Running Out of Options

One of the things I’ve learned in my years of flying is to make sure my airplane never gets somewhere my brain hasn’t already been. We stress the importance of getting a good weather/NOTAM briefing before flight, but sometimes the equally important task of keeping up with the weather during the flight gets lost. Back in the day, it was often difficult on an instrument flight to get controller permission to go off frequency to get weather updates from Flight Service.

Today’s technology (like the portable ADS-B In device I carry in my headset bag) allows us to learn of changes to the weather situation during flight not only without leaving the ATC frequency, but in a textual format that saves writing it all down with one hand while trying to fly the plane with the other. Another thing I’ve learned back in my youthful flying days (a duration that should probably be shortened now in my sixth decade of flying) is that there were really good reasons behind the 12-hour crew duty day limits the Air Force imposed on its fighter crews. Tired brains have too much trouble keeping up with—no less staying ahead—of the airplane. Most important, I’m reminded of the saying that an aircraft accident is what happens when a pilot runs out of altitude, fuel, and ideas all at the same time. It’s important to remember that “fuel” and “options” are pretty much interchangeable words. No fuel, no options.

Not to Plan

Last November, a pilot flying from Anniston, Alabama (ANB) to Quincy, Illinois (UIN) got bitten by a major change in the forecast for his planned destination and the flight ended in a fatal accident. The NTSB preliminary report is out (CEN26FA049), but what that report doesn’t talk about is a dramatic change in the forecast for UIN after the flight took off. The preflight forecast for good VFR conditions on arrival changed shortly after takeoff to a forecast of low IFR, but it appears the pilot was unaware of this change for his planned arrival some four hours later. In the end, instead of an easy night visual approach, the weather was below minimums at UIN when the pilot attempted his approach, resulting in a missed approach.

The pilot then attempted to divert to another airport about 25 nm away, but the weather there went below minimums while the pilot was trying to perform a visual approach and the airplane crashed just short of the runway, killing the pilot and seriously injuring both passengers.

When the flight departed ANB at 16:42 local for the planned 4.15-hour flight, the UIN TAF called for 2,500 feet broken with 6-mile visibility. Given the wind, the RNAV(GPS) RWY 17 approach with LPV DH of 349 would seem to be the proverbial “piece of cake” (or as we called it in the Navy, “Cake, Piece of, 1 each”). So from a weather standpoint, it would seem that there was nothing to worry about—the 1-2-3 rule easily met and no alternate required. Adding 45 minutes to the ETE gives a fuel requirement of 5+00. Based on an estimated 11 gph fuel flow from the PA-28R-201’s POH, that would be a minimum of 55 gallons. This Piper Arrow III has a fuel capacity of 77 gallons (72 usable)—plenty of fuel for this trip with a generous reserve. However, from the NTSB report:

According to fueling documentation, after landing at ANB the pilot initially requested that the airplane be fueled with 40 gallons of fuel (20 gallons per wing fuel tank) but he subsequently amended his fuel request to only 36 gallons (18 gallons per wing fuel tank). According to the employee who fueled the airplane, the fuel level in each wing tank was below the 25-gallon indicator tab after the requested 18 gallons were added to each wing fuel tank.

Pushing the Limits

Even at the 25-gallon tabs (50 gallons total), they were about five gallons short of the FAA’s 91.167 requirement. Why would they short-fuel the plane for a long, night, IFR cross-country flight? The basic empty weight for an Arrow III is around 1,800 pounds, and max gross is 2,750 pounds, leaving about 950 pounds of useful load. Three average-sized adults and bags is probably about 650 pounds, leaving only about 300 pounds for fuel, or about 50 gallons. That might well explain the decision to depart with as little as they did. But why push the limits like that?

My personal minimums for all flying is always landing with one hour of fuel, whether that be at my destination or, if an alternate is required, at the filed alternate, and to land short if at any time it appears I’ll land with less. If I’d been making this flight, given the limited amount of fuel to carry that load, I would never have taken off with the intention of going all the way to UIN. I would have landed for refueling about halfway there (the St Louis, Missouri, area presents itself about halfway with good fueling options). Of course, that would have extended their day (which started about 7 a.m. in Florida) maybe another hour, and even as the trip was flown, the pilot had been awake for around 17 hours at the time of the accident. Perhaps the pilot was trying to outrun the hands of time—rarely a race you win.

When the flight arrived in the UIN area at 0227Z (about four hours after takeoff), the reported weather available to the controller was already nearly 20 minutes old, showing the visibility having deteriorated to 1-1/2 BR. What was actually happening was rapidly thickening fog, but the weather system doesn’t call it “fog” until it goes below 5/8 statute miles, and the difference in terminology might create a failure to recognize the advancing threat. Also, that information did not include the fact that over the preceding 30 minutes the visibility had dropped from 10 to 4 to 1.5 statute miles. Further, what the controller (and apparently the pilot) didn’t know was that in the eight minutes since the last METAR the controller had received, the visibility had dropped further to ¼ FG with vertical visibility of 300 feet. The airport was already well below the minimums of 400-1¼ and by the time the approach was commenced, the ceiling had dropped further to 200 murky feet of vertical visibility.

It’s all pretty shocking given the TAF of 2500-6+ the pilot saw before takeoff. But what the pilot apparently did not know was that about two hours after takeoff, the new 00Z TAF came out with a radically reduced forecast for their arrival to an expected below mins 300-1/2 for their ETA. At this point, they still had over two hours’ of fuel worth of options, as opposed to having run one tank dry just after the missed and not much left in the other. The pilot then tried to divert to PPQ (which they’d overflown about 30 minutes earlier) but by then the weather there had dropped from clear and five as they went by down to 200-1/4—leaving no chance to get in on the LPV with published minimums of 250-1. He was able to see the airport lights from about 2,000 feet AGL, but lost sight when they descended into the ground fog and crashed.

You Must Monitor the Weather

Had they been monitoring the weather during the flight (including TAF updates), they would have learned that while the weather in their destination area was headed for the toilet, the weather in St. Louis (which they passed not long after that new TAF was issued) was good IFR at 800-5 and would hold well above ILS minimums for several hours. This was the time to exercise some preventive discretion, by diverting there to spend the night, heading for UIN in the morning when the weather would be 600-3 (not great, but well above the UIN mins). After that, they were beyond the point of no return.

The takeaway? We all know that when planning a flight, it’s better to be down here wishing we were up there rather than the opposite, and once in flight discretion is the better part of valor. The chain of events started with the very long day proposed (link number one), then the limited fuel (link number two), and finally the failure to stay ahead of the weather (link number three). There were three bad links in that chain. Maybe application of the “Rule of Three” might have cued the pilot to knock it off had he not been too tired to put the links together.

The following were actual weather reports for reference: