From pv magazine 12/2021
Green hydrogen is the topic of the year. From day-long sessions at COP26 to the oil and gas company advertising spree, to the policies hammered out in Brussels and Washington, DC, talk of green hydrogen – created using electrolysis powered by renewable energy – is everywhere.
Hydrogen is needed to decarbonise sectors of the economy that cannot fully be electrified, including international shipping, potentially flexible power generation, long-haul aviation, and the production of fertilisers, methanol and steel. But while hydrogen is either already used or has proven its ability to deliver in most of these sectors, it must also do so cost-effectively.
Critics are quick to point out that for all the talk of green H2, nearly all hydrogen production today is either made directly from fossil fuels or as a byproduct in petroleum refining. They also note that the dominant steam methane reforming (SMR) process in most places is currently many times cheaper than making green hydrogen.
This is widely expected to change over the next decade. Gigawatts of green hydrogen projects are underway and dozens more are planned by 2030. And major industry analysts – including BloombergNEF, Agora Energiewende, and RMI –expect that as these come online over the next decade, green hydrogen will become cost-competitive or cheaper than hydrogen from SMR. RMI, formerly known as Rocky Mountain Institute, says that this could happen even sooner with the deployment of as much as 25 GW of electrolysis capacity.
But when you look deeper into these forecasts, things get a lot more complex. Both current green hydrogen costs and the potential for cost reduction vary widely depending on a host of factors. These include the electrolyser technology, assumptions of learning curve effects, manufacturing and installation location, what the end-use application is, and what policies are in place to support deployment.
The core of hydrogen production is the machine that splits water into hydrogen and oxygen: the electrolyser. There is a broad consensus that as the scale of this central technology grows, they will get cheaper, and that this can be forecast as a learning curve effect.
Learning curves have been observed in dozens of products. Researchers have found that whether you are making airplanes or cellphones, there is a mathematical relationship between how many are made and how cheap they are to make. This has been most dramatically proven in solar manufacturing, where each doubling of deployment led to costs falling by between 28% and 36%.
BloombergNEF and the International Renewable Energy Agency have already estimated learning curve rates for green hydrogen with ranges from 12% to 21%, but these should not be taken as final. Many analysts warn that there are insufficient data points to work from.
“Learning curve is a super-high level metric that helps analysts frame and understand cost reductions,” explains Raffi Garabedian. Garabedian is no stranger to the solar learning curve, as he was formerly the chief technology officer at First Solar. He has since founded and serves as the CEO of Electric Hydrogen Company, a startup that seeks to bring down costs for green H2. “What is more helpful is to peel it apart into its constituent contributors,” he notes.
Alkaline vs. PEM
The first significant split is in technology. Most of the electrolysers made and installed today are based on either alkaline or proton exchange membrane (PEM) technology, and BloombergNEF reports a bigger market share for alkaline technology in current shipments.
Alkaline electrolysis of water has been conducted at industrial scale since the late 1930s. As such, it is a mature technology and unsurprisingly cheaper per unit of hydrogen produced. These machines are also easier to make and can last longer than PEM electrolysers.
PEM was developed in the 1960s and 70s for space and undersea applications. It is a faster-evolving technology, and while more expensive and harder to make, it takes up a smaller footprint. And it has another significant advantage: PEM electrolysers can respond better to the fluctuations in electricity supply that can come with solar and wind.
Garabedian says that PEM has significant cost-reduction potential. He cites “smarter choices of materials, and also productivity improvements in the electrochemistry itself” as important areas for work. But to cut costs PEM must also reduce or eliminate the use of rare and costly metals, including platinum and iridium. “We can’t make platinum-plated electrolysers and achieve the capital cost points that are required,” quips Garabedian.
Garabedian sees fewer opportunities for alkaline electrolysis. “Alkaline isn’t going to get cheap fast enough to keep up with PEM,” he argues. In particular, he notes that alkaline electrolyser cost is dominated by stainless steel, which isn’t getting cheaper. However, Garabedian says that there is a still a technology learning curve around performance.
And there are other factors that apply regardless of the technology used. “What will drive the costs of stacks down is primarily economies of scale and automation,” argues Gniewomir Flis, a hydrogen expert at Agora Energiewende. Flis notes that many electrolysers are currently made in manual processes, but the new gigafactories being built will be automated.
And it isn’t just the water-splitter itself that must get cheaper – it is also the integration of the electrolyser into larger systems, which Flis describes as being highly important. “In most applications, electrolysers are part of a much more complicated system, a chemical plant, for instance,” he explains. Much of the cost here is in power conversion, and as this represents a more mature technology, there are more limited routes for cost reduction. However, industry observers see opportunities for more efficient system design, including Garabedian, who insists that the current approach to electrolyser integration will not deliver the cost reductions needed. “You have to productise electrolysis the way that solar plants have been productised, so that they are easily and cheaply deployable,” he states.
Location, location, location
Electrolyser and integration costs are only part of the puzzle. The cost of electricity is a major input into the cost of green hydrogen, and as other costs fall, this is expected to make up much of the total costs.
Due to the challenges of transporting H2, both green hydrogen setups already installed and most of what will come online over the next few years is bespoke; on-site projects for specific end-uses. This ties the cost of green hydrogen to the availability of renewable energy in specific locations. And as solar and wind prices vary widely by location, so do these costs.
RMI has found that green H2 is already nearing cost-competitiveness with hydrogen made from SMR in some locations. And while many of these are far from demand centres, some are not. In Trinidad and Tobago, RMI found an estimated cost of green H2 production of only US$2.30 (AU$3.2) per kilogram, even with a US$700 (AU$980) per kilowatt electrolyser capex (including integration costs). And there are local markets for this hydrogen in both fertiliser manufacture and potentially in the restart of the island nation’s sole refinery.
Thomas Koch Blank, who heads RMI’s hydrogen work, argues that the advantages of deployment in optimal locations could shape the growth of the hydrogen economy, with green hydrogen taking off in places that have both local demand and rich renewable resources. This makes global average prices less meaningful for actual hydrogen uptake. “Nobody builds a new tech in a place with an average resource,” notes Koch Blank.
But what may be an even bigger factor for cost reduction is where electrolysers are built. While RMI estimates 2021 all-in costs of US$700 to $1,400 (AU$980 – $1960) per kilowatt for PEM electrolysers and US$500 to US$1,000 (AU$700 – $1400) for alkaline, it also notes that an alkaline system can be built for US$200 (AU$280) in China, using Chinese-made components.
Currently, Chinese electrolysers are largely confined to the Chinese market. “A Western bank looking at a Western project isn’t going to take the risk on a Chinese electrolyser maker,” Meredith Annex, head of heating and hydrogen at BloombergNEF, tells pv magazine. However, she expects that to change soon, and when Chinese electrolysers begin to be used in Western projects, this could greatly reduce costs for green hydrogen.
While most near-term green hydrogen projects are co-located with industry, there are several export-oriented projects planned to come online in the 2025-30 time frame. These are often orders of magnitude larger than the co-located projects, and many are looking to make ammonia using green hydrogen.
Transporting hydrogen, either by ship or pipeline, represents losses and/or extra costs, such as the cost of converting hydrogen into ammonia. However, if ammonia is the end-use, this cost will be borne either way. Additionally, transporting hydrogen allows it to be made in lower-cost locations and moved to higher-demand areas.
Most of the projects currently planned involve maritime shipping, as hydrogen pipeline networks are currently limited to a few locations like the Gulf Coast in the United States. However, new hydrogen pipeline networks including on the North Sea coast of Germany and the Netherlands are planned. Agora Energiewende’s Flis expects these to come online around 2030.
“It will take quite a lot of time to have a pipeline network to be sufficiently big enough and long enough to connect to places where there is a renewable advantage and no issues of public acceptance impeding mass rollout of renewables,” notes Flis.
Where you deploy hydrogen also matters greatly in terms of the policy support available. Policies to support hydrogen are being finalised and implemented by many governments including in the European Union, India, and the United States. Meanwhile, China’s 2060 carbon-neutrality goal appears to be sending a strong signal to businesses looking to adopt green hydrogen.
Other policies that are not specific to hydrogen can be strong drivers. In the European Union, the Red 3 Directive requires hydrogen refineries to integrate a certain quantity of renewable fuels of non-biological origin. But perhaps the most effective to date in driving green hydrogen uptake is the combination of the EU Emissions Trading Scheme (EU/ETS) and the Cross-Border Adjustment Mechanism (CBAM).
There is much speculation about what future EU/ETS prices will be. However, even the threat of a higher EU carbon price – particularly when free EU/ETS allowances phase out – is already driving a move to green H2 in sectors such as iron and steel making.
Across markets, BloombergNEF has identified the need for more policy certainty. “A lot of people are waiting on policy to make a final investment decision,” notes BloombergNEF’s Annex. “Why go ahead and do an unsubsidised project when you know you can do a subsidised one?”
The way forward
More can be done to bring down costs for green H2. For the near term, direct subsidies can help to get more projects off the ground, particularly in countries like Germany, where renewables costs are higher.
As costs come down, there is a need for other kinds of support. Several analysts that pv magazine spoke to stress the importance of standards and regulations to help build a mature market, particularly rules to define what qualifies as “green” hydrogen.
There are also solutions outside of policy, such as stronger co-location with wind and solar. As noted previously, much of the cost of integrating electrolysers comes from power conversion. This is also a cost for wind turbines and solar plants. If wind and solar projects are co-located to produce hydrogen on-site and then transport the hydrogen instead of electricity, this can reduce the need for costly power conversion components.
But ultimately the biggest factor may be sheer scale. Demand for electrolysers is growing at a dizzying rate. BloombergNEF expects 400 MW to 500 MW of capacity to be shipped this year, but 1.75 GW to 2.5 GW next year, with roughly 1 GW in China alone. The company is further tracking a total of 40 GW of projects. Beyond these there are MOUs signed for individual projects in the 26-30 GW range, but BNEF is not tracking these as there aren’t enough details yet.
Meanwhile, members of the trade group Green Hydrogen Catapult have set a goal to put online another 45 GW of green hydrogen production by 2027. Electrolyser makers are racing to catch up by putting gigawatt-scale factories online, and Agora Energiewende’s Gniewomir Flis expects the current shortage of electrolysers to turn into overcapacity around 2025.
The math of learning curves shows that what drives down costs is not how many years elapse, but how many electrolysers are made. As such, the ambition being shown by players in this industry is accelerating the rate of change. All of this quickens the day when fertilisers, steel, shipping, and potentially other sectors eliminate emissions, and when global industry is transformed to a clean future.
|Name||Nation||Main Companies||Electrolyser capacity||Technology||Status||Commission date||Applications|
|Hybrit (Luleå)||Sweden||LKAB, Vattenfall, SSAB||4.5 MW||Alkaline||Online||2021||Iron, direct reduction|
VERBUND, Siemens, others
|6 MW||PEM||Online||2019||Iron, blast furnace|
|Commercial Plant Svartsengi||Iceland||Carbon Recycling International||6 MW||Unknown||Online||2011||Methanol|
|Refhyne||Germany||Shell, ITM Power||100 MW||PEM||Online (first phase)||10 MW—2021,|
|Aqualyzer||Japan||Asahi Kasei, Toshiba Energy, Iwatani||10 MW||Alkaline||Online||2020||Stationary fuel cells, transport|
|Canada||Air Liquide||20 MW||PEM||Online||2021||Exports: liquid and gas, for industry and transport|
|Cachimayo Plant||Peru||Industrias Cachimayo||25 MW||Alkaline||Online||1965||Ammonium Nitrate (fertilisers, explosives)|
|Green Lab Skive||Denmark||Green Lab Skive A/B||100 MW||Unknown||Under construction (first phase)||6 MW—2022,|
|Puertollano Green Hydrogen Plant||Spain||Iberdrola, Ingeteam, Fertiberia||830 MW||PEM||Under construction (first phase)||20 MW—2021,|
|1000 MW||PEM||Under construction (first phase)||20 MW—2022,|
|Baofeng Energy||China||Ningxia Baofeng Energy Group||100 MW||Unknown||Under construction||30 MW (installed),|
|Leuna Chemical Complex||Germany||Linde, ITM Power||24 MW||PEM||Contracted||2022 (planned)||Exports via pipeline: chemical production, transport|
|Varennes Carbon Recycling||Canada||Thyssen Krupp,|
|88 MW||Unknown||Contracted||2023 (planned)||Biofuels (transport)|
|H2 Green Steel||Sweden||H2 Green Steel|
|800 MW||Unknown||Acquiring permits||2024 (planned)||Iron, direct reduction|
|Hybrit (Gällivare)||Sweden||LKAB, Vattenfall, SSAB||600 MW||Unknown||Acquiring permits||2026 (planned)||Iron, direct reduction|
|Porsgrunn Ammonia Plant||Norway||Yara, Nel||25 MW||Unknown||FID||2023 (planned)||Ammonia (fertilisers)|
|Westküste 100||Germany||EDF Germany, Ørsted, ThyssenKrupp, others||300 MW||Alkaline||FID||30 MW—2023,|
|Neom Zero-Carbon City||Saudi Arabia||Air Products, ACWA, ThyssenKrupp||4 GW||Alkaline||FID||2025 (planned)||Green ammonia for export, transport|
Note: This spreadsheet is not a complete accounting of all existing or planned green hydrogen projects, and only includes projects that have reached FID or similar milestones.
By Christian Roselund
This content is protected by copyright and may not be reused. If you want to cooperate with us and would like to reuse some of our content, please contact: email@example.com.
By submitting this form you agree to pv magazine using your data for the purposes of publishing your comment.
Your personal data will only be disclosed or otherwise transmitted to third parties for the purposes of spam filtering or if this is necessary for technical maintenance of the website. Any other transfer to third parties will not take place unless this is justified on the basis of applicable data protection regulations or if pv magazine is legally obliged to do so.
You may revoke this consent at any time with effect for the future, in which case your personal data will be deleted immediately. Otherwise, your data will be deleted if pv magazine has processed your request or the purpose of data storage is fulfilled.
Further information on data privacy can be found in our Data Protection Policy.