Deep Stops Increases DCS

[rossh]

Interestingly, to me, the Navy 2008 profile is an identical runtime to VPM-B but with the shallower stops extended dramatically. I suspect that if the NEDU test had been run on those 2 schedules, the results would have been the same but we might spend a 1000 posts less arguing about actual vs "fudged" profiles etc etc.
It also seems that the conservancy settings on VPM seem to less about adding a padding factor (like lowering GF High would achieve) and more about undoing the tissue stress from the deeper stops. So why overdo the deeper stops to start with?

The Nedu test was looking at VVAL18 model and BVM(3) model profiles. Neither of these can generate anything close to a ZHL-C or VPM-B dive. Same in reverse; ZHL-C and VPM-B cannot generate VVAL18 or BVM(3) profiles.

nedu graphs

***************

GF on ZHL lets you make some very big changes. So much, that It allows you to completely abandon the underlying ZHL models concepts. The way its used today - that's pretty much what happens.

Conversely, VPM-B conservatism will always make a VPM-B style plan with the same model attributes - no escaping that. But taken too far, it looses connection and context to the pressures and stresses its trying to manage - hence faked up +7 profiles are rubbish.

Adding more conservatism to VPM-B, lowers the stress across the whole dive, and reduces the surface starting pressure - quite evenly too.

 

[UWSojourner]

Are you saying that GF 100/100 is a better option than VPM?

I'm having a hard time seeing how you came up with that question from my post here?

That post was a response to the issue of 'insufficient shallow time' from Diver0001. It's purpose was to show in the deeper/longer dives Diver0001 brought up that VPM indeed (after the effects of the CVA diminish) simply increases risk as you progressively lengthen bottom time. So with VPM you need to to either 1) insert shallow time, 2) increase VPM conservatism, or 3) accept the steadily increasing risk as you lengthen bottom time.

The shortcomings of VPM highlighted in that post will likely be part of any postmortem critique of VPM IMO (in addition to the too heavy emphasis on deep stops). It's an oversight, or maybe a failure of curiosity, in not asking: If I choose a critical radius and it cuts off growth in part of the bubble distribution (according to VPM theory), what is happening to the part of the bubble distribution that does start growing? As highlighted in that post the answer is "VPM just accepts that it will continually grow", or another way to say it is VPM just accepts continually increasing risk as dives get more extreme (e.g. as bottom times get longer).

Like I said, that's ok if the diver knows that as dives get more extreme the DIVER is responsible to continually pad shallow stops, increase conservatism, etc. But that's hard for divers to know when you have someone running around saying "VPM scales consistently and in the right proportions." It doesn't (as discussed in that post). Luckily, I think most divers have figured that out and ARE padding shallow time, or merging Buhlmann tails to VPM, or simply moving to another method like GF in response to the NEDU study.

 

[RainPilot]

The Nedu test was looking at VVAL18 model and BVM(3) model profiles. Neither of these can generate anything close to a ZHL-C or VPM-B dive. Same in reverse; ZHL-C and VPM-B cannot generate VVAL18 or BVM(3) profiles.

nedu graphs

***************

GF on ZHL lets you make some very big changes. So much, that It allows you to completely abandon the underlying ZHL models concepts. The way its used today - that's pretty much what happens.

Conversely, VPM-B conservatism will always make a VPM-B style plan with the same model attributes - no escaping that. But taken too far, it looses connection and context to the pressures and stresses its trying to manage - hence faked up +7 profiles are rubbish.

Adding more conservatism to VPM-B, lowers the stress across the whole dive, and reduces the surface starting pressure - quite evenly too.

Wow. Just..wow. Like I said, IF they had run these other algorithms as they were, we would be having a lot less arguing about it like you have just squeezed in even though it was apropos of nothing.


On your second point:
In the table I quoted, the difference between VPM and VPM + 2 was almost entirely shallow time extra so yes, increasing conservatism does reduce stress by reducing the percentage of total deco spent deep, or so it seems.

Also, GF to the best of my knowledge doesn't "abandon" the ZHL at all. It allows (for me, maybe Im way off base) a true conservatism option by allowing the user to determine their own margin between the absolute M values and the ones they are prepared to accept, by adjusting GF High. So, for a GF of for example 80/80 the profile would be very similar to a GF 100/100 but the total saturation would have a 20% margin built in above the original ZHL values. As I age or as I have other issues (cold dives etc etc) I could shift to 75/75 or 60/60, obviously at the expense of longer deco times but thats the price of doing business for me.

Of course, when fiddling with GF lo, the potential for skewing the stops deeper exists, I suspect that the result of this thread for many (and definitely for me) will be a raising of GF Lo to a number closer to the GF Hi which, in my case, is quite low to start with.

 

[scubadada]

As I follow this interesting thread, I keep struggling to figure out how it applies to the average diver, diving acceptable profiles. At the end of his presentation, “Decompression Controversies”, Dr Mitchell states that the data has resulting in him diving a profile of 50/70 or 50/75. He appears to put his money where his mouth is and has raised the GF lo above that many typically use

It is very interesting, that for the NEDU profile of 170 feet for 30 min on air, the decompression time for 50/70 is very close to that for VPM+4 and 50/75 is very close to that for VPM+3. The difference is in the distribution of the stops between deep and shallow, and is pretty striking. This has given me something tangible to think about. Now, if there was only some good data to compare these strategies... I'd love to see @UWSojourner's analysis of supersaturation associated with these profiles.

 

[EFX]

....... In contrast, GF tries to account for the longer exposure time by reducing SS (i.e. Buhlmann reduces M-values in slower tissues for this reason). That's rational.

VPM-B is a dual phase deco model. For it's dissolved gas phase does it base it's calculations on Buhlmann M-values (neglect GF's and other conservation factors for now)?

 

[UWSojourner]

VPM-B is a dual phase deco model. For it's dissolved gas phase does it base it's calculations on Buhlmann M-values (neglect GF's and other conservation factors for now)?

No.

See this article for a basic description. From that article,

"The VPM stages divers by limiting tissue tensions during ascent. In this sense, VPM produces its own “vpM-Values”. But, rather than being derived empirically, these vpM-values are determined by the no-bubble-growth equation. The no-bubble-growth equation is dependent on the specific characteristics of your dive, the parameters of VPM, and their combined effect on the Yount bubble. But, it is important to understand that the VPM still stages divers by limiting dissolved gas tensions, the same basic staging paradigm used by Buhlmann."

Full disclosure just so the-person-who-shall-not-be-named won't freak -- I wrote the article.

 

[rossh]

VPM-B is a dual phase deco model. For it's dissolved gas phase does it base it's calculations on Buhlmann M-values (neglect GF's and other conservation factors for now)?

A picture is worth a thousand words:

 

[rossh]

Wow. Just..wow. Like I said, IF they had run these other algorithms as they were, we would be having a lot less arguing about it like you have just squeezed in even though it was apropos of nothing.


On your second point:
In the table I quoted, the difference between VPM and VPM + 2 was almost entirely shallow time extra so yes, increasing conservatism does reduce stress by reducing the percentage of total deco spent deep, or so it seems.

Also, GF to the best of my knowledge doesn't "abandon" the ZHL at all. It allows (for me, maybe Im way off base) a true conservatism option by allowing the user to determine their own margin between the absolute M values and the ones they are prepared to accept, by adjusting GF High. So, for a GF of for example 80/80 the profile would be very similar to a GF 100/100 but the total saturation would have a 20% margin built in above the original ZHL values. As I age or as I have other issues (cold dives etc etc) I could shift to 75/75 or 60/60, obviously at the expense of longer deco times but thats the price of doing business for me.

Of course, when fiddling with GF lo, the potential for skewing the stops deeper exists, I suspect that the result of this thread for many (and definitely for me) will be a raising of GF Lo to a number closer to the GF Hi which, in my case, is quite low to start with.

If GF was used to just add a little bit... like 90/90. then one could easily keep the assumption of association to it base ZHL model.

But its not used that way any more. It can be made to mimic a bubble model plan, and shallow stop navy model plan. Or an inflated doubled sized anything plan. None of those things are associated to ZHL model characteristics. When you start adding 30% ++ (x/70) and all the associated compounding errors of GF, the only connection you have is a loose association to the Haldane gas tracking. You have completely overridden ZHL intentions and attributes. You can play percentages, but in reality, such large changes means you have invented your own new attributes. Good luck.

 

[David Doolette]

If GF was used to just add a little bit... like 90/90. then one could easily keep the assumption of association to it base ZHL model.

But its not used that way any more. It can be made to mimic a bubble model plan, and shallow stop navy model plan. Or an inflated doubled sized anything plan. None of those things are associated to ZHL model characteristics. When you start adding 30% ++ (x/70) and all the associated compounding errors of GF, the only connection you have is a loose association to the Haldane gas tracking. You have completely overridden ZHL intentions and attributes. You can play percentages, but in reality, such large changes means you have invented your own new attributes. Good luck.

So you are agreeing with RainPilot, that increasing the GF-Lo to a larger value (larger than 30%) is a good idea?

 

[EFX]

No.

See this article for a basic description. From that article,

"The VPM stages divers by limiting tissue tensions during ascent. In this sense, VPM produces its own “vpM-Values”. But, rather than being derived empirically, these vpM-values are determined by the no-bubble-growth equation. The no-bubble-growth equation is dependent on the specific characteristics of your dive, the parameters of VPM, and their combined effect on the Yount bubble. But, it is important to understand that the VPM still stages divers by limiting dissolved gas tensions, the same basic staging paradigm used by Buhlmann."

Full disclosure just so the-person-who-shall-not-be-named won't freak -- I wrote the article.

Thanks for the link to the article. It supports what I already knew from previous reading that VPM uses Buhlmann with it's M-values to limit SS on ascent. Moving on, you wrote in a previous post .....

No. Read the post. VPM just keeps the surfacing supersaturation constant (within the constraints I discussed in the post). Are you exposed 45 minutes? VPM: Use SS=600mb. Are you exposed 90 minutes? VPM: Use SS=600mb . Are you exposed 120minutes? VPM: Use SS=600mb.

Obviously you will just get progressively more bubble formation (i.e. risk) as you do this. In contrast, GF tries to account for the longer exposure time by reducing SS (i.e. Buhlmann reduces M-values in slower tissues for this reason). That's rational.

If VPM is using the Buhlmann algorithm or something essentially the same won't it do the same SS limiting as a Buhl-GF? That is, it's going to limit the supersaturation at the limit as prescribed by the M-values for each controlling TC as adjusted by GF-Hi. It would seem your comment in parenthesis above would apply to both VPM and Buhl-GF since they both are using the same algorithm and m-values.