Credit: McPhy | Alkaline electrolyzers like this one made by McPhy Energy currently boast the largest installed capacities. But other technologies are in hot pursuit.
Electrolyzers are the electrochemical cells that split water into hydrogen and oxygen. The systems occupy an industrial niche today but may be a crucial part of how the chemical industry and several other sectors decarbonize in the near future. Two electrolyzer types share the current growing market, and two more are rapidly commercializing. The next handful of years will tell the tale, as corporate targets and government subsidies look to create a new, green, hydrogen economy.
Read the sustainability reports released by chemical companies, and you’ll encounter one phrase over and over: “green hydrogen.” That means H2 produced by splitting water with renewable electricity. And it’s not just the chemical industry. In materials, steel, biofuels, carbon dioxide utilization, heavy transportation, and even grid-level electricity supply, green hydrogen sits at the center of countless decarbonization plans.
But the truth is that green hydrogen barely exists. Less than 1% of the 10 million metric tons (t) of hydrogen produced in the US today counts as green, according to a May 2023 report by Carbon Solutions, a greenhouse gas reduction consultancy.
Instead, 76% is derived from natural gas or coal, a process that emits as much as 18 kg of carbon dioxide for every kilogram of hydrogen produced, and 23% is a by-product of petroleum refining or other chemical processes. Globally, the hydrogen market is about 96 million t per year, with a similarly tiny portion made in a green way.
A multitude of challenges stand in the way of green hydrogen as a climate-saving workhorse, the first being the availability of renewable electricity. “You have to have that energy come from a green source to actually make it viable,” says Amanda Morris, a chemistry professor at Virginia Tech whose research includes catalytic materials for energy applications. “The solar capacity of the United States is nowhere close to being able to create a green hydrogen economy.”
Infrastructure is another hurdle, as hydrogen’s small molecular size lets it leak through pipeline and container materials that work fine for other gases. Most of the hopeful new applications for H2 will also need to mature from pilot projects to commercial scale. And electrolysis will compete both with other low-carbon routes to hydrogen and with other uses for low-carbon electricity.
The industry will also need to build more electrolyzers—a whole lot more. Electrolyzers are the core pieces of a chemical kit that splits a mole of H2O into a mole of H2 and a half mole of O2.
The report from Carbon Solutions puts the number of electrolyzers operating in the US at just 42, with a combined hydrogen production capacity of about 3,000 t per year. The US Department of Energy (DOE) aims to have 10 million t of clean hydrogen flowing per year by 2030, 20 million t by 2040, and 50 million t by 2050. About half that production will come from renewably powered electrolysis.
And that’s just the US. The market intelligence group Rethink Technology Research projects that by 2050, the global demand for clean hydrogen will reach more than 580 million t. Blue hydrogen, made by converting fossil fuels to hydrogen and capturing and sequestering the resulting CO2, is included in the definition of clean hydrogen used by Rethink and the DOE. But blue hydrogen enjoys weak support outside the US and has a shaky track record for real greenhouse gas reductions, so most industry watchers expect green hydrogen to claim the lion’s share of global demand.
All that new hydrogen adds up to $2 trillion in global electrolyzer purchases over the next 27 years, Rethink says. Those purchases will be the backbone of an annual green hydrogen market that could be worth $850 billion.
It’s a massive, rapid scale-up. Electrolyzers are often described in terms of their electrical draw in watts, and there is more than 230 GW of planned capacity across 1,000 global projects as of January, according to an analysis done by the consulting firm McKinsey on behalf of the Hydrogen Council, an industry group. That number equates to about 38,000 t per year of hydrogen at current efficiencies. The count includes almost 800 electrolyzer units that are scheduled to start cranking out hydrogen by 2030 at a combined capital expenditure of $320 billion—if they’re all built. Less than a tenth of that investment has final approval from the companies planning it.
Many electrolyzers will be bigger than those in the current fleet. According to Carbon Solutions, the average electrolysis unit in the US today yields 0.20 t of hydrogen per day. The US government wants to invest $8 billion in several hydrogen hubs across the country by 2026, and they will be required to produce about 250 times as much hydrogen—at least 50 t per day.
The bullish zeitgeist around electrolysis is the product of more than just projections and government targets. Electrolyzer makers say the business environment around green hydrogen is different from other times in recent history when pundits predicted the coming of a hydrogen economy.
“We feel a step-change difference versus what we’ve seen in the past,” says Everett Anderson, vice president of advanced product development at the Norwegian electrolyzer maker Nel.
The orders are much larger, Anderson says, and the entities planning electrolytic hydrogen production plants are a different breed from before. “We were contacted before by one-off developers. Now they’re significant, multinational industrial and energy companies that are in this space looking to do these projects.”
Almost all electrolyzers operating today are based on one of two technologies: alkaline cells and proton-exchange membrane (PEM) cells. The companies making a third type, solid-oxide electrolysis cells (SOECs), say they are already taking orders as they move quickly to build factories to make them. A fourth electrolysis chemistry, based on an anion-exchange membrane (AEM), could one day combine the advantages of alkaline and PEM cells.
The term water electrolysis literally means splitting water using electricity. The net chemical reaction is always the same, H2O → H2 + ½O2, and voltage always pulls electrons out of the anode and through an external circuit and pushes them into the cathode. What varies between cell types is how the chemicals and materials inside the cell balance that electron flow by passing charge carriers—ions—between the electrodes.
Alkaline cells are the oldest and most mature technology for splitting water with electricity. The chemical industry has been making hydrogen this way since the 1800s, when the gas enjoyed a boom filling dirigibles and other floating aircraft.
Norway’s Nel sells systems using both alkaline cells and PEMs. The firm’s main experience is with alkaline chemistry, which it has offered at megawatt scales to customers in Europe since the 1950s. Nel acquired Connecticut-based Proton Energy Systems in 2017 both to get into PEM and to establish a foothold in North America, Anderson says.
In a commercial alkaline cell, concentrated aqueous potassium hydroxide or sodium hydroxide serves as an electrolyte, carrying electrical current in the form of –OH ions. The cell generates O2 at the anode and H2 at the cathode; a separator material often made of ceramics such as zirconium oxide keeps the gases from mixing while allowing water and ions to move freely.
PEMs, by contrast, conduct charge via H+ ions that move through a solid polymer membrane. The polymer features negatively charged functional groups, most commonly sulfonic acid, that can pass cations through but reject anions, according to Simon Cleghorn, a global product manager at the membrane maker W. L. Gore. The anode and cathode are attached directly to the two sides of the membrane.
“For both technologies, we have road maps to be able to improve both efficiency and reduce cost,” Nel’s Anderson says. “Because of the maturity and level of industrialization already achieved, the slope of the cost reduction curve on the alkaline is shallower than on the PEM. I think there’s more opportunity on the PEM side,” he says, in part because of R&D overlap with fuel cell technologies.
Virginia Tech’s Morris says electrode materials are one R&D front for alkaline cells. The incumbent materials are metallic foams made of nickel and iron alloys, she says. Academic researchers are exploring alloys with three, four, and five components. Industry research includes new cell configurations that increase the output pressure of the hydrogen.
Cleghorn says making membranes thinner is important in PEM R&D. Thinner membranes should offer less resistance, which would boost the hydrogen output from a given amount of surface. At the same time, electrodes that use less raw material would save on cost. Standard PEMs need electrodes made of the precious metals platinum and iridium to withstand the acidic conditions inside the cell, which would eat nickel and iron electrodes.
All those R&D efforts form a basis for competition among suppliers. Each company also has to incorporate ease of manufacture into the engineering details of cell and device design. Anderson, for example, says Nel is working on roll-to-roll PEM manufacturing to enhance throughput, quality, and uniformity, which will enable thinner, more efficient membranes and reduce precious metal use. And the corporate patent literature is active with methods to reduce the physical distance between electrodes in alkaline cells, a change that would cut the voltage required to drive electrolysis.
It’s reasonable, at least for now, to think of alkaline systems as big and powerful and PEM kits as small and nimble. Alkaline electrolysis systems are also cheaper, at $500 to $1,000 per kilowatt versus $1,100 to $1,800 per kilowatt for PEM systems, according to estimates from the International Energy Agency. One kilowatt is roughly equivalent to 0.5 kg of hydrogen per day at current efficiencies.
PEM electrolyzers, however, can ramp hydrogen production up and down quickly and easily, which makes them attractive for projects powered directly by wind or solar because they can automatically decrease production when the wind isn’t blowing or the sun isn’t shining.
The electrolyzer purchase decisions of the start-up Air Company provide a window to the current state of the market. Air Company is commercializing a suite of catalysts that convert carbon dioxide and hydrogen into methanol, ethanol, and jet fuel. And it’s working on a newly announced project with the US space program to make edible proteins, also using electrolytic hydrogen.
The firm has a pilot plant in Brooklyn, New York, that runs on CO2 captured from ethanol factories in New York State and green hydrogen made on-site. “Today, our business uses PEM electrolyzers,” says Stafford Sheehan, the firm’s cofounder and chief technology officer. “In New York City, the physical footprint of the electrolyzer is a real consideration, and PEM systems offer the most compact design.” A 140 kW PEM unit from Nel provides the 40 kg of H2 per day the pilot plant needs, he says. The unit is a self-contained box about the size of a Volkswagen Beetle.
Alkaline systems are much bigger for two main reasons: The cells have a lower output per square centimeter, a parameter often described as current density, so bigger units are needed. They also operate at close to ambient pressure, which means they need a lot of supporting tanks, pipes, and compressors to store and supply high-pressure hydrogen, which is what most hydrogen customers use. Membrane and solid-oxide systems operate at high pressures and have fewer caustic components, so they require less infrastructure.
But the constraints at the demonstration-scale plant that Air Company is hoping to open in 2024 will be different. It will need almost 160 times as much hydrogen per day as the pilot plant and be located in an area where real estate won’t be as expensive. “As we scaled up our process, we had trouble finding anyone selling 5 MW PEM electrolyzer trains with 10,000 or more operating hours,” Sheehan says. The firm needed that track record to attract financing on good terms.
“Industry requires electrolyzers that are proven to be robust and durable, that we can trust to have minimal downtime over several years, so we had to turn to alkaline systems,” he says.
By the time Air Company is ready to start building the commercial-scale plant it plans after 2024, the situation may be different. “PEM water electrolysis is a mature technology, but as a buyer of megawatt-scale systems, I felt it still has a lot of ground to cover to catch up with alkaline electrolysis,” Sheehan says. “I do think it will get there.”
Electrolyzer systems based on solid-oxide cells are poised to enter the fray around that same time. SOECs work at 600–900 °C, much hotter than alkaline, PEM, or AEM systems, which all operate between 80 and 120 °C. That high temperature lets SOECs use the unstable O2– ion to carry charge inside the cell instead of the H+ and –OH ions that are chemically accessible near water’s boiling point.
SOEC is the most efficient cell type, with about 90% electrical efficiency, compared with about 60% for alkaline and PEM systems.
“Most of the 10% inefficiency is boiling the water,” says Tony Leo, chief technology officer at FuelCell Energy. So in situations in which SOECs can harvest waste heat from other processes, they can approach 100% efficiency. FuelCell Energy is commercializing an SOEC electrolyzer technology, building on its main business of making fuel cells that use ceramic solid oxides or molten carbonates as their electrolytes.
A fuel cell is basically the reverse of an electrolysis cell; it uses electrochemistry to convert chemical energy to electricity. Leo says FuelCell Energy has been working with solid-oxide cells for more than 20 years, but “for a long time, electrolysis was a real niche application. There wasn’t really a big market for electrolyzers,” he says.
Today, systems optimized for hydrogen production are growing faster than fuel cells thanks to all the corporate targets and government incentives. “We’ve demonstrated durability and efficiency levels to our satisfaction in our lab. Now our task is to basically prove it to the world,” Leo says.
The firm is completing construction on a 270 kW, 100 kg per day demonstration system, which it will deliver to Idaho National Laboratory for testing. And it is taking orders for 1.1 MW systems that yield 600 kg of hydrogen per day. “It’s an exciting time, really, for us,” Leo says.
The International Energy Agency estimates that SOECs cost $2,800–$5,600 per kilowatt. Leo says he expects his firm’s early electrolyzers to be closer to $1,250—and eventually as low as $200 if the demand for very large sizes emerges.
Rethink hydrogen analyst Bogdan Avramuta expects SOECs will claim about 16% of the market by 2030 as FuelCell Energy, Bloom Energy, Sunfire, Cummins, and others build factories capable of churning out gigawatts of cells per year. “SOECs are in a strong position at the moment,” Avramuta says. Around 20% of announced electrolyzer manufacturing plants will be making SOECs, “and orders are starting to come in,” he says.
AEM is the fourth, and least well developed, electrolyzer type, but some industry watchers say it has the most potential. “When it comes to AEM, technology is just about starting to emerge after being in a development stage for the last few years,” Avramuta says.
The physical design of an AEM cell is similar to that of a PEM cell: an anode and a cathode on either side of a polymer membrane. The difference is that an AEM passes –OH anions rather than H+ cations through the membrane.
AEMs leverage “the elegance of making these things polymeric instead of using a liquid electrolyte, so that makes the whole system easier and cheaper to design” than alkaline cells, Gore’s Cleghorn says. “And by using alkaline conductors, you can use non-precious-metal catalysts. The combination of those two makes it the holy grail.”
Unfortunately, although alkaline conductors allow the use of base-metal electrodes, they eat through the carbon-carbon bonds in polymers. Finding materials that can conduct –OH ions and avoid being ravaged by them has been the major obstacle for large-scale introduction of AEM electrolyzers, according to a recent review (Chem. Rev. 2022, DOI: 10.1021/acs.chemrev.1c00854).
The electrolyzer maker Enapter would argue that AEMs are ready for prime time. The firm has taken orders for multiple-megawatt systems that promise to deliver hundreds of kilograms of hydrogen per day, though its current systems produce roughly 1 kg per day.
Virginia Tech’s Morris is skeptical that scale-up will be straightforward for SOECs and AEMs. “One of the things we are learning is that when we do it at the lab scale on our small little benches, then scale it up, it’s drastically different. Your product distributions are drastically different. Your efficiencies are drastically different,” she says. “And that’s just because, now, mass transport is massively important.”
The challenge for SOEC and AEM manufacturers is that the optimization and economies of scale that come with a massive expansion are happening now, and alkaline and PEM systems are in a better position to reap the benefit, Morris says. “The second that we ‘pick’ a technology, we figure out how to make that in the most efficient, lowest-cost way possible. And then it gets very hard to bring in new technology because that new technology is going to come in at that high initial cost.”
If the projections about the growth of green hydrogen are correct, there should be plenty of room for everybody. Indeed, every electrolyzer maker says its main corporate priority right now is rapidly expanding manufacturing capacity.
“We think that the technology is in a good place for market introduction,” Leo says about SOECs. “All the heavy lifting right now for us is manufacturing scale-up.”
Similar to electrolyzers themselves, electrolyzer factories are often measured in the watts their annual cell output will consume per year. In a year, a gigawatt factory would crank out a quantity of electrolyzer cells with an aggregate electricity demand of 1 GW. Globally, electrolyzer makers today can produce about 8 GW of electrolyzers per year, according to the International Energy Agency. It expects that number to grow to more than 60 GW by 2030.
It’s difficult to find agreement on how that 8 GW of manufacturing capacity is divided among the four major cell types, but some experts say alkaline cells are roughly two-thirds of it, and PEM cells take the balance. Rethink expects that ratio to roughly invert over the next decade, on the basis of project announcements from electrolyzer makers. Inroads by SOECs and AEM cells are hard to predict.
Also hard to predict is the cost, which the businesspeople behind any green hydrogen project want to know. “What’s the dollar per kilogram of hydrogen? That’s what people are really looking for,” Gore’s Cleghorn says. “There’s the operating cost, which is going to be really around electricity cost, and then the capital cost is the machine,” including the supporting equipment and the land to put it on, he says. “So how do you balance those to get the lowest cost of hydrogen?”
The DOE wants to see clean hydrogen cost $1.00 per kilogram, about what petroleum-based hydrogen costs today, and the Joe Biden administration is deploying production tax credits and other subsidies toward that goal. McKinsey’s analysis says the policies should work, with room to spare. It projects that the US will eventually be the cheapest place to get green hydrogen, at $0.50–$1.80 per kilogram.
Today, according to the recently released U.S. National Clean Hydrogen Strategy and Roadmap, green hydrogen costs between $5.00 and $7.00 per kilogram, whereas blue hydrogen costs $1.25–$2.10.
That seems like a big gap, but it doesn’t have to close all at once. The road map includes an analysis detailing the price that green hydrogen has to hit to be competitive with incumbent fossil fuels in several applications.
Hydrogen-fueled forklifts already make sense because of the logistical advantages for that niche, and 115 facilities using electrolysis that way are up and running. Buses and long-haul trucks are in reach at $5.00 per kilogram, and a suburb of Washington, DC, is in the final planning stages of a solar-powered filling station for a fleet of hydrogen fuel-cell county buses. The biofuel, iron, steel, and chemical industries will adopt green hydrogen when it is priced between $2.00 and $3.00, the road map says. And every time a new market opens up, the economies of scale accelerate the journey to the next inflection point in electrolyzer cost.
Rethink’s Avramuta sees an industry transformed. “The sun has set on the days when we were trying to convince you that the hydrogen industry is the next big thing,” he writes in the firm’s most recent analysis. “The world has moved on from disbelief to strategic planning.”