Closed-Circuit Rebreathers (CCRs) are complex. Fewer moving parts than a Formula One car, and less mind-boggling than a Heath Robinson machine, but as mysterious and confusing as both to some folks.
Here’s one thing that certainly doesn’t help. When open-circuit scuba goes pear-shaped, the situation usually announces itself with gusto. Events such as a high-pressure seat failure, an o-ring giving up the ghost, a hose failing, or a manifold or burst-disk leaking, make themselves known immediately. Divers spend a huge percentage of the time during any technical training program, rehearsing a variety of valve shutdowns, regulator switches, and one or more options intended to deal with this type of failure, preserve what gas they can, and get their backsides out and to the surface with the least fuss possible.
By contrast, a CCR is not only quieter than open-circuit in normal operation, a whole category of failures arrive unannounced and quietly too. Certainly CCRs are still prone to many of the issues that plague their bubble-making dive buddies. Ruptured hoses, extruded orings, faulty handwheels, and free-flowing first stages are all possible. But in addition, there’s a whole category of sly, furtive malfunctions unique to closed-circuit diving; and each of these has the potential to cause real harm.
The default and simplest solution is to “bailout to open-circuit.” In other words, stop using the rebreather and switch to breathing from open-circuit gear to get back to the surface as rapidly as circumstances allow.
Advanced training for CCR divers puts strong emphasis on keeping the diver in CCR mode for as long as safety allows, and only bailing out as the primary option for scenarios like catastrophic loop failures or full floods, widely divergent oxygen cell readings, carbon-dioxide breakthrough, mechanical damage to primary components, etc. Cave CCR students, for example, are expected to consider all the options available to them in the event of a system failure – real or simulated. A full-cave CCR course is an exercise in complex navigation, and disaster scenario management. However, for the sake of overall safety, CCR cavers are also encouraged to bailout to open circuit if they have doubt about what needs fixing and how best to do so.
A useful phrase worth remembering is: THERE’S NO SHAME IN BAILING OUT!
Of course, as with most pieces of advice about diving, particularly cave diving, and more specifically about diving a CCR in a cave, there is a limitation. There’s no shame in bailing out… provided you have more gas then you need to get back to dryland in one piece.
And this begs the question: How much bailout gas is enough?
Calculating the answer to this is simply a question of using average depth (expressed in bar or ata), and multiplying that number by how much time it will take to get back to open water. In addition, one is advised to factor in some contingency volume for heightened gas consumption due to stress, hypercapnia, and so on. One suggestion is to work with a basic SAC rate of 30 litres / one cubic foot per minute. So using this baseline for a cave with an average depth of 20 metres / about 65 feet / 3 bar or ata, the bailout consumption rate would be 3 X 30 litres or 3 X 1 cubic feet per minute.
This calculation suggests an 80 cubic-foot cylinder (11 L charged to 200-210 bar) would last approximately 25 minutes. Penetrations therefore would be no deeper than a 25 minute swim to the exit… where one might normally stage a small cylinder of decompression gas: usually pure oxygen.
Some divers use a slightly more conservative baseline, some slightly more aggressive. Some calculate a slightly lower consumption rate after the first 10 minutes on bailout, on the understanding that a diver will begin to regain control of his or her breathing after that time.
Another approach is the “one-hour rule.” Following this guideline, divers each plan to surface with one hour of all consumables in reserve, which includes lights, oxygen and diluent gases, scrubber, and bailout.
Whichever guideline one opts to use, the strong recommendation is to backup any seat-of-the-pants calculations by conducting simulated bailouts from various points in caves one dives regularly. These actual real-world data – with an added factor for stress – can then be inserted in calculations to arrive at a more accurate estimate.
Once one has an idea of how much bailout gas is enough, the next decision is how to carry it. Options include, about one’s person, shared among team members, drop-staged at various points in the cave.
The NSS-CDS, one of the original cave diving training agencies, suggests a dive team carries 1.5 times the volume of gas required to get a single diver out of the cave. Therefore, in the example above and a three-person team, each member would carry a fully charged 40-cubic foot bottle.
The logic behind “team bailout” is that there is, for the diver with a gas emergency, a greater level of conservatism than the acceptable norm for open-circuit cave divers. It does however demand that team stays in contact, swap tanks during their exit, and that only one unit has a problem that requires bailout.
Except in exceptional circumstances – with seasoned team members and when the basic bailout scenarios are inappropriate or impractical – I choose to carry on my person, enough gas to swim out of the cave on my own. Depending on the unit I am diving, I find that carrying two, 80-cubic foot sidemount cylinders is easy, comfortable, streamlined, and allows for plenty of time to exit from the vast majority of tourist cave dives. On occasion, for “smaller” dives or shallower profiles, I’ll strap on smaller aluminum tanks for bailout. If a dive requires a bailout volume approaching my normal carried volume, or a greater safety margin, I’ll drop stage bailout gas and/or work out a kind of hybrid personal-carry-team-dropped-stage strategy.
More than any other factor, one should be aware of the elevated gas consumption that typically follows an incident that demanded coming “off the loop” (bailing out). One also has to consider, especially if open-circuit diving is no longer part of your regular dive menu, typical consumption rates for a CCR diver using OC gear are often higher than expected. Something to do with the sudden shock of breathing cold, dry air I suspect.
In any event, remember to always have something appropriate to breathe, and plenty of it. You will never regret carrying more gas than you need.
Doppler's Tech Diving Blog 
Here, here. I couldn’t agree more. The 1.5 times rule not only requires everyone stay together but also does not account for dealing with partner who is experiencing a carbon dioxide breakthrough. Having brought someone out of the cave in that situation, I can attest to the fact that the individual used a TON of gas. Long hoses were also way more practical than swapping tanks. When someone is almost drowning between breaths, the last thing they can mange is calmly swapping a tank with another diver. They will suck down a tank until they reach a vacuum.