PADI RDP... shallow water data extrapolation

[Javik]

So I was wondering just what the actual bottom times are in water more shallow than the 35 ft PADI RDP table minimum.

It occurs to me that the table data can simply be entered into a spreadsheet, and then we can extrapolate the resulting curves to see what the pressure groups actually are.

So then, Microsoft Excel 2013...

Hmm, looks like these are power curves. Some of the extrapolated power curves weirdly overlap, but that's probably because the compact plastic chart is rounding off values to whole minutes. Also, I didn't plot groups Y and Z because the extrapolated curves look weird due to so few provided data points.

Also, it's not clear what exactly a required safety stop is if you're already at 10 ft underwater, though I realize this is a "cheater table" meant to hide decompression complexity from beginning divers and so likely has a huge fudge factor.

So, let's just look at a maximum of group V which doesn't require a safety stop, and clip off the rest beyond it...

Looks like the max open circuit recreational no-decompression bottom time for group V is approximately....
- 5 ft ........ 4000 minutes ..... or 66 hours, 40 minutes
- 7.5 ft ..... 2000 minutes ..... or 33 hours, 20 minutes
- 10 ft .......1200 minutes ..... or 20 hours
- 12.5 ft ..... 800 minutes ..... or 13 hours, 20 minutes
- 15 ft ........ 580 minutes ..... or 9 hours, 40 minutes
- 17.5 ft ..... 450 minutes ..... or 7 hours, 30 minutes
- 20 ft ........ 350 minutes ..... or 5 hours, 50 minutes
- 22.5 ft ..... 290 minutes
- 25 ft ........ 250 minutes
- 27.5 ft ..... 210 minutes
- 30 ft ........ 180 minutes

So to try sleeping underwater on a hookah, I'd probably recommend a depth of 17 ft or less, to have enough time to get a good night's rest.

And finally, there's probably some huge error I'm overlooking that isn't obvious, because the RDP table probably tries to simplify things for the depth range it does cover, and likely it doesn't tell the whole story for out of range values.

 

[Storker]

I'd be extremely careful to actually use any of those data. You're essentially fitting a curve to moderately non-linear data, and then reading values in a range where the extrapolated curve deviates very significantly from linearity. That type of approach is generally a very good recipe for screw-ups.

If this is just an academic exercise for the fun of it, that's quite another story.

 

[Javik]

Well, feel free to jump down my throat and tell me how this is way off, and why. I would interested to know if there are more accurate numbers.

Presumably the dive computers are capable of calculating this data more accurately since they constantly monitor depth / altitude changes.

 

[Neilwood]

I think what @Storker is saying is that you are generating data based on an assumption which appears to fit but without the actual calculations to back up that assumption, you could be a long way out from the true data. Dive tables are made with a lot of actual data (generally obtained in deco chambers with a lot of science behind them) and with certain conclusions made from that data.

This sort of extrapolation is fine in theory but to rely on that theory without the fact to back it up it could be dangerous. You don't want to find that your theory is significantly out by getting bent and having a chamber ride.

Remember that decompression theory is just that - even scientists involved in the production of data accept that:

Decompression is an area where you discover that, the more you learn, the more you know that you really don't know what is going on. For behind the "black-and-white" exactness of table entries, the second-by-second countdowns of dive computers, and beneath the mathematical purity of decompression models, lurks a dark and mysterious physiological jungle that has barely been explored.

— Karl E. Huggins, 1992

 

[Storker]

If stating that I'd be extremely careful using those extrapolations is "jumping down your throat", we have quite different opinions about what jumping down somebody's throat entails.

The point is that when you're extrapolating like that, you're essentially in an area of the map where you don't have any real data to back up your assumption that your fitted curve is valid in that region. That's an inherently risky practice, and I wanted to point out just that. Since I'm generally rather risk adverse, I wouldn't trust those extrapolations. Particularly when the fitted curve deviates so strongly from linearity.

The Danger of Extrapolation in Regression Analysis
The Second Version: Dangers of Extrapolation

And Wikipedia says, quite correctly:

extrapolation [...] is similar to interpolation, which produces estimates between known observations, but extrapolation is subject to greater uncertainty and a higher risk of producing meaningless results.

Extrapolation - Wikipedia, the free encyclopedia

 

[guruboy]

I think the Navy doesn't have a MDT for 20ft or less.

 

[Hawkwood]

Well, feel free to jump down my throat and tell me how this is way off, and why...

Goodness, that is a pretty extreme reaction to a reasoned comment.

 

[AlexL]

since this is in basic, sandbox rules apply, so there should not be any 'jumping of throats'.

then reading values in a range where the extrapolated curve deviates very significantly from linearity.

isn't the case (based on ZHL 100/100) that for anything shallower than 30', for any length of exposure, its considered 'reasonably safe' to direct ascend, even when fully saturated?

 

[seeker242]

I would interested to know if there are more accurate numbers.

US Navy has tables down to 10'

http://umaine.edu/scientificdiving/files/2012/09/Navy_air_tables.pdf

 

[stuartv]

@OP: The point you seem to be missing is that there are depths where even if you stay down long enough to become fully saturated you can still ascend at 30 fpm directly to the surface without exceeding the M value in any compartments - to put it in terms of a Haldanean model. So, at that depth and shallower, you could stay forever and still go directly to the surface.

The Navy tables suggest that that depth is 20 feet, as they don't give an NDL until you get to 25'.

You're clearly mathematically inclined and interested in the subject, so I highly recommend Mark Powell's book, Deco For Divers. I suspect you would really enjoy it and also find it very illuminating. I certainly did.