-1.5 Oxygen Monitoring

1.5 Oxygen Monitoring

At this writing, we seem to be on the eve of a new paradigm in oxygen monitoring via solid-state (IR) oxygen sensor technology, and I will amend this section once there’s more definitive advantage to present.

However, tried and true remains galvanic oxygen cells. Manufacturers advise reasonable expectations of oxygen cell lifespan, and in recent years, we’ve been well educated on how to check cell health via output voltage linearity. Despite this, cells do fail from time to time and have vulnerabilities that we cannot ignore from the physics of water and electricity not mixing. As with the balance of the rebreather system, redundancy is important – both at the display level and at the cell level

Oxygen Monitoring Components & Functions

Starting from the bottom (inside the unit), the commonly accepted practice has been to incorporate up to 3 sensors. I’m partial to sensors that use standard molex terminal connections. These are the most vulnerable part of the entire oxygen monitoring system, though also easily field serviceable which is an important feature for my style of diving.

 

 

 

The wire harness then passes through a bulkhead fitting on the rebreather ‘head’ which is logically the inhalation side of the unit such that the sensors are exposed to well homogenized breathing gas. A waterproofed cable then extends to a display directly (for a single display) or may be split via a sensor isolation board (to be discussed below). From there, out to one, two, or three displays.

Washdown gland on the RD1 Rebreather, allowing easy removal of the molex wiring harness from the head for service. The gland itself fits to a US garden hose for ease of washing the interior of the rebreather.

 

Today’s displays are robust, though are often battery powered, and can fail albeit infrequently. That means we strive for redundancy of displays, which is commonly accepted as standard. Critical to this redundancy is proper cell isolation circuitry which allows for each display to be unaffected by a failure of the other display. This isolation circuitry is typically in line between the oxygen sensors and displays. Common placement for isolation circuitry are y-connectors that solder a resistor in line with the wiring junction. These y-junctions are delicate and create a rat’s nest of wires within the rebreather head. A cleaner option is for a cell isolation board, with the isolation circuity fit to a PCB. This isolation circuitry permits failure (or even severing) of one display while leaving the remaining displays operational.

 

From the isolation board, output options provide for various combinations of off-the-shelf displays to be utilized. This is a nice feature given the evolving state of rebreather electronics. By using a mechanism to plug and play at a junction block or bottle external to the rebreather head, end-users can quickly configure the unit for mission specific requirements or make repairs in the field. Only simple wiring to the molex cells rests within the rebreather head section. 

Obviously, there are all kinds of logic to be employed here with up to three sensors in, and up to three displays out. It can become very unnecessarily complicated, and effort must be placed on the desired functionality, redundancies, vulnerabilities, and style of diving.  In all cases, the diver must be well trained and intimately familiar with the displays utilized. 

In my current configuration, up to three displays are utilized where:

1. Primary display and/or computer - provides main pO2 display and monitoring, as well as monitors and calculates decompression requirements when using an integrated computer.
2. Secondary display, or remote display – provides a sensor validation mechanism, an in-water ‘buddy’ pO2 readout, or data feed to surface via serial telemetry up the diver umbilical or data cable. 
3. HUD display – provides heads up display of each of the three sensors’ PO2 values, and primarily only used on task-loaded dives where looking at the wrist can be inconvenient when hands are occupied (such as during photography).

Creating display redundancy is still not a straightforward plug and play and can require some modest know-how in electrical wiring. It is indeed an acceptance bottleneck, though will quite likely remain a barrier to mitigate risks to manufacturers - if the diver makes a change (which most do), they are bearing a degree of responsibility.

Useful pin-outs and wire coloring schemes found in rebreather electronics. These should always be verified before making connections as variations have been found in the marketplace.

 

Oxygen Monitoring Must Be Truly Redundant

The commonly accepted practice of three oxygen cells utilized came from “voting logic” employed in early electronically controlled systems that require at least two cells to be in agreement on pO2 within an acceptable standard deviation in order to actuate a solenoid and inject oxygen into the breathing loop. Interestingly, this manner of thinking for eCCR systems is only a recent evolution (past 30-40 years) within the span of rebreather technology itself. The earliest of rebreathers didn't have electronics of any type, relying solely on an understanding of oxygen metabolism (for O2 units), and thereafter the physics of gas flows for SCR units. Indeed, modern electronics taking advantage of galvanic sensor technology have changed the game from a CCR standpoint. 

As mCCR has become more popularized, some of this three-cell logic has become challenged, as the ‘voting logic’ uses the diver’s brain rather than a computer to make an informed decision on adjusting the loop pO2, or not. These concepts should be reinforced in all rebreather training, even for eCCR, as it relates back to improved understandings of atmospheric management for all rebreather diving. 

A Means for Oxygen Sensor Verification Must Be Present

Now, of course, we can have all of the redundant displays we want to the Nth degree, but their value comes back to the proper functionality of the oxygen sensors themselves. So, we have three oxygen sensors, but what if one fails, or two fail, or all three fail?

Verifying the breathing gas composition via loop flushes, or capillary injections across sensor faces has been explored more carefully in recent years. If sensors can be routinely validated, do we really need three of them? This is not yet a clear answer to this, though I very firmly believe that sensor validation [verification] is a must-know skill and become as habitual as checking a pressure gauge.

A single cell placed with capillary for validation via diluent MAV injections.

 

This verification starts with calibration to a known value, typically pure oxygen at the start of a dive. Throughout the dive, there is value to the diver in verifying that the cells are a) responsive to changes in the atmosphere, and b) output voltages remain linearly responsive to oxygen throughout the dive.

A means of cell verification, in my opinion, is an absolute must. This can be achieved with simple diluent supply addition routed across the sensor faces via a MAV and periodically cross-checking with expected pO2’s for the given depth.

So, how many oxygen cells are needed in a rebreather? Well, only one if it’s working properly. The key we’ve pointed out is verifying its function, then adding redundancy, and then a degree of voting logic (be it automated or manual) to reduce errors in managing the atmosphere.

Calibration

The most opportune time to calibrate the electronics with pure oxygen is following the negative pressure/vacuum check at the surface during pre-dive. Once the vacuum is pulled, fill the breathing loop with pure oxygen one or several times and calibrate as needed.

Many display computers offer the option of reading cell voltage from the display. It is important to verify cell integrity in both air and oxygen, and to determine the linearity of cell voltages, which can degrade over the cell’s useful life.

The R22D oxygen cells commonly used should read about 10 mV in air. A five-fold increase of oxygen pressure (i.e. air fO2 to oxygen fO2; ~20% x 5 = 100%) should result in about 50 mV displayed. Conveniently, a simple gain of x2 applied to sensor mV will result in the corresponding pO2 value. This relationship is critical to recognize and can be put to work by the average diver during pre-dive and even during the dive when conducting in-water sensor verification checks.

Depth (ATA)	Depth (fsw)	mV (air)	pO2 (air)	mV (oxygen)	pO2 (oxygen)
1	0	10	.20	50	1.0
1.6	20	16	.32	80	1.6
2	33	20	.40	100	2.0
3	66	30	.60
4	99	40	.80
5	132	50	1.0
6	165	60	1.2
7	198	70	1.4
8	231	80	1.6

 

 

Oxygen sensor mV and expected pO2 linearity with depth as a guide for informed decision making during a dive event. Green highlights our pO2 target of 1.0 used throughout, red indicates dangerously high pO2. Yellow indicates self-check at 20fsw on descent/ascent.

Quite simply, as a diver, it is important to carry in the back of your mind, if not written on a slate, what the target pO2 will be if you conduct a diluent flush or an oxygen flush in the shallows. Likewise, for good measure, it is advantageous to know the anticipated cell mV output resulting from that gas flush and then cross-check with your display if mV’s can be displayed. If mV targets cannot be reached, there is a chance that the cell is current limited and should be further tested on the bench or replaced.

It is good practice to periodically test the cell voltage output in a pressurized chamber before diving, while taking the cells to 20 fsw. pO2 should read 1.6 bar with 100% oxygen. Alternatively, this sensor verification can be conducted at the start and finish of a dive at 20fsw by flushing the loop with oxygen. This practice is hotly debated and has its own limitations but is a good routine to implement as part of an S-drill.

Oxygen Sensor Golden Rules

As we strive to be rebreather diving dummies in making these things more commonplace, too much information is as equally dangerous to not having enough information. Gurr (2013) provides a set of oxygen sensor “Golden Rules” which are probably the best set of summary recommendations available.