From Haldane to Bühlmann
The decompression model in your computer has a family tree going back more than a century. A short history of how we learned to come up.
Read this first Information, not instruction. Nothing here replaces training with a qualified instructor.
Every time your computer calculates a ceiling, it's running a descendant of an experiment with goats. The history of decompression is a history of people trying to predict something they couldn't see, and slowly getting better at it.
- 1670
A bubble in a viper's eye
Robert Boyle puts a snake in a vacuum chamber and sees a bubble form in its eye. It's the first recorded observation of gas coming out of solution in a living body.
- 1878
Paul Bert identifies nitrogen
Bert shows that bubbles of nitrogen cause the illness seen in caisson workers, and that slow decompression prevents it.
- 1908
Haldane's staged decompression
Boycott, Damant and Haldane publish the tissue-compartment model and the idea of staged stops, based largely on experiments with goats.
- 1956
US Navy tables revised
The US Navy publishes revised air decompression tables that remain the basis of recreational tables for decades.
- 1965
M-values
Robert Workman defines maximum tolerated inert gas pressures for each compartment, the M-values still used today.
- 1983
Bühlmann's ZH-L16
Albert Bühlmann publishes a 16-compartment model with half-times from about 4 to 635 minutes, including helium.
- 1980s
Bubble models
David Yount and colleagues develop the varying permeability model (VPM), which aims to limit bubble growth rather than dissolved gas.
- Late 1990s
Gradient factors
Erik Baker publishes gradient factors, giving divers two dials to control conservatism in Bühlmann.
- 2011
The deep stops trial
A US Navy trial finds more decompression sickness on a deep-stop schedule than a conventional one.
Haldane's big idea
John Scott Haldane was asked by the Royal Navy to make diving safer. His insight was that the body doesn't absorb or release gas all at once. Different tissues take up gas at different rates. He modelled this with five theoretical compartments, each with its own half-time.
He also proposed that a tissue could tolerate a certain amount of supersaturation, which he set at roughly a two-to-one ratio, before bubbles caused harm. That let divers ascend quickly at first, then slow down with stops as they neared the surface. Every staged decompression schedule since follows that shape.
From ratios to M-values
Haldane's fixed ratio worked for the dives of his day but not for deeper, longer ones. Robert Workman's work for the US Navy in the 1960s replaced it with M-values: a maximum allowed pressure for each compartment that changes with depth. That made the model more flexible and more accurate for a wider range of dives.
Bühlmann and the helium problem
Albert Bühlmann, a Swiss physician, worked on deep mixed-gas diving from the late 1950s, including with Hannes Keller on his record-breaking deep dives. His ZH-L16 model, with 16 compartments and separate values for nitrogen and helium, was designed to handle the range of dives technical divers now do. The C variant, adjusted for use in dive computers, is the one most tech computers run today.
Bubbles enter the picture
Doppler ultrasound, developed for diving in the 1960s and 70s, showed that bubbles form after many dives with no symptoms at all. That undermined the idea of a clean line between safe and unsafe. Bubble models like VPM tried to control bubble growth directly. They usually put the first stops deeper, and they became popular with technical divers in the 1990s and 2000s.
Where we are now
Today most technical divers use either Bühlmann with gradient factors or a bubble model like VPM-B. The 2011 US Navy trial pushed many away from very deep first stops, but the debate isn't settled. Our article on deep stops covers the current arguments.
What hasn't changed since 1908 is the basic humility of the exercise. Every model is a prediction, and individual divers vary.