The Ouroboros Rebreather. A Technical Review.

My first encounter with rebreathers was in the late 1980’s and consisted of a 200m chamber dive on an experimental unit. It failed. This brief experience and a range of deep open circuit diving projects convinced me that rebreathers were the future. Almost eight years ago, having decided the choice was limited, I decided to build my own. This is a brief summary of what I have learnt and how combined with a review of recent incidents, I have finalised on the current design.

Fundamentals

In order to understand what makes a good rebreather, firstly one must understand the fundamental properties that govern a rebreather design. This can be divided into 4 areas.

1. The resistive work of breathing (WOB) within the rebreather
2. The hydrostatic WOB of the unit when submerged
3. The absorbent duration
4. The oxygen control dynamics

Looking at each in turn.

The Resistive Work of Breathing

This is purely a result of the gas flow restrictions within the unit. In other words how much the size of the pipes and orifices generate a resistance to breathing. Such things as small mouthpiece mushroom valves, small hoses, counter lungs with insufficient volume or room to expand and long absorbent paths within a canister are common elements which go to make up a resistive breathing circuit.

Restive WOB is also a function of gas density and hence depth. The deeper the dive and the higher the gas density the greater the WOB. WOB is also a function of ventilation or breathing rate. The more gas flow (higher breathing rate) the mores resistance is generated. A rebreather that breaths OK on the surface may well not at 40m on an air diluent. This why current CE (Central European test standards) and military test standards insist on a resistive WOB measurement at depth and with different ventilation rates and normally in two orientations (swimming positions).

The Hydrostatic WOB

This is the result of the resistive WOB and the effects of the position of the counter lungs about the body when the rebreather and diver are submerged in water. For example, a back mounted counter lung rebreather may have a good resistive WOB but when in a horizontal (face down) swimming position the distance, hence pressure difference between the counter lungs and the lung centroid, may when combined with the resistive WOB create an excessive pressure which the diver has to suck against in order to take a breath. In this case the inhale pressure would be excessive (because the diver is inhaling gas from a lower pressure) and the exhale would be easy having breathed out into a lower pressure.

Chest mounted counter lungs have the reverse affect in the same swim position.

It would seem a perfect solution is an over-shoulder counter lung upon which any hydrostatic effects have little result. However if a rebreather has a poor resistive WOB anyway, the combination of this and any minimal hydrostatic WOB can still mean the unit has a highly restive breathing circuit and hence have a laboured breathing ‘feel’.

In summary a rebreathers WOB can only be quantified under a range of hydrostatic (rotating) positions with additional measurements at depth. Furthermore, assessing a unit under a range of surface conditions at undefined ventilation rates is inadequate as it is only under stressfull conditions (and normally at depth) that we generate high work-rates and this is precisely when the rebreather needs to have a low WOB.

Absorbent duration

This is a greatly misunderstood area. Many manufacturers quote durations based on simple surface trials which is insufficient and potential dangerous. Absorbent life is primarily affected by the following;

1. Amount (Kg or Lbs) of material
2. Type of absorbent material
3. Water temperature
4. The absorbent canisters ability to insulate against the water temperature
5. The amount of CO₂ generated by the diver
6. The gas density/depth
7. The style and design of canister

The standard CE test for a canister is done at 40m with air as a diluent in 4 degrees centigrade water temperature at a CO₂ generation rate of 1.6 l/min and a ventilation rate of 40l/min. Some navies test canister at around 18m and with as low as 0.5l/min CO₂ generation. The CO₂ breakthrough figure is referenced to 5mb.

The range of depths, gas densities, CO₂ rates and water temperatures used in each case can, dependant on the canister design, give markedly different durations. What is certain from the data available is that canister durations measured at the surface are dangerously inaccurate for predicting overall dive durations. It can be easily shown that for most axial and even radial canisters the efficiency of the canister decreases significantly with depth. One canister measured went from 77% efficient in 15m to 49% efficient in 40m. The 15m duration was 3 hours while the 40m duration was 1 hour and 50 minutes.

The ‘saving grace’ of currently available designs is that most people cannot maintain rates of 1.6 l/min CO₂ or seldom dive in 4-degree water. However data suggests that as a method of specifying canister duration, a single hourly rate independent of primarily depth and gas density, is insufficient and a range of diving parameter curves may be better employed to specify a unit’s duration. In other words in order to assess a rebreathers’ suitability for one type of diving over another, it would seem appropriate to test a unit at a range of depths with at least air and trimix as gases. To provide a safety margin water temperature and CO₂ generation rate should remain constant.

Oxygen control system

This can be a mechanical or an electronically controlled device. Which ever is chosen it is important that the control of the oxygen level is maintained within certain limits. If decompression is to be conducted using tables or a fixed PO₂ dive computer, the limits must be accurately maintained. Rapid excursions to and from depth must also not generate excessively low or high PO₂’s. Ideally any such limits reached should generate an alarm, as it cannot be relied upon that the diver will notice especially in a multi-tasking situation. Some rebreather designs, due to high flow rates within the oxygen circuit, can generate massive PO₂ spikes sufficient to cause convulsions in a short space of time if left unchecked, this is an undesirable design feature should an addition valve fail. Within the CE and most Navy tests is a PO₂ tracking control test as well as a upper and lower limit test after rapid depth changes. With units employing a constant minimum feed of oxygen, which is supplemented by the diver, then this minimum should not generate the upper and lower test limits specified.

The Ouroboros

With all of the above in mind and a review of rebreather accidents I started designing my unit.

People had problems because;

1. They started the dive with their electronic control system off
2. They started the dive with their oxygen turned off
3. They descended with diluent off and then panicked when they could not find the manual addition
4. They did surface swims on hypoxic diluents
5. They did not pack the absorbent canister correctly or the design of the canister allowed CO₂ to bypass if O rings where incorrectly greased or assembled.
6. With insufficient guidance on canister durations, people exceeded the duration limits
7. Temporary floods made the breathing loop unusable
8. Insufficient filtering produced oxygen solenoid failures
9. Hose attachment systems produced stress points which perished the hoses
10. Electronics in the loop was affected by moisture
11. Gas supplies were accidentally switched
12. Failures in the electronics made the unit unusable
13. They become stressed at high work rates
14. DCI occurred as a result of the units inability to maintain a near constant PO₂
15. They did not follow pre-dive procedures

The Ouroborus is designed with the above in mind and as a result has the following features;

1. A system that protected all ‘soft parts’ and pressure lines
1. As a result, back mounted counter lungs are employed in the primary design. All internal pipe work is stainless steel, breathing hoses are internally sprung with a rubber and nylon covering to prevent abrasion and crushing.
2. A low WOB

3. A canister that would auto pack with little or no chance of channelling due to mis-packing
1. Axial and tower radial canisters suffer from large amounts of ‘pack down’ due to the long length of absorbent. Short ‘doughnut’ radial canisters have less of a problem. Axial canisters especially, when not packed correctly, can produce a bypass channel when tipped flat. The Ouroboros uses a doughnut radial design
4. Efficient water removal systems
1. The split counter lungs provide an efficient water trap. Water can also be dumped from the exhale counter lung automatically
5. A canister with high efficiency even at depth and with high CO₂ rates
1. The unit has been tested at and will be rated for a range of depths
6. A ‘no tools’ unit for general maintenance
1. Daily maintenance requires no tools. Even electronic parts can be ‘field stripped’ with a multi-tool. No specialist tools are required.
7. Electronics with manual overrides
1. In the event of an electronic failure, a separately powered and isolated means of monitoring PO₂ is vital. In units where dual electronics provide all the control it is advisable that two different software writers are employed to avoid common ‘bugs’. The Ouroboros has a separately powered, triple PO₂ meter display, even cable severance will not interfere with the main electronics.
8. On board decompression
1. This reduces the effects of units not accurately tracking PO₂ when hard tables or fixed PO₂ computers are used
9. No electronic lockouts
1. Given that units will be used for cave and deep diving, there should not be an instance where the diver cannot over ride the electronics and complete the dive. All automation on the Ouroboros is de-selectable to allow manual over rides.
10. Ease of assembly with no chance of misplacing parts or gases
1. All primarily maintenance is ‘tool free’ with no possibility of connecting gases incorrectly. External gas supplies are also coded to avoid swapping.
11. A modular system for different diving styles
1. The primary design is a back-mounted in a carbon fibre case. Different canister durations are available. A ‘travel’ system with a soft pack design and even a chest mounted counter lung format is available
12. Intuitive alarms
1. Alarms available through a head up display (HUD) are backed up by ‘on screen’ detail on the current alarm
2. A rear facing display is available for ‘buddy’ diving and instruction
3. Alarms are visual and also via a vibration system in the HUD
4. Digital HP transducers ensure cylinders that are turned off generate alarms as do any leaks in the LP or HP systems
13. Automatic turn on systems once in-water
1. Wet turn on contacts, a pressure activated turn on and turn on via switches provide redundancy to ensure the chance of an hypoxic incident is reduced
14. Minimum life support
1. Once turned on, irrespective of the ‘set point’ a minimum PO₂ of 0.4 is always maintained
15. No non hermetically sealed electronics in the breathing loop
16. Low flow gas paths to reduce PO₂ spiking but high enough for good diluent flushes
1. All gas flow lines are flow matched to avoid ‘spiking’. The automatic diluent addition valve can be user adjusted for smooth descents
2. Both the Auto diluent addition and oxygen solenoid are protected by in-line filters
17. Easy interchange ability with open circuit safety equipment
1. The unit will accept a range of back plates and wing/harness system. Designed especially for instructors that often have to switch back to open circuit.
18. Data logging
1. Data logging provides useful feedback in training and incident scenarios. Every parameter is logged including when the user aborts a pre-dive sequence
19. Isolation of automatic circuits which then allows manual over rides
1. Automatic polling of sensors as well as manual isolation in the event of failure can be achieved. Closed circuit decompressions can be disabled in favour of an open circuit decompression.
20. A simple but effective pre-dive sequence with short set-up and breakdown times
1. An electronic pre dive sequence is available. Simple canister packing and ‘no tools’ assembly ensure rapid preparation for diving.

The Ouroboros is a modular design. The primary layout is a back mounted system in a hard carbon shell. Lightweight cordura covers are available with differing canister sizes as well as a chest mounted counter lung configuration. The unit will be in production first quarter 2005.

Kevin Gurr