TL&DR Summary: blue hydrogen will be cheaper than green hydrogen for the foreseeable future. But the risk remains that the product will be much cheaper if we ignore the fact that it’s not blue, but rather blackish-blue and bruise coloured due to methane and uncaptured CO2 emissions. But that can also be said about green hydrogen, which can only be considered green once we’ve eliminated fossil burning in the areas where it is being produced.
Today a connection asked me if I agreed with Michael Liebreich that blue hydrogen is the “only answer”.
I’ve read enough of Michael’s work, and watched enough of his presentations, to know that he’s never said anything of the sort.
Michael and I have gone back and forth over these topics for many years. And today it’s difficult to fit more than a couple sheets of paper between our opinions in relation to hydrogen and its role in decarbonization.
I think it’s fair to say that both of us agree- as his famous “hydrogen ladder” shows, that the only meaningful decarbonization uses of hydrogen are pretty much applications where hydrogen is already used- as a chemical. The “new” uses of hydrogen that make sense in a decarbonized future, are in a sense, really existing uses of hydrogen, simply expanded. One example is replacing some of the carbon monoxide currently used in direct iron reduction, with hydrogen and electricity, to produce lower GHG iron for steelmaking- what people have been calling “green steel”. Another is dealing with the oxygen in biomass, running the “hydrogenolysis” reactions to effectively burn the oxygen out of the biomass to leave behind a stable fuel that can be used in existing engines.
We agree that hydrogen won’t be wasted in heating, transport, or as an international energy export vector. That “hydrogen economy” stuff is just hyperbole, being put forward by interested parties and their useful idiot hangers-on.
While Michael gives credence to the use of hydrogen for applications like “long duration storage”, I don’t. I think he’s just throwing the hydrogen fanatics a bone, frankly, because the idea of making excess hydrogen in summer to burn in winter, cannot ever make economic sense.
But what to do about the GHG emissions of existing hydrogen production?
That’s another thing that Michael and I agree on. And while I’ve long said that our focus must be on replacing black hydrogen (fossil derived hydrogen- 99% of world current H2 production- is derived from fossils with no attempt at carbon capture and hence is BLACK, not gray or brown) with green hydrogen, Michael has been suggesting that blue hydrogen would be the better choice.
The reasons for that are pretty straightforward: blue hydrogen is going to be cheaper than green, for the foreseeable future. And on that, we agree.
The problem I have with blue hydrogen is that it’s very likely to be called “blue”, when it’s really blackish blue and bruise coloured. And Michael acknowledges that this is true, too. He’s just unabashedly more optimistic about how regulated the industry can be. He hopes that we can demand that any blue hydrogen project be done under stringent regulatory control that ensures that not only is the right technology selected to allow high percentage CO2 capture, but also the upstream methane emissions will be properly measured and abated.
And I have no hope of that at all.
Why is that?
Because of Shell Quest.
Shell Quest- A Story of Public Cost and Partial Carbon Capture
Shell Quest, which I talk about quite a bit in my piece about blue hydrogen:
https://www.linkedin.com/pulse/blackish-blue-bruise-coloured-hydrogen-spitfire-research-inc
…makes quite blackish blue, bruise coloured hydrogen. But I LOVE that project! Why is that? Because it was publicly funded, and hence the data about the project is, largely, public. Or at least it was until 2023. No 2024 report has come out on the Government of Alberta website, and since this is near the end of 2025, it would not surprise me if 2023’s data is the last we ever get.
What Quest proves, beyond a shadow of a doubt, is that “blue” hydrogen, done at considerable scale (1 million tonnes of CO2 captured and sequestered per year), by very competent people (Shell and Fluor, both of which have some of the world’s best chemical engineers working for them), under nearly ideal conditions, still costs a fortune.
Despite the fact that Quest only goes after the CO2 in the syngas stream of the steam methane reformer (SMR), which has to be removed to make useful pure hydrogen anyway- which amounts to only ~55% of the CO2 emissions from hydrogen production in an SMR.
Despite the fact that its target was to capture only 80% of that CO2, and it eventually had to be adjusted down to 78% to avoid burner stability problems in the firebox of the SMR unit.
Despite the fact that heat and electricity to run the CCS equipment is produced by burning fossil gas, with no CCS attempted.
Despite the fact that the other ~ 45% of the CO2, which comes out of the SMR’s firebox flue, is just vented, just like it was before.
Despite the fact that there’s a perfect hole in the ground- a deep saline aquifer- only ~ 60 km away from the plant, so only a short CO2 pipeline needed to be built.
Despite all these advantages, it STILL cost, in 2023, $145/tonne of CO2 captured and sequestered. Those are Canadian dollars however, so that’s a paltry $106 USD/tonne.
Truly Blue Hydrogen is Hard and Expensive
At first blush, you might think: “well, that’s not so bad”.
Typical SMR hydrogen has direct CO2 emissions of around 10 kg CO2 per kg H2 produced. Multiply that out by $106 USD/tonne of CO2 emissions, and you might conclude that blue hydrogen could be produced for an extra $1.06 USD per kg, driving up black hydrogen cost from $1 to $1.50/kg wholesale to perhaps $2.50/kg at most. And since green hydrogen will never get that cheap (see my article about electrolysis to understand why I conclude that):
https://www.linkedin.com/pulse/scaling-lesson-2-water-electrolysis-paul-martin
…you might conclude that blue hydrogen will be cheaper than green will ever be.
And you’d be right.
But you’d also be wrong.
To do truly “blue” hydrogen, you need to do things differently than they did at Quest. And by “different”, I mean “things that are much more costly”.
https://www.linkedin.com/pulse/part-2-ghost-blue-hydrogens-future-spitfire-research-inc
As my piece explains, you need to either run electric SMR or an autothermal reformer (ATR). An electric SMR, with ~ 30% of the energy in the hydrogen, coming from electricity instead of fire. This energy is required at high temperature, but electric heating can definitely do that- it’s challenging, and costly, and must be run continuously as must all truly high temperature equipment- but it’s far from impossible. But while that would require a lot less electricity than electrolyzing water (about 1/3 as much), it would still leave you with the cost, and geological and logistical challenges of CCS.
Or, you can do autothermal reforming, so that the “fire” that you need to supply the endothermic heat of reforming reactions, is produced by partial combustion inside the reactor instead of in a firebox, by feeding oxygen in with the steam and methane before passing the gas over a reforming catalyst. We know how to do that- it’s fully TRL9 commercial technology and has been for decades. But ATR is less efficient than SMR, and would still require clean electricity to make oxygen and to run the CCS equipment.
And you would still end up with the problem of gas purification. Because pure hydrogen is required, and by “pure” I mean CO+CO2 less than 10 ppm total if you want to feed a fuelcell without killing it, so you will end up with streams of “mostly hydrogen” which contain those unwanted gases. In a conventional SMR, they are fed to the firebox where fire sorts them out. But in an electric SMR or an ATR, you must either burn those streams and emit the CO2, or do costly post-combustion CCS on them.
Neither is an attractive option.
Undeterred, Shell and its partner ATCO, are planning a project called Polaris, which will attempt to recover another 650,000 tonnes/yr of CO2 to dump into the proposed Atlas Carbon Storage Hub, which will use the Basal Cambrian Sands as the CO2 repository. This is the same deep saline aquifer that has been used by Quest to store around 10 million tonnes of CO2 so far since 2015. Polaris apparently is going after post-combustion CO2 from the Scotford upgrader, probably including the firebox flue of the SMR. Whereas the CO2 in the syngas at Quest is ~3.8 bar partial pressure, the firebox flue effluent will be somewhere between 8 and 15% CO2 and at atmospheric pressure. And basic physics tells us that capturing CO2 from a 0.08 to 0.15 bar partial pressure stream, will cost a LOT more per tonne of CO2 than capturing it from a 3.8 bar CO2 stream. Carbon capture costs from this project have not been announced, but rest assured- we Canadians will be paying whatever it costs. Sadly, we won’t likely get the cost data from this project, either.
Methane Emissions
But the elephant in the room is methane emissions. The gas industry and gas users alike, simply pretend that it’s not a problem.
In Canada, you can’t get good data on methane emissions per unit of gas delivered. I’ve looked. Thoroughly. The data that there is, was provided by the industry with no independent verification and is totally incredible, i.e. not believable.
There’s nothing magical about gas in Alberta, either. It’s just as likely to be vented or to leak as gas produced anywhere else in the world. There are no magical regulations in Canada that make this unaffordable or illegal, either. And regulatory capture in Canada by the fossil fuel industry is no less of a problem than it is in many other places in the world- though we fortunately have no equivalent to the utterly indecent and corrupt Texas Railroad Commission.
A little methane goes a long way. With a global warming potential 33x that of CO2 on the 100 year timeline, and 84x that of CO2 on the 20 year timeline, it’s a powerful greenhouse gas. And if you don’t ignore it, and especially if you see hydrogen as a transitional energy resource (rather than as a permanent chemical that we will always need), you really should be using the 20 year GWP figure.
Hydrogen produced by SMR without carbon capture has CO2 emission intensity of about 10 kg CO2 per kg H2. Add in methane emissions at the worldwide average estimate of 1.5% of delivered gas, and that jumps to 14 kg CO2e per kg H2. And CCS, and ATR, and even electric SMR if it is operated on grid electricity, make this worse, because these approaches sap the energy efficiency of the process, wasting more gas, and hence dragging along more methane emissions. Howarth and Jacobson’s paper, linked in my article about bruise coloured hydrogen, made this argument more eloquently than I can manage.
Michael Liebreich’s point is that we, as the public, can and in fact should demand that blue hydrogen be produced properly, with high CO2 capture and low upstream emissions. And while we agree that we should do this, there is every incentive for the gas industry to simply ignore the methane emissions, and for government to blissfully pretend that they’re not real too.
In conclusion, we know that it will take very vigorous regulation to ensure that any “blue” hydrogen produced, will be truly blue, rather than the blackish blue, bruise coloured shit the industry will inevitably try to get away with producing and passing off as if it were “blue”. And frankly, aside from being confident that it will cost more than $3 USD/kg, we really don’t know for sure what it will cost. That depends on many things, most notably the secure supply of ultra-low methane leakage source gas.
Green Hydrogen is No Saint, Either
But green hydrogen isn’t necessarily green, either. It can vary from bright green to blacker than black, depending on how it is made, too.
This excellent paper by de Kleijne et al, explains that green hydrogen’s carbon intensity varies greatly with the conditions under which it is made- even when it is to be used for inarguably good applications such as replacing black hydrogen in ammonia production for use as a fertilizer.
https://www.linkedin.com/pulse/how-green-hydrogen-lifecycle-basis-paul-martin-pbhyc
To be truly green, hydrogen must be produced in “islanded” facilities fed only low GHG electricity (wind, solar, nuclear, cold water hydro etc.), with any excess electricity fed to the grid to offset fossil generation. The second a green hydrogen project takes grid electricity as an input, a shell game is underway, by which the operator pretends that they are feeding the project only green electricity, whereas in reality, the electricity the project is being fed, has a carbon intensity identical to that of the grid from which it is drawing. Even under these ideal conditions, the resulting hydrogen has a rather high GHG emission potential, of 2.9 kg CO2e/kg H2 produced, with a range between 0.9 and 4.3.
While this is of course better than the ~ 14 kg CO2e/kg of black H2 calculated using 1.5% methane leakage on the 20 year time horizon, only under the very best conditions would this green hydrogen meet the Hydrogen Science Coalition’s criterion for truly “clean” hydrogen- 1 kg CO2e/kg H2- which is necessary for hydrogen to be produced under conditions suitable for a “net zero by 2050” type decarbonization scenario.
To mitigate this problem, regulators have attempted to put controls on what conditions are considered acceptable for green hydrogen production subsidies. These include the so-called three pillars:
Additionality: hydrogen must be produced only from new renewable sources that would not otherwise be built. This is obviously a large and speculative test, and is really an ideological nonsense on any grid which still burns fossils for more than a trivial fraction of its power generation
Proximity: power used must be produced within a real wire’s reach of where it is consumed. Projects which pretend to consume Labrador hydropower in Nova Scotia, despite no excess grid capacity interconnecting the two locations, would be considered a fraud, for instance, because it really would be a fraud, regardless what power purchase agreements might say
Temporal Matching: instead of taking a basket of kWh over a year, under a power purchase agreement, and claiming that the plant consumed only green electricity despite being grid connected, the power used by the electrolyzer must be green during every hour of its operation. Otherwise, there would be no incentive to shut down the electrolyzer when the grid is dirty
These conditions are restrictive, burdensome from a regulatory perspective, and hence are the subject of vigorous lobbying to dilute them. Green hydrogen producers are in business just like everybody else, and know that they need to make cheap hydrogen to have a hope of having an economic value proposition. If the result is that their green hydrogen becomes olive drab or even an intense green barely distinguishable from black, then so be it- at least in the minds of some project developers.
Green hydrogen, just like blue hydrogen, can be made either well or poorly, with low or extremely high GHG emissions. And regulatory controls are required to ensure that any green hydrogen produced, is meaningfully green.
The difference is that even under rather lax conditions, the resulting hydrogen, which is not particularly green or low in GHG emissions, will be expensive.
Even in the best locations in the world, where there is nobody within economic reach of an HVDC cable who needs electricity, islanded wind+sun fed electrolysis plants are built to make ammonia for export, and built at giant scale- GW, not MW- will make very expensive ammonia. And that ammonia will represent a very high cost per tonne of CO2 emissions averted.
“Turquoise” Hydrogen is No Salvation, Either.
Some have argued that the solution to “blue” hydrogen is to pyrolyze methane to make carbon and hydrogen. I’ve already written about this on the basis of what portion of my considerable experience with clients, that I am at liberty to share publicly. Rest assured, methane pyrolysis will never represent a significant fraction of world hydrogen production. It can be, if done correctly, a good way to make valuable carbon products, with low GHG hydrogen as a valuable byproduct. But every little thing, helps only a little. And throwing away ½ of the energy and ¾ of the mass of the feedstock into low value applications like asphalt extenders, or worse still, burying it in reverse coal mining operations, makes no economic sense to me.
https://www.linkedin.com/pulse/hydrogen-methane-pyrolysis-paul-martin-vacnc
There Are Better Things to Do
Michael and I agree that despite the fact that hydrogen production represents somewhere in the neighbourhood of 3-4% of world GHG emissions (calculation provided in my article here: https://www.linkedin.com/pulse/hydrogen-from-renewable-electricity-our-future-paul-martin/ ) there are much more urgent and better things to worry about, than about whether we use blue or green hydrogen to replace black hydrogen.
Michael’s work shows that there are far lower cost options for reducing GHG emissions, which include cleaning up electricity supply, and electrifying land transport and low temperature heating. And my own work for clients, only some of which I am not permitted to talk about publicly, confirms those conclusions.
Given that CO2 and other GHGs have a “time value” in the atmosphere, pursuing the fastest, lowest cost gains we can make, is the only decarbonization strategy that makes any sense.
And by that thinking, as sensible as the decarbonization of hydrogen production for chemical uses durable post decarbonization might be, they really should not be any kind of near term priority. Just like the various carbon-negative technologies touted by many who have read the various IPCC reports, only some of these strategies make any technical or economic sense- and the ones that remain are really appropriate only for use AFTER we’ve tackled the “easy and cheap” to decarbonize sectors.
Disclaimer: this article was written by a human, and humans are known to make mistakes. And while I’m sure Michael will correct any misperceptions about his positions that have made their way into my piece, I count on my readers to correct anything that I’ve gotten wrong, by providing good references which steer me in the right direction.
However, if what you don’t like is that I’ve taken a dump on your pet idea, you are directed to contact my employer, Spitfire Research Inc., who will promptly tell you to piss off and write your own article.