Perdix CEIL, does it take into account the estimated off-gassing on ascent (with an assumed ascent rate)?

[dmaziuk]

That was at 30/75 -- it works to illustrate the problem, but IRL I wouldn't be using 30/75 (if I had a ZH-L16 computer) for no-stop dives.

For a more realistic no-stop setting in the 90/90 ballpark, you'd probably have to add another 10 metres or more to the depth, to get the same effect. That's gonna push you past 40 msw, among other things.

There's probably plenty of "AOW" wrecks where these "transient ceilings" could be an issue, but with those I'd say: join BSAC and get properly trained for those type of dives so you don't need to care much about them. And if you plan your dive for staged deco, your ZH-L computer and its GrapeFruits will work as intended, too.

 

[Rogerdd]

My take is the point of this thread is whether CEIL assumes an ascent. IMO, it doesn't, so it's on the diver to manage this grey area you're describing. (I agree with everything you've said, BTW.)

The initial question was whether CEIL assumed an ascent or not.
My guess was that it does NOT, and the purpose of this thread was to get a confirmation.

After some discussion, the question became wider and now it’s not just about CEIL but also NDL: does it assume an ascent rate or is it an ‘instantaneous’ ascent?

If CEIL is instantaneous but NDL is not, then we can have a ceiling without stops.

I believe it’s important to first agree on whether both CEIL and NDL are assuming an ascent or not, because that will impact whether the statement above is true or not.

 

[Rogerdd]

I’d like to refresh this thread as I still think the question is open/not resolved.
Shearwater CS hasn’t really come back to me with the technical details I was expecting.

To sum it up, with Shearwater:
- The mandatory deco stops will appear once NDL becomes less than 0.
- SurfGF is the gradient factor assuming a teleportation to the surface (no time to off-gas)

But how are both NDL and CEIL calculated? The doc mentions an ascent rate of 10m/min, but it doesn’t clearly say which parameters are using an ascent rate.

NDL:

NDL < 0, is it when:
1- SurfGF is > GFHi
Or
2- [estimated GF99 at the surface if going to the surface now at an estimated ascent rate of 10m/min] > GFHi

Basically, does NDL take into account the off-gassing that will take place on ascent (projected to be 10m/min)?

CEIL:

Similarly is CEIL the depth at which the GF99 is equal to GFHi assuming:
A- an instantaneous ascent (teleportation) to that depth
Or
B- an ascent of 10m/min


-> If it’s 2 and A, then there could be some cases where we’re not in deco yet (NDL>0, no stops needed) but with a ceiling below the surface; though that would be a ceiling that we would not reach if maintaining an ascent of 10m/min (without any stops).

-> if it’s 1 and A, or 2 and B, then when NDL goes below 0 is also exactly when CEIL depth goes below the surface

-> if it’s 1 and B, I don’t think such implemention would make sense

 

[iointerrupt]

Basically, does NDL take into account the off-gassing that will take place on ascent (projected to be 10m/min)?

From observing it during dives, I believe it does. If I have a GF Hi of say, 80, NDL will not hit 0 minutes until SurfGF is showing 82 or 83.

Could be other explanations for this, but I always assumed it was because it was taking ascent time into account.

 

[dmaziuk]

Ceiling should be calculated by Buhlmann's formula. Off-gassing on ascent only really matters for deeper no-stop dives: your ascent time has to be comparable to the leading compartment's half time for it to make appreciable difference.

Shearwater hardware should have enough CPU cycles to calculate off-gassing during ascent and factor it in the NDL display. If I were marketing a DC to no-stop divers who are not expected to understand how the computer works, I would want to implement that to avoid the "YOU'RE NDLS TO SHORT" screaming and wailing from the likes of ScubaLab "computer testers" and such.

 

[Rogerdd]

From observing it during dives, I believe it does. If I have a GF Hi of say, 80, NDL will not hit 0 minutes until SurfGF is showing 82 or 83.

Could be other explanations for this, but I always assumed it was because it was taking ascent time into account.

That’s my understanding as well for NDL.
I think @stuartv agreed too earlier in this thread but others like @inquis didn’t seem to..

Ceiling should be calculated by Buhlmann's formula.

Shearwater hardware should have enough CPU cycles to calculate off-gassing during ascent and factor it in the NDL display.

An interesting article about the possible differences in implementation

Why is Bühlmann not like Bühlmann – The Theoretical Diver

thetheoreticaldiver.org

And if shearwater cpu can factor in the off-gassing during ascent in the NDL display, why can’t they do it to adjust the ceilings?

In the graph here produced by them, the lines between stops are slightly sloped, so for the next ceiling (and corresponding stop after rounding) after the current stop, should it not be calculated including some off-gassing that will occur while reaching it? If it doesn’t then each stop indicated on the PDC would be a bit deeper than actually needed?

 

[dmaziuk]

And if shearwater cpu can factor in the off-gassing during ascent in the NDL display, why can’t they do it to adjust the ceilings?

The ceiling is the ceiling, given by C = (P - a) * b

On no-stop dives it can be "transient" i.e. it will disappear by the time you get there. That's the nits we're picking here.

You have to remember that RGBM is the only model that tried to cater to no-stop divers. The rest of them, ZH-L included, are designed for planned decompression and this whole NDL business is duct-taped on top. There is a discontinuity at the point where NDL reaches 0, that's how the math works. Whatever implementation band-aids you add to keep customers happy, does not change that.

 

[inquis]

if shearwater cpu can factor in the off-gassing during ascent in the NDL display, why can’t they do it to adjust the ceilings?

It doesn't make a lot of sense to report a ceiling that depends on ascent rate. Think about a faster ascent than assumed and remember the ceiling is updated throughout. You could make it to a depth shallower than the offgas-included stop + 10 ft, but then the higher than expected loading mandates the deeper stop (i.e., 10 ft deeper than originally stated). Woops, you're already shallower than that.

The conservative approach is an instantaneous ceiling. My thought is Shearwater would have thought of this.

 

[inquis]

My other thought is what does it matter? Is this to avoid violating the boat's "no deco" mandate? You have some amount of time after NDL switches from 1 to 0 min before "official deco" sets in. Of that I'm certain, as I just saw this happen on my last dive. Unfortunately, I don't recall whether the ceiling was nonzero at that time.

(For the curious, it was a planned deco dive but buddy was cold enough to call it done when I checked with him at NDL=0, a convenient prompt.)

 

[inquis]

If it’s 2 and A, then there could be some cases where we’re not in deco yet (NDL>0, no stops needed) but with a ceiling below the surface

It's also a simple matter to report a ceiling of 0 as long as NDL >= 0. In fact, this is almost certainly the case since the initial ceiling & stop is controlled by GFLow. My point is that your 4 combinations are not the only options.

To clarify my earlier comment regarding NDL, that was in conjunction with looking at the log in Subsurface. That could be slightly different than what was reported in the moment.

I'm actually inclined to defer to @iointerrupt's (in-dive) observation: off-gassing is considered for NDL, which is obviously not a problem for a faster than assumed ascent. [Edit: actually it is a problem, as then you'd surface with GF99 > GFHigh.]