Pilots say stall may have brought down Wellstone's plane

Tony Kennedy and Paul McEnroe

Star Tribune

Published Dec. 29, 2002

Ray Juntunen flies every weekday morning into Eveleth-Virginia Municipal Airport as a pilot for United Parcel Service.

In the five years he has been at it -- mostly in a twin-prop, piston-powered Beechcraft Queen Air -- Juntunen keeps his airspeed at 120 knots when approaching the airport. When he has the runway "made," meaning he could glide in if both engines failed, he trims his speed to 110 knots.

This standard approach and the safety cushion it provides is a fundamental teaching from the earliest days of a pilot's training. And that's why aviators and crash experts around the country wonder why a similar-size plane carrying Sen. Paul Wellstone and his entourage was traveling as slow as 85 knots as it neared the Eveleth-Virginia airport Oct. 25.

At that dangerously slow speed, observers say, the plane's wings could have lost their aerodynamic lift, causing the plane to fall from the sky.

In aviation parlance, they say the airplane may have "stalled," a loss-of-lift condition that can happen even if the plane's engines and propellers are operating correctly.

"Eighty-five knots is way, way too slow," said Tom Kirton, former chief flight instructor at Embry-Riddle Aeronautical University in Daytona Beach, Fla.

Said U.S. Rep. Collin Peterson, D-Minn., who is a private pilot: "They should have been at 120 or 130 [knots]. That's the mystery: Why were they at 85 knots?"

The stall question lingers as the National Transportation Safety Board (NTSB) continues its investigation with computer simulations of the Wellstone flight to evaluate aircraft performance and the cockpit activities of Capt. Richard Conry and copilot Michael Guess. Both pilots died in the crash, along with Wellstone, his wife, Sheila, their daughter, Marcia Wellstone Markuson, and three Wellstone campaign workers.

Other pilots' theories

Although all the facts are not in and an official ruling on the cause of the crash is still months away, experienced pilots and veteran air safety investigators offer several possible explanations. One of the theories is that the pilots may have been distracted from the airspeed indicator while looking out the cockpit windows in search of the airport. The plane, a Beechcraft King Air A100, was fully loaded and may have collected some ice. If present, the icing would have increased both the weight and the aerodynamic drag on the airplane.

"If he had ice on the plane, it would have taken more power than normal to maintain whatever speed he thought he wanted to maintain," Kirton said.

As a pilot with more than 10,000 hours of flying experience and more than 1,000 hours in the King Air A100, Kirton said 85 knots for a fully loaded King Air in possible icing conditions and configured for landing was extremely close to the plane's stall speed. In normal cruising, that kind of plane can stall between 60 and 70 knots, Kirton said. (Eighty-five knots is equal to about 100 miles per hour on land.)

Former NTSB investigator Chuck Leonard agreed with Kirton that Wellstone's plane probably couldn't stay aloft at 85 knots. "That's right down there" at the stall speed, Leonard said.

Leonard said the stall theory is prevalent for several reasons.

For one thing, he said, the weather was "grungy," or less than ideal. There were possible icing conditions and the NTSB has said a weather report eight minutes before the 10:22 a.m. crash indicated that light snow was falling in a calm sky. At the Eveleth-Virginia airport, visibility was three miles and the temperature was just above freezing. Clouds were scattered at 400 feet and the sky was overcast at 700 feet.

Leonard said the gray conditions coupled with the lack of a precision landing system at the airport could have meant that both pilots were scanning the horizon in search of the airport as soon as the plane emerged from the clouds.

"Honestly, what they are doing is that they get [out of the clouds] to within eyesight [of the ground] and they are saying to each other, 'Do you see the airport?' " Leonard said.

"They are both up in their seats looking out, and if you get a little ice and you get a little slow, it doesn't take much. I've seen that many times.

"If I'm putting this together as both a pilot and investigator, I'm thinking that's where it went."

Losing speed, altitude

Kirton also said that it's possible both pilots may have been looking out the window before the plane was properly leveled off below the clouds. The last radar sweep of Wellstone's flight showed his chartered plane about 425 feet above the ground, traveling at 85 knots close to the airport. Three minutes earlier, radar indicated the plane was traveling at 155 knots, about 1,600 feet above airport elevation. The speeds and altitudes were released by the NTSB based on its study of the radar data.

According to the NTSB, the plane drifted off its westerly course to the south and crashed two miles southeast of the airport at a steep nose-down angle after striking trees.

The landing gear was down, the flaps were extended for a landing approach and the engines and propellers appear to have been working properly, NTSB investigators have said.

In this configuration, pilots call the airplane "dirty," in part because of the added aero dynamic drag created by the lowered landing gear.

Maintaining proper airspeed during this time is crucial, because if the airplane stalls, there is typically not enough altitude for the pilots to successfully recover flying speed -- even if the engines are working properly.

"If he were going at 85 knots, he screwed up," Kirton said of the pilot who was flying the plane at that point in the flight. (NTSB investigators have not identified which pilot -- Conry or Guess -- was actually flying.)

"It sounds like he came down, leveled off, maybe didn't add enough power and then stalled it," Kirton said. "If you're looking out there for the runway, it's very easy to be a little bit distracted and maybe not turn the power levers up as far as you should."

At that point, he said, correct interplay between the two pilots would have been crucial.

"A lot of accidents happen because the crew doesn't work together like they should, a lot of misunderstanding," he said. "With two pilots, one is supposed to fly and the other look outside."

Kirton said the experience level of Guess "absolutely . . . adds a whole new dimension" to the crash investigation. Guess, 30, had a total of about 700 flight hours and had flown as an on-call co-pilot at Eden Prairie-based Aviation Charter since July 2001, a job he viewed as a training ground for future employment at a major carrier.

"Bottom line, I'll bet money on the crew resource management issue -- they weren't coordinated or they distracted each other somehow or somebody was responsible to set power and they didn't," Kirton said.

Hampering the investigation is the fact that the King Air did not have a cockpit voice recorder or a flight-data recorder to document the flight's final moments. Neither device was required. No radio transmissions were received from the flight after it acknowledged receiving clearance for the approach.

Leonard said investigators also are likely to consider whether pilot fatigue played a role in the accident. As previously reported by the Star Tribune, Conry's sleep pattern was disrupted in the wee hours of Oct. 24 when he was called unexpectedly by Aviation Charter to pilot a medical emergency flight to North Dakota and back. About 24 hours before Wellstone's 9:30 a.m. takeoff, Conry was driving home to Minnetonka after being up all night.

Instead of resting all day, he went to work late that afternoon at his second job as a dialysis nurse -- a job that Conry's bosses at Aviation Charter said they didn't know he had.

Stalls

Airplane stalls have been a focus of crash investigations almost since the Wright brothers pioneered powered flight in 1903. A common misperception about stalls is that they result from engine failures. Rather, stalls occur when the airplane loses the force of lift that is created from the air flowing over the top and bottom of the wings.

In 1972, the safety board published a special report on 744 accidents from 1967 to 1969 that were classified as "stall/spin" accidents that were not preceded by other occurrences, such as engine failure.

The pilot was considered a "broad" cause or factor in 97 percent of the accidents, including 667 cases where they "failed to obtain/maintain flying speed," according to the NTSB report. A spokesman for the safety board said last week that no follow-up studies have been conducted.

But the spokesman said that from 1983-2001, stalls were determined to be the cause or a factor in 87 air charter accidents, or about 4.5 a year. In those crashes, 97 people were killed, the NTSB said.

Peterson, who flies a private plane extensively in his sprawling Minnesota congressional district, said he suspects that Wellstone's pilots found themselves too far to the left of the runway when they emerged from the clouds. Under Peterson's theory, the pilots attempted to circle back to the left while still under the clouds to try again.

"It looks to me like they were too slow," Peterson said.

He and other pilots have speculated that ice may have clogged a probe under the plane's wing that measures airspeed. If the probe, called a Pitot tube, were clogged, that could have provided a faulty airspeed reading in the cockpit. The tubes are heated to prevent ice buildup, but the heater may not have been working or not turned on, Peterson and others have speculated.

Discounting icing

But Juntunen, the UPS pilot, discounts that theory by pointing to the backup systems on the airplane. The King Air A100 has airspeed tubes under each wing that operate independently of each another, feeding separate gauges in the cockpit. Any discrepancy in the airspeed readings should have prompted precautionary action by the pilots, Juntunen said.

The NTSB said earlier this month that it is collaborating with weather specialists to "more accurately define the icing conditions that existed along the accident flight's route." Although lots of icing could result in an airplane stall, the A100 was equipped with de-icing equipment, and Juntunen said his own flight into Eveleth on the morning of Oct. 25 ended with just a few specks of ice on his plane.

"I've landed with three inches of ice on unprotected surfaces [of the plane]," Juntunen said.

Juntunen, who provided the NTSB with a written description of manageable icing conditions he encountered that morning, said he landed in Eveleth about 2 1/2 hours before Wellstone's plane crashed. A second UPS plane also landed in Eveleth that morning, about 8:30 a.m., with the same description of icing conditions.

Juntunen said the wings on his plane never collected enough ice for him to activate the pneumatic de-icing boots. Instead, he flew out of the conditions by dropping to a lower altitude.

While conditions in the air can change quickly, Juntunen said he doesn't agree with speculation that icing alone brought down Wellstone's plane. He is more interested in the radar reading that showed the plane traveling at 85 knots.

"It makes me wonder," he said of the approach speed. "It's hard to figure" why they were going that slow.

-- Staff writer Greg Gordon contributed to this report.

-- Tony Kennedy is at tonyk@startribune.comor 612-673-4398. Paul McEnroe is at pmcenroe@startribune.comor 612-673-1745.