Friday, November 12, 2010

Enough is Enough


Aviation has lots of folksy sayings, most of which, in my opinion, are more often wrong than right. One of the more familiar is, “The only time you have too much fuel is when you are on fire.” It’s clever and it’s funny, but it isn’t true.

Of course we always want to have plenty of fuel, and figuring out what “plenty” is is the point of this post, but that doesn’t mean more is always better. More is certainly not better after a partial loss of power, for instance, regardless whether that loss was caused by a twin engine aircraft losing one engine or a blown cylinder or burned exhaust valve in a single. Nor is more the answer to trying to out climb trees on takeoff, to carry a load of ice, to try to top severe turbulence, or to fly out of down drafts from mountain waves. Fuel has weight, weight carries a performance penalty, and there are any number of situations where performance is more critical than endurance. So it’s just kind of stupid to say, “The only time you have too much fuel is when you’re on fire,” and, what I really don’t like about it is that it rationalizes careless flight planning.

One of the more familiar discussions pilots of high performance aircraft typically have is, how much fuel do you want to land with? What are your personal minimums for fuel? At what point does fuel get so low that you say, “Enough. I’m landing at the nearest suitable airport”? I remember having this discussion with several other pilots years ago when I was flying Citations and Falcons for a company in New Hampshire. The Chief Pilot, Stu Jones, was there too, but didn’t get involved in the conversation. That didn’t surprise us because he was not a big talker but when he did say something it was often quite clever and usually very funny. Finally someone asked him what his number was, and he said, “I always try to land with some.” That was funny. And, of course, directly to the point. Landing with “some” is what it is all about. One of the sayings in aviation that I do like is, “Takeoffs are optional, landings are mandatory.” Once airborne, sooner or later the aircraft is going to land, with or without the benefit of power. It’s always better to do it with power.

I started out in aviation in gliders and my first solo was in a glider. Gliders are great fun and a great way to get into aviation, and they are neither scary nor dangerous just because they don’t have engines. They do have “power", but it isn’t in the form of an engine, it is in the form of excess speed and altitude. From Day One in gliders you learn to enter the pattern with more altitude than needed to simply glide to a landing, and with more speed than needed to simply stay in the air. Both are then gradually dissipated, using spoilers, to maintain the desired glide path. You “add power” by reducing the amount of spoilers, and you “reduce power” by increasing the amount. The rough equivalent of a dead stick landing in a glider is arriving over the airport with minimum airspeed and altitude. In that case, you better get it right the first time, just like any dead stick landing.

The reason I started thinking about this again was because I noticed several ads for fuel monitors included something to the effect of being “GPS ready” or “GPS capable”. What, I wondered, did GPS have to do with fuel flow? So I looked into it a little further, and apparently what GPS does for fuel monitors is provide estimated fuel remaining at destination. I thought that was why we flight planned in the first place—so we would know estimated fuel remaining on arrival, and I thought that was why we kept a flight log enroute, so we could see how that estimate was holding up? If the trend was seriously negative, we could do something about it before we got to dry mouth time. There’s nothing wrong with having a running estimate of fuel remaining at destination, but it told me that if pilots were willing to spend extra money for it, it was because they didn’t have a very good idea otherwise. Any kind of a decent flight plan, one that includes winds aloft forecasts along the route of flight, is going to be better than a simple GPS estimate based on current groundspeed and the assumption that it will stay constant for the remainder of the flight. Now, a flight management computer, or any kind of computerized flight log with position input is the best of both worlds, but I don’t see where having a simple fuel remaining estimate based on current winds tells you a whole lot. It would be pretty disconcerting, for instance, to see that number start out big, and then watch it get smaller and smaller as the headwinds increased, or the tailwinds decreased, along the route of flight. Some information isn’t always a good thing.

I also know that, as a practical matter, most reciprocating aircraft have their tanks topped off after every flight. And I know that topping off the tanks is not always done just to simplify flight planning or to pay homage to the “You never have too much fuel unless you’re on fire” rule. Avgas and water don’t mix, the water settles in the low points and causes all kinds of problems, and full tanks minimize condensation. Also, most general aviation fuel tanks are rubber bladders and fuel keeps them from drying out. So topped off fuel tanks may be a practical fact of life for most general aviation pilots. But that doesn’t eliminate the need for good flight planning and for maintaining an enroute log, and it certainly doesn’t eliminate the question, “How much is enough”?

We know what the regs say (FAR 91.151 and FAR 91.167), but we don’t worry about them too much because, again, the tanks are always full and we don’t try to stretch our range, so we know we will always have enough, if we are VFR, to fly to our destination plus 30 minutes of fuel at normal cruise, 45 at night, and if we filed IFR we need that plus enough to “complete the flight” to the destination—that means descend, approach, and land—and after that fly to an alternate, as required. Assuming you do have only these minimum amounts of reserve fuel, what does that do for you?

VFR, not very much: 30 minutes of fuel is barely enough to see on the gauge, and who knows how much of that is water, gunk, or even usable as you turn from base to final? Forty five minutes at night isn’t much better. I’m not a big fan of VFR at night anyway, it’s hard to see the clouds, and pilotage is a whole lot more difficult at night—better hope that Garmin doesn’t take the night off. So day or night, these reserves are minimal and don’t allow for any contingencies, which is what the reserves are for.

The fuel requirement for IFR is a little better: enough fuel to actually fly to your destination, shoot an approach, do a missed approach, climb back to a safe cruising altitude, proceed to your alternate, land and still have 45 minutes of fuel at normal cruising speed to use, presumably, for your approach and landing there. (You don’t always have to have an alternate, but you’re nuts not to.) That’s quite a bit more fuel, but more to the point, there is a real purpose to it: fly to your destination and land; if unable, fly to your alternate and land there. I don’t really know where the FAA came up with “45 minutes at normal cruising speed”, but these rules go back a long long ways, and they provide the starting point for that question, “How much is enough?”

The NBAA (National Business Aviation Association) has established a minimum fuel profile that is frequently cited in ads and articles on business aircraft: “Range based on NBAA IFR reserves,” is a typical footnote, for instance. The NBAA profile includes fuel for a missed approach, climb to 5000 feet, hold for five minutes, climb at best rate of climb to the optimum altitude to fly 200 nm to an alternate at long range cruise, descend to sea level and land with 45 minutes of fuel. In other words, it flight plans fairly accurately and realistically a typical missed approach at the destination, diversion to the alternate, descent and landing. It is very useful for comparing maximum range of one aircraft to another, and it is useful as a model for realistic flight planning, but it doesn’t really address the issue of how much is enough when you finally do land; it assumes that, ultimately, the FAR 91.167 minimums are adequate. So it’s great for comparing range for one aircraft to another, but not so great at telling you whether you would actually want to fly that aircraft that far.

There are two other interesting lessons implied by this profile, one is that there is no fuel allotted to try the approach again—it assumes you try once and if you miss you go as directly as possible to your alternate and you land there—and it assumes you always will have an alternate. My view on this, discussed in some detail in a previous post, “The Other Part of Flight Planning,” January 2010, is that that is exactly what you should always do: shoot one approach, if unable to land divert to an alternate, one with a precision approach and weather forecast well above minimums, and land there. Too many general aviation accidents occur on the second and third try at the destination; professionals just don’t fly that way and professionals don’t have those kinds of accidents.

So we still haven’t completely answered the question, “Is 45 minutes of fuel based on normal cruise fuel flow adequate or not?” and the only way, I think, that question can be answered is to ask, “What is that reserve fuel for?” The first thing it is for, the most basic, is to make sure there is always “some” fuel remaining at the end. That means, we want there to be enough fuel in the tanks that the needles aren’t resting on the bottom of the gauge—we want a little day light there. Again, who knows what’s in the bottom of those tanks? We want some fuel in the tanks when we finally land, and we want to know for sure that it is fuel and not water, sludge, rubber bits, dirt or anything else that won’t burn. (ATA had an engine flame out on a 1011 just before takeoff. Further investigation revealed that the fuel filter was clogged with bits of cloth and foam padding—a mechanic, working inside the fuel tank, had left his cushion behind. Probably not going to happen to Cessna 152, but strange things do find their way into fuel tanks.)

Let’s look at both 30 and 45 minutes of fuel at “normal cruising speed”, whatever that means. I happen to have five different pilot operating handbooks at home, and the only one that lists a “normal cruising speed” is the Piper Turbo Twin Comanche and that is also its highest power setting (other than max continuous). That’s okay if Piper wants to call it that, but it doesn’t necessarily make it “normal.” (“Old” Piper, that is. New Piper, with the Malibu in any case, another one of the manuals I happen to have on hand, goes back to calling it High Speed Cruise.) So, what is normal? The one your instructor told you to always use? The one that gives you a nice round fuel flow number like 10 gph? The one you happened to flight plan for that day? As far as I’m concerned they are all normal, so take your pick: the higher the power setting, the more fuel will be required. What about the airplane that normally burns 10 gph? Thirty minutes of fuel would be five gallons, typically split into two tanks: two and a half gallons per tank. That might get the needles off the peg, but not by much. Forty five minutes is only a little better, three and three quarters gallons per tank.

If we look at the Piper Turbo Twin Comanche performance charts, Normal Cruise is supposed to be 22.6 gph. Thirty minutes of fuel would be 11.3 gallons, 5 2/3 gallons per tank. Forty five minutes would be almost 17 gallons (16.95) or about 8 ½ gallons per tank. I can probably see that on the gauge, but I wouldn’t like what I saw. And remember these numbers come from using the highest normal power setting; economy or long range power settings would yield even less.

So I think we can safely say that the FAR minimums are not enough. How much more would we like to have, and, again, for what reason? First, we have to agree that the professional way of operating a flight, if we can call it that, is assumed here, or else we’re back to, “You can never have too much fuel.” Let me explain: If we agree that the way to flight plan and operate a flight is to always have a good alternate, and to proceed to that alternate after one try at the destination, then the amount of fuel we would like to have above and beyond that necessary just to fly that profile is a legitimate and serious question. If we don’t agree with that premise, if we want to have the option to try the same approach again, hoping for better weather, or the chance to try to fly the approach better and not have to go around the next time, or to try a different approach, or to hold for awhile hoping the weather improves, or if none of that works, to try the nearest airport to the destination and if we can’t get in there either try another one, then we’re back to taking as much fuel as we can and hoping that will be enough. So if you don’t agree with the basic premise, you can stop reading right now because the rest is irrelevant to you. And good luck.

But let’s assume you have flight planned carefully and you know exactly how much fuel it should take to go to your destination, shoot an approach, do a missed approach, proceed to your alternate, shoot an approach there and land with 45 minutes of fuel remaining. Is that enough, or do you want a margin above that, and for what reason? Well, I think it is obvious that you want more that just enough, if for no other reason that these are flight planned numbers, not real world. In the real world, fuel burns are often higher than book values and airspeeds slower. In the real world you don’t always get the routing you requested or the altitude, the winds aloft are sometimes less favorable than forecast, thunderstorms and icing conditions force detours, and congestion results in vectoring and holding. You don’t often get all of these adverse factors on every flight, but almost all flights will be affected by one or more of these factors, and they all require more fuel. So again, how much?

There are, I think, three ways to look at this. The first is to try to come up with a reasonable and conservative number to account for several, but not all, possible adverse factors. (If you have a really bad day and all of these negatives combine, your enroute log will say to you, “This isn’t going well. You need to land now.”) The second way is to increase the minimum fuel by a given percentage, and the third way is to increase the fuel in terms of additional flying time. All three work, and I’m not sure any one is better or worse than the others; for sure the first one is the most complicated, but it is possibly also the most accurate. Each will be more or less conservative depending on the assumptions made: How many adverse factors do I expect, and what will be their effect?

The airlines, under FAR 121, are required to increase their enroute fuel by 10% when operating internationally. That can be a bunch of fuel for a long flight. (The airlines have found a way around this in the form of a re-release flight plan, which is way too complicated to get into here, but it shows the limitation of the simple percentage method: too much extra fuel on long flights, not enough on short ones.) For the Piper Turbo Twin Comanche, a long flight might require 80 gallons of fuel from takeoff to destination. Ten per cent would mean adding another eight gallons, or about 30 minutes, plus or minus a few minutes depending on the power setting. That seems pretty reasonable to me: 30 minutes of deviating, of groundspeeds slower than flight planned, of getting vectored around, of having to fly lower or higher than desired to stay out of ice, or some combination of all of the above, sounds pretty reasonable to me. (Remember, this is fuel above the 45 minutes required by regulation.) If you want more, or experience shows that that still leaves too many sweaty palm situations, increase it. But have a number, keep track of it enroute, keep records and review them to see how your number is working out, and amend that number as real world experience dictates.

Why do we do this, if, as is usually the case, the aircraft is already full of fuel? What’s the point? You get the fuel you get, some days just a little more than what you need and some days a lot more, but what’s the point of figuring a specific fuel load for that flight that day if the tanks are already full? I think there are several good reasons, but perhaps the best is that you may find out that, even with full tanks, you don’t have enough fuel. That is, what looks alright based on rules of thumb and past experience isn’t enough when you take a more accurate and complete look at it: takeoff and climb, enroute, descent, approach, miss, climb, cruise to alternate, descend again, shoot another approach, land, add 45 minutes of fuel to that (FAR 91.167) and then your own personal number on top of that. And if full tanks won’t let do all of that, then a safe outcome is no longer guaranteed. It may seem like a lot of fuel, and most of the time, when everything goes pretty much as expected and you make a routine arrival and landing at your destination, it probably is, but sooner or later it’s not going to be.

So that’s good reason number one. Good reason number two is that it forces you to have a plan, one that you have assured yourself you have enough fuel for, and one that virtually guarantees a safe outcome. Any fuel above and beyond that required for the plan carries a weight penalty, but at least you know what that weight is. Good reason number three is that it might get you to thinking about ways to not top off the aircraft after every flight without encountering condensation problems or bladder drying. That’s a subject for another time, but suffice it to say that a regularly flown airplane, stored in a heated hangar, is a good candidate for fueling as required prior to flight.

My chief pilot was right: you always want to land with some fuel in the tanks. But topping the tanks off and relying on rules of thumb won’t guarantee that; the only way to guarantee that is to have a plan and stick to it. And fires have nothing to do with it.