(The previous story ‘Lukla flight crash’ was filed at 9am on Tuesday morning, two hours after the plane went missing. This is an update with some facts corrected, and more details pieced together from interviews with pilots and civil aviation sources.)

A file photo taken in 2008 at Lukla airport of 'Hotel Echo', the Agni Air Dornier 228 that crashed on Tuesday morning.

Capt Lucky Shah was an upbeat person, cheerful and popular with his colleagues. He was regarded by peers as a confident pilot: he had over 30 years of experience flying in Nepal and India. The son of a senior pilot for Nepal Army, he was no daredevil.

On Tuesday morning, word came that Lukla had good visibility for the first time in a week. Domestic airlines serving the gateway to Mt Everest all scrambled to get their planes in the air. There was a huge backlog of passengers and cargo at both Kathmandu and Lukla as the Khumbu geared up for the autumn trekking season.

The trouble was that it was raining heavily over central Nepal and the Kathmandu Valley itself was overcast with low clouds. Capt Shah, with co-pilot Sophia Singh, was the first off the domestic apron with a Buddha Air Beech 1900D that was also taxiing out for a Mt Everest sightseeing flight.

On board Agni Air’s German-built Dornier 228 with the call sign 9N-AHE (’Hotel Echo’), were 11 passengers. Five were Nepalis, four American, one British and one Japanese.

The plane took off, and made a standard instrument ‘Igris-1 Alpha’ departure, involving a climbing circle overhead, then heading north east. Despite the heavy rain and turbulence, the pilots must have been encouraged by Lukla reporting good visibility and high clouds. But 30 miles out and cruising at 12,500 ft, one of the generators on board packed up.

Capt Shah told Kathmandu air traffic control he was heading back, but didn’t at first tell them about the generator malfunction. As long as the engines are running, the plane can keep flying. But generators supply power to cockpit instruments, and Dorniers have a backup generator and also a standby battery pack.

Somewhere between the time that Capt Shah turned back and followed a 20-mile arc to intercept the approach to Kathmandu runway 02, the back-up generator also quit. With intermittent battery power, and steering only by compass, Capt Shah seems to have decided to head to Simra.

There are conflicting reports about whether he informed Kathmandu about his decision, or whether he told other Agni Air pilots on his company frequency that he had decided to make an emergency landing in Simra. Piecing together initial sketchy evidence, aviation sources say Capt Shah knew he would not be able to make the VOR-DME approach to Kathmandu without his distance measuring and directional equipment in the cockpit.

But even to land in Simra, he would first have to break through cloud in order to get some ground references. It appears that is what he was trying to do, descending steeply, trying to get visual with ground. The nature of the impact site near a school in Shikharpur of Makwanpur, a crater 10 metres in diameter, and the altitude of the crash (1,700ft), seem to corroborate this.

A lot of the details will have to come from the inquiry commission that has been set up by the Civil Aviation Authority of Nepal (CAAN) and questions will obviously be asked about the state of maintenance of aircraft in domestic operations. A Dornier is built with triple redundancies for most systems, and especially for onboard electrical supply.

Capt Shah, disoriented without instruments in a white-out, seemed to know as he dived to get below the clouds that this was his last chance to find an airport to land. His last words to his fellow Agni pilots over the radio were: “Bye bye.”