Evolving Thoughts on Deep Decompression Stops

[tursiops]

I wonder if diving in a chamber watching theoretical tissue loading graphs on a big screen TV would help us understand more.....

I hope not. Otherwise we'd all want to do it.....

 

[dmaziuk]

A model lacks general context for the thing it models. It works the same way for Boeing stall prevention software as it does for "racially-profiling" facial recognition software as it does for a decompression model: it doesn't know what's really going on with the actual physical entity. Dr Pyle may have dived 100/100 to the letter and the big theoretical tissue loading graph showed he was perfectly OK. Only he didn't feel that way.

 

[dmaziuk]

where did this idea come from?

I work in risk assessment, we use thresholds to decide what is/isn't acceptable all the time...

I'm talking logic, not risk assessment. The base premise is under 100/100 is "safe enough", over 100/100 isn't. For whatever numerical definition of "safe enough" one happens to be using. 100/70 is not pushing you over 100/100 therefore it's also "safe enough" -- it may be "safer" and that "safer" may be statistically identical to "safe enough", but it can't be "less safe". By definition.

Or the base premise is wrong: enter the other models.

PS. by the same logic, if 50/75 is "safe enough", then any combination of GF Lo <= 50 and GF Hi <= 75 is also safe enough: 5/75, or 50/7.5 can't be "less safe" than your upper limit of 50/75.

 

[tursiops]

I'm talking logic, not risk assessment. The base premise is under 100/100 is "safe enough", over 100/100 isn't. For whatever numerical definition of "safe enough" one happens to be using. 100/70 is not pushing you over 100/100 therefore it's also "safe enough" -- it may be "safer" and that "safer" may be statistically identical to "safe enough", but it can't be "less safe". By definition.

Or the base premise is wrong: enter the other models.

PS. by the same logic, if 50/75 is "safe enough", then any combination of GF Lo <= 50 and GF Hi <= 75 is also safe enough: 5/75, or 50/7.5 can't be "less safe" than your upper limit of 50/75.

Wow, talk about a post adding zero value to the discussion.....

 

[ChuckP]

The fundamental problem is the basic premise of the dissolved gas model: if you stay under M-value line, you run acceptably low risk of clinical DCS. It follows that any combination under 100/100 runs acceptably low risk of clinical DCS and 100/70 must be "no less safe" and nor is 10/100. Conversely if 100/70, or 10/90, is "riskier" than 100/100, then the model is built on a faulty premise and is therefore wrong.

I'm trying to understand this direction. Basically arguing deep stops are OK because you're below the M value still:


DIVE PLAN
Surface interval = 1 day 0 hr 0 min.
Elevation = 0ft
Conservatism = GF 100/100

Dec to 200ft (2) Trimix 19/32 80ft/min descent.
Level 200ft 20:30 (23) Trimix 19/32 1.34 ppO2, 111ft ead, 125ft end
Asc to 100ft (26) Trimix 19/32 -30ft/min ascent.
Asc to 60ft (27) Nitrox 37 -30ft/min ascent.
Stop at 60ft 0:20 (28) Nitrox 37 1.04 ppO2, 41ft ead
Stop at 50ft 1:00 (29) Nitrox 37 0.93 ppO2, 33ft ead
Stop at 40ft 3:00 (32) Nitrox 72 1.59 ppO2, 0ft ead
Stop at 30ft 3:00 (35) Nitrox 72 1.37 ppO2, 0ft ead
Stop at 20ft 7:00 (42) Nitrox 72 1.16 ppO2, 0ft ead
Stop at 10ft 12:00 (54) Nitrox 72 0.94 ppO2, 0ft ead
Surface (54) Nitrox 72 -20ft/min ascent.

OR

DIVE PLAN
Surface interval = 1 day 0 hr 0 min.
Elevation = 0ft
Conservatism = GF 10/100

Dec to 200ft (2) Trimix 19/32 80ft/min descent.
Level 200ft 20:30 (23) Trimix 19/32 1.34 ppO2, 111ft ead, 125ft end
Asc to 140ft (25) Trimix 19/32 -30ft/min ascent.
Stop at 140ft 1:00 (26) Trimix 19/32 0.99 ppO2, 74ft ead, 85ft end
Stop at 130ft 1:00 (27) Trimix 19/32 0.94 ppO2, 68ft ead, 78ft end
Stop at 120ft 1:00 (28) Trimix 19/32 0.88 ppO2, 62ft ead, 71ft end
Stop at 110ft 1:00 (29) Trimix 19/32 0.82 ppO2, 56ft ead, 64ft end
Stop at 100ft 1:00 (30) Nitrox 37 1.49 ppO2, 73ft ead
Stop at 90ft 1:00 (31) Nitrox 37 1.38 ppO2, 65ft ead
Stop at 80ft 1:00 (32) Nitrox 37 1.27 ppO2, 57ft ead
Stop at 70ft 1:00 (33) Nitrox 37 1.15 ppO2, 49ft ead
Stop at 60ft 1:00 (34) Nitrox 37 1.04 ppO2, 41ft ead
Stop at 50ft 1:00 (35) Nitrox 37 0.93 ppO2, 33ft ead
Stop at 40ft 3:00 (38) Nitrox 72 1.59 ppO2, 0ft ead
Stop at 30ft 4:00 (42) Nitrox 72 1.37 ppO2, 0ft ead
Stop at 20ft 7:00 (49) Nitrox 72 1.16 ppO2, 0ft ead
Stop at 10ft 13:00 (62) Nitrox 72 0.94 ppO2, 0ft ead
Surface (62) Nitrox 72 -20ft/min ascent.


You spend 6 more minutes or ~10% of dive time under the shallower 60' stop.......

This is above my pay grade, I'm just trying to learn

 

[Fastmarc]

This was a good read. Thanks @boulderjohn

 

[dmaziuk]

I'm trying to understand this direction. Basically arguing deep stops are OK because you're below the M value still:

It's the reason pendulum swung towards all these "other models". Because you can fairly easily argue the dissolved gas one into self-contradiction. People who turn this stuff into computer programs really hate that: it doesn't compute.

PS whe you say "deep stop", what do you mean exactly?

 

[Manatee Diver]

It's the reason pendulum swung towards all these "other models". Because you can fairly easily argue the dissolved gas one into self-contradiction. People who turn this stuff into computer programs really hate that: it doesn't compute.

Personally I think both sides are right, the bubble model guys are right that the dissolved gas model doesn't really account for how bubbles are really formed. But their solution is shown empirically to be wrong, and that the tables derived from the dissolved gas models have about a 100 years of empirical testing, and right now are empirically shown to be safer, but can't explain how micro bubbles are formed even though they are well below the critical M values.

But TBH that is something well above my pay level, I'll let the PhDs figure it out. And dive within limits that are shown to be safe.

 

[ChuckP]

PS whe you say "deep stop", what do you mean exactly?

I mean deep, deeper than otherwise - I dunno - again, trying to learn and understand.

I forget who's presentation it was, Dr Pollock maybe, showing tissue saturation graphs after two different deco strategies, one with deeper stops than the other. The graphs really made it easy to see the different end result and made sense, that's why I was serious about sitting in a chair watching tissue graphs in a chamber......

 

[dmaziuk]

Personally I think both sides are right, the bubble model guys are right that the dissolved gas model doesn't really account for how bubbles are really formed. But their solution is shown empirically to be wrong, and that the tables derived from the dissolved gas models have about a 100 years of empirical testing, and right now are empirically shown to be safer, but can't explain how micro bubbles are formed even though they are well below the critical M values.

What's missing is Hill's "economical on time": the "bad" deep stops is where bubble models create deeper stops at the expense of the shallow ones. That is entirely different from ZHL+GF because when Baker came up with the idea, he wrote "the addition of deeper stops in a profile will generally increase the time required at the shallow stops as well as the overall decompression time. However if truly 'sufficient decompression' is the result, then the concept of 'economic decompression' is not really compromised."

So when someone argues 10/90 "bad" what they're really arguing is it's "uneconomical" -- exactly as designed. That's operator failure: if you were optimizing for time, you should have picked a different set of gradient factors.

"10/90 bad because bubble models" is not even apples and oranges, it's simply... confused.