This report is to promote aviation safety and not to establish legal liability. The CAA’s report contains padding, repetition, poor English and incompetence. So, in the interest of clarity and readability, I have had to correct and paraphrase extensively.

Jim Davis

Aircraft Registration: ZS-KMS

Time of Accident: 12 September 2012 0545Z

Type of Aircraft: 1980 Beech Baron 58P

Type of Operation: Private

Licence Pilot-in-command: Commercial

Age: 27 Valid: Yes

Pilot-in-command Flying Experience:

Total Flying Hours: 1 145.7

Hours on Type: 4.6

Last point of departure: Pietermaritzburg airport (FAPM): KwaZulu Natal province.

Next point of intended landing: Kruger Mpumalanga International (FAKN): Mpumalanga province.

Location of accident: Outside the boundary of Pietermaritzburg airport, approximately 200m from Runway 34 at 2423 feet AMSL.

Meteorological Information Surface wind: 240° at 4 knots; Temp: 19°c; Visibility: 200 meters

Number of people on board: 1 + 2

No. of people injured: 0

No. of people killed: 0

ZS-KMS in better days


On Wednesday 12 September 2012, the pilot, accompanied by two passengers departed Pietermaritzburg aerodrome, bound for Kruger Mpumalanga International Airport under Instrument Flight Rules (IFR). According to the Air Traffic Controller visibility was very poor due to fog at 200m.

0508z ATC reported that the pilot requested start clearance for a flight to FAKN at FL150. The start was approved at pilot’s discretion. 0523Z, ZS-KMS reported ready for departure. 0527Z, ZS-KMS pilot was advised to take-off at his own discretion and report when airborne.

The pilot reported that during the take-off roll from runway16, an audible warning came on which he interpreted to be a turbo over-boost warning. The pilot said all instrument readings were normal, and continued with the take-off. After lift-off, the pilot stated that he retracted the landing gear at approximately 300 feet AGL. After retracting the landing gear, the aircraft speed began to decrease and the aircraft began to lose altitude. The pilot confirmed the manifold pressure and power levers were in a normal take off position. He further stated that he tried to correct the situation and gain altitude but was unsuccessful. The aircraft continued to lose altitude and crashed on an open field, approximately 400 meters from the threshold of runway 34.

The aircraft was substantially damaged and all three occupants evacuated without assistance and were unharmed. Post interview with the pilot revealed that he performed pre-departure checks and everything was normal. Prior to take-off he released the brakes and applied full power gradually. The accident happened at approximately 0545Z, in an open field, approximately 400m from the threshold of runway 34, at FAPM. The aircraft was substantially damaged after impact with the ground.

Meteorological Information:

The South African Weather Service reported that, based on the observations at the two recording stations closest to the accident site, weather conditions were as follows: Wind direction 240 Wind speed 4 Knots Visibility 200m. Temperature 12°C. Cloud cover Overcast Cloud base 3500 ft. Dew point -2°C.

‘It’s very easy to say no if the pax are prepared’

Wreckage and Impact Information:

The aircraft’s initial impact point was a steel pole which ripped the right wing off. The aircraft skidded over the field for approximately 130 meters before it came to a stop and went through a fence before coming to rest.

Flight control cable continuity and pre-impact control integrity could be established. Damage was limited to the propellers, the undercarriage, the underbelly and the right wing. The flap lever in the cockpit and the flaps were found in the UP position. The cabin/ cockpit area was still intact and the battery was found in an OFF position. The undercarriage lever was in the UP position but still in transit with the fuel selector valve in an OFF position. All the aircraft seats were secured on their anchors with the safety harnesses still secured and not failed/snapped.

Tests and Research:

On-site investigation revealed that the aircraft collided with a steel pole which ripped the right wing off, this is an indication that the engines were producing a significant amount of power from take-off up to the last point of impact.

The wreckage revealed that all of the structural damages were consistent with the impact, nothing was found to suggest that there had been any pre-impact failure of the primary structure. The fuel tanks ruptured after impact and damage was limited to the vegetation caused by fuel spillage.

Examination of the engines and propellers:

The aircraft is equipped with the Engine Data Management (EDM)-760 monitor. After the accident the EDM-760 monitor was downloaded and all parameters on both engines were consistent with both engines operating normally.


Available information obtained from the ATC revealed that poor visibility prevailed at the time. The aircraft was properly maintained in accordance with the manufacturer specifications. Documentation did not reflect any defect or malfunction that could have contributed to, or have caused, the accident. The investigation revealed premature retraction of the landing gear by the pilot created higher aerodynamic drag during take-off rendering ground impact inevitable.

Probable Cause/s:

The pilot failed to maintain flying speed.

Contributing factor/s.

Premature retraction of undercarriage by the pilot following rotation.

Jim’s comments.

This report stands out for three reasons:

  1. It’s the worst SACAA accident report I have ever seen – and I have seen a lot of bad ones.

2. It inadvertently highlights an extremely interesting, and little understood, hazard of low visibility takeoffs.

3. It opens the door to a whole class of accidents which I will call having the guts to say NO.

‘having the guts to say NO’.

Okay let’s deal with the SACAA first. To correct this report I must introduce you to a guy who I will call Mr X. I have known him for 50 years so I don’t doubt what he tells me. He is an extremely experienced pilot and was intimately connected with the accident and with the pilot. He was asked by the SACAA’s five-person accident investigation team, first for his assistance, and subsequently to actually write the report. He did assist them, but naturally refused to write the report. He told me that there was not a single pilot amongst the investigators and they largely ignored his input.

I will go through this quickly so we can get on with the good stuff to be learned from this accident.

The impact sliced off the right wing.
  • The pilot had around 1782 hours – not 1145.
  • Although he had less than five hours on the Baron 58P, he had over 568 hours on Barons.
  • He did not retract the gear at 300 ft. – he never got more than a few feet off the ground.
  • Initially they say he crashed 200 m from the threshold of runway 34, then later they say 400 m.
  • They say, under Probable Cause, that the pilot failed to maintain flying speed and stalled. Tests show that the aircraft was doing approximately 280 km/hr when it hit the steel fence post. That’s just over 150 knots.
  • They say the temperature was +12C and the dew point was -2C. And the visibility was 200ft in fog. That’s simply impossible.
  • They say, “the aircraft skidded over the field for approximately 130 meters before it came to a stop and went through a fence before coming to rest.” This is meaningless.
  • They say, “the undercarriage lever was in the UP position but still in transit.” The gear was fully retracted.
  • They say, “the aircraft collided with a steel pole which ripped the right wing off, this is an indication that the engines were producing a significant amount of power from take-off up to the last point of impact.” First, what has the wing being ripped off have to do with engine power? Second, significant power in a level attitude would hardly be consistent with a stall.
  • They say that a contributing factor was the premature retraction of the undercarriage by the pilot following rotation. Really? It crashed more than a kilometer after rotation.

‘It’s the worst SACAA accident report I have ever seen’

These ten points are just a sample – there are many others. They show a terrifying level of incompetence.

I believe the pilot was caught out by three things:

  1. Artificial horizon acceleration error.
  2. Somatogravic illusions.
  3. Pressure from the pax.

AH acceleration error makes the A/H horizon bar descend while the aircraft is accelerating. This gives a false nose-up reading so the pilot tends to lower the nose to compensate. Obviously, this is likely to be disastrous just after liftoff. It has caused many night takeoff accidents.

Somatogravic illusions are caused by apparent G effecting your inner ear and your muscular system. They come in two flavours – during acceleration they give you the incorrect impression the nose is rising – so you are tempted to ease the stick forward. And if you move your head in a nodding plane, such as looking down to find the undercarriage selector, and then look up again, exactly the same thing happens – you get the impression that the nose is rising. In addition, the feeling that you are being pushed against the backrest of your seat is exactly the same feeling you get with the nose pointing up.

This muscular reaction to acceleration is used on full-motion simulators to make you feel you are accelerating – they simply tip the simmi backwards. So it joins the other illusions that tell you to lower the nose.

Further, the strong and unfamiliar takeoff acceleration of a twin is always a surprise to pilots accustomed to flying singles. This magnifies the nose up illusions

There is little doubt in my mind that this was the physical cause of the accident. But I haven’t quite finished with the illusions. If you move your head up and down, as if nodding, you can make things a whole lot worse. I suspect the pilot did this on two counts.

First there was the warning buzzer which he thought was overboost. This is because he had a lot of time on turbocharged Senecas which do give overboost warnings – the pressurised Baron doesn’t because it has automatic waste gates that prevent overboosting. So the pilot would probably have looked down to see what was going on. It turns out that the warning was actually for a radar altimeter that the pilot should have switched off before takeoff.

Second, Beechcraft are notorious for swapping levers and switches around on different models – so it’s likely that the pilot looked down to find the undercarriage lever.

Don’t underestimate this business of looking down. The US navy lost a number of fighters which went out of control during their turn on to final approach on aircraft carriers. They eventually found that the pilots were looking down to make a frequency change that was required at that stage. This head movement was causing disorientation and vertigo. The problem was fixed by moving the frequency selector up to the top of the panel.

The somoatogravic illusion.

Pressure from the pax. I have long believed that this is the major underlying cause of many accidents. Pax don’t even need to express pressure. The fact that they have driven to the airport automatically puts pressure on the pilot to meet their expectations. I will call this whole class of accidents, HAVING THE GUTS TO SAY NO. It includes active and passive pressures from your boss, other pilots and from friends and family. Things like taking off in adverse weather because other pilots are doing it; or pushing the limits to get home for your kid’s birthday.

Let me tell you about an incident I had in PE when I was a new boy to instrument flying. I had a few thousand hours total, but my instrument rating was brand new.

I was to take three charter pax to Bloomies in the famous CNN – Charlie Never Never – a battered, red, short-nosed Aztec. This aeroplane was jinxed. It had given other pilots, and myself, a number of fragrant moments over the past couple of years.

We taxied out in conditions very similar to those described in this accident report. I was far from happy – it would be the first time I had taken off with the cloud on the ground. Had I been alone I would not have dreamed of doing it, but my pax were impatient, and I didn’t have the guts to say No.

I was hoping for a mag drop, but no such luck. As I lined up and looked along the runway into the murk, I knew this was a very bad idea. I could picture the burning wreckage in the military base just past the far end.

I took power slowly and smoothly and as we accelerated the image of the crashed aircraft became more vivid. There was only one way out. I pulled the throttles back and brought the aircraft to a halt while explaining to the pax that the left engine was not behaving itself. We taxied clear of the runway and I ran up the left engine. It was indeed conducting itself very badly. This is because I pulled the mixture back to the point that caused rough running.

Later I had to explain to Ian Ritchie, the engineer, what I had done, and why. Ian was an old hand, and an excellent guy – he understood my dilemma and kept his mouth shut.

The incident shook me badly. I knew it would happen again, in different ways, unless I learned to say no.

Eventually, I discovered the secret. It’s very easy to say no if the pax are prepared for it. I should have phoned them the previous day and explained that fog was forecast for the morning and we may be delayed. It would have prepared them to expect a no-go and it would have given them the chance to make alternate arrangements. But most importantly, it would have made it easy for me to say no.

That lesson has served me well and possibly saved my bacon on many subsequent flights.

Prepare your pax for it and then have the guts to say No.

The memory of the Engine Data Monitor was invaluable – yet the investigation still came to the wrong conclusion.

Take-home stuff:

  • Be very careful when two or more things are uncomfortable. In this case the weather and the unfamiliar cockpit.
  • Plan how you will handle a distraction on takeoff.
  • You will find it easy to say no to pax if you prime them the day before.
  • For night, or low viz takeoffs, go on to instruments as the wheels leave the ground, and expect to have an acceleration error on the A/H and a physical temptation to lower the nose.
  • Only raise the undercarriage when you have a positive climb shown on the VSI and the altimeter
  • Be very careful of head movements when flying on instruments.