When the Air Stops Flowing

The dive began the way most dives do — quietly and uneventfully.

The sea was calm, visibility was good, and the preparation routine was familiar. My cylinder pressure showed 206 bar, measured through a wireless tank pressure transmitter connected to my regulator. My buddy and I completed the usual pre-dive checks and confirmed the cylinder valve was fully open.

One thing was slightly unusual that morning: I did not have my analogue pressure gauge attached. Earlier in the week, a dive buddy’s computer had failed and he had borrowed the SPG from my regulator set so he could safely complete his dives. That meant that for this dive my only pressure indication was the wireless transmitter.

Transmitters are generally reliable, so it didn’t feel like a major concern at the time, but it did remove a layer of redundancy.

We began our descent.

As we went down, the pressure indication from the transmitter started behaving strangely. Occasionally it jumped to values that clearly didn’t make sense before returning to something normal.

At first I assumed it was radio interference. In aviation there have been increasing reports of GPS interference and signal disruptions in various parts of the world, sometimes producing erratic or unrealistic readings on navigation equipment. The behaviour looked somewhat similar. Since tank transmitters also rely on wireless communication, it didn’t seem impossible that interference could occasionally disturb the signal.

Still, it was enough to make me slightly more alert.

We continued descending to about 20 metres.

Something Isn’t Right#

After levelling off we spent a few minutes exploring. The dive initially felt completely normal.

Then the breathing changed.

It wasn’t dramatic. There was no sudden failure or violent rush of bubbles. Instead, the regulator began to feel slightly different.

The first part of each breath felt normal, but the final portion of the inhalation felt weak — as if the regulator couldn’t quite deliver the last bit of air. It had the familiar sensation of breathing from a tank that is nearly empty.

Subtle, but unmistakable once you notice it.

Before even checking the computer again, I had already signalled my buddy to move closer. Something about the breathing simply didn’t feel right.

When I looked at the pressure reading again, the display briefly showed 8 bar.

That was clearly impossible.

Within a few seconds, the regulator stopped delivering air altogether.

Switching to the Backup#

Because my buddy was already nearby, the response was straightforward. I switched to her octopus, which immediately delivered normal airflow.

We signalled the situation to the rest of the group. To draw attention underwater, I used my buddy’s magnetic shaker, producing the distinct rattling sound that divers recognise as a signal for attention.

Everyone quickly understood that something unusual was happening.

We began a controlled ascent.

During the ascent, one of the other divers in the group deployed a surface marker buoy (SMB) early. We were outside the designated swimming area, so the buoy served primarily to increase visibility for passing boat traffic and to signal that divers were surfacing in that location.

The ascent itself remained calm and controlled. With the octopus regulator delivering air normally, there was no urgency, simply a clear decision to end the dive.

Within a few minutes we were back at the surface.

The Strange Behaviour at the Surface#

Once on the surface I tried the regulator again.

Pressing the purge button produced a very brief burst of air, after which the flow stopped almost immediately. It behaved exactly like a tank that had run empty.

But the tank had started at 206 bar and the dive had only lasted a few minutes.

When I removed the regulator from the cylinder valve, the explanation became clearer: air flowed freely from the tank.

The cylinder was not empty.

Something inside the regulator system had restricted the airflow so severely that it eventually stopped delivering gas altogether. The most likely explanation is a partial internal failure or obstruction in the first stage, allowing limited airflow initially but progressively restricting it.

The regulator set will now go for servicing before it is used again.

The Part That Happens Afterwards#

Interestingly, the most uncomfortable moment of the entire experience didn’t happen underwater.

It happened later.

When I was packing the regulator set for the flight home, the dive began replaying in my mind, almost frame by frame. This is a very common human response after unexpected events in safety-critical environments.

Underwater, the brain operates in task mode. Training, procedures, and muscle memory take over. The priority is clear: solve the immediate problem. There is very little emotional processing happening in that moment.

Only afterwards, when the situation is safely resolved, does the mind begin to analyse what happened.

In human factors terms, the brain runs through alternative scenarios:

• What if the failure had occurred faster?
• What if it had happened deeper?
• What if the buddy had been further away?
• What if the symptoms had been misinterpreted?

From a risk-management perspective, this mental replay is actually useful. It is part of how humans update their internal safety models. The brain is effectively asking: What did I just learn from this?

But the same mechanism can also produce anxiety if it turns into uncontrolled speculation.

The key is to convert that reaction into structured reflection rather than rumination.

In aviation, this is exactly what formal debriefs and safety reporting systems are designed to do. They transform a potentially emotional experience into a learning event. The same approach works just as well in diving: analyse the sequence calmly, identify what went well, identify what could improve, and then move on.

In this case, the outcome was clear. The anomaly was recognised early, the dive was terminated, and the team response worked exactly as intended.

The system failed, but the safety system did not. That is the prime example of Safety II: "the surprise is not that things occasionally go wrong, but that they go right so often"¹.

What Worked Well#

Looking back, several things worked exactly as they should.

First, the early recognition of the breathing change. Instruments were unreliable in this situation, but the physical sensation of breathing told the real story.

Second, I brought my buddy closer early when something felt unusual. That meant that when the regulator stopped delivering air, the gas-sharing response happened immediately.

Third, the group awareness. Using the magnetic shaker ensured the other divers quickly understood that something was wrong, and the early SMB deployment increased visibility for boat traffic during the ascent.

Most importantly, the situation remained calm and controlled from beginning to end.

Lessons Worth Remembering#

A few practical lessons stand out from this dive.

• Breathing feel is the most reliable indicator of a problem. If a regulator feels different, assume something is wrong.
• Redundancy matters. Having both a transmitter and an analogue SPG provides an important cross-check.
• Instrument anomalies should raise your alert level. The fluctuating pressure readings were an early signal that something wasn’t normal.
• Bring your buddy closer early when something feels off. That small step significantly reduces response time if the situation deteriorates.

And perhaps the most important rule:

If your breathing system behaves abnormally underwater, end the dive.

A Familiar Pattern#

As someone who works in aviation, the parallels are hard to miss.

Both environments rely on complex systems to sustain life. Failures rarely appear dramatically at first. More often they begin as small deviations from normal behaviour — an unusual indication, a subtle performance change, or a system that simply doesn’t feel right.

The key skill is recognising those signals early and acting before they escalate.

On this dive, the anomaly was detected, the dive was terminated, and everyone surfaced safely.

In complex environments, that is exactly how safety systems are supposed to work.