View this post on the web at [link removed]

Hydrogen is the most abundant element in the universe. As commercial developers of fusion power [ [link removed] ] technology like to remind us, it is hydrogen fusion that powers the sun. However, despite recent advances, practical fusion power remains a technology of the future, never seeming to get any closer.
At the same time, the smallest element looms large as a potential source of more conventional power, either as a clean-burning alternative to fossil fuel in combustion engines or, when combined with oxygen, as a source of electricity. On Earth, hydrogen is tantalizingly available in the form of water. But like the condemned Greek mythological figure [ [link removed] ] Tantalus, we seem unable to realize its promise.
Many governments remain keen to develop a hydrogen economy that could displace the existing fossil fuel-based one. The U.S. Department of Energy, for example, is offering billions of dollars [ [link removed] ] in incentives to establish “hydrogen hubs” across the country. Yet the allure of hydrogen as a limitless fuel is confounded by how elusive it seems to be in practical applications.
According to the International Energy Agency, global demand for hydrogen [ [link removed] ] could reach 100 million metric tons this year (if used for energy, about 3,300 terawatt hours). This is a fraction of the amount of natural gas [ [link removed] ] used in the world per year, which was about 4 trillion cubic meters (about 35,264 terawatt hours) in 2020. The reason for the disparity is that hydrogen is not a common energy source. While clean-burning, it contains only about 30% of the energy per volume as methane, the main component of natural gas.
Also, hydrogen is only about 12.5% as dense as natural gas, so moving it through pipelines efficiently requires higher pressures. In fact, one method of moving hydrogen without compressing it is to mix it with natural gas to move through pipelines. This, however, requires blending in at one end and filtering out at the other—both of which add costs. Likewise, compressing hydrogen into liquid form for transport hikes the cost, particularly at the scale needed for use in energy production.
If hydrogen is going to rise to its full potential as a fuel for widespread use, it will be because researchers, investors and businesspeople perceive opportunities to use its unique properties to solve problems that have no other cost-effective solution.
Spin the Color Wheel
Hydrogen can play a big role in generating electricity. One way to obtain hydrogen is through the process of electrolysis, which uses electricity to split water into hydrogen and oxygen molecules. The hydrogen is collected and contained for use or transport. Hydrogen fuel cells produce electricity by introducing oxygen to mix with contained hydrogen, producing water as a byproduct. But the conundrum here for hydrogen as an energy source is apparent: You have to use energy to produce the hydrogen that presumably you will use to generate electricity someplace else.
Hydrogen is used in many industries [ [link removed] ], importantly for producing ammonia for fertilizers and in petroleum refining. It is used in metalworking, glass production, medical and laboratory processes, and other sectors. Many types of rockets use liquid hydrogen for fuel. Existing industrial demands for hydrogen mean that there would be hydrogen producers regardless of its potential as a source of electricity.
The problem is that electrolysis is not the most efficient way to produce hydrogen, particularly at scale. The vast majority of the hydrogen produced in the world uses natural gas through the steam reforming process that results in carbon dioxide (CO2) as a byproduct. Another process makes hydrogen from coal through gasification, involving partial combustion that also ultimately results in CO2 as a byproduct.
According to MIT [ [link removed] ], about 95% of hydrogen is produced from fossil fuels. This releases 830 million metric tons of CO2 to produce about 74 million metric tons of hydrogen annually in recent years. Also, the various industrial processes used result in the escape of methane and/or CO2 into the environment, the former being regarded as a potent if relatively short-lived greenhouse gas.
The advantage of steam reforming in particular is that the natural gas and heat were going to be on hand anyway. And coal is abundant. Why not make all the hydrogen needed from these methods?
The answer, of course, is climate change. The prioritization of climate change and its association with greenhouse gases such as CO2 and methane resulting from human activities have completely transformed the way the modern world looks at energy. Climate change is one of the main drivers of the so-called energy transition away from fossil fuels toward carbon-neutral sources, such as renewable (solar, wind, geothermal), nuclear and hydroelectric power.
As a result, the energy industry and its pundits have come to categorize how hydrogen is produced, assigning each of these methods a color [ [link removed] ] and a value judgment according to how much greenhouse gas is released into the environment from its production. For example, green hydrogen is produced through electrolysis using renewable energy, while blue hydrogen comes from steam reforming where the CO2 is captured and sequestered in some way. These methods produce what is considered “good hydrogen.” The “bad hydrogen” comes from coal gasification (black and brown) and from steam reforming where the CO2 is not captured (gray). Other colors crop up as new categories and processes are devised (e.g., hydrogen produced via electrolysis using nuclear power is pink).
It is interesting to note that “white hydrogen” refers to gas that exists in its natural state trapped underground and doesn’t require any artificial formative process, only extraction and containment. At the moment there are no reliable technologies for the extraction of white hydrogen; however, a number of research teams around the world are studying the issue as potentially vast sources of natural hydrogen are being discovered, including an extensive one [ [link removed] ] in the United States.
Entrepreneurs Lift Off
While it may be tempting to scoff (or at least smirk) at the hydrogen rainbow pinwheel, the fact remains that concerns about the byproducts of any process are placing a premium on green and blue production methods. Moreover, demands to “decarbonize” the energy and transportation sectors are offering new opportunities to use hydrogen as a means of reducing dependence on fossil fuels. The Inflation Reduction Act of 2022 is financing numerous incentives [ [link removed] ] related to the production of hydrogen and its use as an energy source.
However, not all new industries are embracing hydrogen because of government subsidies. Yuval Bachar, CEO of California-based data center developer ECL, said he and his partners settled on hydrogen as the best method for producing electricity to meet that burgeoning industry’s power demands while also avoiding the trend toward diminishing grid access. Bachar’s company uses hydrogen fuel cells to generate electricity stored in batteries that power the servers in its data center facilities.
“When we started work at ECL, it was very clear to us that we had to actually operate off-grid, because we had a sustainability mission and the grid is not going to be green for a very, very long time—probably 25 to 35 years, if at all,” Bachar said, indicating that use of non- or low-carbon sources was important for the company’s clients. “The second aspect that came into play is that the availability of the grid was diminishing very quickly from 2021 up to where we are right now, where it’s almost impossible to get access to the grid for our power requirements in a timely way. So, our solution is to go off-grid. And the best way to do that is with hydrogen.”
It is difficult to overestimate the impact artificial intelligence [ [link removed] ] has had on demands for electricity worldwide, and this is especially true in the United States, which hosts about a third of the 8,000 data centers in the world, the most of any country. According to Bachar, the world’s existing data centers built over the last 50 years or so consume about 105 GW of electricity per year. With the emergence of AI, the industry will be demanding up to 50 GW more per year within the next five years.
These demands are also reawakening interest in nuclear power as a reliable source. In September, energy provider Constellation announced it had signed a 20-year power purchase agreement [ [link removed] ] with Microsoft that will lead to restarting one of the reactors at the Three Mile Island nuclear power plant in Pennsylvania. The TMI Unit 1 facility was not involved in the 1979 accident—that was TMI Unit 2—and was shut down five years ago for economic reasons. Microsoft will be the sole off-taker [ [link removed] ] at the nuclear facility, and all that power will be used for the company’s AI data center needs.
Earlier in the year, Amazon acquired a 960 MW [ [link removed] ] data center campus near the Susquehanna nuclear power plant that is powered by that 2.5 GW facility, although regulators have since put the brakes [ [link removed] ] on a generous power purchase agreement. Furthermore, Google has signed a deal [ [link removed] ] with California-based Kairos Power to develop small modular nuclear reactors to supply its future data center needs.
Of course, not every data center customer has the throw weight of Microsoft, Amazon or Google to go nuclear. However, it reveals an important fact: As attractive and green as renewable energy may be, data center operators do not consider solar and wind reliable enough for their literally 24/7 power demands. Bachar said hydrogen is an excellent technology for data centers because it can provide reliable, around-the-clock power and can be built on a short timescale.
“Small modular nuclear reactors will be available maybe six, seven, eight years out,” he said. “It will not solve our problem right now, and the AI problem is right now. It’s not whoever gives you a solution in 2035. I was in a conference and I said, ‘We will not have this problem in 2035. We will have different problems, to be sure.’”
ECL’s approach to solving the AI data center problem “now” involves designing facilities powered by hydrogen fuel cells in 1 MW modules that can be scaled up to meet requirements. Hydrogen is either trucked in or extracted from pipelines, where it is mixed with natural gas. The resulting electricity produced in the fuel cell stacks has a byproduct of fresh water that is used to cool the data center’s servers. ECL opened its first facility in Mountain View, California, in June, using compressed hydrogen brought in from suppliers.
In September, ECL announced it is building a 1 GW facility near Houston that will run on fuel cells with hydrogen taken from pipelines on the site. The first 50 MW phase of the project is scheduled to go online next year at a cost of $450 million. The full project is expected to cost $8 billion. Bachar said ECL is financing the project with its partners, and that clients have already signed up to operate their servers in the facility.
Suppliers Open the Taps
If this seems like a lot of money, industry sources [ [link removed] ] say data centers cost between $7 million and $12 million per MW of IT capacity. ECL’s 1 GW project falls into this range, for now at least. One of the ongoing challenges will be procuring hydrogen to make cost-effective electricity, particularly from desirable and environmentally friendly green and blue sources.
If sourcing hydrogen were just a matter of cost per kWh, then the gray variety would do nicely. However, as indicated above, the production methodology means a lot to many of the customers responsible for its increasing demand as an alternative fuel.
Edmonton, Alberta-based Aurora Hydrogen is applying new technology to produce hydrogen from natural gas without emissions and without the need for carbon capture. The company's process uses microwaves to heat methane, producing the desired hydrogen as well as solid carbon, which can be collected for other industrial purposes. Because there is no gaseous CO2, there are no emissions.
Aurora CEO Andrew Gillis said it is his company's patented use of microwaves [ [link removed] ] that separates it from other methane pyrolysis processes, where very high heat is used to break down the target molecule. For those keeping track on their color wheels, hydrogen produced using pyrolysis is coded “turquoise.”
Gillis and his company co-founders, chief technology officer Erin Bobicki and chief science officer Murray Thomson, combined their academic and commercial work on microwave technology and high-temperature pyrolysis of hydrocarbons to pursue a better way to produce hydrogen, which all of them saw as an emerging alternate fuel source.
“There’s two levers of efficiency we pull,” Gillis said “The first is that the conversion of electricity to microwave energy is very efficient. The second is that the carbon particles in our reaction chamber readily absorb microwaves and heat very readily. This enables us to produce hydrogen cheaply without emissions on an industrial scale.”
A key element of Aurora’s strategy is to take advantage of the existing natural gas infrastructure, which is why the company is headquartered in the heart of Canada’s fossil fuel industry. At the same time, the company wants to deploy containerized microwave reactor systems on the site of potential customers. It is finishing a demonstration plant with a nickel refiner in Alberta that uses hydrogen in its process.
“This is an example of direct industrial application,” Gillis said. “It's a great partnership for us as we scale up: modest hydrogen demand where it is needed in an onsite process and an interest in decarbonization.”
Embrace the Slippery Molecule
While demand for hydrogen may be increasing in some sectors, it has remained dormant in many others, such as transportation. There was a time before the ascendancy of lithium-ion batteries when automakers pursued hydrogen as a promising alternative fuel. I attended the Detroit Auto Show in 1997, where Chrysler unveiled its LHX concept sedan [ [link removed] ] powered by hydrogen fuel cells. Lauded for its sleek design, it was also criticized for using the “wrong” fuel. Hydrogen purists wanted cars to fill up on hydrogen. But there was no infrastructure for that, and hydrogen is hard to store. So Chrysler used gasoline, which is easy to store: Chrysler’s technology extracted hydrogen from gasoline so prospective drivers could refuel at existing gas pumps, even though it would have resulted in pollutants as byproducts, which purists didn’t want.
In addition to being difficult to store, the great challenge of hydrogen as a fuel is that it is also very difficult to transport at ambient temperatures. It is the smallest molecule and therefore has a tendency to permeate metals and polymers [ [link removed] ], increasing brittleness and possibly exacerbating fatigue and cracking. The fact is, the effects of hydrogen on piping, containers, valves and fittings are not completely understood. U.S. national labs [ [link removed] ] are partnering with industry to research how to make, move and store hydrogen. It’s a slippery molecule.
This is an important reason why hydrogen producers and consumers are motivated to find ways to co-locate on the same sites. Once hydrogen has to move from place to place, costs go up. Aurora’s Gillis said hydrogen fuel for long-distance trucking makes sense because fueling stations can host microwave methane pyrolysis units. He said the same argument could be made for freight rail. Refueling is also preferable to recharging electric vehicles.
“Our cost-per-kilogram of hydrogen on a per-mile down-the-road basis is less expensive than diesel,” Gillis said, mixing his measures in that Canadian way. “And that is absent incentive. Plus, why carry all those batteries around with you and wait around for charging?”
In early November, New Jersey-based Avina Clean Hydrogen broke ground on a facility [ [link removed] ] in Southern California that will generate green hydrogen and provide a fueling point for fuel-cell-equipped trucks serving the nearby port of Long Beach. The plant is expected to produce up to 4 metric tons of compressed hydrogen per day, enough to refuel 100 heavy trucks, without releasing CO2 or other pollution. However, the economics of such infrastructure are subordinate to California’s rigorous decarbonization goals [ [link removed] ].
Ultimately, industries will embrace hydrogen if it is more attractive than the alternative. The sectors that need it for their particular processes are already taking their fill from a wide variety of producers across the hydrogen color spectrum. Governments and pundits may spur more demand and preferences for some production methods over others. But a true hydrogen economy is waiting for those entrepreneurs who can unlock its potential as a limitless source of energy.

Unsubscribe [link removed]?