australian aviation logo

New details come to light in 2017 seaplane crash

written by Hannah Dowling | July 3, 2020

The Australian Transport Safety Bureau has released an interim report on its ongoing investigation into a Sydney Seaplanes Beaver accident from December 2017, which saw the pilot and five passengers lose their lives.

On 31 December 2017, a Sydney Seaplanes DHC-2 Beaver floatplane departed from Cottage Point in Sydney’s north, on a planned return trip to Rose Bay.

Shortly after take-off, the aircraft deviated from the operator’s standard flight path, stopped climbing, and entered the confines of Jerusalem Bay below the height of surrounding terrain. 

The aircraft continued along the bay, made a steep right turn, and then collided with the water.

The recent interim report has suggested that the pilot and passengers could have been subjected to carbon monoxide within the aircraft cabin, likely causing the pilot to lose control of the plane.


“During the draft review process for the investigation’s final report, the aviation medical specialist engaged by the ATSB recommended that carbon monoxide toxicology testing be undertaken on blood samples of the aircraft occupants,” said ATSB chief commissioner Greg Hood.

The results of that testing, provided to the ATSB in March 2020, indicated that the pilot and two of the passengers had elevated levels of carbon monoxide, a likely culprit for the cause of the crash.

“From that consultation with medical experts, and research into the effects of carbon monoxide on aircraft operations, the ATSB considers the levels of carbon monoxide were likely to have adversely affected the pilot’s ability to control the aircraft,” Hood said.

The ATSB then re-examined the aircraft and undertook testing on an exemplar Beaver aircraft to replicate the potential source of carbon monoxide and ingress into the aircraft cabin, and it appears that the aircraft’s engine exhaust had a crack that existed prior to the plane crash.

“Having discounted other potential sources of carbon monoxide exposure, the ATSB considers it likely that the pilot and passengers were exposed to carbon monoxide inside the aircraft cabin,” Hood said.

The ATSB found a pre-existing crack in the engine exhaust collector ring, which is believed could lead to “exhaust leakage” into the engine bay.

“Further, the ATSB found a breach in the firewall from missing bolts used to secure magneto access panels in the firewall under the instrument panel in the cabin. Any breach in the firewall can allow the ingress of gases from the engine bay into the cabin,” Hood said. 

The investigation will continue and a final report will be released over the coming months.

“This investigation is on-going, and our final report, which will contain specific findings, is anticipated to be released in coming months, so we are limited in discussing specific details,” Hood said.

“However, if at any time during an investigation, should the ATSB identify issues that are critical to safety, we will immediately notify relevant stakeholders so proactive safety action can be taken to help prevent similar occurrences.”

In response to the findings, the ATSB has now published two Safety Advisory Notices, with a focus on the prevention and detection of carbon monoxide in piston-engine aircraft.

Although the accident aircraft involved a DHC-2 Beaver, these issues are relevant to piston-engine aircraft in general, Hood noted.

“The ATSB is reminding aircraft maintainers that the primary mechanism for the prevention of carbon monoxide exposure to aircraft occupants is to carry out regular inspections of aircraft exhaust systems to identify and repair holes and cracks, and to detect breaches in the firewall,” he said.

The ATSB is also highlighting the limitations of disposable carbon monoxide chemical spot detectors, as used commonly in general aviation, and was fitted to the accident aircraft.

Spot detectors have a limited shelf-life, can be affected by factors such as direct sunlight and cleaning chemicals, and are passive, relying on pilots to regularly monitor them.

“In contrast, electronic active carbon monoxide detectors are designed to attract the pilot’s attention through auditory and/or visual alerts when carbon monoxide levels are elevated,” Hood said.

“These detectors are now inexpensive and widely available. Had there been an alert of the presence of carbon monoxide, the pilot would have been able to take measures to reduce the risk to those on board.”

The Civil Aviation Safety Authority (CASA) has been informed as to the investigation’s progress. 

To date, CASA has contacted all operators and owners of the 20 DHC-2 Beaver aircraft registered in Australia to emphasise the importance of inspections of the exhaust system, to confirm that the scheduled inspections were being conducted, and to seek information pertaining to the number of exhaust ring segments requiring repair or replacement.

In addition, CASA has published an Airworthiness Bulletin in order to highlight the risks and dangers of carbon monoxide poisoning to all piston-engine owners, operators and aircraft engineers, and advising of the fitment of active carbon monoxide detectors.

You need to be a member to post comments. Become a member today!

Comment (1)

  • Marum


    I watched footage of that crash over and over. It was immediately obvious that he did not have sufficient power on, to make a turn of that degree and maintain his altitude. Normally, in under powered piston engine aircraft of that vintage, full power has to be applied to even maintain one’s altitude in a sharp turn. The aircraft was too low and too slow. From the report, it appears to be a maintenance issue. If the firewall had been sealed properly the accident may never have happened.

    I am constantly amazed by people’s lack of appreciation of the dangers of CO. Even qualified people such as motor mechanics, have been found dead in Pump Sheds, and Generating Rooms, due to CO gases present.In this day and age, a good CO detector can be purchased for under AUD $100.00.

    Has anyone set a standard for CO detectors yet? Similar to the one they set a few years ago, for home smoke detectors? ie. suitable for unpressurized light aircraft at all altitudes. I am not sure whether altitude makes a difference to the sensitivity of a CO detector.


Comments are closed.

You don't have credit card details available. You will be redirected to update payment method page. Click OK to continue.