Pelican’s Perch #60:V-Tail at Flight Level 270

WARNING: The techniques and procedures described in this column can definitely be hazardous to your health and life. I do not encourage anyone to fly at these altitudes unless they have experience in an altitude chamber, have AT LEAST the equipment described, and know the subject and their limitations thoroughly.

Mountaineers call altitudes above 25,000 feet "The Death Zone," and they're not kidding. They are superbly conditioned athletes, very well-acclimated to very high altitudes, often spending weeks there to get that last little "edge" of altitude tolerance. Deaths still occur to the best of them. Any time they do not use oxygen at those altitudes, brain cells are dying. And yes, I've been told I've spent too much time at high altitude without a mask!

Aside from oxygen deprivation, there is also some slight risk of "Decompression Sickness," also known as "the bends." This is the same problem divers face when they have been too deep, too long, or come up too fast. A rapid climb to high altitudes may allow some of the trapped gasses (mostly nitrogen) to expand into bubbles in the blood stream, and those bubbles may get caught in the narrow places, especially in the joints. This can be very painful at the least, and fatal, at worst. Those in poor health, or the obese, are at the most risk. Pre-breathing oxygen, slow climbs, and intermediate periods of level flight will prevent all but the very rarest problems with DCS.

It's a hostile world, up there. I have explored it and my own limitations slowly and cautiously over the past two years, as research for this column. That included two flights with several people present, all on full-time oxygen.

The Great Circle distance from Camarillo, California, to Ada, Oklahoma (ADH), is 1,107 nm. It is a fair challenge to make that distance nonstop in a Bonanza without tip tanks, and unless there is a strong tailwind, it takes a fair number of techniques and tricks to do it. There is also R-2501, an "always hot" restricted area, so the 18-mile deviation north of the direct track to bypass that adds another nautical mile to the overall flight. Direct HEC (Hector) VOR, direct ADH was my rough plan.

I had done it easily once with my airplane (a 1982 V-35B, Serial D-10402), when taking it to Tornado Alley Turbo for the installation of their turbo system. As I recall, I landed with just over 10 gallons of fuel remaining. That is my personal minimum, but only under very special conditions. To "push it" that far, I need excellent weather at the destination, and the final hour must be within easy gliding distance of good airports. The logbook entry for that flight shows a flight time of 5.96 hours, and the remarks are:

I log time from the Horizon Electronic Tach, which essentially records engine run time, so the airborne time would have been about five minutes or so less.

For the fun of it, I wanted to see if I could duplicate that previous flight, this time with the turbo. Checking DUATS, my heart sank, for the winds were not at all favorable, at any altitudes. There were either direct light crosswinds, or very light headwinds. On another whim, I decided this would be an excellent opportunity to try a long flight at FL290, a most unusual altitude for a V-tail. So I punched in "direct" and took a peek.

If the DUATS computers were surprised, they didn't show it, because they spit out the following flight plan, with RNAV waypoints off various VOR/DME stations along the way:

The weather was forecast "magnificent" (official term) all the way, but the flight data and fuel burn were downright sickening. 99 gallons of fuel required? My airplane only holds 82.4 gallons! Even with every ounce of it usable (notwithstanding the lawyerese in the POH), that's a no-go. But, wait. The parameters for that were based on flight at 210 knots TAS (true air speed), and 15 GPH, or about 75% power - lean of peak (LOP), of course. I preloaded DUATS with those numbers right after I got the turbo, and they're not really very close, I've just been too lazy to change them. In reality, I commonly burn 17.4 GPH, for 85% power and 200 knots anywhere above about 12,000 feet, to 210 knots or more at 20,000 feet. I'm usually a "go fast" kinda guy, and that power setting and speed suits my purposes, most of the time. But it clearly wouldn't work on this flight, unless I planned to run dry well short of ADH, and glide the rest of the way in.

Pull a Lindbergh?

What would happen if I pulled a Lindbergh, and pulled the power back? That would reduce the fuel flow a lot (in percentage terms), while reducing the true air speed only a little, thus increasing range. It was also apparent that DUATS was telling me to expect about a 10-knot headwind, which pretty well killed my chances of making the long nonstop. I thought I'd try anyway, and just see what I could do. There are many good fuel (and food) stops along the way, with Albuquerque an easy stop if things didn't look good. That became my official destination, and we'd see how it went.

I've had the turbo'd Bonanza to FL240 and 250 a number of times on research flights and a few cross-country flights, but had never gone above that, and had no idea of what the numbers would be. WWII pilots (young and robust) routinely operated unpressurized aircraft up to 35,000 feet for many hours at a time, so I knew it could be done with the proper equipment and a healthy individual. I'm very healthy, but I am 63 years old, and 40 pounds overweight, so the usual safeguards are even more important.

The Right Stuff

Pulse Oximeter ...

In my opinion, no one should attempt serious altitude flights without some sort of electronic monitor that shows the percentage of saturation of oxygen in the blood, in real time. Dr. Brent Blue of Aeromedix first made the little Nonin "pulse oximeter" device widely available to pilots several years ago, and in doing so, did an enormous service to all of us. At first, the company that invented and manufactured it - Nonin (which is derived from "non-invasive") - didn't want to sell it to pilots, possibly for fear of liability. They put up barriers by requiring a doctor's prescription. No doctors (to my knowledge) would write the Rx, also for fear of liability. As far as I know, Brent was the first to bypass all the fear by first becoming a dealer, then selling the device with the prescription. Thanks Brent, we need more people like you!

To their credit, Nonin ultimately recognized the error of their ways, and now offers a non-prescription version of the unit (the "FlightStat") intended specifically for pilots. The Nonin FlightStat is a little cube a bit more than one inch on a side, that clips onto the end of a finger. You can find an excellent report on this little box, complete with pictures, right here on AVweb. (For the actual manual for the unit, see http://www.mhoxygen.com/images/EDS-D1-manual.pdf)

Hospitals have been using these subminiature pulse oximeters for some time, replacing much larger and more expensive electronics that require power from the wall. All these magical devices work by shining a couple of beams of light from one half of the sensor through the tissue to the other half. From the differences in the light, the electronics determine the color of the blood, and from that, the oxygen saturation level, displaying it digitally. It's fairly accurate, leaving only the question of what is "normal" saturation. Brent makes it simple. He suggests first checking for your normal saturation at the altitudes where you live, and using 10 points below that as a safe lower limit. I run around 97% at sea level, so about 87% is my personal minimum. With experience, I've found this works well for me. I feel better at the end of a long trip when I've used oxygen, and I don't think I have any problems, at least none I admit to. When I get below 80% for very long, I can start to notice symptoms of hypoxia and slight discomfort.

The Nonin pulse ox uses two AAA batteries, which last quite a long time, perhaps several dozen flights, or several months. The batteries may last a bit longer if removed between uses, if you can remember to do that. In my opinion, the NONIN and a set of spare batteries are mandatory above 10,000 feet for any significant length of time, for me. It is astonishingly quick to respond, showing changes in blood saturation within a breath or two. I was amazed to find that talking while breathing oxygen (cannula or mask) drops the blood oxygen level very quickly! Even a couple exchanges with ATC will do this. For some women I know… well, let's not go there.

... And O₂ Gear ...

I run a rather small 22 cu. ft. portable tank, filled from my own two big welding oxygen tanks and the usual cannula, with the superb little EDS 1 electronic oxygen regulator sold by Mountain High.

This device senses the very small fluctuations of pressure in the oxygen supply tubes from your breathing, and puffs just the right amount of oxygen for the altitude into the cannula at the precise point when you start to inhale. The rest of the time, no oxygen is flowing, a great savings. Mountain High claims this will reduce the requirement for oxygen by up to 90%, and while that may be a bit optimistic, it's really not too far off the mark, for me. I used to make the long haul between Southern California and Seattle a lot, and would need a fill on each end, often as much as $50 a pop. Since airline pilots can be banned from the brotherhood for so much as actually buying a copy of USA TODAY, my soul was in torment every time I forked over good money, just for something to breathe. Once I began using the EDS, I routinely used much less than of a tank for the same trip, often doing two and three round trips before needing a fill (note this was not at FL250, but more reasonable altitudes in the mid-teens).

[Dr. Brent Blue has taken all this a step further, and developed a device that combines the blood oxygen sensor and something like the D-1. The blood oxygen sensor tells the black box what the saturation level is, and the black box automatically increases or decreases the total flow of oxygen, while still pulsing it to occur at just the right moment. Pretty slick. I've used it, and it works well. Last I heard, Brent is making it smaller and less expensive, so look for it soon.]

Once the investment is made in the EDS, a "trans-filler," and a pair of welding tanks, the use of oxygen becomes much easier and painless (even for an airline pilot), and I normally put it on climbing through about 5,000 feet, if I'm going anywhere near 10,000 feet or above. For this long flight at very high altitudes, I started breathing oxygen from the Aerox mask as I taxied out.

... And a Mask (Not a Cannula) ...

I'm also fairly conservative about the use of cannula. Conventional wisdom holds that they are good up to FL180, and I used them up to that (and higher) for years - until I got the oximeter, and saw some real numbers! They just don't do well for me at maintaining a blood oxygen level of 87% or better above about 15,000 feet or so. I haven't seen them do very well for anyone else, either, regardless of condition, weight, or ego. There really is a sharp drop-off, too. Let me explain. If you check your oxygen saturation at the elevation where you live, you'll see 95% or better. This doesn't drop appreciably until rising above 10,000 feet, where it starts dropping very sharply. In other words, it is not linear.

What's that? You there, in the back? What happens when you live above 10,000 feet? Go away, I'm busy, here. Move, maybe! Some people…

Similarly, when I'm on oxygen (cannula) at and below about 15,000 feet or so, the oxygen level stays up to 90% and better pretty well. But at 17,500, I find it very difficult to keep my oxygen level above 80% with the cannula. That's that drop-off, again.

It is useless to kid yourself, thinking this won't happen to you. Even if you're in very good physical condition, you MAY be even more susceptible to hypoxia than some fat old ex-airline pilot who writes on the Internet!

During one of my research flights with Doug Ritter, I took a strong liking to the Aerox oxygen mask with the internal electret microphone. It fit me very well, did better than the other masks we tested, and I found the microphone by far the best of the bunch that we tested for his article. (Yes, the picture there is me, at FL240, with my Aerox.)

When the test was done, I asked for a discount for the "now-used" mask, and held onto it, mumbling something about prying it from my cold dead hands with blue fingernails. Aerox gave me a small discount, so I was happy. Airline pilots like discounts second only to freebies. I've used it a number of times since, and won't go above 15,000 feet for any length of time without it. Using it, I had been to FL240 several times, and to FL250 once without problems, but this flight to Ada would be my first foray above that.

... And a Backup O₂ Supply

One final piece of equipment that made me a lot more comfortable going so high. I was concerned about a possible malfunction of some part of my oxygen system, so I have a completely independent supply that is good for long enough to set up an emergency descent, and keep me conscious for it.

This is a green high-pressure cylinder like a large CO2 bottle, about six inches high, with a screw-on cap and regulator valve, and a primitive rubber mask. For storage, the cap is screwed on very lightly, barely to the point of resistance. When needed, the cap is given a couple of quick hard turns, and oxygen flows immediately. One hand will be required to hold this over the nose and mouth, but I think it's an excellent backup device, and I feel a lot better having it along.

Oodles of Data ...

Here's a Microsoft Excel graph of the flight, as recorded by the JPI EDM-800 engine monitor.

Click for higher-resolution graphic

That chart looks like a chaotic mess, doesn't it? There are six hours and ten minutes of data at six-second intervals, 3,654 data sweeps, all jammed into one graph, and there were a LOT of changes during the flight, so it's even more confusing than usual. For that reason, I'll break the same chart down into six individual flight portions, and we'll see if we can sort it out a little more clearly. My thanks to Microsoft for the magnificent Excel program, without which we'd not have these beautiful graphs. I cuss 'em for Windows, but a couple of their Office programs are just phenomenal.

Please take a moment to get an overall idea of what you're looking at. (It'll help a lot if you look at the high-resolution version.) I converted the time to elapsed time from the JPI readout of GMT, and that is the "x axis" across the bottom. So, the entire flight is portrayed in time from left to right, with each individual trace rising and falling with the passing time. Only the EGT and TIT are measured against the left side (y axis), all other parameters are measured against the right side of the chart. The scales have been adjusted so that everything shows on the same chart, with less interference. A couple of the parameters have been multiplied or divided by 10, so that they will plot on the chart with everything else. Fuel flow, for example, is in gallons per hour, but is multiplied by 10 to make it expand, and show in a better location. It is also shown as a heavy red line to make it stand out as one of the most important parameters. In most cases, the absolute numbers are not as important as the changes, relative to each other.

You may also note that MAP (Manifold Pressure) varies a lot, which is very unusual in my airplane, because I'm always at full throttle, about 31.0 inches MAP. In this one, I'm still at full throttle the entire time, until just a few minutes before landing. On this special flight, I was either at such a high altitude that the turbo couldn't maintain sea level MAP, and/or I was so severely leaned out that there wasn't enough energy left in the exhaust stream to drive the turbo.

Takeoff and Climb to 17,500

This chart is expanded horizontally, and only shows the first 40 minutes of flight:

Click for higher-resolution graphic

One of the interesting things about the recording feature of the JPI EDM-800 is that you can't get away with anything! Everything is recorded, and there's a reason for everything that happens. That's what makes it such a valuable tool. Someday, we'll have airspeed, altitude and heading included in the data stream, and it will serve as a flight recorder! It's pretty close as is, and as we see more and more of these recording instruments in more and more airplanes, they'll become very valuable accident investigation tools, even if they do reduce accidents.

Note the CHT values at the start? Remember, CHT is plotted against the right vertical axis, so all the CHTs pretty well start around 60F, or ambient temperature (OAT) that day. The CHTs are slightly above that, and the EGTs start out at about 500 to 800F. What's going on, here? Well, the JPI is hooked up through the avionics bus, and doesn't start recording until I turn the avionics master on, several seconds after the start, and the JPI itself goes through a self-test that takes perhaps 15 seconds. So by the time the JPI starts recording data in my airplane, the EGTs have gotten a fair head start, and the CHTs have started a small rise, too.

The runup (marked) shows clearly, with a double fuel flow pulse, corresponding to the two functional checks of the prop at 1700 RPM. RPM does a double dip, and there is a corresponding rise in the MAP, although the second pulse is not as strong, probably because I didn't wait for the RPM to drop off as much.

Takeoff roll starts at 00:06:30, and full power is applied before 00:06:48. 18 seconds? No way, Jose! I take the power up much more quickly than that! A quick check of the raw data for manifold pressure at that point shows, at six-second intervals, 15.1, 19.0, 26.7, and 33.0 inches. I don't know the delay between the actual data measurement and the data burst, but I'd guess I started up with the power just before that 19.0 measurement was taken, and got it to the full 33.0 inches very shortly after the 26.7 was taken, for something just over six seconds, which is about right. "Snapshots" at six-second intervals is the minimum interval allowed by the JPI, and while there have been a few times I've wanted a faster sampling rate, for the most part it's about right.

The redline MAP is 31.0 inches, so 33.0 is a bit of an overboost, but this is common on most turbos that are controlled by using engine oil for control of the turbo system. We've watched that on the test stand, and it's harmless. It drives the CHT up a tiny bit faster, but by the time the airplane has enough airspeed for cooling, the CHTs have not become critical. For those who aren't comfortable with this momentary overboost, it's easy enough to pull the throttle back to the redline at 31.0 inches MAP, and add it back later.

As the engine oil warms, and the MAP drops back, the fuel controller reduces the fuel flow accordingly. The reduction to 2500 RPM (for noise) right after gear retraction also reduces the RPM of the engine-driven fuel pump, and this reduces the fuel flow a bit, too. These effects contribute to making this a very easy turbo system to run, it pretty well takes care of itself.

Takeoff was on Runway 25 at Camarillo, about 10 a.m. That's obviously the wrong direction, so I asked for an early left turn for a downwind departure. That was approved, and I turned very early, out over the huge farmlands to get headed in the right direction.

Since fuel burn was absolutely critical, I elected to climb lean of peak (LOP). This is a perfectly valid technique, and will save fuel, but we do not recommend it as a routine maneuver. With the rapidly changing altitude, changes in speed and climb rate, ATC, and initial maneuvering, there's usually too much going on to add to the workload. The warning system built into the JPI makes it much better as that system will start flashing if any pilot-set limits are exceeded.

We call that rapid transition from ROP (Rich of Peak) to LOP "The Big Mixture Pull," or "BMP" for short. I just pulled the mixture knob back from 31.0 GPH to 15.8 GPH over about two or three seconds. For the next minute or two, the CHTs stayed rock solid, with the hottest at 353F. After a little fiddling around, I settled on 14.2 GPH as a good setting.

Note the inversion in the OAT? 61F at sea level, and 92F at about 6,000 feet? That rise in OAT may have affected the other engine parameters a bit more than usual, even climbing at 130 knots indicated.

At about 23 minutes, I noticed the CHTs had dropped a bit, so I increased the fuel flow to about 15.2 GPH. The TIT and all EGTs jumped up in response, and the CHT climbed back to about 370F over then next minute or two.

At about 32 minutes, note the fuel flow starts dropping off? Look at the MAP, too. I'm guessing (based on time and about a 500 fpm climb) that this was at about 16,000 feet, and the exhaust at this very lean mixture no longer had the energy needed to drive the turbo fast enough to maintain sea level power. Normally, at very rich mixtures, this turbo will hold sea level power to 20,000 feet and above, and the airplane will climb at 1,000 fpm to that altitude. The hundreds hand on the altimeter normally runs in perfect unison with the sweep second hand on the clock. Very nice to see a V-Tail do that!

But this was a fuel-saving flight, not a "go fast mode" flight. At 14.2 GPH, horsepower was 212, while the normal climb would have been 300 HP. The climb rate suffers, but the airplane goes "downrange" almost as quickly, so little time is lost.

At about 37 minutes, I ran the fuel flow back up to 15.2 GPH, which brought the MAP back up a bit. Finally, almost 40 minutes after takeoff, I leveled out at 17,500 feet, my initial cruise altitude. During the climb, I picked up radar advisories from SoCal Approach, then transitioned to Los Angeles Center, while tracking directly to Hector, using GPS.

The estimated fuel remaining at Ada, according to the JPI and GPS, was a -9.0 (minus 9 gallons). I very nearly threw in the towel on the nonstop right there.

Up to FL230

This chart shows one hour and 14 minutes of flight:

Click for higher-resolution graphic

The first part of this chart reflects a bit of indecision on my part. At this point, I knew that making the nonstop was nearly hopeless, so at one point in cruise at 17,500 feet, I got disgusted, and just shoved the fuel flow up to 17.0 GPH, for my usual "go fast" mode, truing out at 212 knots TAS, and about the same ground speed. MAP was back up to 30.5", RPM 2500, and the hottest CHT (#3) was 375F. I began planning lunch in ABQ, then remembered I had a nice sandwich, a granola bar, and a couple of Diet Cokes in the little cooler. Yum, better'n any restaurant!

But it was only just past 11:00, so I manfully resisted, and began thinking of my intended climb to higher altitudes. At my request, ATC converted radar advisories to IFR, and with no further ado, (eat your hearts out, East Coasters!) cleared me direct HEC, direct "Will Rogers," with a climb to FL230. He did specify that he wanted me OVER HEC to avoid the restricted areas to the south, and I refrained from telling him I could skirt that area far more accurately than he could portray it. I'd been aiming for HEC anyway, and it only added one mile to the trip. I'd like to think that even airline pilots see logic every once in awhile. I pitched the nose up to climb at 130 knots or so, and up we went, still LOP, of course. I made no changes to the power setting, and expected to see CHTs rise, due to the loss of some cooling airflow.

This is important, and it's one of the reasons we don't encourage LOP climbs! That loss in cooling airflow can have a major effect on CHT, and very high temperatures can result. It appears that I pulled off about 2 GPH to allow for this (remember, when LOP, leaner is cooler).

At about that point, I got into the discussion about the restricted area with the controller, very friendly, and also a comment or two about a Bonanza going to 230. My response was a snappy, "You ain't seen nuthin' yet, I'll be requesting 290, shortly!"

When I first plotted this portion in Excel, I was quite startled at the raggedy pattern during the climb. Could I have done that bad a job? Well, in a word, yes. But it's interesting, nonetheless.

At time 00:58:06, the oil temperature takes a sudden small jump upwards. That is probably where I pitched up and started the climb, at the reduced airspeed. At that same time, I pulled off 2 GPH, and got mildly distracted by ATC, and by checking my blood oxygen.

At 01:02 or so, I realized CHTs were down, EGTs were down, fuel flow was down to 13 GPH! This got my attention, as I thought at first I'd had a malfunction of some kind, and it didn't make much sense. One of the things that happens to the mental mindset after you've flown the turbonormalizer for a time is that you're used to just setting full throttle and forgetting it, and the MAP just sits on 31.0", essentially "forever." So I wasn't really looking at the MAP in my scan.

I shoved the fuel flow up to over 16 GPH, to see what would happen. Sure enough, everything headed up again, and when I finally looked at the MAP, it was up to 29.0. I checked the throttle, still wide open, and I remember thinking, "critical altitude," but paid no more attention. Over the next few minutes, the same thing happened again, and again I was startled to see EGT, CHT, MAP, and fuel flow dropping. Again, I added fuel flow, and this time thought it through, and realized what had happened.

The next couple of spikes in the fuel flow reflect where I played around with it for a bit, even increasing the RPM to 2662 at one point. I wanted to see what those power settings were doing to my range, and predicted fuel remaining. At about 01:27:00, I suddenly noticed that predicted fuel remaining at Ada was PLUS! It was only a gallon or two, but things were improving! Maybe, just maybe, with a little shift in the winds, and a little help, I might get that up to my required 10 gallons, and make it after all! I also began regretting some the of the fooling around I'd been doing, and the short period at high-speed cruise, wondering if I'd shot myself in the fuel supply department.

I fooled around a bit, and ended up with 11.2 GPH, 21.8" MAP (wide open throttle), and 2173 RPM (a rough stab at 2200, I think). My indicated speed was down to just over 115 knots, and a true airspeed of right at 170, with a predicted fuel of 4.0 gallons in Ada. As you can see on the right end of the chart, that gave a nice stable situation.

That also put me close enough to noontime that the rumbling in my tummy and the tantalizing thought of that sandwich overcame me, and I decided to eat. I don't recall trying to eat while wearing a mask before, and it's a lot more difficult that with a cannula. A mask should be pretty secure to prevent leaks, and mine was. I disconnected the strap that goes around the nape of my neck, and managed to lift it just enough to sneak the end of the sandwich into my mouth, and surveyed the world from 230 with great contentment as the first bites hit bottom. Life is good.

At least, until I checked the oximeter, and discovered I was below 80! Eating drops the blood oxygen level even more than talking! Steady deep breathing in between bites brought the oxygen level back up nicely, above 90, but it did slow down my sandwich consumption. I can also state unequivocally that sandwich consumption reduces intelligibility on a mask mike. There must be another corollary to Murphy's Law which states that "ATC will always call right after you take a big bite, when wearing an oxygen mask."

(Picture me, chomping desperately, sandwich in one hand, the other hand holding the mask on in the absence of the strap.)

(Geez, those guys are so IMPATIENT!)

(More chomping in silence, faster.)

I'm not only unreadable, I'm unprintable, but we finally get it sorted out, and I explain what had happened, much to the amusement of everyone in the airspace. Damned smart-mouth airline pilots…

Anyway, then there were comments on what type of airplane I'm flying, and the controller remarks he's never worked a V-Tail at 230, before.

Ever the smart ass, I shot back:

There was a most satisfying silence as he thought that over.

He was silent again, no doubt wondering what he'd done in a previous life to deserve a 170-knot airplane cluttering up his airspace full of 500-knot airplanes.

That REALLY sparked some comments.

Finally, Up to FL270

This chart displays only 53 minutes of the flight, edge to edge. The OAT was at +10F at FL230, and dropped to -8F at 270, with the climb taking only 12 minutes, an average of 333 fpm. Not too shabby!

Click for higher-resolution graphic

For the climb to 270, I first pushed the RPM up to the full 2700 (actually, 2160 to 2689, from the raw data). That large increase in engine speed caused the fuel pump to turn faster, which ran the fuel flow up to 18.8 (the sharp spike near the left edge of the chart). I knew that would be too high, so I leaned it back to 16 GPH, and that worked pretty well, giving 25" MAP, 2700 RPM, and CHTs topping out at exactly 400F on the hottest cylinders.

Oddly enough, the #6 CHT started up with the others, but within the first few seconds, seemed to sag behind, staying quite cold, relatively. At FL230, the "spread" between the hottest and coolest CHTs was only about 30 degrees, but by the time I got to 270, the spread was over 50 degrees.

The scale on the chart makes this look more serious than it is, but I'm still curious about that one. The airplane went in for annual at GAMI, and I'll be curious to see if they saw anything that might have caused that.

At this altitude, I had the pulse oximeter on my finger full-time, and was scanning it every few seconds. I also had the EDS-D1 in "R/M" mode (see manual), and was watching my oxygen bottle pressure very closely. This was NOT a time to relax, or enjoy the view, although it was spectacular. My focus was on my oxygen level, my engine, my oxygen level, and my engine. Oh, and my oxygen level. Comments to ATC were short and to the point, and I was VERY aware of what I was feeling - which was quite normal, at first. To keep the oxygen saturation up above 85%, I had to take very deep, very steady breaths, and make a point to blow out as much as I could. At first, I was able to hold it above 90%, but it was getting more and more difficult.

Predicted fuel remaining at Ada was up to 7 gallons, but at this point, I was not worrying about that. In the back of my mind, I was thinking that a long, slow descent for the final hour might put me over my required 10 gallons, but I'd worry about that later.

The good news was that the airplane was running great! 290 would have been no problem for the airplane, and I'd bet it would go a lot higher. If I can get the oxygen problem licked, I may someday do a column with "370" in the title!

At about the 02:20:00 point, I experimented with various power settings, to see if I could improve that fuel remaining, but nothing seemed to work very well. The fuel remaining was dithering, and dropping, showing only 3 or 4 gallons remaining. Knot by knot, the effective headwind was increasing, too, up to 18 knots, at one point. The best I could get on fuel remaining was at 10 GPH, so I left it there, still with the faint hope the winds would change, and that a long descent would help enough.

It became more and more difficult to hold the oxygen level above 85%, and I only rarely saw 90%. Even a quick exchange with ATC would cause it to drop into the low 80s, and it would take a dozen breaths to get it back to 85% or better. This was too much work! So I asked for a descent to FL250, and that was quickly given.

Backing Down

This chart is only 43 minutes of the flight, so it too is expanded horizontally. I left the fuel flow at 10 GPH, and the fuel remaining stayed around 5 gallons.

Click for higher-resolution graphic

This is a relatively boring chart, and I haven't marked it up much. But there is one interesting point - note the fuel flow is very steady across the chart at 10.0 GPH, which means 149 HP. But look at the CHTs? They take a rather pronounced drop during the descent, even though the MAP rises half an inch, and the OAT comes up, too. Then the CHTs rise again, after leveling at 250. Everything else stays pretty steady.

The answer, of course, is cooling airflow during the 2,000-foot descent from 270 to 250, which took eight minutes. Even that small rise in airspeed (perhaps five or ten knots) showed up on the CHT. The predicted fuel remaining jumped up a bit, too, but I knew that was phony.

Upon arriving at 250, my oxygen saturation was back up in the low 90s, and I figured I'd finish out the flight there if the oxygen level stayed up, and my supply held out. It wasn't the 290 I'd hoped for, but I felt I'd gathered some useful data.

Time to Pull the Plug

This chart represents one hour, 50 minutes of the flight, so the data is a bit more compact, horizontally.

Click for higher-resolution graphic

Much to my surprise, as the minutes at FL250 went by, my blood oxygen began dropping again! What was this? Plenty left in the tank, all settings were right, but I was, once again, hovering around 85%, and nothing I did seemed to raise it. Once again, I made sure the little emergency system was right there, ready for use, and reviewed what I needed to do to get on it quickly. Headset would have to come off, mask would be ripped off without regard for the straps, screw the bottle into the regulator hard, and put it up to my face. I noticed that it seemed to take more effort to figure numbers out, and suddenly when I shifted in my seat, I realized there was numbness in my left leg!

My reaction was instant. I put the gear down, and started down, HARD, at 5,000 fpm, speed increasing from 110 IAS to 155, the max gear speed, all on the autopilot, because if I did lose it, I wanted the airplane on the autopilot to get me down. While pushing the autopilot pitch button with one hand, I was pushing the mike button with the other, telling ATC I had a problem with oxygen and needed an immediate descent to 15,000 feet. Had he come back with anything but a clearance, I would have declared the emergency, but he sensed the (deliberate) urgency in my voice, and instantly cleared me down. A quick glance at the oximeter showed 77%, and I started to use the breathing pattern that had worked for me so far. It came back up slowly, and within a minute, it was back in the 90s, long before I reached 15,000 feet. The leg numbness went away within seconds of leaving 250, too. As best I can tell, I lost the first 6,000 feet of altitude in one minute, 48 seconds, without touching the engine controls.

(From the raw data, the last OAT at 0F was at 03:25:18, and the first reading of an OAT of 23F was at 03:27:06.)

When things go bad at those altitudes, they go bad very quickly!

I was prepared to cut the power for the descent, but once I saw how fast I was able to come down with the gear down and resulting slow rise in the speed, I thought I'd just leave things alone. As I recall, I broke the descent at about 180, pulled the gear up, and made the rest of the descent a bit more sedately. This shows up in the much more gradual slope of the OAT rising to 35F at 15,000 feet.

Look at the rest of the traces on the chart. Note how precipitously the CHTs drop with the huge increase in airspeed, and no change in power. Even the EGTs dropped a bit, which surprised me.

During the final stage of the descent, the MAP rose enough to drive the fuel flow back up a little, from 10 GPH to 12.6 GPH. Remember, at no time during this flight did I ever touch the throttle, it was wide open the entire time. Any change in the MAP is due to other factors, mostly those driving the turbo.

It took me about five minutes to catch on to the increased fuel flow and remember I'd had 10 GPH set, so I pulled the mixture back to 10.0 GPH. I was surprised to see the predicted fuel remaining had risen to 8 gallons! The wind had changed enough during the descent to nearly "create" enough fuel to make the long nonstop! If only I hadn't wasted some fuel, early on!

The fleeting thought "Maybe landing with 8 is okay?" I swatted that thought like the bug it was.

Alas, it was not to be, as the wind steadily began shifting around to my nose again, and I began the search for a good fuel stop. Hobart (Oklahoma) looked good, and the JPI/GPS told me I'd have about 13 gallons left if I landed there. It was the only one around with "Fuel, 24 hours," so I figured it was a good bet.

Alert readers may spot the little downward spike in most of the parameters just after time 04:57:06. What was that? Here's the raw data for it:

Note the first 18 seconds (three time snapshots) are normal, but on the fourth row, the fuel drops from 9.1 to 8.6, to 8.4 six seconds later, and to 7.9, before turning right around and going back to 9.0.

That's where my right tank ran dry, of course.

I had flown the first hour (exactly) on the left tank, then used the right tank for 3:57, for an average burn of 10.43 GPH, which seemed about right. Knowing I'd climbed that first 40 minutes or so at 17 GPH (LOP), I now knew I had to have about 24 gallons remaining, and the JPI confirmed that. Finally, the left fuel gauge showed just over , another confirmation, as I know from experience my gauges are exactly correct.

With all this, I had ample evidence that I truly did have 24 gallons remaining, and this satisfied one of my requirements for pushing range to land with only 10 gallons remaining. At this power setting, it was about an hour's fuel, which even satisfies the FAA rules!

Descent and Landing

So there I was, fat, dumb, and happy, at 15,000 feet, resolved to land at Hobart, burning 10 GPH, and poking along at about 170 knots TAS.

Click for higher-resolution graphic

Knowing I now had nothing to lose, I thought I'd see how much leaner I could go. I leaned, and leaned, and leaned, and finally found just a trace of roughness at 6.5 GPH, or so, or 97 HP. I enriched just a hair, and the engine was baby-bottom smooth at 6.9 GPH, or 103 HP! This wasn't even enough power to stay in the air, I had to begin a gentle descent to stay above 110 IAS! This is just about 250 F lean of peak EGT. Yes, the engine WAS running, but just barely.

It bears repeating, this is still at full throttle! This demonstrates the extreme power control from mixture alone.

6.9 GPH didn't even give enough power to make the airport, so I had to boot it up to about 12 GPH, and finished with a fairly normal 45 minutes on the end of the flight.

The long slow descent was uneventful, except the headwinds continued building, and I ended up landing at Hobart with just 11 gallons remaining. Yes, I did reduce throttle to land! I did have one scare, as I turned short, very high final, I realized there was NOTHING at this airport! No airplanes, no cars, no sign of human habitation or activity, just 30 knots of blowing dust! The one redeeming feature of this desolate airport was a gleaming new 24-hour gas pump, self-service. What a relief!

As I taxied up, I wondered, "Will my credit card work?" When it did, and I heard the machinery come to life, I wondered, "Is there enough gas?"

There was, and I finished up with the 20-minute flight to Ada. Sorry, no data for that leg.

Be careful up there!