22 of 24 - Fatal accident

Back to the national construction company

After several years during which I worked in Qatar for a consultant and co-ran my own safety consultancy in the UK, Insight Safety, I re-joined the group and was offered the senior safety role in Dubai with dozens of officers, high-profile builds, and a safety culture still finding its footing. It was here, amid the dust and heat, that I realised just how far I’d come from those early mornings in Cheltenham.

The work demanded not just technical oversight but cultural fluency, logistical coordination, and the ability to navigate layered hierarchies—contractors, consultants, and clients, each with their own expectations and interpretations of “safe practice.” It was a place where heat, haste, and hierarchy collided daily, and where the margin for error was often razor-thin.

One of the projects I worked on involved very deep trenching—deeper than anything I’d seen in the UK. Unsupported, but properly and safely stepped, carved into compacted sand that would have had a British inspector reaching for the stop notice. But here, it was normal and expected. Dewatering pumps thrummed in quiet rhythm, and the excavation process moved with a kind of choreography. I asked for an explanation of the control measures. Then I asked again. Eventually, I was convinced—not just by the paperwork, but by the sequencing, the redundancy, the inspection regime and the quiet competence of men who knew this ground.

Then my UK boss arrived. He frowned at the depth, the absence of shoring, the unfamiliar soil. “You’re comfortable with this?” he asked.

“I wasn’t,” I said. “But I’ve seen it work.”

He paused. Not agreement, but a kind of grudging trust. The trench held.

Al Ain, a few years earlier.

The trench was narrow, unstable, and threaded through a congested urban corridor. The spoil was piled right to the edge—compacted deliberately to store more. The ground had been fractured by decades of rework, and the signs were there: minor cave-ins, tension cracks, the quiet language of collapse. The local safety man saw it. He issued a stop work notice. The site manager ignored it.

Three men died when a large chunk of the trench wall collapsed and engulfed them.

I walked the trench the day after. The silence was different—not procedural, but moral. The excavation had spoken. No one had listened. My report was scathing. I didn’t soften the language. There are moments when diplomacy is complicity.

Dismissals followed. Not enough, perhaps. But something.

A final fatality

Our work in Dubai included road layouts, buildings, and a specialist tower to accommodate a scenic ride. The ride featured a tower encircled by a doughnut-shaped observation platform, engineered to ascend and descend, offering passengers a panoramic view. This is the scene of the last fatal accident that I investigated.

The tower itself was a straight-sided tube, 55 metres high, with an internal diameter of just under 3 metres. Its external face housed a rack-and-pinion system alongside the power train—an articulated cable control device that rose and fell with the platform. This power train resembled a flexible spine, composed of hinged, interlinked segments that guided electrical and communication cables as the platform moved. One end delivered power and data to the platform; the other terminated at the tower’s summit, where the cables were routed internally down to the control centre, located below ground level at the base.

External construction was arduous: major excavation, heavy rebar installation, and complex concrete pours preceded the intricate assembly of the platform itself. But it was the internal work that proved most extreme. Daytime temperatures inside the tower tube rendered most tasks untenable. While forced-air ventilation helped at the base, it was ineffective higher up. Much of the internal work was therefore carried out at night. The most demanding operations occurred near the summit, where the cables entered the structure. Accessing this zone required highly skilled workers—electricians and fitters trained in rope access techniques—operating in confined, overheated conditions.

This ride was a one-off. There was no tried-and-tested assembly plan, no detailed study of how to fit the hundreds of bespoke components. Many of the more unusual elements were installed through trial and error. One such component—critical to housing the internal cables at the point of entry—proved especially problematic.

So, consider this. You’re a seasoned technician, confident in your rope access skills, having worked on countless rides. But this one is different. You’re fifty metres up, inside a pitch-black, narrow metal tube. The air is hot and stale. Just reaching the work face is a trial. And when you arrive, you realise your rigging doesn’t give you the access you need. The component lies on the far side—180 degrees across the 3m void.

You unclip. Just for a moment. You reach across, misjudge the handhold, or the foothold. And you fall.

I arrived on site the morning after the fall. The tower stood as it always had—silent, vertical, indifferent. There were no signs of disruption from the outside. No broken panels, no scattered tools. Just the knowledge that someone had died inside it.

The investigation began with the usual protocols: access logs, crew rosters, equipment checks. But the truth lay not in the paperwork, nor in the schematics. It lay in the geometry of the space inside the tube, the choreography of the task, and the quiet assumptions that had gone unchallenged.

The component in question had been modified late in the build. A minor adjustment, made to accommodate a cable bend radius that hadn’t been accounted for in the original design. It wasn’t flagged as a risk. It wasn’t even documented properly. But it changed the work face—shifted it just far enough to make the standard rigging ineffective.

We reconstructed the sequence. Rope access, partial reach, uncoupling. A moment’s misjudgement. A fall.

Just a worker doing what he’d done a hundred times before, in a structure that hadn’t been built a hundred times. The ride was a one-off. So was the accident.

Halfway through the investigation, I had to leave Dubai. The reasons were practical, but the timing was brutal. The site—the tower, the heat, the geometry—remained behind. I continued the inquiry from a desk thousands of miles away, speaking into a screen early in the morning , relying on an assistant still on the ground.

It was surreal. The tower I knew intimately now appeared only in grainy video feeds and still images. The sounds of the site—metal on metal, the low hum of ventilation—were replaced by buffering audio and the occasional dropped call. I found myself asking for angles, for measurements, for gestures that I would normally take in with a glance. My assistant became my proxy: climbing, photographing, describing. I trusted him, but it wasn’t the same.

The stress came not from the technical complexity, but from the dislocation. I was still responsible, still accountable, but no longer present. I couldn’t feel the heat inside the tube. I couldn’t trace the cable path with my own hands. I couldn’t stand at the base and look up.

We reconstructed the sequence. We reviewed the rigging. We examined the component. But the work was mediated—filtered through screens, compressed into reports. The human cost, though, remained unfiltered.

I remember one call in particular. My assistant described the reach, the misalignment, the probable misstep. I listened, took notes, asked questions. And then I sat back and stared at the screen. It was silent. Just a still image of the inside of the tube. Black, narrow, unforgiving.

That was the moment I understood the limits of remote investigation. You can analyse, interpret, report. But you cannot bear witness. Not properly. Not fully.

We wrote the report. The tower was operational and the ride ascended and descended as designed. Passengers marvelled at the view.

But I still think about the geometry. The heat. The silence inside that tube. And the moment when a man reached across the void, and didn’t come back.

In the months that followed, I found myself withdrawing. Not dramatically. No grand declarations. Just a quiet erosion of appetite—for the meetings, the reports, the next project. That empty, disembodied feeling I’d carried through the remote investigation didn’t fade. It settled.

I retired earlier than I’d planned. Not because I couldn’t do the work, but because I no longer wanted to. The distress wasn’t just emotional—it was ethical. The conclusion I reached was simple, and depressing: that although there were countless reasons why the job was almost impossible to build safely, and although the designers, planners, technical experts and we contractors could all have prevented that death, none of us did.

It wasn’t malice. It wasn’t neglect. It was something quieter. A shared assumption that someone else had checked, someone else had tested, someone else had thought it through. And in that silence, a man fell.

I’ve carried that with me. Not as guilt, exactly, but as a kind of residue. A reminder that technical mastery is not enough. That safety isn’t just a checklist—it’s a culture. And that sometimes, the most dangerous thing on site isn’t the machinery or the geometry. It’s the absence of a voice saying: “This isn’t safe.”

Final Briefing

On my last day at work, I attended a group safety meeting held in a corporate room inside a football stadium. A few of the old faces were there—colleagues I’d worked alongside for years—but there were many more younger ones I didn’t recognise. New recruits, perhaps. Or just new to me.

I stayed through the morning session and lingered until lunch was done. Just before I left, the Group Safety Director invited me to say a few words. I kept it simple. I told them that over the years I’d seen real change: better awareness, safer tools, and a shift in boardroom attitudes that had made working lives safer and more dignified. I reminded them that they were doing important work in one of the most dangerous industries, and that they should feel proud of their contribution.

Then I stepped away—quietly, without ceremony—leaving the room to those who would carry it forward.