DCIEM and Equivalent Buhlmann Gradient Factors

[VE7DAC]

With Shearwater supporting DCIEM upgrades for commercial and military applications, I thought this might have been discussed here but haven't found much with searches.

I live and dive around Victoria BC, and have done some contract work on the boats that the Canadian Navy Fleet Dive Unit use. (They have some SWEET gas mixing setups, and they all dive Shearwaters BTW. Probably a big part of the reason DCIEM was ported.) I've heard some things about DCIEM tables and how they're more conservative than most recreational algorithms, but I don't see that when I actually look at the tables. I ran a few comparisons in Subsurface, and the DCIEM no-stop and deco tables seem to align quite closely with a 90/90GF, a far cry from the 50/80 I dive, and a lot closer to the US Navy dive tables. It makes sense for a military application, with the bend-em-and-mend-em attitude, but I would have expected a more conservative algorithm for commercial diving, and this is the one they use.

Before a spend a few hours/days of my life in excel comparing various Buhlmann gradient factors and the DCIEM data there an established way to model/evaluate different tables/algorithms? Surely someone has a MATLAB template or something? Maybe a standard set of dives to compare?

I'm specifically interested in seeing what set of gradient factors comes closest to DCIEM. More for education than anything else, but also because CAD$450 is goddamn criminal for a software update/algorithm addition. If I could be doing less deco, safely, with a rigorously tested algorithm developed right around the corner, I would be a very happy diver.

EDIT: Comparing NDLs is pretty simple, I'm doing that now. It doesn't seem too far off from the PADI recommendations, falling somewhere in between the no-stop and "mandatory safety stop" times. Deco comparisons seem trickier to do in a meaningful way.

 

[tursiops]

With Shearwater supporting DCIEM upgrades for commercial and military applications, I thought this might have been discussed here but haven't found much with searches.

I live and dive around Victoria BC, and have done some contract work on the boats that the Canadian Navy Fleet Dive Unit use. (They have some SWEET gas mixing setups, and they all dive Shearwaters BTW. Probably a big part of the reason DCIEM was ported.) I've heard some things about DCIEM tables and how they're more conservative than most recreational algorithms, but I don't see that when I actually look at the tables. I ran a few comparisons in Subsurface, and the DCIEM no-stop and deco tables seem to align quite closely with a 90/90GF, a far cry from the 50/80 I dive, and a lot closer to the US Navy dive tables. It makes sense for a military application, with the bend-em-and-mend-em attitude, but I would have expected a more conservative algorithm for commercial diving, and this is the one they use.

Before a spend a few hours/days of my life in excel comparing various Buhlmann gradient factors and the DCIEM data there an established way to model/evaluate different tables/algorithms? Surely someone has a MATLAB template or something? Maybe a standard set of dives to compare?

I'm specifically interested in seeing what set of gradient factors comes closest to DCIEM. More for education than anything else, but also because CAD$450 is goddamn criminal for a software update/algorithm addition. If I could be doing less deco, safely, with a rigorously tested algorithm developed right around the corner, I would be a very happy diver.

EDIT: Comparing NDLs is pretty simple, I'm doing that now. It doesn't seem too far off from the PADI recommendations, falling somewhere in between the no-stop and "mandatory safety stop" times. Deco comparisons seem trickier to do in a meaningful way.

The DCIEM and GF 45/95 (only 95 is important, of course) NDLs for 60 to 110 ft are within 1 minutes of each other. Oceanic PZ+ is also close.

 

[VE7DAC]

The DCIEM and GF 45/95 (only 95 is important, of course) NDLs for 60 to 110 ft are within 1 minutes of each other. Oceanic PZ+ is also close.

That's also what I'm seeing in my own comparisons. I just graphed 10-40M for PADI, DCIEM, and Buhlmann and the differences are vanishingly small with depth. The PADI table NDLs actually seem to be the most aggressive of them all until around 40M, probably because those come with a "mandatory safety stop". Because "deco stop" would be scary.

I played around with a few air deco scenarios from the DCIEM manual, and they all seemed to fit a 90 to 100GFHi and GFLow as well. I find myself wondering why DCIEM has a reputation as a conservative algorithm when it's so close to pure Buhlmann. Most people I know, and most of the recommendations from the docs on SB, seem to recommend a more conservative approach for cold water or higher exertion levels. I personally wouldn't choose to use it for a cold water recreational dive, much less for commercial diving applications. But it has been rigorously tested with doppler measurements in working divers in cold water.

All this to say, I have more questions now than when I started. Science!

 

[tursiops]

That's also what I'm seeing in my own comparisons. I just graphed 10-40M for PADI, DCIEM, and Buhlmann and the differences are vanishingly small with depth. The PADI table NDLs actually seem to be the most aggressive of them all until around 40M, probably because those come with a "mandatory safety stop". Because "deco stop" would be scary.

I played around with a few air deco scenarios from the DCIEM manual, and they all seemed to fit a 90 to 100GFHi and GFLow as well. I find myself wondering why DCIEM has a reputation as a conservative algorithm when it's so close to pure Buhlmann. Most people I know, and most of the recommendations from the docs on SB, seem to recommend a more conservative approach for cold water or higher exertion levels. I personally wouldn't choose to use it for a cold water recreational dive, much less for commercial diving applications. But it has been rigorously tested with doppler measurements in working divers in cold water.

All this to say, I have more questions now than when I started. Science!

I think much of DCIEM's reputation is from early on....especially in comparison to the old very aggressive USN tables. We are smarter now.

 

[dmaziuk]

@atdotde has a nutshell analysis here: The DCIEM model – The Theoretical Diver