Chapter 16

Dragon Peaks Resort.

The King Air was on its last trip. This time, by road. Crated inside a shipping container on the back of a container transporter, it left Dragon Peaks Airfield at 10:25. I stood there and watched it go, feeling a strange mixture of relief and sadness as the truck disappeared down the access road.

We had been at this for just over two weeks.

Two weeks of wreckage, paperwork, interviews, measurements, photographs, weather reports, flight planning data and enough coffee to keep a small army awake. It was the first time I had conducted both the site investigation and the complete wreckage reconstruction at the same location. Normally, the wreckage would be recovered and moved somewhere more suitable for examination. Here, we’d had the luxury of being able to work from the accident site and the reconstruction area without losing the geographical context.

There was something strangely final about seeing the King Air leave Dragon Peaks.

It had arrived here in pieces. Now it was leaving in a box. The investigation had reached its conclusion.

The team was unanimous on the broad cause of the accident. Atmospheric conditions had caught the flight crew by surprise, and by the time the implications of those conditions became apparent, there was no realistic opportunity left to intervene and prevent the outcome.

It was a sad conclusion. There was no dramatic mechanical failure to blame. No single act of pilot recklessness. No convenient villain sitting in a cockpit waiting to be identified. Instead, there had been a series of small holes.

The Swiss Cheese Model.

The first hole in the slice to line up was the delayed departure.

That one had not been pilot error. The crew had been operating within company expectations. The problem was the company’s policy itself, a well-intentioned attempt to keep passengers happy.

Keep the PAX happy. Wait on a delayed PAX.

It sounded harmless enough when written in a company policy manual.

In the real world, however, two hours was a long time for conditions to change in mountainous country.

I remembered an article I’d read once about the Concorde.

The comparison wasn’t entirely fair, of course, because the King Air was no Concorde. Not even remotely. But the story had stuck with me because it demonstrated just how much speed could change the relationship between an aircraft and the world beneath it.

A Concorde departed Boston for Paris at about the same time a Boeing 747 left Paris for Boston. The Concorde crossed the Atlantic, landed at Charles de Gaulle, turned around in roughly sixty-eight minutes, departed again for Boston and still arrived there about eleven minutes before the Boeing 747 had reached the runway threshold.

That was the Concorde. Fast enough to make geography look negotiable.

The King Air was no Concorde. Its world was measured in hours rather than minutes, and its flight path was constrained by terrain, weather, fuel and the limitations of the aircraft and its crew.

The second hole in our Swiss Cheese Model to line up was the atmospheric pressure discrepancy.

The pressure had changed significantly enough to affect the relationship between the aircraft’s indicated altitude and its actual altitude. The recalculated pressure information was not communicated to aircraft operating in that part of the Drakensberg in time to prevent the accident.

By the time the information became available, the crew and passengers of the King Air were already dead.

That was something I struggled with. Not because I expected weather information to be perfect. It never was. But because information that could potentially save lives is only useful if it reaches the people who need it while there is still time to act on it.

One of my recommendations in the report would therefore be that significant atmospheric pressure changes in mountainous operating areas should be communicated to aircraft at regular intervals, with a fifteen-minute maximum interval as a proposed operational benchmark where practical.

Four times an hour. It wasn’t foolproof. Nothing in aviation was.

But it would be another layer in the Swiss Cheese Model, and sometimes another layer was all you needed to stop the holes from lining up.

Another recommendation would address operations in mountainous areas when aircraft encountered deteriorating weather or IMC.

Rather than relying on an arbitrary increase of one thousand feet above whatever altitude the aircraft happened to be flying at, my recommendation would focus on ensuring that crews obtain an altitude that provides adequate terrain clearance in accordance with the applicable minimum safe or sector altitude and the prevailing conditions.

In simple terms, if the weather closes in around you in the mountains, going a little higher isn’t enough. You need to know exactly how high you need to be.

I sat down at the little desk in my room and opened my weather app and flight-planning system on the laptop.

The investigation was finished. The King Air was gone. Now I had to get myself home.

I started planning my flight back to Wonderboom.

The route gave me a magnetic course of approximately 005 degrees. I planned to cruise at 8,500 feet MSL and around 138 KIAS.

The weather application showed a light two-knot tailwind component.

Nothing spectacular. Just enough to help.

The calculated distance was 212 nautical miles. Based on Misty’s fuel planning figures, the flight would require approximately 330 pounds of fuel, which worked out to about 49 US gallons.

That number made me pause for a moment.

Forty-nine gallons sounded like a lot until I compared it with what Misty could actually carry. Her usable fuel capacity was 146 gallons of Jet A-1, giving her a range of 533 nautical miles and an endurance of three hours and fifty-two minutes.

At 49 gallons consumed, I’d be using roughly 34 percent of the available fuel, leaving about 97 gallons before allowing for whatever reserves and operational considerations I wanted to apply.

That was a comfortable margin.

More importantly, it gave me options.

I never liked planning a helicopter flight on the assumption that everything would go exactly according to the spreadsheet. The weather could change. ATC could delay me. I could need to divert. A temporary restriction could appear. A headwind could replace the tailwind.

Fuel was not there to make the planned flight possible.

Fuel was there to give the pilot choices when the planned flight stopped being the flight he was actually flying.

At the planned cruise speed and allowing for the climb, descent and normal operational variations, I estimated the flight would take somewhere in the region of one hour forty to one hour forty-five minutes.

If everything behaved itself, we’d be back at Wonderboom at roughly 10:40.

Give or take a few minutes.

I looked at the route again. At 8,500 feet, and assuming the weather remained within VFR limits, the flight could technically be conducted under VFR. But I wasn’t going to let the fact that I could fly it VFR become the reason I did so.

I would file IFR.

That would give me a structured flight plan and ATC services for the route, rather than simply wandering north under my own navigation while hoping the weather stayed friendly.

After the last two weeks, I had developed a healthy respect for hoping. Hoping wasn’t a flight plan. I closed the route page and looked at the clock.

The weather forecast was good. For now.

I would do one final weather check at 08:45 tomorrow morning before starting Misty’s engine.

Because after everything that had happened here, I wasn’t about to let a computer forecast have the last word. The mountains had already taught me that lesson long ago. Tomorrow, I would listen to the weather.

And if the weather had something to say, I would listen carefully.

A soft knock at the door saved me from staring at my laptop screen any longer.

I didn’t need to ask who it was. The tea tray made its characteristic faint rattle as Megan balanced it against the doorframe.

“Come in...”, I called out, closing the route planning window but leaving the nav log open.

Megan pushed the door open with her elbow, carefully carrying two steaming mugs and a small plate with four Ouma rusks—condensed milk flavour, if the sweet aroma was anything to go by. She wore a comfortable oversized jersey, her slightly damp hair cascading over her shoulders, looking far removed from the sharp, clinical investigator who had spent the last two weeks measuring impact marks in the mud.