When a fuel sample is all water

An AAIB investigation into an engine power loss shortly after takeoff highlights a deceptive fuel-sampling hazard: a sample consisting entirely of water can still carry an aviation-fuel odour.

What the investigation found

The aircraft suffered engine power loss shortly after takeoff and departed controlled flight. Examination found significant water contamination throughout the fuel system. A cracked wing-filler-cap grommet could have allowed rainwater to enter. The AAIB also highlighted that a sample tube filled entirely with water may still smell of aviation gasoline.

Why an apparently clear sample can mislead

Pilots are often taught to look for a visible interface between fuel and water. That cue may be absent when the container holds only water. Smell alone is not a reliable identification method, especially when the sample has contacted fuel.

Defences to review

• Use the sampling method and every drain point specified for the aircraft.
• Know the expected colour and other approved ways of confirming that the sample is fuel.
• Investigate unexpected quantity, repeated contamination or damaged seals rather than repeatedly draining and accepting the result.
• Consider exposure to rain, outdoor parking and recent refuelling when assessing the threat.
• Have defects inspected and rectified under the applicable maintenance requirements.

Learning takeaway

A completed drain action is not the objective; positively identifying an uncontaminated fuel sample is. Follow the aircraft manufacturer's instructions and current regulator guidance rather than relying on a single sensory cue.

Warning signs to notice earlier

• A clear sample is accepted because no fuel-water boundary is visible.
• Fuel smell is treated as proof even though water has contacted fuel in the sampling system.
• Repeated contamination, rain exposure or a damaged filler seal is normalised instead of investigated.

Decision analysis

The occurrence challenges a familiar visual test. A boundary appears only when both liquids are present in the sample. If the container holds water alone, it can look uniformly clear, and residual aviation-fuel odour can reinforce the wrong conclusion. Positive identification must therefore use the aircraft manufacturer's sampling method and known characteristics of the expected fuel.

Sampling is part of a system that includes sound caps and seals, correct parking and refuelling practices, every specified drain point and maintenance response to repeated contamination. Draining until an expected appearance emerges does not repair an ingress path.

Safer alternatives to discuss

If quantity, colour, contamination or seal condition is unexpected, stop the departure and obtain competent inspection. Compare the sample with a known fuel sample only where the approved procedure supports that method, and handle drained fuel safely. Do not return a suspect sample to the tank.

Use the report to audit the local routine: are pilots trained to identify an all-water sample, are drains and sample tools suitable, and is repeated water recorded as a defect? A checklist tick is not evidence that the substance was correctly identified.

Instructor or group debrief

1. Why might no interface be visible?
2. Which approved cues identify the expected fuel?
3. What aircraft defect could allow repeated water entry?
4. At what point must sampling stop and maintenance begin?