Reformation (Blue) H2 under 45V

This is part 2 of my deeper analysis on the 45V tax credit for hydrogen. This post may be updated a bit as I am traveling and it might need some tweaks when I am finally stationary.

The 45V rule fundamentally makes it impossible to economically get $3/kg credit for reformation by blocking blending of renewable natural gas. Generally speaking, this is a good thing. An Autothermal reformer to produce hydrogen costs ~$1B to build and at $3.50/mmbtu would use $1B of natural gas over 10 years, and pay $250M[1] in natural gas transport and storage costs for a total nominal CapEx and most of the variable Opex of ~$2.2B[2]. These ATRs produce 500+ tons of H2 per day and over 10 years would generate $5B in 45V credits for a 100% profit of over $2.5B on the shoulder of taxpayers before charging customers for their H2. There over ~100 large steam methane reformers in the US, and if there is a tax credit to make a massive profit by replacing them, they will be, to the cost of $500B to the taxpayers.

45Q – the carbon capture credit that has been in place since 2008 and updated with IRA – allows for $85/ton CO2 capture, would be worth $600M, and is far easier to get.

These credits are meant to soften the expense of clean H2 production – not make a massive windfall at taxpayer expense. Here is a helpful chart to show what it is supposed to be for reformation:

Very rough nominal CapEx and OpEx of 45V vs 45Q credits

The subsidy is meant to cover part of CapEx and OpEx for carbon capture to make carbon capture break-even with no capture– not cover the entire plant and make a 100% profit on top.

What about the upstream and other emissions?

I will get into differentiated upstream emissions in another post, but suffice it to say that it will be hard to zero out upstream emissions (see chart below). In addition, ATRs require significant electricity to operate, and if these aren’t using renewables it makes them emit more than is allowable for any 45V credit. Even if an ATR has three-pillars appropriate power around-the-clock, with upstream emissions the CI of an ATR or SMR are at best 3.5 kg CO2 per kg H2, putting it firmly in the territory of 60 cents per kg H2 for a subsidy – or $1B. Alternatively, if the plant opts to use 45Q (the carbon capture credit instead of the clean hydrogen credit) with its 1500 tons of CO2 per day capture, it can get ~$600M with much more reliability. Generally speaking, the low tier of 45V may be achievable, but is risky, whereas the 45Q is pretty easy to hit.

From the NETL report on emissions from reformation-based H2 production - the right two columns are carbon capture. Even if all upstream emissions are eliminated (pretty much impossible) and all power requirements are made with renewables with 45V compliant renewables (that isn’t easy) - it’s still not possible to achieve $3/kg 45V credit

The original law was never meant to provide $3/kg to reformation – it was meant to do so with electrolysis to move the electrolyzers down the cost curve and make them able to stand on their own without government subsidy. With massively price negative blue H2, there is no place for electrolytic H2. This is a problem because there will be no CapEx or OpEx reductions of blue H2. For electrolytic H2, an infant technology, deployment means moving down the cost learning curve and moving up the efficiency curve, much like solar has. Electrolysis will experience massive CapEx reductions and moderate OpEx reductions from the subsidy, providing a significant ROI on taxpayer dollars.

In other words, spending taxpayer dollars on blue H2 doesn’t move us towards a future where the blue H2 is less expensive once the tax incentive runs out. For electrolytic hydrogen the incentive could drop green H2 from the current $5-11/kg to $2/kg in most places, with some locations actually achieving the mythical $1/kg of the H2 Earthshot[3].

I suspect that the blockage of blending will be challenged in courts, but it likely won’t matter. In the face of a potential $500B unintended tax subsidy, congress will act. Historically, a PTC have never been repealed, but a $500B subsidy to oil and gas would challenge that precedent. More importantly, there is precedent for congress making specific entities ineligible for subsidies. Expect recurring action to prevent blue H2 from achieving $3/kg subsidies.

Now that we’ve set the stage, lets get into the numbers behind the reformation and why the current guidance makes $3/kg economically infeasible.

No blending means $3/kg is economically impossible – even with upstream emission abatement

It will be largely impossible for large-scale reformation to economically achieve the $3/kg threshold – the cost of using any type of RNG for all the feedstock would add more to the cost of H2 than the marginal increase in PTC per kg produced.[4]

The reason is that 45VGREET has two types of data: background data that can’t be changed by the user, and foreground data where the user can have input if it’s validated and audited by a third party. Background data is plant processes. The NREL background data indicates that the emissions of remnant CO2 after capture in an ATR are 0.51 kg CO2/kg-H2 with an additional 0.14 kg CO2/kg-H2 for carbon transport and storage. For SMR it is 0.38 kg CO2 /kg-H2 base with an additional 0.15kg from carbon transport and storage. Owing to this, any “responsibly sourced methane”/low-emission methane/ differentiated upstream pathway (take your pick of marketing name for methane with less upstream emissions) still precludes $3/kg.

Advances in capture technology like cryo capture may rectify this, but cryo capture requires an immense amount of additional energy (with commensurate 45V issues), has a large parasitic load on the reformation process, and is not commercial yet. Even if they were, reducing the background data of an ATR’s 0.65 kg CO2 per kg H2 by the permitted 16% will make it 0.54 kg co2 per kg H2, still not enough to get $3/kg.

Generally speaking, however, this reference will not change

LCOH Price implication of no blending

It takes ~0.17 mmbtu natural gas to make 1kg of hydrogen in an ATR or SMR. This equates to $0.85/kg additional H2 cost for every $5/mmbtu. Current LCFS prices for RNG average around $15-20/mmbtu, meaning $2.50 to $3.50 additional per kilogram if blending is disallowed. The marginal cost is more than the marginal gain in credits.

What if the courts vacate no blending?

With Treasury’s mandated -51 g CO2e/MJ for manure RNG, the CI for all RNG in reforming results in -9kg CO2e/kg-H2. An ATR with carbon capture would need to only a 5% blend of RNG to achieve $3/kg, which would instead results in an marginal increase in LCOH of $0.15/kg H2 for an marginal additional revenue of $2/kg H2 PTC. 

At scale, the impact is economically troubling. With $3/kg PTC, a standard 500 ton per day ATR would generate $1.5 million credits per day, ~$500M per year, and $5B over a 10-year credit period. An ATR of this size costs ~$1-1.5B of CAPEX to build, and at $3.50 per mmbtu an ATR will use $1B in natural gas in 10 years for a profit of $2.5-3B on credits alone before charging anything for the hydrogen. There are ~100 large steam methane reformers in the US that could be replaced with ATRs, making this a credit that could balloon to $500B.

That being said, if the courts invalidate the decision, Treasury has to go back and rewrite the rule, not just invalidate the one part. Depending on who is in office then, it could become more lax or it could find several other ways to make it more strict. Afterwards, it has to go back to congress for re-approval. What do you think Congress will do if they see a rule that will cost taxpayers $500B? I, for one, don’t think they will approve it.

What’s the big picture?

Allowing reformation to get $3/kg H2 will be egregiously expensive. It’s probably not going to happen. Getting $1/kg will require upstream emissions to be abated while being regularly observed and validated by third parties and that all the auxiliary power requirements of production are met with three-pillars compliant power – this won’t be easy and anyone that thinks it will be is going to be in for a world of shock when they realize what upstream verification and validation entails. Getting $0.60/kg is within reach, but will require all that upstream verification still.

45Q is very easy to get, comparatively. The smart move is for large-scale reformation to go that way.

On a funny note, this makes Exxon’s threatening to scrap Baytown Blue over not getting a $2.5B government-sponsored profit really special. Note that this would be profit before charging anyone for their H2. I’m only picking on Exxon here because they were public about this, and threatened to throw their toys out of the toybox in a tantrum if they didn’t get it, but everyone pushing for $3/kg for reformation H2 is in the same boat – they are trying to get an immense profit at taxpayer expense that is well beyond the intended coverage of the costs of carbon capture.

What about novel reformation technologies, like pyrolysis?

If the input is methane, getting $3/kg for any technology will be difficult to impossible. If you are investing in early stage tech and your investments are counting on $3/kg H2 with methane as a feedstock, and their prospects don’t look good otherwise, you should find a way to exit that investment.

[1] For every kg of H2 created, an ATR creates about 3 kg of CO2. So at 500 tons per day H2, it’s 1500 tons per day CO2. At $50/ton to capture and store it (this number is actually very high, it’s less expensive than this in most cases) we get $250M

[2] Cost of capital, fixed OpEx, and some parts of variable cost like water, electricity, and process gases are not accounted for here. These numbers don’t swing the total amount much, and I’ve under-indexed on how much credits this could produce. An ATR can produce 600 tons per day, so the credit would closer to $6B.

[3] This would require total installed cost of electrolyzers to drop to ¼ of where they are now, which is feasible, and 2 cent per kwh renewable power with at least 25% uptime directly connected to the electrolysis

[4] This was confirmed with the analysis team at the DOE that many scenarios were ran, and this path would not produce $3/kg with any combination of factors – including zero upstream emissions and high capture rates.