![An Embraer EMB-500 Phenom 100 taxis at London Southend Airport, which is covered in snow during blizzard. [Alamy]](https://www.flyingmag.com/wp-content/uploads/sites/2/2025/08/2DGBH31-scaled.jpg)
On a January morning at Provo, Utah, an Embraer Phenom 300 was towed out of its heated hangar onto a ramp whitened by falling snow. It was fueled, and the pilot and three passengers boarded.
It taxied to the runway and, some 40 minutes after it had emerged from the hangar, the Phenom began its takeoff roll. It rotated, began to climb, then rolled to the left. The wing struck the runway, pulling the airplane sideways. The fuselage came to rest largely intact, only its left front portion crushed. There was no fire. All of the passengers survived the crash—only the pilot died.
If you're not already a subscriber, what are you waiting for? Subscribe today to get the issue as soon as it is released in either Print or Digital formats.
Subscribe Now
The airplane was equipped with a flight data recorder. Its histories of speed, angle of attack, and bank angle seemed to confirm what National Transportation Safety Board (NTSB) investigators must have suspected from the outset: This was a case of a wing stalling prematurely because of ice contamination.
And there was something more. The pilot had selected wing and stabilizer leading edge deice as part of the pretakeoff checklist, but through some trick of memory or performance—he was narrating his actions to a passenger beside him—he had then unintentionally turned deice off again.
- READ MORE: Casual Decision Turns Deadly for Pilot
That any kind of contamination on the upper surface of a wing may reduce its maximum lift is well known. The effect is hard to quantify and to separate from that of leading edge ice accumulation, but even if the leading edges are clean, frost or frozen water droplets farther back on a wing compromise its lifting ability. It’s contamination of the upper surface that matters, when “upper” means the direction in which the lift force is exerted. Thus, T-tail stabilizers, although hard to inspect and clean, are in less danger from contamination than wings, because the consequential surface is their underside.
There was no direct evidence that the wings of the Phenom were contaminated, and there was some reason to suppose that they might not have been. The airplane had been in a heated hangar, the fueler reported that he noticed drops of liquid water on the wings, and the ambient temperature was not far below freezing.
On the other hand, the pilot had not, so far as NTSB investigators could tell, inspected the wings visually from the cabin as the POH suggests doing, and he had not chosen to have the airplane deiced and anti-icing fluid applied. There was, therefore, at least no evidence that the wings were not contaminated when the airplane took off. Their surfaces might, for example, have cooled just sufficiently during boarding and taxiing for water droplets to congeal or for falling snow to form a crust.
The role of leading edge deice, or rather the lack of it, was unclear. The airplane would most likely not have developed much leading-edge contamination during taxi and takeoff. If ice contamination was the reason for the left wing stalling, it was probably upper-surface ice rather than leading edge ice that was the culprit.
However, there is another aspect to this accident that pilots should be aware of. It is the role played by ground effect during the first moments airborne.
Pilots usually think of ground effect in terms of its reduction of lift-dependent drag, which results in floated landings and underpowered airplanes lifting off and then being unable to climb. But ground effect influences lift as well. In some cases, the maximum lift available from the wing diminishes slightly in close proximity to the ground. In others it may remain unchanged or even increase slightly. What is more significant is that lift increases more rapidly than usual in relation to angle of attack. Even if the maximum lift remains roughly the same, the stall occurs at a lower angle of attack in ground effect than out of it.
A classic instance of this phenomenon was the crash of a prototype Gulfstream 650 undergoing single-engine takeoff performance testing at Roswell, New Mexico, in 2011. The airplane dropped a wing and crashed, killing all aboard. Its behavior was very similar to that of the Provo Phenom, but at Roswell neither weather nor ice was a factor. The NTSB faulted Gulfstream for failing to take ground effects into account in its calculation of target pitch angles and maximum lift coefficients and in the programming of the airplane’s air data computer and stall warning system. The error in stalling angle of attack, the agency determined, could have been as large as 3 degrees.
Three degrees doesn’t sound like much, but in jets, where target pitch angles are used for certain maneuvers, it’s not a negligible error. The usual pitch angle for first segment climb in the Phenom is 10-12 degrees, but several witnesses, including the pilot of a Gulfstream V who watched its takeoff from the FBO’s office, reported that the Phenom appeared unusually nose-high as it began to climb. Indeed, the Phenom’s flight data recorder showed the angle of attack rising to 16 degrees and triggering the stall warning as the jet rotated. Angle of attack and pitch angle are not synonymous, but during takeoff, absent a strong crosswind, they are for all practical purposes the same.
Another small but not negligible factor is the acceleration produced by rotation. The Phenom experienced an acceleration of 0.1 G as it gained vertical speed, reflecting the fact that during a rapid rotation the wing’s angle of attack is briefly greater than it is once a steady climb has been established.
Thus, a combination of factors—wing contamination, ground effect, and upward acceleration—all pushed the wing toward stalling.
While taxiing, the pilot remarked to the passenger beside him, “Let’s hurry and get down there so we can get this thing in the air.” Witnesses reported the jet kicking up a rooster tail of slush as it accelerated. If he felt some impatience to get the airplane off the slushy runway as soon as possible, it’s possible that the pilot rotated a little earlier, with a little more vigor, or to a slightly steeper pitch angle than usual.
The NTSB blamed the accident on the pilot’s failure to deice the wings, but, as I noted, that conclusion is based on unproven suppositions. It’s possible that the accident could have occurred even without any contamination on the wings. The Roswell Gulfstream accident did. Most likely, however, there was some contamination, and it, combined with the subtle influence of ground effect, turned a slightly excessive pitch angle into a fatal error.
It’s perhaps noteworthy that the Gulfstream V pilot who witnessed the accident at Provo delayed moving her airplane out of the heated hangar because she wanted to expose it to the elements for as short a time as possible. She also delayed her departure to take advantage of a forecast window of improved weather.
The improvement was expected to arrive in 20 minutes.
This column first appeared in the August Issue 961 of the FLYING print edition.
