Operational Failures: Execuflight 1526 Part Two
On the 10th of November 2015, a BAe 125 operated by Execuflight crashed into an apartment building while on approach to Akron Fulton International Airport. The captain, the first officer and seven passengers died.
I looked at the flight in Trouble in the Cockpit: Execuflight 1526 Part One. The key question, which we discussed in the comments, was how the situation in cockpit degenerated so badly that a routine approach turned into a fatal accident. Clearly the CRM (Cockpit Resource Management) had fallen to pieces. Fatigue was mentioned, as well as lack of leadership skills and flat out incompetence.
One unexpected issue that came up during the investigation was the weight and balance. The British Aerospace HS 125-700A serial number 257072 had a basic weight of 13,815 pounds (6,266 kg). The operatating weight was documented as 17,276.92 pounds.
Then in 2014, the APU auxiliary power unit was removed, reducing the basic operating weight by 300 pounds (136 kg). This is noted on the weight and balance record, with the basic operating weight amended to 13,976 pounds. Except that there’s no mention in the aircraft maintenance records of this ever happening. There is a mantenance record from 2015 which refers to an unserviceable APU which was removed and then replaced with a serviceable APU. Certainly, an APU was found in the wreckage. But somehow, that 300 pounds was never added back to the weight and balance.
Not that it mattered. Because, when they checked the weight and balance computation for the flight and also the previous flights on the two-day trip, it was clear that the computations used the aircraft’s basic empty weight of 13,815 pounds instead of the actual operating weight of 14,276 pounds. So they’d already lost over 300 pounds to their starting weight when doing the calculations.
They used a program called Ultra-Nav which had the basic empty weight as the default setting, instead of the basic operating weight. The chief pilot said he didn’t know what the default setting was in the program, as any crew member who had access to the system could change the default. There’s no way of knowing who entered the basic empty weight instead of the basic operating weight.
The other mistake seems almost trivial in the scheme of things but I think is worth mentioning. The captain hadn’t filed an alternate airport for the flight plan of the accident flight. More importantly, this was not a one off.
From the official accident report
A review of the available past flight plans filed by the captain while operating as PIC for Execuflight between August 28, 2015, and November 10, 2015, revealed four other flights for which an alternate airport was required but not filed, including the flight before the accident flight from LUK to MGY. This indicated a recent history on the part of the captain of failing to file an alternate airport on a flight plan as required by federal regulations. In addition, the captain’s history of failing to file an alternate airport when one was required indicates that Execuflight management had a history of failing to verify that an appropriate flight plan was filed before departure of each flight as required by company procedures.
The investigation found that the pilots sent the Ultra-Nav printouts to the chief pilot and not the completed flight logs.
Flight logs for 9th and 10th November were found at the accident site and it turned out that the weight and balance on these didn’t match the numbers on the Ultra-Nav print outs sent to the chief pilot. There were no entries on the flight log for the accident flight at all. The NTSB reconstructed the weight and balance as best they could, although they didn’t have the weights of the passengers and the baggage, subtracting fuel burn and came to an estimated landing weight of 22,286 pounds. The BAe 125’s maximum landing weight was 22,000 pounds.
At their actual weight, as opposed to the logged weight, VREF (landing reference speed, or, the speed at which you should cross the threshold in a landing configuration at a height of about 50 feet) was 125 knots at flaps 45°.
With that information, we can revisit their descent. At the FAF crossing altitude they were going 120 knots. At the point when they were descending at 2,000 feet per minutes, the airspeed was down to 118 knots. At the MDA they were flying 113 knots and, as they descended past it, the stick shaker activated. The first officer levelled off but did not add thrust, and the recording ended.
Now, it’s not exactly surprising that they stalled and the weight and balance can’t be said to have caused the crash. But I think it gives some interesting insights into the operations.
While we’re looking at the flight logs, let’s take a look at the previous flights of the crew. The captain’s details are straight forward: he was off duty on the 6th and 7th of November. He flew a day trip on the 8th of November, on duty from 07:50 to 13:05. Then he was off work until the two-day charter of seven legs. So he was nicely rested and ready to go.
The first officer’s week was a little bit more hectic.
On the 6th and 7th of November, the first officer was paired with a different captain for a two-day trip from Fort Lauderdale, FL to Teterboro, NJ and back.
On the 6th, he was on duty from 0800 to 1300, when they arrived at Teterboro. The following day, he and the captain checked out of their hotel at 12:10 and took a shuttle from the hotel to the airport. They arrived some time after 14:00 and the first officer began work, conducting all the pre-flight preparation for the next leg of the trip. The flight log shows that they came on duty at 15:00. The captain sent a text message to the dispatcher at 16:53 stating doors closed for the flight back to Fort Lauderdale.
The allowable length of a flight duty period depends on when the pilot’s day begins and the number of flight segments he or she is expected to fly, and ranges from 9-14 hours for single crew operations. The flight duty period begins when a flightcrew member is required to report for duty with the intention of conducting a flight and ends when the aircraft is parked after the last flight. It includes the period of time before a flight or between flights that a pilot is working without an intervening rest period. Flight duty includes deadhead transportation, training in an aircraft or flight simulator, and airport standby or reserve duty if these tasks occur before a flight or between flights without an intervening required rest period.
Execuflight’s president was the operations management person on duty. He said that the first officer did not come on duty at 15:00 on the 7th but instead came on duty at 17:15, which would mean that the first officer had ten hours of consecutive rest in the preceding twenty four hours. He said he’d specifically instructed the first officer not to go to the airport until shortly before the flight. When asked if anyone had checked up on this, he said that the chief pilot and the first officer discussed the duty time and whether it would be OK or not OK for the first officer to make the trip.
Execuflight’s president gave sworn testimony that the officer came on duty after the captain, starting at 17:15, and thus had his ten hours rest. But the hotel records, the doors closed message and the sworn testimony of the captain of that flight tell a different story.
Execuflight then asked the two pilots to fly an overnight trip to Mexico. The captain of the two-day flight said that they couldn’t, as it would cause both of them to exceed their duty time.
The flight arrived back at Fort Lauderdale at 20:06. When they landed, the chief pilot met them and told the first officer that he was needed for the overnight trip to Mexico, departing at 20:40. The captain of the flight was the chief pilot.
The first officer finished his duty period at 07:15 on the 8th. The two-day trip captain spoke to the first officer that evening. The first officer said that he was tired from the overnight trip and that he’d expected to stay overnight in Mexico, but they ended up flying all night and coming back.
Execuflight’s president said that they always had the option to stay in Mexico if they were breaking duty time.
At this stage, he’d had 7 hours and 45 minutes consecutive rest in the preceeding 24 hours.
From 09:17 to 17:04 he had a rest period and the opportunity to sleep. He sent a text message at 12:53. Other than that, we don’t know anything about his rest pattern but it doesn’t look good.
The FAA overhauled the pilot fatigue rules in 2011. As of the current rules for passenger airline pilots, the flight time limits are eight or nine hours (depending on start time) with a ten-hour minimum rest period prior to the flight duty period, with an opportunity for eight hours of unterrupted sleep within the ten hours. They must also have at least 30 consecutive hours free from duty on a weekly basis.
On the evening of the 8th of November, the first officer finally had a rest period of over ten hours, with a sleep opportunity from 19:15 that evening to 05:50 the next morning, when he rose to meet the captain of the accident flight for their seven-leg charter flight. That day, the two pilots were on duty from 05:50 to 20:30, a total of 14 hours and 40 minutes.
The following day, they got up at 09:00 and departed Cincinnati for the half hour flight to Dayton. The captain filed the IFR flight plan for Akron, Ohio, departing at 13:30. They actually departed at 14:13. Not a tough day in itself but it isn’t hard to see from this schedule that the first officer must have been tired and really, really ready for a day off.
Now his fixation and inability to take in what the captain was saying makes a lot more sense. At that stage, he just couldn’t keep track of anything any more.
The thing is (and I want to say this gently, because the poor man lost his life over this) he wasn’t an excellent pilot in the first place. When he started at Execuflight, he was fifty and had worked as a first officer for eight years.
From 2007 to 2011, he was employed by Personal Jet as a first officer on the Hawker and Learjet. From 2012-2014 he worked as a first officer for Chauff Air and in 2014 he was hired as a first officer on a Boeing 737 for Sky King.
Sky King reported that he started to fall behind and that he “struggled” with memory items and weight and balance problems. The Sky King check airman described the first officer’s performance in the simulator as “ridiculously weak”.
After a few months, they struggled with his “lack of acceptable progression” and had him fly as a jump seat observer for a week to pick up some 16 hours of observation experience. At the end of that, they ranked his progression as “significantly below acceptable standards.” He was let go, with “unsatisfactory work performance” given as the reason.
Three months later, he was hired by Execuflight.
Now, there’s specific legislation in the US to ensure that operators adequately invesitgate a pilot’s background before allowing him or her to conduct commercial air carrier flights. Before you can put the pilot in the cockpit, you have to review the last five years of the pilot’s background and other safety-related records. The Pilot Records Improvement Act (PRIA) requires all previous air carriers to provide employee records pertaining to the individual’s performance as a pilot, with a specific focus on training, qualifications, proficiency and professional competence.
The Execuflight General Operations Manual stated that the president was responsible for recruiting and terminating all company personnel. The duties of the director of operations included conducting personnel interviews and recommending personnel actions to the president.
The director of operations (who has since left the company) said he was not involved in hiring of either of the pilots. The president agreed that he’d hired both pilots but said that he’d delegated the review of PRIA records to the chief pilot. The chief pilot said that the hiring was done by the president. He had no idea that the first officer had been fired by Sky King.
Execuflight’s president stated that he hired the first officer based on a recommendation from another pilot and a Part 91 flight that he made with the first officer. When asked about the first officer’s PRIA background check, the president stated that he was “not too familiar” with the PRIA records because the chief pilot was the one who “does all the due diligence in that respect.” He further stated that he did not “really home in on” the first officer as he was hiring him as an SIC and “didn’t get in deep into his file.”
The Sky King check airman confirmed that no one from Execuflight ever contacted him to follow up or to ask for additional information.
No remedial training was offered to the first officer. He was simply put to work. With his history of struggling with memory items and weight and balance, in an aircraft above its maximum weight, fatigued after a week of heavy flying, it’s hardly surprising he couldn’t make sense of the Akron approach.
However, this still doesn’t explain why the captain let him have control on a revenue flight and why the captain didn’t take over when he realized that the first officer wasn’t coping. He was 40 and had no prior accidents or incidents, just a single enforcement action for operating an aircraft contrary to an ATC instruction.
Execuflight supplied PRIA documentation which referred to itemized records from his previous employer but did not actually include any copies of those records.
Execuflight’s president stated that he’d hired the captain based on his background, including his previous experience in the Hawker. He said that the background check showed an “administrative issue” related to a communication course and that the captain left his previous post voluntarily. He did not contact the previous employer to ask for additional background information.
However, he had too been fired from his previous employment. He was scheduled for recurrent training on the Hawker 800A and didn’t attend the training, for which he was terminated.
The FAA requires every operator to offer an approved CRM (Cockpit Resource Management) training, which at Execuflight was offered through lectures and a PowerPoint presentation by the chief pilot. The content was copy and pasted from the appendix of the Code of Federal Regulations (CFR) which is meant to assist organizations in program and curriculum development. In other words, it’s a list of issues that should be taught and not meant as the actual training content. At the end of the training, each pilot took a ten question test and was required to have a ‘non-corrected passing grade of 80%’. If a pilot scored under 80%, then the pilot was to review the ten questions and receive a corrected grade to 100%. The test questions and answers were provided in the CRM manual, making revision easy.
The captain’s answer sheet had 100% written in the box for the uncorrected grade. The box for the corrected grade was blank. Looking at the answers, however, the captain only got four questions correct: his uncorrected grade should have been 40%.
Specific areas of weakness: responsibilities of the Pilot in Command, flight deck management and aeronautical decision-making.
(For what it’s worth, the first officer got 70% uncorrected and 80% corrected.)
Although both pilots had attended sim training at Execuflight, the captain was paired with a pilot from a different operator and the first officer was paired with an instructor. As a result, neither pilot had the opportunity to practise CRM. There was no scenario-based training or line-oriented flight training, thus no reinforcement of CRM principles as a part of the simulator training.
The FAA principal operations inspector responsible for the oversight of Execuflight clearly dropped the ball. He had certified the operator in 2004 and, in the intervening time, he oversaw 16 certificates of which 14 were Part 135 operators like Execuflight. His workload was “very busy” and, although he’d visited the operator a few times a year and conducted line and ramp checks, he didn’t know what pilots did in normal operations. He only saw them during line checks, which weren’t performed on revenue flights. When the captain received his line check when he started his employment in June 2015, the inspector did not directly administer it, instead, he observed Execujet’s chief pilot who conducted the captain’s line check.
The POI considered Execuflight to be a “very good operator,” and, in general, he felt Execuflight did a “pretty good job.” According to the POI, he was aware that the position of director of operations was vacant at the time of the accident and that Execuflight was in the process of trying to fill it. The POI stated that the chief pilot temporarily took over some of the duties and did an “excellent” job.
The report concluded that the captain should have taken control or called for a missed approach before the aircraft ever reached the Final Approach Fix (FAF) when the first officer reduced the airspeed, putting the aircraft in danger of a stall. After they descended past the MDA, the captain said ‘okay, level off guy’ instead of calling for a missed approach. The stick shaker went off. The last radar data of the aircraft shows that the aircraft was about 1,300 feet above mean sea level, with an estimated speed of 98 knots and an estimated angle of attack of 20°. The critical angle of attack for the aircraft, that is, the point at which any further increase will decrease lift and cause the aircraft to stall, was 15.5°.
Although the captain clearly recognized that the first officer was having difficulty flying the approach, he did not intervene either before the airplane reached the FAF when he noted the first officer was not maintaining the proper airspeed or after the airplane passed the FAF when he noted that the first officer was exceeding the proper vertical speed. Rather, the captain inappropriately adopted the duties of a flight instructor and did so ineffectively as demonstrated by his failure to ensure the safety of flight as he coached the first officer through the approach. The NTSB concludes that the captain’s failure to enforce adherence to SOPs and his mismanagement of the approach placed the airplane in an unsafe situation that ultimately resulted in the loss of control.
Before we finish, I just want to refer back to the Execuflight’s comment to the press at the time of the accident.
Planes just generally don’t fall out of the sky. I can tell you that they were very well seasoned pilots, both of them. They like to fly together. We monitor the flights leg by leg since it started and it’s typical for them to give us a doors open, doors closed message, we’ve got them all.
The two pilots had flown together three times in the 90 days before the accident. But more importantly, this throw-away statement to the press reinforces another finding by the NTSB: the staff at Execuflight kept track of the aircraft through the doors open and doors closed text messages, but they did not appear to consistently review other aspects of the flight: the fuel, the weight and balance, the flight plan.
3.2 Probable Cause
The National Transportation Safety Board determines that the probable cause of this accident was the flight crew’s mismanagement of the approach and multiple deviations from company standard operating procedures, which placed the airplane in an unsafe situation and led to an unstabilized approach, a descent below minimum descent altitude without visual contact with the runway environment, and an aerodynamic stall. Contributing to the accident were Execuflight’s casual attitude toward compliance with standards; its inadequate hiring, training, and operational oversight of the flight crew; the company’s lack of a formal safety program; and the Federal Aviation Administration’s insufficient oversight of the company’s training program and flight operations.
In addition to the recommendation, there was a Board Member Statement which I found uncommonly powerful.
I believe the organization that chartered this aircraft expected to get a professionally managed aircraft. I suspect they expected to get a professionally flown aircraft. And, I further believe they expected that when the regulator, the Federal Aviation Administration, issued an Air Carrier Certificate to Execuflight, they expected the FAA to provide adequate surveillance of Execuflight. Tragically, as this investigation found, those charter customers did not get what they expected or deserved in any of these respects. Their expectations were based on a house of cards that created an illusion of safety.
The statement focuses on this illusion of safety and calls out both the lack of regulatory oversight and failure of industry audits, as the operator held an ARGUS Gold rating and was Wyvern Registered, both seals of approval which offer customers the security that they are choosing a respected and safe air travel provider.
I’m glad to see the statement. It’s depressing to see an American accident in 2015 with all the hallmarks of a cowboy operation of what I had hoped was a time gone by. There’s no excuse any more for disregarding the importance of good CRM and of protecting pilots from fatigue. I’m also glad to see that the FAA was called out for not seeing the operator’s issues, for failing to spot the standard disregard for SOPs that appears to have been endemic within the company.
Paying passengers should be able to feel secure that flying is safe, whether in a Boeing 747 or a ten-seater business jet. I spend a lot of time trying to convince the general public that they are not taking their lives into their hands every time they consider a flight. Situations like this really does aviation’s reputation no good at all.
The full NTSB report, including the board member’s statement, can be found online on the NTSB site: Crash During Nonprecision Instrument Approach to Landing Execuflight Flight 1526