As a hyperbaric physician and dive instructor, I’ve been around a lot of dive accidents, and I’ve spent decades educating divers and treating decompression sickness (DCS). A misunderstanding I consistently see among the divers I treat pertains to no-decompression limits (NDLs), with divers repeating the common refrain that they were “diving within the limits” and that their DCS must therefore have been a random event.
DCS is a probabilistic condition, in which injury risk increases with proximity to NDLs. The fundamental variables we (and our computers) use are depth and time. We also appreciate that additional factors influence DCS probability, including breathing gas, dive profile characteristics, decompression time, thermal stress, exertion, hydration status, age, sex, fitness, body composition, acclimatization, and circulatory health.
Our current understanding of these many variables does not yet translate to quantifiable risk values that we can use to prevent DCS cases. Nor are there countermeasures, other than entirely avoiding exposure, that can eliminate all risk. That said, we are not lost in the wilderness and subject to random occurrence. Indeed, our ability to predict DCS is quite good. Our current decompression models, however imperfect, have resulted in very low DCS incidence among those who stay within their NDLs.
NDLs, also known as no-stop limits, provide depth and bottom time parameters. When we remain within these boundaries we can ascend directly to the surface without an anticipated risk of injury. Indeed, if we remain within 50% of the allowable bottom time, we have a negligible risk of DCS with direct ascent to surface.
In the precomputer era, dive tables provided strict parameters for bottom times based on the maximum dive depth, regardless of time spent. Understandably, the advent of dive computers that enabled longer bottom times by accounting for ascents to shallower depths (with less inert gas loading) were a welcome change. Longer dives, however, result in an overall increase in inert gas load, and we should appreciate that the decompression algorithms used in computers are mathematical models. They’re not based on controlled trials or an endpoint of postdive DCS symptoms.
The U.S. Navy’s experimental work created NDLs based on the two parameters of depth and time (on air) with an endpoint of injury (DCS). The testing was based on square profiles, necessary to determine safe limits for Navy divers. The growth of technical, rebreather, and altitude diving has led to advances in our understanding of decompression theory and utilization of different models, all with the common aim of avoiding injury and facilitating safe exploration.
Despite the minimal incidence of DCS associated with no-decompression diving in the precomputer era, many of us added safety buffers. As deep dives commonly employ square profiles, when I was teaching we would plan dives with the NDL set two or three “boxes to the left” of the actual black-box (NDL) limits. This practice was employed both to keep divers safe and to underscore the increase in DCS risk associated with proximity to NDLs.
The computer version of this approach to safety is the capacity within dive computers to adjust NDL parameters — through conservative settings or gradient factors — based on risk tolerance. Modifiable NDLs should elevate our appreciation for the range of NDL possibilities (these limits are not set in stone) and should prompt us to respect these limits by being thoughtful about our closeness to them.
The upside of computer use includes continual tracking of depth and bottom time and continual updating of the NDL. The downsides include the potential for overconfidence in the protection offered by NDLs (where dives within the NDLs are considered 100% safe and risk is thought to be elevated only when NDLs are exceeded). We should take heart in the impressive safety record of diving and the rarity of DCS, especially severe neurological DCS. There remain, however, and will remain “error bars” on either side of NDLs, wherein the probability of DCS increases with proximity to the NDL.
Another aspect of computer use can affect safety and potentially cause trouble for divers: different computers with different settings. Not long ago divers used a common set of tables, and everyone breathed air. Today divers use different computers, different algorithms, different computer safety settings, and different gas mixtures. As such, unless divers use the same computer with the same breathing gas and safety settings, each will receive different NDLs throughout the dive.
One factor I see as a likely contributor to many DCS cases is the practice of remaining close to NDLs throughout dives (and throughout dive series). While a diver may remain “within the limits,” they are nonetheless exposed to more decompression stress (inert gas loading) than a dive table would allow. Some computer models now minimize DCS risk by extending safety stops from three minutes to five minutes (or more), especially when a diver gets within five minutes of their NDL or when they dive deeper than 100 feet.
It is important to remember that the more risk factors you have — such as cold water, exertion, and deep dives — the further from NDLs you should remain and the longer your safety stops should be. When your dives push NDL boundaries or involve these risk factors, strongly consider incorporating longer surface intervals and fewer dives per day into your dive planning.
Nick Bird is the medical director at the Center for Hyperbaric Medicine at Virginia Mason Medical Center in Seattle, Washington.
© Alert Diver – Q1 2025