Redistribution of decompression stops to optimise integral supersaturation

[David Carron]

.

 

[David Carron]

Did you look at the ISS when you do that?

There's no real reason to; the idea was to find the trade-off between the faster and greater desaturation at 3m against the overall effect on integral supersaturation. On Oxygen, executing the last stop shallower doesn't change the desaturation and can only increase the integral supersaturation so the trade-off is very much determined in advance.

 

[dmaziuk]

My point was that if you want to reduce the ISS, deco on pure O2 is probably the last thing you want.

And no, there should be no difference between 3 and 6 msw as on pure O2 the ambient inert gas pressure is 0 at either depth.

 

[David Carron]

My point was that if you want to reduce the ISS, deco on pure O2 is probably the last thing you want.

And no, there should be no difference between 3 and 6 msw as on pure O2 the ambient inert gas pressure is 0 at either depth.

I'm concerned here only with the optimizing the decompression strategy for single gas dives.

For decompression stops from 6m and shallower there is no doubt that, with Oxygen as a decompression gas, the optimal (as measured by ISS) decompression strategy is to remain at 6m.

 

[inquis]

These simulations are all for single gas air or nitrox dives

Please forgive me, but I vaguely recall that you were holding the total deco time constant, varying only the distribution thereof. Is that correct, and is that the time when the 3m stop is allowed?

I guess at this point, I'm wondering about the surfacing supersaturation (gradient factor). Are you ignoring that in favor of simply minimizing ISS? For that matter, with the exhaustive search of time distribution, it seems like you may be violating the traditionally computed ceiling at one or more stops in many of the trials. Can you please comment on whether ceilings are a consideration during the ascent?

 

[David Carron]

Please forgive me, but I vaguely recall that you were holding the total deco time constant, varying only the distribution thereof. Is that correct, and is that the time when the 3m stop is allowed?

I guess at this point, I'm wondering about the surfacing supersaturation (gradient factor). Are you ignoring that in favor of simply minimizing ISS? For that matter, with the exhaustive search of time distribution, it seems like you may be violating the traditionally computed ceiling at one or more stops in many of the trials. Can you please comment on whether ceilings are a consideration during the ascent?

So, in the original set of simulations I started with the total decompression time and then tried all combinations of stop depths and times, without reference to the instantaneous GF, retaining the one with lowest ISS.

Clearly a lot of these candidate ascents involve massive ceiling violations.

Perhaps surprisingly though, even though it doesn't involve any Bühlmann ceiling calculations, the optimal ISS based ascent is very similar to the ZHL 80/80 ascent although typically with a small ceiling violation in the middle stops and systematically with the last stop at 6m and then a small ceiling violation on surfacing.

The complete optimisation is way too intense to have any chance of running in real time on my wrist computer and since the profiles are basically the same anyway, I can simply use standard zhl programmed for and the last stop at 6m.

Executing the last stop at 6m for the same GFhi extends the total deco time, so this last round of simulations was to confirm that this distribution of stops is still basically optimal for that extended time.

Long story short, using ZHL with last stop at 6m appears to be close to optimal in terms of ISS without requiring any real-time optimisation calculation.

 

[inquis]

Long story short, using ZHL with last stop at 6m appears to be close to optimal in terms of ISS without requiring any real-time optimisation calculation.

Thanks for the elaboration. I'd be curious as to the times and ISS values for a 80/80 ascent with last stop at a) 6m and b) 3m. That seems to be the front-runners, practically speaking. Oh, and which gas you assumed for those cases.

 

[David Carron]

Thanks for the elaboration. I'd be curious as to the times and ISS values for a 80/80 ascent with last stop at a) 6m and b) 3m. That seems to be the front-runners, practically speaking. Oh, and which gas you assumed for those cases.

I had to reassemble the simulator and haven't actually looked at the numbers so although I have no real reason to think so, it's not impossible that it's total gibberish.

First column is depth in cm
Second column is descent plus bottom time in seconds
Third and fourth columns are total deco time in seconds and ISS assuming last stop at 3m
Fifth and sixth columns are total deco time in seconds and ISS assuming last stop at 6m

These are all 80/80 air dives with air deco, ascent 10m/min, descent 20m/min, 1030g/l sea water, 1013mb surface pressure.

 

[inquis]

@David Carron - Thanks so much for those runs. Generally speaking, there's a fairly consistent reduction in ISS of about 5-10% with a 6m final stop (compared to the 3m). The penalty one pays for this benefit is an average of 50-100% more time in the water. For a bottom-gas deco dive, though, I suspect a realistic limit would be about 20 minutes of deco time before multiple gases are far more attractive. On the other hand, I'd expect these general trends to hold for a CCR dive, though with significantly reduced benefits/penalties.

Zooming in on the 20-minute region:

 

[David Carron]

Might be worth plotting things against the traditional PrT measure of engagement to see if any pattern emerges

ISS(6m)/ISS(3m) against PrT