R. J. Waldron & Company (1987) Ltd.
110-5920 No. 2 Road, Richmond, B.C. V7C 4R9 (604) 270-2722
30 October 2008
07-313
Summary Report Wreckage Examination
MJM Air Ltd
DHC2 Beaver, C-GAQW
Near Quadra Island
28 February 2005
Introduction
The aircraft departed from Campbell River spit on a VFR flight to Frances Bay/Knights Bay with four passengers and the pilot on board. The aircraft did not arrive at either destination. The official search did not locate the aircraft but floating debris was identified and the body of David Stevens, who died from exposure and drowning, was found on shore. The remaining occupants have not been recovered.
A search and recovery effort financed and conducted by the victims’ families in July 2005 located the crash site and the aircraft was recovered on 28 July 2005. All of the occupants had exited the aircraft following the crash, indicating that the impact did not induce any incapacitating injuries.
R.J. Waldron & Co. was initially retained by Work Safe BC to attend the Transportation Safety Board of Canada examination of the engine at Valley Aero Engines in Langley, BC on 25 Sept. 2007. Subsequently, WSBC released George Heath to investigate for the victim’s families.
This summary report will document the observations, findings and conclusions of the R.J. Waldron & Co. investigation.
Scope of R.J. Waldron & Co. Investigation
R.J. Waldron & Co. was directed to complete the examination of the airframe physical evidence that had not been carried out by the Transportation Safety Board of Canada. Examination of the engine had not revealed any evidence of mechanical failure; however, examination of the propeller indicated rotation at low power at impact. The TSB noted that the propeller control was set to fine
pitch and that the counterweight assembly and blade shim impact marks were in the fine pitch range. The propeller is normally set to Fine PITCH for landing and during emergency power requirements – the propeller control is set toward the Decrease RPM setting in normal cruise flight. The propeller evidence suggested a possible precautionary or forced landing situation.
Various witness accounts reported hearing abnormal engine sounds as the aircraft passed nearby. These witness testimonies were documented by the families and the information was used to successfully locate the wreckage.
The primary objectives of this investigation were;
• Examination of the all fuel system components, including tanks, fittings, plumbing, filters, selector valve system and vent system.
• Examination of engine control systems.
• Examination of primary flight control systems
· Examination and evaluation of the oil deposits on various parts of the aircraft.
In addition, a detailed examination and analysis of the non-standard carburetor plumbing, and an assessment of the floats was requested.
Airframe Components
1. Airframe Overview
Time is the enemy of effective aircraft accident investigations. Depending on the environment and other factors, as time passes
the condition of the physical evidence changes and deteriorates. The Coast Guard failed to locate this wreckage during the initial
search, but it was found by the families through private searching about five months later. The subsequent recovery, storage and
transportation of the wreckage unavoidably caused further change and deterioration to the wreckage.
The Transportation Safety Board of Canada carried out three separate examinations in response to pressure by the families to
conduct an investigation. The airframe inspections were narrow in scope and cursory in most regards.
The exception was the engine, which was recovered 2 ½ years after the crash – the engine examination was complete and
thorough. If the cause had turned out to be readily apparent – a catastrophic engine failure for example – the deterioration might not
have presented a difficult problem. Since the actual cause is more obscure and may have been lost, the deterioration was a significant
obstacle to the investigation. Definitively separating post and precrash damage and corrosion was not possible in many areas.
The interior and floor were removed to allow access to the concealed system components.
2. Aileron flight control system
The aileron flight controls were secure and functional through the fuselage and the left wing. The right wing has not been recovered,
but its’ location on the ocean floor is documented. Separation of the right wing occurred during the initial impact sequence and rollover.
3. Elevator flight control system The elevator flight controls were secure and functional through the fuselage.
4. Rudder flight control system
The rudder flight controls were secure and functional through the fuselage.
5. Flap flight control system
The flap system was secure and functional through the fuselage and the left wing. The right wing has not been recovered. The flap
actuator was found in the retracted position indicating flaps UP. The flap selector was in the neutral (normal cruise) position.
6. Fuel Selector and Fuel filter
As examined the fuel selector would not rotate between the three tank positions – it was seized in the REAR position. The cable
system between the selector handle and the selector was secure and capable of normal operation. When the cable was
disconnected from the selector, the cockpit handle moved freely.
The fuel selector was removed, examined and disassembled. It appeared normal externally with no impact damage or signs of
corrosion. It was seized and could not be rotated. It was completely blocked and did not flow through from inlet to outlet. Disassembly
disclosed that the unit was packed with internal corrosion byproduct material.
The fuel filter lockwire was intact, indicating that the filter was not inspected during the TSB investigation. Examination of the filter
revealed that it was about 20% blocked by corrosion materials and fragments of the disintegrating filter. These contaminants were
judged to be post crash.
7. Fuel tanks, fuel plumbing & venting
The three fuel tanks were opened for internal inspection. The quantity transmitters for all three tanks were intact and capable of
normal movement. The three fuel outlet screens were open and the plumbing from the tanks to the selector valve was unobstructed.
The rear tank was removed with difficulty due to deformation of the tank and the enclosing structure – the bottom surface of the
fuselage was hydro-formed upward and the bulkheads were displaced. Overall, the tank was free from corrosion; however,
there were two small holes corroded through the bottom of the tank. There is no conclusive method to determine if these holes formed before or after the crash.
The plumbing from the fuel filter outlet to the firewall was open and secure.
The fuel tank venting system consists of a vent pipe from each tank merging into a single tube that extends up into the left wing root.
The inlet/outlet of the vent tube was found crushed and sealed, which would have prevented normal function. However, examination of the damaged end determined that the crushing occurred during recovery, storage and/or transporting the aircraft.
Unexpectedly, no fuel was found in any tank, line, or in the filter housing or fuel selector. It was anticipated that at least small
quantities of fuel would be found in closed and protected areas such as the filter housing. The pilot’s hand primer was found in the
open and unlocked position, indicating that he may have been hand pumping fuel.
8. Airframe Structures
There was substantial crushing (hydroforming) of the aircraft belly from the fire wall to empennage. The floor did not deform
significantly and the seats remained relatively undamaged and attached. Overall, the impact damage was moderate and
concentrated on the front and bottom of the aircraft. The left wing remained attached to the aircraft and was undamaged except for
the tip.
The right wing remained at the accident site.
There was substantial corrosion of steel parts and the magnesium control wheel had completely dissolved. However, overall corrosion
was not as severe as expected.
9. General wiring and plumbing
There were several non-conforming wiring installations associated with accessory electronic parts. The non-conformities included the
use of uncertified non-aircraft wire, bundling of large coils of surplus wire length and inappropriate use of crimped connectors. There
was an automotive style cigarette lighter wired into the aircraft system and attached to the pilot side pedestal.
Oil Deposits
An oily film was found on the windscreens and the right side of the fuselage. It was suspected that these deposits may have been an
indication of major inflight oil loss due to engine failure or hose rupture. The deposits on the windscreens were examined with an optical zoom stereo microscope. Five samples were collected for scanning electron microscope and chemical analysis from the inside and outside of both left and right windscreens and a representative fuselage sample.
Visually, the inside and outside surfaces of the windscreen looked similar. They appeared to have a relatively uniform oily/greasy layer with sporadic filiform organics in some areas. On top of that was a variety of environmental contaminants – dust, salt, pollen etc. As confirming evidence, we looked for the unique chemical signatures of used engine oil – specifically lead and bromine, using the EDS (energy dispersive spectrometer) capability of the scanning electron microscope. We found these elements in all five specimens. There was a relatively uniform layer of engine oil on the inside and outside of both right and left windscreen and the right side of the fuselage from the firewall to the tail cone.
However, I do not think that this oil was deposited on the aircraft in flight for the following reasons;
- • The deposit on the fuselage is present in the protected wing root area. That area only became exposed when the right wing
separated at impact. Therefore the oil deposit on the right side of the fuselage could only have occurred sometime after the wing
separation.
• The oil deposit being on both the inside and the outside of the windscreen reduces the likelihood of inflight oil leak.
• The pattern of the oil deposit on the right side of the fuselage is inconsistent with inflight oil loss. When oil is lost in flight, it streaks
rearward, leaving clean spots behind rivet heads, sheet metal joints and other protuberances. These typical features were absent.
Carburetor Fitting
A brass reducer was found separated from the inlet to the carburetor when the aircraft was recovered. The carburetor inlet had been
repaired by installing a brass insert into the body, but this would not allow the standard 90 degree steel elbow fitting to be installed. This
brass reducer was apparently installed to bridge these parts.
The brass insert in the carburetor body is apparently a repair scheme developed by the overhauler without manufacturer, type certificate holder or Transport Canada approval. It was reported that the carburetor was removed from another aircraft and installed on C-F C-GAQW without proper documentation or certification. The airworthiness of this installation is suspect.
The reducer is a 3/8 NPT (National Pipe Thread) on the male end and ½ NPT on the female side. It was not a certified aircraft part and was not part of the normal installation. The reducer is likely a commercial plumbing fitting for boilers or similar device. The reducer constricted the fuel flow at that location by about 40%.
Floats
The aircraft had a set of EDO 58-4580 floats installed. The right float was substantially damaged by impact forces and contact with the propeller while the left float was relatively intact. Both floats exhibited fairly even compression damage at the rear, which indicates a wings-level attitude at initial contact.
Both floats exhibited numerous patches and large areas of thin zinc chromate paint spray. The zinc chromate was a questionable repair for areas with missing paint that had already exhibited signs of corrosion. For reference, the float compartments in each float were numbered from one through five, front to back.
The right float compartments 1, 2, 3, and 4 had sustained substantial impact damage. Propeller slash marks were visible on the deck and sides of compartments 1 and 2, and the float had been torn into two sections through the step compartment. The sides and bottoms of compartments 1 and 2 were skewed left; consistent with the forward half of the right float being side loaded inward under the engine and propeller during the impact sequence. The bottom skin on compartment 4 was torn open. Compartment 5 was intact but the keel and bottom skins were displaced inward. The impact damage to the right float would have precluded compartments 1, 2, 3 and 4 from providing any positive buoyancy after he accident.
The left float (SN 1441), was relatively intact. There was a hole in each sidewall of compartment 1 introduced during salvage. There was a large hole in the float deck above compartment 3, near the rear strut attachment fitting that could be either impact or salvage damage. There were small recovery damage holes in the outboard bottom skin of compartment 4 and 5.
All five left float compartment hatch covers had sealing defects; sealant was absent under the edges of the hatch covers on compartments 1, 2, and 3 and the 4 and 5 hatch cover seals were badly deteriorated and ineffective.
Most of the pump-out ports on both floats were open and the ones that were sealed used halved rubber balls instead of the expanding and locking plugs called for by the type design.
Buoyancy
Total float buoyancy for the EDO 58-4580 floats was 9,160 pounds in fresh water and about 9,435 pounds in salt water. The actual empty
weight of the aircraft as recovered (with right wing detached) is unknown but would be less than the book figure of approximately 3,300 pounds. All of the materials that are present in the aircraft weigh less when submerged in water than in air – even the heavy materials like steel. One hundred pounds of magnesium becomes 41 pounds when submerged and 100 pounds of aluminum – the major constituent of the aircraft weighs only 63 pounds when submerged – in fresh water. Submersion in salt water decreases the weight a further 3 %.
The displacement of water by submerged material lowers its weight but an even greater effect is achieved by trapped air, fuel and oil which have positive buoyancy. It is likely that between the loss of weight due to the separated right wing and the displacement of water by the wreckage, the submerged weight was less than 1500 pounds.
Investigator Bill Kemp from the Transportation Safety Board of Canada determined that the left float’s three good compartments were capable of providing about 2748 pounds of buoyancy and the one unbreached compartment in the right float would add another 916 pounds. Therefore, about 3664 pounds of buoyancy should have been available - more than sufficient to support the wreckage at the surface. Keeping the wreckage afloat would have provided the five initial survivors with a means of floatation and perhaps a perch out of the water. Furthermore, the floating wreckage would be far easier and quicker to locate by search aircraft and boats than a few bobbing heads. In a situation like this time is critical.
The separated right wing and the rest of the wreckage were found within 150 feet. Current at the surface was reported to be in the 5 to 7 knot range, so the fact that all the wreckage is close together indicates that the main aircraft and float portion sank almost as fast as the wing which would not have floated.
Obviously, the four good compartments flooded and lost buoyancy quickly. Water likely entered through unsealed or poorly sealed hatch covers, unplugged pump-out ports, leaks between watertight compartments and through various seems and joints of the float structure. The aircraft likely sank in one or two minutes.
Conclusions
The R.J. Waldron & Co. investigation determined that the aircraft crashed while the engine was at a low power condition, possibly indicating an intention to land – either forced or precautionary. There was no evidence of mechanical failure of the engine but it was not possible to determine if the carburetor, fuel, or ignition systems were functioning properly.
The rapid and unexpected sinking of the floats was the likely cause that turned this survivable accident into five fatalities.
Recommendations to prevent a recurrence of this accident and to remedy identified safety deficiencies are contained in a separate document addressed to the BC Coroner Service.
Prepared by:
George Heath
R.J. Waldron & Co. Ltd.