Watching a documentary and reading up on the Air Transat 236 incident, I would have thought that even with the tech available in 2001, it would have been simpler to spot a fuel leak, even with the strange warnings going on. Don't aircraft have a "pounds per hour" real-time display that indicates rate of fuel expenditure?

If so, would such a device have been taken to have malfunctioned or showing incorrect information, or would the difference it showed at the time of the leak be unnoticeable? It is said that the pilots were checking fuel balances half hourly or so before the first alarm. Would a "pounds per hour" display be a good up-to-the-second indicator, that even if the last fuel check was ok, there was now a steady problem? Would it remove be necessary to make any other calculations or do any other guesswork? Was this figure/display not available on this flight?


1 Answer 1


Fuel flow is measured by flow transducers in line with the fuel pipes. And they do a very good job at measuring flow, but unfortunately a fuel leak bypasses them. There is no direct measurement of the leak flow anywhere.

The pilots of Air Transat 236 had two indications of the fuel leak:

  • Increased fuel flow through the oil/fuel heat exchanger caused a drop in oil temperature and a corresponding increase in oil pressure.
  • Drained fuel on the leaky side caused a fuel imbalance.

It would only be possible to positively identify increased flow as a leak if the leak occurs in between two flow transducers, both mounted in the same fuel line. If the leak occurs after a single transducer, the increased flow gives an indication. Otherwise, the leak flow would have to be observed from available volume, minus legitimately used engine flow. There are fuel tank probes distributed in airliner fuel tanks that give an indication of fuel volume per tank. Picture below shows the six fuel probes in the 747 centre tank (source).

enter image description here

The difficulty here is that leak flow must be derived from volume measurement: unit of flow is $m^3$/sec, the derivative of volume. The wing tanks are long and flat and the probes measure vertical distance, with an accuracy that suffices for Qty indication but lacks the precision of the fuel flow meters which directly measure impulse of the mass flowing through them. Plus they measure mass directly, a more useful entity than volume. Volume per mass changes with fuel temperature, tank levels change with aircraft pitch and roll, the level needs to be derived over time: all of this can only give an indication over the course of hours, when a large amount of fuel has leaked out already, and with a degree of uncertainty that would leave room for doubt.

You're right though, in wondering why there seemed to be no automated way for monitoring discrepancies in fuel quantity. @Waked mentions the estimation messages of the B737NG, which would also only seem to occur after much fuel has been lost aready. The A330 now has an ECAM caution message: FUEL FU/FOB DISCREPANCY, put in place after the Air Transat incident. Lessons learned from this incident would now be incorporated in the fuel management systems and aircrew procedures. The brilliant glide flight of the pilots demonstrates once again why we need pilots in front of the aeroplane.

  • 1
    $\begingroup$ Fuel qty discrepancies are indeed monitored in some aircraft. The following applies to B737 NG: FMC will present the message "USING RESERVE FUEL" if predicted landing fuel is below the minimum FMC reserve fuel (typically final + alternate fuel). The prediction is based on FQIS measured fuel - FMC predicted fuel burn. When predicted landing fuel @ dest. is less than zero, "INSUFFICIENT FUEL" is shown. An "IMBAL" caution is presented if sufficient discrepancy exists between left and right fuel tanks. The relevant QRH checklist instructs pilots to confirm no fuel leak exists before crossfeeding. $\endgroup$
    – Waked
    Commented Jul 23, 2017 at 13:11
  • 1
    $\begingroup$ @Waked, well, the A330 checklist at the time of that incident did say not to use it if suspecting fuel leak. They even tried to look whether they are leaving a trail behind them, but they couldn't see it in the dark of the night. And what the checklist did not say, and they did not make that connection, was that if the imbalance does not improve in a couple of minutes, it should be considered a sign that you do have a fuel leak and turn it off and start planning a diversion. $\endgroup$
    – Jan Hudec
    Commented Jul 23, 2017 at 13:45
  • 1
    $\begingroup$ @JanHudec The purpose of my comment was to state that some form of "automated" monitoring does exist, not to in any way evaluate the performance of the crew. $\endgroup$
    – Waked
    Commented Jul 23, 2017 at 14:26
  • 1
    $\begingroup$ @Waked, yes. The crew can't be really blamed for not realizing the imbalance, if it does not fix quickly, does imply a fuel leak. The author of the check-list had a lot more time to think about it then they and didn't realize either (and therefore any automated system couldn't help either, because automated system can only do what the author of the requirements thought about). $\endgroup$
    – Jan Hudec
    Commented Jul 23, 2017 at 14:29
  • 2
    $\begingroup$ You don't need two transducers to see a leak, but the real problem is whether the leak occurs before or after the transducer which is measuring the flow to the engines... $\endgroup$
    – Lnafziger
    Commented Jul 24, 2017 at 18:08

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .