DK182028B1

DK182028B1 - Alkaline electrolyser and a method for its operation

Info

Publication number: DK182028B1
Application number: DKPA202330316A
Authority: DK
Inventors: Stiesdal Henrik; Vinter Voigt Niels; Gellrich Florian
Original assignee: Stiesdal Hydrogen As
Priority date: 2023-11-03
Filing date: 2023-11-03
Publication date: 2025-06-04
Also published as: DK202330316A1

Abstract

An alkaline electrolyzer comprising a stack (17) of electrolytic cells (1) is used for producing hydrogen gas (8). Each of the cathode compartments (5) comprises a cathode gas outlet (23A) into a cathode electrolyte return conduit (22A), the downstream end (41) of which is connected to a hydrogen purifier (33) configured for providing purified hydrogen gas by removing oxygen from the gas received from the cathode electrolyte return conduit (22A). A cathode gas recirculation system (38) connects a downstream end of the hydrogen purifier (32,33) to an upstream end (40) of the cathode electrolyte return conduit (22A) for supplying purified hydrogen gas to the cathode electrolyte return conduit (22A). Alternatively, or in addition, each of the anode compartments (6) comprises an anode gas outlet (23B) into an anode electrolyte return conduit (22B), the downstream end (41) of which is connected to an oxygen purifier (33), configured for providing purified oxygen gas by removing hydrogen from the gas coming from the anode electrolyte return conduit (22B). An anode gas recirculation system (38) connects a downstream end (41) of the oxygen purifier (33) to an upstream end (40) of the anode electrolyte return conduit (22B) for supplying purified oxygen gas to the anode electrolyte return conduit (22B). By recirculating purified gases through the electrolyte return conduits, the electrolyzer can operated at part load, for example below 10% of the nominal load.

Description

DK 182028 B1 1

Alkaline electrolyser and a method for its operation

FIELD OF THE INVENTION

The present invention relates to an alkaline electrolyser for water electrolysis, for ex- ample for hydrogen production, and to a method for the operation of the electrolyser.

BACKGROUND OF THE INVENTION

Electrolysis is a promising technology for the production of hydrogen from renewable energy sources. So-called electrolysers are used for this purpose. The primary objec- tive of an electrolyser is for production of hydrogen gas. Hydrogen is collected for later use, for example in fuel cells or industrial applications. However, due to the split- ting of water in the electrolyte when applying electrical power, oxygen is also pro- duced. The oxygen may also be collected for later use.

Several types of electrolysers exist, each exhibiting advantages and disadvantages. Al- kaline electrolysis is generally seen as a mature, low-cost, robust technology, but it is disadvantaged by limited ability to operate at part load. In this context, “load” means the electric power supplied to the electrolyzer to drive the splitting of water into gase- ous oxygen and hydrogen. “Nominal load” means the application of a level of power to the electrolyser corresponding to the nameplate power rating of the electrolyser, and “part load” means the application of a level of power to the electrolyser that is lower than the nameplate power rating of the electrolyser.

Due to the limited ability of alkaline electrolysers to operate at part load below a cer- tain limit, other electrolyser technologies such as PEM or SOEC are sometimes pre- ferred over alkaline electrolysers due to their capability to operate at lower part load than alkaline electrolysers, even though this capability is often acquired at the expense of higher costs or other disadvantages, such as the need for rare metal or the need for high operational temperatures.

DK 182028 B1 2

The inability of alkaline electrolysers to operate at part load below certain limits is re- lated to the permeability of the separators used in the electrolytic cells that are at the core of the electrolyser.

In an alkaline electrolyser, multiple electrolytic cells are connected in the so-called electrolyser stack. Each electrolytic cell comprises an anode and an associated com- partment filled with electrolyte, a cathode and an associated compartment filled with electrolyte, and a separator sandwiched between the two compartments. Typically, the electrolyte is an aqueous potassium hydroxide solution of 25-35 percent concentration by weight, but other electrolyte compositions are also used.

In traditional alkaline electrolysers, each electrolytic cell has two electrode plates sep- arated by a certain distance, and with the separator located about halfway between the two electrodes. The gap between the electrodes is filled with the electrolyte. When sufficient voltage is applied, hydrogen is released on the cathode surface and oxygen is released on the anode surface. The gap between the electrodes needs to be of suffi- cient width to allow the escape of hydrogen and oxygen bubbles without excessive blocking of the conductive path through the electrolyte from the anode to the cathode, and to allow electrolyte circulation without excessive pressure loss.

In later years, the configuration of traditional alkaline electrolysers has been replaced by a so-called zero-gap configuration. In the zero-gap configuration the cell design works by pressing two porous electrodes onto either side of the separator. This achieves a gap between the two electrodes equal to the sum of the thickness of the separator, typically 0.5 mm or even less, and any additional distances from each elec- trode to the separator, rather than the 2-5 mm required for the traditional gap configu- ration. The smaller gap reduces the ohmic resistance contribution of the electrolyte to the losses in the electrolytic cells. The electrodes have pores in order to allow the es- cape of hydrogen and oxygen bubbles to the side of the electrode not facing the sepa- rator.

Hence, an electrolytic cell of an alkaline electrolyser is typically composed of a first solid metallic plate, an anode electrically connected to this first solid metallic plate and separated from it at sufficient distance to facilitate the escape of oxygen bubbles through the electrolyte without excessive blocking, a separator, and a cathode electri- cally connected to a second solid metallic plate and separated from this second solid plate at sufficient distance to facilitate the escape of hydrogen bubbles through the electrolyte without excessive blocking.

In the electrolyser stack the electrolytic cells are placed back-to-back, and the second solid metallic plate of a first electrolytic cell is the first solid metallic plate of a second electrolytic cell.

During operation of the electrolyser, liquid water is split into gaseous oxygen and hy- drogen through electrochemical reactions at the electrodes of each of the electrolytic cells in the stack.

At the cathode, the following reaction takes place: 2H0 (9) + 20 — Hz (2) + 20H (aq)

At the anode, the following reaction takes place:

ZOH (aq) — 402 (2) + HOH (1) + 2e'

The overall reaction is:

H0 (1) > Ho fø) + Mb fg)

This equation represents the splitting of water into hydrogen and oxygen gases using electrical energy. The hydrogen gas is collected at the cathode, while the oxygen gas is collected at the anode.

The electrolyte solution, which contains hydroxide ions (OH), plays a crucial role in facilitating the movement of ions and ensuring the completion of the electrochemical circuit. The hydroxide ions created at the cathode migrate from the cathode compart ment to the anode compartment through the separator, allowing the overall electrolysis

DTOCESS TØ OCCur.

DK 182028 B1 4

Water is consumed from the electrolyte at a rate corresponding to the gas production rate. In order to maintain an approximately constant concentration of the electrolyte an electrolyte circulation arrangement is established whereby the electrolyte is circulated through the electrolytic cells of the electrolyser stack, through two gas separation tanks for oxygen and hydrogen, respectively, replenished with fresh water, and pumped back into the electrolytic cells of the electrolyser stack.

The electrolyte circulation is enabled with conduits connecting the electrolytic cells in the stack. Typically, one electrolyte feed conduit connects all the anode compartments in the stack, and another electrolyte feed conduit connects all the cathode compart- ments in the stack. Other arrangements are also used, e.g., having a single feed con- duit connecting both the anode compartments and the cathode compartments in the stack, or having several feed conduits each connecting only part of the compartments in the stack.

Electrolyte is returned from the electrolytic cells through two or more electrolyte re- turn conduits connecting the electrolytic cells in the stack. Gas produced in the elec- trolytic cells is also removed through these electrolyte return conduits. Typically, one electrolyte return conduit connects all the anode compartments in the stack, and an- other electrolyte return conduit connects all the cathode compartments in the stack.

Other arrangements are also used, e.g., having several electrolyte return conduits each connecting only part of the compartments in the stack.

During operation of the electrolyser, gaseous oxygen bubbles form at or near the an- ode electrode in the anode compartments of the stack, and gaseous hydrogen bubbles form at or near the cathode electrode in the cathode compartments of the stack. In each half-cell compartment, the bubbles mix with the liquid electrolyte to form an electrolyte-bubble mix that is evacuated to the electrolyte return conduit of the rele- vant gas, the oxygen electrolyte return conduit for the anode compartments and the hydrogen electrolyte return conduit for the cathode compartments.

As mentioned above, the hydroxide ions created at the cathode migrate from the cath- ode compartment to the anode compartment through the separator, allowing the

DK 182028 B1 overall electrolysis process to occur. For this transport of hydroxide ions to occur, the separator is typically porous and hydrophilic. As a consequence, gas diffusion across the separator is unavoidable. Even with the emergence of new dense separator types such as anion exchanging membranes, gas-containing electrolyte will be exchanged 5 between the two half cells over the separator. Hence, gas crossover may be reduced but not avoided in alkaline electrolyzers. This gas diffusion across the separator, nor- mally designated crossover, occurs at all load levels. The gas crossover is mainly driven by the differences in partial pressure at the two different sides of the separator.

In the anode compartment, the electrolyte is saturated with oxygen, and in the cathode compartment, the electrolyte is saturated with hydrogen. It is the differences in partial pressures resulting from the saturation that causes diffusion of the gases across the separator. Oxygen diffuses from the anode compartment to the cathode compartment, and hydrogen diffuses from the cathode compartment to the anode compartment.

Saturation of the electrolyte occurs rapidly when the production of oxygen and hydro- gen commences, and following an initial build-up of saturation, the rate of crossover is fairly constant, irrespective of the electrolyser load. Additional crossover may occur as a result of supersaturation close to the electrodes, and this additional crossover may to some extent be load-dependent, but for all practical purposes, the crossover is inde- pendent of the electrolyser load.

Oxygen-hydrogen gas mixtures are not combustible if the concentration of either of the gases in the gas mixture is lower than 4 percent. If the concentration of either of the gases in the gas mixture is higher than 4 percent, the gas mixture may be combus- tible. Hence, a concentration of more than 4 percent of one of the gases in the gas mixture is considered to be critical.

At high load, the gas production in each electrolytic half-cell is much larger than the crossover of gas from the opposite electrolytic half-cell, and the gas contamination is well below the critical level. However, at low part load operation, the gas contamina- tion may exceed the critical level due to the lower gas production, causing the gas mixture in the individual gas bubbles to become combustible.

DK 182028 B1 6

Combustibility is, generally, not a problem in the electrolyte-bubble mix that occurs in an electrolytic half-cell, since the volume of each bubble is very small, typically on the order of fractions of cubic millimetres to single-digit cubic millimetres. No source of ignition exists in a bubble, and even if detonation should occur in one or few bub- bles, the likelihood of chain-reaction detonations is small, since the multitude of bub- bles in the electrolyte-bubble mix serves as a spring-damper system that reduces the pressure waves of any small detonations.

The situation changes when the electrolyte-bubble mix is evacuated from the half- cells through orifices to the electrolyte return conduit of the relevant gas. In the elec- trolyte return conduit the gas bubbles will coalesce to form large bubbles, often com- pletely separating from the electrolyte. If the gas is contaminated to a level where it is combustible, i.e., if the concentration of oxygen in the hydrogen electrolyte return conduit is higher than 4 percent or the concentration of hydrogen in the oxygen elec- trolyte return conduit is higher than 4 percent, then, the risk of spontaneous combus- tion or even detonation is imminent.

An even worse situation may occur in the gas separation tanks. Here, the gases are separated from the electrolyte, and gas volumes may be on the order of hundreds of litres per tank, or even cubic metres. Combustion of such volumes of contaminated gas may lead to catastrophic explosion.

Consequently, it is not possible to operate an alkaline electrolyser at a part load lower than that at which the gas contamination caused by crossover reaches the critical level of 4 percent. In practice, at level of gas contamination of 2 percent is typically set as the operational limit.

The lower threshold for safe part load operation of alkaline electrolysers that will meet the operational limit of 2 percent gas contamination is often stated to be in the range of 10-25 percent, as demonstrated on p. 185 of the publication, Innovation landscape for smart electrification: Decarbonising end-use sectors with renewable power, pub- lished by the International Renewable Energy Agency (IRENA), Abu Dhabi, ISBN: 978-92-9260-532-2 (2023).

DK 182028 B1 7

When forced to operate close to or below the lower threshold for safe part load opera- tion of alkaline electrolysers, e.g., as a result of variations in the input power caused by the variability of renewable energy sources, such as solar or wind power, it is cus- tomary to shut down the electrolyser.

Shutdown of the electrolyser at low load has several negative effects. First, hydrogen production is lost. Second, restarting the electrolyser takes time, and the loss of hydro- gen production may be sustained for a while at very considerable load levels, aggra- vating the loss of hydrogen production. Finally, frequent shutdowns and start-ups ac- celerate electrode degradation, reducing the electrolyser's expected lifespan and in- creasing the costs for maintenance and replacement.

Alternative strategies may be implemented to reduce the negative effects of shutdown.

Such alternative strategies may involve using batteries or other energy storage devices to absorb power fluctuations, or the implementation of cascaded start-up and shut- down of a string of electrolysers.

Gas crossover and contamination cause additional problems at the restart of alkaline electrolyzers after shut-down.

During operation prior to shut-down the electrolyte in each of the compartments will have been saturated with the gas produced in the respective compartment; in the anode compartments, the electrolyte is saturated with oxygen, and in the cathode compart- ments, the electrolyte is saturated with hydrogen. As described above, gas crossover is mainly driven by the differences in partial pressure arising from difference in gas con- centration at the two different sides of the separator. The differences in partial pres- sure will cause diffusion of the gases across the separator.

During standstill, gas crossover, 1.e., the diffusive flux of gas across the separator, with oxygen diffusing from the anode compartment to the cathode compartment and hydrogen diffusing from the cathode compartment to the anode compartment, will continue until the partial pressures of the two gases are the same at both sides of the separator.

DK 182028 B1 8

During operation prior to shut-down, gaseous oxygen bubbles will have formed at or near the anode electrode in the anode compartments of the stack, and gaseous hydro- gen bubbles will have formed at or near the cathode electrode in the cathode compart- ments of the stack. Some of these bubbles will remain on the electrodes after shut- down. Initially, the bubbles will contain clean or almost clean gas, but soon an equi- librium will be established where the gas mixture in each bubble will correspond to the ratio of partial pressures in the electrolyte. Consequently, once crossover has caused the partial pressures of the two gases to be the same at both sides of the separa- tor, all bubbles will have the same mixed gas composition, which is invariably com- bustible.

As described above, the presence of combustible gas mixtures in small bubbles sur- rounded by electrolyte does not in itself pose a risk. However, once the electrolyser is restarted, the bubble formation resulting from the gas evolution of the water-splitting process will push the highly contaminated bubbles from the electrode surfaces. As a consequence, the gases initially delivered to the electrolyte return conduits will be contaminated to a level that poses imminent danger of combustion or explosion in the electrolyte return conduits and the gas separation tanks.

Typically, the problem of high contamination of the gases initially delivered to the electrolyte return conduits at startup is overcome through strong flushing of the elec- trolyte return conduit with nitrogen, diluting the contaminated gases to a level where combustion is no longer possible. The duration of flushing may be on the order of minutes or tens of minutes. The IRENA reference quoted above specifies the start-up time to be in the range of 1-10 minutes.

Flushing with nitrogen has several disadvantages. First, significant amounts of nitro- gen are required to ensure safe operation, and this requires substantial nitrogen supply arrangements, typically, comprising batteries of high-pressure nitrogen flasks or even nitrogen generators connected through complex arrangements of pipes and valves. All of this adds both investment and operational costs to the system. Second, the hydrogen produced during startup is by default contaminated with nitrogen. In many use cases for hydrogen such contamination is not permissible, and since it is difficult to separate

DK 182028 B1 9 nitrogen from hydrogen it is often necessary to simply vent the contaminated hydro- gen to the atmosphere, representing a manifest loss of economic value.

US10865486 discloses a method for reduction of the partial gas pressure in the elec- trolytic half-cells of a high-temperature electrolyser through the flushing with gases.

The actual aim of the flushing is to reduce the respective partial pressure at the flushed electrode. For example, flushing the anode leads to a reduction in the oxygen partial pressure at the anode, which reduces the corrosive properties of oxygen at the anode.

Furthermore, it is noted as an advantage that the removal of the gas formed on the electrodes avoids reduction in the electrochemical performance of the electrolysis cell through removal of excess gases.

Generally, the flushing gases disclosed in US10865486 are inert gases such as argon or nitrogen, but it is noted as an option that a part of the hydrogen generated at the cathode can be recirculated, that is to say, the hydrogen gas can again be fed in the re- circulation circuit to the cathode for the flushing thereof. A similar approach can be taken for the anode. It is stated that the overall conversion rate of the electrolysis sys- tem is increased by recycling an unused reactant or a product gas mixture containing the reactant to the process inlet, which is to say, re-feeding the unused reactant to the cathode of the electrolysis cell.

KR102526673B1 and the equivalent US2024/218527A1 disclose recirculation of hy- drogen gas into the cathode for cooling purposes.

The method disclosed in US10865486 aims at high-temperature electrolysis but may in principle also serve to reduce the concentration of the gases produced in the electro- lytic cells of an alkaline electrolyser and thereby indirectly reduce the contamination level caused by gas crossover, allowing operation at lower part load levels than the usual 10-25 percent of nominal load. However, the method disclosed in US10865486 has several disadvantages. First, in alkaline electrolysers it is not straightforward to ensure equal distribution of the flushing gases to all the electrolytic cells of the elec- trolyser stack, since the gases will need to be mixed with the liquid electrolyte fed to the electrochemical cells. It is notoriously difficult to ensure even distribution of gas and liquid when feeding a mixture of the two phases to a long string of recipients.

DK 182028 B1 10

Second, if inert gases are used for the flushing, these gases will need to be separated from the hydrogen gas before supply to the user of the gas. Such gas separation is al- ways difficult and often impossible. Third, if the hydrogen generated at the cathode is recirculated for the flushing of the cathode compartments the oxygen content in the hydrogen will remain at the initial level, and no benefit is achieved regarding part load operation. The same applies on the oxygen side.

Hence, there is an unfulfilled need for alkaline electrolysers that can continue safe op- eration to very low levels of part load operation, and that can startup at short notice with little or no contamination with nitrogen.

DESCRIPTION / SUMMARY OF THE INVENTION

It is therefore an objective of the invention to provide an improvement in the art. In particular, it is an objective to provide an alkaline electrolyser that has a high degree of gas purity facilitating part load operation at a very low fraction of nominal load and rapid startup leading to little or no contamination of the produced hydrogen with nitro- gen. Furthermore, it is an objective to provide a method for the operation of the elec- trolyser. These objectives and further advantages are achieved with an alkaline elec- trolyser for production of hydrogen gas as described below and in the claims.

The objectives and further advantages are achieved with a gas recirculation system that provides cleaned gas to the electrolyte return conduits connecting the electrolytic cells of the electrolyser stack.

The stack comprises a series of electrolytic cells. Each electrolytic cell comprises a cathode compartment and an anode compartment that are both containing liquid alka- line electrolyte and are separated by an ion-conducting separator, Typically, the elec- trolyser is operated with an alkaline electrolyte based on NaOH or KOH, for example having a temperature in the range of in the range of 50-90 °C.

Each of the cathode compartments comprises a cathode electrolyte outlet into a cath- ode electrolyte return conduit, the downstream end of which is connected to a

DK 182028 B1 11 hydrogen gas separation tank, which separates the hydrogen and the electrolyte com- ing from the return conduit.

Advantageously, a hydrogen purifier is connected to the downstream end of the hy- drogen gas separation tank and configured for providing purified hydrogen gas by re- moving oxygen from the gas received from the cathode electrolyte return conduit.

Alternatively, or in addition, similarly, each of the anode compartments comprises an anode electrolyte outlet into an anode electrolyte return conduit, the downstream end of which is connected to an oxygen gas separation tank, which separates the oxygen and the electrolyte coming from the return conduit.

Advantageously, an oxygen purifier is connected to the downstream end of the oxygen gas separation tank and configured for providing purified oxygen gas by removing hy- drogen from the gas received from the anode electrolyte return conduit.

Gas purification systems for hydrogen and oxygen are typically established with a gas recombination system comprising a heterogenous catalyst such as palladium or plati- num that facilitates rapid recombination of oxygen and hydrogen to form water. In practice, a gas recombination system typically comprises one or more steel tanks filled with a granular or fibrous heterogenous catalyst. Other systems for gas purification exist, including pressure swing absorption systems and other systems known from the process industry.

In the following, the term “purifier” will be used for a gas purification system irre- spective of the actual technology applied.

The feed of one type of gas to the purifier may have up to 2 percent contamination with the other type of gas. Gas purifiers may be designed for varying levels of cleanli- ness at the outlet; typically, a cleanliness of 99.99 percent may be a target value. In the following description the gas delivered from the outlet of the gas purifier is considered to be clean, since the contamination is one or more orders of magnitude lower than the contamination of the gas prior to purification.

DK 182028 B1 12

Each of the anode and cathode compartments comprises an electrolyte inlet and an electrolyte outlet, advantageously being identical to the respective anode and cathode gas outlets and leading into the respective electrolyte return conduits. For recircula- tion, the gases are separated from the electrolyte in corresponding separation stages.

For example, the electrolyzer comprises an electrolyte recirculation system for recir- culation of electrolyte from the electrolyte outlets to the electrolyte inlets. The electro- lyte recirculation system comprises an alkalinity adjustment system for adjusting the alkalinity of the alkaline electrolyte to predetermined specifications prior to re-injec- tion of the electrolyte through the electrolyte inlets. In particular, water has to be sup- plied to the electrolyte for the hydrogen production. In operation, electrolyte is ex- tracted from each anode compartment and each cathode compartment through the electrolyte outlets, which are also the respective gas outlets. The mix of gas and elec- trolyte flows into the respective electrolyte return conduits. Then, the electrolyte is separated from the respective gases and collected, and the alkalinity of the alkaline electrolyte adjusted to predetermined specifications prior to re-injecting the electrolyte into the anode and cathode compartments through electrolyte inlets.

In practical embodiments, the electrolyser comprises a separation tank for each of ox- ygen and hydrogen, respectively, a water replenishment system and a pump system to return the replenished electrolyte to the electrolytic cells of the electrolyser stack.

The gas outlet from the separation tank for the hydrogen or oxygen side or both is connected to a corresponding purifier.

Typically, a gas cleaning system is installed between the separation tank and the gas purifier. The gas cleaning system may comprise a scrubber and a demisting system, preventing electrolyte aerosol from contaminating the gas purifier.

The electrolyser according to the invention further comprises one or two gas recircula- tion systems, one connecting the outlet of the hydrogen gas purifier to the upstream end of the cathode electrolyte return conduit, which functions as a hydrogen return conduit in the electrolyser stack, the other connecting the outlet of the oxygen gas pu- rifier to the upstream end of the second electrolyte return conduit, which functions as

DK 182028 B1 13 an oxygen return conduit in the electrolyser stack. Each gas recirculation system com- prises means for recirculating the gas, e.g., a gas pump or compressor, and a piping system.

The gas recirculation system or systems is/are used to provide clean gas to the electro- lyte return conduits, diluting contaminated gas with clean gas of the same type. The dilution level is set to deliver gas at a safe level of contamination irrespective of the degree of contamination delivered from the electrolytic cells to the electrolyte return conduit.

Advantageously, the electrolyzer comprises both a cathode gas recirculation system and an anode gas recirculation system.

However, in certain circumstances, not both but only one of them is provided. For ex- ample, if an already existing system that has been in operation is upgraded, it may de- sirable to provide only one of the gas recirculation system as a retrofit. For example, the already existing system may have a hydrogen purifier, making it relatively easy to establish a recirculation system for hydrogen, but space constraints may make it unat- tractive to also establish a recirculation system with an oxygen purifier, and it is more desirable to use an already existing inert gas purging system for the anode electrolyte return conduit. Other considerations may apply for establishing an oxygen recircula- tion system but using a different safety mechanism for the cathode side of the electro- lyzer.

In some operation, gas from the downstream end of the cathode electrolyte return con- duit is received by the hydrogen purifier through the gas separation tank, and purified hydrogen gas is provided by removing oxygen by the hydrogen purifier. At least a portion of the purified hydrogen gas may be recirculated by the hydrogen gas recircu- lation system through the cathode electrolyte return conduit while operating the elec- trolyser for hydrogen production.

In some operation, gas from the downstream end of the anode electrolyte return con- duit is received by the oxygen purifier through the gas separation tank, and purified oxygen gas is provided by removing hydrogen by the oxygen purifier. At least a

DK 182028 B1 14 portion of the purified oxygen gas may be recirculated by the oxygen gas recirculation system through the anode electrolyte return conduit while operating the electrolyser for hydrogen production.

Advantageously, the electrolyzer also comprises one or more gas analysing systems, for example one configured for measuring oxygen contamination levels in hydrogen gas flowing from the cathode electrolyte return conduit and optionally another for measuring hydrogen contamination levels in oxygen gas flowing from the anode elec- trolyte return conduit. These gas analysing systems can be used in a control system for determining the recirculation flow of the purified gases.

For example, the gas recirculation system may only be in operation if the measured contamination levels of one or both of the gases, hydrogen and oxygen, are above a threshold level. Alternatively, it may be in continuous operation.

In some practical embodiments, gas recirculation is operated at variable flow rates, for example by operating recirculation pumps or compressors at variable speed, adjusted in dependence of the measured contamination levels, wherein different speeds are used for different contamination levels.

The relative amounts of the two gases that are recirculated may be identical, but they do not need to be so. For example, the diffusion of hydrogen across the separator may not be the same as the diffusion of oxygen across the separator, and therefore the gas contamination may not be the same on the anode side and the cathode side. Conse- quently, the typical 2 percent gas contamination operational limit may occur sooner on either the anode or the cathode side than on the other side. In this case it may be pref- erable to operate recirculation on the anode side while not operating recirculation on the cathode side, or to operate recirculation on the cathode side while not operating re- circulation on the anode side. In another embodiment it may be preferable to operate recirculation on both sides, but at different flow rates.

In some cases, it may be preferable to establish recirculation on only the anode side and not the cathode side, or on only the cathode side and not the anode side.

DK 182028 B1 15

An estimate of the dilution level and the amount of gas that is required to be recircu- lated can be determined on the basis of simple assumptions.

It is assumed that gas crossover is independent of the electrolyser load. It is also as- sumed that the 2 percent contamination threshold for safe operation is reached at a certain load when recirculation is not implemented. The relative load at which the threshold occurs is designated as X, where X is the load relative to nominal load.

The gas contamination at any load can now be calculated:

Contamination = 2 percent * (X / actual load in percent)

For example, it follows that for X = 20 percent the contamination level will be 20 per- cent at 2 percent load, 4 percent at 10 percent load, 2 percent at 20 percent load, and 0.4 percent at nominal load. Calculating the contamination level at very low part load leads to nonphysical results.

It is now assumed that a certain percentage of the gas produced at nominal load is re- circulated. The recirculation percentage is designated as C, where C is the percentage of gas recirculated relative to the gas produced at nominal load.

The gas contamination at any other load can then be calculated:

Contamination = 2 percent * (X / (C + actual load in percent))

If follows that if C is set to the same value as X then the contamination at zero load will be 2 percent, and that it will be lower than the 2 percent threshold at all load val- ues from zero up to nominal load.

For example, it follows that for C = X = 20 percent the contamination level will be 2 percent at zero load, 1.8 percent at 2 percent load, 1.3 percent at 10 percent load, 1 percent at 20 percent load, and 0.3 percent at nominal load.

DK 182028 B1 16

Consequently, the implementation of the recirculation arrangement according to the invention enables the safe operation of alkaline electrolysers through the nominal load range, from zero to 100 percent of nominal power.

The designer of an alkaline electrolyser according to the invention is not restricted to select a value of the recirculation percentage C that is equal to the value of the thresh- old percentage X.

The designer of the alkaline electrolyser according to the invention may elect to use a value of the recirculation percentage C that is lower than the value of the threshold percentage X.

For example, for C = 0.5 X = 10 percent the contamination level will be 4 percent at zero load, 3.3 percent at 2 percent load, 2.0 percent at 10 percent load, 1.3 percent at percent load, and 0.4 percent at nominal load.

In this case operation to zero load will not be possible, but the dynamic range of the electrolyser will have been extended. 20 The designer of the alkaline electrolyser according to the invention may alternatively elect to use a value of the recirculation percentage C that is higher than the value of the threshold percentage X.

For example, for C =2 X = 40 percent the contamination level will be 1.0 percent at zero load, 1.0 percent at 2 percent load, 0.8 percent at 10 percent load, 0.7 percent at 20 percent load, and 0.3 percent at nominal load.

In this case operation to zero load is not only possible but will happen at a contamina- tion level that is even lower than the threshold percentage X.

Hence, the designer of an alkaline electrolyser according to the invention will have a hitherto unknown design flexibility available, permitting the selection of target con- tamination levels through the load range from zero to nominal load.

DK 182028 B1 17

As recirculation of the purified gases through the corresponding electrolyte return conduits is done while operating the electrolyser for hydrogen production at reduced load, for example at a part load lower than 25% or lower than 10% of the nominal load, it is possible to continue operating the electrolyzer and produce hydrogen at such part loads without facing risk for explosion. A threshold value, for example at such 25% or 10%, can be used as a trigger for starting recirculation, but it is also possible to use the recirculation at all times and potentially use the levels of the part load as cri- teria for adjusting the recirculation flow rate, for example by adjusting the recircula- tion speed.

It is also possible, to reduce and then stop the hydrogen production, and provide a quick and safe restart by recirculating hydrogen gas through the hydrogen purifier and the cathode electrolyte return conduit and recirculating oxygen gas through the respective oxygen purifier and the anode electrolyte return conduit while the operation is stopped and until restart of hydrogen production by the electrolyzer.

This improved dynamic range achieved with the invention has several advantages.

First, the stopping of electrolysers at a lower threshold for safe part load operation, e.g., as a result of variations in the input power caused by the variability of renewable energy sources, such as solar or wind power, can be avoided with obvious economic benefits. Second, unnecessary waiting time at restart can also be avoided, enhancing the economic benefits. Third, the reduction of the electrolyser's expected lifespan and the increased costs for maintenance and replacement that typically result from fre- quent shutdowns can be reduced or eliminated.

The implementation of the recirculation arrangement according to the invention will also enable the flushing of the electrolyte return conduits, with the relevant clean gas prior to startup and restart, thereby avoiding the need for nitrogen flushing.

For example, when the hydrogen production is stopped, at least a portion of purified hydrogen gas through the cathode electrolyte return conduit and at least a portion of purified oxygen gas through the anode electrolyte return conduit is continued for a predetermined time after having stopped the hydrogen production. This keeps the

DK 182028 B1 18 contamination levels down and the system ready for a quick restart within the prede- termined time. If the time is longer, the recirculation is stopped.

For starting hydrogen production, at least a portion of purified hydrogen gas is recir- culated through the cathode electrolyte return conduit and at least a portion of purified oxygen gas is recirculated through the anode electrolyte return conduit.

This recirculation prior to start, such as restart, can be done for a predetermined length of time, which is determined on the basis of experience for time that is sufficiently long to avoid explosion risk

Alternatively, or in addition, the method comprises measuring oxygen contamination levels in hydrogen gas coming from the cathode electrolyte return conduit and hydro- gen contamination in oxygen coming from the anode electrolyte return conduit, and continuing the recirculation until the hydrogen and oxygen contamination levels are below predetermined levels, and only then starting hydrogen production by the elec- trolyzer.

This new flushing arrangement achieved with the invention has several advantages.

First, the need for nitrogen flushing may be significantly reduced or even eliminated, leading to obvious economic benefits. Second, contamination with nitrogen of the hy- drogen produced during startup is avoided, preventing the need for venting contami- nated hydrogen to the atmosphere, again leading to obvious economic benefits.

On a higher level of aggregation, the preference of other electrolyser technologies such as PEM or SOEC, due to their capability to operate at lower load level than alka- line electrolysers and in some cases their faster startup time, is no longer justified, and the higher costs and other disadvantages, such as the need for rare metal or the need for high operational temperatures, can be avoided.

When selecting the method of operation, the designer of an alkaline electrolyser ac- cording to the invention will have a range of options available for normal operation.

DK 182028 B1 19

In preferred embodiment, at least one gas recirculation pump or compressor is oper- ated at constant speed through the full load range, delivering an approximately con- stant flow rate.

The flow rate of at least one gas recirculation system may be selected to cause the value of the recirculation percentage C to be lower than the value of the threshold per- centage X. In this case, the electrolyser will need to be shut down when the contami- nation level reaches the threshold percentage X as determined either by calculation or by measurement of the contamination level in or at the outlet of the electrolyte return conduit or in the separation tank.

The flow rate of at least one gas recirculation system may alternatively be selected to cause the value of the recirculation percentage C to be equal to or higher than the value of the threshold percentage X. In this case, there is no reason to shut down the electrolyser at any load level.

In another preferred embodiment, at least one gas recirculation pump or compressor is shut down if the calculated or measured contamination level is below a certain limit.

This limit may be equal to the threshold percentage X, or it may be lower than the threshold percentage X.

In yet another preferred embodiment, at least one gas recirculation pump or compres- sor is operated at variable speed, delivering a variable flow rate. The speed may ad- vantageously be varied so as to keep the calculated or measured contamination level at or below a certain limit. This limit may be equal to the threshold percentage X, or it may be lower than the threshold percentage X.

Permutations of the above methods for operation of an electrolyser according to the invention may be selected to achieve particular operational targets desired by the de- signer.

Similarly, when selecting the method of operation, the designer of an alkaline electro- lyser according to the invention will have a range of options available for startup and shutdown.

DK 182028 B1 20

In one preferred embodiment, at least one gas recirculation pump or compressor is op- erated during the entire period of shutdown.

In another preferred embodiment, the operation of at least one gas recirculation pump or compressor is discontinued at shutdown or at a later time determined by a time set- ting or by the calculated or measured level of contamination.

In yet another preferred embodiment, the flow rate of at least one gas recirculation pump or compressor is increased from zero or from an idling level at or prior to startup. The flow rate may be selected as a function of the calculated or measured level of contamination.

In yet another preferred embodiment, the electrolyser load is gradually ramped up as a function of the calculated or measured level of contamination.

Permutations of the above methods for operation of an electrolyser according to the invention at startup or shutdown may be selected to achieve particular operational tar- gets desired by the designer.

SHORT DESCRIPTION OF THE DRAWINGS

The invention will be explained in more detail with reference to the drawings, where. — FIG. I Ais a sketch of an electrolytic cell in a gap configuration; — FIG. 1B is a sketch of an electrolytic cell in a zero-gap configuration; — FIG. 1 Cis a sketch of an electrolyser stack comprising a series of electrolytic cells in zero-gap configuration; — FIG 2A is an exploded view of an electrolytic cell; — FIG 2B is a view of the assembled electrolytic cell; — FIG 3A is a sectional view of an electrolyser stack according to the state of the art;

DK 182028 B1 21 — FIG 3B is the same sectional view of an electrolyser stack according to the state of the art, showing the inwards flow of the electrolyte and the outwards flow of the produced gas; — FIG 4A shows the arrangement of an alkaline electrolyzer according to the state of the art, using the cathode side with the production of hydrogen as example; and — FIG 4B shows the arrangement of an alkaline electrolyser according to the inven- tion, again using the cathode side with the production of hydrogen as example; — FIG. 5 illustrates a practical example of an alkaline electrolyser stack according to the invention.

DETAILED DESCRIPTION / PREFERRED EMBODIMENT

FIG 1 A, B, and C show principle sketches of general variants of alkaline electrolytic cells and an electrolyser stack.

FIG. I A is a sketch of an electrolytic cell I in a gap configuration, comprising a cath- ode 2, an anode 3, and an ion-transporting separator 4. Hydrogen gas 8 is produced at the side of the cathode 2 facing the separator 4 in the cathode compartment 5, and ox- ygen gas 9 is produced at the side of the anode 3 facing the separator 4 in the anode compartment 6. A power supply 7 drives the electrolytic process.

FIG. 1B is a sketch of an electrolytic cell 1 in a zero-gap configuration, comprising a porous cathode 12, a porous anode 13, and an ion-transporting separator 4. The cath- ode compartment 5 and the anode compartment 6 are contained within metal separator plates 14. Electrical connections are provided between the electrodes 12, 13 and the metal separator plates 14 with conductive elements, not detailed in the sketch. Hydro- gen is produced at the side of the porous cathode 12 facing the separator 4 and is con- veyed to the cathode compartment 5 through holes or pores 16 in the cathode 12, and oxygen is produced at the side of the porous anode 13 facing the separator 4 and is conveyed to the anode compartment 6 through holes or pores 16 in the anode 13.

FIG. 1 C is a sketch of an electrolyser stack 17, comprising a series of electrolytic cells 1 in zero-gap configuration of the type as illustrated in FIG. 1B.

DK 182028 B1 22

FIG. 2 A and B illustrate details of a practical example of an electrolytic cell for an al- kaline electrolyser.

FIG 2A is an exploded view of an electrolytic cell 1. The cell 1 comprises a first solid metallic plate 14A, a porous cathode 12 electrically connected to the first solid metal- lic plate 14A with connectors 15, a cathode gasket 18, a separator 4, an anode gasket 19, a porous anode 13 and a second solid metallic plate 14B electrically connected to the anode with connectors. The solid metallic plates 14A and 14B and the gaskets 18 and 19 are penetrated with holes 21A, 21B forming the cathodic and anodic electro- lyte feed conduits along the stack, respectively, and with other holes 22A, 22B form- ing the cathodic and anodic electrolyte return conduits 28 along the stack, respec- tively.

The general term “electrolyte return conduit” and the specific terms “anode electrolyte return conduit” and “cathode electrolyte return conduit” are used interchangeably in the following, as it also appears from the context.

The gaskets 18 and 19 are fitted with pathways 23A and 23B functioning as electro- lyte feed inlets for the cathode compartment 5 and the anode compartment 6, respec- tively, connecting them with their respective holes 21A and 21B, and with pathways 24A and 24B functioning as electrolyte gas outlets from the cathode compartment 5 and the anode compartment 6, respectively connecting them with their respective holes 22A and 22B.

FIG 2B is a view of the assembled electrolytic cell, with a section (left) cut through an electrolyte feed conduit and an electrolyte return conduit, and a complete view (right) of the cell. The cathode compartment 5 is contained within the first solid metallic plate 14A, the cathode gasket 18, and the separator 4, and the anode compartment 5 is contained within the second solid metallic plate 14B, the anode gasket 19, and the sep- arator 4.

FIG. 3 A and B illustrate a practical example of an alkaline electrolyser stack 17 ac- cording to the state of the art, assembled from electrolytic cells 1.

DK 182028 B1 23

FIG 3A is a sectional view of an electrolyser stack 17 according to the state of the art.

The stack 17 comprises ten electrolytic cells 1 sandwiched between a first endplate 25 and a second endplate 26. The first end plate 25 has holes corresponding to the same holes forming the electrolyte feed conduits 27 in the electrolytic cells 1, and holes 22A, 22B corresponding to the same holes forming the electrolyte return conduits 28 in the electrolytic cells 1. The second end plate 26 has no holes.

The holes 21A, 21B passing through the first end plate 25, and corresponding holes through all the solid metallic plates 14A, 14B and all the gaskets 18 and 19 of the electrolytic cells 1 join to form the electrolyte feed conduits 27.

The holes 22A, 22B passing through the first end plate 25, and corresponding holes through all the solid metallic plates 14 and all the gaskets 18 and 19 of the electrolytic cells 1 join to form the electrolyte return conduits 28.

FIG 3B is the same sectional view of an electrolyser stack 17, showing the inwards flow 29 of the electrolyte and the outwards flow 30 of the produced gas. In the exam- ple shown, the anode compartments 6 are fed with the inwards flow 29 of the electro- lyte, and the outwards flow 30 of the electrolyte and gas mainly comprises oxygen with a certain contamination of hydrogen.

FIG 4A illustrates the general arrangement of an alkaline electrolyser according to the state of the art, and FIG. 4B an alkaline electrolyzer according to the invention.

FIG 4A shows the arrangement of an alkaline electrolyzer according to the state of the art, using the cathode side with the production of hydrogen as example. The anode side would be likewise, and it is not shown separately.

The electrolyser has a stack 17 that comprises a series of electrolytic cells 1 sand- wiched between a first endplate 25 and a second endplate 26. Hydrogen produced in the stack is fed, together with some electrolyte, through the electrolyte return conduit (illustrated as 28 in FIG. 5), which at its downstream side 40 is connected to the sepa- ration tank 31. Here, the hydrogen is separated from the electrolyte. The hydrogen is

DK 182028 B1 24 then fed to a gas cleaning system 32 and from there to the purifier 33 before it is sup- plied to the end user 34 or to a collection tank. The electrolyte is fed back through conduit 27A to the electrolyte feed conduit ( shown as 27 in FIG. 5) by means of the electrolyte pump 35. Feed water is provided from a feed water reservoir 36 by means of a water pump 37. Power to the electrolysis process is provided by the power supply 7.

The anode side of this prior art system with the production of oxygen has the same general arrangement as on the cathode side, with the exception that, if the oxygen is merely vented to the atmosphere and not provided to any end user, the purification system is sometimes omitted.

FIG 4B shows the arrangement of an alkaline electrolyser according to the invention, again using the cathode side with the production of hydrogen as example.

The arrangement is generally the same as for an alkaline electrolyser according to the state of the art, which was illustrated in FIG. 4A, however, with the exception that a gas recirculation system 38 has been added, connecting the outlet of the hydrogen gas purifier 33 to the second end plate 26, which has now been fitted with holes (not shown) corresponding to the same holes (shown as 22A, 22B in FIG. 3) forming the electrolyte return conduits 28 (shown in FIG. 5) in the electrolytic cells. The gas recir- culation system 38 is fitted with means 39 for recirculating the gas, e.g., a gas pump or compressor. The gas is recirculated to holes in the endplate 26 and into an upstream end 41 of the electrolyte return conduits (shown as 28 in FIG. 5).

The anode side with the production of oxygen has the same general arrangement as on the cathode side. In contrast to the prior art, when implementing the arrangement ac- cording to the invention, the purifier cannot be omitted, even if the oxygen is merely vented to the atmosphere and not provided to any end user. This is so because at least some of the purified oxygen is recirculated into the anode electrolyte return conduit 28.

FIG 5 illustrates a practical example of an alkaline electrolyser stack 17 according to the invention. Whereas the second endplate 26 may have no holes in an alkaline

DK 182028 B1 25 electrolyser stack 17 according to the state of the art, as illustrated in FIG. 3, this em- bodiment according to the invention, as illustrated in FIG. 5, has holes at an upstream end 41 of the electrolyte return conduits 28 corresponding to the holes 22A, 22B in the first endplate 25 that are forming the electrolyte return conduits 28, as illustrate in

FIG. 3. The inwards flow of the electrolyte 29 to the electrolyte feed conduits 27 and the outwards flow 30 of the produced gas from the electrolyte return conduits 28 are similar to the arrangements for the alkaline electrolyser according to the state of the art, but in addition, recirculated flow 42 of clean gas is now provided to the electrolyte return conduits 28 through the holes at the upstream end 41 of the electrolyte return conduits 28 by the gas recirculating system 38 (as shown in FIG. 4B).

Claims

DK 182028 B1 26 REQUIREMENTS

1. Alkaline electrolyzer for producing hydrogen gas (8) and comprising a stack (17) of electrolytic cells (1), each electrolytic cell (1) comprising a cathode chamber (5) and an anode chamber (6) containing liquid alkaline electrolyte and separated by an ion-conducting separator (4), each of the cathode chambers (5) comprising a cathode gas outlet into a cathode-electrolyte return line (28) and each of the anode chambers (6) comprising an anode gas outlet into an anode-electrolyte return line (28), the electrolyzer comprising at least one of A and B; wherein in A there is a hydrogen purifier (33) connected to a downstream end (40) of the cathode electrolyte return line (28) and configured to provide purified hydrogen gas by removing oxygen from the gas received from the downstream end (40) of the cathode electrolyte return line (28), wherein in B there is an oxygen purifier connected to a downstream end of the anode electrolyte return line (28) and configured to provide purified oxygen gas by removing hydrogen from the gas received from the downstream end of the anode electrolyte return line (28), characterized in that in A there is a cathode gas recirculation system (38) connecting a downstream end of the hydrogen purifier (33) to an upstream end (41) of the cathode electrolyte return line (28) to supply purified hydrogen gas to the cathode electrolyte return line (28); wherein B is an anode gas recirculation system connecting a downstream end of the oxygen purifier to an upstream end of the anode electrolyte return line (28) to supply purified oxygen gas to the anode electrolyte return line (28).

2. An alkaline electrolyzer according to claim 1, wherein the electrolyzer comprises both the hydrogen purifier (33) and the oxygen purifier and the corresponding cathode gas recirculation system (38) and anode gas recirculation system.

3. Alkaline electrolyzer according to claim 1 or 2, wherein the electrolyzer comprises a control system capable of controlling the operation of the gas recirculation system by varying parameters of volume flow or operating time or both as a function of parameters measured on the electrolysis system, including at least one of load, operating time and degree of contamination of the produced gases.

DK 182028 B1 27

4. An alkaline electrolyzer according to any preceding claim, wherein the electrolyzer comprises the cathode gas recirculation system, and the cathode gas recirculation system is configured to recirculate at least a portion of the purified hydrogen gas to the upstream end (41) of the cathode electrolyte return line (28), wherein the electrolyzer also comprises a gas analysis system configured to measure oxygen contamination levels in the hydrogen gas coming from the outlet of the cathode electrolyte return line (28), and wherein the cathode gas recirculation system is configured to operate at a variable flow rate, adjusted in dependence on the measured oxygen contamination levels.

5. The alkaline electrolyzer of any preceding claim, wherein the electrolyzer comprises the anode gas recirculation system, and the anode gas recirculation system is configured to recirculate at least a portion of the purified oxygen gas to the upstream end of the anode electrolyte return line (28), wherein the electrolyzer also comprises a gas analysis system configured to measure hydrogen contamination levels in the oxygen gas coming from the outlet of the anode electrolyte return line (28), and wherein the anode gas recirculation system is configured to operate at a variable flow rate, adjusted in dependence on the measured hydrogen contamination levels.

6. A method of operating an electrolyzer according to any one of the preceding claims, wherein the method of operating the electrolyzer with A comprises providing cathode gas recirculation through the cathode electrolyte return line (28) and the hydrogen purifier (33), wherein the cathode gas recirculation comprises the hydrogen purifier (33) receiving gas from the downstream end (40) of the cathode electrolyte return line (28) and providing purified hydrogen gas by removing oxygen by the hydrogen purifier (33) and supplying at least a portion of the purified hydrogen gas to the upstream end (41) of the cathode electrolyte return line (28); wherein the method of operating the electrolyzer with B comprises providing anode gas recirculation through the anode electrolyte return line (28) and the oxygen purifier, wherein the anode gas recirculation comprises the oxygen purifier receiving gas from the downstream end of the anode electrolyte return line (28) and providing purified oxygen gas by removing hydrogen by the oxygen purifier, and supplying at least a portion of the purified oxygen gas to the upstream end of the anode electrolyte return line (28).

DK 182028 B1 28

The method of claim 6, wherein the method comprises both A and B.

A method according to claim 6 or 7, wherein the method comprises recirculating at least a portion of the purified gas through the corresponding electrolyte return line (28) during operation of the electrolyzer for hydrogen production.

9. The method of claim 8, wherein the method comprises recirculating at least a portion of the purified gas through the corresponding electrolyte return line (28) during operation of the electrolyzer for hydrogen production at a partial load below a nominal load, wherein the partial load is less than 25% of the nominal load.

10. The method of claim 9, wherein the method comprises operating the electrolyzer for hydrogen production at a partial load that is less than 10% of the nominal load.

11. The method of any one of claims 6-10, wherein the method comprises stopping hydrogen production; and wherein the method of A comprises recycling at least a portion of purified hydrogen gas through the cathode electrolyte return line (28) for a predetermined time after stopping hydrogen production; and wherein the method of B comprises recycling at least a portion of purified oxygen gas through the anode electrolyte return line (28) for a predetermined time after stopping hydrogen production.

12. The method of any one of claims 6-11, wherein the method comprises A and further comprises recirculating at least a portion of purified hydrogen gas through the cathode-electrolyte return line (28) before commencing hydrogen production, and measuring oxygen contamination levels in hydrogen gas coming from the cathode-electrolyte return line (28), and continuing the recirculation until the oxygen contamination levels are below predetermined levels, and only then commencing hydrogen production by the electrolyzer.

13. The method of any one of claims 6-12, wherein the method comprises B and further comprises recycling at least a portion of purified oxygen gas through the anode-electrolyte return line (28) prior to commencing hydrogen production, and

DK 182028 B1 29 painting hydrogen contamination in oxygen coming from the anode electrolyte return line (28), and continuing the recirculation until the hydrogen contamination levels are below predetermined levels, and only then starting hydrogen production by the electrolyzer.

14. The method of any one of claims 6-13, wherein the method of A comprises recirculating at least a portion of purified hydrogen gas through the cathode-electrolyte return line (28) and continuing the recirculation for a predetermined time before starting hydrogen production; and wherein the method of B comprises recirculating at least a portion of purified oxygen gas through the anode-electrolyte return line (28) and continuing the recirculation for a predetermined time before starting hydrogen production.

15. The method of any one of claims 6-14, wherein the hydrogen purifier (33) in A comprises a cathode gas purifier (33) containing a catalyst for catalytic recombination of oxygen and hydrogen to form water, and wherein the method comprises purifying hydrogen gas coming from the cathode-electrolyte return line (28) with the cathode gas purifier (33); wherein the oxygen purifier in B comprises an anode gas purifier containing a catalyst for catalytic recombination of oxygen and hydrogen to form water, and wherein the method comprises purifying oxygen gas coming from the anode-electrolyte return line (28) with the anode gas purifier.

16. The method of any one of claims 6-15, wherein the method of A comprises recycling at least a portion of the purified hydrogen gas from the hydrogen purifier (33) to the cathode electrolyte return line (28) and measuring oxygen contamination levels in hydrogen gas coming from the cathode electrolyte return line (28), and only operating the hydrogen gas recycling system if the measured oxygen contamination levels are above the contamination thresholds; wherein the method of B comprises recycling at least a portion of the purified oxygen gas from the oxygen purifier to the anode electrolyte return lines (28) and measuring hydrogen contamination in oxygen coming from the anode electrolyte return line (28), and only operating the oxygen gas recycling system if the measured hydrogen contamination levels are above the contamination thresholds.

17. The method of any one of claims 6-16, wherein the method in A comprises recycling at least a portion of the purified hydrogen gas from the hydrogen purifier (33) to the cathode electrolyte return line and measuring oxygen contamination levels in the hydrogen gas,

DK 182028 B1 30 coming from the cathode electrolyte return line (28), and operating the hydrogen gas recirculation system at variable flow rates, adjusted in dependence on the measured oxygen contamination levels, where different flow rates are used for different oxygen contamination levels; wherein the method in B comprises recirculating at least a portion of the purified oxygen gas from the oxygen purifier to the anode electrolyte return lines and measuring hydrogen contamination in oxygen coming from the anode electrolyte return line (28), and operating the oxygen gas recirculation system at variable flow rates, adjusted in dependence on the measured hydrogen contamination levels, where different flow rates are used for different hydrogen contamination levels.

18. A method according to any one of claims 6-17, wherein the method comprises operating the electrolyzer with an electrolyte having a temperature in the range of 50-90°C.