The design of an Oxygen Rebreather for exploring the St Clair Cave, Jamaica

1. Introduction

In this document we describe the different parts of an oxygen rebreather for exploring and surveying the Acheron with its foul air. The main purpose of the design is to make it possible to visit in the Acheron during three hours (1 hour go, 1 hour return and 1 hour reserve), this with a double safety margin.

This means the rebreather should be reliable for minimal 5 hours doing light too moderate exertion. When in rest, consuming much less oxygen, the dwell time should rise to 15 hours and higher.

Other requirements are: easy to build, light, cheap and general available components, more or less cave proof.

2. How an oxygen rebreather works

2.1. The very short:

An oxygen rebreather is a closed breathing system filled with 100 % oxygen which recycles the user’s breathing gas. It provides oxygen to, and absorbs carbon dioxide produced by the user. The main parts of a rebreather are: the mouth piece, breathing hoses, counter lungs, scrubber, oxygen supply and regulator.

2.2. The not so very short:

Let's say the rebreather is ready to be used and the counter lung is filled with 100 % oxygen. When we now inhale from the counter lung it will be emptied for a great part. Our body will use the inhaled oxygen it can use and convert it to carbon dioxide. When we breathe out into the mouthpiece there will be now about 4 % CO₂ in the oxygen. The exhaled gas passes through the scrubber filled with lime where the CO₂is absorbed and then the gas (now mostly oxygen again) is going into and filling the counter lung again where it is ready to be inhaled the next time.

As you can see, 4 percent of the gas volume has gone since it has been absorbed by the lime in the scrubber. That is why a constant flow of oxygen is added into the loop. So the volume we lose by absorbing the CO₂ is refilled by a constant oxygen flow from the compressed gas cylinder.
The trick is now to determine a good flow of oxygen. Too much oxygen and the gas will be wasted through a relief valve. This happens when we are in rest consuming less oxygen and as result producing less CO₂. So the constant flow of oxygen will surpass the amount of CO₂ absorbed. This excess of gas will be vented to the air. This valve is located before the scrubber. Saving valuable scrubber and oxygen.

If the oxygen flow is too low the counter lung will be empty after a certain time and we can't breathe anymore. This happens when we do heavy work. then we need more oxygen and produce more CO₂than added by the constant flow. A manual bypass valve will put more oxygen into the loop.

There are a lot of complicated tricks to regulate the oxygen level in a rebreather but for our goal a constant flow of oxygen is good enough. This is rather simple in construction at the cost of having some venting when we are resting. A manual bypass will aid us when we need more oxygen than the normal flow.

We breathe through a mouthpiece. But once the rebreather is in use we can't take it out of our mouth for e.g. talking, drinking ... And a nose clip is a must!! Otherwise the risk is too great to bring nitrogen into the system. And that is something we certainly don't want. Because once in the system it stays there (the absorber can't absorb nitrogen) and can reduce rapidly the effective oxygen volume we have in the breathing loop. Here is an example:

Let’s assume the total volume of the breathing loop (this is mouthpiece, hoses, canister and counter lungs) together with our lungs is 10 litres (0.36 cu ft). Now we take one breath from the outside through our nose and exhale into the breathing loop. That makes about 3 litres (0.1 cu ft) of air. The composition of air is 79 % N₂and 21 % oxygen. So our 3 litres of air contains 2.4 litres (0.08 cu ft) of nitrogen. 2.4 litres N₂ on a total volume of 10 litres means that 24 percent of our breathing loop will be filled with nitrogen gas!!! As you can see any addition of outside air will make the gas in the loop very fast unbreathable. The user of the rebreather will lose consciousness rapidly once there is 84 % percent inert gas (nitrogen and/or CO₂)!

There are several causes why the percentage of CO₂ in the loop can rise:

1. when there is a breakthrough
2. the absorber has worked out
3. there is channelling through the absorber bed

Only the breakthrough is not too bad. This happens during heavy exertion when more carbon dioxide is produced than the absorber bed can handle. Once the workload returns to normal the breakthrough will stop. The other causes are fatal and must be avoided by careful planning. Always fill the scrubber with a new load of lime when using the rebreather. We never know what happened with the lime in the scrubber and old fillings must always be disposed. Only a new load of lime will guarantee us the good working of the scrubber. And the filling of the scrubber must be done very careful and in a controlled way. The lime must be stacked firm enough but still allowing gas to be passed but no rattling must be heard when the scrubber is shaken.

Any left over air in the rebreather must be purged before use. Normally purging goes like this: we breathe through the mouthpiece and exhale through the nose (without the nose clips). This we do till the counter lung is empty. When ready we must put the nose clips again on. But to shorten this purge cycle (should be done three times!) the system will be pulled vacuum through a manual pump before filling it with oxygen.

I confess, scuba gear is easier but for what we want to do a rebreather is the way to go. In principle a rebreather can be made very simple,  but I don't want to gamble with lives. So I want it to be safe too. That's why I also put all the electronics into it to measure all the parameters to check the quality of breathing gas. In theory one can say you do not need all these gizmo's but hey, we are not jumping into the swimming pool in our backyard but we are going into the Acheron! It's something like exploring the other side of the moon, only a lot cheaper ;-)

3. Parts of a rebreather

3.1. Mouthpiece

The question was if we were going to use full face mask or a mouthpiece. The full face mask seems to be the most favourable but in the hot humid climate where we will use the rebreather the mouthpiece got the preference. The main problem with a face mask is the seal between the face and the mask which is not 100 % air proof. The risk for contaminating the loop with nitrogen from the air is possible. Another problem is fogging of the face mask and excessive sweating, making it all very uncomfortable. We need a good view during the exploration, so everything hampering our sight must be banned.

The mouthpiece of a rebreather has two hoses. One for exhaling and one for inhaling. There are also two valves inside forcing the air in one direction through the breathing loop, enabling the right functioning of the rebreather.

If possible a shutoff valve should be installed. This prevents losing oxygen and polluting the loop with nitrogen when pulling the mouthpiece out of the mouth. But this also makes the construction more difficult. The shut off valve is a must when diving where the risk of flooding the loop with water is prominent. But a plug can do the same job in out of the water situations. The mouthpiece can be extracted from the mouth for a short time enabling the plug to be inserted without too much worries. The loop pressure is higher than the environment preventing nitrogen to rush into the system.

An elastic strap will be attached to the mouthpiece giving some relief to the jaw muscles.

3.2. Scrubber

The scrubber is a container with a bed of lime. The air is forced through the lime and will remove the user's exhaled CO₂ through an exothermal chemical reaction.

We go for the vertical mounted axial scrubber. The radial scrubber seems to be more efficient but the axial type is a much simpler construction. Also the vertical axial scrubber is less prone to channelling. We can go for this design because the rebreather will be used most of the time in upright position, this in contradiction with a diver which has much more axis of freedom. The diameter of this canister must be > 15 cm. This to assure an optimal surface for the reaction front in the scrubber.

3.3. Breathing hoses

These are part of the breathing loop and connect the mouthpiece with the scrubber and counter lungs. They have to be flexible enough to assure a certain degree of comfort during use. Also the hoses must have a minimal diameter of 2 cm (3/4 inch) enabling minimal breathing resistance. A higher diameter is even better but this will raise the weight of the machine. Remember we carry the rebreather on the back. For a diver weight is not really a problem because of the law of Archimedes. Some dive rebreathers do indeed weight more than 100 pounds!

3.4. Lime

The rule of thumb is that 1 kg of lime is needed for 1 hour rebreather time doing moderate exertion. The scrubber will be filled with 5 kg of lime. This gives us a large enough safety limit.

3.5. counter lung

A rebreather is a closed circuit. So when we exhale the air has to go somewhere to be temporary stored. This is done in the counter lung. It is a flexible bag which expands when we exhale and contract when inhaling. So the total volume of the loop remains constant during the breathing cycle.

Instead of using one counter lung we are going to use a split counter lung. This means there is a breathing bag before and after the scrubber. The split design allows a longer dwell time of the air in the scrubber, increasing the ability of the lime to absorb the carbon dioxide. Only one half of the gas volume goes through the scrubber during exhalation. The other half goes through when we inhale. In the case of one counter lung all the air is forced through the scrubber bed during exhalation. Resulting in faster gas velocities and bigger volumes, reducing the reactivity of the lime.

3.6. Oxygen cylinder with constant flow mechanism

The oxygen is delivered in the rebreather as a compressed gas. See chapter 4 for more details.

3.7. Relief valve

When oxygen consumption is lower than the constant flow, the counter lung is filled to capacity and the system must vent through a relief valve. The pressure for opening the relief valve must be lower than 1.6 bar. At this pressure oxygen becomes toxic!

3.8. Manual demand valve

There is the possibility to bypass the constant flow of the oxygen when the consumption is higher than the delivered amount. This valve should be placed in a convenient place but safe enough to prevent accidental activation or damage. The right side of the rebreather at the lower end seems right.

3.9. Cooler

A lot of heat is generated in a rebreather. A way to dispose the heat is necessary. There are different ways to cool down the air. Some constructions use sealed tubes with low melting point salts like potassium phosphate hydrate. This is rather heavy and only useable till all the crystals have melted. Other designs use ice instead of salts. It has been proved that this method is not feasible for our goal. The ice provokes condensation generating more heat than the ice can take absorb by melting. another way of cooling is a heat exchanger. This with the extra help of water evaporating on the surface seems to be the way to go. This passive method can be given an extra boost by adding an extra fan forcing the evaporation of the water. See also chapter 7.

3.10. Oxygen sensors

We need to know what we breathe. This should be almost 100 % oxygen. But there will be a certain percentage of nitrogen in the gas too. This must be kept as low as possible. Nitrogen can't be absorbed by the scrubber and will accumulate in the loop till the percentage of oxygen will be to low to be breathable. This is also the reason why we purge the system before use. To aid this purging we will pull the system vacuum. Another source of nitrogen will be the gas dissolved in our blood. A certain amount of nitrogen will be freed from our blood since the partial pressure of nitrogen in the loop will be less than that in our blood.

Another gas in the loop will be CO₂. Normally this gas should be absorbed by the scrubber. But there is the possibility of a breakthrough of the scrubber or that all the lime in the scrubber is depleted. 10 % CO₂ in the loop means a certain death. Problem is that for the moment there is no decent (cheap) way of measuring CO₂ in moist air. The only way we can check this is to be on alert for the symptoms of CO₂ intoxication. Therefor we will monitor a list of different parameters of the system each five minutes. These parameters could be monitored automatically but it is better that the explorer checks the system himself. This as a self-test. If we don't succeed in this simple task due to concentration problems or fatigue then there are indeed troubles. That's also the reason why we go with two persons, to check each other with this task. The moment there are problems we have to end the exploration and return to base.

3.11. Pressure gauge

The pressure gauge is to monitor the pressure in the oxygen cylinder. The pressure gives an indication of how much oxygen we still have and is a good indication of how long we still can go.
The pressure gauge must be easily accessible. So a long flexible tube will connect the oxygen cylinder and the gauge. With a configuration like this the gauge can be attached to the chest straps.

3.12. Water traps

There are different sources of water available in a rebreather:

- The chemical reaction in the absorber produces water, roughly 157 ml of water is formed per hour!

- a lot of saliva is produced because of having the mouthpiece in the mouth. This amount will differ per individual.

- there is also a lot of evaporation of water through our lungs. What we exhale is moist air.

All this water must be kept out of the system because it can cause serious harm. Too much water in the scrubber can influence the efficienty of absorber in a bad way. another problem that can happen is the flooding of certain parts of the loop making breathing harder and hampering the good working of the system.

Throughout the rebreather temperatures goes up and down forcing the air to contain more or less water vapour. This vapour transports water through the loop. Condensation is to be expected at the coldest places of the loop. In the scrubber the temperature can rise to 100 degrees C and can indeed produce steam. What we exhale will always be 36 degrees C. The coldest point in the rebreather will be the cooler, which will have a temperature of probably 30 degrees C. So the cooler must be built as a water trap. Another water trap will be the bottom of the scrubber. This is a big device and can hold a lot of water.

Through the exhale tube saliva will dribble down and must also be collected or purged. The relief valve will be put where the saliva collects. So when the loop purges this source of water will also be disposed.

4. Human oxygen metabolism

Table of oxygen metabolism of a human (male 70 kg)

(Source Mastering Rebreathers, Jeffrey E. Bozanic, p 83)

According to Ake Larsson (www.teknosofen.com) oxygen consumption as high as 2 L/min is rarely seen in real dives.

In the NIOSH report RI9650 “Performance Comparison of Rescue Breathing Apparatus” the rebreathers are tested against:

VO₂ (O₂consumption) = 1.35 l/min
VCO₂ (CO₂production) = 1.10 l/min
Ve (ventilation) = 30 L/min
RF (respiration frequency) = 17.9 breaths/min

Tests are terminated when CO₂> 10 % or O₂ < 15 %

For the exploration of the Acheron an oxygen consumption of 1.1 l/min will be used. Surveying does not take much energy and can be catalogued as a light activity. Of course exploration can be moderate to heavy exertion. But we have to prevent this and do everything in a slow and relaxed way (the Jamaican way…). A temporary high oxygen demand can be catched by the extra addition through the manual demand valve. But what we have to prevent is the breakthrough of the scrubber bed. A breakthrough happens when the volume of carbon dioxide is bigger than the scrubber can take. Resulting in carbon dioxide passing through the bed, polluting the loop.

5. Oxygen cylinder

We had to find a local source of oxygen in Ja. Thanks to Jan this source was found at IGL (industrial gases ltd) Kingston.

D Size aluminium Oxygen cylinder containing 425 l of oxygen (15 cu feet)

CGA 540 port valve for bigger volumes
DOT 3AA or DOT 3AL (DOT: department of transportation)
where 3AA =  chrome molybdenum and 3AL = aluminium

Medical oxygen is shipped at a pressure of maximal 2015 psi (139 bar) for safety reasons with the special CGA870 valve.

The real volume of this bottle (when filled e.g. with water) can be derived by

Boyle's law: P1.V1 = P2.V2
pressure 1 x volume 1 = pressure 2 x volume 2, temperature must be the same

Let's say atmospheric pressure is 14.6 psi, so we get

real vol x 2015 psi = 15 cu ft x 14.6 psi
or
15 cu ft x 14.6 psi / 2015 psi = 0.11 cu ft or 3 litre

So Model D has a water capacity of 0.11 cu ft (3 l) and can contain 15 cu ft (420 l) of oxygen compressed at 2015 psi (140 bar).

If we breathe 1 cu ft/min (28 l/min) then our body will use 4 vol% of the oxygen (this is the same in air or in pure oxygen, the body takes what it needs). So we breathe out 4 vol% CO₂. It's this CO2 which has to be chemical absorbed and topped with the oxygen from the tank.
So each minute 0.04 cu ft (1.1 l/min) of oxygen has to be put in the closed loop. And since we have 14.8 cu ft (420 l) we can go 14.8 cu ft / 0.04 cu ft/min = 370 minutes or more than 6 hours! This is in a perfect rebreather. A certain pressure is necessary in the interstage pressure regulator to work and deliver the constant flow of 1.1 l/min. To play on the safe side a pressure drop of 10 bar is supposed over the regulator. This gives a netto volume of 390 liters instead of 420 liters! This gives us a net time of 354 minutes or 5.9 hours, which is still in the safety limit of 5 hours as postulated in the introduction.

Since we use a constant mass flow of 1.1 liter or 0.04 cu ft it's possible some of the oxygen will be purged in moments of low activity and at moments of high activity we'll have to bypass the constant mass flow to add more oxygen in the loop. But we are still on the safe side of the requirement of a 3 hours rebreather knowing we have an 5 hours oxygen supply. And that's the way I like it. With normal diving gear we would never get this done with such a safety margin.

6. Lime

Soda lime is composed of water (16-20 %), sodium hydroxide (NaOH) or potassium hydroxide (KOH) and Calcium hydroxide (Ca(OH)₂) where calcium hydroxide is the most abundant compound (70-80%). The bulkdensity is about 0.9 kg/l

Commercial Soda lime a specially prepared mixture of calcium and sodium hydroxides. The granules are creamy white in color, hard and processed to minimise dust formation. It is also porous, and irregularly shaped to provide for a larger surface area.

When put in contact with a acidic gas like CO₂, a strong, exothermic (heat producing) reaction takes place which gives off water and binds the CO₂ by forming a stable Calcium Carbonate.  When binding to other acidic gases, other calcium or sodium salts are formed. Sometimes a color indicator is added which aids in indicating the useful life of product.  As the soda lime becomes less effective at binding CO₂, the color of the product changes to a violet color. This color change should not be your sole method of determining when to replace your absorbant, but it is an indication of use.

Commercial products are e.g.: Sofnolime (Molecular Products ltd), SodaSorb (W.R. Grace & Co.), Divesorb (Drager)

6.1. Reaction

The general description of the reaction is as follows: First the gaseous CO₂ reacts with water to form carbonic acid - H₂CO₃. Then the NaOH reacts with the carbonic acid to produce Na₂CO₂ and H₂O. The Na₂CO₂ reacts with the Ca(OH)₂ which has been disassociated into Calcium and Hydroxide Ions. (Ca⁺⁺ and OH^-) to produce CaCO₃ (calcium carbonate, otherwise known as limestone.) The CO₂ is now in a relatively stable state. There is a net production of three H₂O molecules for every molecule of CO₂ which is taken in.

Here is the reaction in more detail.
The carbon dioxide first reacts with the moisture in the soda lime to form carbonic acid: CO₂+H₂O-> H₂CO₃(-19.36 kJ/mol)
The carbonic acid then reacts with the sodium and potassium hydroxides to form sodium and potassium carbonate and regenerate water (neutralisation reaction)

2H₂CO₃+NaOH+2KOH -> Na₂CO₃+K₂CO₃+4H₂O (-89.4 kJ/mol)

The caustic alkalis draw the acid gas out of the passing mixture and hold it. The sodium and potassium carbonates then react with the hydrated lime, forming the unsoluble calcium carbonate and regenerating sodium hydroxide and potassium hydroxide

Na₂CO₃+K₂CO₃+2CA(OH)₂-> 2CaCO₃+2NaOH+2KOH (+12.3 kJ/mol)

each mole of absorbed CO₂ (44g) will produces three mole of water (54 g)

1 mole of carbon dioxide weights 44 g and has a volume of 22.4 l at atmospheric pressure.

1 l of CO₂ weights 1.97 g

1.1 l/min of oxygen is added to the loop which is converted by our metabolism in CO₂, giving 1.1 l/min x 60 min/h / 22.4 l/mol = 2.9 mol/h

2.9 mol CO₂/h x 3 mol H₂O/mol CO₂= 8.7 mol H₂O/h

8.7 mol H₂O/h x 18 g H₂O/mol H₂O = 156.6 g H₂0 = 156.6 ml H₂O

about 157 ml of water is formed/hour through the chemical reactions of CO₂ absorbtion.

6.2. Which mesh size to use?

4-8 grade (granule size of 2.5 to 5 mm or .098" to .197") is the most popular for recreational dives. The 8-12 (1.0 to 2.5 mm or .039" to .098") will last about 1.3 to 1.5 times longer then 4-8 given the same dives. If you are doing extreme dives you probably want the 8-12, if you are paddling around on reefs, the 4-8 will do just fine.

7. Designing a Cooler

Two sources of heat are available in a rebreather:

• Our own body heat. A rebreather is a closed system, so we breathe our own recycled air. The temperature inside the rebreather will rise to 36 °C (96.8 °F) without any other input of heath.
• The chemical absorption in the scrubber of the expired carbon dioxide is highly exothermic. This can heat the air in the loop to temperatures higher than 80°C (176 °F), making it impossible to breathe.

As tested by myself (on June 30 2008) air hotter than 50 °C (122 °F) is unbearable. So the air in the rebreather must be cooled to a temperature were we can do the job as comfortable as possible.

Normally, in a rebreather there is no direct source of cooling available. But if the system was used during diving then a substantial amount of cooling would be delivered by the surrounding water. But since we are using the rebreather in the hot moist air of a Jamaican cave no cooling power can be suspected from our environment.

So we have to add a cooler. The first design was using ice as demonstrated in the professional rescue breathing apparatus by Drager, the BG4. In this machine a 1.2 kg (2.65 lbs) block of ice sealed in a plastic container is used. Test conducted by me on 30 June 2008 proved that this is a bad system. One of the weak points is that the cooling ice invokes condensation of the moist air inside the rebreather. Condensation of water vapor releases almost 7 times more heat than the heat needed to melt the same amount of ice. So most of the cooling power is wasted in condensating water. A second drawback is that 1.2 kg (2.65 lbs) of ice can only absorb a limited amount of heat till it is all melted and the water has the same temperature of the inside air of the rebreather. This means that the cooling power is finite in time ( Heat of fusion of ice 335 J/g, Heat of vaporization of water 2260 J/g).

Only reason I see why they use ice in the BG4 is the psychological factor. Something like: it’s ice so it cools the best. And the machine is from Drager and it costs a fortune so it must be good.

On 3 July I did test with another type of cooler. The air was send through a 3 meter (9.8 ft) long copper tube covered with a wet cloth. This setup of passive cooler was impressive. The cooling power was 30 °C (86°F) between both ends of the tube and the system did not need any extra power to cool. Extra ventilation to stimulate the evaporation of the water in the cloth gave an extra boost of 5 to 10 °C.

The construction of a copper spiral was not very practical and after several trails (wasting some meters of good copper tubing) and other design was necessary. The final design is a long thin walled plastic tube which worked remarkably well.

8. Cleaning the rebreather

All parts of the breathing loop should be disinfected (except the scrubber). There are different brands of disinfectant like poloxamer-iodine (Wescodyne), Virkon S or RelyOn MDC (Dupont), but povidone-iodine (10%) (Betadine) can also be used.

9. Purging the rebreather before use

A rebreather has to be purged before use. Otherwise a large volume of inert gas (N₂) can cause low oxygen levels leading to hypoxia (lack of oxygen).

The system has to be purged by repetitively inhaling all the gas in the loop through the mouthpiece and exhaling through the nose. Do not inhale from the ambient atmosphere during this procedure. Once the loop is as empty as possible, the loop has to be refilled with oxygen from the supply cylinder. Repeat this three times without removing the mouthpiece. Once this is done the unit ready to use.

Also if you take a breath of air, for example to speak to someone, you will have filled your lungs with a quantity of nitrogen. Before resuming use of the rebreather, you must purge your lungs of the inert gas by repetitively inhaling from the rebreather and exhaling through your nose. If you exhale into the rebreather without doing this, you will add inert gas to the system, creating the potential of hypoxia.