Helium Drives - Long-term use

[thomas-hn]

Nowadays, the bigger harddrives are often filled with Helium instead of air. However, Helium is a very volatile gas and I have problems in believing that the Helium stays inside the drive chassis forever (also the S.M.A.R.T. ID 22 "Helium Level" supports my fear).

Does anyone know for which duration those drives usually keep their Helium before it is gone by leakage?
What will happen to the drive after the Helium is gone (and is replaced by air)? Will the drive fail?
Does someone here have Helium drives running since a long period?

 

[Ericloewe]

What will happen to the drive after the Helium is gone (and is replaced by air)? Will the drive fail?

Short answer, yes.

Does someone here have Helium drives running since a long period?

Not really long, but coming up on two years, my WD100EFAX-68LHPN0 units are at 100% helium level.

 

[Constantin]

23K hours later, multiple units here are at 100%. How they determine that 100% is 100%... now that's a better question. My guess is drag and the attendant increase in motor power needed over time.

 

[thomas-hn]

As some time passed since the start of this thread, are there any further experiences with Helium filled drives so far?

 

[Jailer]

Helium is a very volatile gas

What??? By what measure do you consider helium volatile?

 

[Ericloewe]

Yeah, the same disks as above have been doing great. Cool, reliable, and 100% helium.

 

[Ericloewe]

What??? By what measure do you consider helium volatile?

Certainly not in the common sense, but in the technical sense of evaporates easily - although the danger, as it were, is more in terms of Helium atoms sneaking their way through seals and stuff, being rather small atoms. Still a lot easier to deal with than Hydrogen, which is small enough to insert itself into typical crystal structures of metals, thereby ruining the material's properties.

 

[danb35]

By what measure do you consider helium volatile?

in the technical sense of evaporates easily

Not really, helium is already a gas, so it's already evaporated unless you started with liquid He (unless that's what you meant, in which case I've missed the point). The real problem--and I'm sure there's a technical term for this too, but I don't know what it is--is that He is very hard to contain.

My 8 TB He-filled disks have been in service for nearly 5 years, and are all reporting 100% He remaining. But I'm also doubtful about how that's determined.

 

[Constantin]

I reckon the most likely measure is drag - measure power needed, or how long it takes to spin up a disk, or the PWM cycle needed to keep a disk at a given speed, etc. Many different indirect ways to assess drag and from there infer how much air has replaced the helium.

 

[Ericloewe]

Not really, helium is already a gas, so it's already evaporated unless you started with liquid He (unless that's what you meant, in which case I've missed the point).

Of course, though it is admittedly a weird thing in this context, since we're talking about room temperature, not 4 K. I need to make an effort to separate thoughts into separate sentences. It makes things easier to understand. Easier to separate concepts. Better for the reader.

 

[Arwen]

One consideration in regards to helium seepage, would be that the outside should have lower pressure than the inside. If both inside and outside are basically equal, any successful escape by some helium would cause a pressure differential. Unless regular air, (oxygen, nitrogen, or CO2), could replace the escaped helium, further escapes by helium would be prevented.

I may not be saying that clearly. But, hopefully you get the idea. (Elves aren't perfect with mortal science... Magic is so much easier, for us.)

 

[rogerh]

One consideration in regards to helium seepage, would be that the outside should have lower pressure than the inside. If both inside and outside are basically equal, any successful escape by some helium would cause a pressure differential. Unless regular air, (oxygen, nitrogen, or CO2), could replace the escaped helium, further escapes by helium would be prevented.

I think, from my dim memories of skool science in the 1960s, that the rate of helium loss will slow as the partial pressure (concentration) of the Helium falls, but the absolute pressure in the disk will make no difference. As there is virtually no Helium (zero partial pressure) outside, this process will in theory continue until there is a vacuum inside the disk. In practice, I suspect it will get very slow as the Helium gets very sparse inside, and in practice some air will leak in. But, to a first approximation, the pressure differential will make no difference.

 

[Johnny Fartpants]

I’ve ran literally thousands of these 8TB drives for the last 6-7 years and they have been great. Recently taken out of production but no issues related to them being Helium filled . https://documents.westerndigital.co...astar-sas-series/data-sheet-ultrastar-he8.pdf

 

[Constantin]

Here we are, two years later from my 23k hour post and all those He10s are still fine, passing smart tests, etc. Presumably, that means they’ve run about 40k hours by now. I’m going to stock up on qualified replacements ahead of time but I’m not going to replace them unless they actually fail.

 

[Yorick]

44k8 hours reported by smartctl, and 100 fill. No reallocations, no errors. That's just north of 5 years. Any year now for failures?