45V - Hourly Averaging
Hydrogen 45V Analysis – Part 1 – Three Pillars and Averaging
Treasury released the final 45V guidance a few days ago on Friday January 3. There were some surprises, but the major points are:
1. Treasury made it almost impossible to get the full $3/kg for reformation hydrogen. This is a good thing, because this would have become a $500B credit that would have been cancelled (next post)
2. The three pillars mostly stand, with a carve-out for nuclear, for states with high decarbonization goals, and a slight amelioration to allow some grid backup without wrecking all project credits
3. Litigation risk remains high – particularly on the reformation side given how much money is at stake. This will continue to have some chilling effect – mitigated by the carve-outs from the three pillars
4. CCS in power production is allowed to produce credits for 45V and get 45Q carbon capture credits – the stacked credits can be quite lucrative
This article is about some big picture “so what” background and the hourly averaging option for carbon intensity with the time-matching requirement for the credit.
Short version: the three pillars will make electrolytic hydrogen hard. Over the next few days I will post analysis about some of the ameliorating factors.
The big picture – what is at stake for a developer and why current 45V remains a risk
The two year delay of 45V created a chilling effect in the H2 space by preventing projects from reaching Final Investment Decision (FID). Projects that could have gone forward without 45V were no longer viable because 45V adherent projects might have significant competitive advantage – but know one could know until the guidance finally arrived. The uncertainty around how strict the December ’23 guidance was – combined with how vulnerable it is to litigation – means a project build adhering to strict 45V would be blown out of the water in cost if should parts of 45V be relaxed. The asset would then be stranded. Thus, limited investment with 45V uncertainty.
The current iteration of 45V doesn’t abate these risks in most locations. The chance of litigation and overturn resulting in a project that becomes non-competitive in a post-litigation 45V structure is high in many or most cases.
Smart analysis is necessary to identify projects and locations where the risk is low. If there are sufficient volume of these projects, it may be enough to spur the US electrolyzer industry to reduce cost sufficiently to allow for low cost islanded H2 production.
The real issue with the three pillars – it means we only have H2 when the weather is cooperative
The variable renewables and hydrogen storage problem
All hydrogen offtake requires a reliably delivered supply. Without grid backup, renewable powered hydrogen production requires up to several days of hydrogen storage – often more expensive than the electrolyzers if the H2 provider promises 98% uptime[1]. The additional cost makes the delivered cost of hydrogen significantly higher for new end uses outside of existing networks.
Time-matching in the new guidance
In the draft guidance, emission offsets (EACs) needed to match hourly with H2 production, but emissions were averaged annually. That means if a project is producing 95% clean H2 but then has 5% grid backup, nearly all my credits go away, because it’s all blended together.
The new guidance is complex but addresses this to some extent. Provided the annual average of emissions is less then 4 kg CO2 per kg H2, the option for hourly averaging for CI opens up. That being said, it’s not pretty. I crunched a series of scenarios where H2 is made by renewables part of the time and uses grid backup the rest of the time.
In the chart below, the percentages are the uptime you are aiming for with your zero emission power, IE renewables. On the top row is the kg CO2 per kg H2 you would expect from making H2 from the grid average power mix. The percent on the left means “this is what percent of the year the project has EACs for.” The colored percents in the box are how long a project can have uptime with EACs plus that grid mix as backup before invalidating all credits. Everything in the red means extra storage is necessary or customers get random H2 deliveries (note that this doesn’t work). Note that most of the US is currently around 20 kg CO2 per kg H2 using grid power, and solar is about 25%, so we’re all the way at the top left.
For all the red zones, the three pillars create a major variable production problem that remains unaccounted for. The green zones indicate where electrolytic production can continue, albeit at lower or zero subsidy, while maintaining firm H2 production without invalidating credit for EAC compliant H2.
The red zone is where our problem lies. Currently nearly all of the US is in the red zone in terms of trying to use renewables and grid.
45V strongly favors incumbents – who have held the sector back for decades
All is not lost! An upcoming post discusses the other two mitigations to 45V – both dialing back on additionality
That being said, this iteration of 45V three pillars favors existing hydrogen distribution networks. As an example, a 200mw electrolyzer with 30% uptime would produce 30 tons per day of H2. With this, an industrial gas companies can turn down a single 300 ton per day steam methane reformer by 10%[2], which doesn’t upset operations of the reformer at all, and drop the electrolytic hydrogen into their existing gray hydrogen distribution network without any disruption. [AK1] This allows them to use their distribution network as storage for drop-in replacements of grey hydrogen by doing a one-for-one switch. It does not expand the hydrogen economy, however, and it prevents new end uses as the industrial gas companies have shown a stark inability to innovate and operate in the hydrogen economy – just look at California where Air Products can’t bother delivering hydrogen to fueling stations despite saying they won’t build new projects because they already have too much hydrogen capacity.
New entrants and new end uses are barred by the existing hydrogen entities from sharing existing infrastructure, so they can’t just do drop-in replacement of existing hydrogen. They need immense amounts of expensive hydrogen storage. On the demand side, new offtake that isn’t industrial scale struggles to get deliveries of hydrogen from industrial gas companies – and the headwinds of 45V that don’t allow for reliable electrolytic production won’t help.
The math of why hourly averaging with an annual average cap doesn’t work
If your eyes glaze over with math, it’s time to stop reading.
45V guidance allows for hourly averaging of H2 CI provided the annual average of all H2 production from a facility is less than 4kg CO2 per kg H2. In practice, this means a plant that operates 25% of the time with solar EACs to achieve $3/kg for this H2 can operate an additional number of hours with grid connection and still get the $3/kg for only the solar-EAC backed H2, but only if the further produced H2 doesn’t bring the annual average CI beyond 4 kg CO2e per kg H2. This could, for example, allow one tranche of H2 production to get $3/kg, another tranche from the same facility to get $1/kg, and another tranche be ineligible for any subsidy without disqualifying the subsidized H2.
Achieving an annual average of 4kg CO2 per kg H2 allows for a choice of hourly averaging to obtain higher credits for lower-emission H2 production. After producing 45V compliant H2, the amount of additional H2 that can be produced from grid support is given by the equation:
where
h is allowable grid H2 production
x is the amount of $3/kg 45V compliant H2 produced
m is the emissions from the 45V compliant H2
z is the kg co2 per kg H2 of grid electrolysis
As an example, if 100kg of H2 is produced with 0.45 kg CO2/kg to know how much can be made with grid hydrogen at 20kg CO2/kg H2, the equation above shows us a result of 22kg max. With variable renewable H2 production, an additional 20% is insufficient to cover times daily consistent production needs, much less multi day when there is weather.
If we instead plug in a grid intensity of 5.25 kg CO2 per kg H2, we get 300 kg of additional production. If we are using 25% base solar, this extra 300 kg of production will fill out the other 75% of the time. The chart above plots a whole range of scenarios.
Again, for that chart, we are in the top left territory. Which, referring back to my long-ago analysis with extremely outdated (and low) cost of renewables (thanks, data centers) requires significant storage:
The chart here shows 4 cents per kwh, so just add $2 to each cost stack. The important part is that the black line is hydrogen costs with storage and the red line is hydrogen costs without storage. Owing to 45V restricting production, electrolytic hydrogen becomes significantly more expensive – we’re talking $1.40/kg more to make and store it, compared to the current total cost of ~$1/kg.
We’re getting long here, so I’ll follow up tomorrow with some important discussion on the blue side.
[1] Government analysts have shown that 98% uptime from renewables-only H2 production would require one week of H2 storage in most of the US, and 99% uptime would require up to two weeks of storage. One week of storage for a one-ton per day project is seven tons. At $500,000 per ton for storage it is $3.5M for storage. The total installed cost of a 2MW electrolyzer that can produce 1 ton per day of H2 with a 90% utilization is on the order of $4M and dropping.
[2] Turning down a SMR or ATR by only 10% is very easy and won’t greatly affect broader H2 economics. In some plants, turning down beyond this amount can begin to cause issues with efficiency and operations, and turning down beyond a certain point will cause damage to the system
