Feedback on recent two-tank and dive limits

[old frogman]

If you gentlemen have bothered to follow the history of decompression development since Haldane, you may note that up until the late 1980s most recreational divers were using the US Navy dive tables which were based on probabilistic modelling with a DCS risk of about 3% through the expected depth range of scuba diving. In the late 1980s PADI came out with the RDP tables which were based on the US Navy tables with a safety margin for each depth/time profile. In addition, a safety stop was recommended for the deeper depth or for times that approached the NDL limit. I hope you guys are following this.

The dive computer I use for recreational dives employs the DSAT algorithm which very closely follows the PADI RDP table. I know this because I have checked the dive plan mode on the dive computer against the RDP for square dive profiles. Again, I hope you guys are following this.

I therefore make the educated assumption that my dive computer operates within the + or - 3% risk probability for DCS. Conversely, you could then say that the risk probability of not getting a DCS hit is about 97% (which sounds better).

Now let's look at the Buhlmann table. It is based on the deterministic model. It has been clinically validated in Mr Buhlmann's laboratory but has had nowhere the extensive empirical testing the US Navy tables have received.

From reading Erk Baker's paper on the Buhlmann GF algorithm my take is that it is based on a formula which mathematically makes an interpolation between a low GF and a high GF resulting in an ascent profile which produces a dive ascent profile below the M value ceiling.

Consequently, the Buhlmann GF algorithm is deterministic and hence theoretical. It has not been as thoroughly validated (as the US Navy tables and its associated Thalman algorithm). In other words, each Buhlmann GF calculated ascent profile is a theoretical solution.

The US Navy EDU tested the Buhlmann GF algorithm for various depth/time profiles and found considerable variance from the desired 3%. DCS risk. That is why the US Navy did not adopt it.

Consequently, for basic recreational dives I prefer on of my trusty old computers with DSAT. I also vary my safety stop based on knowledge, experience, common sense and situational awareness.

From a number of posts, I strongly suspect most of you have bought expensive dive computers and are probably using them only on the preset or default GF.

 

[inquis]

I therefore make the educated assumption that my [DSAT] dive computer operates within the + or - 3% risk probability for DCS

No, DSAT does not bend someone on 3 out of 100 NDL dives (on average) -- far from it, actually. You should read up on it's development sometime. For diving within one day, over 900 dives (some repetitive, some single) were tested with zero DCS incidents. For tests involving 6 consecutive days of diving with 4 dives every day, there were nearly 500 dives conducted, also with zero DCS incidents.

Additionally, Buhlmann+GF with the common presets of x/75, x/85, or x/95 have NDL bottom times either less than or comparable to DSAT. On average, it's even safer than DSAT.

 

[boulderjohn]

In the late 1980s PADI came out with the RDP tables which were based on the US Navy tables with a safety margin for each depth/time profile. I

Well, here's another thing you don't know a whole lot about. The PADI RDP was not based on the US Navy tables. Far from it.

The PADI RDP was created after extensive research on hundreds of divers using doppler bubble imaging. Their work led to many publications in scientific journals. One of the researchers on the project, Dr. Michael Powell, was a decompression specialist at NASA, and he was Dr. Deco on ScubaBoard for many years.

 

[old frogman]

Well, here's another thing you don't know a whole lot about. The PADI RDP was not based on the US Navy tables. Far from it.

The PADI RDP was created after extensive research on hundreds of divers using doppler bubble imaging. Their work led to many publications in scientific journals. One of the researchers on the project, Dr. Michael Powell, was a decompression specialist at NASA, and he was Dr. Deco on ScubaBoard for many years.

The US Navy Tables were the basis for developing the RDP. The doppler bubble imaging was used to "validate" the ability of each modified schedule for effectively reduce bubbles.

The doppler was a validating tool not a development tool. It was used at the end of each dive sequence to see the level of bubbling. Dr Spencer who developed the tables also developed the Spencer Bubble Grading Scale for measuring bubble propagation.

Several other tables independently preceded the PADI RDP, for example the Huggins tables which also evolved from the US Navy Diving tables.

If you read your history correctly, things do not suddenly develop out of nowhere, development is a continuum based on previous ideas.

Mate, get you facts straight.

 

[Blackcrusader]

The dive computer I use for recreational dives employs the DSAT algorithm which very closely follows the PADI RDP table. I know this because I have checked the dive plan mode on the dive computer against the RDP for square dive profiles. Again, I hope you guys are following this.

I therefore make the educated assumption that my dive computer operates within the + or - 3% risk probability for DCS. Conversely, you could then say that the risk probability of not getting a DCS hit is about 97% (which sounds better).

You educated assumption is not correct as recreational divers are doing multilevel dives not square profile dives. Therefor your claim about getting a DCS hit is also not trustworthy based on your assumptions. When I do a dive vacation I oten do 30 - 40 dives 3 - 4 dives a day for 2 days.
Never had a DCS hit. This includes doing NDL and Deco dives on the same day.

This is about as "square" as any of my recreational / deco dives would ever get. I doubt your Padi RDP or DSAT tables would even come close for this dive for me knowing what my NDL is going to be.

I have not seen any guide or diver looking at any tables for recreational diving for the last 15 years.
Everyone uses their DC and NDL given in real time or divers without DC's simply follow a guide or their dive buddy. Also what does it matter if divers are using default settings on their DC's? First you claim they should not change GF's then you go on about divers using default settings as if that is also wrong.

 

[old frogman]

No, DSAT does not bend someone on 3 out of 100 NDL dives (on average) -- far from it, actually. You should read up on it's development sometime. For diving within one day, over 900 dives (some repetitive, some single) were tested with zero DCS incidents. For tests involving 6 consecutive days of diving with 4 dives every day, there were nearly 500 dives conducted, also with zero DCS incidents.

Additionally, Buhlmann+GF with the common presets of x/75, x/85, or x/95 have NDL bottom times either less than or comparable to DSAT. On average, it's even safer than DSAT.

Probability is the mathematical potential of something happening, not that something is going to happen and I did include a + and -.

Looks to me that you young lads do not read things properly.

 

[Blackcrusader]

Probability is the mathematical potential of something happening, not that something is going to happen and I did include a + and -.

Looks to me that you young lads do not read things properly.

We just go doing multilevel NDL and Deco dives and not get DCS hits.
So if I have over 3500 dives and no DCS hit's hows that work out for your mathematical probablility of me getting a DCS hit?

 

[JMarc]

If you gentlemen have bothered to follow the history of decompression development since Haldane, you may note that up until the late 1980s most recreational divers were using the US Navy dive tables which were based on probabilistic modelling with a DCS risk of about 3% through the expected depth range of scuba diving.

I'd like your source for the probabilistic modelling used for tables in use in the early and mid 80s. I'm just aware of probabilistic modeling done during the mid to late 80's and not aware of any table nor dive computer which used those models (although the probabilistic models were used in experiments and in the parameter determination and validation of other models).

BTW, I know of no used model which keep a constant DCS risk on the range of depth and time. Everything that I'm aware of indicates that the risk increase with the depth and time of the dive. US Navy table used in the 80's are not an exception.

 

[JMarc]

So if I have over 3500 dives and no DCS hit's hows that work out for your mathematical probablility of me getting a DCS hit?

To be fair I doubt that most of your dives were to the limit of the tables (which would mean square dive profile), or of your computer settings (which would mean immediately surfacing after reaching a NDL of 0 or just after clearing your deco obligation if you had any). When expressing the risk of a deco method, it is the risk at just the limit of what the method allows.

 

[Blackcrusader]

To be fair I doubt that most of your dives were to the limit of the tables (which would mean square dive profile), or of your computer settings (which would mean immediately surfacing after reaching a NDL of 0 or just after clearing your deco obligation if you had any). When expressing the risk of a deco method, it is the risk at just the limit of what the method allows.

To be fair you make a good point. Even so.. maybe I got lucky never had a DCS hit.

Still none of this is relevant to the OP"S post. All the OP did was not pay attention to his DC, got close to NDL and ascended and did a 5 minute safety stop due to adpative settings. He was not in and danger and the dive was safe. No DCS. The rest of the discourse on GF over DSAT or RGBM DC's is totally irrelevant.

Who immediately surfaces even if you go to 0 NDL? Not I. Not anyone else I know.
I don't know anyone that does square profile dives even on deco dives even when planned for on tables the dive differs in reality on the DC.

Still.... in my early days Deco dives were done on tables and were planned as square profile dives but no DC back then. Long deco stops on air were pretty boring lol.