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

[EFX]

That's not really correct. As mentioned by @iointerrupt some numbers are calculated using an estimated ascent rate (10m/min for the Perdix). TTS is one of them for example. My understanding is that NDL is as well. My question was about CEIL, and I believe that unlike NDL, it is not calculated using an ascent rate. It’s an instantaneous ceiling.

I forgot about TTS (time to surface). Of course, an assumed ascent rate is needed to calculate the time it takes to get to the surface without exceeding GFHi. I was thinking only of the calculations for ceiling and NDL NDL doesn't need to predict the ascent rate because it is concerned solely with how much time you can spend at the current depth.

Erik Baker, P.E. does not mention assumed ascent rates in his paper "Calculating the no-stop time" attached here for your convenience. He solves the Haldane (otherwise known as the "instantaneous") equation for t (time) to get the NDL time remaining. This method can only be used under specific conditions where:

Mo > Pi > Po
Mo < Pi < Po

;Mo is the m-value, Pi is the current inert gas tissue pressure, and Po is the last inert gas tissue pressure.

Another method which I use in my spreadsheet calculates the ceiling starting with t = 1 minute in the Haldane equation and loops increasing t by 1 minute until the ceiling becomes zero. The value of t at this point becomes the NDL. For what it's worth the NDL values in my spreadsheet compare favorably with the Perdix's dive planner.

 

[Rogerdd]

So we basically agree on the rest if we take the same definition of NDL, we just need to confirm how shearwater implemented their calc of NDL; whether they take into account ascent time (like a lot of tables out there as mentioned by @crofrog) or assume an instantaenous ascent as you said.

If it’s the latter, then the 3 SB posters in the screenshots I posted might be misled if they’re using a Shearwater dive computer.

I sent an email to Shearwater, let’s see what they answer.
Maybe they have a presence here on the forum too? @Shearwater

 

[VikingDives]

It looks like from the documentation (I spent time in that rabbit hole today) that ascent is assumed to the next stop at a rate of 10M/33F per minute, and it assumes you leave at that moment.

 

[stuartv]

The dive computer makes calculations based on the current conditions of your dive. No calculations are done based on assumed future data (i.e. ascent rates).

False.

That's correct. At the current time, there are no dive computer options you can download that mimic crystal balls.

False.

I forgot about TTS (time to surface). Of course, an assumed ascent rate is needed to calculate the time it takes to get to the surface without exceeding GFHi. I was thinking only of the calculations for ceiling and NDL NDL doesn't need to predict the ascent rate because it is concerned solely with how much time you can spend at the current depth.

Erik Baker, P.E. does not mention assumed ascent rates in his paper "Calculating the no-stop time" attached here for your convenience. He solves the Haldane (otherwise known as the "instantaneous") equation for t (time) to get the NDL time remaining. This method can only be used under specific conditions where:

Mo > Pi > Po
Mo < Pi < Po

;Mo is the m-value, Pi is the current inert gas tissue pressure, and Po is the last inert gas tissue pressure.

Irrelevant. I mean, I'm willing to stipulate whatever you say that Baker's paper says. I haven't read his paper. I'm looking at his Fortran code that implements Gradient Factors. The code takes an input of "ascent rate". It allows the user to input what ascent rate to assume.

Baker's code uses the Haldane equation (in a Fortran function oddly enough called "HALDANE_EQUATION") to calculate gas loading for dive legs that are at a constant depth.

For parts of the dive that are ascending or descending, he uses a different Fortran function called "SCHREINER_EQUATION". In his implementation of the Schreiner equation, he uses whatever ascent rate is passed into the program as part of the dive data to calculate gas loading during that leg.

So, while he does not have a hard-coded ascent rate in his code, his code does use a predetermined ascent rate. I.e. whatever rate the user tells the code to use. Here is part of the comments in his code for calculating the start of the "deco zone". Note the reference to the time it would take to ascend, which can only be determined by assuming an ascent rate - in this case, the ascent rate the user supplies as input.

the time when the gas tension in the compartment will be equal to ambient pressure. The low bound for time is set at zero and the high bound is set at the time it would take to ascend to zero ambient pressure (absolute).

His code does not calculate an NDL at all.

Shearwater computers assume an ascent rate in order to give you the longest NDL they can - which is what they do, rather than having the user enter a planned ascent rate. By making an assumption about your ascent rate and forecasting how much you will off-gas during that ascent, it allows the computer to give you a slightly longer NDL than it could otherwise.

The difference between Baker's code and the Shearwater implementation is that Baker's code would yield an NDL based on when CEIL is 0. Shearwater takes things a step further and calculates how long you can stay at depth until CEIL would be at 0 after an ascent at 10mpm.

Baker's code does not dictate that either way is "correct".

Another method which I use in my spreadsheet calculates the ceiling starting with t = 1 minute in the Haldane equation and loops increasing t by 1 minute until the ceiling becomes zero. The value of t at this point becomes the NDL. For what it's worth the NDL values in my spreadsheet compare favorably with the Perdix's dive planner.

You are welcome to calculate NDL that way. The way you're doing it, you would only exceed your GFHi by instantaneously transporting to the surface. That is impossible, so you will always arrive at the surface with a GF99 that is less than your GFHi. That's fine. That is more conservative... but less accurate*.

Shearwater computers do it better. If I can stay an extra couple of minutes and still ascend (at the rate I know my computer to be assuming) to the surface without exceeding my GFHi, then that is better, in my personal opinion.

But, it really doesn't matter, as long as you know how your computer works and dive accordingly.

* "accuracy" being defined by how close you will be to your GFHi when you actually reach the surface.

 

[inquis]

Shearwater computers assume an ascent rate in order to give you the longest NDL they can

I agree they include an ascent for some calculations, but it's not stated in the manual that it's applied to the NDL calculation (as it does for TTS, for example). Have you seen something that states that?

That's also not what I see in downloaded logs: the time at which the computer-reported NDL goes to 0 is the same moment the ceiling becomes non-zero. This would not be the case if NDL were extended by considering off-gasing during ascent.

 

[inquis]

ascent is assumed to the next stop at a rate of 10M/33F per minute, and it assumes you leave at that moment

That is for the calculation of TTS. (It also assumes you will switch to more appropriate gases if any are turned on.) But none of that is relevant to NDL. AFAICT, remaining NDL is the time at the current depth on the current gas until the ceiling becomes non-zero. That's also the more conservative value and the easier implementation.

 

[stuartv]

I agree they include an ascent for some calculations, but it's not stated in the manual that it's applied to the NDL calculation (as it does for TTS, for example). Have you seen something that states that?

That's also not what I see in downloaded logs: the time at which the computer-reported NDL goes to 0 is the same moment the ceiling becomes non-zero. This would not be the case if NDL were extended by considering off-gasing during ascent.

Yes, but I can't give you a citation right now.

Maybe they calculate CEIL factoring in ascent time, too...?

 

[inquis]

Maybe they calculate CEIL factoring in ascent time, too...?

I don't believe so -- it matches the calculation in Subsurface which is clearly the instantaneous ceiling.

 

[VikingDives]

That is for the calculation of TTS. (It also assumes you will switch to more appropriate gases if any are turned on.) But none of that is relevant to NDL. AFAICT, remaining NDL is the time at the current depth on the current gas until the ceiling becomes non-zero. That's also the more conservative value and the easier implementation.

That's what I thought, but after @iointerrupt's post, I went and looked at the manual. It does say that deco calculations (and therefore presumably the NDL) are calculated with an assumed ascent rate of 10 meters (33 feet) per minute.

That's at the bottom of page 13. There's discussion elsewhere that indicates that all ascents.

 

[inquis]

It does say that deco calculations (and therefore presumably the NDL)

Interesting qualifier... I think the most compelling evidence is this:

1. Shearwater has stated that surfGF assumes a teleport ascent.
2. In my logs at the time sample when reported NDL clicks from 1 minute to 0 minutes, surfGF equals the configured GFhigh.

Since teleporting takes no time, I conclude that Shearwater's NDL calculation assumes zero off-gassing.