Human Factors: Cool as a cucumber

The first link in the accident chain was a good decision. Or, at least, a well-reasoned one at the time.

Despite a high overhead observer pattern, and two low passes over a grass-covered creek bed on an expedition to scout backcountry fishing locations in Georgia, the bush plane’s right main hit something hiding in the grass on landing. On roll out, it became clear that the monster 31-inch tundra tire was losing air.

So there they were, the pilot and his passenger, out in the middle of nowhere with a perfectly functioning airplane about to ground itself. Although the pilot had landed at this spot before, he wasn’t sure if there was cell phone service or not, and he knew that land-based access to the site was nil.

In his accident report to the NTSB, he wrote, “With the decreasing tire pressure and a few moments to make a decision, I considered the uncertainty of access to the location, unknown cellular service for rescue/assistance, and a potential survival situation.”

Weighing all of that, he makes the snap decision to get the hell out of Dodge while he still can, and quickly takes off again.

No, that’s not when the accident happens. The accident won’t be for another couple of hours.

The takeoff is fine. And once safely in the air, the pilot starts considering his next steps. First he sets his engine power aggressively to start burning off fuel, because you need extra fuel in an emergency landing like you need a hole in your head.

And then he starts making phone calls. Four of them, in fact. More about those in a moment, but first, some more about the pilot himself.

The Pilot

The pilot was a 48-year-old male. He was a member of the STOL Bandits group, and a frequent competitor (and winner) in STOL events. At the time of the accident he had just over 5,000 hours. He was fully current, with his most recent flight review being in an Airbus A-320.

Right. I might have forgotten to mention that he held an ATP certificate with type ratings in both the Boeing 737 and the Airbus, and that his day job was as a pilot for one of the majors.

Before the airline he served in the U.S. Air Force as a F-16 jockey. He saw combat in the Middle East, where he was awarded an Air Medal and a Bronze Star. He later became an F-16 instructor pilot.

The Airplane

The airplane was an interesting little beast. It was a 1950s-era Cessna 172 converted to tailwheel configuration and fitted with a six-cylinder Franklin 180-hp engine driving a controllable-pitch two-blade McCauley prop. It had a host of STOL mods, long-range tanks, aileron and flap gap seals, stall fences, domed cabin door windows, and more.

An archived Aircraft.com listing says, “this gorgeous, ‘no excuses’ bird had a fortune invested in modifications.”

The accident photos suggest the spending spree wasn’t over after the pilot bought it. The panel was beautifully remade after the sale, and documents show the airplane was IFR equipped and certified.

And, naturally, it was outfitted with modern coms that let our pilot make those four phone calls.

Formulating a Plan

The pilot calls the owner of a grass strip that he knows, but they decide after discussing options that the pilot should head for an airport with better services, just in case the worst happens trying to land with one 31-inch bush tire full and one an unknown percentage flat.

Next he calls someone he knew that flies a similar airplane with similar tires, a pilot who recently suffered a blowout on landing. They discuss at length what might be expected in terms of handling when the airplane touches down. From this, the consensus is that grass is better than pavement under the circumstances.

The third person he calls owned a business and had a hangar at Chattanooga’s Lovell Field (KCHA), a Class C airport about 20 miles northeast from the creek bed where the bad day started. This contact provided the 4-1-1 on emergency services there. An emergency landing at Chattanooga becomes the plan.

Still with fuel to burn, literally — in both the productive and in the scary sense — he makes a fourth call to Cleveland Regional Jetport (KRZR), a non-towered airport northeast of Chattanooga, to arrange for some practice approaches before the main event at Chattanooga. These are “roll-on” fly-bys over both the runway and the adjacent grass to begin to feel out the handling characteristics.

And no, this isn’t where the accident happens either. But we’re getting closer.

Sufficiently confident that he’s done all he can in the way of advanced prep, he heads to Chattanooga and declares an emergency with approach control. After being handed off to the tower, he requests, and is granted, clearance for multiple approaches to the grass areas adjacent to Runway 20 to determine the best area for the full-stop emergency landing.

The Plan Goes Amiss

Like Goldilocks and the Three Bears, he tests different patches of grass at the Class C airport. The first area feels rough on the roll-on fly-bys. The grass is also high and during two attempts at the spot he flushes birds out. A bird strike on landing is all he needs to add to his woes.

The next spot was recently mowed and long and wide and seems mostly smooth. He again does his signature pair of passes, then powers up for the final go-around prior to the emergency full-stop landing.

And that’s when the accident happens.

The Accident

Climbing at 65 mph, at 350 feet above the ground, the engine stumbles. I imagine the pilot’s heart does, too. The engine recovers for a second, then goes silent. The pilot is low and slow with no energy to manage. The view ahead isn’t pretty. He pitches down and executes a right turn through 135° to try to reach an area slightly behind him that looks better.

It’s not.

He comes down on uneven ground and smacks into a ditch just outside the airport perimeter fence. The time, from power loss to touchdown, according to the pilot, was somewhere between 10 and 15 seconds.

One of the many upgrades put into the airplane were four point restraints, which turned out to be an excellent investment, as neither the pilot nor his passenger were injured.

The NTSB

The NTSB would declare the cause of the accident to be a lack of fuel to the engine.

But the pilot didn’t misjudge his fuel burn-off. There was fuel in both tanks, in all the lines, and in the carb. No contamination was found anywhere in the excellently maintained system. Nor were there any mechanical failures. The engine performed flawlessly in all post-accident tests.

Instead, the NTSB said it was a clear case of fuel starvation caused by carburetor icing.

As a semantic reminder, fuel exhaustion is when you run out of gas, while fuel starvation is when you have gas but something prevents the fuel from getting to the engine — in this case a carb venturi packed like a snow cone cup full of ice.

Carb Ice 4-1-1

Carb ice is a repeat offender in GA accidents, despite the efforts of the FAA and pilot organizations to do something about it.

As a quick refresher, the heart of a carburetor is a venturi where air and fuel are mixed to become the vapor that powers our magic carpets. That’s all good and fine, but there are two incidental matters of physics at play that we need to consider.

The first is that the vaporization of fuel sucks heat out of the air. The second is that as air exits the pinch point of the venturi it expands and that causes it to cool yet more. This dual-action drop in temperature inside the carb happens in a fraction of a second, and if cold enough, any water vapor in the fuel/air mixture can freeze and turn to ice. That ice can actually build up on the interior surfaces of the carb just like ice builds on your wing in icing conditions aloft.

In this case, the NTSB report stated that, “atmospheric conditions at the time of the accident were conducive to serious icing.”

Which leads to the question of what are “icing conditions” for carburetors? And how do we know?

Unlike airframe icing conditions, there is no SIGMET for carb icing. Where did the NTSB get that info?

The NTSB investigators used a carburetor ice probability chart.

Now for those of you who are saying, “Carburetor ice probability chart? What carburetor ice probability chart?” I’m happy to report that there’re actually a bunch of them out there.

The first, and worst, is in the Pilot’s Handbook of Aeronautical knowledge, which is a simple boxy chart showing that there is a high potential for carb icing between 20℉ and 70℉ when the humidity is above 80%. While true, that’s not the whole story (and in fairness, the text teases it out a bit better, and there is a disclaimer in the figure’s caption saying “carburetor icing is possible under conditions not depicted”).

A much better icing probability chart appears in Special Airworthiness Information Bulletin CE-09-35. Don’t feel bad if you’ve never heard about it. SAIBs are issued by the certification branch of the FAA, so they don’t always cross the radar of pilots the way other information, like Advisory Circulars, do.

This SAIB was actually issued back in 2009 and, at that time, looking back over a decade of data there were 212 carb ice crashes, 13 of them fatal. Despite ongoing pilot education efforts by both the Feds and the Aircraft Owners and Pilots Association (AOPA), that accident rate remained stubbornly stable and is still — clearly — an issue today.

Anyway, this new chart featured a three axis graph of ambient temperatures, dewpoint, and relative humidity. Overlaid on it is the risk of carb ice by another variable we haven’t discussed yet — engine power settings — and this new risk graph compared glide and cruise power, quantifying the risk level over various conditions.

Carb ice risk goes up at lower power settings, due to the facts that the butterfly valve has restricted the venturi further and, with the engine at a lower setting, the air passing through the induction system is less heated — colder going in to start with.

Boldmethod created a very nice alternate version of this graph with the colors reimagined and more user-friendly (they always have the most splendid graphics).

As you can see, the yellow band of “serious” carb ice risk at glide power eats up a huge swath of the graph. Which is another way of saying there are a lot of risky combos of temperature and dewpoint that can snowcone-up your carb — and a good chunk of it is north of the PHAK’s 70℉, even with humidity down into the 30% range.

That’s a lot of flying days. Just sayin’

At the day and time of the accident the temperature was 77℉ with a dewpoint at 52℉. The NTSB investigators used the glide power in their analysis, as even though the pilot had just powered up for his go-around, the ice would have been building before that point.

Looking at the Boldmethod graph, you can see that the carb ice risk is smack dab in the middle of the serious zone with that temperature and dewpoint.

Analysis & Discussion

In his well-written and detailed accounting of the accident to the NTSB, not too long after it happened, it’s clear that the pilot was still completely stumped about what caused his engine issues. He reported, “the cause of the engine power loss is yet to be determined. No known mechanical failure has been found.”

He goes on to detail that the mags, mag wires, timing, and spark plugs all check out. The tanks had fuel, as did the fuel lines, gascolator, and carb. You can almost see him scratching his head as he writes.

Even after having time to process it, carb ice simply doesn’t enter his mind as a possibility. If he’d been a career airline pilot who suddenly decided to take up STOL (not that there’s anything wrong with that) this accident would be easier to understand. We might then assume that many years of jet time could cause carb heat to fade from the mind. But he was an active GA pilot and airplane owner. His logbook shows over 800 hours in make and model.

So what are we to make of a highly experienced aviator with military, commercial, and backcountry experience — who was intimately familiar with his airplane — being brought down by carb ice? And why is it that, even after the fact, something fundamental like carb ice didn’t enter his mind as a possible cause?

I suspect it’s because carb ice isn’t well taught. Even with colorful charts that show the wide ranging effect, no one seems to be using them, and carb ice awareness — as an operational risk management norm — seems lacking.

Many pilots still believe you can’t get carb ice when it’s hot out. So let me take a crack at it.

Forget the graphs, and the charts, and the probabilities and consider one simple fact and one simple formula. The fact is that the dual forces at play in the venturi can drop the air temperature inside the carburetor by as much as 70° Fahrenheit. So, as we know that water freezes at 32℉, with some kindergarten math we can determine when the air will be too warm for carb ice to ever form. Ok, let’s see here…32 + 70 = 102.

So that means even with ambient air temps at 102℉, you can get carb ice.

Realistically, that’s any time you fly. Every time you fly. It’s really not even a variable. All you need is a little water in the air, because the carb will almost always chill the air enough for snow cones.

The Takeaway

So what to do?

Well one simple thing is to use carb heat as an anti carb icing prophylaxis any time you are at glide power. Sure, carb heat reduces power some, but if you are at glide power you obviously don’t need full power.

At altitude, if you get ice, you can probably melt it off by pulling the carb heat, but it — pardon the pun — takes a hot minute to do so. If you are low and ice up, you may not have time.

The accident pilot did not pull his carb heat. But with the engine-out to in-the-grass time being so short, I doubt he could have cleared the carb even if he had done so. Not that there would have been any harm in trying.

But if he’d had his carb heat on during his low-power descent to his roll-ons, it’s likely the accident wouldn’t have happened. Or if it had, it would have been a different accident, one from the deflated right main tundra tire.

It’s a good reminder for all of us that for any landing — or landing-like operation (low fly-bys to check landing surface, etc.) — when the throttle comes back, so too should the carb heat knob.

The Numbers

Want to read more? Download the NTSB’s final report here or view the items on docket here.