Lessons Learned from TSB Investigation of Helicopter Accidents (1994-2003)

Lessons Learned from Transportation Safety Board Investigations of Helicopter Accidents (1994-2003)

Joel Morley and Brian MacDonald

Transportation Safety Board of Canada

Ottawa, Ontario, Canada

Abstract

Between 1994 and 2003, an average of 53 Canadian registered helicopters were involved in reportable accidents each year. During this period, the Transportation Safety Board of Canada conducted formal investigations and produced reports into more than 100 of these occurrences. The paper presents an analysis of the types of accidents investigated for the purposes of providing a snap-shot of the most significant risks faced by the industry during this period. For comparison purposes, a similar analysis was conducted for the Canadian military helicopter accidents during the same period. A number of well recognized hazards were represented in the ten year sample of occurrences. A little over half of the occurrences examined had mechanical cause factors and improper maintenance was implicated in one quarter to one third of those cases suggesting industry efforts to address human factors in maintenance must be maintained and reinforced. Occurrences involving loss of visual references accounted for only 15% of the occurrences investigated. However, 80% of these accidents were fatal and these accidents accounted for more lives lost than any other category. As such, efforts to address this type of accident must also be maintained. It is hoped that this paper will present some food for thought with respect to the risks faced by the industry and the most effective means to address them.

Introduction[1]

A review of Canadian Transportation Safety Board (TSB) Accident Statistics indicates that between 1994 and 2003, an average of 53 Canadian registered helicopters were involved in reportable accidents each year, ranging from a high of 68 aircraft in 1995 to a low of 44 aircraft in 2003.

This represents an average accident rate of 9.30 accidents per 100,000 flight hours for helicopter operators. This rate is higher than accidents involving airliners (0.63 per 100, 000 flight hours), commuter aircraft (3.27 per 100, 000 flight hours), and aerial work operators (7.99 per 100, 000 flight hours). It is similar to the average accident rate for air taxi operations (9.81 per 100, 000 flight hours) although the accident rate for this sector of the industry has been lower than that for helicopters for the last several years. The helicopter accident rate is significantly lower than that for corporate and privately operated aircraft (28.1 per 100, 000 flight hours). The accident rates for all civil aircraft in Canada for the ten year period are shown at figures 1 and 2. When compared to the Canadian Air Force’s helicopter accident rate the civil rate is slightly higher. Over the past 10 years (1994-2003) the military fleet has experienced an accident rate of 6.84.

Such statistics lack meaning unless they are grounded in an understanding of the risks associated with a particular operation and the measures taken to manage those risks.

In the ten year period described above, the TSB has investigated over 100 occurrences involving helicopter operations in Canada. This paper will report upon a systematic review of civil and military occurrences in order to identify recurring issues and present a picture of the most significant risks facing the industry.

Figure 1: Accident Rate for Canadian Registered Helicopters (1994-2003)

Figure 2: Accident Rate for all Civil Aircraft by Aircraft Category (1994-2003)

Method

This investigation was limited to aviation occurrences, involving helicopters, which were reported to the TSB and which were subjected to formal investigations under the Canadian Transportation Accident Investigation and Safety Board (CTAISB) Act[2]. There were 103 such investigations conducted in the subject period.

For the purposes of comparison, a sample of Canadian Forces helicopter accidents investigated during the same time period was analyzed using the identical methodology as for the civilian sample. There were 32 such occurrences. It should be noted that the policies for classifying occurrences are quite different in the military and as such, the two samples may reflect these differences.

The methodology employed for the analysis of the occurrences was as follows.

Step 1: Initial Occurrence Categorization

In order to provide a general picture of the types of occurrences which were being investigated, these occurrences were reviewed and initially categorized into one of the following principal occurrence types:

Power Loss: Loss of power to main rotors due to insufficient engine power or malfunction in drive-train.

Structural failure: Any mechanical problem related to airframe or rotors.

Loss of visual reference: Due to weather or degraded visual conditions (e.g. whiteout, smoke, dust, precipitation, cloud, smoke).

Struck object: Struck terrain or obstacle when sufficient visual cues were available to see and avoid it.

Loss of control: Pilot or environmentally induced (e.g. dynamic rollover, vortex ring state) – aircraft serviceable and sufficient visual cues available.

Loss of separation: Less than adequate traffic separation with other aircraft.

Training for emergencies: Occurred while training for emergency procedures (e.g. autorotation training and is not due to another reason).

Other: All other occurrences.

Step 2: Examination of Proportion of Fatal Accidents

To provide an indication of the human cost of each category, the proportion of fatal accidents in each category was calculated. An accident was considered to be ‘fatal’ where one or more passengers or crew were fatally injured as a result of the accident.

Step 3: Further Break-down of occurrence categories

The categories described in step 1, were further broken down into the following sub-categories which were selected because they seemed to provide a good picture of the types of events contributing to the accidents.

Power Loss

·  Component Failure: Failure for reasons other than improper maintenance

·  Contaminated Fuel: Self Explanatory

·  FOD: Foreign Object Damage

·  Improper Maintenance: Maintenance procedures not followed or not properly completed

·  Other

·  Undetermined

Structural Failure

·  Component Failure: Failure for reasons other than improper maintenance

·  Improper Maintenance: Maintenance procedures not followed or incorrectly completed

·  Other

Loss of Visual Reference

·  Snowball: Reduced visibility in snow due to rotor wash

·  Lack of Contrast: Whiteout or glassy water

·  VFR into IMC: Aircraft continued into IMC conditions

Loss of Control

·  Dynamic Roll-over

·  Loss of Tail Rotor Effectiveness

·  Flight Control Obstruction

·  Rotor Decayed

·  Vortex Ring State

·  Environmental: Unpredicted winds (e.g. downdraft)

Struck Object

·  Trees

·  Wires

·  Terrain

·  Slung Object Caught Obstacle

Due to their small size, the categories of ‘Loss of Separation’, ‘Training for Emergencies’ and ‘Other’ were not subject to further sub-categorization as it was thought that this would not add significantly to the understanding of these accidents.

Step 4: Conclusions of Analysis

The occurrences within the sub-categories were examined for potential areas where safety efforts are best focused.

For the purposes of comparison, a sample of Canadian Forces helicopter accidents for the same time period was analyzed using the methodology described above.

Results

The sample of civilian occurrences used for this analysis included 103 investigations conducted by the TSB between 1994 and 2003. Of the aircraft represented in these occurrences roughly one quarter were multi-engine.

The initial categorization of occurrences by category indicated a large proportion of accidents were related to some sort of mechanical event. Losses of engine power accounted for 35% of occurrences and problems with the airframe or rotor system accounted for 18% of occurrences. Together, these two categories accounted for 53% of occurrences. The number of occurrences in each category is shown at table 1.

Table 1: Occurrences by Category

Category / Number / Percent
Power Loss / 37 / 35
Structural / Component Failure / 18 / 17
Loss of Visual Reference / 15 / 15
Loss of Control / 10 / 10
Struck Object / 8 / 8
Training for Emergencies / 6 / 6
Loss of Separation / 4 / 4
Other / 5 / 5
Total: / 103

The proportion of fatal accidents in each category can be seen at Table 2 and is shown in Figure 3. For comparison purposes, the number of lives lost in each category is shown in Figure 4. These data clearly demonstrate that the most frequent accidents do not necessarily carry the greatest human cost. Accidents in the loss of visual reference category have the greatest probability of resulting in fatalities. Eighty percent of these accidents were fatal, accounting for 31 lives lost in 15 accidents.

Table 2: Proportion of Fatal Accidents by Category

Category / # / # which were fatal / % fatal
Loss of power / 37 / 8 / 22
Structural / Component Failure / 18 / 6 / 33
Loss of Visual Reference / 15 / 12 / 80
Loss of Control / 10 / 4 / 40
Struck Object / 8 / 5 / 63
Training for Emergencies / 4 / 1 / 25
Loss of Separation / 6 / 0 / 0
Other / 5 / 2 / 40

Table 3: Occurrences by Sub Category

Category: Power Loss
Sub-Category / # / %
Component Failure / 16 / 43
Improper Maintenance / 9 / 24
Undetermined / 6 / 16
FOD / 3 / 8
Other / 2 / 5
Contaminated Fuel / 1 / 3
Category: Structural Failure
Sub-Category / # / %
Component Failure / 10 / 53
Improper Maintenance / 6 / 32
Other / 2 / 11
Category: Loss of Visual Reference
Sub-Category / # / %
VFR into IMC / 8 / 53
Lack of Contrast / 5 / 33
Snowball / 2 / 13
Category: Loss of Control
Sub-Category / # / %
Loss of Tail Rotor Effectiveness / 3 / 30
Rotor Decay / 3 / 30
Dynamic Roll-Over / 1 / 10
Flight Control Obstruction / 1 / 10
Vortex Ring State / 1 / 10
Environmental / 1 / 10
Category: Struck Object
Sub-Category / # / %
Slung Object Caught Obstacle / 3 / 38
Wires / 2 / 25
Terrain / 2 / 25
Trees / 1 / 13

Table 4: Military Occurrences by Category

Category / Number / Percent
Loss of power / 4 / 12.5
Structural / Component Failure / 8 / 25.0
Loss of Visual Reference / 3 / 9.4
Loss of Control / 2 / 6.2
Struck Object / 5 / 15.6
Training for Emergencies / 9 / 28.1
Loss of Separation / 0 / 0
Other / 1 / 3.1
Total: / 32

The military sample contained three fatal accidents, all of which occurred in the structural failure category.

Given the small number of military occurrences examined, further breakdown by the sub-categories used for the civilian sample will not be presented here.

Figure 3: Ratio of Accidents to Fatal Accidents by Category

Figure 4: Number of Lives Lost by Category

Concluding Remarks

Loss of Visual Reference Accidents: Although this category represented only 15% of the accidents in the sample, the data presented here serve to reinforce a well known hazard in aviation. While this type of accident was third in terms of frequency of occurrence, it was the single largest killer. Eighty percent were fatal and a total of 31 lives were lost to this type of accident over the ten year period.

It can be argued that commercial helicopter operations in Canada are predominately oriented towards VFR flying. In October of 2004, there were 3600 individuals in Canada holding either Commercial or Air Transport Helicopter Licenses. Only 600 of these people held a valid instrument rating. Unless crews are trained, equipped and prepared to enter IMC or other types of degraded visual conditions, avoiding such situations is essential to safety.

Given the human cost, the reduction of this type of accident must remain a priority for the industry. Potential counter-measures include:

·  steps to increase awareness of the hazards presented by situations of adverse weather and restrictions to visibility on the part of managers, crews and, equally important, customers;

·  increasing the capability of crews to manage such conditions through increased instrument competency and improved equipment, and;

·  improved use of cockpit technology such as radar altimeters and altitude alerting devices.

Power Loss and Structural Failure Accidents: Given the similarities in the sub-categories selected and the patterns observed, these two categories will be addressed together here. Power loss was the most frequently occurring type of accident investigated with 37 accidents or 35% of the sample. Approximately one quarter (9 of 37) power loss accidents involved multi-engine aircraft.

Structural failure accidents were the second most frequently observed with 18 accidents, or 17% of the sample. Twenty-two percent of power loss accidents were fatal resulting in 17 deaths while one third or 33% of structural failure accidents were fatal leading to 12 lives lost.

When taken together, these two accident categories account for slightly more than half of the total sample (52%). The most common sub-category in both cases was component failure with 43% and 53% of power loss and structural failure accidents being classified as such respectively. The scope of this analysis precludes a detailed explanation as to why the components failed (i.e. manufacturing defect, abnormal operations, operational stressors etc.).

The second most frequent sub-category in both power loss and structural failure occurrences was improper maintenance, representing approximately one quarter of power loss occurrences (24%) and one third of structural failure occurrences (32%). These occurrences included improper overhaul or improper installation of components. The potential risk which maintenance error presents has been widely recognized in aviation and these data serve to underscore the importance of efforts to understand and prevent human error in maintenance operations. Such mitigating actions can include enhanced procedures to detect and trap errors, actions to reduce factors which are known to pre-dispose people to errors (e.g. fatigue) and human factors training to increase awareness of these factors among management and maintenance personnel.

It is interesting to note that power loss accidents, while representing the largest category of accidents, shows the lowest proportion of fatal accidents. Thus, many of the power loss accidents are being effectively handled by crews. This fact, when taken in concert with the low number of accidents observed while training for emergencies would suggest that in-flight training for emergencies is of value.