Will someone explain this "fuel bug"

[MET?]

I've heard a lot of reference to the 0.16 fuel bug. I hear people talking about it all the time, but what it is? How does it affect my rockets? Can it be utilized/avoided?

[John Bedlam]

The fuel bug comes from a wrong calculation of the amount of fuel used by engines. For example, if you burn at 50% power, you only use the equivalent of burning at 25%, and at 10%, you only use what you'd normally use at 1%. So your stages last a lot longer if you can use minimal thrust. This means that any ship that you build now that seems to do what you want will possibly not work in 0.17. You can mostly avoid it by only burning at full thrust, and I believe there's a workaround plugin.

[MET?]

I see...I assume this applies to all the engines? Not just the ones that were added in 0.16 and the new fuel tanks as well?

[John Bedlam]

I do believe so. In the video Scott Manley did on it, he uses a one-man pod and an old tank and engine.

[semininja]

This bug applies to any engine which has Isp values instead of a fuel consumption parameter.

[quarthinos]

And when the devs were in the code, they also tracked down some other fuel consumption bugs. So all the fuel consumption logic was rewritten for .17.

[rkman]

Bug report and Harvester's explanation here:

[0.16] [MAJOR] Fuel consumption is wrong at low throttle

http://kerbalspaceprogram.com/forum/showthread.php/15196-0-16-MAJOR-Fuel-consumption-is-wrong-at-low-throttle/

Unofficial bug fix here:

[0.16.0] Hacky Fuel Consumption Bug Fix

http://kerbalspaceprogram.com/forum/showthread.php/16439-0-16-0-Hacky-Fuel-Consumption-Bug-Fix

[PakledHostage]

Thanks everyone for the explanations! I only first tried out v0.16 late last week so I missed the initial discussions about the fuel bug. When I orbited my first rocket in v0.16, I ended up with more fuel than expected left over. I attributed the difference to overly conservative assumptions... Maybe they weren't so conservative after all?

But something else I did notice during my own test firings was that the new Isp system seems to result in a variation in fuel flow vs. static pressure rather than thrust vs. static pressure (either could vary with static pressure to yield the Isp values quoted in the VAB).

Intuitively, it makes more sense to me that thrust should vary with static air pressure due to back pressure on the engine's nozzle, while fuel flow would be constant for any given throttle setting. Can anyone confirm this, or is there another physical explanation for why it is the way that it is in the game?

[Markus Reese]

Dang! Now many of my designs might cease to work later on since I do alot of sub 100% thrust maneuvers in my flights. Hmm... Well, moar boosters I guess.

[TOROWY]

ya this bug has kinda kept me from playing since all my rockets are pretty much cheating.

[MET?]

ya this bug has kinda kept me from playing since all my rockets are pretty much cheating.

I wouldnt let it stop you from playing...Just do flight tests at 100% throttle...

And if you want to make sure your 0.16 rockets will be able to achieve certain feats, just calculate the delta-v. I believe I've read that delta v required for the mun and return is 7600 km/s.

[Antsiscool]

I wouldnt let it stop you from playing...Just do flight tests at 100% throttle...

And if you want to make sure your 0.16 rockets will be able to achieve certain feats, just calculate the delta-v. I believe I've read that delta v required for the mun and return is 7600 km/s.

You mean 7600 m/s.

[vexx32]

Haha, if it's 7600km/s, you're probably heading for solar escape xD

[bsalis]

And when the devs were in the code, they also tracked down some other fuel consumption bugs. So all the fuel consumption logic was rewritten for .17.

Anyone know if the problems involved how fuel tanks drain? Since they sometimes drain in odd, unbalancing ways. Some designs of mine seem almost random in how the fuel drain works.

[CrashnBurn]

Anyone know if the problems involved how fuel tanks drain? Since they sometimes drain in odd, unbalancing ways. Some designs of mine seem almost random in how the fuel drain works.

I don't believe so. I believe it was posted that they reduce the amount of fuel used during reduced throttle in two separate places in the code. So they cut fuel use, say, 50%, then cut it 50% again. Just an oversight that has been discovered and will be corrected in .17

[xorvious]

Anyone know if the problems involved how fuel tanks drain? Since they sometimes drain in odd, unbalancing ways. Some designs of mine seem almost random in how the fuel drain works.

I believe that was mentioned in the thread linked earlier where Harvester talked about the bug. While looking for it he found a number of issues with how the craft decides where to pull fuel from and basically re-wrote the entire fuel source system, so it sounds like many things got fixed.

[quarthinos]

He did rewrite the fuel source system, but I'm not sure it will fix balancing issues. Keeping the center of mass/thrust in the correct place requires more logic than just generating a list of fuel sources.

[Ziff]

Anyone know if the problems involved how fuel tanks drain? Since they sometimes drain in odd, unbalancing ways. Some designs of mine seem almost random in how the fuel drain works.

All HarvesteR has said was that he went through and found a bunch of bugs in the fuel code. Since that was a previous complaint, we can only hope that it was part of the fix. I think the only time they drain incorrectly is when they are laterally attached, directly, without the use of radial decouplers, because of the parent/child rule that fuel tanks follow. Hopefully that got squashed though.

[Znath]

That might explain why some of the time I have a lot of extra fuel...

I think my lander currently will work even when this is fixed. But some of my more minimal landers will probably have to be updated in 1.7 that kinda sucks.

My current setup worked in 1.3 and is just slightly modified for 1.6. So I think it'll work mostly once this bug is fixed. I'm going to just enjoy the current bug while I can

[Clouds]

There's another bug whereby fuel use for an engine is halved if the fuel source goes "past" another engine, but is not directly connected to that engine, yet must be connected by a "node" to whatever that engine is on. Coincidentally, this is a common setup for efficient rockets, where all the fuel is fed from outer stacks (with their own engine) to inner stacks (also with engine). Except now, a setup like that is "cheating" as well...

Yes, it's been reported. I hope it's solved with the .17 overhaul. I hope .17 comes out soon! This whole fuel use fiasco seems like it could use a .16.1 hotfix.

I've built a rocket that uses not even 1/20 of a 200L tank to visit the Mun, Minmus, and return to Kerbin with a powered landing, running on the biggest 1200 thrust engine...at full throttle.