Speaking of NASA a fair # of the "early" manual metric to imperial conversions had errors.
Seems I remember a Candair crash that was the result of a metric, imperial conversion fuel load error.
Yup - it was an Air Canada 767 on which the pilot muffed the calculation of the fuel load due to a metric/imperial error and an inoperative fuel quantity instrument.
https://en.wikipedia.org/wiki/Gimli_Glider
The plane had a double-engine
shut-down (not an engine
failure) due to fuel starvation near Winnipeg on the flight from Montreal QC to Edmonton, Alberta - which is quite the shortfall (more than 1300 km or over 800 miles).
The copilot knew the area and the pilot was an experienced glider pilot - which saved the lives of everyone aboard. The co-pilot recalled that there was an old RCAF base nearby at the town of
Gimli Manitoba - and they landed the airplane on it after gliding for quite some distance. The major problem was that the runway was being used as a dragstrip at that very moment and so the car race people had to be warned and evacuate the runway as the aircraft was on its (VERY silent) final approach.
The aunt of a friend of mine was a passenger on the
Gimli Glider and she told everyone that there was no panic, but that the cabin got
very quiet after both engines shut-down. The nose gear didn't lock-down due to low hydraulic pressure (all they had was the ram-air turbine pump but they weren't flying fast enough to get full system pressure) and so there was some superficial structural damage. However, the aircraft was repaired fairly soon and flown out and re-joined the Air Canada fleet. In fact, I flew on that very plane several times over the years before it was retired and scrapped at the Mojave boneyard.
The entire affair was a close-run thing, but nobody got seriously hurt and it was a he!! of a piece of airmanship - caused by a Grade IX arithmetic error and a failed cockpit instrument.
Pete