45V Loophole with blending?
The recently defined hydrogen 45V tax credit may have a massive loophole in its treatment of renewable natural gas (RNG) for reformation vs electrolysis. This is a direct result of doing such tortured logic in the rulemaking in their attempts to turn the 45V credit into a transmission grid credit. This loophole would allow for subsidies alone to pay for nearly all the CapEx and OpEx of a CCUS power plant and an electrolyzer to create market-competitive hydrogen with most of the taxpayer-funded subsidy going to RNG producers, not to the goal of cost reduction for electrolyzers.
I would remind everyone that the goal of 45V was to expand the manufacturing base of electrolyzers and create standardized, modularized parts so the costs come down by 4x with American manufacturing so we don’t cede this ground to China like we did with solar. Using the subsidy to grow the RNG industry is bad because it would incentivize environmentally hazardous manure management practices instead of helping the US come to the forefront of electrolysis manufacturing.
Some background: Renewable natural gas has a negative carbon intensity; if a project uses it, part of the emissions are considered negative. This is why 45V disallows blending on hydrogen reformation - if blending were allowed, it would take only $0.25 extra purchase price of a small amount of RNG to get $2 in additional credit - wrecking the tax credit for everyone. Congress never intended blue H2 to get $3/kg
But because the Biden administration (mainly Podesta in his climate role) decided to hamstring electrolysis in defiance of the law’s original intent, they opened up RNG blending for electrolysis through the power sector. Here’s how it works:
A natural gas power plant needs a construction date no earlier than 18 months prior to an electrolyzer Commercial Operation Date (COD) to achieve the additionality requirement of 45V
For the other pillars, hourly matching isn’t an issue because nat-gas plants produce power whenever they want. Regionality remains an issue - a carbon capture power plant in Texas can’t provide EACs to Maine
The electrolyzer buys grid power in a region with a qualifying carbon capture natural gas plant
The power plant buys gas EACS with book-and-claim for RNG to create an EAC with a negative carbon intensity. For those unfamiliar with book-and-claim, it means an RNG developer anywhere in the US injects RNG into a gas pipeline, and a buyer pays for the rights to claim that RNG as theirs - so it doesn’t have to be directly delivered
Because an electrolyzer can choose hourly averaging of carbon intensity, they only buy sufficient EACs to offset their grid-based emissions to 0.45kg CO2 per kg H2 OR they buy a combination of RNG power EACs and CCUS power EACs (ie without RNG) to tailor their average credit
Collect $3/kg for the H2 produced around the clock
Based on the treatment of RNG in 45V, this has two possible results
Following RNG precedent in 45V: The precedent for reformation is that blending of RNG isn’t allowed, and RNG is considered an entirely separate process. Therefore, a carbon capture gas plant that gets book-and-claim RNG for 5% of its power doesn’t have a lower carbon intensity for all power it produces at that time, it has a much lower CI for 5% of the power produced at that time
Not following the RNG precedent in 45V: The precedent for electrolysis is that averaging is allowed for EACs - and while RNG blending isn’t allowed, the DOE may be forced to do some maneuverings to prevent the existing “separate molecule, separate pathway” RNG rules to try to prevent the scenario I describe here. So a 5% blend of RNG with significantly negative carbon intensity should show a ~15% reduction in overall carbon intensity - then it is much harder to make RNG—>power—>electrolysis play economical, but it will entirely undermine the current 45V logic for no blending of RNG and probably invite a very quickly resolved court challenge. It can make it so that a carbon capture power plant can provide power to jump subsidy tiers for a relatively low cost.
Example to make this clear:
Inputs:
RNG at the 45V mandated -51 g CO2 per MJ carbon intensity in a 50% efficient combined-cycle power plant is about -18 kg CO2 per kg H2
1 mmbtu is 293 kwh, and a brand new combined cycle natural gas turbine with carbon capture is about 50% efficient, so 1 mmbtu of natural gas yields 150kwh of usable power to the grid
It takes 45-50kwh to make one kg of H2 with most electrolysis tech
Assume RNG cost is $25/mmbtu
Carbon capture will be required - any new plants require carbon capture per EPA’s power plant guidelines, and the plant has to be new or retrofitted with CCS to count for additionality purposes
This RNG power came from a natural gas plant with carbon capture and assumed the rest of the power was grid power. But a carbon capture power plant without any upstream emissions abatement produces baseload power with EACs at 5.73 kg CO2 per kg H2 (source: GREET jan 2025 model). So to achieve a mix of 0.45 kg CO2 per kg H2, this project needs 1kg H2 from RNG EACs per every 3.5 kg from CCUS-NG EACs. SInce each mmbtu of power yields about 3 kg of H2, That means 1 mmbtu of RNG power combined with CCUS power yields 13.5 kg of H2 at $3/kg for the price of one mmbtu of RNG. So our $25 per mmbtu RNG cost means a marginal cost of $1.86 yields $3/kg H2 subsidies for a net variable OpEx margin of $1.24/kg.
In addition, for the carbon captured the power plant will achieve $85/ton of CO2 captured - thus allowing for the only way to stack 45Q and 45V credits.
While this example uses a whole power plant, it could easily be just a fraction of a power plant’s output that produces the necessary EACs. A power plant need not have 22% of its entire capacity powered by RNG, it could be just a small fraction, provided the electrolyzer is buying RNG EACs at the 22:78 ratio. In this case, an existing power plant that is providing baseload power could gain even better economics from partnering with an electrolyzer.
Shown here is the basic process for using RNG for the power sector one gas EACs are available. For a standard CCUS plant’s emissions, at $25/mmbtu adds a marginal cost of $1.86/kg to the H2 and yields $3/kg in subsidies at the taxpayer expense - with most of the money going to subsidize RNG production for an inefficient end use rather than going towards moving electrolyzer CapEx down the cost stack
Now let’s make it worse
Let’s assume we have enough upstream leak abatement of methane to make our carbon capture power plant have emissions associated with 3 kg CO2 per kg H2 rather than the current 5.5 kg.
H2 from this EAC will achieve $0.60/kg right off the bat. But if RNG is blended, it only takes a 12% RNG EACs to hit $3/kg for all the H2 - and 1 mmbtu supports 25 kg of electrolytic H2 production at $3/kg levels. For a marginal cost of $1/kg, we get an additional $2.40/kg now - for a net variables OpEx margin of $1.40/kg.
As mentioned above, the economics get even better for a plant that is already built and has extra capacity - most of the CapEx part of the cost stack will be covered by the other offtake so there is more margin available in the LCOE and thus LCOH.
Not available immediately
First, this would require RNG gaseous EACs are available. 45V guidance lays out what this takes, and it is pretty onerous. It will be several years before they are available. Second, it requires that the three pillars are in place. Third, if electrolyzer costs come down significantly by 2030 then the need for shenanigans like this will be obviated. Fourth, and the big one, this pathway isn’t represented in 45VGREET - the model that is required for getting 45V credits.
The absence of this pathway in 45V-GREET quite possibly means that the writers of the rules did not consider this path, and it’s a loophole.
45VGREET, however, allows for submitting a pathway if it is not represented in 45VGREET. This is a Provisional Emission Rate process. Once a project hits FID, it can submit this specific pathway that includes a qualifying power generating asset using RNG.
Here are the steps to get there if this isn’t addressed in 45VGREET:
Wait until RNG-EACs are actually available (this g-EAC trading system proposed in the actual 45V draft doesn’t yet exist)
Hit FID with an electrolyzer that plans to get support from a CCUS plant that can use book-and-claim on RNG (hitting FID is a requirement for
Apply for a PER
Watch the DOE struggle to prevent this ridiculous subsidy while being hamstrung by the ridiculous mental gymnastics that Podesta’s team pulled in writing the 45V rules - because they made clear that RNG is allowed, can’t be blended and has to be a separate process, and didn’t explicitly disallow it as a feedstock. Just to reiterate, this mess is not on the DOE - this all all Biden’s executive office.
At large scale this would be a mind-boggling cost
The maths (nominal, not discounted) show that building a CCUS power plant just to power electrolysis is a great investment, at the expense of taxpayers. Shoring up an underutilized CCUS power plant could make eve better for the owners - at the expensive of taxpayers.
To keep things simple, the starting example will be a power plant that solely is for hydrogen. Assuming a grid CI with a resulting 10 kg CO2 per kg H2 were blended with EACs from CCUS-RNG power to achieve 0 CI – IE around a 40-60 blend.
45V credits: This 1GW of power would generate on the order of 400 tons of H2 per day for $420M in credits per year and $4.2B credits over the ten year period (1.4 MT of H2 in that time).
45Q credits: The entire power plant should generate about $2B in 45Q credits over the 12 year 45Q period, but assuming 50% of those credits are eaten by transport and storage costs it’s more like $1B net
Power plant CapEx: These power plants cost about $3B for 1GW, and we are going to do a rosy assumption that they can plug into a CO2 pipeline
Grid CapEx: Grid upgrades for 1GW could be on the order of $1B if it’s local, or up to $3B if it’s a long distance power line (using grainbelt express and
Electrolyzer CapEx: 1GW of electrolyzer would be $2B today or probably $1-$1.5B in 2030 when CCUS plants and RNG EACs are available
Power plant OpEx: ~$4.3B. at ~8400GWh per year, a 50% efficient 1GW plant would use 56 million mmbtu per year. If 22% were $25/mmbtu RNG and the rest were $3/mmbtu natural gas, the RNG would be $3B over 10 years, and the standard gas would be $1.3B over 10 years
45Q OpEx: Assume 50% of the 45Q credits are consumed by transport and storage of CO2, so $1B.
Very rough results: LCOH around $1.50/kg H2, plus or minus 50 cents, which puts it in competition with grey H2. Subsidies are $6.2B in credits for $8-$11.5B in CapEx and OpEx. On the low end, that is a gap of ~$2B for 1.4MT of H2, which is where we get our $1.50/kg
It’s actually worse than this. These subsidies are for 10 and 12 years, the assets are 25-50 year assets. Pretty much these subsidies could provide the H2 for one or two ammonia facilities and all the H2 and power assets would be paid for by subsidies within 10 or 12 years. The assets could operate for 15-40 more years with all their CapEx paid off, meaning massive ongoing profits, all at taxpayer expense.
If this goes forward, the taxpayers are functionally subsidizing RNG in order to expand electrolysis. This is the exact issue we would have seen with reforming getting $3/kg, just with much higher efficiency losses in the power production and electrolyzers.
What should have happened with the money
Instead of paying $6.2B in taxpayer credits for a mere 1GW of electrolysis, grandfathering or hourly carbon intensity averaging without the asinine annual emissions requirement could have resulted in that $6.2B putting out 6GW or more of electrolyzers in 2030.
There is only one major barrier here, and cost barriers are overstated
Geographically, carbon capture and storage is not available in all regions. It’s available in the places where highly subsidized large-scale H2 production would make sense, however, such as California and Texas.
A commercial barrier that can be overcome more easily is that and in many regions the owners of the CO2 pipelines are charging significantly higher fees than break-even plus a small margin would imply. We’re talking about Exxon. When Denbury owned the CO2 pipeline in the gulf, projects were quoted $35/ton as the price to move the CO2. I’ve heard that when Exxon bought Denbury the price increased to closer to $60/ton. This is not a physical barrier, it’s a commercial barrier.
Carbon transport and storage outside the gulf may be more expensive, and in these cases the economics change - but the fact still remains that the subsidies can pay for CapEx and OpEx of an electrolyzer and an entire power plant with carbon capture within a decade while still producing reasonably priced H2.
Is a full discounted cash flow model of this coming?
Not any time soon. This is a very specific model and I would have to be accordingly compensated to make that.
