Accident Probe: Preflight, Interrupted

We all get distracted sometimes, but we must learn to recognize when it happens and respond appropriately.


The airline industry long ago figured out that one of the most dangerous things in aviation is two pilots trying to fly the same airplane at the same time. One inevitable result of such an arrangement is that there are times when no one is flying, and one of the ways we know this is from the accident record. Airlines evolved the pilot-flying/pilot-not-flying concept to acknowledge this characteristic of crewed cockpits and established clear responsibilities for each pilot.

In single-pilot general aviation cockpits, however, we rarely have the discipline to carve out specific roles or tasks for a pilot-rated passenger, who probably is eager to help and/or demonstrate his/ her skills to the pilot in command. This often means delegating tasks like untying the airplane or looking up a CTAF and getting the destination’s automated weather. When there’s a clear delineation of responsibilities, the pilot-rated passenger can be a huge asset to ensuring the flight’s safety and efficiency.

It’s still up to the PIC to ensure the flight’s safe outcome, though, and this often means double-checking anything the pilot-passenger does. It also means checking for things that didn’t get done.


On September 24, 2017, at about 1829 Central time, a Beech C35 Bonanza collided with terrain during an uncontrolled descent after takeoff from the Harrell Field Airport in Camden, Ark. The commercial pilot and the pilot-rated passenger were fatally injured; the airplane was destroyed. Daytime visual conditions prevailed.

The pilot/owner and the pilot-rated passenger had just added 27.35 gallons of fuel to the airplane and were taking off for the final flight of the day, back to the airplane’s base. One witness questioned why the airplane was not gaining altitude after takeoff. Another witness saw the airplane flying just above the treetops, begin a left turn, and then descend and crash. Security camera video from about a mile away recorded the airplane in a steep, left-turning dive just before it impacted the ground and caught fire. No evidence of an inflight fire was observed in the video.


The airplane came to rest upright in a field on airport property about 172 feet east and 1000 feet south of the takeoff runway’s departure end. With one exception, all airplane components were contained in an area 33 feet long and 35 feet wide. The grass and bushes immediately surrounding the wreckage were burned.

Much of the wreckage was consumed by the post-crash fire. Examination revealed the landing gear was down and the flaps were fully retracted at ground impact. One propeller blade remained attached to the hub and was bent aft about 60 degrees beginning about eight inches outboard from the hub. The blade showed no signs of S-bending or chordwise scratches. The other propeller blade was broken at the hub mounting clamps and was bent aft about 10 degrees beginning about 12 inches from the hub. The blade showed chordwise scratches and leading-edge rubbing from midspan to the blade tip.

Flight control continuity was confirmed from the forward cabin area to all control surfaces. Many of the engine accessories were too fire-damaged to verify their pre-crash condition. Borescope examination of the engine cylinders revealed an exhaust valve was worn but functional. All spark plugs showed normal operational signatures. The primary fuel selector was disassembled and found to be in the right main fuel tank feed position. According to the POH, the fuel selector should be on the left main fuel tank for takeoff. For landing, the selector should be on the tank with the greatest amount of fuel.

At 1815, the departure airport’s automated weather observation station recorded calm winds and 10 miles of visibility in clear skies. Data recovered from a handheld GPS device showed the airplane reaching its maximum GPS altitude of 298 feet (about 170 feet AGL) at 1828:52. According to the NTSB, “a relatively flat, open, grass-covered area extended for about 2323 ft in the takeoff direction from the accident site to the airport perimeter.”

That exception to all of the airplane’s components being at the accident site? The airplane’s left main fuel tank cap was found about 4500 feet south of the rest of the wreckage, on the left side of the runway at the 1000-foot marker. The locking lever was engaged, and the cap showed no fire or impact damage.

Probable Cause

The NTSB determined the probable cause(s) of this accident included: “The pilot’s improper decision to return to the runway instead of landing straight ahead when the engine lost power and his failure to maintain adequate airspeed while maneuvering for an emergency landing, which resulted in an exceedance of the airplane’s critical angle of attack and an aerodynamic stall. Contributing to the accident was the pilot’s failure to properly secure the left main fuel tank cap after refueling, which resulted in a loss of engine power due to fuel starvation during the takeoff climb.”

The NTSB added: “It is likely that the left main fuel tank cap was not secured after the airplane was refueled and fell off the airplane’s left wing onto the runway during the takeoff. Without the cap in place, fuel escaped from the left main fuel tank and subsequently starved the engine of fuel during the climb, resulting in the power loss. The pilot likely switched the fuel selector to the right main fuel tank in an attempt to restart the engine. When the pilot tried to turn back to the airport, he failed to maintain a safe airspeed, and the airplane exceeded its critical angle of attack and entered an aerodynamic stall.”

The NTSB’s scenario is a likely one, and we can imagine reacting in similar way to the situation. At that altitude, however, the apparent attempt to turn back to the airport simply wasn’t going to work—it would have been better to land straight ahead. And we’re rather impressed at the speed with which a fuel tank can be emptied when its cap is missing.

Although we’ll never know how and why the left main fuel cap wasn’t properly secured after refueling, it could have involved miscommunication between the two pilots or the pilot-rated passenger’s unfamiliarity with the cap itself. But it likely came down to an interruption in the pilot’s routine, and a failure to verify everything was secured after the refueling.

Distractions During Preflight

According to the Flight Safety Foundation (FSF), “Interruptions and distractions often result in omitting an action and/ or deviating from standard operating procedures (SOPs).” A task force assembled by the FSF to examine the impact of interruptions and distractions in the cockpit found that their primary impact was to “break the flow pattern of ongoing…activities (actions or communications)” including SOPs, normal checklists, communications and problem-solving activities. The FSF says to reestablish situational awareness, we must:

  • Identify the task being performed previously;
  • Ask when during that task you were interrupted;
  • Decide what’s necessary to complete the task;
  • Prioritize the steps required to complete the task;
  • Plan the sequence for those steps; and
  • Act.

Aircraft Profile: Beechcraft C35 Bonanza

Image: Aleksander Markin – CC BY-SA 2.0

OEM Engine: Continental E-185-11

Empty Weight: 1650 lbs.

Maximum Gross Takeoff Weight: 2700 lbs.

Typical Cruise Speed: 148 KTAS

Standard Fuel Capacity: 39 gal.

Service Ceiling: 18,000 feet

Range: 510 NM


This article originally appeared in the September 2019 issue of Aviation Safety magazine.

For more great content like this, subscribe to Aviation Safety!

Joseph E. (Jeb) Burnside
Jeb Burnside is the editor-in-chief of Aviation Safety magazine. He’s an airline transport pilot who owns a Beechcraft Debonair, plus the expensive half of an Aeronca 7CCM Champ.


  1. The NTSB added: “..Without the cap in place, fuel escaped from the left main fuel tank and subsequently starved the engine of fuel during the climb,”

    WHAT climb? They never even reached 300′.
    Tell me HOW a missing fuel cap on the top of a tank will starved the engine of fuel right after takeoff.

    • Probable:
      1. likely to occur or prove true: He foresaw a probable business loss. He is the probable writer of the article.
      2. having more evidence for than against, or evidence that inclines the mind to belief but leaves some room for doubt.
      3. affording ground for belief.
      Okay, what’s your best W.A.G.?

    • I agree with Mark. I had the experience of leaving a fuel cap off one time, in the only tank with fuel in it. This was in a Cessna 180, and the tank with fuel only had about 18 gallons in it for a 20 minute flight. I did not realize that the cap was off, hanging by the chain, until preparing to re-fuel at my destination! Nice stain on the wing too. Remember the articles called “I Learned About Flying From That”?
      I think that the fuel siphoning eventually reduces or stops when the fuel level is quite low, but that takes time. In a Bonanza, if fuel pressure is affected, there is always the electric fuel pump.
      Ian Hollingsworth
      DER Flight Test Pilot

      • Best WAG is that there was plenty of fuel in the left tank that was just filled and the cap not fully secured. Gobs of fuel. A huge post impact fire confirms there was still plenty of fuel on board the aircraft.

        The real question becomes what was in the right tank (that still had a cap on it and was found selected at the time of the crash). Lack of fuel in the perfectly capped right tank is my “WAG”, not the left tank.

  2. 20 gallon capacity, 17 usable in level flight. There is an AD that requires placarding the fuel gauges with a yellow line. That yellow line starts at the 10 gallon mark prohibiting takeoff when the fuel level gets down to the yellow as you can have fuel starvation problems. Not difficult to burn 1-2 gallons during run up and taxi, especially when you have the electric prop as checking it takes longer and is totally different than a hydraulic. That leaves approximately only 5 gallons to siphon off causing a potential fuel starvation event depending on take-off / climb attitude and the nuances of that particular air-frame’s fuel bladder installation. A missing cap would siphon off five gallons very quickly. Not saying every Bonanza quits when the fuel reaches the yellow, but take off is prohibited when in that range.

  3. Yup Stuff like this happens. Was an instrument instructor 40 years ago and had a flight instructor who wanted his IFR rating. Took a PA-23-250 which he rented and he left the left fuel cap loose on a very cold morning at YYZ.

    Whooosh goes the fuel. Barely made it to YHM but he learned his lesson!

    Reminded him to tell himself and his students to complete the preflight without interuption!

  4. I dunno, the siphon explanation doesn’t do it for me either. The yellow-line restriction is related to turning-induced un-porting of the aft-located fuel pickoff point, causing air ingestion. That would have occurred basically at the start of the takeoff roll and before any siphoning could occur. Assuming he started with enough fuel to preclude the initial un-porting (did he?) I can’t visualize siphoning radical enough to empty the tank to the point of pulling all fuel completely away from the pickoff point.

    But, like the explanation that he was interrupted and that resulted in the fuel cap falling off, we’re just speculating.

    • I agree that the idea of an entire 20 gal tank being drained in a matter of seconds or even a few minutes is highly unlikely. Fuel would have to be gushing out like a firehose! I’m ashamed to admit I’ve made this mistake (twice!), and I lost significant fuel, but not the entire tank – maybe a third – and that was over a long flight. However, if the tank’s only vent was in the cap, then the low pressure over the wing could stop the fuel from flowing even if the tank was full. Most Cessnas have dual vents, one in the cap and a ram-air vent. Not sure about the older Bos. Maybe someone with specific knowledge could comment.

  5. First off a few reminders on the definition of the word “probable”:
    1. likely to occur or prove true: He foresaw a probable business loss. He is the probable writer of the article.
    2. having more evidence for than against, or evidence that inclines the mind to belief but leaves some room for doubt.
    3. affording ground for belief.
    After extensive inspection of the accident site, airplane remains and other verifiable factual information the “probable cause” is stated (refer to above definition). The NTSB does not state absolute cause as the is often unknown or provable possibilities that cannot be substantiated.

    • The examination that a cap was off the left tank means one of 2 things:
      1) that ~30 gallons of fuel drained out UPWARDS in mere seconds and that the pilot had lightning good emergency decision skills to recognize the problem and fast reflexes to switch to the right tank or…
      2) he took off on a near-empty right tank, it sucked air when the plane pitched up, and his first reaction was to turn back instead of switch tanks and the post crash fire was from the ~30 gallons and no fuel cap.

      Tell me which sounds “probable” based on the information?

  6. Love it! Been accused of being “punctilious” and sometimes I wear it like a medal. For all we know there’s probably a mathematical way to compute the volume of avgas that would be sucked out of the top os a Bonanza wing with the fuselage cap off. Regardless, I trust the exhaustive analysis done by the NTSB to come up a probable cause more than some armchair astrophysicist’s. Just sayin.

    • This has been the second consecutive story on fuel; both “conclusions” seem to not add up based on my limited 40 years of flying. The simple explanation here was that the pilot had the right tank selected on landing and subsequently forgot to switch to the left for takeoff. When the right tank suck air on takeoff, the pilot “knew” he had fuel so his first reaction was to turn back, NOT to look down and fiddle with the fuel controls. He crashed with the fuel selector still on the right tank and the good chance of a post crash fire from an open fuel tank with still at least 27 gallons.

      The missing cap only played a part in the post crash fire.

  7. Possibility of water in the fuel or tanks wasn’t mentioned. If the left tank was empty then the fire would be from only the fuel in the right tank. Was their any forensic evidence that this right tank fuel was or was not solely responsible for the fire?

  8. “According to the NTSB, “a relatively flat, open, grass-covered area extended for about 2323 ft in the takeoff direction from the accident site to the airport perimeter.” Assuming the pilot departed the +5000ft RWY 1, he he would see about 1200ft of grass ahead and houses to the left. I can’t find the 2323 ft grass field the pilot is apparently accused of rejecting. That said, if a mile is required to get a Bo to “treetop” altitude something likely more sinister than a loose cap is in play IMHO.

    • That, and it was getting really dark at that time of evening. The view ahead into the darkness was probably not very good. Unfortunately it sounds like the engine did not regain any power or surges so the turn back did not happen. It’s real tempting to turn back to the “known good” than head forward into the “dark unknown”.

      • The report states that “daylight visual conditions prevailed”. The sunset time for the departure airport on September 24,2017 would have been between 7:15-30 pm approximately. The NTSB make some suppositions with regards to the pilots selection of the fuel tank selection and fuel cap affecting fuel starvation as they were unable to make a definitive conclusion as to why the engine quit or lost power. The focus of the article strayed from distractions and interruptions during a pre-flight and secondly the cause of the resulting crash (stall) to why the engine quit, which was not determined conclusively. Here’s a link to the actual full report.
        Nevertheless interesting discussion.

  9. Probable in this context means that’s the most likely explanation we’ve been able to think of. I was involved in an accident that the NTSB’s best guess didn’t fit the timeline. They were relying on some erroneous “witness” testimony.

  10. I’m not familiar with the Bonanza in question (Air Force and airline heavies, mostly), but I ‘d like to contribute the following: guessing that the fuel filler is visible from the cockpit, is there a chance that the pilot got a glimpse of the departing fuel cap, or was doing a quick glance-around just after lift-off, and kick-started the response sequence which led to the resulting crash?

  11. I can’t speak to the rest of the accident sequence, but an open gas tank on a Bonanza will drain in a real hurry. A year or two ago I watched one of our pilots have a gas cap pop off his A-36 at takeoff. There was quite the plume of gas coming off the wing. I don’t remember how long it lasted, but by the time he made it around the pattern and landed the tank was empty. Fortunately the cap was on a safety chain, so all he lost was the gas. (Also, fortunately, he was flying on the other tank, or got to the other tank in time.) His remark afterwards was “Well, that’s the most expensive pattern I’ve ever flown.”

  12. I just recently left the oil filler cap off after adding a quart of oil to the engine on my C172. I took off and flew for an hour with the oil cap off. I had a nice mess to clean up off the side of the aircraft after landing. The probable cause was interrupting my normal flow of pre-flight and going back to add oil to the engine after completing my pre-flight. I learned a thankfully harmless, but messy, lesson to review the preflight again after any interruptions or deviation from the normal flow of the checklist.