I have tried to find the basis for the published limits on CNS, and I have not found it. (I have to admit that I have not tried all that hard.) I strongly suspect that is was something of a guess, somewhere in the range between "somewhat educated" and "wild."
Now think what that does to me as a technical instructor. The course materials require students to make sure that students do all the CNS and OTU calculations for their dives, and they learn, of course, that it takes some pretty big dives, dives that we won't be doing in training, to exceed those limits. But they know that people do indeed do dives that exceed those limits, sometimes by a wide margin.
So when they ask me how that happens, what do I say? I answer honestly but very carefully. I certainly don't tell them that they should go ahead and ignore those limits with my blessing. I have developed something of a spiel that walks the tightrope.
What I am really saying in this post is that it is time for someone in the scuba agency to do some kind of investigation and report on the origin and accuracy of those numbers. I am not expecting it, though, because I am sure they are walking the same tightrope I am. If they were to write anything that sounded like a blessing to ignore those limits and someone were to do that and die, ....
The current NOAA Diving Manuals (Sixth Edition-2017; Fifth-2013; Fourth-2001; Third-1991) all have the same table (15.1) in them for oxygen exposure; namely the infamous “Normal Exposure Oxygen Partial Pressure Limits” that allow 45 minutes at PPO2=1.6, 180 minutes at 1.3, etc. The table is "adapted from Butler and Thalmann, 1986." The text discussing that table says:
“single exposures somewhat longer than those shown in Table 15-1 can be conducted without episodes of central nervous system (CNS) oxygen toxicity. However, the more conservative exposure times shown in Table 15-1 take operational safety into consideration and are sufficient in duration for anticipated NOAA dives. At the same time, the limits shown in Table 15-1 extend the limits published in the second edition of the NOAA Diving Manual.”
I’ve not been able to find a Second Edition (1979) of the NOAA Diving Manual, but presumably it had even lower limits than the admittedly conservative limits in Table 15.1.
The Butler and Thalmann (1986) work in Undersea Biomed Res, 1986 Jun;13(2):193-223, has this abstract:
"Central nervous system oxygen toxicity is currently the limiting factor in underwater swimming/diving operations using closed-circuit oxygen equipment. A dive series was conducted at the Navy Experimental Diving Unit in Panama City, FL, to determine whether these limits can be safely extended and also to evaluate the feasibility of making excursions to increased depth after a previous transit at a shallower depth for various lengths of time. A total of 465 man-dives were conducted on 14 different experimental profiles. In all, 33 episodes of oxygen toxicity were encountered, including 2 convulsions. Symptoms were classified as probable, definite, or convulsion. Findings were as follows: symptom classification is a useful tool in evaluating symptoms of oxygen toxicity; safe exposure limits should generally be adjusted only as a result of definite symptoms or convulsions; the following single-depth dive limits are proposed: 20 fsw (6.1 msw)--240 min, 25 fsw (7.6 msw)--240 min, 30 fsw (9.1 msw)--80 min, 35 fsw (10.7 msw)--25 min, 40 fsw (12.2 msw)--15 min, 50 fsw (15.2 msw)--10 min; a pre-exposure of up to 4 h at 20 fsw causes only a slight increase in the probability of an oxygen toxicity symptom on subsequent downward excursions; a pre-exposure depth of 25 fsw will have a more adverse effect on subsequent excursions than will 20 fsw; a return to 20 fsw for periods of 95-110 min seems to provide an adequate recovery period from an earlier excursion and enables a second excursion to be taken without additional hazard; nausea was the most commonly noted symptom of oxygen toxicity, followed by muscle twitching and dizziness; dives on which oxygen toxicity episodes were noted had a more rapid rate of core temperature cooling than dives without toxicity episodes; several divers who had passed the U.S. Navy Oxygen Tolerance Test were observed to be reproducibly more susceptible to oxygen toxicity than the other experimental divers."
Note that they propose a PPO2=1.6 oxygen limit as 240 mins, not 45 mins!
It would be easy to conclude that the NOAA tables are VERY conservative, and (as they say, quoted above) appropriate for anticipated NOAA dives, but perhaps a bit too conservative for those exploring the edge of their comfort zone.