In part 2 of my discussion with Paul Martin on hydrogen (read or listen to part 1 here), we pivot from the places on Michael Liebreich’s hydrogen ladder where hydrogen is less like to be used to more productive places for it. And then back again.
We start this part of our discussion at the top of the ladder instead of the bottom, which includes the vast majority of all hydrogen consumed annually today, approaching 100%.
Two of the five uses, hydrogenation for vegetable oils and the like, and methanol, are going to continue to be with us, but they are very small portions of annual hydrogen use.
Two of the five, hydrocracking and desulphurization, are in petroleum refineries per both IEA and IRENA, and hydrogen use for those purposes will drop by 75% to 80% in Paul’s estimation as we eliminate fossil fuels. IEA and IRENA count different things, with IEA including syngas use and IRENA excluding it, so while I’ve reported on 55% of hydrogen used in refineries and 37% in fertilizer in the past, that’s the IEA perspective, and IRENA’s is more nuanced.
Roughly half of the IRENA hydrogen use is for ammonia-nitrogen fertilizer. That’s been flat for 30 years and petroleum use was the only substantial growth market during those decades. It’s likely going to diminish, Paul and I agree, due to significant expansion of precision agriculture, a lot more low-tillage agriculture and biogenetics such as Pivot Bio’s nitrogen-fixing microbes, all of which displace fertilizer, a major source of global greenhouse gas emissions not only in manufacturing and distribution, but in nitrous oxides from fields.
In other words, three of the five unavoidable uses for hydrogen, the places we use black hydrogen today, representing the large majority of hydrogen consumption, will be dropping by 50% or more in the coming decades. And as we discussed in the first half, almost all of the sections of the ladder related to transportation will electrify or use biofuels, with minimal additional hydrogen.
Paul posits that we don’t use hydrogen as a fuel today because it’s not a good fuel, and we shouldn’t use it for that purpose in the future. I agree. If you want Tucker Carlson’s early bow-tied career of talking heads disagreeing loudly with one another, this isn’t the podcast for you.
Our attention then turns to long-term storage, something Liebreich puts on the B row, very high up in the ladder. Paul and I disagree with Liebreich, at length.
Dunkelflaute, doldrum periods with no sun or wind, comes up a lot in storage discussions, with hydrogen advocates thinking that it is fit for purpose for seasonal storage. The seductive idea is that we’ll have too much in the summer, and we’ll make hydrogen then and use it in the winter. However, it makes no economic sense. Electrolyzers are very expensive, Paul points out, so running them at 10% annual capacity makes the cost per kg of hydrogen very expensive. We can’t afford to make green hydrogen unless they are closely linked to very high-capacity hybrid wind and solar, and then green hydrogen is only 3-4x the cost of black hydrogen.
And those locations are in the wrong places, such as the Saharan Desert or the north coasts of Brazil. Today we make hydrogen where it is used, in the same chemical plant it is used in, because it’s hard and expensive to move. Chilling and compressing hydrogen were explored at length in the first half of the discussion, and the problem only gets worse when you try to put hydrogen on ships and transport it thousands of kilometers. There are liquid carriers like toluene, but you have to convert it back and forth, with a lot of heat entropy, hence even more efficiency losses. Heat losses are important, as heat isn’t particularly useful most of the time.
Paul has coined a phrase, the First Sin of Thermodynamics, which is “Though shalt not compare two kinds of energy just because they have the same units.” The useful portion of energy is called exergy. Electricity is almost all exergy, in that you can convert it into mechanical work with high efficiency. Heat has bad exergy unless it’s very hot, and the heat lost in hydrogen compression, release and use is low-grade heat. A kilowatt hour of heat, in other words, is not equal to a kilowatt hour of electricity, but a kilowatt hour of electricity is equal to a kilowatt hour of heat or more.
Hydrogen requires pre-existing large sealed caverns that are fairly close to high capacity factor wind and solar farms. Thermal management of cooling hydrogen as it is released for use wastes more energy. Fuel cells are at best 60% efficient. Both Paul and I like to give failed technologies as much benefit of the doubt as possible in calculations. I use 43% maximum best case scenario for hydrogen, but Paul accounts for the storage losses as well, and arrives at 37% best case efficiency.
Hydrogen’s problems are the properties of the molecule and its thermodynamics as a chemical reagent, neither of which we can change.
We can’t innovate our way out of the fundamentals of physics and chemistry.
The storage alternatives are pushing water uphill, pumped hydro, and batteries, especially redox flow batteries in my opinion, along with converting existing hydro facilities to run on a much more on-demand model instead of a baseload model. But we don’t need much storage until closer to the end game of decarbonization, and then not nearly as much as many expect.
Paul has the heretical idea that we should burn fossil fuels for the dunkeflaute. If we are only using fossil fuels for the very hardest parts of the year, we’ve already addressed global warming. I agree — with nuances — that we’ll slowly diminish capacity factors for gas generators through 2100 until they are no longer used at all.
Our point is that most dunkelflautes are regional, and continental and intercontinental connections, some of which already exist, allow electricity to flow from where it is generated to where it is needed. Electricity travels from northern Canada and Europe’s hydro facilities to major industrial and population centers far south and from Africa into Europe already.
Despite this, many people seem to believe that you can’t ship electricity from Africa to Europe, or between large countries, or the 124 miles from Japan to the mainland of Asia , so you have to make hydrogen and synthetic fuels and ship it by tankers. This is empirically false, and easily disproven just by the vaguest awareness of what we have been doing for decades.
Paul thinks that a lot of people have a burner box on their head. They burn stuff for energy now, so when they need to replace it, they ask what else they can burn. But electricity is trivially easy to use in most cases, so just use electricity. Paul and I agree that HVDC electrical transmission lines running east, west, north, and south are the energy corridors, pipelines, and energy transports we actually need, and will end up building.
Paul also asserts that we need to get our heads around efficiency that matters vs efficiency that doesn’t matter. Everybody who has off grid power inherently understands this. They have more panels than they need in the summer because panels are cheap. They buy panels based on winter requirements, and minimize storage. Overbuilding renewables is something we’ll do, just as we overbuilt a lot of coal and gas generation, hence the reason it’s near the top of my short list of climate actions that will work. But the nice thing about wind and solar is that the energy itself is free, so there’s no cost for fuel, including shipping it, storing it, and getting rid of the wastes from burning it.
The net of all of this is that while Germany may do some long-term energy storage using hydrogen because they have an odd national fixation on the molecule similar to Japan as well as salt caverns, virtually nobody else in the world will be using hydrogen as a long-term energy store.
We pivot our discussion then to another area purported to be a big growth market for green hydrogen: steel. Recently steel made using hydrogen from renewable electricity was delivered to a car plant in Sweden. Per Paul, about 10% of hydrogen is used in syngas for steel reduction already, prior to this latest effort. Normally iron ore is reduced to steel in blast furnaces with coking coal with lots of CO2 emissions. But we both point out that a great deal of steel is already being made with electric steel minimills, which can run off entirely renewable electricity. Paul has also been tracking using electricity directly to reduce iron, similar to what we do already with aluminum.
As energy gets less environmentally impactful, we’ll be able to use more energy to recycle things without a negative environmental impact and mine less ore for new steel. As I like to point out, there’s an absurd amount of steel in fossil fuel infrastructure including freighters, pipelines, storage, and refineries that we aren’t going to need in the coming decades, so ripping them up for scrap steel and putting them through electric steel minimills will meet a lot of demand.
My summary of our conversation was that a lot of hydrogen is being pushed by fossil fuel companies for purposes it isn’t useful for. Hydrogen demand is going to decline before it rises, as major uses today are going to drop a lot. And that we do have to supply high value processes which use black hydrogen with green hydrogen.
Paul closed with the assertion that he’s not anti-hydrogen. It has a big role in decarbonization, but that’s in decarbonizing hydrogen we use for chemical and industrial processes itself. Assuming we need to decarbonize the 90 million tons we use today, we would need 513 gigawatts of electrolyzers running at 100% capacity factors all year. A more reasonable 50% capacity factor would require a terawatt of electrolyzers.
Right now there isn’t a single gigawatt of electrolyzers operating in the world today, so we are off by three orders of magnitude.. Similarly, we would need to power them with 4,500 TWh of green electricity every year, but we’re still around half of that at present, and that electricity is eliminating the use of coal and gas for electricity, a much more effective use.
The net is that hydrogen is not a decarbonization solution, it’s a huge decarbonization problem that’s being sold aggressively by organizations and people who want to perpetuate the fossil fuel industry.