Gaseous Hydrogen (GH2) systems costs now much lower than Liquid Hydrogen (LH2) part 2

May 1

Liquid and Gaseous H2 (LH2 and GH2)– Refueling differences

In my last post I ran the numbers on moving hydrogen as a gas vs as a liquid –the liquefaction process is an egregiously expensive way to move H2 compared to using new compressors and gaseous tube trailers. At $250M for a 30 ton per day liquefaction facility compared to an estimated $20M for 30 tons per day of compressor (20 bar to 550 bar), liquefaction is not cost competitive for most use cases.

I received a lot of comments specifically on hydrogen fueling for vehicles regarding material advantages of liquid refueling stations. This article covers why most of these advantages are obsolete or make up only a fraction of the cost difference created by liquefaction.

Up until the new tube trailers and compressors came out just this year, liquid hydrogen deliveries to fueling stations had several major advantages that resulted in much lower costs and often better station performance compared to gaseous (first column below). Of those advantages, only one remains, and in almost no cases will it material compared to the $4/kg cost stack addition that liquefaction and vaporization losses add (columns two and three).

Chart depicts costs of legacy hydrogen stations (column 1) costs excluding H2 production costs. Column 2 shows distribution and station costs for commercially operating systems in the EU. Column 3 shows liquid value chains - ‘real world’ includes H2 losses in transit and delivery that end up in contracts. Numbers show are *levelized cost*. Prices paid by customers will include markups at delivery and at refueling.

More details on the refueling stations in later sections, but the main points here are that the design, throughput, and inefficiencies of the early gaseous stations resulted in staggering CapEx (2x compressors at every site because they break so often, expensive 900 bar H2 storage to address the terrible compressors, pre-coolers that leaked prodigious amounts of energy to the environment, etc). LH2 had a profound cost advantage. The new gaseous equipment not only eroded that advantage, it blew past LH2 costs to be more than competitive in commercially operated systems that already exist in the EU.

While LH2 stations can cost less in many cases the liquefaction upstream burdens the entire cost stack in ways that can’t be overcome. Major takeaways first:

LH2 refueling stations focused on space constraint advantages are an edge case, not a market: For a 1 ton per day fueling station the additional land for GH2 to do a 40’ drop-and-swap would have to be $10M to make the LCOH at a GH2 station cost as much as an LH2 station. For the space required, this is 40-80x the cost of land in prime Manhattan. LH2 will work in cases where building at existing fueling stations is the only option – something that isn’t required anywhere.
1. Corollary and history lesson: Trying to build GH2 stations on existing fueling sites is highly unlikely to be economical. Achieving sustainable H2 prices will require new-build stations on lots designed for drop-and-swap. Economic stations need more space than existing gasoline stations provide. This is not a hypothetical - this is a major part of why Shell failed in its commitments and state funding to build H2 refueling stations in California. Either management or Shell H2 bizdev committed to to building refueling stations only on existing shell retail station sites. There isn’t enough space on these sites to do drop-and-swap or cost-effective gaseous deliveries, so there was no way to realize cost savings. With the Industrial Gas Companies charging $20/kg for liquid value chains (still with gaseous delivery - just in paltry volumes), there was no economic way to run a station. The changeover of leadership at Shell cut these as wasteful in part because the “build only on existing sites” precluded gaseous deliveries and was a massive money loser.
LH2 driving distance advantage: supply lines from production would need to be over 1000 miles to the end use for LH2 to be more cost effective than GH2
There is nowhere in the US where the longer supply chains permitted by LH2 provide a energy arbitrage opportunity: Electricity would need to be $0.12/kwh lower at a theoretical LH2 electrolysis production site compared to a local GH2 production site to break even with local GH2 production. There are few places in the US where the average total price of industrial power is this high, much less the delta between two sites.
Compression and pre-cooling: the new compressors are highly efficient. Combined with drop-and-swap, they eliminating the need for 900 bar cascading storage – reducing station costs by over $500,000 – and making the CapEx on part with any cost savings from LH2 having lower space and electricity requirements. The high pressure of drop-and-swap also reduced the cooling requirements for fueling – making the cooling far more efficient

Unless the cost of liquefiers comes down by a factor of nearly 5x there is no way for LH2 to make up this gap. The only mitigating factor would be consumer preference allowing for higher pump prices for LH2 – but with the H2 ended up as compressed gas in their car either way, consumers will not pay this.

The investing “so what”

This change opens up a massive opportunity for investment through the entire value chain.

Project financing: The potential thick margins for fueling while hitting competitive prices with gasoline could allow sufficient cash flow to entice project financiers. While project financing usually requires long term offtake, they will fund projects with projected free cash flow that is over 2x debt servicing.

More accessible investments: Requiring liquefaction while no liquefiers are commercially available disaggregates production and offtake while concentrating power in the three companies capable of making liquefiers. Moving to gaseous systems means project investors can invest in production, distribution, and offtake without being beholden to liquefaction plant gatekeepers.

More supply chain investment opportunities: Liquefiers are really only made by three companies and they aren’t accessible to investment. While new liquefaction tech may become available with new startups, it’s not a sure bet. Comparatively, the H2 compressor space and tube trailer space is accessible and will become more crowded. Opportunities to invest in the supply chain from a PE, VC, and M&A standpoint are much better here.

More investment size opportunities: Liquefaction projects are have massive headline size. Gaseous projects can get that big – with pipelines – but generally can scale from $1M to $100M – providing a lot more scale of opportunity depending on risk appetite.

More induced offtake investment opportunity: Gaseous systems are mostly modular and thus highly flexible. This flexibility will allow more offtake opportunities to arise – an area where VC and PE could see immense opportunity.

Into the details - skip if you don’t like the engineering and operations

Major LH2 cost adders

Liquefaction CapEx is 10x that of compression. Liquefiers come in only a few sizes - 15 or 30 tons per day (tpd) is typical. A 30 ton per day facility is $250M. 30 tons per day of compressor should be $20M.
- Liquefaction maintenance is much higher than compressors owing to the 10x CapEx - the parts are expensive and a single broken component will cost more than the entire compressor line
Liquefaction OpEx is 10 that of compression . A 30tpd liquefier uses 12kwh/kg to liquefy H2. State-of-the-art compressors use less than 1kwh to compress H2 from 20 bar to 300/500 bar for transport.
Liquefiers waste energy when H2 isn’t available because they can’t easily power cycle. 45V compliant hydrogen production forces intermittent production - new compressors can load follow and accommodate without being destroyed (the old ones break under these conditions). Liquefiers can’t turn on and off frequently without breaking - so they keep cycling cooled hydrogen and using liquid nitrogen even when they aren’t actively outputting liquid H2 because no new H2 is being produced.
Vaporization losses on transport and delivery: Vaporization losses are one of the largest hidden cost contributors in liquid value chains. They are not included in most models and marketing material - but are included in the bill of anyone buying LH2. Buyers have to pay for losses in H2 during and at delivery. This may be a solvable problem - but currently LH2 trucks vent H2 in transit, vent H2 to help push LH2 while filling some stations, and must vent LH2 in the receiving storage to pre-cool the receiving storage. Each additional stop incurs more losses - so a 3.5 ton drop-and-swap would incur almost no losses, but delivering to five 500kg storage stations could lose up to 30% of the H2.

Any savings in LH2 refueling can’t make up for these.

Space savings for LH2 stations – relevant but an edge case

In a location where there isn’t space for gaseous trailer drop-and-swap or even stationary storage. LH2 has a smaller station footprint. This will only be sufficient in edge cases for LH2 to maintain an edge. In nearly all locations, LH2 will remain much more expensive and thus vulnerable to a neighbor opening a GH2 refueling site at much lower fueling cost. The cost advantage of gaseous supply chains is so vast that the land for a new one-ton-per-day GH2 refueling station would have to cost $10M for the 1/30th an additional acre to make up for the added cost of liquefaction. In other words, there is no place in the world where a refueling station would be located where the extra space required for gaseous could not be met by a competitor buying a nearby lot.

Nonetheless, these LH2 stations will exist, despite hydrogen fueling prices up to 50%-100% more expensive than gaseous hydrogen. Gasoline refueling stations in LA on Fairfax, Beverly hills, parts of Manhattan, and many other places show people willingly pay 50% higher fuel prices to refuel near home rather than driving a few miles to less expensive fuel.

Low-volume refueling stations between cities will also possibly need LH2 – because the driving distances to deliver fuel will be so far that the marginal per-mile LH2 delivery costs offset the much higher liquefaction CapEx.

Most refueling stations, however, will move towards much simpler and more cost effective gaseous ecosystems. Especially heavy duty stations. Heavy duty trucking is entirely driven by profitability, there is no consumer choice preference. If a heavy duty fueling station is offering fuel at $6/kg H2 (~30% margin) using GH2 value chains delivery or $12/kg for liquid delivery (~30% margins), the lower fuel cost will win and the refueling station will figure out how to work with greater space requirements of GH2. If they don’t – competitors will figure it out undercut them.

These GH2 cost changes are already commercially deployed in the EU at over 50 sites where gaseous refueling stations with 96% hardware uptime are being installed for less than $2M – equivalent or less than the cost of most liquid H2 stations. The new “type IV” carbon tube trailers, only recently available in the US, have also been successfully deployed in the EU.

The result is that the levelized cost of gaseous distribution and refueling at a 500kg/day station has dropped from $12/kg to slightly above $4/kg, while the combined with liquid hovers around $9/kg:

With the gaseous refueling station making up a significant reduction from around $8/kg to around $3/kg:

What’s changed to make gaseous fueling more cost effective?

Five major changes with gaseous reduced fueling costs:

Production Compression: Piston compressors solves the seal-degradation problem by using better seal materials or designs. Diaphragm compressor solved the spalling problem on the diaphragm head by changing to corrosion resistant steel and having dynamic oil pressure valves (this is deep engineering issues that plagued stations through 2020). Compressors are 1/10th the CapEx of liquefiers per ton H2 moved and use 1/5th the energy of liquefiers.
Distribution trailers: The major changes are moving from steel tubes (AKA type I) or aluminum-lined carbon tubes (AKA type III) to polymer-lined carbon tubes (AKA Type IV). These allow for very high pressure hydrogen to be transported in significant volume at low cost. Trailer H2 volume has gone from 200kg to over 1 ton as a result. The new tube trailers move 1/3 the H2 of a liquid trailer – but they don’t experience the vapor losses or the massive upstream overhead costs of liquid value chains
Drop-and-swap: Drop and swap combined with Type IV tubes allows for the delivery of high-pressure hydrogen very quickly to an end use. For a high-pressure end use like refueling stations, this means that the average starting refueling pressure is 140 bar for 300 bar tubes or 240 bar for 500 bar tubes. This has two major effects
1. Reduced compression needs: The average compression ratio for fueling drops to only 4-6x – drastically reducing compressor and energy requirements
2. Reduced pre-cooling needs: The much smaller average compression ratio compared to old systems means less pre-cooling energy is required to be J2601 compliant
New fueling hardware:
1. Compressors: The new compressors are demonstrating much better reliability than those used in stations up until a few years ago – allowing much better uptime at lower cost. They use the same compressor seal tech as the newest production compressors
2. Pre-coolers: new pre-cooler designs largely eliminate waste heat to the environment reducing electricity for cooling needs to less than 0.4kwh/kg. The new pre-coolers are also extremely high reliability
Elimination of high-pressure cascade tanks: Older stations required 900 bar tanks on-site to be able to fuel quickly. These are extremely expensive. The new compressors can push H2 quickly enough to fuel in-line – IE directly to the vehicle without needing these high pressure tanks. The elimination of this system saves nearly $500k-$1M from a small station.

Refueling advantages with LH2 are either mitigated with new technology or obviated by cost factors

LH2 refueling had enjoyed four main advantages at refueling stations:

No need for pre-cooling
Less energy required for compression
Less space requirements (already covered)
Lower cost of hydrogen delivered from 600-1000 miles away

All of these are either mostly or entirely eliminated by new hardware available

Pre-cooling

Covered above – the new systems are effective and inexpensive. Pre-cooling adds a total of about 10 cents per kg to the cost of GH2.

Less energy required for compression

The new compressors approach 90% thermal efficiency. This makes it take less than 1kwh/kg to fuel – even less with higher pressure tube trailer deliveries. The added cost here with both CapEx and OpEx is around $0.10-$0.20/kg, barely moving the needle on the costs compared to liquefying H2.

The new valves on the gaseous compressors also push their reliability into the mid 90s. The liquid compressor valves remain and unresolved problem – introducing reliability issues into the fueling stations.

The higher use of electricity for gaseous compressors contributes to about $0.10-$0.15/kg higher cost compared to LH2. This is offset by the much higher reliability of the new GH2 compressors – the LH2 compressors have still not resolved their valve issues and have lower uptime and thus higher maintenance and lower station utilization.

Better at longer supply chains

LH2 can be transported over longer distances at less expense than GH2. The break-even point is around 1000 miles. New modular small-scale H2 production equipment allows for distributed production to entirely obviate the need for long supply chains. In most cases, new smaller production hardware – such 10 ton per day steam methane reformers, pyrolysis systems, and even local electrolysis can produce and distribute hydrogen locally. While there are certainly cases where long supply chains have cost advantage – like electrolysis in areas with extremely low power cost – the driving distance needs to be very far to overcome the advantages of locally produced hydrogen moved at high pressure. The longer supply chain opportunity for liquid H2 really has no more applicable use cases other than extremely expensive backup supply.

There is no “lower electricity input cost” arbitrage opportunity anywhere in the US that would justify liquefaction supply chain costs. Liquefiers add $4/kg to the cost stack. Every 2 cents per kwh adds $1 to the electrolysis cost stack, and another $0.20 to the liquefaction cost stack. To break even compared to gaseous systems, the cost savings to justify liquefaction would need to be a 10 cent per kwh decrease in power price to cover the additional costs of liquefier alone. Adding in the presumed 500+ mile supply chain to get these lower power prices would also add $1/kg to the cost stack – pushing that trade-off to 12 cents per kwh lower power price required to make LH2 more cost effective than GH2.

Is this set in stone? What would need to happen in the LH2 ecosystem to reverse this?

Liquefiers would need to come down in price by at least 3x-5x and also nearly halve their energy usage. These are metrics that compressors for gaseous have accomplished in the past decade.

I am not counting on these changes happening in the short term for liquid H2. There is no indication in any adjacent market that the expansion turbine systems for superchilling gases can achieve these cost reductions – hundreds of deployed Air Separation Units to produce pure CO2 and pure oxygen worldwide have not come down in cost – there is no evidence that the much more complicated hydrogen liquefiers of this scale can achieve better.

A breakthrough energy technology for cooling and liquefying hydrogen would need to hit market to alleviate these cost issues.

Until then, I will continue to recommend to my clients that they move towards the gaseous systems – doing otherwise risks a competitor with much lower systems costs moving in down the street.

In addition, liquid transport needs to deal with the venting issue. Right now LH2 needs to cool the receiving tank to -250C and is gasified and vented in the process. Moreover, the overpressure from some of the LH2 becoming gaseous in the trailer is used to push the LH2 into the receiving tank – and this plus any other gasified H2 needs to be vented before getting underway again. An LH2 tanker will lose 5-15% of its H2 at every stop. New equipment to rectify this needs to be made.

One caveat: Industrial gas companies and oil companies are working on 500 ton per day liquefiers. If these come to fruition, the LH2 costs could become more competitive. The 5x less expensive is highly suspect - an engineer at PraxAir told me that a doubling of size only nets 10% CapEx gains on an scale with a liquefier, so five doublings would get a 40% cost reduction - not the 80% required. There are some potential step-gains rather than incremental gains with liquefiers splitting the cold boxes (a level of engineering detail I should not even mention here), but 80% reduction is still a big gap. Worse: If one of these companies makes 500 ton per day system, it’s for someone else, not for mobility, and the leftovers will go to mobility. This is not a path to grow hydrogen for refueling. Investing implications: a 500 tpd system is not an ecosystem that any investor has access to. This is reserved for oil and gas companies and industrial gas companies alone.

Important: levelized cost is not cost at the pump

Typically levelized cost assumed only a 7%-10% cost of capital – margins aren’t included. Even more importantly, levelized cost doesn’t account for commercial shenanigans. Many commenters on the last post said that incumbents are offering LH2 for lower prices than I’ve listed. This isn’t true. They are marketing LH2 at lower prices, they are contracting the LH2 at much higher prices. In other words, in bizdev conversations they may say they can achieve $5/kg but when it comes to the contract, it is $20/kg or even $36/kg in one bus contract I’ve heard about.

I’ve seen contracts with delivered hydrogen prices in the EU of €6-8 per kg using gaseous hardware.

Part of the reason for this high contract price is that the barrier to entry with LH2 is impossibly high – only three companies can do it, and they won’t sell liquefiers. They will only sell liquid hydrogen molecules – and when you are locked into their ecosystem they have shown they will increase prices. No one else can provide LH2 – so they can get away with it. Chart is working on making itself a fourth provider for liquefiers and hopefully will sell the hardware. If they do, LH2 could move back into competition if they can provide a much more cost effective liquefier.

That being said, the barriers to entry with gaseous are near zero – the compressor and tube trailer companies will sell the equipment to anyone with good credit. I know because one of them provided a final RFP for a $1M project. With barriers to entry this low, expect price competition to emerge – preventing the 6x marketing/contract price divergence that project developers currently deal with.

None of this is a thought exercise - LH2 has failed the hydrogen ecosystem for the past decade already – it will continue to fail

The liquid supply chains have utterly failed to reliably supply hydrogen to fueling stations, resulting in the meltdown of hydrogen fueling in California. Toyota was sued for lack of hydrogen availability in cars. This is not because the cars were bad. This is not because the stations were bad (though they were – but the stations had redundancy). This is entirely because the industrial gas companies and their liquid H2 supply chains entirely failed to deliver on their contracts to deliver H2 to the stations.

Similarly – Plug Power was unable to get enough liquid hydrogen to sell to its customers. So it decided to vertically integrate into liquid, electrolysis, and all over the rest of H2. It is now on the cusp of being de-listed from NASDAQ and circling the drain towards bankruptcy. Either way it was in trouble – either the liquid supply chains could not provide the hydrogen it needed, or they had to expand into unfamiliar territory and are now paying the price.

The commercial reality is that liquid hydrogen providers have failed to provide reliable cost-effective hydrogen to refueling. Gaseous providers in the EU have already deployed the new gaseous hardware with lower costs and better results. Expect more of that here in the US.

Appendix

Model parameters:

Distribution is same as last time

Economic:

15 year capitalization timeline w/ 1 year of construction

2% inflation

7% cost of capital (assumption is balance sheet building - loans may cost more depending on credit and equity will cost a lot more)

No contingency included - assuming a later-stage build rather than first builds

10 refueling sites rated for 1440 kg/day each (60 kg per hour) and dispensing 700 kg per day

Fueling:

Gaseous

$2M per station - $1M per 60 kg per hour compressor and $1M for construction

$1M for an extra trailer to have a second day of H2 storage and associated safety walls

-trailer is considered parked at the production site for space reasons

700kg per day (lower than the average daily usage at Shell’s only good station in Torrance in 2020-2021)

Stations swaps GH2 every day which means 300 kg of H2 gets moved and not used - this is a significant cost addition

Two fueling positions

0.8kwh/kg for fueling (commercial data with systems in the EU)

0.4kwh/kg for pre-cooling (commercial data in the US already)

8kwh/day standby power

1% losses from purging (commercial deployments show this is a max, it can be lower)

$100k/yr maintenance

$150k/yr land rent

Important note on gaseous:

The max fueling rate changes throughout the day. With an overnight fuel swap, the storage pressure is 300-500 bar requiring only 2x compression - the compressors can go fast - dispensing 10kg in five minutes. As the hydrogen is used more and pressure drops, the fueling rate will drop to a minimum of 60kg/hr

Liquid

$2M per station for 60kg per hour compression (from the California applications for funding for stations) these costs go higher if the station needs to bury the storage for space savings

$500k for an additional day of storage to make even with GH2 analysis

700 kg per day demand

All non-vented H2 is used compared to GH2 whereby a significant amount goes back to the production site in the drop-and-swap

No pre-cooling energy requirements

0.2kwh/kg for fueling

$150k/yr maintenance (liquid systems require more annual safety checks and current compressors break more often – operating at near absolute zero remains difficult)