Shipping Liquid Hydrogen Would Cost 5-7x LNG Costs Per Unit Of Energy
- Dec 29, 2021 6:13 pm GMT
Recently a former boss of mine, with whom I worked in Canada and Brazil, reached out to me with a question about a Wall Street Journal article he’d read on a proposed Namibian project to manufacture green hydrogen and ship it to the developed world. He was wondering if this was economically viable. It isn’t, of course, but it gave me an excuse to finally do the math.
Liquid natural gas (LNG) is the best comparison, as it requires liquification, which consumes about 10% of the amount of energy embodied in the LNG and oceanic shipping. A Wartsila Q-Max ship, the biggest LNG tanker I’m aware of, carries 266,000 cubic meters of LNG. The liquified gas is cheap, with average delivered import prices in the US of $109 per cubic meter of LNG, or about $0.18 per cubic meter of natural gas, although prices are obviously spiking in late 2021 around the world. Those 266,000 cubic meters amounted to about $29 million in value before the recent spikes.
The first thing to know is that while hydrogen is energy-dense by mass, it isn’t energy-dense by volume. Assuming the same-sized ship, the delivered BTUs of energy would be about 27% of the LNG. This is because even liquified, hydrogen has less energy by volume than LNG, but also because liquifying hydrogen takes about 33% of the energy in the liquified hydrogen, as opposed to the 10% required for LNG. Different gases, different temperatures required for liquification. Amazing stuff with liquid oxygen for space travel, but not so much anywhere less exacting.
The second problem is that hydrogen by itself is expensive. Using the absolute best case from Lazard’s hydrogen LCOE, with dirt cheap electrolyzers, $10 per MWh electricity and 100% capacity factors, hydrogen is projected to be about $0.78 per kilogram. These numbers are as likely as unicorns appearing worldwide and granting free rides to children. A cubic meter of liquified hydrogen masses 71 kg. That means that just the cost of the liquified hydrogen, excluding the energy costs of liquification or getting it into the ship, would be 1.9 times as high as the delivered price of LNG. And that delivered price includes extended ocean journeys with daily LNG ship charter costs averaging of $150,000 per day, currently peaking over $200,000 per day, with a high of $350,000 per day in January 2021. There’s a bit more on the cost of a cubic meter of liquid hydrogen for liquification, but as we are pretending that electricity will cost $10 per MWh, it’s not very much, about another dollar per cubic meter.
Currently, the average for trip durations, excluding loading and unloading periods, is around 23 days, and the pre-berthing, loading, and unloading add another 4-5 days. Calling it 28 days results in an additional cost of $4.2 million.
The next thing to consider is the boil off rate. LNG best of breed boil off rates are 0.1% per day. While NASA demonstrated Brayton cycle refrigeration of bulk LH2 storage with approaching 0% boil off rate, that was at great expense and with a lot of additional equipment. The projected best case for in-voyage boil off rates is 0.2% for liquid hydrogen per a recent South Korean KAIST publication. Boil off rates during transfer are also higher than for LNG. Liquid hydrogen would lose another ~5% of its total delivered energy due to that in the best case scenario, and likely more.
That brings the total costs of the delivered hydrogen to roughly $19 million for less than 27% of the delivered energy. One assumes that the people doing this would like to make a profit as well, so assuming 10%, the total price delivered will be in the range of $21 million for the load.
That means the price per delivered unit of energy is roughly 5 times that of liquid natural gas, in the best possible case.
More realistic numbers from Lazard’s LCOE with still low $20/MWh electricity, still high 90% capacity factors, and more realistically priced but still cheap electrolyzers, would be 40% more expensive, around 7x the price of LNG per unit of energy. Doing the math with the numbers that most favor hydrogen shows how starkly bad the economics of using it for energy really are.
Of course, this is before the liquid hydrogen is converted to actually useful energy, at a maximum of 60% efficiency, which is about the same as a good combined cycle gas turbine.
What does all this mean for the actual sunshine falling in Namibia? Well, take 20% off for electrolysis, 33% off for liquification, 10% for efficiency losses for long-haul cooling and handling, and 40% off for conversion back into electricity, and the solar energy energy in Namibia turns into perhaps 29% of the energy being useful.
Of course, this doesn’t address the problem of distributing hydrogen in the country that imports it either, or the complete and utter lack of any large-scale hydrogen distribution network. 85% of hydrogen consumed globally is manufactured onsite because it’s so expensive to ship.
Delivering the same energy in the form of electricity through high-voltage direct current cables is vastly more efficient, with HVDC losses running about 3.5% per 1000 kilometers. Of course, Namibia being in southern Africa next to South Africa, it would make more sense for it to deliver the electricity there instead. Northern African solar, wind, and storage linked to fat HVDC pipes crossing the Strait of Gibraltar, crossing from Tunisia to Italy, and crossing the Bosphorus — as transmission lines already do today — makes much more economic sense. There’s a Moroccan project proposed with 20 GW of wind and solar firmed by storage, and a proposed 3,500-kilometer HVDC transmission line to the UK. Assuming the entire route is underwater, that would deliver about 88% of the generated electricity to market, not 29% — more than 3 times as much. Similarly, the Australian green hydrogen proposal originally ran HVDC transmission to Singapore, but has been distracted by the hydrogen hype and now proposes to manufacture hydrogen and ship it.
Of course, there are even less efficient ways to ship hydrogen. It could be chemically converted into a stable liquid at room temperature, and then extracted at destination, stepping on the electricity a few more times, at great loss of useful energy. It could be converted into a liquid hydrocarbon with the addition of CO2 from somewhere and upgraded to a useful plug-compatible fuel, then shipped, at only multiples of the cost of just running things off of electricity, and with the added “benefit” of air pollution. For those who say, "But ammonia!", this is the problem with it. It has 51% of the energy density, but has more efficiency losses in manufacturing and use, so it's a wash.
This isn’t to say that Namibian green hydrogen can’t be useful. The country is dependent on South Africa for urea fertilizer, and could be manufacturing that itself to supply the 9% of its economy dependent on agriculture. Fertilizer needs lots of hydrogen, and once manufactured is easy to distribute, unlike hydrogen. That’s part of the reason why my projection of hydrogen demand is contrarian, with it falling over the coming decades.
Yet another straightforward analysis that accounts for both the physics, the engineering, and the costs makes it clear that hydrogen is not an economically viable store of energy for our future decarbonized economy. All of the projects proposing to manufacture hydrogen where sunshine and wind are constant and cheap, and ship it to where energy is consumed, are clearly based on hand-waving, ignorance, sheer #hopium or outright larceny.
A version of this article previously appeared in CleanTechnica.
Get Published - Build a Following
The Energy Central Power Industry Network is based on one core idea - power industry professionals helping each other and advancing the industry by sharing and learning from each other.
If you have an experience or insight to share or have learned something from a conference or seminar, your peers and colleagues on Energy Central want to hear about it. It's also easy to share a link to an article you've liked or an industry resource that you think would be helpful.