When a Blue Plastic Bag Becomes the Last “Safety Barrier”

Happened just now…

When a Blue Plastic Bag Becomes the Last “Safety Barrier”

Good morning from the plant.

We have a gasoline pump and its associated piping system. The pipe is clearly marked “Benzine”.

The system is leaking.

The leak is known, visible and very real.

The temporary solution? Wrap the leaking section in a blue plastic waste bag and tape to catch the gasoline.

Then start welding approximately four metres away.

But apparently everything is under control. The workers have personal gas detectors. There is a Hot Work Permit. And the person authorising the permit?

He is sitting in an office about two kilometres away.

This is an Operational Ex story.

A gas detector does not make hot work safe.

One of the most dangerous misconceptions in hazardous-area operations is treating the presence of one protective measure as proof that the situation is safe.

“We have gas detectors.” Good. But what does that actually prove?

A personal gas detector is an important detection barrier. It can warn the wearer when the concentration at the sensor reaches a defined alarm level.

But it does not stop the gasoline leak. It does not isolate the process. It does not prevent evaporation. It does not control where the vapour travels. And it certainly does not remove the ignition capability of welding.

Gas detection is therefore part of the control strategy. It is not a substitute for controlling the release.

Four metres is a distance, not a risk assessment.

Looking at this situation, the obvious question might be: “Is four metres far enough away to weld?”

Wrong question.

The correct question is: On what basis have we demonstrated that a flammable atmosphere cannot reach the hot-work location - and remain controlled for the entire duration of the work?

Gasoline vapour does not understand the boundary drawn on a hazardous-area classification drawing. It does not care where the welding permit says the worksite begins.

With an active release, vapour behaviour depends on the release characteristics, evaporation, ventilation, wind direction, equipment geometry, low points, obstructions and numerous other site-specific conditions.

Therefore: “The welding is four metres away” is not a safety argument. It is a measurement.

Then there is the blue plastic bag

Perhaps this is the most powerful part of the photograph.

The bag probably appeared because somebody needed an immediate practical solution:

“It is leaking. Let’s catch it somehow.”

That may reduce liquid dripping onto surrounding equipment or the ground.

But collecting leaking liquid and controlling a flammable release are two very different things.

A plastic bag is not process isolation. It is not engineered containment. It is not a verified repair. And it certainly does not mean that the source of release has disappeared.

In fact, the leak should trigger exactly the opposite thought: Our operating condition has changed.

The hazardous-area classification drawing may correctly describe the installation under the assumptions used when the classification was prepared.

An abnormal active gasoline leak is another condition.

The drawing tells us what we designed for. The leak tells us what is happening now.

A Hot Work Permit is not a piece of paper.

This brings us to perhaps the most important management question.

If the person issuing the Hot Work Permit is sitting two kilometres away, what exactly has been authorised? Has that person seen the leak? Have they seen the temporary plastic containment? Do they know the actual condition of the plant? Has isolation been physically verified? Do they know where the gas measurements were taken? Do they know when they were taken? Do they understand how frequently monitoring will continue? Has the potential migration of gasoline vapour from the release towards the hot-work location been considered?

And perhaps most importantly: Is there a clearly defined STOP WORK criterion if conditions change?

A permit-to-work system is supposed to control hazardous work. The signature is evidence of a process.

Look at it as barrier management.

Put the situation into a simple BowTie logic.

Hazard: gasoline.

Threat: loss of containment.

Top event: release of gasoline/vapour with the potential for a flammable atmosphere.

On the preventive side we would normally expect layers such as mechanical integrity, containment, isolation, repair, control of the work area, appropriate gas-testing strategy and an effective permit-to-work system.

Then, only a few metres away, we deliberately introduce an extremely capable ignition source: hot work. (Side note: plastic bag - static discharge)

At this point the interesting question is: “How many barriers had already degraded before an active gasoline leak and welding four metres away could exist in the same operational decision?”

Here it becomes a management problem.

The Head of Operations question:

For the Head of Operations, the decision should be: “Can I demonstrate that the plant and the work are in a controlled, safe condition?”

Where there is an uncontrolled or inadequately controlled flammable release and we intend to deliberately introduce a high-energy ignition source nearby, the sequence matters.

Leak → plastic bag → gas detector → permit → weld.

That difference captures much of what Operational Ex is about.

Explosion protection does not prove itself when everything is operating normally.

We discover whether the system really works when something is no longer normal.

Cafe break is over, let us get back to work… ☕️