maintained in accordance with the method of operation, it is not necessary for reactant flows to be interrupted or redirected to selected SOEC stacks. The reactant flows are distributed evenly to all SOEC stacks and do not affect the capacity factor of the SOEC stacks or the system as long as the switching frequency is appropriately selected.

[0073] In the illustrated example, two underlying assumptions are made: 1) in any particular cycle, power is distributed to the stack (electrical branch) for which power has not been requested for the longest period of time. This ensures a rotating power demand that minimizes the continuous on and off times of any particular stack. 2) Each stack (or electrical branch) supplies power in a precise proportion to the system level power demand. Neither assumption is necessary for an overall discipline strategy. For example, random or semi-random decisions of which stacks to turn on in a particular cycle could be used. Individual power branches do not need to operate in precise proportion to the overall system demand, but instead could operate in response to both system demand and individual branch capacity (or health), so that weaker branches could be loaded less than more powerful branches, or certain branches could be preferentially heated (or cooled) in preparation for or in response to a slower power transition.

[0074] Referring to Figure 1, the above operating method allows the SOEC stack to always operate at point d or point e from a thermal and power standpoint. Both points d and e are thermally neutral. At 50% power, the SOEC stack spends half the time at point d and half the time at point e. The SOEC stack is maintained at full operating temperature and can be ramped up to full power as fast as reactants become available, without any fluctuation in operating voltage or temperature.

[0075] Furthermore, consider the range of 20% to 30% of rated power. In this region, even allowing the SOEC stack to cool down to 650°C is insufficient to move the SOEC stack out of the severe endothermic state. Process heaters or very low stack temperatures are required, which presents control challenges and severe limitations in dynamic response capabilities. As a result, a further advantage of the actuation method is the extended turndown capability compared to conventional control strategies.

[0076] By avoiding severe endothermic part-load conditions, the balance of the system can be simplified since there is no need to support severe endothermic or severe exothermic conditions. Instead, the SOEC system and SOEC stack operate at a specific design offset toward near thermal neutrality, or toward endothermic or exothermic operation as desired. Similarly, the SOEC system can compensate for the stack's degradation over time while maintaining a narrow operating window at conditions most favorable to the stack's life and performance. In addition, the SOEC system can operate at part-load conditions while holding the SOEC stack in conditions that facilitate near instantaneous loading up to full load, allowing for a much faster transient response than would otherwise be possible.

[0077] Depending on the method of operation, the SOEC stack will be maintained at operating temperature even when operating at partial load conditions. This means that power absorption from the SOEC stack can be increased to full power on the sub-second scale for short duration transients and on the second scale for long duration transients requiring flow rate changes. This greatly increases the potential for grid voltage and frequency stabilization since the available modulation is much larger than in conventional controlled SOEC systems.

[0078] The above operating method allows for sustained operation of the SOEC system under partial load (i.e., not full load) conditions. The operating method does not rely on auxiliary heating. Instead, the operating method relies on self-heating of the SOEC stack. This allows for separation of heating requirements, thereby simplifying balance of plant in SOEC systems where the balance of plant is an auxiliary device required to ensure the SOEC system operates as a reliable power source. In accordance with the operating method, heat is generated where it is needed (within the SOEC stack), eliminating the need to optimize heat transfer from a separate heater to the SOEC stack, providing simplification to the overall hot module. Thus, the operating method also reduces the thermal load on the SOEC stack for a potential increase in stack efficiency and life.

[0079] As shown in FIG. 1, the performance / temperature curve is not symmetrical. If the SOEC stack is allowed to cool with a part-load operation strategy, the SOEC stack is limited in its ability to ramp power upwards. For example, if operated at 50% power, the SOEC stack may only be able to quickly increase to 60% power, and may take several minutes to ramp to 100% power while maintaining a reasonable temperature within the SOEC stack. In contrast, the above operation method allows for fast ramping from any power level to any other power level as fast as the process gas can be ramped.

[0080] The actuation method may be particularly suitable for highly modular stack arrays, where the impact of electrical switching of the power bus can be mitigated by sequentially taking the stacks out of phase. For example, at 50% load conditions, half of the stacks are on at any particular time, and there is no time when all stacks are active. In a single stack, the current switching may impose noise on the resulting DC bus, which may have adverse consequences on the efficiency of the system power electronics. In a highly modular system (i.e., systems where many stacks share the same DC bus), the transition times of each stack may be offset relative to the other stacks to minimize bus noise. The more stacks that operate electrically in parallel, the less the actuation method impacts the bus voltage and noise. In some aspects, the switching is systematic. In particular, the order and timing of the SOEC stacks being on or off at any given part-load condition is the same. In other aspects, the switching is pseudo-random. In particular, the order and timing of the SOEC stacks being on or off at any given part-load condition is not the same. For example, if one SOEC stack is weaker than another, the pseudorandom approach may smooth out the resulting electrical noise. In terms of eliminating forced noise and unplanned bias in the system, the pseudorandom approach may provide advantages.

[0081] In some embodiments, the application of pulse-width modulated power can be used to control the net thermal operating state of the SOEC stack in a manner independent of the net electrical state. This concept is described herein in terms of maintaining a net thermally neutral state at partial electrical load conditions that would be endothermic. This concept can be equally applied to targeting a particular thermal state, for example, to apply a net heat generating state as a way to help heat the stack. The application of pulse-width modulated power at partial load conditions allows for independent control of the stack thermal state and, as a result, can be used to move the stack temperature up or down in response to other system demands.

[0082] In some embodiments, the operating frequency of an SOEC stack can vary between a lower boundary dictated by the stack geometry and the net flow period for any fluid particle to move from the cell active area inlet to the cell active area outlet, and an upper boundary dictated by the point of diminishing efficiency recovery due to switching losses, and ultimately, the point at which cell electrical performance begins to equal part-load conditions and thermal benefits begin to disappear.

[0083] As used herein, the terms "approximately," "about," "substantially," and similar terms are intended to have a broad meaning consistent with commonly accepted usage by those of ordinary skill in the art to which the subject matter of this disclosure pertains. It should be understood by those of ordinary skill in the art who review this disclosure that these terms are intended to enable the description of the particular features described and claimed without limiting the scope of those features to the precise numerical ranges provided. These terms should therefore be interpreted to indicate that insubstantial or insignificant modifications or variations of the subject matter described and claimed are considered to be within the scope of the invention as recited in the appended claims.

[0084] As used herein, terms such as "coupled," "connected," and the like, mean that two members are joined directly or indirectly to one another. Such joining can be fixed (e.g., permanent) or movable (e.g., removable or releasable). Such joining can be achieved by the two members or the two members and an additional intermediate member being integrally formed with one another as a single, unitary body, or by the two members or the two members and an additional intermediate member being attached to one another.

[0085] References herein to the location of elements (e.g., "top", "bottom", "upper", "lower", etc.) are merely used to describe the orientation of the various elements in the figures. It should be noted that the orientation of the various elements may differ according to other exemplary embodiments, and such variations are intended to be encompassed by the present disclosure.

[0086] It is important to note that the construction and arrangement of the various exemplary embodiments are illustrative only. Although only a few embodiments have been described in detail in this disclosure, those skilled in the art who review this disclosure will readily appreciate that many modifications (e.g., changes in size, dimensions, structure, shape, and proportions of various elements, parameter values, mounting arrangements, use of materials, colors, orientations, etc.) are possible without substantially departing from the novel teachings and advantages of the subject matter described herein. For example, elements shown as integrally formed may be composed of multiple parts or elements, the location of elements may be reversed or otherwise changed, and the nature or number of individual elements or locations may be altered or changed. The order or sequence of any process or method steps may be changed or rearranged according to alternative embodiments. Other substitutions, modifications, changes, and omissions may be made in the design, operating conditions, and arrangement of the various exemplary embodiments without departing from the scope of the present invention. For example, the heat recovery heat exchanger may be further optimized.

[0087] The operations described above may be performed by a computer programmed to execute the steps of an algorithm. Embodiments of the subject matter and operations described herein may be implemented in digital electronic circuitry, or in computer software embodied in tangible media, firmware, or hardware including the structures disclosed herein and their structural equivalents, or in a combination of one or more of them. Embodiments of the subject matter described herein may be implemented as one or more computer programs, i.e., one or more modules of computer program instructions encoded on one or more computer storage media for execution by or for controlling the operation of a data processing apparatus. Alternatively or additionally, the program instructions may be encoded in an artificially generated propagated signal, e.g., a machine-generated electrical, optical, or electromagnetic signal generated to encode information for execution by a data processing apparatus or for transmission to a suitable receiving apparatus. The computer storage medium may be or may be included in a computer readable storage device, a computer readable storage substrate, a random or serial access memory array or device, or a combination of one or more of them. Furthermore, although a computer storage medium is not a propagated signal, a computer storage medium may be a source or destination of computer program instructions encoded in an artificially generated propagated signal. The computer storage medium may also be, or may be contained in, one or more separate components or media (e.g., multiple CDs, disks, or other storage devices). Thus, the computer storage medium may be tangible and non-transitory.

[0088] The operations described herein may be implemented as operations performed by a data processing device or processing circuitry on data stored in one or more computer-readable storage devices or data received from other sources.

[0089] The device may include special purpose logic circuitry, such as an FPGA (field programmable gate array) or an ASIC (application specific integrated circuit). In addition to hardware, the device may also include code that creates an execution environment for the computer program in question, such as code that constitutes processor firmware, a protocol stack, a database management system, an operating system, a cross-platform runtime environment, a virtual machine, or a combination of one or more of these. The device and execution environment may implement a variety of different computing model infrastructures, such as web services, distributed computing, and grid computing infrastructures.

[0090] A computer program (also known as a program, software, software application, script, or code) may be written in any type of programming language, including compiled or interpreted, declarative or procedural, and it may be deployed in any form, including as a stand-alone program or as a module, component, subroutine, object, or other unit suitable for use in a computing environment. A computer program may correspond to a file in a file system, but need not. A program may be stored in a portion of a file that holds other programs or data (e.g., one or more scripts stored in a markup language document), in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files that store one or more modules, subprograms, or portions of code). A computer program located at one site or distributed across multiple sites and interconnected by a communication network may be deployed to run on one computer or on multiple computers.

[0091] The processes and logic flows described herein may be performed by one or more programmable processors or processing circuits executing one or more computer programs to perform actions by operating on input data and generating output. The processes and logic flows may also be performed by, and an apparatus may also be implemented as, special purpose logic circuitry, such as an FPGA or ASIC.

[0092] Processors or processing circuits suitable for executing computer programs include, by way of example, both general-purpose and special-purpose microprocessors, as well as any one or more processors of any kind of digital computer. Generally, a processor will receive instructions and data from a read-only memory or a random access memory, or both. The essential elements of a computer are a processor for performing actions according to instructions, and one or more memory devices for storing instructions and data. Generally, a computer will also include one or more mass storage devices, such as magnetic, magneto-optical, or optical disks, for storing data, or be operatively attached to receive data from or transfer data to, or both. However, a computer need not have such devices. Furthermore, a computer may be embedded in another device, such as a mobile phone, a personal digital assistant (PDA), a mobile audio or video player, a game console, a global positioning system (GPS) receiver, or a portable storage device (e.g., a universal serial bus (USB) flash drive), to name a few. Suitable devices for storing computer program instructions and data include, by way of example, all forms of non-volatile memory, media, and memory devices, including semiconductor memory devices, such as EPROM, EEPROM, and flash memory devices; magnetic disks, such as internal hard disks or removable disks; magneto-optical disks; and CD-ROM and DVD-ROM disks. The processor and the memory can be supplemented by, or incorporated in, special purpose logic circuitry.

[0093] To provide interaction with a user, embodiments of the subject matter described herein may be implemented on a computer having a display device, e.g., a CRT (cathode ray tube) or LCD (liquid crystal display), OLED (organic light emitting diode), TFT (thin film transistor), plasma, other flexible configurations, or any other monitor for displaying information to a user, and a keyboard, a pointing device, e.g., a mouse trackball, or the like, or a touch screen, touch pad, or the like, through which a user can provide input to the computer. Other types of devices may also be used to provide interaction with a user, e.g., feedback provided to the user may be any form of sensory feedback, e.g., visual feedback, auditory feedback, or haptic feedback, and input from the user may be received in any form, including acoustic input, speech input, or tactile input. In addition, a computer may interact with a user by sending documents to and receiving documents from a device used by the user, e.g., by sending a web page to a web browser on the user's client device in response to a request received from the web browser.

Claims

1. 1. A method for operating a solid oxide electrolysis cell system at part load, the solid oxide electrolysis cell system comprising a plurality of branches electrically connected in parallel, each branch comprising at least one solid oxide electrolysis cell stack, each solid oxide electrolysis cell stack comprising a plurality of solid oxide electrolysis cells, the method comprising: determining a thermal-neutrality target voltage, for a given operating temperature, below which operation of the solid oxide electrolysis cell system is endothermic and above which operation of the solid oxide electrolysis cell system is exothermic; performing pulse width modulated current control by cycling on and off phases for each of the branches such that during an operating cycle of the solid oxide electrolysis cell system, the solid oxide electrolysis cell system operates at an average operating power equal to a selected percentage of an operating power at the thermal neutral target voltage; during the on-phase, all of the solid oxide electrolysis cell stacks in a given branch operate at the thermally neutral target voltage, and during the off-phase, all of the solid oxide electrolysis cell stacks in the given branch are unloaded to an open circuit voltage and operate at 0% of rated power while steam flow to the stacks in the given branch continues uninterrupted; each of the branches is configured to operate in the on-phase or the off-phase independently of the other branches; and performing, in said operating cycle, each branch entering a first off phase before any branch enters a second off phase.

2. The method of claim 1 , wherein all of said branches are in said off phase during at least one of said operating cycles.

3. The method of claim 1 , wherein all of said branches are in said on-phase during at least one of said operating cycles.

4. 2. The method of claim 1, wherein during at least one period of the operating cycle, at least one branch is in the ON phase while at least one other branch is in the OFF phase.

5. In said operating cycle, a first branch is in the on-phase for the duration of a first period, while a second branch is in the off-phase for the duration of the first period; the second branch is switched to the on-phase at the beginning of a second period and remains in the on-phase for the duration of the second period; 2. The method of claim 1, wherein the first branch remains in the on phase for the duration of the second period or is switched to the off phase at the start of the second period and remains in the off phase for the duration of the second period.

6. The method of claim 1 , wherein each branch comprises a plurality of solid oxide electrolysis cell stacks.

7. Implementing pulse width modulated current control 2. The method of claim 1, comprising selecting a pulse width modulation frequency and a duty cycle such that a transition period between an on phase and an off phase or between an off phase and an on phase occupies less than 10 percent of the duration of the operating cycle and such that the prevailing thermal condition is thermal neutral.

8. 8. The method of claim 7, wherein the pulse width modulation frequency and the duty cycle are selected such that each branch switches from the on phase to the off phase before cell starvation due to reactant depletion occurs.

9. the total power cycles is a first predetermined percentage of the rated power; each of said branches is operated in said on phase for a predetermined percentage of time; The method of claim 1 , wherein the first predetermined percentage of rated power is equal to a predetermined percentage of the time during which each of the branches operates in the on-phase.

10. 2. The method of claim 1, wherein at least one of the branches operates at 100% of rated power in a first ON phase and at 0% of rated power in a subsequent OFF phase immediately following the first ON phase.

11. 10. The method of claim 1, wherein at least one of the branches is unloaded and operated at 0% of rated power at regular intervals in an operating cycle of the solid oxide electrolysis cell system.

12. The method of claim 1 , wherein at least one of the branches is unloaded at irregular intervals in an operating cycle of the solid oxide electrolysis cell system.

13. 1. A solid oxide electrolysis cell system comprising: a plurality of branches electrically connected in parallel, each branch including at least one solid oxide electrolysis cell stack, each solid oxide electrolysis cell stack including a plurality of solid oxide electrolysis cells; determining a thermal-neutrality target voltage, where operation of the solid oxide electrolysis cell system is endothermic below the thermal-neutrality target voltage and operation of the solid oxide electrolysis cell system is exothermic above the thermal-neutrality target voltage; a controller programmed to perform pulse width modulated current control by cycling on and off phases for each of the branches such that the solid oxide electrolysis cell system operates at the thermally neutral target voltage during an operating cycle of the solid oxide electrolysis cell system; In the on-phase, the branch operates at the thermal neutral target voltage; In the off phase, a branch is unloaded to an open circuit voltage and operates at 0% of rated power while steam flow to the stack in the branch continues uninterrupted; each of the branches is configured to operate in the on-phase or the off-phase independently of the other branches; A solid oxide electrolysis cell system, wherein in said operating cycle, each branch enters a first off phase before any branch enters a second off phase.

14. 14. The solid oxide electrolysis cell system of claim 13, wherein each of the plurality of solid oxide electrolysis cells comprises an anode, a cathode, and a solid ceramic electrolyte between the anode and the cathode.

15. 15. The solid oxide electrolysis cell system of claim 14, wherein the solid ceramic electrolyte comprises a non-porous metal oxide.

16. A method for operating a solid oxide electrolysis cell system at partial load, the solid oxide electrolysis cell system including a plurality of branches electrically connected in parallel, each branch including at least one solid oxide electrolysis cell stack, each solid oxide electrolysis cell stack including a plurality of solid oxide electrolysis cells, the method comprising: determining a thermal-neutrality target voltage, for a given operating temperature, below which operation of the solid oxide electrolysis cell system is endothermic and above which operation of the solid oxide electrolysis cell system is exothermic; performing pulse width modulated current control by cycling on and off phases for each of the branches such that during an operating cycle of the solid oxide electrolysis cell system, the solid oxide electrolysis cell system operates at an average operating power equal to a selected percentage of an operating power at the thermal neutral target voltage; during the on-phase, all of the solid oxide electrolysis cell stacks in a given branch operate at the thermally neutral target voltage, and during the off-phase, all of the solid oxide electrolysis cell stacks in the given branch are unloaded to an open circuit voltage and operate at 0% of rated power while steam flow to the stacks in the given branch continues uninterrupted; each of the branches is configured to operate in the on-phase or the off-phase independently of the other branches; and performing, in the operating cycle, a frequency of the off phase for each branch based on a capacity or health state of the at least one solid oxide electrolysis cell stack in the branch, such that a stronger or healthier branch has a lower frequency of off phase than a weaker or less healthy branch.

17. A solid oxide electrolysis cell system comprising: a plurality of branches electrically connected in parallel, each branch including at least one solid oxide electrolysis cell stack, each solid oxide electrolysis cell stack including a plurality of solid oxide electrolysis cells; determining a thermal-neutrality target voltage, where operation of the solid oxide electrolysis cell system is endothermic below the thermal-neutrality target voltage and operation of the solid oxide electrolysis cell system is exothermic above the thermal-neutrality target voltage; a controller programmed to perform pulse width modulated current control by cycling on and off phases for each of the branches such that the solid oxide electrolysis cell system operates at the thermally neutral target voltage during an operating cycle of the solid oxide electrolysis cell system; In the on-phase, the branch operates at the thermal neutral target voltage; In the off phase, a branch is unloaded to an open circuit voltage and operates at 0% of rated power while steam flow to the stack in the branch continues uninterrupted; each of the branches is configured to operate in the on-phase or the off-phase independently of the other branches; A solid oxide electrolysis cell system, wherein in the operating cycle, the frequency of the off phase of each branch is determined based on the capacity or health of the at least one solid oxide electrolysis cell stack in the branch, such that a stronger or healthier branch has a lower frequency of off phase than a weaker or less healthy branch.

Images