> The work may be niche, but the impact could be high. About 1 percent of SF6 leaks from electrical equipment. In 2018, that translated to 8,200 tonnes of SF6 emitted globally, accounting for about 1 percent of the global-warming value that year.
This figure is for the electricity sector only, not overall global emissions. Still, considering the sheer volume of CO2 puffing up from power stations, it's impressive that the normal operation of SF6 breakers accounts for an integer percentage of their GHG impact.
----
Global emissions were 53 Gt CO2 equivalent in 2023 [0]. 38% of CO2 emissions are attributed to the electricity sector in 2023. [1] This figure seems to be strictly CO2, not including other GHG, and I can't quickly find a sector-by-sector breakdown for that year. Per IPCC reports in 2022, electricity production and heating accounted for 34% of global GHG in 2019 [2], so for back-of-the-envelope math, it's reasonable.
Per the article, the GHG impact of SF6 is 25k CO2, so 8.2k tons SF6 emitted annually is 205 million tons CO2e. This is 0.39% of 53 Gt CO2e (the global value), or nearly exactly 1% of the electricity sector's 38% share.
[0] https://www.statista.com/statistics/1285502/annual-global-gr... [1] https://www.statista.com/statistics/1129656/global-share-of-... [2] https://www.epa.gov/ghgemissions/global-greenhouse-gas-overv...
edit: replaced typos of C02 with correct CO2
Not just academic: Hitachi have already deployed a different gas (mix) last month
https://www.hitachienergy.com/news-and-events/press-releases...
Thank you for doing these calculations! 0.39% of anthropogenic global warming is surprisingly large, but it doesn't sound like a big impact to me. I mean, it sounds like about the same greenhouse effect impact as the San Bernardino metropolitan area (5 million people) or the Chongqing metropolitan area (12 million people).
Retric: I regret not having responded in time to your comments, which I agree with.
In terms of impact, it doesn't sound major, I agree. But it's still work that will need to be done. As industry and transport become more electrified, and electric generation gets decarbonized, impacts like high-multiple GHG gases will become a bigger and bigger share of CO2e still happening.
Atmospheric CO2 is already too high to avert terrible long-term impacts of global climate change. Unless we manage to make massive cost reductions for atmospheric CO2 sequestration, weaning entirely off of fossil fuels will not be sufficient to avoid climate change impacts if we still add long-lived GHG molecules to the atmosphere. (SF6 has an atmospheric lifetime on the order of 1000s of years.)
Think of it like the tide going out in a rocky bay. As the water level recedes, rock pillars that used to be too deep underwater to worry about now are close enough to the surface to cause you trouble. On the other side of the same coin, putting in the work to clear them helps give you as big a space to operate in as you had before.
> Unless we manage to make massive cost reductions for atmospheric CO2 sequestration,
We will. The main reason atmospheric carbon capture is expensive is that it requires a lot of energy, and the cost of energy is falling through the floor because of cheap renewables. Expensive high-efficiency chemistries for carbon capture will cede to simpler, energy-hungrier chemistries, the ultimate reductio ad absurdum being something like soda lime. Soon enough synfuel from atmospheric carbon capture will be an attractive alternative to fossil fuels for transport (within 15 years), and then it's just a question of capturing the combustion products from the fuel. We may need to start adding high-multiple GHGs to the atmosphere to compensate for carbon dioxide we remove to make plastic. Hopefully shorter-half-life GHGs than sulfur hexafluoride, though.
The US has taken a very aggressive policy stance against renewables and in favor of fossil fuels, but ultimately it can't prevent the inevitable. If it continues to punish the importation of renewable energy equipment, US subjects will import cheap synfuel, or, failing that, they'll import electrolytic iron, zinc, or magnesium to use as fuel, from countries like Chile, China, and Dubai.
I am bullish on the rapid global uptake of low carbon electrical generation (global solar PV deployment alone is almost at 1 TW/year), but bearish on carbon sequestration being viable unless sucking the CO2 out of the ocean (due to energy required via atmospheric capture). It’s not just energy required, but how much carrier (whether that’s air or water) you need to process per unit of CO2 removed.
Stanford Study: Renewable Energy Beats Carbon Capture on Cost and Climate Impact - https://carbonherald.com/stanford-study-renewable-energy-bea... - June 11th, 2025
Energy, Health, and Climate Costs of Carbon-Capture and Direct-Air-Capture versus 100%-Wind-Water-Solar Climate Policies in 149 Countries - https://pubs.acs.org/doi/10.1021/acs.est.4c10686 | https://doi.org/10.1021/acs.est.4c10686
Climeworks’ capture fails to cover its own emissions - https://heimildin.is/grein/24581/ - May 15th, 2025
I'm skeptical of the Heimildin article, because it contains obvious factual errors:
> equivalent to almost four times Iceland's electricity production, which is about 20 terawatts per year.
You can't measure electricity production in terawatts per year; you can measure things like solar panel deployment speed in those units, as you correctly did. This makes me wonder how many other factual errors I failed to spot in the article.
But, yes, we should not expect atmospheric carbon capture to be economically feasible yet, and when it is, we should expect most companies that attempt it to fail, just as most solar panel companies have failed. But remember that solar energy is free when the sun is up; there's no economic benefit to curtailing your electric production because your batteries are full. So we should expect vastly more energy-intensive approaches than Climeworks' to be viable.
Atmospheric carbon capture isn't an alternative to renewable energy. It's what you do in response to the much lower energy costs resulting from renewable energy, and to reverse the damage already done.
> Expensive high-efficiency chemistries for carbon capture will cede to simpler, energy-hungrier chemistries, the ultimate reductio ad absurdum being something like soda lime
I haven't looked into the topic much at all but that does resonate. It reminds of the way solar farms are becoming less fine-tuned (e.g., no sun-tracking tilt motors anymore) as panel costs drops through the floor.
But large enough that if you did similar 100 times you'd almost halve it!
SF6 is a significant greenhouse gas but what is troubling is that "SF6 is a very stable compound in the atmosphere, with an average residence time of about 3,200 years".
From https://www.energynetworks.org/assets/images/Resource%20libr... :
It is estimated that SF6 currently contributes about 0.2% of the global warming potential of the annual greenhouse gas emissions in the UK.
SF6 in electrical equipment is contained and not intentionally released during operation. Recycling of SF6 from electric power equipment, using commercial gas reclaimers, is becoming common practice [in the UK].
> About 1 percent of SF6 leaks from electrical equipment. In 2018, that translated to 8,200 tonnes of SF6 emitted globally, accounting for about 1 percent of the global-warming value that year.
There's something wrong with that figure. Global production was about 7500 tons in 1995, and countries have been trying to use less since then. 50% of production used in OEM equipment and 31% by Electric Utilities (maybe recharging breakers after inspections?). See: https://www.epa.gov/sites/default/files/2016-02/documents/co...
Interesting but just not practical, especially when there is a well proven system in use.
I get that we don't want SF6 released unnecessarily, but grid components need to be reliable, and a breaker that could turn into bomb if it loses temperature control seems loke a very bad idea.
PS: I had a small amount of SF6 injected into my eyeball a few years ago, and it helped save my sight.
I’m certain that I have seen CF3I (trifluoroiodomethane) used or proposed to be used as an arc quenching agent in HV breakers. It also may soon be replacing Halon (CF3Br, bromotrifluoromethane) for aircraft fire extinguishing.
At first I thought that would be wildly impractical but CO2 goes supercritical at just shy of 90°F and a little over 1kpsi. Those kinds of pressures are pretty well unknown in a substation, but occur all the time in refineries and the like. It's basically just a pressure vessel with a relatively mild heater requirement. Eminently doable. I think we will probably see these deployed in the next few decades. SF6 is an amazing material but really, really horrible as a greenhouse gas.
Super critical c02 is a regular extraction process method for certain molecules! You can find it in a number of laboratory applications. Anyway, yes, doable - essentially for the reasons you mention. But also literally done at the moment, for the general extraction process capabilities of supercrit CO2
For years they had an SF6 tank in my building because somebody had a wind tunnel (the first photo)
https://batl.mae.cornell.edu/facilities/
the machine is still there but they took out the SF6 tank.
Here's a guy playing with a supercritical CO2 pressure vessel on a table (seems a little brave to be shaking something at such high pressures!) https://www.youtube.com/watch?v=2d7RGQMCX24
SF6? The DRIE etchant gas?
Yes, but for DRIE you need plasma so you ionise it.
The article's title doesn't seem to be accurate.
- It doesn't exist (there is mention of a prototype - but that's not really a debut, all the images in the article are generic).
- The $3.9M funding ended May 18th https://arpa-e.energy.gov/programs-and-initiatives/search-al... and there doesn't seem to be a related paper yet (other than this IEEE "maybe" from April)
- The project is (unfortunately) called TESLA: Tough and Ecological Supercritical Line Breaker for AC 2022: https://www.sf6andalternativescoalition.org/wp-content/uploa...
OK, we retitled it to a more modest subtitle from the article.
Is there a word for this? Subpar lead-free solders and piezoelectrics, subpar coolants etc. Greenshittification?
Engineering involves trade-offs. It's easier when you don't consider all of them. But it's poor workmanship.
Isn't SF6 very dense?
Why would its greenhouse warming potential matter if it's never going high up into the atmosphere in the first place?
Gases in the atmosphere separate by molecular weight only at very high altitude, above the "homopause". This is from 80 to 120 km. Below that altitude, turbulence remixes gases faster than they can separate.
It will mix throughout the troposphere like any other gas.
Hey CO2 is still greenhouse. I guess not while it's contained and in its supercritical state, but if it's released to atmosphere then gas it will be.
Excited for the guys at Chalfont to test this out eventually, another fun chance to see if something goes boom :)
> Hey CO2 is still greenhouse
SF6 has 25,000 times the 100-year global warming potential as CO2.
You’d have to accidentally vent thousands of these breakers to equal the same global warming potential as one SF6 breaker
When it comes to topics like this, reducing it to a binary is-or-isn’t greenhouse gas is the wrong way to look at it.
I was responding to the badly editorialised title which read "without greehouse gases". I think I am on pretty solid ground here, the title misled.
Now the title reads "Reinventing". Is this a pejorative? Engineers make firsts too. TFA's title works just fine.
These breakers are carbon neutral because the CO2 comes either straight from the atmosphere or is harvested from sources that would otherwise be vented to the atmosphere.
This. I don't doubt that the new design is a vast emissions improvement. But the title clearly says without greenhouse gas, and CO2 is the poster child for greenhouse gases.
The thing that baffles me is folk in this community so willingly put their critical thinking skills to one side to downvote your post.