While I was doing my ANDP course I was interested in how the relationship of deco gas O₂ percentage had a non-linear relationship with the time it would take to get to the surface.

Naïve me thought that if 50% O₂ got me to the surface in x minutes, then 100% O₂ would get me there in half the time, but that’s super wrong. You can use a leaner mix at a deeper stop, meaning that it’s useful for longer; that stepped relationship means that there’s some weird steps in the graph:

The graph above assumes that your back gas is air, but if you were to go for a nitrox back gas, what’s the sweet spot there? The problem is that it’s a 2D problem now, and that makes it hard to show on one complicated plot, so here’s an even more complicated table of plots:

Back gas ppO₂ 1.3, Deco gas ppO₂ 1.6 - Decompression Profiles

Depth	18 min	21 min	24 min
35m
40m
45m
50m

And here’s that again, but at a ppO₂ 1.6 of 1.4 for those who like to live dangerously:

Back gas ppO₂ 1.4, Deco gas ppO₂ 1.6 - Decompression Profiles

Depth	18 min	21 min	24 min
35m
40m
45m
50m

These surprised me, I didn’t expect the low % nitrox to have such an effect on deco.

I was interested to see that the benefit of the highly optimised mixes isn’t a wild improvement on just having a 50% and being done with it. With a worst case scenario of 24 minutes at 45m, nitrox 28 back gas, and nitrox 83 for deco gets you out of the water less than 10 minutes faster than air and 50%.

I’d imagine that this process would be a lot more useful beyond the realm of what can be plotted on a 2d graph, when you’re dealing with helium, multiple deco tanks, and much longer bottom times/depths. Then it might be able to knock a fair bit off, and 20 minutes less deco might be the difference between freezing and having a good day. I guess I’ll find out later when/if I do my trimix course.

---

All decompression calculations use the decotengu library, a pure Python implementation of the Bühlmann ZHL-16B-GF algorithm.