Professional Documents
Culture Documents
A Primer For Decompression Theory
A Primer For Decompression Theory
TECH
TECHNICAL DIVING
DECOMPRESSION
PHILIPPINES THEORY
andy@scubatechphilippines.com
History .............................................................................................................................................................. 15
M-Values .......................................................................................................................................................... 19
Consistency of M-Values................................................................................................................................. 37
Let’s start off with a disclaimer and an outline of some assumptions made
while working on this blog post.
First off, I did not start out intending to write a definitive piece on
decompression theory or on the stellar work of Professor Albert Bühlmann.
Also, this contains no detailed explanation of the internal workings of the
maths behind a decompression algorithm and the challenges it meets while
trying to model human physiology. Simply put, this was written to help the
average punter better understand what gradient factors describe; and the
potential impact of playing around with the GF settings on your personal dive
computer (for example, a Shearwater Petrel).
Secondly, I’ve made several assumptions… not the least of which is that
readers have a basic understanding what happens to an individual after
spending more than a couple of minutes sub-surface breathing compressed
gas. (That’s just another way of saying that I’m writing for certified and
reasonably experienced divers who understand that diving can result in
decompression stress of one flavor or another.)
All dives are decompression dives, ergo all divers are decompression divers.
Our acknowledgement of this and the depth of that acknowledgement can
greatly influence our behavior when we dive: most specifically, our ascent
behavior. How deep we dive, what gases we breathe, and how long we spend
at depth vary considerably from dive to dive; but all dives share one common
threat for individual divers: the risk of getting bent.
5|P ag e ©Andy Davis 2015 www.scubatechphilippines.com
Smart divers consider many factors when planning dives, many of which help
to alleviate that risk. One is choosing to dive a “conservative profile.” I guess
there are many interpretations of what that means… but for our purposes
here, it means choosing a decompression algorithm that’s proven, and
choosing a setting for that algorithm that will generate stress-tolerant dive
tables.
Yet Bühlmann tables remain popular. They are in fact perfectly functional,
and are helping to keep tens of thousands of divers safe from the bends
every week. The secret is that with very simple tweaking, a Bühlmann
schedule can be made to follow a time-and-place curve very similar to those
produced by bubble models slowing ascent and beginning staged stops
deeper in the water column.
Clearly, Now that we have some concept of the relationship between variable
times and various group saturations, let’s look at M-values.
One hundred percent is the point where M-Value is on the verge of critical
bubbling (1), and zero percent is the same as ambient pressure (M-Value of 0
where there is no force driving gas out of solution at all). Therefore, effective
decompression can be found somewhere between those two points.
When one uses GFs to set the “conservatism” for one’s decompression, one
uses two numbers: these represent the Low Gradient Factor and the High
Gradient Factor.
In simply terms, the Low Gradient Factor (LGF) defines how deeply in the
water column one takes one’s first decompression stop. The High Gradient
Factor (HGF) defines how close to the 100 percent M-Value one surfaces with
at the end of the dive.
This setting promotes off-gassing but stops one’s ascent deep in the water
column but above the gas transition point (the theoretical spot on the water
column at which more gas is eliminated than is taken up according to the
algorithm).
Most of all, be aware, nobody and no organization can predict precisely the
outcome of any dive!
Experience has shown that bubbles can form and decompression sickness can
develop when supersaturation is within established M-value limits. This
indicates the imperfection of simple limits. In response, many dive computers
offer some degree of user-selectable conservatism. Conservatism can be
important for both physical safety and psychological comfort. The person
who believes himself or herself to be bends-resistant may be happy with
settings that would not be comfortable for a person less confident about
being bulletproof or one simply wanting additional peace of mind.
Gradient factors are typically applied in two steps. The first number (GFlow)
represents that percentage of the M-value used to determine the first stop
during ascent. The second number (GFhigh) is the percentage of the M-value
that cannot be exceeded at any point during surfacing. The dive computer
effectively draws a straight line between the two, creating the ascent slope.
The alteration of limits is applicable to both technical and non-technical
exposures.
Gradient factors adjust the limits displayed on the computer screen. Low
GFlow values (e.g., ≤20) are for those who believe in deep stops. Higher
GFlow values (e.g., ≥30) get divers off the bottom to reduce inert gas uptake
in tissues that are not saturated. Low GFhigh values (e.g., ≤70) limit the
maximal decompression stress experienced on the dive. High GFhigh values
(e.g., ≥80) get divers out of the water quickly.
One practical note is that adding extra shallow stop time is not the same as
reducing GFhigh. Extra shallow stop time will reduce the effective GF at the
point of surfacing, but higher effective GF values could have been reached
earlier during ascent. The impact of reducing GFhigh is greatest near the
surface, but it will moderate the ascent profile throughout. Extra shallow
stop time provides additional protection and is well worthwhile if time, gas,
and conditions allow.
The thoughtful and well-informed diver knows that conditions during a dive
affect real-time risk. Having a dive computer that can help manage exposures
to keep them within zones of comfort can also help ensure good outcomes
on every dive.
Remember your first diving classes and the lesson about bubbling soda bottle
and too rapid ascents? No matter how deeply you study the decompression
theory, this soda bubble analogy is still valid. However, it’s time to introduce
some more fundamentals of the issue. But let’s start from the history:
History
Decompression theory is a relatively old science. Already in late 1800’s,
French physiologist Paul Bert (1833-1886) discovered decompression sickness
and the need for decompression stops and slow ascend speed. Bert also
studied the effects of oxygen to the humans, as he was more interested in
the physiological effects of mountaineering and hot air ballooning. He also
extended his studies to cover high pressure environments, and found out
later about oxygen toxicity.
Bert made a conclusion that high oxygen partial pressures affect humans
chemically, not mechanically, as he described the causes of Central Nervous
System (CNS) oxygen toxicity. When Bert studied air and nitrogen, he
correctly determined the cause of the Decompression Sickness (DCS) to be
caused by the nitrogen bubbles in the blood and other tissues (mechanical
effects). Bert also did experiments on recompression therapy and oxygen
administration in DCS cases. The most famous of Bert’s books is “La Pression
barometrique” 1 , published in 1878, which dealt with the human physiology
in low and high air-pressures.
While Bert laid the fundamentals to the decompression studies, it was John
Scott Haldane (1860-1936), a Scottish physiologist who approached the
problem of decompression theory with more scientific approach. In 1905,
Haldane was appointed by the Royal Navy to perform research about Navy’s
diving operations.
Decompression Basics
Let’s start from basics: A diver goes down and breathes compressed air from
his/her cylinder. Air contains nitrogen, which, as an inert gas, dissolves into
the diver’s tissues. When the diver starts ascending, the ambient pressure
decreases and dissolved nitrogen transfers from other tissues to the blood,
from there to the lungs and finally leaves the body with each exhale cycle.
Simple as that, is it?
When the ascent starts, the diver cannot ascend above the ceiling without
risking the possibility of decompression sickness. The decompression stops
are clearly visible in the dive profile when the line goes below the ceiling
depth. The closer one goes to the ceiling, the less margin of safety remains.
The ceiling depth does not yet indicate on-gassing or off-gassing. Bühlmann
used 16 tissue compartments to model inert gas dissolving in our body. These
compartments either take more dissolved gas in (on-gassing) or expel
dissolved gas out (off-gassing).
The ceiling depth indicates the pressure change from current depth, in which
the leading compartment off-gasses so fast, that further increased pressure
drop would risk the possibility of DCS.
During the ascent phase, a TC can go supersaturated (exceed 100%). The key
of the decompression is to be supersaturated, but not so much that the
dissolved gas would form excess bubbles to our tissues and blood.
The bigger the pressure difference (i.e. pressure gradient), the faster the gas
dissolves, in both directions. This leads to an obvious question: Why not just
come up? What are the limits of supersaturation, and how are they defined?
M-Values
Back to the history: Robert Workman introduced the term M-value, which
means Maximum inert gas pressure in a hypothetical tissue compartment
which it can tolerate without DCS. As mentioned, Haldane found out in his
research that M-value is 2, and Workman refined it to be 1.58 (this number
comes from pressure change from 2 ATA to 1 ATA, and taking into account
that air has 79% inert gases, mainly nitrogen).
Bühlmann used the same method than Workman to express the M-values,
but instead of using the depth pressure (relative pressure), he used absolute
pressure, which is 1 ATA higher at depth. This difference is shown in Figure 3,
where Workman’s M-value line goes above Bühlmann’s M-value line.
To make things a bit more complex, it should be noted that while the M-
values vary by tissue compartment, also two sets of M-values are used for
each TC; M0-values (of depth pressure, indicating surfacing pressure. M0 is
pronounced “M naught”) and M-values of pressure ratio (ΔM, “delta-
M” values).
These sets of values are listed in literature 4. However, they concern the
same thing: maximum allowed overpressure of the tissue compartments. It is
also important to know, that decompression illness does not exactly follow
the M-values. More sickness occurs at and above the pressures represented
by the M-values, and less sickness occurs when divers stay well below the M-
values.
Gradient Factors
Gradient Factors are meant to offer conservatism settings for Bühlmann’s
decompression model.
As mentioned in the previous chapter, M-value line sets a limit which is not
supposed to be exceeded during ascent and decompression. However, as no
decompression model can positively prevent all DCS cases, and because both
dives and divers are individual, additional safety margin should be applied.
On the other hand, we do not want to go too close to the M-value line for
safety reasons. Gradient Factors define the conservatism here.
Some diver’s did not like the idea of using the same conservatism factor
throughout the ascent. Instead of having one GF, there was need to change
the safety margin during the ascent. This led to two GF values; “GF Low” and
“GF High”.
Low Gradient Factor defines the first decompression stop, while High
Gradient Factor defines the surfacing value. Using this method, the GF
actually changes throughout the ascent.
This is illustrated in Figure 4, where GF Low and GF High forms start and end
points to a Gradient Factor line. In that graph, decompression starts when
the inert gas partial pressure in diver’s TC’s reaches 30% of the of the way
between Ambient Pressure line and M-value line.
Then the diver spends time in that stop until partial pressure drops in the
TC’s enough for enabling ascent to next stop, which again has a bit higher GF.
In fact, modern decompression science has proven that there are bubbles
present in our tissues even when there are no DCS symptoms after a dive.
Therefore, M-values neither represent a bubble-free situation,
but tolerable amount of “silent” bubbles in tissues.
Figure 5: Silent bubbles are present in our tissues even when no DCS
symptoms are present. It is important to know personal safety margin and
individual susceptibility to DCS.
A diver goes to 50m/165ft for 20 minutes bottom time, using Trimix 18/45
(18% oxygen, 45% helium) as back gas, and oxygen for decompression from
6m (20ft) on. Descent rate is 15m/min (50ft/min) and ascent rate is 10m/min
(33ft/min). Decompression algorithm is based on Bühlmann ZH-L16B and the
different decompression tables, based on five different GFs, are shown in
Table 1.
These GF parameters are commonly used for different types of dives (e.g.
rebreather, deep/cold dives, default values in some decompression SW) and
GF 100/100 is shown here as a reference, since it is pure Bühlmann table
(containing no margin, so it is also not very safe!).
It is easy to modify the dive plan even drastically by using different gradient
factors.
Now, an easy but dangerous choice would be altering the gradient factors so
that the decompression time decreases, leading to lower decompression gas
need.
Too many divers simply use the default settings or copy their GF parameters
from other divers or even from the Internet, no matter what kind of a dive
they are doing.
Some divers have higher susceptibility to DCS and some dives are physically
more demanding than others. Although the gradient factor method provides
substantial flexibility in controlling the decompression profiles and thus the
dive plan and gas logistics, it just might be worth to hang there a bit longer
sometimes.
REFERENCES
M-values are representative limits for the tolerated gradient between inert
gas pressure and ambient pressure in each compartment.
Other terms used for M-values are "limits for tolerated overpressure,"
"critical tensions," and "supersaturation limits." The term M-value is
commonly used by decompression modelers.
28 | P a g e ©Andy Davis 2015 www.scubatechphilippines.com
Historical Background
In the dissolved gas or "Haldanian" decompression model, gas loading
calculations for each hypothetical "tissue" compartment are compared
against "ascent limiting criteria" to determine the safe profile for ascent.
In the early years of the model, including the method developed by John S.
Haldane in 1908, the ascent limiting criteria was in the form of
"supersaturation ratios."
For example, Haldane found that a diver whose "tissues" were saturated by
breathing air at a depth of 33 fsw could ascend directly to the surface (sea
level) without experiencing symptoms of DCS.
Because the ambient pressure at 33 fsw depth is twice that at sea level,
Haldane concluded that a ratio of 2:1 for tolerated overpressure above
ambient could be used as the ascent limiting criteria.
Most of the U.S. Navy decompression tables were calculated using this
supersaturation ratio method.
However, there was a problem. Many of the tables produced by this method
were deficient when it came to deeper and longer dives.
First of all, he recognized that Haldane's original ratio of 2:1 (based on air)
was really a ratio of 1.58:1 if you considered only the partial pressure of the
inert gas in air - nitrogen.
Bühlmann continued his research for over thirty years and made a number of
important contributions to decompression science. In 1983 he published the
first edition (in German) of a successful book entitled Decompression _
Decompression Sickness.
As a result, the "Bühlmann algorithm" became the basis for most of the
world's in-water decompression computers and do-it-yourself desktop
computer programs.
Three more editions of the book were published in German in 1990, 1993,
and 1995 under the name Tauchmedizin or "Diving Medicine." [An English
translation of the 4th Edition of the book (1995) is in preparation for
publication].
The major difference between the two approaches was that Workman's M-
values were based on depth pressure (i.e. diving from sea level) and
Bühlmann's M-values were based on absolute pressure (i.e. for diving at
altitude).
The ZH-L16A set has sixteen pairs of coefficients for sixteen half-time
compartments and these M-values were mathematically-derived from the
half-times based on the tolerated surplus volumes and solubilities of the inert
gases. The ZH-L16A set of M-values for nitrogen is further divided into
subsets B and C because the mathematically-derived set A was found
empirically not to be conservative enough in the middle compartments.
In addition to air diving, the 11F6 M-values have worked well for trimix diving
and are the basis for many custom decompression tables in use by technical
divers.Many sport divers are familiar with the Recreational Dive Planner
(RDP) distributed by the Professional Association of Diving Instructors (PADI).
The M-values used for the RDP were developed and tested by Dr. Raymond E.
Rogers, Dr. Michael R. Powell, and colleagues with Diving Science and
Technology Corp. (DSAT), a corporate affiliate of PADI. The DSAT M-values
were empirically verified with extensive in-water diver testing and Doppler
monitoring.
Comparison of M-Values
Tables 1 thru 4 (below) present a comparison of M-values for nitrogen and
helium between the various Haldanian decompression algorithms discussed
in this article. All M-values are presented in Workman-style format. An
evolution or refinement in the M-values is evident from Workman (1965) to
Bühlmann (1990).
The general trend has been to become slightly more conservative. This trend
reflects a more intensive validation process (empirical testing) and includes
the use of Doppler ultrasound monitoring for the presence and quantity of
"silent bubbles" (bubbles which are detectable in the circulation but are not
associated with overt symptoms of decompression sickness).
This format is ideal for computer programming since it allows the M-values to
be calculated "on-the-fly" as they are needed. The linear format permits the
display of M-value lines on the pressure graph as well.
M-values can also be expressed in the form of a matrix or table. This is simply
where the M-values for each half-time compartment and each stop depth are
pre-calculated and arranged in columns and rows.
This format is useful for detailed comparisons and analysis. Some of the early
dive computers and desktop computer programs used the table format to
"look up" M-values for each stop during the calculation process.
M-Value Characteristics
M-value sets can be classified into two categories, no-decompression sets
and decompression sets. No-decompression M-values are surfacing values
only.
The DSAT RDP M-values are an example. No-stop dive profiles are designed
so that the calculated gas loadings in the compartments do not exceed the
surfacing M-values. This allows for direct ascent to the surface at any time
37 | P a g e ©Andy Davis 2015 www.scubatechphilippines.com
during the dive. Some no-decompression algorithms account for ascent and
descent rates in the calculations.
A fixed slope of 1.0 means that the compartment will tolerate the same
overpressure gradient regardless of depth. In all cases, the value of the slope
can never be less than 1.0. Otherwise, the M-value line would cross under
the ambient pressure line at some point and this would represent an illogical
situation whereby the compartment could not tolerate even the ambient
pressures.
This is important because when the inert gas loading in a compartment goes
above the ambient pressure line, an overpressure gradient is created. An M-
value line represents the established limit for tolerated overpressure gradient
above the ambient pressure line.
Generally, the faster compartments will cross into the decompression zone
first and be leading (gas loadings closest to M-value lines) and then the rest
of the decompression profile will be controlled by the slower compartments
in sequence.
Mixed gas decompression algorithms must deal with more than one inert gas
in the breathing mix, such as helium and nitrogen in trimix. M-values for this
situation are handled differently by the various algorithms. Some
In the M-value linear equation, the Coefficient a (He+N2) and the Coefficient
b (He+N2) are computed in accordance with the partial pressures of helium
(PHe) and nitrogen (PN2) as follows:
This might explain why some divers routinely push the limits of their tables or
dive computer.
The experience of diving medicine has shown that the established limits (M-
values) are sometimes inadequate. The degree of inadequacy is seen to vary
with the individual and the situation.
Overall, the dissolved gas model has worked well for divers and the
knowledge base has continued to grow. For example, it was originally
presumed that all inert gas had to remain dissolved in solution and that any
bubbles were indicative of DCS.
However, we now know that silent bubbles are present even during
symptom-free dives. Thus, the reality is that there is a combination of two
conditions during a dive _ most of the inert gas presumably in solution and
some of the inert gas out of solution as bubbles. An M-value, therefore,
Even though good data is obtained about the approximate threshold for
symptoms of decompression sickness (M-values), this process cannot
accurately predict or guarantee an absolute threshold for everyone.
Also, we know from experience that certain factors are predisposing for DCS:
Many divers would like to be in the range of "no symptoms" and "very low
level of risk" when it comes to their decompression profiles. Fortunately, it is
well understood among decompression modelers and programmers that
calculations based on the established M-values alone cannot produce
sufficiently reliable decompression tables for all individuals and all scenarios.
M-Value Relationships
Some fundamental relationships involving M-values and decompression
calculations are indicated on the pressure graph in Figure 3
.
43 | P a g e ©Andy Davis 2015 www.scubatechphilippines.com
The Percent M-value calculation has been used by various decompression
modelers over the years. Professor Bühlmann, for example, evaluated many
of his decompression trials on a Percent M-value basis and reported the data
as such in his book(s).
Analysis of Profiles
Many divers would like to know precisely what the effect is of the
conservatism factors in their desktop decompression program(s). They realize
that longer and deeper profiles are generated with increasing conservatism
factors, but more fundamental information is desired.
Both the Percent M-value and Percent M-value Gradient relationships are
useful for the analysis and evaluation of decompression profiles. Using a
standard set of reference M-values, different profiles can be evaluated on a
consistent basis. This includes comparison of profiles generated by entirely
different programs, algorithms, and decompression models.
At 100% Conservatism Factor, the profile is in the 77% M-value range and has
entered approximately 20-35% into the decompression zone. Note that the
values given in Table 5 are upon arrival the respective stops which is the
45 | P a g e ©Andy Davis 2015 www.scubatechphilippines.com
worst-case condition. This correlates with the edges of the "stair-steps" in the
gas loading profile on the pressure graph (see example in Figure 3).
The highest values across all profiles are calculated upon arrival at the
surface which illustrates why a very slow final ascent from the last
decompression stop to the surface is always prudent.
Margin of Safety
Using the M-value relationships and a standard set of reference M-values,
divers can determine personal decompression limits which are both well-
defined and transportable. The margin of safety selected will depend on
individual disposition and prior experience with profiles.
- Bennett PB, Elliott DH, eds. 1993. The physiology and medicine of diving. London: WB
Saunders.
- Boycott AE, Damant GCC, Haldane JS. 1908. The prevention of compressed air illness. J
Hyg (London) 8:342-443.
- Bühlmann, AA. 1984. Decompression_ Decompression sickness. Berlin: Springer-
Verlag.Bühlmann, AA. 1995. Tauchmedizin. Berlin: Springer-Verlag.
- Hamilton RW, Muren A, Röckert H, Örnhagen H. 1988. Proposed new Swedish air
decompression tables. In: Shields TG, ed. XIVth Annual Meeting of the EUBS. European
Undersea Biomedical Society. Aberdeen: National Hyperbaric Center.
- Hamilton RW, Rogers RE, Powell MR, Vann RD. 1994. Development and validation of no-
stop decompression procedures for recreational diving: The DSAT Recreational Dive
Planner. Santa Ana, CA: Diving Science and Technology Corp.
- Schreiner HR, Kelley PL. 1971. A pragmatic view of decompression. In: Lambertsen CJ, ed.
Underwater Physiology IV. New York: Academic Press.
- Wienke BR. 1991. Basic decompression theory and application. Flagstaff, AZ: Best
Publishing Co.
- Wienke BR. 1994. Basic diving physics and applications. Flagstaff, AZ: Best Publishing Co.
- Workman RD. 1965. Calculation of decompression schedules for nitrogen-oxygen and
helium-oxygen dives. Research Report 6-65. Washington: Navy Experimental Diving Unit.
- Workman RD. 1969. American decompression theory and practice. In: Bennett PB, Elliott
DH, eds. The physiology and medicine of diving and compressed air work. London:
Baillière, Tindall & Cassell.
Andy
47 | P a g e ©Andy Davis 2015 www.scubatechphilippines.com