Liquid Hydrogen is Hard
Hot take: if a startup or company is trying to do liquid hydrogen (LH2) and isn’t partnered with Air Liquide, Air Products, or Linde, it’s at high risk of going bankrupt like Universal Hydrogen did last month. Those are the only three companies that do LH2 without introducing company-ending technical risk. On the downside, in the US there is a lot of commercial risk of non-delivery of H2 associated with some of these Industrial Gas Companies (IGCs) that also increases risk of venture failure. In other words, liquid hydrogen is high-risk as an investment or business decision - and these risks need to be mitigated to help assure project/company success.
Here’s the article flow:
1. Why liquid hydrogen is hard from a technical view
2. Why liquid hydrogen is hard from an operations perspective
3. Why liquid hydrogen is commercially a non-starter
4. Why liquid hydrogen is monumentally more dangerous than gaseous H2
5. Why the IGCs still like LH2 anyway
6. Working with the IGCs is a double edged sword
7. Very brief “what are the other options”
8. Conclusion
9. PS for the extremely advanced reader: For real though, how can we make LH2 happen, at what scale, and who can do it?
Why Liquid Hydrogen is Hard from a Technical perspective
Let’s compare LH2 to liquid natural gas (LNG). Hydrogen becomes a liquid at -253C, or 20 degrees above absolute zero. Natural gas becomes LNG at a mere -161.5C. Keep in mind that the rate of heat loss scales with temperature gradient (q = -k * ΔT, or heat loss rate = conductivity times temperature gradient) so making H2 it liquid and keeping it liquid is far harder than LNG.
Worse yet, liquid hydrogen wants to switch back to gaseous hydrogen very quickly. The only way to prevent this is to push the hydrogen through an quantum electron transition called an ortho-para transition whereby the rotational direction of one of the electrons in an H2 molecule is flipped. This makes it so the liquid hydrogen can be stable as a liquid, instead of reverting back to gas, for up to 10 days as opposed to a few hours. In other words, making liquid hydrogen involves quantum chemistry.
Why is liquid hydrogen hard from an operational perspective
1. It’s cold, and handling something 20 degrees above absolute zero is dangerous. Truck drivers would need specialized cryogenics training just to refuel their truck. Even having untrained anyone around LH2 is dangerous, so handling and moving it comes with commensurate safety and cost.
2. Because it is extremely cold, there are no pumps or valves that really work reliably for LH2. We don’t have the valves and pumps to actually move LH2 easily and repeatedly. They are under development, but it’s not the likes of Universal Hydrogen that will solve this problem.
3. Because of the lack of pumps, we vent a lot of liquid hydrogen when it’s transported. An LH2 tankers delivers LH2 by allowing some heat in to the tank and letting some of the liquid hydrogen vaporize. This builds up gas pressure in the tank. The pressure then pushes the LH2 from the tank to the buyer’s storage unit. Before the LH2 tanker truck gets back on the road, they vent the overpressure of gasified LH2. A truck can lose 10-15% of its LH2 per stop in this way, easily losing up to 30-40% of LH2 per delivery run if there are 4+ stops.
4. Building liquid hydrogen containers is hard. Everything becomes brittle at -253C and it’s not particularly safe.
One caveat: a 40 foot LH2 trailer can move 3500kg of LH2 vs 1200kg for a gaseous H2 trailer. So if the trailer is going more than 500 miles, LH2 can start to make sense.
Another caveat: GH2 is working through its own growing pains with valves and compressor seals, but many have solutions already, they just aren’t DOT certified yet.
Why LH2 is commercially a non-starter
In an earlier post, I wrote about how to move hydrogen. Here is more detail on why LH2 is a non-starter. It’s just too expensive. From a prior post, see the results of a detailed model I built on relative costs:
Break-even volumes and distances for LH2, GH2, and pipelines
A hydrogen liquefier is at least $5000/kg-day capacity ($150M+ for a 30tpd capacity) and it can’t efficiently be turned down to operate at low capacity. It also currently takes 11kwh/kg to liquefy. Keep in mind a family of four uses 30kwh per day, so 3kg of LH2 uses as much energy to liquefy as a family of four uses in a day. The CapEx alone means that the levelized cost LH2 is $4/kg before we’ve even bought a molecule of H2. The OpEx with $0.10/kwh electricity means you add another dollar of OpEx on top. Then there is the cost of the H2, which for grey H2 is $1/kg and for green is $6/kg.
Let’s compare to gaseous. Gaseous compressors cost $500/kg-day capacity, can cycle up and down easily, and take 1kwh/kg to get onto a compress gas H2 (GH2) truck. The CapEx contribution is $0.40/kg and the OpEx is closer to $0.04/kwh – if tied to an electrolyzer it can cycle up and down and use very inexpensive solar and wind. This is an important concept – only GH2 can be phase-tied to H2 production for renewable H2 production, LH2 cannot. This means that you’re adding $0.10 for OpEx from compression on top of the levelized cost stack, compared to $6/kg for LH2.
The difference can be made up in CapEx and OpEx of trailers, but only over very long distances. The distance from production to offtake has to be on the order of 500 miles in most cases to be cost-effective, and goes to 1000 miles when you consider the venting issues with LH2.
The real killer, however, is that compressors can scale up or down and liquefiers can’t. Compressors currently operate great up to rates of 2 tons per day with 30x compression ratios, and new 10 ton per day models are coming out. They can work in parallel for larger throughput. Larger than 30 tons per day and you consider a pipeline. LH2 gets much worse at smaller scales and isn’t appropriate to scale up or down.
LH2 is monumentally more dangerous than GH2
GH2 rises in the atmosphere at 70 mph because it is light. LH2 pools on the ground, causes oxygen to liquefy out of the air, then you have a mixture of liquid and gaseous H2 and O2, IE rocket fuel, and you get a vapor-gas explosion.
LH2 is far more dangerous than GH2, and the cost to manage it commensurately higher.
Why do IGCs like LH2?
As I said above, the IGCs are the only ones that can do LH2. In the US the deployment competing technology of GH2 is being held up by the Department of Transportation, and as a result the IGCs have an impenetrable moat in LH2. In the EU where restrictions are less harsh, GH2 is the main, preferred, and most cost-effective way to move H2. As long as the US DOT holds up certification GH2 hardware, it remains difficult for GH2 to get the foothold it needs to take off.
The double edged sword of working with an IGC
This is a double edged sword for everyone else, because the IGCs can be very hard to work with. To the point that they can make it impossible to succeed in the space. I’ll handle this issue in a separate post, but the general point is that there is no way to win with LH2, the deck is stacked against commercial or technical success.
What are the other options?
See this post. Generally, gaseous hydrogen at low volume is ready to go, we’re just waiting on the US Department of Transportation to develop real approval rules for trailers and for local jurisdictions in the US to allow them to move through tunnels and over bridges. The EU already has them available. China doesn’t allow them because they can’t manufacture their own composites to the proper spec and they don’t want imports.
Beyond low volume, pipelines and gaseous are great.
Conclusion for investors: stop trying to make LH2 happen
Liquid hydrogen is too expensive, full stop, and it will be for the foreseeable future. So, to paraphrase the movie Mean Girls, “Stop trying to make Liquid Hydrogen happen. It’s not going to happen.”
We will not see a hydrogen economy take off on the back of LH2. LH2 should be reserved for niche use cases, and anyone looking to play with it now without proper supervision of an IGC is probably going to end up belly-up.
That is not to say that all companies that work in LH2 will go belly-up. The global hydrogen market is estimated to be $250B per year. Niches can make a lot of money here. Just don’t expect a company that isn’t an IGC or IOC to be able to enter the LH2 production and distribution supply chain and have an easy time of it.
Post-script for the advanced readers: how LH2 could work in the future
We’re past 1500 words, so this is just an appendix section. It will also be complicated.
Liquid hydrogen can come down in cost. It just needs to scale massively. A few years ago, the largest liquefiers produced 30 tons per day of LH2. In terms of energy, this is 1% of the energy throughput of an LNG terminal. LH2 production at a single facility probably needs to scale up to about that size in order to be cost-competitive.
Beyond hand-waiving: Hydrogen is liquified in three steps inside a container called a cold box. At 30-60 tons per day, all three steps happen in the same cold box. Expanding to separate cold boxes without hitting proper scale will result in the CapEx of a plant going from $5000/kg-day to $8,000/kg day or more. Until the plant gets big enough.
Somewhere around 500 tons a day, it makes sense to split all three cold boxes out, and then everything starts to get a lot more efficient. Both the CapEx and the OpEx drop significantly. As the systems scale beyond 500 tons per day, these efficiencies both rise.
It’s feasible to liquefy hydrogen at this scale for $1-$2 per kg. This would be the equivalent of $9-$18/mmbtu. It starts to make it cost-effective to get to places like the EU where fossil natural gas is more expensive than this.
I don’t think this level of scale and innovation is something the IGCs can accomplish, and no startup will ever accomplish it. Only a few companies in the world could do this, and they are all oil companies. I would expect to see Shell or Saudi Aramco accomplish it, probably by 2030 or 2035.
Once hydrogen is delivered as a liquid onshore, it can be kept as liquid. In other words, if the EU is a liquid hydrogen importer, they could feasibly use the liquid hydrogen in trucks, provided they worked out how to automate liquid fueling so no human is involved anywhere.
