Diving with gradient factors for a new recreational diver

[scubadada]

All dives are deco dives, this is the very limit of what most people would consider a recreational profile, and is the max no-stop time on the NOAA 32% table.

Quantify the risk of the different GF factors, are you 100% less likely to get DCS with a GF high of 50 versus GF high of 100?

All dives are either no stop or have mandatory deco stops. A 30 min dive to 100 ft on 32% is a deco dive at any GF high of 99 or below using Buhlmann ZH-L16C with GF (Teric NDL and Deco planner at 1014 mbar)

The SAUL probabilistic dive planner is currently unavailable, so I could not rerun the risks of DCS at the NDLs from my post #235. I have previous calculations for the risk of DCS at 100 ft on 32%:

28 min, 0.32%, 1/312, 1 min less that the NDL for a GF high of 99
26 min, 0.22%, 1/455, the NDL for the GF high of 95
21 min, 0.04%, 1/2500, 1 min less than the NDL for a GF high of 85
17 min, <0.01%, <1/10,000, the NDL for a GF high of 75
9 min would be even less, the NDL for a GF high of 50

Yes, the risk of DCS is lower with less exposure.

Thanks to @Angelo Farina for the mention above

 

[crofrog]

All dives are either no stop or have mandatory deco stops. A 30 min dive to 100 ft on 32% is a deco dive at any GF high of 99 or below using Buhlmann ZH-L16C with GF (Teric NDL and Deco planner at 1014 mbar)

The SAUL probabilistic dive planner is currently unavailable, so I could not rerun the risks of DCS at the NDLs from my post #235. I have previous calculations for the risk of DCS at 100 ft on 32%:

28 min, 0.32%, 1/312
26 min, 0.22%, 1/455
21 min, 0.04%, 1/2500
17 min, <0.01%, <1/10,000
9 min would be even less

Yes, the risk of DCS is lower with less exposure.

Thanks to @Angelo Farina for the mention above

What is the ascent profile for those dives?

 

[scubadada]

What is the ascent profile for those dives?

The Saul decompression algorithm uses an ascent rate of 60 ft/min and includes a 3 min safety stop in the probability of DCS.

Hopefully, the Saul website will be available again so others can check it out

http://moderndecompression.com/

Probabilistic decompression algorithms and the SAUL decompression algorithm have been discussed in The Theoretical Diver

About The Theoretical Diver – The Theoretical Diver

thetheoreticaldiver.org

 

[dmaziuk]

I'm pretty sure you have the logarithm on the wrong side of the equation. See the formula above, or one of the references for a more exact specification. A reduction in the integrated time at a given gradient results in an exponential reduction in risk in these models.

Recall that off-gassing is driven by delta-P and reducing the delta-P (as in lower GF) results in longer off-gassing times. On staged decompression dives that leads to the longer Time To Surface with the attendant hassles. On no-stop dives it should in lower integral supersaturation (which is fine except gradient factors were not intended for use in no-stop dives to begin with, but we'll let that slide for clarity), at the expense of reduced bottom time.

In either case the two questions are:
- is the reduction of risk touted by various analyses "real"? -- as in having the actual measurable physiological impact as opposed to being the artifact of the models themselves (i.e. the numerical approximations of those processes -- that are good as designed: at GF100, but were never really studied at lower GF levels), and
- is the real reduction of risk, whatever that may be, worth the trade-off?

For some dives/models you can look at the model's target and say: yeah, USN tables, or ZH-L16 at 100/100 is probably OK for a single dive for a Navy diver, but a fat old diver may want to dial it down for a "recreational" dive trip.

But once you're on, say DSAT M-values on a multi-day multi-dive trip, there's whole lot of empirical evidence that those are "plenty safe enough" and not much reason to believe dialing it down further will give you anything other than shorter bottom times.

The models seem to converge in the GF85..70 range (google Fraedrich study) on staged decompression dives, and I don't know of any studies other than DSAT report, for no-stop dives. You're welcome to believe you're exponentially safer at GF<90-ish on your vacation trip (<70 if do staged deco), of course, if it makes you sleep better. Good night's sleep is important, it'll likely do more good for your decompression stress than a gradient factor will.

 

[tursiops]

The Saul decompression algorithm uses an ascent rate of 60 ft/min and includes a 3 min safety stop in the probability of DCS.

Hopefully, the Saul website will be available again so others can check it out

http://moderndecompression.com/

Probabilistic decompression algorithms and the SAUL decompression algorithm have been discussed in The Theoretical Diver

About The Theoretical Diver – The Theoretical Diver

thetheoreticaldiver.org

https://web.archive.org/web/20230224024503/http://moderndecompression.com/?page_id=493

 

[goldfishtornado]

Recall that off-gassing is driven by delta-P and reducing the delta-P (as in lower GF) results in longer off-gassing times. On staged decompression dives that leads to the longer Time To Surface with the attendant hassles. On no-stop dives it should in lower integral supersaturation (which is fine except gradient factors were not intended for use in no-stop dives to begin with, but we'll let that slide for clarity), at the expense of reduced bottom time.

I think I understand better what you mean by logarithmically decreasing risk based on this clarification. I still maintain the original phrasing was imprecise. Your risk decreases exponentially with decreased integrated pressure difference gradient. It also takes exponentially longer to reach smaller and smaller pressure differences. But your original statement was: "That would mean first halving of the M-value results in, what, 4% reduction of risk?" The answer to this is no, M stands for Maximum. Look at Figure 1 from the paper you reference (Fraedrich D. Validation of algorithms used in commercial off-the-shelf dive computers. Diving Hyperb Med. 2018 Dec 24;48(4):252-258. doi: 10.28920/dhm48.4.252-258.) Halving the maximum tension you reach in a given compartment is is equivalent to an exponential decrease in bottom time in the figure, or a translation of more than 1/2 of the x distance on the plot. Clearly you cross more than a 2x reduction on the P(DCS) isopleths from even a 1/2 x-axis displacement. A 2x reduction is much greater than 4%.

It is somewhat accurate to say that you only gain a logarithmic benefit in an incremental decrease in bottom time for a fixed total decompression time, but the point is that even in no-stop dives decompression still occurs. This means that you can reduce risk without giving up substantial bottom time. Take for example a dive to 100ft, on EAN32%, with a bottom time of 24 minutes. In the first case assume you ascend and descend at 60 ft/min and do not take a safety stop. The maximum gradient is at the surface at 94%. In the second case assume you ascend and descend at 30 ft/min and take a 5 minute safety stop. Your maximum gradient is again at the surface, but only 72%. You have substantially reduced risk without decreasing bottom time at all.

The models seem to converge in the GF85..70 range (google Fraedrich study) on staged decompression dives, and I don't know of any studies other than DSAT report, for no-stop dives. You're welcome to believe you're exponentially safer at GF<90-ish on your vacation trip (<70 if do staged deco), of course, if it makes you sleep better. Good night's sleep is important, it'll likely do more good for your decompression stress than a gradient factor will.

The paper does not claim surfacing with a GF100 under ZHL-16C is perfectly safe, it claims that is it likely to be below the 3% DCS risk isopleth for a single recreational dive. Six divers diving twice a day for four days with a 3% chance per dive would have a 77% chance of at least one DCS case. Obviously this is much much higher than observed incidence. This is because if you actually follow your average dive computers recreational dive profile suggestions maximum gradients reached will be much less than 100% of M. From personal experience, staying until the NDL on a DSAT computer with no consirvitasim then assending normally witha 3-5 minute stop results in a maximum gradient of about 70% under a ZHL style calculation, simmilar to the example above. People being "plenty safe enough" on DSAT doesn not contridict the fact that there is an expoential decrease in risk with a incrimental decrease in intigrated pressure difference gradient.

 

[L13]

If now you reduce the GF to 50 (making a slower ascent and longer deco stops), it means that your critical compartment now contains half of the maximum allowed content of nitrogen.

That is definitely erring on the safe side. But, as @goldfishtornado stated, the actual reduction is risk (assuming the Buhlman model is even remotely close to useful) is significantly greater than 50% ( probably greater than 75%) and no where near the fantastical reduction of 4% that someone else suggested.

 

[L13]

All dives are deco dives, this is the very limit of what most people would consider a recreational profile, and is the max no-stop time on the NOAA 32% table.

Quantify the risk of the different GF factors, are you 100% less likely to get DCS with a GF high of 50 versus GF high of 100?

All other factors being equal risk at GF=100 >> risk at GF=50.

 

[L13]

The problem with the SAUL results is that you can not find the actual algorithm. While you can find some of the foundational papers establishing the principles on which it is based, it itself is proprietary making it hard to independently verify its accuracy.

Fundamentally it is the following:

• parameterize a 3 tissue compartment parallel /serial model that presents it's result as a probability of DCS.
• Use the large body of well documented research dives to derive the optimum parameters for that model (he tried several models and picked the one that worked best).
• Verify that the risk results closely match the statistics of the the body of dive data it is based on.
• run the new dive through that model and assume that the indicated risk is a good approximation of the actual risk of the dive.

Given the difficulty in funding the quality of research dives necessary for other approaches, I think this is a great idea. However, it would be nice if the actual algorithmic details were public domain.

 

[dmaziuk]

The paper does not claim surfacing with a GF100 under ZHL-16C is perfectly safe, it claims that is it likely to be below the 3% DCS risk isopleth for a single recreational dive. Six divers diving twice a day for four days with a 3% chance per dive would have a 77% chance of at least one DCS case.

Are you sure about that? Good thing nobody is diving 100/100 twice a day for four days in a row, then. (That wasn't the question, the question was is the reduction of risk from no-stop diving, say, near-DSAT 93/93, to, say, 70/80, is meaningful in practical terms. For bonus points: on the proposed twice a day for 4 days schedule. Since we know DSAT users don't get bent on that schedule without outside help, the 70/80 divers can rejoice in the knowledge that they're getting "exponentially more" not bent.)

Anyway, try this one Setting Gradient Factors based on published probability of DCS – The Theoretical Diver and follow the reference #3 to see how deep the rabbit hole goes.