Tonight I did my best to follow a real world 747-8 flight (CLX7606 / LX-VCC) from Luxembourg to Rio de Janeiro (ELLX-SBGL).
I loaded 90% fuel (all fuel tanks full, except the stabilizer tank and the reserve tanks, which are unusable at the moment, due to bugs), and I only loaded enough cargo to balance the center of gravity properly, for all phases of flight. The fuel and payload page looked like this:
That fuel should have been more than enough to reach the destination, since the 747 has a 16 hour autonomy, and even with 90% fuel it should fly over 14 hours, especially with no passengers and limited cargo. This flight was 10.5 hours long, and the map confirmed that:
So basically I should have had about 4 hours of extra fuel. In theory.
I departed 30 minutes after the real world flight, directly from the runway, in order not to waste additional fuel and time by taxiing. I replicated all the maneuvers of the real flight, I did all the altitude changes they did, and I kept the same Mach speed (.84 to .85). I began the flight climbing to FL290, 20 minutes later I climbed to FL310, 2 hours later to FL320, 6 hours later to FL340, and 30 minutes later to FL360.
I was plagued by the usual autopilot bug, that prevents it from being able to keep the altitude and speed, and makes the aircraft oscillate up and down like a roller coaster. I fought that by using the FLCH mode every time the altitude oscillations got to an amplitude larger than +/- 300 feet over/under the set altitude, which were triggering ATC to warn me about not keeping my altitude. But I doubt that autopilot bug is enough to explain the huge fuel usage.
In the end, as the fuel remaining reached about 7%, I decided to divert to Salvador. But even that was not close enough. I had to re-divert to the Arcaju airport. While doing that, on LNAV, the autopilot did a 360 turn, for no reason. When it attempted a second 360 turn, I enabled heading select towards the runway.
It was too late. Because of that 360 turn the remaining fumes in the tanks have been exhausted, the engines flamed out, and I crashed in the ocean, just 4NM from the runway. The correct decision would have probably been to divert to Recife. And if I was a real life pilot I would have probably been fired, if I survived, for not diverting much earlier.
At the time of the flameout I have flown for 9 hours and 8 minutes. I have flown just 4300NM, and the 747 has in theory an autonomy of 8000NM. Even with 90% fuel, due to the 3 unusable fuel tanks, it should easily reach at least 7000NM, especially with no passengers and limited cargo. It barely reaches 60% of that.
At this point I fail to see how you can do a flight lasting more than 10 hours in this aircraft, without cheating and adding more fuel during the flight.
Not sure exactly what caused this high fuel usage. I noticed the calculated gross weight doesn’t decrease during the flight, maybe the simulator does the performance calculations based on that, which means the fuel usage doesn’t decrease as it should, when the weight of the fuel carried decreases.
And this is just one of my many attempts to fly the 747 properly. I consistently experienced this high fuel usage issue on each of my attempts.
The gross weight calculation issues also seem to affect the aircraft pitch attitude. As you burn more fuel in the 747, you should be able to climb higher, which is desirable as the jet engines are more efficient at high altitudes, where the air is colder. Instead the aircraft pitches up too much, up to 10 degrees, if you attempt to reach high altitudes, like FL430. Even when you used most of the fuel.
So most likely that’s the mechanism that causes the high fuel usage: the calculated gross weight doesn’t decrease as fuel is used, the aircraft pitches up more than it would if the gross weight was calculated correctly, increasing drag, which reduces efficiency, increasing the fuel usage. Either that, or they didn’t model correctly the increased efficiency of the engines at low air temperatures. Or both.
And the problem doesn’t seem to be 747 specific, the A320 has a similar behavior. The previous weekend I attempted a flight in the A320 from Seattle to Ireland (KSEA to EINN). That’s a 3900NM flight, so a bit over the 3500NM autonomy of the A320. But with no cargo and passengers on board, and full fuel tanks, I should have been able to do that easily.
From what I heard from a Ryanair 737 pilot, that’s usually how they fly the newly delivered 737s from Seattle to Europe: empty, with a refueling stop in Ireland. So I tried to replicated that in the A320, which has similar autonomy to the 737.
I started from the gates at KSEA, departed, climbed directly to FL390, which probably is a bit too high for an aircraft full of fuel, but I was too lazy to do step climbing that particular day, especially since I’m not very familiar with the A320. But it seemed to have no trouble reaching that altitude. I cruised at Mach .78, the normal cruise speed.
The autopilot in the A320 was much better than the one in the 747, it had no trouble keeping the altitude and speed, and followed LNAV properly.
Long story short, I wasn’t even close to being able to reach Ireland. I considered diverting to Iceland, but even that seemed risky, although the total distance would have been just 3200NM, so well within the 3500NM theoretical range of the A320, especially with no passengers and cargo.
In the end, to be on the safe side, I decided to divert to Greenland / BGSF, so just 2500NM total flight distance. About 70% of the 3500NM autonomy. I landed there with just 4% fuel remaining, after a 6 hour flight, so it was a good decision. I would have never been able to reach Iceland. But normally the aircraft should be able to fly for at least 7 hours, especially with no passengers and cargo, so clearly the fuel usage was way too high.
So the fuel issue of the 747 seems to affect the A320 as well, and probably for the same reason: the calculated gross weight didn’t decrease at all during the flight, which was probably the cause of the high fuel usage.
After fully refueling in Greenland at BGSF, I departed for Bucharest, Romania, LROP, with the same empty aircraft, no passengers or cargo, a planned 6 hours 2700NM flight, well within the A320 range of 3500NM, especially with no passengers and cargo.
I may have wasted a bit of fuel by starting my descent about 150NM too early, but other than that, it was a flight similar to the previous leg, and I should have had plenty of fuel upon my arrival at Bucharest. Instead, this is how the fuel looked after I landed at Bucharest:
Only 106 KG of fuel left, so about 0.6%. About the fuel capacity of two small cars. Yes, if I was a real pilot I should have probably been fired for not diverting to refuel somewhere along the route, but I wanted to do a thorough test of the capabilities of the MSFS implementation of the A320.
For me right now the high fuel usage is the most frustrating part about flying the airliners. Especially in the 747, that has 3 unusable fuel tanks, as a bonus. The autopilot issues in the 747 are a close second, in my top issues list. The low FPS when start with long flight plans, especially in the cockpits, would be the third, but it’s something I could live with.