One, two, three, four, what's the magic number?
By: Dale Bletso
The magic number of PO2 sensors in a manual closed circuit rebreather (MCCR) loop has ignited a great deal of controversy on the web. The number of sensors depends on many factors, some of which are personal choices. I will attempt to cover the various modes of failure, and you can determine if the addition of a third sensor is needed as many think. First a little background:
On a semi-closed circuit rebreather (SCR), PO2 monitoring has often been offered as an "option" and not a requirement. I have always asserted that PO2 monitoring is essential with any rebreather. On an SCR it is a kind of redundant equipment. After diving a semi-closed system for a while, most divers can hear the gentle hiss of the gas injector as well as the occasional exhaust of bubbles. If either of these indicators is absent, PO2 can be verified by a PO2 monitor. Most SCR's become very quiet at deeper depths, and more reliance on a PO2 monitor becomes necessary. There have been instances of partial orifice blockage at depth after doing a positive pre-dive check. In my opinion, one PO2 monitor is a requirement for a SCR. Two would lend nothing to system reliability as the monitoring of PO2 is a secondary function on SCR with listening for bubbles being the primary way of monitoring system function.
In an electronically controlled CCR (ECCR) system, three sensors are the norm as the diver is essentially out of the decision loop. The PO2 is monitored by the electronics, and a voting logic is used by the electronics to control the addition of the oxygen needed by the system. The diver establishes setpoints for the addition of O2 into the loop to maintain a constant PO2. When O2 is needed, a solenoid is activated, O2 is added when the upper set-point is reached, the solenoid is dropped out. PO2 in the loop is held relatively constant without diver intervention. Most systems will alert the diver when there is a voting logic problem. The main problem with these systems is that the electronics can be fooled into thinking everything is OK and start adding O2. The diver may be completely clueless until a problem is already critical. ( i. e. if moisture causes a PO2 error on all sensors, it will start adding O2 to bring it back into line.) It can become critically high without the diver ever knowing. These systems are complicated and require a subconscious awareness to discern a potential problem. (Is the add solenoid picking up too often or not often enough?)
On a manual CCR system (MCCR), as with the ECCR, there are no bubbles. On an MCCR the O2 orifice is so small that the injection of the gas cannot be heard. Unlike SCR and ECCR there are no audible signs of system functionality. One PO2 device certainly would not be adequate as its failure would leave the diver totally without any way of knowing the PO2 in the breathing loop. Electronics and water, particularly salt water, do not mix. It is not a question of if a sensor will fail but when the diver will have a sensor failure. So are two or three PO2 devices needed? After all, monitoring your PO2 is the only way you can verify if the system is functioning as expected. This is where the controversy begins. I contend that two is sufficient. I certainly don't think three is bad, just not necessary. Why is that so?
While diving any rebreather, it is generally acknowledged that any problem warrants aborting the dive by switching to open circuit. This is certainly sound advice. But before you switch you should determine if there is indeed a problem and where it is to avoid it in the future. There is little a rebreather diver can do to clear a problem; and as with open circuit when you have a gear malfunction the scenario is the same. (i. e. terminate the dive.) The only procedural difference when using a rebreather is whether you terminate the dive open circuit or take a chance and terminate the dive closed circuit. This is where two or three monitors can have an impact. Using three monitors, the tendency for the diver accept the two like readings as correct and attempt to continue the mission. This certainly sounds reasonable. But why could this procedure be considered foolhardy?
When one of a group of PO2 monitors varies from the other(s), one must first try to find which one is displaying incorrect information. The assumption with three sensors is that the two that agree must be correct. This I believe to be just human nature. There are some failures which can be seen over two or more sensors simultaneously. My self (as well as others) have had simultaneous multiple false readings. The proper response always starts by purging the system. In purging the system you can immediately verify which units are working by simple math. (i. e. 30 m is 4 ata, diluent 0,21 means 0,21 * 4 = 0,8 ish ata.) The monitor that does not read in this range is the malfunctioning one. How close do monitors have to agree? They should be reasonably close. A sensor indicating 0,05 ata from the other would not make me unduly concerned. If on the other hand they are 0,1 ata or 0,2 ata off, I would do a flush and reverify. More and I would certainly terminate the dive.
On the other hand, there are any number of things which can cause disparate readings. Placement of the sensors, the amount of humidity, temperature, and battery condition all can affect the sensors. Of course, in a rebreather humidity is the greatest enemy. A thin film of water over a sensor can cause it to read lower than what is in the bag. The lower reading is caused by a film of water isolating the sensor from the gas stream. The sensor then slowly uses up the O2 under the film and causes a low reading. Even placing three sensors in a loop won't eliminate this potential problem! On a manual CCR system, a simple flush will verify a sensor's accuracy. A system purge may also help evaporate excess moisture. A static reading on a single or multiple sensors irrespective of exercise level is usually a sign of a moisture trap. Improper assessment of low readings could lead an inexperienced MCCR diver to think his injection is low and to start adding O2, thereby placing himself at risk of a CNS incident. Due to the nature of a MCCR system, a flush will establish a base from which to make your decision to bail.
A system purge should bring one or all sensors to a correct value. If it doesn't, then the assumption is a moisture trap or sensor failure; and OC bailout is warranted. If the sensor or sensors stabilize and then continue to drop, a possible obstructed O2 orifice can be suspected. A termination of the dive is then indicated using the manual O2 addition valve but only if both sensors are tracking uniformly. If they are not, go to OC and bail. Does the addition of a third or fourth sensor add significantly to system reliability or does it add more potential failure points?
The possibility of a sensor failure giving a higher reading has been proposed by various people on the web. The actual modality of this failure mode has been explained and may be due to poor construction techniques which will allow more oxygen through the membrane thus allowing higher than anticipated readings. Mechanical shock may also play a role in some manufacturers units. The potential for multiple failures of this nature seem likely. However, high readings can and do occur even without a cell failure. A high reading can indicate a cell failure, electronic failure, a defective manual O2 addition valve or a creeping or high intermediate pressure (IP) to the O2 orifice. In any case, it cannot be dealt with underwater. An OC bailout is warranted if there is any suspicion of any of these failure modes. Being able to ascertain if a failure is electronic or mechanical will determine if you terminate the dive CCR or OC. If you experience a high PO2 alarm, it is important to remember that high PO2's can cause convulsions and death. If it is a mechanical problem, it would be very important to verify this quickly.
The verification process is the same as with a low reading and is simple: flush, read, and do the math. As mentioned above, a high reading can only be caused three ways: First, an electronic malfunction; second, by a defective manual addition valve; and last, a high IP from the drive regulator. An addition valve failure or IP problem can be verified by shutting off the O2 supply, flushing the system and verifying that the PO2 stabilizes and begins to slowly drop. Depending on depth, this can take some time. If, on turning the O2 supply back on, the PO2 starts to climb, it is time to OC bail. Remember that in a manual CCR system O2 for metabolism is usually added via an orifice supplying less O2 than is required to maintain metabolism in a resting state. Any additional O2 need is added via a manual addition valve. If flushing doesn't lower the reading, you have a sensor or electronic problem and its time to go OC. Again, does the addition of a third or fourth sensor add significantly to system reliability or does it add more potential failure points?
I have attempted to outline the types of failures encountered on a MCCR. The most powerful computer (your brain) and a little common sense can make a manual system as safe as, if not more so, than an ECCR. In my opinion, more than two sensors will not necessarily increase the reliability of the system. On the contrary, it may actually lead to a false sense of security because any sensor failure should lead to the termination of the dive. In all cases, each sensor should be completely independent and isolated from the others. No common power supply (battery), cabling, housing, etc. should be used. Further, all sensors should always be regarded with a healthy degree of skepticism Whenever different readings are encountered, it is important to determine the probable cause and to take the appropriate steps. During a dive, frequent monitoring and an occasional flush will enable you to verify proper system function. Even if using three sensors, never assume the two that agree are correct. It could cost you your life.
Above all, remain calm and take decisive methodical steps in determining the next course of action. Remember that not all failure modes involve a single sensor. Training and experience are the keystone to surviving while diving any type of rebreather. With a MCCR as with an ECCR the diver must take the time to know his system under various conditions.