In: Pollock NW, Sellers SH, Godfrey JM, eds. Rebreathers and Scientific Diving.

Proceedings of
NPS/NOAA/DAN/AAUS June 16-19, 2015 Workshop. Durham, NC; 2016.

Respiratory Physiology of Rebreather Diving

Gavin Anthony1, Simon J. Mitchell2*
1
QinetiQ, Gosport, Hampshire, United Kingdom
2
Department of Anaesthesiology, University of Auckland, Auckland, New Zealand
[email protected]
* corresponding author

Abstract

The use of rebreathers imposes a number of stresses on the respiratory system that frequently
provoke retention of carbon dioxide (CO2) during diving. The most important physiological
mechanism leading to CO2 retention is a derangement of the control of breathing which is usually
responsible for subconsciously adjusting lung ventilation to keep the arterial CO2 (PaCO2) at a
normal level. When the work of breathing increases during diving there is a tendency for this
breathing control system to become insensitive to rising PaCO2. An elevated PaCO2 can cause
unpleasant and dangerous symptoms, increase inert gas narcosis, and predispose to cerebral oxygen
toxicity. It follows that strategies to mitigate the risk of CO2 retention in rebreather diving are
important. These include minimising the work of breathing through appropriate rebreather design,
taking account of respired gas density when planning rebreather dives, minimising physical exertion
(particularly when deep), and meticulous attention to equipment preparation and adherence to best
practice guidelines for replacement of CO2 absorbent material.

Keywords: diving, rebreather, carbon dioxide, breathing, ventilation, hypercapnia

Introduction

The principal function of the lungs is to bring venous blood and gas in the lung alveoli into close
proximity so that carbon dioxide (CO2) in the blood may be exchanged for oxygen (O2) in the alveoli. In
healthy individuals the lungs are remarkably efficient at this task, and ventilation (the volume of gas
moved in and out of the alveoli per unit time) is 'automatically' controlled (see below) to maintain
adequate oxygenation (an arterial blood PO2 [PaO2] between 80 and 100 mm Hg) and normal CO2 levels
('normocapnia' – an arterial blood PCO2 [PaCO2] around 38±7.5 mm Hg [2SD]).

During diving the inspired PO2 is almost always elevated to planned and safe levels of 'hyperoxia'. Thus,
in the absence of equipment malfunction or diver error, hypoxia or symptomatic hyperoxia are
unexpected. In contrast, both immersion and the use of rebreathers (or other underwater breathing
apparatus) impose challenges to maintenance of normal respiratory control and CO2 homeostasis. As a
result, a PaCO2 higher than normal (hypercapnia) is frequently encountered in the absence of any error or
equipment related problem. This is important because hypercapnia can augment inert gas narcosis,
increase the risk of oxygen toxicity, and produce unpleasant symptoms such as shortness of breath,
confusion, anxiety, and ultimately unconsciousness.

This article will focus on the physiological mechanisms which may lead to hypercapnia during diving. It
will begin with a brief account of normal CO2 physiology. It will then examine the reasons why the work
of breathing may increase when a diver is immersed using rebreathers, and the physiological basis for this

66
to cause hypercapnia. Finally, it will examine the strategies divers may use to mitigate these physiological
challenges. With the target scientific diver audience in mind, the article is deliberately written in a
didactic style and does not assume detailed prior knowledge. It is not intended as a comprehensive
academic work on the subject. Such treatments can be found elsewhere (Doolette and Mitchell 2011).

Normal CO2 Physiology

Carbon dioxide is a by-product of metabolism of oxygen in cells. It is a volatile acid and will produce
unwanted biochemical derangements (and symptoms as mentioned above) if levels in the body are
allowed to increase. CO2 diffuses from tissues to venous blood and is carried to the lungs where it
diffuses from blood to alveoli and is breathed out. Maintenance of the diffusion gradient that drives this
process is entirely dependent on movement of fresh gas in and out of the lungs ('ventilation'). Thus,
greater ventilation will remove more CO2 from the alveoli, thus maintaining an increased partial pressure
gradient for CO2 diffusion from the venous blood. Conversely, less ventilation will remove less CO2 from
the alveoli and less CO2 will be removed from the blood. The crucial message here is that the amount of
CO2 eliminated from the body is directly proportional to ventilation. The relevant processes are depicted
in Figure 1.

Figure 1. Depiction of the process of production and elimination of CO2.

It can be deduced from Figure 1 that the production of CO2 in tissues and its removal by the lungs are
processes that must be balanced. If ventilation of the lungs is inadequate ('hypoventilation') CO2 levels
will increase, and if ventilation is excessive ('hyperventilation') then CO2 levels will decrease. The
process of balancing CO2 elimination by the lungs with production by the tissues is mediated through
control of ventilation by the respiratory controller in the brain stem. Although maintenance of adequate
oxygenation would seem intuitively more important than CO2 regulation, and although both hypoxia and
hypercapnia do provoke the respiratory controller to increase breathing, it is the PaCO2 that is widely
accepted as the primary effector. The respiratory controller indirectly monitors arterial CO2 levels through
sensing of the pH of the cerebrospinal fluid (which is directly influenced by PaCO2). If CO2 levels
increase then the respiratory controller will drive increased ventilation to remove more CO2 and vice
versa. The controller generally 'defends' a PaCO2 around 38 mm Hg ± 7.5 mm Hg (2SD) (5.1 kPa ± 1 kPa
[2SD]), though as will be seen below, this can be disturbed in diving. This is a substantial
oversimplification of a complex and incompletely understood process, but it is adequate for the purposes
of this discussion.

67
The most common derangement of this system during diving is that there may be inadequate ventilation
and an increase in PaCO2; a process often referred to as 'CO2 retention'. The obvious question is 'what
causes divers to hypoventilate thus allowing the PaCO2 to rise?' The answer is not simple or even fully
understood, but a significant contribution to the process occurs because of the increase in the work of
breathing that occurs during diving. Thus, in the following section we briefly consider the causes of
increased work of breathing in diving.

Causes of Increased Work of Breathing in Diving

There are multiple factors that increase the physical effort required to move gas in and out of the lungs
during diving.

Immersion effects

Immersion may cause changes in the mechanical properties of the lungs if the chest is exposed to a
different external pressure than the pressure inside the airways. For example, consider a diver upright in
the water using open-circuit scuba. The regulator supplies gas at a pressure equating to the ambient
pressure at the depth of the second stage (mouthpiece). Since the diver's airways are connected to this
regulator, the pressure inside the airways is therefore the same as the ambient pressure at the depth of the
mouth. The lungs themselves (remember the diver in this example is upright) are slightly deeper than the
mouth and they are therefore exposed to an external water pressure that is slightly higher than the
pressure inside the airways. This difference in pressure 'across' the lung (the pressure within the airways
being slightly less than the pressure on the outside of the lung) is called a 'negative static lung load' or
'hydrostatic imbalance'. The relative negative pressure inside the lung airways encourages blood to
engorge the relatively distensible lung blood vessels, and this renders the lung stiffer than normal. Put
another way, the lung's compliance is reduced meaning that more muscular effort would be needed to
move the same amount of gas in and out. A negative static lung load also exists when a rebreather diver
with a back-mounted counterlung is swimming in a horizontal position. In this setting, the airways are in
continuity with (and contain gas at the same pressure as) the counterlung, which is sitting at a slightly
shallower depth (and lower pressure) than the lungs themselves.

Static lung loads can vary according to the type of equipment (open- or closed-circuit), the position of the
counterlung in the latter, and the orientation of the diver in the water. It is beyond the scope of this article
to discuss the various combinations of circumstances that may arise. Suffice it to say that under some
commonly encountered circumstances, static lung loads (and particularly negative static lung loads) can
increase the work of breathing during diving as described.

Equipment-related resistance

The use of underwater breathing apparatus imposes an external resistance to breathing. It is intuitively
apparent that this would be potentially important in a rebreather. In using a rebreather all of the energy
required to propel gas through the hoses, various connectors, and the CO2 scrubber, must be provided by
the diver's own effort. In this regard, the design of the rebreather (and in particular considerations like the
geometry of the gas flow path, diameter of hoses, and type of CO2 absorbent canister) can make a
substantial difference to the work of breathing. Not surprisingly, there are recommended standards for
maximum work of in underwater breathing apparatus. Relevant standards and testing of rebreathers in this
regard are discussed in more detail by Anthony (2009).

68
Gas density

One of the most important influences on work of breathing in diving is the increase in density of respired
gas that occurs as depth increases. Since any underwater breathing apparatus will supply gas at ambient
pressure, the density of the respired gas increases in direct proportion to depth. Increases in gas density
result in a parallel increase in the resistance to flow of the gas through the diver's own airways, and in
rebreather diving there is also the extra effort of moving dense gas through the hoses, connectors and CO2
scrubber of the unit. Under these circumstances, the associated increase in the work of breathing can be
substantial.

Another relevant phenomenon profoundly affected by gas density is a reduction in the maximal
ventilation that can be achieved even when a diver is consciously attempting to move as much gas as
possible in and out of the lungs. For example, in dry chamber experiments it has been shown that the
maximum amount of air a subject can move in and out of the lungs in one minute is approximately halved
(compared to the surface) at 100 ft (30 m, 4.0 ATA) (Camporesi and Bosco 2003).

This 'ceiling' on ventilation performance appears related to the physiological phenomenon known as
'dynamic airway compression', and it is explained as follows. During maximal breathing effort, the
muscles of the chest wall and diaphragm create a positive pressure inside the chest in order to force gas
out of the alveoli and outward through the airways as quickly as possible. However, as gas passes out
along the airway, the pressure inside airway falls due to frictional forces of the gas on the airway walls.
At some point during a forced exhalation this pressure drop inside the airway is sufficient that the raised
pressure inside the chest exceeds the pressure in the airway, and the airway starts collapsing. This limits
the outward gas flow through the airway, and this restriction on outward flow then becomes the limiting
factor in how much gas can be moved in and out of the lungs each minute.

This actually occurs in air breathing at 1.0 ATA, but the limitation begins at such high flow rates that is
does not significantly hamper work performance (except perhaps in extreme exercise). However, when
breathing a dense gas underwater the resistance to flow is much higher and a significant pressure drop
inside the airway as gas flows outwards occurs at much lower flow rates. Thus, the airway will begin to
collapse at low flow rates, and this limits breathing to a much greater extent than seen during air breathing
at 1.0 ATA. Indeed, it has been shown that if extremely dense gas is breathed, a diver might not be
capable of moving much more gas in and out of their lungs than during normal breathing sitting at rest
(Wood and Bryan 1969). Such situations would be unlikely to be encountered in properly planned dives,
but it is possible (see below). A more detailed and illustrated explanation of this phenomenon can be
found in the DAN Technical Diving Workshop Proceedings (Mitchell 2009).

Physiological Mechanisms of Hypercapnia in Diving

Having briefly considered the causes of increased work of breathing in diving, the discussion moves on to
an explanation of how this increase in work may result in hypercapnia.

With reference to the earlier discussion of control of ventilation, it would be expected that if the PaCO2
began to rise (for example, when a diver starts to exercise and produces more CO2), then the respiratory
controller would automatically increase ventilation in order to remove more CO2 and bring the PaCO2
back to normal levels. This is indeed the classically described ventilation response in experiments using
very low resistance breathing equipment where the PaCO2 is forced to rise by introducing CO2 to the
inhaled gas so that no matter how much the subject breathes, they cannot return the CO2 to normal. Under
these circumstances, curves plotting end-tidal CO2 (an indirect measurement of PaCO2) and ventilation
typically show an approximately direct linear relationship.

69
However, there is some degree of inter-individual variability in the ventilation response to rising CO2, and
this can be markedly exaggerated if there is an unusual increase in the work required to increase
ventilation (as is the case in diving). Under these circumstances, it is as though the respiratory controller
is confronted with a choice: Either to perform the extra work required to maintain a normal PaCO2, or to
avoid the extra work and allow the PaCO2 to rise. There appears to be a spread of individual responses
between these two extremes. This is illustrated in Figure 2 which shows end-tidal CO2 vs ventilation
'curves' for 15 subjects who were breathing on a rebreather circuit with no CO2 scrubber in place (Deng et
al. 2015). In this setting there was no removal of exhaled CO2 and consequently there was substantial CO2
rebreathing. The arterial CO2 was forced to rise no matter how hard the subjects breathed. It is also
contextually important that the diving rebreather used in this experiment imposed an increase in the work
of breathing that was greater than normal. The Figure 2 'curves' are lines interpolated between points
plotted from measurements of end-tidal CO2 and ventilation made 30 s after starting to breathe and on
termination of breathing on the circuit. All of these subjects voluntarily terminated breathing within a
five-minute period citing 'shortness of breath' among their symptoms. The remarkable feature of the data
is the variability in individual ventilation responses. Some subjects did not increase ventilation at all
(indeed in some it actually decreased) whereas others exhibited a more classical linear increase in
ventilation as the end-tidal CO2 rose.

The implication of these data (and those of others that have examined the underlying mechanisms in more
detail (Poon 1987; 1989)) is that in some divers at least, there is a tendency for the respiratory controller
to prioritize the avoidance of respiratory work over maintaining the PaCO2 at normal levels when the
work of breathing is increased. Put another way, during diving an increase in the work of breathing may
provoke a 'naturally occurring flaw' in control of breathing such that the PaCO2 may rise simply because
the diver does not breathe enough to eliminate the CO2 that they are producing. This is the most plausible
mechanism for the frequent finding of CO2 retention in divers using underwater breathing apparatus
(UBA), especially during exercise.

A second physiological mechanism for hypercapnia during diving relates to the potential for respiratory
limitation by dynamic airway compression described above. It is plausible that if sufficiently dense gas
was breathed a diver could find themselves in a situation where they would be unable to ventilate
sufficiently to maintain a normal PaCO2 even at minimal levels of exercise, and even if they tried hard to
do so. The principle of this mechanism is illustrated in Figure 3.

70
Figure 2. Indicative end-tidal CO2 – ventilation curves for subjects breathing on a rebreather circuit with
no CO2 scrubber. Breathing was voluntarily terminated by the subjects when they developed symptoms
of CO2 toxicity (including a perception of shortness of breath in all cases). Each subject is represented
by a straight line linking two paired measurements of end-tidal CO2 and ventilation: the first made at
30 s after starting to breathe on the circuit, and the second on voluntary termination of the breathing
period. PETCO2 (end-tidal CO2) is a conveniently measured approximation of the PaCO2.
Reproduced with permission from Deng et al. (2015).

Figure 3. Notional depiction of the relationship between maximum possible ventilation and increasing
depth and respired gas density. If the maximum possible ventilation falls below the ventilation
required to eliminate the CO2 produced (and therefore to maintain a normal PaCO2) at a given
level of exercise, then the PaCO2 must inevitably increase. See text for further explanation.

In reference to Figure 3, the amount of ventilation (gas movement in and out of the lungs) required to
keep the PaCO2 normal at a given level of exercise (nominally to swim at 0.5 knots) does not change as
depth increases. However, as depth and the respired gas density increase, the maximum ventilation that
can be achieved decreases because of the onset of dynamic airway compression at progressively lower
flow rates through the airways. If the diver progresses deeper than a depth where they can produce the
ventilation required to keep the PaCO2 normal, then the PaCO2 must inevitably rise. To make matters
worse, the rising PaCO2 may trigger increased breathing effort which will only serve to produce more CO2
because once dynamic airway compression occurs, no amount of extra effort will improve ventilation
volumes. There is one published event in which there is reasonable supporting evidence for involvement
of this mechanism, which occurred on a rebreather dive to a depth of 265 m (869 ft) (Mitchell et al.
2007).

For completeness, we observe that 'non-physiological' problems related to equipment (such as an absent,
incorrectly installed or expired scrubber canister, or malfunctioning one way valves in the rebreather
mouthpiece) are also potential causes of hypercapnia during rebreather diving. All of these result in some
degree of CO2 rebreathing, and if CO2 is inhaled the diffusion gradient for elimination of CO2 from
venous blood to lung alveoli is diminished. If a large amount of CO2 is rebreathed this can lead to a
catastrophic impairment of CO2 elimination with rapid development of symptoms of hypercapnia, but
even a relatively small amount of inhaled CO2 is potentially problematic because the associated
impairment of CO2 elimination will compound the physiological predispositions to hypercapnia described
above.

Mitigation of the Risk of Hypercapnia During Rebreather Diving

71
At a practical level, the most important question arising from this discussion is 'what steps can be taken to
mitigate the risk of hypercapnia during rebreather diving'? There are several possibilities.

Manipulation of static lung load

There is some evidence that a negative static lung load is the least desirable condition from a
physiological perspective in rebreather diving, and that mildly positive static lung loads are best tolerated
during hard work underwater (Thalmann et al. 1979). In a horizontal diver these conditions would be
produced by back and front-mounted counterlungs respectively. However, choosing a counterlung
configuration based primarily on concerns about static lung load may be ill-advised because the lung load
will vary according to the diver's orientation in the water. For example, while a back-mounted
counterlung would produce a negative static lung load in the horizontal position, it would be largely
neutral in the upright position. In theory, over the shoulder counterlungs should produce the least extreme
and least variable static lung load, but they have their own set of disadvantages such as cluttering the
space around the diver's front and head.

Minimising equipment-related breathing resistance

All underwater breathing apparatus, including rebreathers, should be designed with the goal of reducing
their external breathing resistance as much as is practicable. Other than choosing a device with good
related design and testing characteristics there is little that divers can do in this regard. However, on a
cautionary note, divers should take great care with making any modifications to a rebreather that might
alter the geometry or resistance of the gas flow path. Common examples include departures from
manufacturer-recommended grade of CO2 absorbent material, the incorporation of extra oxygen cells for
independent PO2 monitoring, changing mouthpiece configuration, and changing the composition of any
moisture pad material.

Consideration of gas density in diving planning

Most rebreather divers are very familiar with specialised dive planning strategies like calculating a
maximum operating depth for a gas in order to avoid an unsafe inspired PO2, or calculating an equivalent
narcotic depth in planning the helium content of trimix to avoid unacceptable levels of nitrogen narcosis
(Mitchell and Doolette 2013). In contrast, one almost universally overlooked dive planning strategy
related to work of breathing is the use of gas density calculations to avoid breathing gases with
unacceptably high density at depth. In no small part this situation prevails because there have been no
definitive guidelines on acceptable gas density in diving.

There is a paucity of related data, though a recent analysis of a dataset of human testing records for UBA
provides some potentially valuable insights upon which some preliminary guidelines can be based.
Among other things, QinetiQ is a UBA testing house located near Portsmouth in the UK. Over some 20
years hundreds of manned test dives have been undertaken utilising ethics committee approved protocols
which incorporate graded levels of underwater work for evaluating performance of a range open-circuit,
semi-closed, and closed-circuit UBA. These dives have been conducted over depths ranging from 4 to 80
m (13 to 262 ft), using a range of gases including oxygen, air, nitrox and heliox. Throughout these tests a
standard set of endpoints have been used to define 'dive failure' including: (any of) equipment or
monitoring failure, diver unable or unwilling to continue because of dyspnoea (shortness of breath) or
exhaustion, and an end-tidal CO2 >8.5 kPa (64 mm Hg) over five consecutive breaths. The latter is
indicative of significant CO2 retention to a level associated with sudden incapacitation in the diving
setting (Warkander et al. 1990).

72
Although this program of testing was not designed to specifically answer questions about tolerable gas
density, the wide range of gas densities that were incidentally used has facilitated an evaluation of the
proportion of work-loaded rebreather dive failures due to end-tidal CO2 >8.5kPa stratified according to
the gas density breathed. These data are reported in Figure 4. With the dual caveats that the trials were not
specifically designed to answer this question and that the number of dives at the higher densities is
comparatively small, there is a clear signal that near a respired gas density of 6.0 g·L-1 there is an upward
inflection in the risk of dangerous CO2 retention during working rebreather dives. A similar analysis of
dive failures in open-circuit underwater breathing apparatus trials produced a virtually identical result.

Figure 4. The proportion of rebreather test dives ending in failure due to an end-tidal CO2 >8.5 kPa
(black) and other causes of failure (dark grey) stratified by respired gas density. Figures refer to
numbers of dives. At respired gas densities >6 g·L-1 there is a sharp increase in the risk of dive
failure, with most failures being caused by dangerous levels of CO2 retention.

For the purposes of planning rebreather dives and in the current absence of more definitive or
contradictory data, it seems prudent to recommend an ideal maximum gas density of 5.2 g·L-1 (equivalent
to air diving at 31 m [102 ft]) and an absolute maximum of 6.2 g·L-1 (equivalent to air diving at 39 m
[128 ft]). Implementation of such a recommendation will require an appreciation of how to calculate gas
density for a given respired gas at a given depth. Such calculations begin with knowledge of the density
of air and the individual components of gas mixes at 1.0 ATA (Table 1).

Gas Density (g·L-1)

Hydrogen 0.090
Helium 0.179
Nitrogen 1.251
Oxygen 1.428
Air 1.293

Table 1. Gas density in g·L-1 for common diluent gases, oxygen and air
at 1.0 ATA. Data from Doolette and Mitchell (2011).

73
Calculation of the density of air at depth is a simple process of multiplying its density at 1.0 ATA by the
ambient pressure at the target depth. For example, the density of air at 30 m (99 ft) is given by 1.293 g·L-1
x 4.0 ATA = 5.17 g·L-1.

Calculation of density for a mixed gas is achieved by using simple proportions to calculate the density of
each component at 1.0 ATA, summing the components, and multiplying this sum by the ambient pressure
in ATA at the target depth. For example, consider trimix 16:50 (16% oxygen, 50% helium, 34% nitrogen)
intended for use at 70 m (230 ft) where the ambient pressure is 8.0 ATA. Calculating density for each
component at 1.0 ATA we use the fraction of gas x its density at 1.0 ATA, thus, substituting in values
from Table 1:

0.16 x density of oxygen (1.428) = 0.23 g·L-1

0.50 x density of helium (0.179) = 0.09 g·L-1
0.34 x density of nitrogen (1.251) = 0.43 g·L-1

The sum of the products of these calculations is 0.75 g·L-1 for density at 1.0 ATA. If this is then
multiplied by 8.0 ATA for the ambient pressure at the planned depth we get 6.0 g·L-1. Therefore, in
respect of gas density this would be an acceptable (but less than ideal) mix at this depth.

Moderating expectations of work capacity at depth

Unsurprisingly (given the above discussion) it is widely recognised among experienced divers that as
depth increases there should be a corresponding moderation of expectation of work capacity. Hard work
(with an inevitable increase in CO2 production) is best avoided on a rebreather at any time, but this is
particularly so at increased deep depths where the respired gas density is likely to be trending toward (or
exceeding) the ideal limit. There are many practical strategies which help with reducing work at depth
including exhibition of basic dive skills (such as maintenance of good buoyancy control and good
trim/streamlining in the water), intelligent task planning, and the use of assistive technology such as diver
propulsion vehicles. However, the use of such strategies is not a substitute for minimising the work of
breathing in a UBA and strategic planning of gas density because events such as an emergency situation
requiring extra work, or failure of a diver propulsion vehicle can occur unexpectedly.

Detection of CO2 rebreathing

We earlier acknowledged the potential for CO2 rebreathing to be caused by an absent, incorrectly installed
or expired scrubber canister, or by malfunctioning one way valves in the rebreather mouthpiece. The
strategies to prevent and detect such problems are issues of rebreather diving technology and practice
rather than physiology. Nevertheless, for completeness, we will briefly discuss them there.

The cornerstone of preventing CO2 rebreathing during use of a rebreather is meticulous adherence to
manufacturer guidelines on both CO2 absorbent duration and preparation of the rebreather before diving.
Function of the mushroom valves in the rebreather mouthpiece should be checked every time the
rebreather is assembled and the unit should not be used if the valves appear to be leaking. Great care must
be taken with packing absorbent into the CO2 scrubber canister to ensure that subsequent settling of the
material does not result in a loose pack and channelling of gas through pathways of low material density.
Similarly, the scrubber canister must be carefully installed in the rebreather avoiding any error that might
result in gas bypassing the canister. Various rebreathers have easily avoidable but known vulnerabilities
in this regard, and users must be aware of these.

As a final check of these good practices, rebreather divers are taught to conduct a five minute 'prebreathe'
of the unit prior to entering the water. The prebreathe has multiple goals, but one of them (and the one

74
upon which the five-minute duration is predicated) is the detection of symptoms of CO2 toxicity should
there be any error in preparation or assembly that allows rebreathing of CO2. The efficacy of this strategy
was recently tested in a randomised single blind study in which divers prebreathed a rebreather which
either had a normal scrubber, a completely absent scrubber, or a partial failure of the scrubber allowing
bypass of a significant amount of CO2. The subjects were asked to terminate the prebreathe as they would
in the real world if they developed symptoms of CO2 toxicity. Twenty trials were undertaken in each
condition. As expected, no diver terminated the prebreathe when breathing on a circuit with a normal
scrubber. However, only 10% (2/20) were able to detect symptoms (and thus terminated) in the partial
failure condition despite an inspired PCO2 of 20 mm Hg. A much higher proportion (75%) detected the
complete absence of the scrubber, but remarkably, 25% did not despite developing an end-tidal CO2
greater than 60 mm Hg. Thus, it was concluded that while a prebreathe is a vital part of evaluating a
rebreather before diving (for example, to check that the oxygen addition system is functioning), it cannot
be relied upon to reveal problems with the CO2 scrubbing function of the unit. Based on reports of a
stressed or breathless appearance of some subjects who did not terminate during CO2 rebreathing in the
Deng et al. (2015) study, it could be concluded that peer observation of the prebreathe might improve its
sensitivity. Whilst such a strategy might be applied successfully in a disciplined military or scientific
diving setting, it is unlikely to be considered practical or executed diligently in 'mainstream' technical
diving.

CO2 sensors placed on the inhale limb of the rebreather circuit (downstream of the CO2 scrubber) are a
relatively recent innovation that represent a potential solution to the problem of detecting CO2 bypassing
the scrubber. These are only available on a small number of rebreathers at this time and there is limited
experience with their use in the field. It is too early to make definitive recommendations on their use.

References

Anthony TG, Diving re-breathing apparatus testing and standards UK/EU perspective. In: Vann RD, Mitchell SJ,
Denoble PJ, Anthony TG, eds. Technical Diving Conference Proceedings. Durham, NC: Divers Alert Network;
2009, pg. 218-36.

Camporesi EM, Bosco G. Ventilation, gas exchange, and exercise under pressure. In: Brubakk AO, Neuman TS,
eds. Bennett and Elliott's Physiology and Medicine of Diving, 5th ed. Edinburgh, UK: Saunders, 2003: 77-114.

Deng C, Pollock NW, Gant N, Hannam JA, Dooley A, Mesley P, Mitchell SJ. The five minute prebreathe in
evaluating carbon dioxide absorption in a closed-circuit rebreather: a randomised single-blind study. Diving
Hyperbaric Med. 2015; 45: 16-24.

Doolette DJ, Mitchell SJ. Hyperbaric conditions. Comprehensive Physiol. 2011; 1: 163-201.

Mitchell SJ, Cronje F, Meintjes WAJ, Britz HC. Fatal respiratory failure during a technical rebreather dive at
extreme pressure. Aviat Space Environ Med. 2007; 78: 81-6.

Mitchell SJ. Respiratory issues in technical diving. In Vann RD, Mitchell SJ, Denoble PJ, Anthony TG, eds.
Technical Diving Conference Proceedings. Durham, NC: Divers Alert Network; 2009, pg. 12-37.

Mitchell SJ, Doolette DJ. Recreational technical diving part 1. An introduction to technical diving. Diving
Hyperbaric Med. 2013; 43: 86-93.

Poon CS. Ventilatory control in hypercapnia and exercise: Optimization hypothesis. J Appl Physiol.1987; 62: 2447-
59.

75
Poon CS. Effects of inspiratory resistive load on respiratory control in hypercapnia and exercise. J Appl Physiol.
1989; 66: 2391-99.

Thalmann ED, Sponholtz DK, Lundgren GEG. Effects of immersion and static lung loading on submerged exercise
at depth. Undersea Biomed Res. 1979; 6: 259-90.

Warkander DE, Norfleet WT, Nagasawa GK, Lundgren CE. CO2 retention with minimal symptoms but severe
dysfunction during wet simulated dives to 6.8 atm abs. Undersea Biomed Res. 1990; 17: 515-23.

Wood LDH, Bryan AC. Effect of increased ambient pressure on flow volume curve of the lung. J Appl Physiol.
1969; 27: 4-8.

QUESTIONS AND DISCUSSION

AUDIENCE MEMBER: What is that depth again?

SIMON MITCHELL: It is not a depth. It is a density. So there are multiple ways you could achieve a
particular gas density. You could have a light gas at a very deep depth or a heavier gas at a shallow depth.
Irrespective of the depth the important thing is the gas density. Depth (or more correctly ambient
pressure) is relevant in that it determines density for a given respired gas, but it is not relevant to these
results.

JEFF BOZANIC: Did you measure PCO2 during the tests for all your subjects?

SIMON MITCHELL: Yes.

JEFF BOZANIC: What was the highest level you were able to measure?

SIMON MITCHELL: The normal level of inhaled CO2 is zero. It went up to over 50 mm Hg. By the end
of a five-minute prebreathe with no scrubber our subjects had an end-tidal CO2 of around 8.5 kPa, over 60
mm Hg, who had no idea (Deng et al. 2015). Unbelievable, is it not? Even with substantial ventilation
increases a lot of these divers did not know. In terms of the ventilation increase, the tidal volume
increased more than the rate.

DAVE CONLIN: Two questions. One, the separation between your counterlungs and your actual lungs,
do you get to a point at depth where that difference in pressure because it is a small difference in the
overall pressure, that your body is physiologically experiencing, is inconsequential? If you are at 33 ft (10
m), 25 cm is a lot different in pressure.

SIMON MITCHELL: I see, you are asking is there context with depth in determining a static lung load?
No -- 20 cm of water is always 20 cm of water, no matter where you are in the water column.

DAVE CONLIN: My other question is if your body is a CO2 retainer and your body does not care about
the CO2 it is retaining, so to speak, what is the problem?

SIMON MITCHELL: That is a good question. I did not think to mention it because high CO2 becomes a
problem in everyone once it reaches a certain point. The trouble is that CO2 retainers do not see it coming.
Dan Warkander and Barbara Shykoff at NEDU have done a lot of work on this. What they see with CO2
retainers is that CO2 creeps up during exercise at depth with an elevated breathing resistance. The divers
are peddling away underwater indicating "I am fine." Then, all of a sudden, they either pass out or stop
responding to hand signals. They pass a threshold and the lights go out. So, eventually it happens to

76
everyone. The non-CO2 retainers get more symptoms as the CO2 levels go up. They sense it and are more
aware of the problem, and under normal circumstances would be more likely to stop what they are doing
and rest to prevent the CO2 rising any further. That is the trouble with CO2 retention. The diver may be
unaware of it so does not respond in the way they should, which is to stop and rest and get their breath
back.

DAVE CONLIN: One final comment about the prebreathe. If you have a temp stick, you are getting a lot
of secondary information, not physiologically based, that shows a value of doing a prebreathe, seeing that
your stacks warming normally, and that your PO2 is dropping and rebounding.

SIMON MITCHELL: Agreed. If it takes five minutes to activate your temp stick and that is what you like
to see, then you should do a five-minute prebreathe. But our message is that divers should not kid
themselves that completing the prebreathe is verifying that they have perfectly functioning scrubber.

MARK KEUSENKOTHEN: So once a CO2 retainer, always a CO2 retainer?

SIMON MITCHELL: That is a great question. There is good news and bad news in that category. There
is some evidence that diving actually teaches you to be a CO2 retainer. None of that data actually applies
to rebreather divers specifically. And I would not be surprised if it is less of an issue with rebreather
divers. The theory is that it is the breathing discipline that you impose on yourself when you are an open-
circuit diver to try to conserve gas which teaches you to be a CO2 retainer. There is some evidence in that
regard. There is also some evidence that you can train yourself not to be one. There is one paper
published by a group at Buffalo which showed that respiratory muscle training can actually lower the
tendency to retain CO2. So, there is some evidence, if preliminary, in both directions.

Pendergast DR, Lindholm P, Wylegala J, Warkander D, Lundgren CE. Effects of respiratory muscle
training on respiratory CO2 sensitivity in scuba divers. Undersea Hyperb Med. 2006; 33(6): 447-53.

DAVE PENCE: In terms of correct mix gas selection for deep rebreather driving, you said that rather than
actually having to do the actual gas density calculations, it is adequate to simply select a mixture that
gives you an appropriate equivalent narcotic depth?

SIMON MITCHELL: No, the opposite actually. I said that by looking after your equivalent narcotic
depth, you do not automatically end up with a gas of sufficiently low density. You may need more helium
for gas density purposes than you need for equivalent narcotic purposes. It should be part of dive
planning; make sure your PO2 and your END are going to be okay. Then when you think you have chosen
your gas, calculate the gas density at the target depth and make sure that is okay as well. If it is not, bump
the helium up appropriately.

DAVE PENCE: They have a practice among some rebreather divers simply to use Heliair as the diluent,
which always leaves you with an oxygen to nitrogen ratio of 1 to 4. Does that tie equivalent narcotic
depth closer to breathing resistance than actually mixing.

SIMON MITCHELL: There will be a sweet spot somewhere where it does, but it will not apply across the
entire depth range. I think it is an individual calculation that probably should be done for most dives. Over
time you will start to get used to what works at what different depths.

JOHN BRIGHT: Is there any information that CO2 retention will increase your sensitivity to other inert
gas narcosis, i.e., you increase your CO2 retention and the threshold at which PN2 and PO2 will have an
additive narcotic effect decreases?

77
SIMON MITCHELL: Unfortunately, CO2 retention is synergistic with nitrogen narcosis. There is enough
evidence to believe that there is a synergy between the two; that is they are actually worse than the sum of
their parts. CO2 retention is bad for narcosis. CO2 is a very narcotic gas. There was even a proposal at one
stage that CO2 could be used as an aesthetic agent because if you breathe enough of it, it will render you
unconscious. It never caught on. But certainly it is worse for narcosis. The other thing that you mentioned
is the interaction between CO2 retention and oxygen toxicity. This interaction probably accounts for the
higher risk of oxygen toxicity when divers are in the water compared to sitting in a chamber breathing
oxygen through a low resistance system. In the water we all probably retain a little bit of CO2, especially
if we are working hard or our breathing resistance is high. For a given PO2 this increases the risk of
oxygen toxicity because CO2 dilates the cerebral blood vessels. So blood flow to the brain is much higher
when CO2 is elevated, and this increases oxygen delivery and thus the oxygen tension in the brain. There
are very good experimental data demonstrating this.

Lambertsen CJ, Ewing JH, Kough RH, Gould R, Stroud MW. Oxygen toxicity: arterial and internal jugular
blood gas composition in man during inhalation of air, 100% O2 and 2% CO2 in O2 at 3.5 atmospheres
ambient pressure. J Appl Physiol. 1955; 8: 255-63.

So, CO2 retention is a risk factor for narcosis and a risk factor for oxygen toxicity. It is an important gas
that people do not understand well enough. Some of what I have been explaining is quite complex, but it
is an important gas to understand because it has far-reaching effects in a lot of what we do.

KARL HUGGINS: In the five-minute prebreathe with the failed loop is the CO2 retention enough to
change physiological pH, say, in the saliva where it is something that could be measured by litmus or
some other type of pH detector?

SIMON MITCHELL: Good thought. If you did an arterial blood gas on most people when they had an
end-tidal CO2 of 64 mm Hg, they will all have a respiratory acidosis. Whether in a five-minute period that
translates into a change in saliva pH is another thing entirely. Intuitively I would say probably not, but I
would not bet my house on it. It is a good thought. It is easy to test exactly the way you suggested it.

PHIL SHORT: Could the condition and the same thing happen with open-circuit?

SIMON MITCHELL: CO2 retention can definitely happen on open-circuit (but obviously not because of
CO2 rebreathing as in a rebreather). CO2 retention can occur if the breathing resistance is increased for
any reason, especially if combined with exercise. This may be less likely on open-circuit because the
breathing resistance is often lower on a good open-circuit set. One of the things I did not share is one of
the other disadvantages of a negative static lung load which can also occur on open-circuit. If you have a
negative pressure in your airways, it increases the risk of dynamic airway compression. This is because
falling pressure along the airway is more likely to promote airway collapse if you have got a negative
static lung load and there a relatively negative pressure in the airway. Returning to rebreathers,
theoretically, if a diver with a back-mounted counterlung flipped over onto his back to create a positive
static lung load might help splint the airways open. Emphysema patients get the same problem at
atmospheric pressure because they have got so little elasticity in their lungs. You watch them breathing
and see that they create a back pressure in the airway by exhaling through pursed lips so it stops the
airways collapsing. This is why a positive static lung load is slightly better for hard exercise because the
positive pressure splints the airway open. I am digressing; CO2 retention on open-circuit could certainly
happen if you breathe a gas dense enough that it significantly increases work of breathing or you get
dynamic airway compression which limits lung ventilation.

RICHARD PYLE: You referred generically to more breathing. What is the difference between increasing
tidal volume versus increasing respiratory rate versus increasing both to maximally achieve that?

78
SIMON MITCHELL: The answer is that minute volume, that is, the product of tidal volume and
respiratory rate, makes a difference in removing CO2. The only caveat is that in diving we have an extra
little bit of deadspace in our mouthpiece. The average rebreather mouthpiece has 50-80 mL of increased
deadspace, which brings our total deadspace to just over 200 mL. So if you really want to increase
ventilation of your alveoli, you need to increase tidal volume more than rate. A very high number of tiny
breaths will just ventilate the deadspace.

RICHARD PYLE: Even if you just count the extra deadspace from diving equipment, it will lean you
toward tidal volume?

SIMON MITCHELL: Absolutely right.

RICHARD PYLE: So I guess both.

SIMON MITCHELL: Both are important.

79