Given the number of components in an electrolyser system, and the multiple options for each component, there are thousands of degrees of freedom that an electrolyser OEM can leverage to optimise their product for cost, efficiency, dynamic response, or durability.

Only the best electrolyser OEMs, using the best components, are able to out-perform their rivals on multiple performance attributes. Component-level choices matter. And on top of that, there must be a cohesive concept which integrates the holistic design, from feed water entering the system to pure, pressurised hydrogen leaving it.

Technology classes

There are four broad categories of electrolyser technology to split water into hydrogen and oxygen

.

1. Alkaline water electrolysers (AWE) operate using an electrolyte which contains about 30% by mass of potassium hydroxide (KOH). This is referred to as lye and is an extremely alkaline medium.

2. Anion exchange membrane (AEM) electrolysers also use an alkaline electrolyte. In comparison, to the AWE technology, the AEM electrolyte will either have a lower concentration of KOH or use a benign bicarbonate solution.

3. Proton exchange membrane (PEM) electrolysers use pure water at the electrolyte. During electrolysis, an acidic environment is created. This is a key differentiator between PEM electrolysers and the AWE and AEM classifications.

4. Solid oxide electrolyser cells (SOECs) use a solid electrolyte. Their design and operation is vastly different to AWE, AEM and PEM systems, which rely on liquid electrolytes. Furthermore, SOECs generally operate in the temperature range 700 to 900 °C – significantly higher than AWE, AEM and PEM technologies which operate in the range of 60 to 120 °C.

Acidity is unavoidable in PEM systems

Materials selection for AWE and AEM electrolysers operating in an alkaline liquid is different to the case for PEM. Additionally, the AWE and AEM stacks are more robust than PEM stacks meaning they are more tolerant of some impurities in the electrolyte.

The PEM membrane must be manufactured from an extremely resilient polymer to avoid it being damaged by the acidity surrounding it. A fluorinated, sulphonated polymer called perfluorosulfonic acid (PFSA) is the default choice to achieve this goal.

Using a fluorinated polymer is a double-edged sword. On the one side, the fluorine makes the polymer robust. On the other side, the acidity in the PEM stack is so extreme that some fluorine is released from the PFSA membrane as hydrofluoric acid (HF). HF is one of the most aggressive acids known to chemistry.

Over time, HF attacks the highly sensitive PEM electrolyser’s proton exchange membrane, which is intentionally very thin to reduce ohmic losses

.

Durable PEM membranes

In essence, a PEM stack with a PFSA membrane is designed to destroy itself. The challenge is to balance the design to achieve several years of stable operation at high efficiency before the stack deteriorates to such an extent that the electricity input required to generate hydrogen rises to such an amount that re-investment in new stacks is cost-effective.

Avoidance of HF would be possible if an alternative membrane were to be used. Selection of a membrane that avoids fluoropolymers would also reduce concerns related to ‘PFAS’: per- and polyfluoroalkyl substances which are known as ‘forever-chemicals’ which are harmful to the environment and extremely difficult to break down in nature.

Toray, and several other companies are working to develop fluorine-free, PFAS-free PEM membranes. However, at present, the state of the art relies on PFSA, due to its resilience.

Feed water purification

The feed water to the electrolyser must be purified to an extremely low level of ionic conductivity. The purification equipment for this can include ultrafiltration, reverse osmosis, electro-deionisation and ion exchange resin beds.

However, even the most robust purifiers allow a few parts per billion of contaminant ions such as calcium, magnesium and iron being introduced to the electrolyser. The presence of iron oxide particles in the size range of 0.1-5 µm in the electrolyser recirculation water can be observed through ‘rouging’. Rouging is caused when pure water at around 60 °C comes in contact with stainless steel.

Some of these metallic ion impurities are very damaging to the membrane. Big cations such as calcium and magnesium are membrane killers because they block the proton transfer ‘pores’ in membrane.

Blockages in the proton exchange membrane lead to a progressive increase in ohmic resistance. In turn, that means a reduction in the stack efficiency and an increase in electricity consumption per kg of hydrogen produced. This is a core reason for PEM stack degradation.

Limitations of stainless steel

Selection of stainless steel for water purification system and PEM electrolyser pipework is common because it is strong and releases fewer ions into the water than carbon steel. However, even the highest grades of stainless steel can release metallic ions into the feed water and circulating water.

Specification of drawn stainless steel tubing with internally polished surfaces is essential if this material is used in PEM electrolysers and ultra-pure water treatment systems. This is potentially a limitation to the scale up of PEM electrolyser stacks because this type of pipework is only available at smaller diameters.

The use of large-diameter stainless steel pipe, fabricated with a welded seam would be fatal for a PEM electrolyser. At a microscopic level, the weld is rough and porous and may be a catastrophic source of ionic contamination. Similarly, when joining stainless steel piping sections, welds must be avoided.

The case for plastic pipework

An alternative approach to minimise the potential for damage to the membrane is to minimise the use of metallic pipework in contact with the feed water and ultra-pure water in the electrolyser recirculation circuit. With fewer impurities reaching the membrane and catalysts, the stack degradation rate is reduced and its usable lifetime is extended.

Similar to the many grades of steel and stainless steel that might be selected, there are many types of plastic that are used to construct pipes. Not all of these are suitable for use in PEM electrolysers. Only appropriate, sophisticated engineering plastics should be used.

Engineering plastics

There are many different plastics that we encounter in our daily lives. Polyethylene (PE) is commonly used to make bags and food packaging. Polypropylene (PP) is denser and is preferred for car parts and sturdy children’s toys such as a garden slide. Applications of Polyvinylchloride (PVC) range from construction materials for window frames to medical devices.

Engineering plastics differ from household plastics due to their enhanced properties of temperature, pressure or chemical resistance. When designing a PEM electrolyser, superior performance in all three of these attributes is essential.

PP and PE are used in ambient temperature applications such as water purification systems and cooling water circuits in electrolysers.

For PEM water circulation, where the temperature is around 60 °C, we select premium grades, such as PP-H. PP-H is constructed from a homopolymer material, and the manufacturing process ensures very smooth internal wall surfaces and ensures virtually no extractable ionic or particulate contamination is present.

Moving beyond the water treatment system, PVDF and ECTFE are highly stable fluoropolymer materials which are suitable for use in the condensate water return line from the oxygen phase separator. They are non-corrosive and therefore minimise iron ion release and avoid rouging of the recirculation water and damage to the stack.

PVDF is supplied to highly sensitive applications in the pharmaceutical, biotechnology and semiconductor sectors for many years. It is specially cleaned to be oil- and grease-free. At present, we are experiencing a high degree of interest from electrolyser OEMs to use these materials in PEM systems at pressures of up to around 10 bar.

PVDF-HP fittings are generally manufactured in a ISO Class 5 cleanroom: the same category that is used to manufacture less-critical semiconductor components.

Pressure limitations

Many PEM electrolysers produce hydrogen at an elevated pressure between 20 and 40 bar. A notable exception to this is Siemens Energy’s Elyzer-P which operates close to atmospheric pressure. The oxygen side of PEM systems can either be at a low pressure in the order of 3 bar, or at an elevated pressure similar to the hydrogen side of the system.

There is a trade-off between temperature and pressure when using thermoplastic plastic components: at higher temperatures, the maximum safe working pressure is reduced. PEM electrolysers generally operate at around 60 °C and the maximum pressure rating of PP is reduced by half at this temperature compared to the allowable pressure for operation at 20 °C.

Similar to PVDF, Ethylene Chlorotrifluoroethylene (ECTFE) is a thermoplastic with excellent chemical resistance. ECTFE has some advantages over PVDF, namely it is less susceptible to chemical stress and can be used at higher temperatures. These characteristics make it ideal for electrolyser piping circuits in PEM electrolyser systems.

Gases, liquids and oxygen compatibility

In PEM electrolysers plastics are generally used to transport liquids, not gases. A primary application is in the feed water purification and conditioning area where the water is purified to remove particles, dissolved organic compounds, gases and ions.

In addition to selection of appropriate engineering plastics, these components must be cleaned to remove any traces of oils or grease that may have been used during manufacturing. For valves, this is especially important since it would not be appropriate for lubricants that might be used on the moving parts of the valve stem to enter the electrolyser feed water.

The phase separators which allow the oxygen and hydrogen gases to be removed from the recirculating water are generally manufactured from stainless steel. They may need to be certified pressure vessels, and the use of plastics in this instance may not be appropriate.

Downstream of the phase separator, some PEM and AEM electrolysers are being built with plastic pipes and fittings in the condensate water return lines from the phase separator to the stack.

Downstream of the phase separators stainless steel is generally specified for pipework and valves in contact with the oxygen and hydrogen gases.

Stainless steel can easily be cleaned to be compatible with oxygen gas. Furthermore, it is non-flammable and can be constructed to tolerate very high pressures.

The main cost driver for green and blue fertilizers is the increased production cost of clean hydrogen. However, as the amount of hydrogen in the fertilizer decreases, the impact of the higher cost for green or blue hydrogen diminishes. The reason is that the downstream stages to produce ammonia and ammonia derivatives remain un-altered, whatever colour hydrogen they are fed with.

Ammonia derivatives dilute the clean premium

Added-value ammonia derivatives such as urea, ammonium sulphate (AS), diammonium phosphate (DAP), ammonium nitrate (AN) and calcium ammonium nitrate (CAN) dilute the green or blue hydrogen cost to reduce the ‘clean premium’ to a more affordable level.

Hydrogen is 17.8% of the mass of ammonia, with nitrogen being the remainder. The hydrogen content in AN is reduced to 5%. And as the AN is converted to CAN, hydrogen falls to only 3.5% of the total mass.

As the mass of hydrogen in ammonia derivatives decreases, so does its cost within the product. In the case of DAP and CAN, hydrogen represents around 41% of the product value, whereas in ammonia it is 73%.

Additionally, DAP and CAN are inert solid fertilizer granules which can be transported more easily than liquid ammonia, which is both toxic and flammable. For safety reasons, the use of AN as a fertilizer is prohibited, driving a high-value regulated European market for CAN.

Adding value to clean ammonia through ammonia derivatives significantly improves the business case for clean fertilizer value chains.

Air gases for ammonia derivatives

The Villeta Green Fertilizer Project is the flagship development of ATOME PLC. Villeta is an integrated complex, where green hydrogen will be produced using 110MW of electrolyser capacity fed with renewable electricity from the Itaipu dam.

Additionally, cryogenic air separation will produce gaseous nitrogen from the source of renewable electricity. This will be reacted with the green hydrogen in an ammonia synthesis plant. Conversion of the ammonia to nitric acid and the ammonia derivatives AN and CAN will also take place at the 30-hectare site.

In some nitric acid plants, oxygen facilitates process intensification in the absorber, where gaseous oxides of nitrogen are dissolved in dilute nitric acid to enrich the acid to the required commercial concentration.

In integrated sites where an air separation unit is included in the scheme to generate nitrogen for the ammonia production, or an electrolyser is present to generate hydrogen for ammonia synthesis, oxygen can beneficially be recovered as a co-product from these unit operations.

CAN and DAP production

Located in Puerto Sara, approximately 50 km south of Asunción, Villeta sits close to the east bank of the Paraguay River close to the Paraguay-Paraná waterway. This connects the site to several MERCOSUR countries through a regional river transport system.

Villeta is designed to produce CAN and will require circa 170 tonnes per day of dolomite. This rock will be sourced from local mines, integrating the facility into the Paraguayan economy.

Following Villeta, ATOME is planning a further ammonia derivatives development in Paraguay. The 300 MW Yguazu project will diversify their product scope to DAP.

ATOME’s Paraguayan projects are designed to reduce fertilizer imports to the MERCOSUR region and replace grey ammonia production from fossil fuel reforming with green ammonia production from 100% renewable electricity.

By producing green fertilizer at cost parity with the landed cost of gas-based grey hydrogen in South America, ATOME’s projects in Paraguay serve climate goals and address the needs of the region’s agricultural sector. And, by producing exportable added-value fertilizers such as CAN, the project will produce an internationally competitive ammonia derivative.

Bankability above all

The economic viability of green hydrogen production is dependent on the availability and cost of renewable electricity, which constitutes 60-70% of the production cost. Unlike many green hydrogen projects that rely on non-programmable wind or solar power, Villeta will utilise the 14 GW Itaipu dam as a firm supply of green electricity.

The Villeta project will utilise Casale’s Flexigreen® technology, which is optimized for renewable electricity-fed ammonia plants. While the hydropower source is remarkably firm, the Flexigreen® design allows the process to turn down to 10% of nominal capacity. This allows the dam to supply other customers during periods of peak demand and ensures that the Villeta plant will be able to adapt to any potential future connection to integrated wind and solar renewable energy sources.

To de-risk the construction phase, in April 2025, ATOME and Casale signed a fixed-price, lump-sum EPC contract valued at $465 million. The contract guarantees that production will commence within 38 months of the construction start date. The use of proven technologies and the experience that Casale has in the CAN value chain maximises the bankability for equity investors and creditors.

A significant commercial milestone for Villeta was closure of a 10-year offtake agreement with Yara International. The agreement, finalised in September 2025, commits Yara to purchase the entire 260,000 tonne-per-year CAN production. This “take-or-pay” structure will fully cover the costs of energy, operation and capital depreciation. It therefore serves as the foundation for the project’s bankability.

Green hydrogen and CCS to combat CBAM tariffs

A benefit of producing green fertilizers is that they have a very low CO2 intensity and are therefore not subject to the carbon border adjustment mechanism (CBAM) when they are imported to Europe. Blue ammonia and ammonia derivatives will have a similarly competitive position.

Due to the CBAM, some of the additional costs incurred when producing reduced CO2 intensity products are neutralised once the product is landed in the EU and all taxes and tariffs have been paid. This supports the international competitiveness of decarbonised products and can help to build a business case for added-value clean ammonia and ammonia derivatives production.

When ammonia is made from steam methane reforming of natural gas, CO2 leaving the reformer must be removed prior to the catalytic Haber Bosch ammonia synthesis reaction.

The active ammonia synthesis catalyst is iron. All molecules containing oxygen, such as water, carbon monoxide and CO2 must be removed from the syngas before it is fed to the ammonia synthesis loop. Otherwise, the iron catalyst is oxidised and becomes ineffective. Therefore, every natural gas-fed ammonia facility integrates a CO2 capture plant.

In some facilities, about 60% of the captured CO2 is combined with ammonia to make urea. However, the residual 40% of the CO2 from the SMR (and all of the CO2 from ammonia plants that do produce urea) can be permanently sequestered.

Permanent CO2 storage reduces the CO2 intensity of their downstream nitrogen fertilizer products, but add only a marginal incremental cost for CO2 conditioning, transportation and geological storage.

Petrobras has operated geological CO2 injection schemes linked to enhanced oil recovery for many years in the Santos Basin. They have sequestered more tonnes of CO2 offshore than any other operator worldwide.

So, in addition to green ammonia being produced in South America, the continent has the potential to produce blue molecules also.

DAP and AS integration for value-added ammonia derivatives

CO2 capture from DAP production will be implemented from 2027 at OCP’s Jorf Lasfar industrial platform in Morocco. The initiative is part of OCP Group’s decarbonization roadmap, which aims to achieve carbon neutrality across scope 1, 2 and 3 emissions by 2040. It also addresses international competitiveness by reducing the CO2 intensity of DAP.

The captured CO2 will be used by OCP Green Water for pH adjustment and re-mineralisation of drinking water in support of a local reverse osmosis desalination plant. This aligns to Morocco’s goal of meeting all of the nation’s water needs through unconventional water sources by 2027.

As an additional use case for the captured CO2, INNOVX will build a facility that combines captured CO2 and waste phosphogypsum to produce AS.

The Merseburg Process is the most common pathway to achieve this: it reacts aqueous ammonia with CO2 to produce ammonium carbonate. Phosphogypsum is then reacted with ammonium carbonate to yield calcium carbonate and the target product ammonium sulphate.

Calcium carbonate is a recognised permanent sink for CO2, so this pathway sequesters CO2 emissions from the DAP process through mineralisation and enables integrated production of DAP and AS.

AS – ideal for sub-tropical climates

AS was the first commercially available nitrogen fertilizer but it was displaced in many markets by ammonia and urea because they have a higher nitrogen content by mass.

However, AS is still used extensively in Brazil where it is suitable for the local soil, which contains very low levels of sulphur. Sulphur is essential to grow protein-rich crops such as soya beans, which are used locally for beef production and are exported.

Additionally, AS has a lower ammonia volatilisation rate than urea. This is important in the warm Brazilian climate and is a factor that leads to its use in South East Asia for growing wheat and rice.

The technologies that will bring fresh water to electrolysis schemes are exactly those that are relied on today to make potable water available in arid locations. Whilst the technologies are known, they must be adapted to make fresh water available for green hydrogen projects.

Consistent quality and reliable supply

With freshwater feed, the capital cost of the equipment for ultrapure water supply for a typical green hydrogen project will be around 4 to 8% of the electrolyser scheme. Where desalination is also required, the capital cost of the whole water processing facility may represent 10 to 15% of the total equipment capex for a green hydrogen electrolyser scheme.

To put this level of capital expenditure into context the power management system involving the transformer and rectifier to provide direct current (DC) electricity for the electrolyser would be around 20% of the total project capex. So, water treatment is an essential element of the scheme, but is unlikely to be the dominant fraction of the green hydrogen project budget.

With a freshwater feed, the operating cost of the ultrapure water plant will generally be less than 1% of the total green hydrogen plant opex. This is a combination of dosing chemical costs and electrical power for high pressure water pumps. The cost of the freshwater procurement may be 1 to 2% of the total operating cost if it is withdrawn from a local pipeline supply or brought in by road tanker. With seawater supply, that operating cost may rise to be around 3% of a green hydrogen scheme due to additional complexity.

The capital and operating costs of ultrapure water are a small proportion of the project total, but the consequences of failure can be very high: reliability is therefore key. An interruption to the water supply would mean the electrolyser must shut down. For a PEM system that is unlikely to result in equipment damage. However, for alkaline or solid oxide electrolysers, an unplanned shutdown may result in degradation of the electrolyser’s performance and a reduction in its efficiency.

Impure water can be extremely damaging to the valuable electrolysers. Calcium or magnesium cations in the water will rapidly damage a proton exchange membrane (PEM) electrolyser membrane due to interaction with the water-splitting catalyst that coats the membrane. High temperature solid oxide electrolysers (SOEs) are also sensitive to poisons in the water.

High temperature electrolysis integrated with ammonia production

Ammonia production also requires a vast amount of boiler water to generate superheated steam. The reaction between hydrogen and nitrogen to make ammonia is highly exothermic and the reaction heat is recovered as steam to ensure energy efficiency. For each tonne of ammonia produced, there are approximately 4 tonnes of ultrapure boiler water required.

The equipment that is used to generate this boiler feed water is similar to that required to generate the water for electrolysis and economies of scale and synergies can be achieved by combining the equipment.

Most of the steam is used to drive the compressor that feeds the nitrogen and hydrogen gas mixture to the Haber Bosch ammonia synthesis loop. Additional steam is used to drive the refrigeration compressor that is required to recycle incondensable gases from the ammonia liquefier. It is common to use steam turbines on the drives of these machines to avoid the losses of converting the steam to power. After expansion in these turbines, some lower-grade steam is available for export.

SOE technology is well aligned to green ammonia projects. The reason is related to process integration potential and energy efficiency. Waste heat from the Haber-Bosch ammonia synthesis loop can be used to reduce the electrical power input requirement to the SOE, which operates at high temperatures and is fed with steam, not water. When fed with steam, the SOE requires about 20% less power than a PEM or alkaline electrolyser to generate the same amount of hydrogen.

The catalysts that are used in the cells within an SOE are highly sensitive to sulphate ions as well as silicates, siloxanes, and aluminium oxides. These impurities must be reduced to less than 5 or 10 parts per billion to avoid degradation of the SOE performance.

The typical feed for an electrolyser is referred to as ‘Grade 2’ in the ISO 3696:1987 standard, with a maximum conductivity of 0.1 mS/m – identical to ASTM D1193-06(2018) ‘Type 2’ water. Both the ASTM and ISO standard have Silica specifications, but the permitted amount differs by almost a factor of 10 between the two standards. Also, neither standard explicitly specifies total sulphur. These points highlight the importance of researching the precise ultrapure feed water specification that the electrolyser manufacturer requires and designing a suitable ultrapure water treatment facility.

Primary water purification unit operations

Electrolyser feed water must be prepared in a similar way to boiler feed water. Many thermal power plants withdraw river water or seawater and purify that raw water prior to it being vaporised and introduced as the working fluid for the steam turbines. The higher the pressure of operation, the lower the permitted levels of impurities. However, ultrapure water for electrolysis must be even more pure than boiler water prepared for 30 bar steam generation operation.

In the early stages of water purification various filtration processes are used. Initially, a rough metal grate can be used to trap fish, leaves, logs, or other debris such as floating plastic waste. After this, additives may be used to flocculate dissolved metals. For example, Aluminium Sulphate addition is commonly used. Pre-chlorination can support the flocculation of heavy metals.

A multi-media, or sand bed filter can then be used to trap the flocculated particles. It will also remove sludge, algae mud, and entrained sand from the raw water. This filter must be backwashed periodically, and the resultant water should be purified prior to discharge to the drain, sea or river.

An activated carbon filter can then be used to trap the chlorine which would corrode membranes that are used in subsequent water purification processes and would also damage the PEM electrolyser membrane. Chlorine might have been used in the Prechlorination stage. It could also have been dosed to mains water to avoid microbial contamination.

The activated carbon filter will also remove dissolved organic material. Spent activated carbon filter cartridges are not regenerated in situ, they must be replaced. Often the cartridges are exchanged by a service provider. The saturated activated carbon is generally incinerated.

Following the activated carbon filter, water softening can be used to remove calcium and magnesium ions in hard water areas. These are especially damaging to PEM electrolyser membrane electrode assembly catalysts.

Reverse osmosis (RO) is a core technology for the final stages of freshwater purification and seawater desalination. To purify fresh water, low pressure reverse osmosis (LPRO) can be used. It operates at around 15 bar. The power consumption of an LPRO plant is in the order of 10 kWh per m³ of water purified. Between 70 and 85% of the water is recovered.

Polishing processes

After LPRO to purify the fresh water, intermediate buffer storage is generally implemented. As water is withdrawn from the storage, it may flow through an electro deionisation (EDI) plant for polishing. Here the power requirement is similar to LPRO at 10 kWh/m³ of water. The concentrate discharge from the EDI plant can be fed to the LPRO inlet to recover water.

EDI is a mature technology that has been applied to create demineralised water for supercritical steam generation, for example in concentrated solar plants. It is also used to purify water for direct injection to gas turbines for NOx reduction. Many petrochemical, pharmaceutical and semiconductor facilities also rely on EDI for ultrapure water supply.

After the EDI process, microbial contamination can be removed. Post-chlorination is not a suitable technology, due to the incompatibility of chlorine with the sensitive electrolyser components, such as the polymer electrolyte membrane that is at the heart of the membrane electrode assembly (MEA) in a PEM electrolyser. Therefore, a UV steriliser lamp is generally used to kill any bacteria that may be present.

Following the UV lamp, an ultrafilter traps the dead bacteria. The power requirement for the ultrafilter may be around 3 kWh/m³ of water. The ultrafilter is periodically backwashed to clear the filter media cake. As with other backwash waters, this waste must be treated prior to discharge.

Degassing of the pure water may also be required. Dissolved oxygen, nitrogen, and carbon dioxide (CO2) gases will always be present in the feed water. CO2 is unwelcome in the electrolyser because it is broken down through electrolysis to form carbon monoxide (CO). This will be an impurity in the hydrogen making the product unsuitable for fuel cell applications.

At this stage, the ultrapure water is finally ready to be introduced to the electrolyser water or lye recirculation circuit.

Sea water for electrolysis in arid locations

Many large-scale green hydrogen projects are proposing to use renewable power generation from wind and solar resources. Integration of wind and solar can improve the firmness of renewable power generation. Hybrid schemes of this kind will be optimal in countries such as Southern Chile, western Namibia, Morocco and the east coast of Oman. Western Australia, Northwestern Saudi Arabia, northeastern Egypt and parts of northern and western China are also blessed with ideal conditions for integrated wind and solar power generation.

However, in each of these locations is the rainfall is neither sufficient nor steady enough to provide water for large scale electrolysis. Seawater is the preferred source of water for electrolysis. It requires desalination, in addition to purification. Desalination is used extensively in the Middle East to make water available for the emerging vertical farming sector and coastal cities. The same knowledge can be transferred to electrolytic hydrogen production.

Seawater reverse osmosis (SWRO) is a common desalination process and has largely replaced less efficient thermal desalination. However, where waste heat from power generation can be used for desalination, thermal desalination techniques using mechanical vapour recovery to upgrade steam can be the most economic option.

During SWRO an operating pressure of around 80 bar is used to achieve desalination. Due to the high-pressure operation, the power draw is around 4 times that of the LPRO system used to purify fresh water.

To generate one litre of ultrapure water, approximately three litres of seawater must be extracted and processed. The remaining two litres can be treated to remove any chemical residues and then returned to the sea.

Seawater processing to ultrapure water will be an essential aspect of MW- and GW-scale green hydrogen and green ammonia projects of the future.

Ends 1812 words

Author credit – Stephen B. Harrison, sbh4 consulting

Link to related high-resolution images

sbh4 GmbH has purchased the rights to some royalty free Shutterstocks images in this selection and these rights may be transferred to gasworld.com ltd / H2 View.

Start with the end in mind

To optimize the initial CO2 capture technology, it makes sense to consider the requirement at the very end of the CCTUS value chain. What is the required purity of the CO2? What is the required temperature and pressure condition?

If the goal is to compress supercritical CO2 into a geological storage site, and the transmission is by pipeline, the capture technology can be selected to produce CO2 gas which can be compressed into the pipeline.

On the other hand, if the end-game is utilisation of the CO2 in an e-methanol or Fischer Tropsch e-fuels production process, and the transportation is using liquid CO2 by rail, there is a requirement to produce liquid CO2 at the capture location, or CO2 transfer terminal where liquid CO2 is loaded onto the train.

Robbing Peter to pay Paul

Consider the case that a liquid solvent-based CO2 capture system, such as a twin-tower amine process, is used upstream of a CO2 liquefier to condition the CO2 for transportation by rail and ship to an offshore geological storage facility.

The given boundary conditions are an overall CO2 capture rate of >90% from the emissions source to geological storage and that the liquid CO2 must comply with the Northern Lights CO2 purity specifications.

If the amine-based CO2 capture system were optimized in isolation, it would capture 90% of the flue gas CO2. This minimises the size of the absorption tower to reduce capex, and minimises the reboiler energy to reduce opex.

The amine system would typically yield CO2 of around 99% purity with nitrogen as the main balance gas. This would then be fed to the CO2 liquefier. In the liquefier, the incondensable nitrogen must be vented from the system. As it is released, it carries with it an amount of CO2 according to the vapour / liquid equilibrium of these two gases. Typically, the vent is between 50% and 60% CO2, with only the balance being the nitrogen which must be purged.

The implication is that the overall capture rate is less than 90% due to the additional CO2 losses. The end-to-end system no longer meets the given constraints defined in the boundary conditions. The solution is to increase the CO2 capture rate in the amine absorption towers, boosting the energy requirement and increasing the equipment size.

Switching CO2 capture technology

Another common mistake when attempting to optimise the end-to-end process is to substitute one technology for another, without due consideration of the downstream implications.

In the case above, it may be decided that the amine system should be replaced with a pressure / vacuum swing adsorption process using a solid zeolite sorbent to capture the CO2. This technique is generally able to produce CO2 of a purity of 95% cost-effectively, with nitrogen being the balance. Whilst the CO2 capture rate might be high, at 98%, the CO2 purity may be low.

The motivation for substituting the amine-based system for the VPSA unit may be sound, for example to reduce the capex or opex of the CO2 capture aspect of the process. However, the knock on effect is that a negative consequence is felt in the CO2 liquefier.

The electrical energy input required to liquefy CO2 of 99% purity from the amine system is lower than that required to liquefy CO2 of 95% purity from the VPSA. This is because the compressors must do un-productive work harder to handle the inert nitrogen gas, and the system pressure is increased due to the presence of a higher amount of incondensable nitrogen. In turn, this increases the compression energy requirement.

The net result could to be that the savings achieved by implementing the VPSA system are negated by the increased capex and opex in the liquefier.

Cryogenic CO2 capture

If CO2 is required as a gas, for example for transmission in a pipeline, the additional energy input required to liquefy the gas would have no value to the process. On the other hand, if liquid CO2 is the requirement, the energy input must be applied at some point.

Where liquid CO2 is the target, it would be relevant to consider integrating the CO2 capture process with CO2 liquefaction. This is the essence of cryogenic CO2 capture.

The physical property of CO2 that makes cryogenic CO2 capture technically feasible is that it liquefies at a much higher temperature than the other major components in a typical flue gas, namely nitrogen and oxygen.

The basic process involved in cryogenic CO2 capture is to cool the entire flue gas stream to the point where the CO2 forms a liquid, but the other components remain as gases. When considering the operability and energy efficiency of this process, several critical aspects come to the fore.

1) Moisture in the flue gas must be removed using a drying process prior to cooling the flue gas below 0 °C. This is essential to prevent ice forming, which would block the process equipment.

2) The nitrogen and oxygen in the flue gas are cooled along with the CO2. This serves no benefit in the overall process, so the cold energy must be recovered from the nitrogen and oxygen. This can be achieved using a cryogenic heat exchanger to cool the incoming flue gas against the cold oxygen and air leaving the process.

3) CO2 forms a solid directly from the gas under certain temperature and pressure conditions. If solid CO2 accumulates, it will block the equipment in a similar way that frozen water would. The transition of CO2 from gas to liquid must be carefully managed through process engineering and a deep understanding of CO2’s phase diagram.

Rising to the challenge

Several pioneers have addressed the issues associated with cryogenic CO2 capture. The first to enter the limelight was Sustainable Energy Solutions (SES), founded in 2008 by Dr. Larry Baxter and based in Orem, Utah. In 2020, Chart Industries acquired SES as part of their drive to focus their technology portfolio and business towards the energy transition.

Chart’s Cryogenic Carbon Capture, or CCC process exploits the gas to solid deposition phase change that is unique to CO2. The entire flue gas stream is cooled to around -120 C causing deposition of gaseous CO2 as solid particle. Once the small CO2 particles are formed, they are converted into liquid CO2 through subtle temperature and pressure changes.

As the flue gas is cooled, condensable pollutants such as oxides of sulphur (SO2) and nitrogen (NOx) are converted into liquids before the CO2 deposits as a solid. These liquids can be separated to purify the emissions from the process. Additionally, any traces of these and other pollutants that break through into the liquid CO2 will solidify and can be removed using filtration.

The simultaneous removal of category pollutants is a feature of the CCC that maximises its economic benefits when being integrated into greenfield site projects, because it reduces the costs of upstream SO2 and NOx removal, which are regulatory requirements. Furthermore, when retrofitting the CCC process to existing plants that do not have adequate pollutant gas removal the investment can fulfil two core objectives simultaneously.

CCC has been piloted at 1 tonne per day of CO2 capture using a containerised, mobile plant at cement factories and power plants. A 30 tonne per day demonstration plant at Central Plains Cement Company in Sugar Creek, Missouri has also been commissioned.

Revcoo’s CarbonCloud

Founded in 2021 by Hugo Lucas and based in Lyon, France, Revcoo’s CarbonCloud is exactly that. Cryogenic liquid nitrogen is sprayed into the pre-cooled flue gas. As the CO2 in the flue gas is cooled by the nitrogen, it solidifies. This CO2 snow cloud is then separated from the gases based on density differences in a cyclone. The solid CO2 is then warmed up to yield liquid CO2.

To create the liquid nitrogen, mechanical energy is required in the compression stages of the liquefier. This is generally provided from electrically powered motors to drive the compressor. To maximize the energy efficiency of the process and minimize the amount of electricity required for the nitrogen liquefier, cold energy from the flue gas is recovered to pre-cool the warm incoming flue gas before it enters the CarbonCloud chamber.

Additionally, cold energy is released from the solid CO2 as it converts to liquid. This is an additional means by which the flue gas entering the CarbonCloud is pre-cooled.

A Caron Cloud pilot plant capable of removing 2 tonnes per day of CO2 was installed at the Eiffage lime kiln in France in 2024. Lime making releases unavoidable geogenic CO2 as the calcite (limestone) rock is converted to lime. Similar to cement, the geogenic CO2 emissions from lime making can only be decarbonized with the use of CO2 capture.

One company seeking to address this challenge is Canadian technology developer G4 Insights, which has developed a bio-hydrogen production process based on woody and cellulosic biomass.

The company has developed a proprietary process known as PyroCatalytic Hydrogenation (PCH), which converts woody and cellulosic biomass into biomethane, bio-syngas, and bio-hydrogen, while also producing biochar and biogenic carbon dioxide (CO2) as co-products.

“Resilience, localised supply chains and energy security are valued more than in the past”, believes Edson Ng, CEO of G4 Insights. “And we must simultaneously focus on affordability and sustainability.”

The technology is designed to unlock energy from cellulosic biomass resources that have historically been difficult to utilise, including residual forestry waste, straw, and waste wood.

While most biomethane production relies on anaerobic digestion, G4 says PCH has been developed to process underutilised cellulosic feedstocks and generate a range of energy products.

Alongside bio-hydrogen, the process can produce biomethane, bio-syngas, biochar and biogenic CO₂, allowing operators to optimise product output according to market demand while maintaining a carbon-negative overall process balance.

The cellulosic challenge

Local energy sourcing is a fundamental driver of supply chain resilience, and bio-based energy is generally produced and consumed locally. This is due to the low volumetric and gravimetric energy density of the biomass feedstocks, such as wood, grass, bagasse, and maize, compared to liquid fuels.

Bio-hydrogen can be produced by reforming biomethane, which primarily comes from anaerobic sludge digesters of purpose-grown crops such as grass or maize.

Despite many efforts to adapt the anaerobic sludge digestion, this process has not yet been successfully deployed at scale for highly cellulosic feedstocks such as straw and waste wood.

Ng described G4’s technology as a “technological breakthrough,” arguing that it could enable the use of cellulosic biomass resources that have historically been difficult to convert into energy products.

Research into converting residual forestry waste and other cellulosic post-harvest wastes has been ongoing for decades. Current technologies centre around various thermal techniques such as torrefaction, pyrolysis, and gasification.

G4 PyroCatalytic Hydrogenation in 10 steps

1) Woody biomass feedstock preparation to create particles with dimensions in the order of 5mm;

2) Pyrolysis of the biomass to create pyrolysis gas (pygas) and biochar;

3) Recovery of the biochar as a valuable co-product;

4) Conversion of the pygas and residual tars to biomethane in a proprietary catalytic hydrogenation reactor;

5) Stream-splitting to recover a portion of the biomethane as product and feed the balance to a steam methane reformer (SMR);

6) Conversion of the bio-methane to bio-syngas reformate in the SMR;

7) Water gas shift (WGS) to convert carbon monoxide (CO) from the bio-syngas to green bio-hydrogen and CO2;

8) Separation of CO2 from green bio-hydrogen using pressure swing adsorption (PSA);

9) Recovery of the biogenic CO2 as a valuable co-product;

10) Recycling green bio-hydrogen from the SMR to the hydrogenation reactor.

Overcoming catalyst fouling

Research into the catalytic conversion of uncleaned pygas has taken place over many years. Until now, this direct conversion of uncleaned pygas has been held back by the propensity of tars in the pygas to coke onto the catalyst, which generally uses a nickel-based compound as the active material.

In the PCH process, conversion of pygas to biomethane is achieved in the catalytic hydrogenation reactor. According to G4, two features distinguish the process from previous approaches. Firstly, bio-hydrogen is introduced into the reactor. Secondly, the reactor design itself is proprietary to G4 Insights.

“The hidden secret of G4’s technology is the internal design of the hydrogenation reactor,” Ng said. “It carefully controls the strong exotherm to avoid catalyst damage. Simultaneously, it achieves an exceptionally high conversion of tars and other chemicals in the pygas to methane through tight control of the temperature profile in this reactor.”

Carbon negative, flexible output

In G4’s PCH process, both the pyrolysis and the steam methane reforming (SMR) stages require heat input to drive the chemical reactions.

The SMR requires heat input, and it also emits heat because the reformate gases are cooled between the water gas shift (WGS) stage and the pressure swing adsorption (PSA) unit. This heat can be recovered and used to drive the pyrolysis reaction.

In the case of PCH, bio-methane produced in the catalytic hydrogenation reactor, or natural gas, can be used. Ng added, “The SMR can be electrically heated. Electrical heating of the SMR with renewable power, such as that from hydro, wind and solar resources, simultaneously reduces the carbon intensity of the bio-methane and maximises the bio-methane yield.”

Electrified reforming and electrically heated SMRs are being developed by Topsøe with its eREACT system and Sypox. In a classical flame-heated SMR, approximately one-third of the methane feed is to the burner. Therefore, if this energy input is replaced by renewable electricity, the CO2 intensity of the products from the SMR can be reduced by 30%.

Throughout the PCH process, biomethane, biosyngas and green bio-hydrogen are generated at various stages. Biochar and biogenic CO2 are also important co-products. The process can be configured to favour different combinations of biomethane, bio-syngas, and bio-hydrogen depending on project requirements.

“Production of more than one bio-energy gas enables revenue diversification from the investment to reduce the commercial risk,” claimed Ng. “During operation, some deviation from the design point can also swing the product mix towards market needs. This enables arbitrage to maximise production of the most profitable product(s) at any moment in time.”

The biogenic CO2 can either be sequestered or liquefied and sold into the merchant CO2 market. The biochar can be monetised as a soil improver, and will also be eligible for CO2 removal certificates since it permanently binds atmospheric carbon.

According to G4, when the overall carbon balance of the PCH process is considered, the resulting bio-hydrogen can be produced with a negative carbon intensity

G4 argues that revenues from co-products such as biochar and biogenic CO2 could help offset production costs and improve the economics of bio-hydrogen projects.

While the technology can produce multiple energy vectors, its ability to utilise low-value forestry residues and other cellulosic waste streams is central to the company’s commercial proposition.

Validation

The company says it has accumulated over 2,000 hours of testing in Canada, generating data on reactor performance and catalyst behaviour.

G4 sees potential for the technology in applications where forestry residues and other cellulosic waste streams are available and where demand exists for low-carbon fuels, carbon removals and biogenic CO2.

The company is now planning to build commercial pilot plant at 40 dry tonnes per day biomass input to produce four tonnes of bio-hydrogen nine tonnes of biochar per day.

Future commercial plants processing 250, 500, and 750 tonnes of biomass per day are also planned.

Cryogenic liquefaction of a gas consumes mechanical energy which is generally provided from an electrical motor through a compressor. Once liquefied, the gas is colder than the ambient surroundings.

Regardless of how good the insulation around the cryogenic liquid storage vessel is, it will not be perfect. So, over time, it is inevitable that heat will ‘leak’ into the liquefied gas and it will progressively ‘boil-off’ to become a vapour, known as boil-off gas (BOG).

When a liquefied cryogenic gas boils-off, the pressure inside the storage vessel increases. To protect the vessel, it must be fitted with a pressure relief vent valve and potentially also a bursting disc. If boil-off is not controlled, BOG will be vented and lost from the system.

The BOG will certainly have a financial value both from the chemical itself and the energy that has been used to liquefy it. Additionally, it may be toxic (ammonia), flammable (hydrogen), asphyxiant (carbon dioxide) or a highly potent greenhouse gas (methane). Therefore, it is essential to mitigate boil-off to prevent monetary loss, and to protect people, assets and the environment.

Liquid hydrogen and super insulation

The temperature difference between cryogenic liquid hydrogen, when stored at circa 5 bar, and normal ambient temperature is about 270 C. That is more than the temperature difference between your hand and a pan of boiling frying oil. In this most severe case, super insulation using a combination of vacuum and Multi-Layer Insulation (MLI) is essential to avoid rapid boil-off and loss of liquid hydrogen.

Cryogenic vacuum insulation is achieved by using two steel, or stainless steel tanks. One is placed in inside the other and a support structure links the two. Between the two tanks there is a gap of circa 100mm. A vacuum is pulled between the two tanks to minimise conductive heat transfer. The presence of the vacuum means that the outer tank is exposed to “under-pressure” so it must be strong to avoid crumpling inwards.

Also, inside the space between the two tanks, there are multiple layers of a reflective material MLI) which prevents radiative heat ingress to the cryogenic liquid. MLI is made of alternating layers of Mylar film, coated with a microscopic layer of aluminium, and a fine mesh, generally made of polyester or nylon. This MLI is like the coating used on satellites, where the film is often coated with a microscopic layer of gold for even better radiation protection.

Closed loop cryogenic reliquefication

Liquid helium and liquid hydrogen are used in superconductivity and deep space research where ultra-low temperatures are required. Deliveries of liquid or gaseous helium and hydrogen incur cost and, in space-constrained research centres, they can be logistically inconvenient. The solution is to charge the system with the require gas or liquid during startup, and then perpetually reliquefy the boil-off gas over the years of operation.

At small-scale, the Stirling Cycle is an ideal process for reliquefication of BOG at the ultra-low cryogenic temperatures encountered with liquid helium and hydrogen. The cycle compresses the helium or hydrogen in a piston in the Cryogenerator. The heat of compression is removed, and the gas is expanded. The process repeats this cycle to create a sufficiently cold environment in the machine to liquefy the gas.

A single Cryogenerator can be used during startup to liquefy the initial supply of gaseous hydrogen or helium and then to reliquefy BOG that is produced during normal operation. During BOG reliquefication, less electrical power is required than during the high-intensity startup phase because the gas entering the Stirling Cryogenerator is cold.

Liquid CO2 and foam insulation

The temperature of liquid carbon dioxide (CO2) at the pressure under which it is normally stored (circa 15 bar) is around -30 °C. Much warmer than the liquid hydrogen, which is around -250 °C. This means the temperature difference, which is the driving force for heat to leak into the storage tank, is around 50 °C: substantially less than in the case of liquid hydrogen.

Cryogenic liquid CO2 can be stored in a vacuum insulated vessel. However, due to the smaller temperature difference to ambient conditions, it is also possible to use a foam insulation layer. The converse does not apply for very cold cryogenic liquids such as nitrogen, hydrogen or LNG. They must be stored in vacuum insulated tanks. In a foam-insulated tank, they would boil off within hours due to the high amount of heat ingress caused by the large temperature difference between the liquid and the ambient conditions.

Use of foam, instead of the vacuum, reduces the cost and weight of the storage tank. Since there is no vacuum, the foam can be covered in a thin, lightweight aluminium shroud which is much lighter, and less costly than the strong outer jacket that is required for the cryogenic vacuum insulation.

Operational considerations

For a industrial gas distribution business which operates with multiple cryogenic liquids, such as nitrogen and CO2, there is a trade-off between storage tank asset flexibility and capex minimisation.

If a business estimates that they will need 10 cryogenic road tanks to support liquid nitrogen logistics, and 10 to support their CO2 distribution requirements, the lowest capital solution would be to purchase 10 super insulated vacuum tanks for the liquid nitrogen and 10 foam insulated tanks for the CO2.

On the other hand, if the demand for either of these two liquefied gas distribution tanks changes due to distribution distances increasing, or the balance of the two products shifting in the market then there is no possibility to use the foam-insulated tanks for liquid nitrogen. Additional assets must either be purchased or leased.

If the operator had initially purchased a fleet of 20 vacuum insulated tanks, the initial capex would be higher, but the operational flexibility would be greater. Additionally, owning fewer assets reduces maintenance costs.

Product transfer

When a cryogenic liquid is transferred from one tank to another, the receiving tank is likely to be either empty of only partially full with the cryogenic liquid product. The implication is that much of the receiving tank may be a slightly higher temperature than the cryogenic liquid that is being introduced. The result is that some of the liquid boils-off.

Also, to transfer the liquid there must be a pressure drop between the source and the receiving tank. This pressure drop also results in cryogenic liquid ‘flashing’ or boiling off during the product transfer. A good way to visualise this is to observe what happens when a bottle of fizzy water is opened. As the pressure is reduced, gas bubbles emerge as dissolved CO2 ‘flashes’ out of the water.

To avoid product loss, with its adverse financial, safety and environmental consequences, the boiled-off vapours are not vented to the atmosphere. Instead they can be used to maintain the pressure in the tank to provide the driving force which transfers the liquid to the receiving tank.

Alternatively, the boil-off during product transfer can be re-liquefied. If the transfer is from port to an LNG tanker at an LNG terminal, the boiled off methane can be transferred to the inlet of the natural gas liquefaction train.

In modern CO2 terminals which are being implemented to support CCS value chains, it may be the case that CO2 is fed to the port as a gas and then liquefied at the port location on a CO2 liquefier. Here also the CO2 that boils off during product transfer to the ship can be routed back to the inlet of the CO2 liquefier.

Shipping LNG

Once LNG is loaded onto the ship, the voyage may be several weeks duration. The short Atlantic crossing from the US Gulf Coast to Rotterdam is approximately 8 days. From the Gulf through the Straits of Gibraltar and the Mediterranean Sea to the Greek LNG terminal on the Islet of Revithoussa west of Athens is more than 20 days. Even the best forms of cryogenic insulation cannot prevent boil-off over this timescale.

To prevent methane losses, onboard the LNG tanker, there will be a re-liquefaction unit such as the Compact Reliq™ offered by Wärtsilä. This uses similar technology that is employed in the main LNG liquefaction train, at a smaller scale. It converts the boiled-off natural gas vapour back to LNG which flows back into the storage tanks on the ship.

The required LNG BOG reliquefication capacity for large tankers means that the thermodynamic cycle and equipment used in the Compact Reliq™ differs from the Stirling Cycle. Instead of a reciprocating piston, high speed rotating centrifugal compressors and expanders are used to chill nitrogen as a refrigeration gas in a closed-loop Brayton cycle.

The cold nitrogen gas is fed to a cryogenic heat exchanger in which the BOG is also present. This causes the BOG to reliquefy to LNG. The nitrogen returns to the compression and expansion cycle in the Compact Reliq™ unit to be re-chilled.

To meet the demand, new generation capacity is being added worldwide. However, in many grids, growth of non-programmable renewable generation is outpacing flexible and firm electricity sources. Long-duration energy storage (LDES) is therefore essential to balance the supply and demand of electricity.

LIBs are the core technology for LDES

For daily balancing with a discharge period of up to 4 hours, lithium ion batteries (LIBs) are unquestionably state of the art. LIBs are the technology of choice for battery electric vehicles (BEVs) and handheld devices such as mobile phones and tablet computers.

Huge utility-scale LDES facilities are being implemented in many countries. They are typically able to supply 100 to 500 MW of power and store up to 2,000 MWh of energy.

For grid balancing, as is the major application in Europe, the discharge period of an LIB LDES scheme is generally optimised at 4 hours. For solar power time shifting in locations such as California, they must work overnight and typically discharge over an 8 to 12 hour period.

CO2 for LIB production and recycling

LIBs are complex electrochemical devices. Throughout their production and re-processing, CO2 has a central role to play.

During charging and discharging electrons flow from the LIB anode and cathode, and vice versa, through an electrolyte which is supported in a carbonate solvent. That solvent is produced utilising carbon dioxide (CO2).

At the end of life, valuable industrial metals such as lithium, cobalt, nickel and manganese are recovered during LIB recycling using hydrometallurgical processes which also utilise CO2.

Carbonate electrolyte solvent production

The electrolyte is supported in a hydrocarbon solvent such as dimethyl carbonate (DMC) or ethylene carbonate (EC).

EC is produced by reacting ethylene oxide with CO2. For every Tonne of EC, 500 kg of CO2 are required. The CO2 purity must be very high to avoid contamination of the EC. For example, the CO2 must contain less than 20ppm of moisture.

DMC has traditionally been produced from the poisonous gas phosgene. The modern process for DMC production avoids phosgene and is therefore much safer. It reacts CO2 with MC and thereby also enables synergy within an integrated ethylene derivatives chemicals complex.

CO2 utilisation for battery electrolyte solvents

In March of 2024, Dow[1] announced their intent to invest in a new ethylene derivatives plant on the U.S. Gulf Coast. This will produce carbonate solvents which are used to support the electrolyte in LIBs and will support growth in BEV and LDES markets.

As part of Dow’s Decarbonize & Grow strategy, the ethylene derivatives facility will capture more than 90% of the CO2 from Dow’s existing ethylene oxide manufacturing process. This CO2 will be utilised to produce the carbonate solvents.

CO2 must be captured during ethylene oxide production to enable process operation, making this one of the lowest cost sources of CO2. Utilisation of the CO2 prevents its direct emission to the atmosphere.

LIB technologies

There are many types of LIBs, each with different active materials in the LIB anode.

The highest energy density is achieved with Nickel, Manganese, Cobalt (NMC) LIBs. These are favoured in premium automotive applications and are displacing Lithium Cobalt Oxide (LCO) LIBs in premium mobile phones to extend their battery life.

Lithium Iron Phosphate (LFP) LIBs are known for their low cost and are regarded as being safer than NMC LIBs. They are heavier than NMC batteries, but for static power storage applications this weight penalty is of no consequence.

Regulatory drivers and strategic materials

NMC and LCO LIBs contain large amounts of cobalt, manganese and nickel. In the EU cobalt and manganese are both recognised as both strategic and critical materials. Nickel is regarded as strategic. Recovery of these metals during LIB recycling is therefore of paramount importance.

Reflecting these points, the EU Battery Regulation (2023/1542) is now in force. It is a very stringent framework for battery life-cycle management.

In 2026, at least 65% of the total mass of an LIB must be recycled. That requirement rises to 70% in 2030. Zooming into cobalt and nickel, the reprocessing recovery at the end of 2027 must be at least 90%. By the end of 2031, that increases to 95%.

To extract nickel and cobalt, it is necessary to separate them from lithium in the LIB.

CO2 utilisation for Lithium circularity

EU Battery Regulation currently requires 50% of the lithium to be recovered. This is aligned to smelting or pyrometallurgical recovery of LIB materials. However, this jumps to 80% for 2032 and beyond. This increase will force the industry to implement chemical-based hydrometallurgical recycling which is more efficient at recovering lithium from LIBs.

The initial stages of LIB recycling are brutally mechanical. The batteries are crushed to release the electrolyte and its solvent. Copper and aluminium foils are shredded and separated using air blowers. The resulting powder, called black mass, is a mixture of anode and cathode materials. It is rich in metal oxides, lithium and graphite.

Hydrometallurgical recycling mixes the black mass with water. Then, CO2 gas is sparged into the liquor. The dissolved CO2 converts lithium ions to soluble lithium hydrogen carbonate, which can then be precipitated and captured using filtration.

Dynamic market

As with any emerging technology, in the field of LDES and batteries, there is rapid innovation which leads to product substitution and obsolescence.

NMC batteries are being challenged by LFP. The economics and process for LFP battery recycling are very different to NMC LIBs. If NMC batteries are no longer used, the CO2 demand for hydrometallurgy will decline.

For stationary LDES, vanadium redox flow batteries and iron-air batteries are challenging the dominance of LIBs. In EV mobility, solid state lithium ion and sodium ion batteries are potentially safer and more economical options.

If these newer technologies dominate, the need for CO2 to produce EC and DMC electrolyte solvents for LIBs may peak 15 to 20 years from now, plateau and then progressively decline.

Place your bets now

In the short- to mid-term, investing in CO2 supply for LIB electrolyte solvent production and hydrometallurgy, would seem to be an extremely good bet to leverage a rapidly growing LDES market for stationary and mobile applications.

On the other hand, as innovation races ahead and substitute LDES and battery technologies establish themselves as challengers to LIBs, it will be essential to manage investments towards the peak of demand to avoid over-capacity in a declining market.

[1] https://corporate.dow.com/en-us/news/press-releases/dow-announces-intent-to-invest-in-new-world-scale-carbonate-solv.html

OCP is the world’s largest producer of phosphate fertilizers with 31% global market share. Jorf Lasfar is the world’s largest phosphate processing plant and the largest fertilizer complex in the world, from which OCP exports phosphate fertilizers worldwide.

Florida is also a major producer of phosphate fertilizers, exploiting locally mined fluorapatite ore. The Mosaic Company and Nutrien both operate fertilizer production plants there. Despite this local production, the USA is the largest importer of DAP worldwide.

Geogenic CO2 from DAP production

DAP is produced in locations where phosphate rock is mined. This ore is rich in a phosphorus-bearing mineral, fluorapatite. The ore is beneficiated to remove sand, clay and as much calcite as possible.

In addition to CO2 emissions from the heat and energy requirements of DAP production, geogenic CO2 is produced during DAP production because the fluorapatite mineral is contaminated with calcite (calcium carbonate, or limestone).

The beneficiated fluorapatite ore is mixed with sulphuric acid, to yield phosphogypsum as a solid waste material and phosphoric acid. The phosphoric acid is reacted with ammonia, then granulated and sieved to produce DAP.

In addition to ore beneficiation, phosphoric acid production and DAP granulation, the end-to-end process generally involves on-site production of sulphuric acid from sulphur. Ammonia, another key input to the process, can be made onsite, but is generally purchased from off-site sources.

High quality ores, such those found in Morocco, and Jordan contain between 10 and 15% calcite. In other main phosphate rock mining locations such as Florida and Saudi Arabia, the calcite content can be up to 20%.

During phosphate rock beneficiation, clay and silica are easily removed. However only about half of the calcite is separated from the fluorapatite and the calcite is mixed with the sulphuric acid, along with the fluorapatite. During the reaction between calcite and sulphuric acid, CO2 is released.

The flue gas from the sulphuric acid mixing chamber is hot, moist and contains hydrogen fluoride (HF), which is released from the fluorapatite ore. It is scrubbed with water to remove the HF. The resultant gas mixture is hot, moist CO2 which is generally vented to atmosphere.

If the calcite content of the beneficiated ore is reduced to 6%, as would be the case in Morocco, approximately 0.05 tonnes of geogenic CO2 per tonne of DAP produced. For Saudi Arabian DAP production, the geogenic CO2 emissions liberated during the production of phosphoric acid are close to 0.1 tonnes of CO2 per tonne of DAP.

To partially decarbonise DAP production, recover and separation of the geogenic CO2 from moist flue gas can easily be achieved using cooling and condensation. The resultant dry, pure CO2 can then be liquefied for rail or sea shipment, or compressed for pipeline transmission at low marginal cost.

From theory to practice

The idea to capture CO2 from phosphate fertilizer production will be implemented from 2027 by OCP Nutricrops in partnership with OCP Green Water and INNOVX at OCP’s Jorf Lasfar industrial platform in Morocco.

The initiative is part of OCP Group’s decarbonization roadmap, which aims to achieve carbon neutrality across scope 1, 2 and 3 emissions by 2040. It also addresses international competitiveness by reducing the CO2 intensity of DAP which is important when exporting to the EU where the Carbon Border Adjustment Mechanism will impose tariffs on carbon-intensive imports starting in 2026.

The captured CO2 will be used by OCP Green Water for pH adjustment and re-mineralisation of drinking water in support of a local reverse osmosis desalination plant. This will support Morocco’s goal to meet 100% of the nation’s water needs through unconventional resources by 2027.

INNOVX will develop a plant to use captured CO2 and waste phosphogypsum to produce ammonium sulphate fertilizer. The Merseberg Process is one of the pathways to achieve this. It reacts aqueous ammonia with CO2 to produce ammonium carbonate.

Phosphogypsum is reacted with ammonium carbonate to yield calcium carbonate and the target product ammonium sulphate. Calcium carbonate is a recognised permanent sink for CO2, so this pathway sequesters CO2 gas emissions from the DAP process and enables integrated production of ammonium sulphate.

Ammonium sulphate was the first commercially available nitrogen fertilizer. However, it was displaced in many markets by ammonia and urea due to their higher nitrogen content. However, it is still used extensively in Brazil, which imports about 4 million tonnes per year, mostly from China.

Ammonium sulphate is suitable for the Brazilian soils which contains very low sulphur levels. Sulphur is essential to grow high protein crops such as soya beans, which are exported to the USA and used locally for beef cattle farming.

Ammonium sulphate also has a lower ammonia volatilisation rate than urea. This is an important factor in the warm Brazilian climate and leads to its use on South East Asia for growing wheat and rice.

Deep decarbonisation

For each tonne of DAP, 0.2 tonnes of ammonia are required to react with the phosphoric acid. Grey ammonia production releases about 1.8 tonnes of CO2 per tonne of ammonia. Thus, 0.36 tonnes of CO2 is released from the ammonia required to produce one tonne of DAP.

Comparing the CO2 emissions from ammonia production to the geogenic CO2 emissions for the Moroccan and Saudi Arabian cases above, this is about 3.5 to 7 times more than the amount of CO2 released from phosphoric acid production. To achieve deep decarbonisation of DAP production, capture of CO2 from ammonia production is also essential.

When ammonia is made from steam methane reforming of natural gas, CO2 leaving the reformer in the syngas must be removed to enable the catalytic Haber Bosch ammonia synthesis reaction to take place. This is normally achieved using an amine solvent based CO2 capture process.

The active ammonia catalyst is iron. All molecules containing oxygen, such as water, carbon monoxide or CO2 must be removed from the syngas before it is fed to the ammonia synthesis loop. Otherwise, the catalyst is oxidised and becomes ineffective. Therefore, every natural gas-fed ammonia plant already has a CO2 capture facility.

In some facilities, about 60% of the captured CO2 is combined with ammonia to make urea. However, the residual 40% of the CO2 from the SMR and ammonia plants that do not utilise captured CO2 for urea production can sequester the captured CO2 to reduce the CO2 intensity of DAP and nitrogen fertilizer production with only a small incremental cost for CO2 sequestration.

In China, coal-fed ammonia production is more common than the use of natural gas. Immediately after coal gasification the raw syngas is fed to a Rectisol unit where CO2 and sulphurous gases are removed.

This captured CO2 from ammonia synthesis coal gasifiers can also be sequestered at low-cost incremental since the capital and operating costs of the Rectisol plant are absorbed into the overall costs of ammonia production. To reduce the CO2 intensity of coal to ammonia, the only incremental costs are CO2 transmission and sequestration.

Process integration

OCP Group‘s third integrated fertilizer production plant, Jorf Fertilizers Company 3 (JFC 3), has been fully operational since 2017. JFC 3 has a capacity of 1 million tonnes per year of granulated phosphate fertilizers.

JFC 3 includes a sulfuric acid line with a capacity of 1.4 million tonnes per year, as well as a phosphoric acid line of 0.45 million tonnes per year. The chemical reactions involved in the production of these acids are exothermic and the heat is used to generate steam for 62 MW of electricity generation on a thermal power plant. This process integration makes the JFC 3 unit autonomous in terms of energy consumption and reduces the requirement to burn fossil fuels for electricity generation.

Phosphogypsum recycling

Ma’aden Phosphate is collaborating with thyssenkrupp Uhde, with additional expertise from thyssenkrupp Polysius, to recycle Phosphogypsum at Ras Al Khair.

Phosphogypsum is the solid waste material that is generated during phosphoric acid production. It is produced when sulphuric acid combines with the phosphate ore to produce phosphoric acid. Effectively, liquid sulphuric acid and solid phosphate rock is converted into liquid phosphoric acid and solid phosphogypsum.

Towering white phosphogypsum mountains are common close to phosphoric acid and phosphate fertilizer plants. Through calcination of phosphogypsum using high temperature heat, this waste can be converted to lime (CaO), sulphur dioxide (SO2) , oxygen and water. The SO2 gas can be converted to sulphuric acid to introduce circularity to the DAP production process.

The lime can be used for cement and steel production. Also it is used in the pulp and paper industry and for pH modification of soils. Alternatively, the lime can be used to mineralise CO2 emissions from DAP production to produce calcite (CaCO3). This mineral is insoluble and is recognised as a permanent CO2 storage mechanism.

Scope 3 emissions occur in the supply chain. They are neither directly emitted from the iron and steel plant (Scope 1) nor do they account for emissions from power generation for the iron and steel plant (Scope 2).

Purchased materials such as coke, oxygen, and lime are high CO2-intensity inputs that contribute to Scope 3 emissions.

Coke substitution with biocarbon

Pulverised coal injection (PCI) is used in some blast furnaces to reduce the iron ore while others rely on coke derived from coal. Biocarbon, produced from torrefaction or pyrolysis of wood, is gaining attention as a substitute for blast furnace coal or coke.

While switching to biocarbon does not reduce the total CO2 produced during iron ore reduction, it shifts emissions from fossil-based to biogenic, which is considered carbon-neutral over the lifecycle of the biomass.

Outokumpu is using biocarbon produced from waste-wood to reduce its carbon footprint in Tornio, Finland. Biocarbon replaces fossil coke in Outokumpu’s ferrochrome stainless steel production. The company has invested €40 million in a biocarbon plant at Mukran on the island of Rügen in Germany. Additionally, Outokumpu has invested in Envigas AB, a Swedish biocarbon producer, to develop a long-term biocarbon supply.

Switching from coke to biocoke reduces Scope 1 emissions at the iron and steel plant. If the biocarbon is sourced from a third-party supplier, its use also reduces Scope 3 emissions, as the CO2 emissions from offsite biocarbon production are biogenic.

Renewable power for over-the-fence green oxygen supply

Over-the-fence oxygen supply can significantly contribute to Scope 3 emissions. Oxygen is used for basic oxygen furnaces and reheating burners. It is produced on power-hungry air separation units (ASUs). The emissions from off-site oxygen production with over-the-fence pipeline supply are related to electricity generation. But when the power is used by a third party oxygen supplier, the emissions are categorised as Scope 3 at the iron and steel factory.

The core process for making oxygen is cryogenic air separation. Ambient air is drawn into the process using a large air compressor. To serve an iron making plant equipped with a blast furnace and multiple basic oxygen furnaces, the ASU may require an air compressor that continuously draws 20MW of power.

After the air is separated, the oxygen gas must be compressed to around 20 to 40 bar and supplied by pipeline to the iron processing facility. High pressure is required to blow the oxygen through the pipeline and ensure there is sufficient pressure to force the oxygen into the furnaces. In addition to the ASU air compressor, an oxygen compressor with a power draw of around 10MW may be required.

To decarbonise the electricity supply to the air and oxygen compressor motors, renewable electricity can be purchased by the ASU operator from wind, solar, or hydro sources. Use of low-carbon nuclear power generation for the ASU would also decarbonise oxygen production and supply.

Power transmission to the ASU via the grid will generally be required. A power-purchase-agreement (PPA) with a suitable guarantee of origin must be put in place to assign the appropriate quantity of renewable or nuclear power to the ASU operator so that the oxygen can be certified as ‘green’.

Low-carbon lime

Lime production is another significant source of Scope 3 emissions. Lime, used for slagging in blast furnaces to remove impurities, is often produced offsite. The CO2 emissions from offsite lime kilns are categorized as Scope 3 emissions for the steelmaker.

Lime making results in geogenic CO2 emissions which are generated when calcium carbonate (CaCO3) is baked out of limestone rock to generate lime (CaO). The thermal energy that is required to heat the rock during this process is also a source of CO2 emissions.

Heat for the lime kiln can be theoretically be decarbonised using biogas or biomass as a fuel source to supplement a natural gas burner or powdered coal burner. Whilst these options may sound good in theory, the reality is that a huge amount of heat energy is required. This stretches each of these decarbonisation options to, or beyond their practical limits.

The geogenic CO2 emissions from limestone baking are unavoidable. To mitigate these, CO2 capture equipment must be used. This feeds back into the decision-making process for decarbonisation of the high temperature heat input to the lime kiln.

The lime kiln burner flue gases are generally produced in the same gas stream as the geogenic CO2 emissions from the calcination process. This offers the potential to capture CO2 from heat generation alongside the unavoidable geogenic process CO2 emissions. A similar situation arises in the cement industry, where CO2 capture is being implemented on several plants. There are transferrable lessons between these sectors.

Liquid solvents to capture CO2 from lime making

The CO2 concentration in the flue gas from the kiln is elevated beyond that of a typical combustion process because it also contains geogenic CO2 from the calcination process. This CO2 enrichment influences the technology selection for the CO2 capture process.

Liquid solvent-based CO2 capture systems are being deployed for cement and lime making decarbonisation. In recent deployments, the selected solvent has often been an amine-based solution. This liquid is sprayed onto a rising plume of flue gas in an absorber tower. The amines chemically bond with CO2 to remove it from the decarbonised flue gas, which is vented to atmosphere.

The CO2-rich amine solvent is then pumped to a second tower called a stripper. At the base of the stripper heat is introduced to boil CO2 out of the solvent. The regenerated CO2-lean solvent is pumped back to the absorber tower and re-used to capture more CO2.

CO2 leaves the top of the stripper at a dry purity of around 99.5%. It is cooled to condense residual water vapour. The dry CO2 can either be compressed to transmit it by pipeline to an underground geological storage scheme or liquefied for transportation by rail or ship to a CO2 storage or utilisation location.

As an alternative to an amine-based solvent, hot potassium carbonate (HPC) solution can be used. Selection of an amine type solvent or HPC is driven by many factors, including the CO2 concentration in the flue gas and the pressure at which this stream is available. Higher concentrations and pressures tend to favour use of HPC, whereas the amine can be more cost-effective in low pressure flue gas streams where CO2 is present at a very dilute concentration.

Other factors which are key when considering the most appropriate CO2 capture technology include the presence of moisture, oxygen and sulphur compounds in the flue gas. Furthermore, if there is limited fuel supply available to provide heat for regeneration of the liquid solvent, an electrified process using solid sorbents may be preferred.

PSA for electrified CO2 capture

Solid sorbents can also be used for CO2 capture in lime making on pressure swing adsorption (PSA) systems. These sorbents may be molecular sieves, solid amines or metal-organic-frameworks (MOFs). In this technique, CO2 from the flue gas is chemically or physically adsorbed onto a bed of the sorbent at an elevated pressure.

The sorbent beads are contained within a pressure vessel. A number of vessels are connected in parallel with a manifold and change over valves to switch the gas flow between the sorbent beds. When a sorbent bed is saturated with CO2, the flue gas is diverted to another bed.

The sorbent bed which is saturated with CO2 is then depressurised. This pressure reduction, or pressure swing, causes the CO2 to be released. The sorbent is regenerated through this process. The process cycles between adsorption and desorption approximately every minute.

The CO2 released from the PSA system is typically less pure than that from a liquid solvent process. Several percent of nitrogen from the combustion air are generally captured with the CO2. A PSA system can generally recover CO2 at 95% purity. This gas may be suitable for compression and pipeline transmission or it can be liquefied and purified for transportation by rail, barge or ship.

CO2 capture for Scope 1 and Scope 3 emissions

CO2 capture techniques can also be used at the iron and steel making site to decarbonise blast furnace gas, basic oxygen furnace gas and other emissions. In these cases, Scope 1 emissions would be reduced.

CO2 capture at the offsite lime-making facility is the responsibility of the third-party lime producer. However, the iron and steel maker must ensure that the ‘green’ lime is appropriately certified with a suitable life cycle analysis to genuinely claim a reduction in Scope 3 emissions.

When it comes to Scope 3 emissions, trust, certification and collaboration throughout the supply chain are essential.

A hybrid approach

Electrof…

Read more

Several of the ongoing natural hydrogen exploration projects may commence com…

Read more

Traditional sources of CO2 are in a major transition. Ammonia plants in Europe are closing. Smaller refineries are at risk. Corn-ethanol production is unlikely to sca…

Read more

At the high temperatures and pressures in the gasifier, these feedstocks are converted into syngas, a mixture of hydrogen and carbon monoxide and carbon dioxide.

Read more

Whilst hydrogen is certainly a powerful tool in the toolbox for a net-zero future, this focus has diverted attention from the unique properties hydrogen offers. There are certain properties that hydrogen, whether it be grey, blue or green, offers as…

Read more

2.Electrolytic renewable/green hydrogen for sustainable e-fuels

3.CO2 capture from Waste to Energy can help to close the CO2 circularity loop

4.The need for new liquid CO2 supply chains

5.Will DAC be required?

Subscribe to view the full (30+ slides) presentation and other premium content from sbh4 consulting.

Read more

2.MSW as a feedstock for hydrogen and fuels production

3.Technologies and projects for MSW to hydrogen and fuels

4.Lessons from the past

5.Waste to energy is the low-risk option for hydrogen generation

Susbscribe to view the full (30+ slides) presentation

Read more

Drones, also known as unmanned aerial vehicles (UAVs) or unmanned aircraft systems (UASs) are increasingly being used in agriculture. In addition to liquid fertilizer application, they are deployed to spray pesticides and h…

Read more

The presentation (30+ slides) behind the paywall examines these areas in detail.

Read more