The Bühlmann ZHL-16C Algorithm: How Dive Computers Calculate No-Decompression Limits

One diver surfaces after a 40-minute descent to a wreck, their wrist computer clearing them for immediate ascent. Their dive buddy, floating right beside them on the exact same profile, stares at a screen demanding a mandatory 15-minute decompression stop at 20 feet. To the first diver, the algorithm is a flexible tool that safely maximizes bottom time by trusting modern dissolved-gas models. To the second, who manually dialed in conservative gradient factors, that same underlying mathematics is a baseline that requires strict safety buffers to prevent micro-bubble formation in slow tissues. Both are relying on the exact same Bühlmann ZHL-16C algorithm, yet they are experiencing two entirely different realities in the water.

In the oppressive silence of a deep dive, the human body undergoes an invisible but relentless physics experiment. As ambient pressure increases, inert gases—primarily nitrogen—dissolve into the bloodstream and tissues. During ascent, as that pressure drops, those gases must safely exit the body. If the pressure drops too rapidly, the nitrogen comes out of solution and forms bubbles, leading to decompression sickness. Managing that ascent is the primary job of a dive computer, and for the vast majority of technical and recreational divers today, the mathematical engine making those life-or-death decisions is the Bühlmann ZHL-16C.

Originally developed in the 1960s by Dr. Albert A. Bühlmann, a Swiss physician specializing in respiratory physiology at the University of Zurich, the model was designed to simulate how the human body absorbs and releases inert gas. The 'ZH' in the name stands for Zurich, the 'L' denotes the linear nature of the oversaturation limits, and the '16' represents the number of theoretical tissue compartments the algorithm tracks.[2]

These 16 compartments do not correspond to specific organs, but rather to mathematical half-times ranging from 4 minutes to 635 minutes. Fast tissues, like the blood and brain, absorb and release nitrogen quickly. Slow tissues, like cartilage and bone, take hours to saturate and just as long to clear. The algorithm continuously calculates the nitrogen pressure in all 16 compartments simultaneously, comparing them against a maximum tolerated supersaturation limit known as the M-value.[2]

The M-value is the absolute threshold. If the nitrogen pressure in any of the 16 compartments exceeds its specific M-value during an ascent, the model predicts that bubbles will form. Therefore, the computer dictates a decompression stop, holding the diver at a specific depth until the gas pressure drops back into the safe zone. For decades, diving right up to the M-value was considered acceptable, but modern physiological understanding has shifted.[2]

"Decompression algorithms are not a single fixed answer - even widely trusted planners disagree with each other," notes the validation documentation from Dive Tools, an independent decompression engine testing platform. The realization that diving to the absolute mathematical limit carried unacceptable risks for many led to the most significant evolution in the Bühlmann model's history: the introduction of Gradient Factors (GF) by engineer Erik Baker in 1998.[4]

Gradient Factors allow divers to artificially lower the M-value, creating a customizable safety buffer. They are expressed as a pair of percentages, such as 30/70. The first number, GF Low, dictates the maximum allowed saturation when the first deep decompression stop occurs. A GF Low of 30 means the diver will be halted when their leading tissue reaches just 30 percent of its theoretical limit, forcing them to stop deeper and earlier.[2]

The second number, GF High, controls the safety margin upon surfacing. A GF High of 70 means the diver will not be allowed to break the surface until all 16 tissue compartments have dropped below 70 percent of their maximum M-value. By adjusting these two numbers, a diver can radically alter the shape of their ascent, prioritizing deep stops to control micro-bubbles or extending shallow stops to ensure slow tissues are thoroughly off-gassed.[2]

This flexibility is why two divers on the same wreck can have entirely different decompression obligations. Dive computer manufacturers implement the ZHL-16C mathematics identically, but they diverge wildly on the default Gradient Factors they ship from the factory. A brand catering to recreational vacationers might lock the algorithm behind a highly conservative preset, while a technical computer allows the user to manually input any GF pair they desire.[5]

The US Navy Experimental Diving Unit (NEDU) has extensively tested commercial off-the-shelf algorithms against their own rigorous man-trials. In a validation study published by the National Center for Biotechnology Information, researchers found that the base algorithm required tweaking to match military safety standards. "ZHL-16C could be adjusted to pass all of the tests with GF-Hi <= 70 and GF-Lo >= 55," the researchers concluded.[1]

Despite the NEDU's findings, the technical diving community often favors a GF Low of 30, believing that deeper stops prevent the initial formation of bubbles. However, recent clinical studies have challenged this practice, suggesting that stopping too deep for too long actually continues to saturate the slow tissues with nitrogen, paradoxically increasing the overall risk of decompression sickness by the time the diver reaches the surface.[1]

This debate has led to a schism in how manufacturers approach the algorithm. Some have historically favored proprietary bubble models that automatically mandate deep stops, but have recently integrated the Bühlmann 16 GF algorithm into newer models to satisfy market demand. Others, like Dive Gear Express, offer firmware upgrades that replace the Bühlmann model entirely with the DCIEM algorithm, a different mathematical approach based on serial tissue compartments rather than parallel ones.[3]

When Dive Tools cross-checked their own ZHL-16C engine against industry standards, they found a 90 percent agreement rate with Shearwater's implementation across 15 complex dive profiles, including multi-level and closed-circuit rebreather scenarios. This high level of parity confirms that the core mathematics of the Zurich model remain robust and universally understood by software engineers.[4]

Yet, the precision of the math masks the fundamental uncertainty of the biology. The 16 compartments are theoretical constructs, not biological realities. The algorithm does not know if the diver is dehydrated, cold, fatigued, or older—all factors that significantly alter how the human body actually processes dissolved gases under pressure.[5]

The Bühlmann ZHL-16C endures not because it is a perfect biological mirror, but because it is a transparent, predictable framework. Unlike proprietary black-box algorithms, its open mathematics allow divers to understand exactly why their computer is demanding a stop. The algorithm provides the data, but the diver, through their choice of Gradient Factors and their understanding of their own physiology, dictates the final margin of safety.[5]