By Joseph Clough
Introduction
The following is a review of agricultural aviation incident/accident causal factors. NTSB incident and accident reports were the sole source of information. The data includes incidents/accidents reported and recorded between January 1989 and November 2009, and included both piston and turbine engine aircraft involved in crop spraying, fire bombing and associated training activities.
Choice of Air Tractor
Air Tractor was chosen purely because it was possible to differentiate between piston and turbine engine aircraft from the NTSB reports. Other agricultural airplane manufacturers used in agricultural aviation were considered; for example Ayers, which has both piston and turbine powered models, however it was not possible to separate piston and turbine data for all incidents/accidents reports.
Despite a difference in the number of accidents between Air Tractor and Ayers probably due to the numbers of each type in service, a brief comparison showed similar trends in accident numbers over the twenty-year period. The percentage of accidents being fatal is also similar, 13% for Air Tractors versus 16% for Ayers S2R models.
The Air Tractor data collected between January 1989 and November 2009 provided 491 accidents of which 63 were fatal. There was also an approximate 50/50 split between piston and turbine models.
Overview
Interestingly over the last twenty years with the introduction of turbine engines and advances in technology the accident rate although variable from year to year has shown no signs of decreasing, with the number of fatalities following a similar pattern.
Within the Air Tractor sample there is an average of 24 incident/accidents a year with 3 being fatal. For all agricultural types a rough estimate is that there is on average 8 incidents a month with an average of 1 fatality a month.
There maybe various reasons for this trend, increased amounts of agricultural flying, aging aircraft etc., however this is beyond the scope of this review and is mentioned for interest.
Causal Factors
Where possible an incident/accident was attributed to a single causal factor. It is noted that although many reports were prescriptive in the main cause aiding this process, there was a degree of subjective judgment. The following categories were used to classify the incidents/accidents and will be further explained in each section.
CFIT - Controlled flight into terrain
FLOC - Loss of control (In Flight)
GLOC - Loss of control (On ground)
TOPERF - Take-off performance
EMAL - Engine malfunctions
SMEC - Structural / Mechanical failure
MET - Met conditions
OTHER - Includes all others not listed including collisions
Findings
The findings were generated to establish an overall picture of what the main influences on safety are. This review assumes that agricultural aviation includes crop spraying, firebombing and training activities in Air Tractor aircraft.
P= Piston Incident/Accidents
T= Turbine Incident/Accidents
C= Combined
CFIT- Controlled Flight Into Terrain
CFIT is a major factor in aviation activities. Obstacles are a particular hazard in low-level flying, as a result terrain includes power lines, irrigation pipe work, tree lines and any other fixed obstacle.
The majority of CFIT incidents and accidents involved flying into power lines during crop spraying activities, and flying into ridges during fire bombing sorties.
P = 19%
T = 31%
C = 24%.
The number of CFIT incidents has increased by a proportion of 10% in turbine-powered aircraft. Making it the most common cause of incidents and accidents in the turbine models. The reasons for this are not immediately apparent, but may include the increase in operating speeds, longer sorties due to increased loads amongst others.
Three scenarios were identified;
- Pilot misjudgement of space around or under obstacles during spray runs.
- Pilots not seeing the obstacles during both spray and fire operations.
- Pilot misjudgement of ridgelines during fire operations.
GLOC- Loss of Control (Ground)
GLOC includes take-off and landing manoeuvres. This category was included as it is an operational hazard of operating high performance tail wheel aircraft.
P = 8%
T = 13%
C = 10%.
Factors identified in contributing towards ground looping and loss of control included;
- Incorrect landing/take-off technique (see note below).
- Wind close to or above aircraft limits.
- Runway condition
Note- There were numerous incidents where pilots seeing that remaining runway was insufficient attempted to lift off early, getting stuck in ground effect or simply mushing back onto the runway. Along with this observation, the decision not to dump/ jettison hopper load was a common theme.
FLOC- Loss of Control (In-Flight)
Like CFIT, LOC is a recognised issue in all forms of aviation, and an increased hazard in agricultural operations. Loss of control was also the secondary result of many of the other categories including CFIT, MET and even EMAL.
P = 13%
T = 15%
C = 14%.
The most common theme in FLOC accidents was the stall spin scenario with the following being causal factors;
- Flight into IMC.
- Over use of flap in turns.
- After dumping.
TOPERF- Take-off Performance
This category is broad in nature and included the following factors: over loading, not accounting for density altitude, wind limits, runway slope and condition, trim and aircraft configuration. Many of these incidents and accidents resulted in GLOC, LOC and CFIT, however it was decided TOPERF issues were the primary causal factor, warranting its own category.
P = 7%
T = 15%
C = 10%.
This category showed a 100% increase from piston to turbine-powered aircraft. There were no stated reasons for this, however it is likely some pilots had misguided faith in the extra power available for the turbine powered airframes, and became blasé about performance and/or accepted greater risks.
As expected the assumed causal factors were reflected in the NTSB reports and included;
- Incorrect or no calculation of maximum load
- Density altitude
- Excessive tailwind
- Poor runway conditions
- Not accounting for runway slope
- Aircraft out of trim
- Incorrect flap selection
Again the decision not to dump/ jettison the hopper load was a recurring issue.
EMAL- Engine Malfunction
The introduction of turbine engines has undoubtedly improved this category. The causes of engine malfunctions are varied and again the category was broadened to include fuel starvation and contamination as well as technical malfunctions.
P = 52%
T = 24%
C = 37%.
Overall engine malfunctions were the most common causal factor and by far the largest category for piston powered models. The introduction of turbine engines has reduced the number of engine malfunctions by approximately 50%. Although the causal factors of failures are different in nature between piston and turbine engines the following were identified as causal factors;
- Intentionally operating engine beyond limits.
- Engine mishandling
- Continuing flight with known engine faults
- Poor maintenance
- Poor understanding of turbine engines
- Fuel contamination
- Fuel starvation
- Icing (Piston)
Fuel starvation accounted for approximately 17% of all engine malfunctions, and cropped up equally in both piston and turbine models.
Distraction as a result of engine malfunctions resulted in loss of control and controlled flight into terrain on several occasions.
OTHERS
The following were some of the causal factors identified in these categories;
- Mid air collision.
- Collision during take-off and landing (aircraft, vehicles, pedestrians)
- In-flight break-up (wing detachment)
- Component failure due to poor maintenance
- Flight into IMC
- Pilot experience
- Wake turbulence
- GPS distraction
- Drugs and alcohol
- Incapacitation
Decision making
It became evident during the review that pilot decision-making and planning were influences in many of the categories. The following are recurring examples:
Fuel starvation- the pilot was aware of a low fuel state and elected to continue a sortie, resulting in an engine failure.
Power lines- pilots stating they were aware of a power line and a lack of manoeuvring area and continuing to try and fly under a line.
IMC- pilots not being suitably qualified or equipped continuing flight into IMC resulting in LOC. Most of these occurrences were fatal.
Engine malfunctions- pilots identifying engine faults and electing to continue a sortie, resulting in failures or severe damage.
Terrain- pilots being aware that insufficient time and performance are available to avoid a ridge, electing not to dump and ‘go for it’.
Performance- a pilot being aware of increasing quartering tailwinds, or changes in conditions and electing to continue as it had 'been all right on previous sorties', or 'that’s the direction we always take-off in'.
Dumping (Jettisoning) the Hopper Load
On many occasions ‘dumping’ may have prevented an incident or accident, however there appeared to be a reluctance for pilots to dump, both on the ground and in-flight.
There were also several instances where in-flight dumping was initiated and resulted in loss of control.
Conclusion
The review did not unearth anything new, as all of the factors are common knowledge in all types of aviation, and most agricultural pilots are well aware of the hazards.
The purpose of the review was to categorise these factors and to see how they have changed with the introduction of the turbine engine.
As mentioned in the overview section the accident rate and fatality rate has remained largely unchanged over the last twenty years.
Engine malfunctions accounted for the greatest proportion of Air Tractor incidents/accidents, with controlled flight into terrain the second. However it should be noted that engine malfunctions accounted for over 50% of incidents/accidents in piston aircraft versus 24% in turbines, where as CFIT was the most common cause of turbine incidents/accidents at 31%.
From this review it would be reasonable to say that accidents will continue to happen and have the same factors involved. So perhaps training methods and recurrent training for ‘Ag Pilots’ may be a solution in reducing the number and impact of incidents and accidents. Decision making and dumping were issues identified as key areas, these combined with better training in operating and managing turbine engines and malfunctions would be ideally suited to simulator training. Where many situations and particularly those identified can be simulated and practiced over and over again, with a high degree of realism and safely.
If such training were to be accepted and become the ‘norm’, the potential savings in engine damage, hull losses and a reduction in fatalities would be notable and cost worthy.
About the Author
Joseph Clough has a Bachelors degree in Aeronautical Engineering and a Masters degree in Human Factors and Safety Assessment in Aeronautics. He has an ATP and is currently employed as a Senior First Officer for a major UK charter airline flying B737’s. He has had a life long association with agricultural aviation as his father, younger brother and many friends are agricultural pilots.
