There are big rebreathers and little rebreathers. So why would you choose one over the other? The four vital elements that are affected when size changes in a rebreather are:
1. The buoyancy
2. The canister endurance
3. The work of breathing
4. Reliability of sensors
Example 1: A manufacturer makes a rebreather with a large canister in order to increase absorbent capacity and endurance (if it is designed it correctly). It now has a large volume (since the counterlungs themselves must be a minimum of 4.5l). I jump in the water and just can't sink. So while my unit might be fairly light on the surface, I need lots of lead to dive it. So the trade here is to accept that the extra endurance is what we want and choose (heavier) engineering materials to make the unit neutral in the water. The ultimate trade off here is traveling weight.
So if you want long endurance, be prepared to:
1. Wear lead
2. Pay for traveling weight
3. Pay for more absorbent
There is, however another twist to this part of story. If the rebreather is designed correctly, it can use smaller absorbent fills and still have long endurance. The trade off again is the size/volume that is needed to provide insulation, but at least you lose one of the downsides. The feature of a well insulated rebreather is that it doesn't suffer from massive endurance reduction with depth as most rebreathers do. The Sentinel (which has one of the smallest absorbent fills at only 2.2kg) does not suffer adversely with endurance reduction with depth. In addition the extra insulation ensures a very dry environment for the sensors, significantly improving accuracy and sensor life over smaller designs. The extra insulation is compensated for by a precise selection of engineering materials that makes the unit near-neutral in the water without additional weighting. Canister design is really a 'black art'. There are several basic concepts that fundamentally work (that is why there are so may 2.5kg axial canisters out there) but these are not really optimized. One or two more radical designs exist but as yet their operational rules do not seem to be based in much science and testing. Multiple canisters can either reduce the work of breathing (WOB) and at least double the duration (parallel systems) or double the WOB (at the canister) and increase the duration to an unpredictable limit (serial/inline). In a parallel system both canisters experience the same gas flow and both run for at least the full endurance of the single unit. The larger surface area of the two canisters may decrease the flow and increase the dwell time of the gas in the absorbent and have a beneficial effect. In the serial system, double the absorbent just increases the WOB and the first canister is preferentially exhausted. But if it is a small canister and/or you work hard and bypass it, then the second canister comes into play and hopefully stops any CO2, but now you don't know where you are from an endurance standpoint, since it is likely your endurance math hasn't taken into account partially using the second canister. If you swap the canisters around part way through a dive sequence, you are dealing with an even larger unknown quantity. So while smaller serial systems may seem specially attractive they can lead to WOB, unknown endurance and sensor care issues. Imagine you have two fuel tanks in your car. the first one gets used for crawling around town but when you need to go faster the 'go faster' inlet jet on the carb is needed and that grabs fuel from both tanks. You can't see what is in the tanks but you know one tank in normal use lasts about 'X' minutes. But now you are street racing! Both tanks are being used but you don't know how much. When you get to the gas station what do you do? Just fill the first in the hope that when you need to go faster again there will be enough left in the second? You don't want to lose the next race so you probably fill both! Hence in the end long/predictable endurance and a stable sensor environment will always affect buoyancy/weight.
Example 2 :
I want the best WOB but I don't want chest mounted counterlungs.
In fact to get things straight from the start; just because a unit has chest mounted lungs it doesn't mean it will have a low WOB. It may gain hydrostically but it may be like breathing through a sock - resistively speaking. So here again size matters. To keep the resistive circuit low you may need bigger mushroom valves, hoses, counterlung ports and counterlungs, but not too much absorbent as the more you add, the harder it gets to breathe. This obviously depends on the canister design. Many people suggest using a radial canister. That will reduce the resistive effort but it may also reduce the canister duration, especially under a high workload where the 'bed length' just isn't enough to cope with increased ventilation. While a well designed radial may deal with these issues, it is a myth that fitting a radial cures all. It is very difficult to design a radial that performs under all conditions. Back mounted counterlungs can also generate their own WOB issues. Long, thin counterlungs can generate a high resistive load in a vertical swim position (ascending/descending) as you are effectively sucking and blowing gas from a long bag with increasing pressure acting along its length. A centrally fitted lung(s) is far better at generating a lower WOB in all positions. Size in a rebreather also affects how it handles moisture, therefore protects the O2 and CO2 sensors. But don't avoid back mounted system because they look big on a diver. Consider the entire diver profile.
An over shoulder counterlung rebreather is a combination of something inflated on your front and a case on your back and the cases are often square in profile. Back mounted systems only carry volume on your back (and lungs don't collapse when you try to squeeze through a hole) and are often triangular in profile so less cross sectional surface area. If you look at the Henkel Restriction in Ginnie Springs with your back mounted rig on and think ' it isn't for me'......think again!
Small rebreathers have their issues and often do not provide the ideal environment for sensors, WOB or reliable canister life prediction.
Kevin Gurr - http://www.technologyindepth.com