The amount of carbon dioxide (CO2) that can be safely inhaled by rebreather divers is a continuing point of conjecture, and vigorous argument. Unfortunately, the U.S. Navy Experimental Diving Unit has confused that issue, until recently.
A non-diver might wonder why a diver should inhale any CO2. After all, the air we breathe contains only a small fraction of CO2 (0.039%). A rebreather is best known for emitting no bubbles, or at most very few bubbles depending on the type of rebreather. It does that by recirculating the diver’s breath, adding oxygen to make up for oxygen consumed by the diver, and absorbing the carbon dioxide produced by the diver. 
The CO2 scrubber canister is vital to keeping the diver alive. As pointed out in the first post in this series, carbon dioxide is toxic; it can kill.
A CO2 scrubber keeps the recirculating CO2 levels low by chemically absorbing exhaled CO2. However, the scrubber has a finite lifetime – it can only absorb so much CO2. Once its capacity has been exceeded, CO2 passing through the canister accumulates exponentially as the diver continues to produce CO2 from his respiration.
The question rebreather divers want answered is, “How much of that bypassed CO2 can I tolerate?” As we’ve discussed in previous posts, 30% CO2 can incapacitate you within a few breaths. I can personally verify that if you’re exercising you may not notice the effect 7% CO2 has on you, until you try to do something requiring coordination. I’d equate it to the effect of drinking too many beers. There is little controversy about CO2 levels of 5-7% being bad for a diver.
For levels below 5-7% CO2, the U.S. Navy has not been real clear. For instance, 2% CO2 is the maximum CO2 allowed in diving helmets. If CO2 were to climb higher the diver would most likely feel a need to ventilate the helmet by briefly turning up the fresh gas supply to clear CO2.
Since at least 1981, NEDU has defined the scrubber canister breakthrough point in rebreathers as 0.5% CO2. That means that at some point, which varies with CO2 injection rate, ventilation rate, water temperature, and grain size of CO2 absorbent, CO2 begins leaking past the canister, not being fully absorbed during its passage through the canister. Once that leakage starts, the amount of CO2 entering the diver’s inspired breath rises at an ever increasing rate unless work rate or other variables change. By the time the CO2 leaving the canister has reached 0.5%, the canister has unequivocally “broken through”.
I pointed out in my last post that even 0% inspired CO2 may be too much for some divers when they are facing resistance to breathing. And all rebreathers are more difficult to breathe than other types of underwater breathing apparatus because the diver has to force his breath through the rig’s scrubber canister and associated hoses. The deeper the dive the denser the breathing gas and the worse breathing resistance becomes.
In free-flow diving helmets like the old MK 5, and the short-lived MK 12, the diver did not breathe through hoses and scrubber canisters. But those helmets had a high dead space and to keep helmet CO2 at tolerable levels a fresh gas flow of 6 actual cubic feet per minute (acfm; 170 liters per minute) was required. The U.S. Navy allowed up to 2% CO2 in the helmet because 1) the helmets did not have a high work of breathing and 2) due to simple physics the helmet CO2 couldn’t be kept very low.
For rebreathers, none of the above apply. A high breathing resistance is inevitable, at least compared to free-flow helmets, and once CO2 starts rising there is nothing you can do to decrease it again, short of stopping work.
In 2000, NEDU’s M. Knafelc published a literature review espousing that the same limit for inspired CO2 which applies in helmets could be used in rebreathers. Nevertheless, in 2010 NEDU’s D. Warkander and B. Shykoff clearly demonstrated that in the face of rising inspired CO2 concentrations work performance is reduced, and blood levels of CO2 rise, in some cases to dangerous levels. More recent work by the Warkander and Shykoff duo have extended those studies into submersion, however those reports are not yet publicly available.
As a result of both physiological theory and confirmatory data in young, physically-fit experimental divers, NEDU has not relaxed the existing definitions of scrubber canister breakthrough, 0.5% PCO2. Furthermore NEDU will adhere to the current practice of using statistical prediction methods to define published canister durations, methods which are designed to keep the odds of a diver’s rebreather canister “breaking through” to no more than 2.5%, comparable to the odds of decompression sickness following Navy multi-level dive tables. Details of this procedure will be explained in later postings.