Does Subsurface penalize for Helium?

[Jiri Pokorny]

Based on following article:
https://www.shearwater.com/monthly-blog-posts/eliminating-helium-penalty/

Does Subsurface penalize for Helium?

I tried to search project history, but didn't found any related change.

[Linus Torvalds]

On Sun, May 14, 2023 at 8:36 AM Jiri Pokorny <j....@post.cz> wrote:
>
> Based on following article:
> https://www.shearwater.com/monthly-blog-posts/eliminating-helium-penalty/
> Does Subsurface penalize for Helium?

Yes, we use the bog-standard Bühlmann tables, which have Helium with a
halflife that is 2.65 times shorter than Nitrogen.

See buehlmann_He_t_halflife[] in core/deco.c (and contrast it to
buehlmann_N2_t_halflife[]).

Linus

[Robert C. Helling]

Hi,

On 14. May 2023, at 18:47, Linus Torvalds wrote:

Does Subsurface penalize for Helium?

Yes, we use the bog-standard Bühlmann tables, which have Helium with a
halflife that is 2.65 times shorter than Nitrogen.

See buehlmann_He_t_halflife[] in core/deco.c (and contrast it to
buehlmann_N2_t_halflife[]).

It is true that Subsurface uses the Bühlmann algorithm (with gradient factors) which prescribes different half-times for helium. But things are more complicated:

The compartments that the model simulates do not correspond to specific prescribed compartments of the human body. Rather, a better way to refer to a compartment than its compartment number is to refer to it by its half-time. So, changing all half-times by a constant factor amounts really only to a relabelling of the compartments. That by itself does not really have any effect (after all, a compartment that takes up a gas faster also releases it faster by the same factor). 

You really have to look at how the limits (that compute the ceiling) on overpressure are related to half-times (rather than imaginary compartment number). And once you do that, even the old Bühlmann model does not show much difference for nitrogen compared to helium. 

I have blogged about this a while ago with some plots that show the effect here: 

https://thetheoreticaldiver.org/wordpress/index.php/2017/11/02/n2-vs-he-whats-the-difference/

The actual difference in treating N2 and He is how you combine those when you have both in the mix. This is where the different half-times really make a difference: According to Bühlmann, you need to add N2 and He for the compartment with the same compartment number (not the same half-time). And this is where with trimix you seem to have an advantage in the model. But I think the empirical basis for this combination rule is not really strong anyways.

The second difference between the gases is that at the start of the dive, you body is saturated with N2 at the surface and has 0 helium, so helium has a starting advantage here.

I should add that I do not know the specifics of the decompression model that NEDU uses. But it is not obvious that it really applies to dives of recreational (as in not professional navy) tec divers: First of all, I think it is really hard to establish probabilistic models (see https://thetheoreticaldiver.org/wordpress/index.php/2023/02/13/probabilistic-decompression-models-probably-problematic/ ) but note also that they are dealing with schedules that predict a probability of a few percent of DCS cases. Once more, it is clear that without that you will not get any significant data from a study with only 50 dives, but on the other hand, this rate is vastly higher than what is acceptable to divers that are not diving with a decompression chamber close by for practically every dive. You probably want your risk to be about two orders of magnitude smaller and it is far from obvious that any conclusions extrapolate to this very different regime of the model.

And finally, as also the article from the Shearwater blog concludes, there is no model available today that would use the findings of this study and thus there is nothing for Subsurface to implement at this time. Of course you can lie to the algorithm about the gas you are using but I would be very hesitant to do that.

Best

Robert

[Jiri Pokorny]

Sure Robert, i red your post, since it is linked in the Shearwater article. I hope i understood your explanation very well.

I only wanted to confirm, that after the the Shearwater article there was no change in the Subsurface behavior (which I didn't found).
Thanks

Dne neděle 14. května 2023 v 21:53:41 UTC+2 uživatel Robert C. Helling napsal: