Hydrogen gas

Not to be confused with nuclear fusion of hydrogen atoms.

Hydrogen gas (H₂) is a combustible fuel that leaves behind nothing but water vapor (H₂O) when burned.

There are no natural resources of hydrogen gas(...)( except in rare and extremely small quantities, not a viable way to supply energy in any meaningful amount ). To make hydrogen gas, you need to use some other energy source. In this way, hydrogen can be understood as a form of energy storage.

This page is about how hydrogen gas could be used with renewable energy.

Energy storage basics

For energy storage of renewable electricity:

• Hydrogen gas would be produced via electrolysis:
  • Electricity is used to convert water (H₂O) into hydrogen gas and oxygen gas.
• Hydrogen gas would be consumed via...
  • Burning it as fuel, producing heat.
  • Using it in fuel cells, producing electricity (and still some heat).
  • In both cases, the hydrogen reacts with oxygen in the air to form H₂O again (water vapor).

This process has more energy losses than charging/discharging a battery, but hydrogen gas is far better suited for long-term energy storage. Hydrogen can be stockpiled in pressurized tanks (if designed properly). It can also be shipped long distances, just like any other fuel. This could help in cases where renewable energy sources are geographically far away from where energy is needed.

The intent would be for hydrogen gas to be used in place of fossil fuels:

• Cars, trucks, etc. would be:
  • Hydrogen combustion vehicles, or
  • Hydrogen fuel cell electric vehicles
• Homes & buildings:
  • For heating: Hydrogen gas could be burned instead of natural gas.
  • For cooking food: Hydrogen gas could probably work with gas stoves. ^{[RESEARCH needed]}
• Factories:
  • Most of the energy used in manufacturing is in the form of high heat needed for processing materials. Factories could burn hydrogen gas instead of burning coal or natural gas.

Energy sources

Main use-case: Storing wind power.
Here's why:

• Wind power is far more intermittent than solar. Whereas solar follows a day/night cycle, windy and not-so-windy seasons can last for months at a time.
• Wind turbines tend to be geographically far away from where electricity is needed, on average. Wind power is more spread out in terms of land, compared to the same amount of energy from local rooftop solar. Hydrogen could be transported long distances that can't be reached with power lines.

Other use-case: Since solar panels produce more energy in the summer, it would still be worthwhile to store some of that energy via hydrogen gas, to be used during the winter. Note, however: Batteries are a better choice for smoothing out the day/night cycle of solar power.

Other use-case: Storing energy from hydroelectricity during long periods of low demand.

Status quo

• Most hydrogen today is produced from fossil fuels (natural gas) via steam reforming. The carbon emissions are as high as burning the natural gas itself.
• Most hydrogen today is used in producing fertilizer.

Platinum-group metals

Problem in some cases

Both electrolysis and fuel cells need platinum-group metals (PGMs):

• [platinum, palladium, rhodium, ruthenium, iridium, osmium]
  • Any of these metals will do, but all of them are extremely scarce (even more than gold), with platinum & palladium being the most available.
  • These metals serve as catalysts in the reactions. They are not used up, but they need to be there, in a thin layer plated onto the electrodes.

Note: It is possible to build fuel cells and electrolysis systems without PGMs, but the energy-efficiency is much lower.^{[QUANTIFICATION needed]} There are scientists trying to overcome this,^{[LINKS needed]} but there's no guarantee that it will be viable in the near future.

How much would be needed, if hydrogen were scaled up?

world.population

8 billion

Number of people alive today, globally

https://www.unfpa.org/data/world-population-dashboard
Last updated in 2023

world.cars

1.446 billion

Number of cars in the world

Last updated in 2022
www.carsguide.com.au › car-advice › how-many-cars-are-there-in-the-wor...
hedgescompany.com › blog › 2021/06 › how-many-cars-are-there-in-the-...

toyota_mirai.pgm

30 grams

Amount of platinum-group metals (PGMs) in a Toyota Mirai (fuel cell vehicle)

The Toyota Mirai is a common example of a hydrogen-powered vehicle.

https://www.heraeus.com/media/media/hpm/doc_hpm/precious_metal_update/en_6/20181031_PGM_Market_Analysis.pdf

catalytic_converter.pgm

2 grams

Platinum-group metals (PGMs) in a catalytic converter of a car

Countless automotive forums say 3 to 7 grams for a typical car.

But ThermoFisher (which is more reputable, perhaps) says "The recoverable amounts of Pt, Pd, and Rh in each [vehicle] can range from 1-2 grams for a small car to 12-15 grams for a big truck in the US." - Are There Precious Metals in Catalytic Converters? https://www.thermofisher.com/blog/metals/platinum-group-metal-recovery-from-spent-catalytic-converters-using-xrf/

I assume they mean 1-2 grams ''total'', not 1-2 grams ''of each'' Pt Pd Rh, right? That would make sense considering they also mention that the ratios vary as metal prices/availability change over time.

1 to 2 grams total recoverable is also consistent with the following study: Yakoumis et al 2018 IOP Conf. Ser.: Mater. Sci. Eng. 329 012009 - Real life experimental determination of platinum group metals content in automotive catalytic converters - https://iopscience.iop.org/article/10.1088/1757-899X/329/1/012009/pdf

Still no word on what percentage this ''recoverable'' is of total PGMs - how efficient is the recycling process? Unknown

platinum.mine_production

186000 kg/year

Global production of new platinum from mining

Using data from 2019.

Source: USGS Mineral Commodity Summaries 2021

palladium.mine_production

227000 kg/year

Global production of new palladium from mining

Using data from 2019.

Source: USGS Mineral Commodity Summaries 2021

pgm.mine_production

platinum.mine_production + palladium.mine_production

Global production of platinum-group metals (PGMs) from mining (status quo)

Assumption: that the other PGMs (iridium, rhodium, osmium, ruthenium) are in such small quantities that it's ok that they aren't counted here (because data is unavailable)

pgm.reserves

70000 tonnes

Global mineral reserves of platinum-group metals

Includes platinum, palladium, ruthenium, rhodium, osmium, iridium.

Platinum-group metal reserves worldwide by country 2021
Statista - https://www.statista.com › statistics › platinum-me...

crop_land

15000000 km^2

Agricultural land used for growing crops - global total

https://ourworldindata.org/land-use

electrolysis.pgm_by_power

0.209 grams per kilowatt

Quantity of platinum-group metals (PGMs) in an electrolyzer

Electrolyzers make hydrogen gas from water & electricity. Platinum and/or similar metals are needed as catalytic plating.

Data source:
Manufacturing Cost Analysis for Proton Exchange Membrane Water Electrolyzers
August 2019 Technical Report NREL/TP-6A20-72740
https://www.nrel.gov/docs/fy19osti/72740.pdf
Pages 4 and 5: Table 1:
Cell plate area: CCM coated area: 748 cm^2
Platinum loading (anode): 7 g/m^2
Platinum-iridium loading (cathode): 4 g/m^2
Single cell power: 1965 W
From this we can calculate:
(7 g/m^2 + 4 g/m^2)/2 * 748 cm^2 / 1965 W = 0.20936387 g/kW
Sucks that the article doesn't directly specify this 'g/kW' value for us to confirm whether my calculations are correct. Still this is the best data source I could find. The article does also provide a lot of specs on total costs (ranging from $561/kW all the way down to $69/kW for some proposed systems with advanced techniques and economies of scale).

wind.capacity_factor

35%

Wind power: ratio: average output / peak power capacity

"The capacity factor of a wind turbine is its average power output divided by its maximum power capability. On land, capacity factors range from 0.26 to 0.52. The average 2019 capacity factor for projects built between 2014 and 2018 was 41%. In the U.S., the fleetwide average capacity factor was 35%."
https://css.umich.edu/factsheets/wind-energy-factsheet

wind.rq_land

34.5 hectares/MW

Land requirements of wind power

Important:
- This is per megawatt capacity (peak), not per average output.
- Stats can vary tremendously based on how windy the location is.
- This stat is based on 172 different wind projects scattered throughout the USA.
- Consider variance: (34.5 +/- 22.4) hectares/MW
- This is the total land use, including the spacing between turbines in a wind farm.
- This is much bigger than [wind.rq_land_disturbed] which is just the land directly impacted by constructing the turbine itself.

Citation:
Land-Use Requirements Of Modern Wind Power Plants In The United States
(Paul Denholm, Maureen Hand, Maddalena Jackson, and Sean Ong)
Page 16

energy.tfc

9937.70 Mtoe/year

Global energy usage - total final consumption (TFC)

Includes: fuel (80.7%) + electricity (19.3%) AFTER it is generated.

Does not include the fuel used in generating electricity. See [energy.tes] for that.

Citation: "Key World Energy Statistics 2020" IEA
- Page 47 - Simplified energy balance table - World energy balance, 2018

fossil_fuels.consumption

11596.92 Mtoe/year

Total consumption of coal, oil, and natural gas (worldwide) (energy units)

Key World Energy Statistics 2020 (IEA report)
- page 47: World energy balance, 2018
- - Total Energy Supply (TES), first 4 columns combined

electrolysis.efficiency

80%

Energy efficiency of producing hydrogen & oxygen gases from water

Hydrogen made by the electrolysis of water is now cost-competitive ...
www.carboncommentary.com › blog › hydrogen-made-by-the-electrolysis...

commercial_factor

2

Without this, we'd be calculating for just personal vehicles. But we also need to factor in commercial vehicles such as buses and trucks. These vary widely in size, and data is hard to find, so for simplicity sake, we just assume that they'd add up to about the same as personal vehicles - thus doubling total energy storage needed. This assumption is based on the fact that freight trucks are a somewhat smaller share of energy demand than passenger vehicles, but the trucks need far more horsepower and probably a longer range.

General principlesGeneral principles

The supply of PGMs is limited to what we can mine from the Earth (mineral reserves / resources), so we have to be mindful of how much would be needed.

The mass of PGMs needed is proportional to peak power:

• For electrolysis systems, the maximum rate of hydrogen production is limited by the amount of PGMs.
• For fuel cell vehicles, the horsepower is limited by the amount of PGMs.
  • But the vehicle can still achieve short bursts of higher horsepower if there's a battery or supercapacitor in parallel with the fuel cell.

Scenario 1: If hydrogen gas (from wind power) were to directly replace all fossil fuels (this implies that people would drive hydrogen combustion vehicles):

(see maths)(see maths)

electrolysis.pgm_by_power / electrolysis.efficiency / wind.capacity_factor * fossil_fuels.consumption % pgm.reserves a (calculation loading) a years pgm.mine_production (calculation loading)

In other words, the amount of PGMs needed is pretty reasonable - but we'd have to start mining a lot faster than the status quo (and find some way to do it without exploitative labor).^{[new page needed]}

And in fact, some(...)( about half, if calculated from the same datapoints used on this page ) of this mining could be avoided, as we could also recycle existing catalytic converters(...)( which also contain some PGMs, though nowhere near as much as fuel cell vehicles ) from the gasoline cars and diesel trucks that would no-longer be used.

Scenario 2: If all vehicles were hydrogen fuel cell vehicles instead:

(see maths)(see maths)

(toyota_mirai.pgm - catalytic_converter.pgm) * world.cars * commercial_factor % pgm.reserves b (calculation loading) b years pgm.mine_production (calculation loading)

One benefit of fuel cell vehicles is that they're more energy-efficient than combustion vehicles (i.e. less hydrogen used per kilometer driven).

The problem is, fuel cells would need 7 times the PGMs of Scenario 1 Template:Light(in the estimate above). Perhaps too much to be scalable. And this is true even though we factored in the recycling of old catalytic converters this time (which have less than 1/10th the PGMs of a hydrogen fuel cell vehicle, on average).

Verdict:

• If we're going to have hydrogen-powered vehicles, most of them will probably have to be combustion only (or some sort of hybrid with just a very small fuel cell).
• PGMs are not a limiting factor for wind-based hydrogen production.

More discussion / research neededMore discussion / research needed

More musings about the calculations above:

• All of this assumes that electrolyzers and fuel cells can be completely recycled at their end-of-life, with all PGMs recovered. If they can't, we're kind of screwed in the long run (at least for hydrogen).
• Hydrogen combustion vehicles are about as energy-efficient as gasoline combustion vehicles. Hence we can assume that the Scenario 1 estimate is accurate enough.
• Home electricity can also be done with fuel cells - this would of course need more PGMs (and more hydrogen to make up for the losses in fuel cells (although those losses could be used as heating in some cases)).
• We didn't count the hydrogen needed in the vehicles that transport the hydrogen (hopefully would be minor, like with fossil fuel transport).
• All this is based on status-quo energy demand, which unfortunately relies on the fact that most of the world currently lives in poverty. If all nations were developed, more resources would be needed.
• But in any case, we probably wouldn't actually use wind/hydrogen for everything anyway. Rooftop solar combined with batteries could probably be a better way to provide electricity whenever hydrogen need not be involved.
• Since vehicle fuel cells use the biggest share of PGMs in this estimate, this is yet another reason to advocate for public transit and walkability.

Energy losses

Lossy but manageable

• Electrolysis is at most 80% efficient.
• Fuel cells are at most 60% efficient.
• Thus, best-case electricity recovery is only 48%(...)( in other words, 60% of 80% ). Far less than most batteries which have a charge-discharge efficiency of 80% to 90%.
  • But for things that just need heat, then the energy recovery is still a good 80%. For example, wind power to produce hydrogen gas to burn for heating homes.
    • Note however: For vehicles, this is outweighed by the fact that hydrogen combustion vehicles are less fuel-efficient than fuel cell vehicles.

TODO: add calculation: knowing the losses, is there still enough land for wind-generated hydrogen gas were to directly replace all fossil fuels, in principle?

Shelf life

^{[RESEARCH needed]}

Chemically, hydrogen is the lightest gas (smallest molecules). This makes it harder to store than other gases, but there are still ways. ^{[ELABORATION needed]}

Pipelines

^{[RESEARCH needed]}

Could existing natural gas pipelines be used for transporting hydrogen gas? Or would it cause too much leakage/corrosion? ^{[RESEARCH needed]}

Safety

Manageable

• Just like natural gas, hydrogen gas is non-toxic and odorless but highly flammable. For safety in consumer applications, small quantities of some non-toxic but smelly gas(...)( such as methyl mercaptan, hydrogen sulfide, or ethyl isobutyrate (Wikipedia has a page "Hydrogen odorant") )should be added to it, so that people would know if there's a gas leak.
• This section needs more safety-related info.

Color terminology

Hydrogen is a colorless gas, but researchers sometimes name it with colors to indicate how it was produced:

• "Grey hydrogen" is made from natural gas (steam reforming) - high greenhouse gas emissions. Currently the vast majority of hydrogen is produced this way.
• "Blue hydrogen" is made from natural gas the same way, but with carbon capture. This is supposed to reduce emissions, but in practice it doesn't help much.
• "Pink hydrogen" is made from electrolysis using nuclear energy.
• "Green hydrogen" is made from electrolysis using renewable energy.

See also

• Methane cracking - another way to produce hydrogen gas. Not worthwhile currently, but in theory the right tech could maybe change that.