There was an interesting conversation about fuel, maybe here or maybe somewhere else, in which precision of fuel management was being discussed. I believe the conversation started with whether or not one should run their tanks dry before switching. Without derailing the conversation, I run my aux tanks dry, at altitude, in level flight, and when I feel it's important to do so, to get every last gallon out of the tanks. It's safe to do so, and both the aircraft mfgr and the engine mfgr agree. In 28 years of flying, I've never had a problem with it. But of course, that's for airplanes I've either owned or knew well, and it's within my personal risk parameters. It probably doesn't fit every last example of flying machines ever made, so YMMV. Anyway, Deakin wrote on this subject for AvWeb back in the day. (He ran his tanks dry, too.)
One person, who no doubt believed he had a conservative view, felt that running a tank dry was just too risky. He said "If one or two gallons makes a difference in your flying, you're cutting it too close." On the face of it, that makes sense, right? But I disagreed. In fact my view is the polar opposite. I think that level of precision -- or at least attempting to attain it -- is the only way you
can operate safely when it comes to fuel management.
I provided an example of how this could be true. Background: I fly a light twin, and I have an electronic engine monitor (a JPI EDM-760) which, among other things, calculates fuel flow and total gallons used via transducers. I have vetted this unit over time and I know it to be accurate within 1 gallon when I'm getting down to my reserve. (I know this because I compare the fuel totalizer to my fuel receipts. The totalizer is pessimistic by about a gallon when I get down below 1/4 tanks on the mains.) I also know that when I use a timer, the gauges, and estimate when I need to switch the tanks, I leave anywhere from 1.5 to 2 gallons per side in the auxes. Because it's such a tiny amount of fuel, I really can't afford the risk of going "back there" to get that fuel except in a true emergency. In my light twin 2 gallons is what I factor for my taxi out fuel. It's not much, especially down low and at higher power settings.
So by estimating the switch, I'm costing myself up to 4 gallons, or so I must assume, out of the 84 usable in my light twin. That may not sound like much, but that's nearly 5% of my go juice.
So the example is this. Even in good day VFR conditions, I do not land with less than 16 gallons total in the airplane. 16 gallons is good to cruise for about an hour at about 75% power at altitude, or a little more down low at lower power settings (55% or so). So I flight plan down to that number. This goes for IFR as well: I need 16 gallons after landing at my alternate airport. Anytime an alternate is needed, the range of my little twin goes down pretty quickly. Instead of looking for a 500-600nm leg I might be looking more realistically at 400-450nm. That reserve plus the alternate eats up a fair amount of endurance.
When I say 16 gallons, that is THE number. Not 15.8, not 15ish. 16. If my planning shows it drops below 16, I must come up with a different option. I fly fairly long legs for a non-tip tank Twin Comanche, so that comes up every now and again. I think it's clear to see where I'm going with this: I can safely plan for that level of conservative precision when I do manage fuel down to the 1-2 gallon range, because I can reliably hit my already-conservative fuel reserve within a gallon. If I have to assume a 4 gallon error, that's now a huge hit on my reserves. I go from 16 gallons OB upon landing reliably (in fact it's actually 17, since the totalizer is a little bit pessimistic, but I never count that gallon) to as low as 12-13 gallons. That's not an acceptable reserve for my flying and my risk profile.
So that level of accuracy affords me the luxury of reliable, conservative fuel planning, and also maximizes the range and endurance of my aircraft.
Precision will pretty much always win when it comes to the big ticket items like fuel. Guesstimating your fuel burn, assuming you have lots of gas and far more than you need, etc. results in a hit on SA and sometimes leads to an eventual accident... maybe not today or tomorrow, but someday, because the true fuel status of the aircraft is not fully understand and
internalized by the pilot. "We have enough." Well, do ya? How much will you have exactly, when you land? All it takes is an unforseen diversion, followed by another unusual occurrence (perhaps a disabled aircraft blocking the runway at the second choice airport) to make the fuel situation very interesting indeed, all of a sudden.
So keep your precise fuel needs in mind when managing fuel. That's my view.
