Hacker News Reader

Comment by JumpCrisscross | Hacker News Reader

> wonder how they would remake the battery chemistry if they were to use all the latest advancements

“While the energy density is only around one third as that of a lithium battery, the distinctive virtue of the nickel–hydrogen battery is its long life: the cells handle more than 20,000 charge cycles with 85% energy efficiency and 100% faradaic efficiency” [1].

[1] https://en.m.wikipedia.org/wiki/Nickel%E2%80%93hydrogen_batt...

View on HN

"It differs from a nickel–metal hydride (NiMH) battery by the use of hydrogen in gaseous form, stored in a pressurized cell at up to 1200 psi (82.7 bar) pressure."

whoah

that's not that much. Hydrogen tanks in cars like the Toyota Mirai for instance, stores it at 10000 psi.

That's nuts. 10k PSI. My car has a a CNG tank that is at several hundred PSI and that makes me nervous when filling it up, but 10,000 PSI is way way more than this.

This is one of the reasons hydrogen cars are unlikely to achieve any widespread deployment. But the new proposal is hydrogen-carrying trucks, which is better (larger tanks and bigger scale means more and better maintenance than a typical passenger vehicle can justify) but also, it means extremely high pressure hydrogen tanks on the road with your traffic.

(Of course, it may not happen. Trucking hydrogen around is pretty inefficient and stupid: hopefully it will prove cheaper to pipe electrons around and generate hydrogen on site for whatever industrial process requires it.)

   I owned a number of 2001 CNG Chevy Cavaliers around 2010, purchased through govt. surplus auctions.  Their tanks were 3600 PSI.  The tanks were certified for 15 years with no recertification.  Nothing would stop working but they would no longer be certified.  The tank would get hot while filling so I imagine the fatigue from many cycles of heating up was one factor in the certification period.  Internal corrosion is another factor.  If the natural gas compressor farm does not dry the compress gas then moisture will get into the tank and over many years will corrode the tank.  There's a video out there of a CNG tank explosion at a fueling station somewhere in South America.  No doubt metal fatigue plus corrosion contributed to that failure.   
   So 10,000 PSI for hydrogen is a lot of pressure to be transporting around in a vehicle for multiple years of heating / cooling and possible corrosion.

Scuba tanks are usually filled to 200 bar/3000 psi

And pressurized water reactors are a huge industrial installation running at 155 bar.

There’s a lot more stainless steel and concrete in a pressurized nuclear reactor than an orbiting space station

that should definitely make it disassemble rapidly

Hopefully only on schedule.

Fwiw, EVE's lf560k lifepo4 cell is finally out now & supposedly good for 12k cycles.

I haven't heard the term faradaic efficiency before but im super super super happy to hear of it! I've been so curious, I just never had the term. Apologies if I mess this up but it seems to be the efficiency of turning input charge to stored charge. So curious but never figured out what kind of figures to expect from lipo or lifepo4.

I'm not sure if there's another term for efficiency of that stored charge being released/sourced, or if that tends to be >99% in most processes or what not.

Faradiac efficiency is essentially electron counting: e.g. at 100pct efficiency you need a mole of electrons to convert a mole of Li+ to Li0. At 50pct 2 moles. Moles of electrons can be converted to amp-hrs for a more conventional unit.

Numbers are probably reasonably high (99+%) for commercial batteries, as the only two losses are ohmic (electric conduction through the cell, which is roughly the self discharge rate) and chemical (side reactions of electrolyte - which result in the degredation of the cell.

If you want so dive further down the rabbit-hole look up the Nernst equation.

Comes up more in electrochemistry than batteries, as energy efficiency is the dominant factor in the latter (watts_out/watts_in), noting that faradiac efficiency doesn't consider voltage.

Reply on news.ycombinator.com