Fuel Leak and Feedline Burst
A fuel leak is the abnormal where doing the obvious thing — balancing, crossfeeding — makes it worse. The QRH answer is a detective procedure: isolate the plumbing, watch which tank keeps losing, and let the arithmetic name the leak. This article walks the recognition cues, the isolation decision tree with its three endings, the loss-reduction follow-up, and the specialised case with its own alert — the engine feedline burst.
1. Recognition — ten ways a leak introduces itself
Suspect a leak on any of: FOB + fuel used significantly below the departure figure, or decreasing; visible spray (engine, pylon, wing tip — a cabin-window check is part of the drill when conditions allow); total quantity falling abnormally fast; a developing imbalance; one tank dropping too fast; a tank overflowing (an in-tank pipe rupture pushes fuel up as well as out); high fuel flow with low N1; fuel smell in the cabin; destination EFOB turning amber or falling abnormally on the FMS.
The standing FCTM habit makes these cues visible: fuel checks every waypoint or 30 minutes, and any quantity/imbalance anomaly treated as a possible leak first (fuel management, fuel imbalance).
Once suspected: LAND ASAP (amber) — the investigation below happens en route to a runway, not instead of one.
2. The decision tree
Leak already pinned to an engine/pylon (visible, or FF far above N1's story):
THR LEVER (affected) ......... IDLE
ENG MASTER (affected) ......... OFF
WING X FEED ............... AS RQRD
DO NOT RESTART THE AFFECTED ENGINE
Leak not located — build the isolation box:
WING X FEED ......... MAINTAIN CLOSED ← each side keeps its own books
L + R CTR PUMP ................. OFF ← (six-tank) stop feeding the gallery
T TANK FEED ................... ISOL ← lock the trim line
INNER TANKS QUANTITIES ...... MONITOR ← now watch the arithmetic
With all transfers stopped and the system split in two, whichever inner keeps falling is on the leaking side — and the tree forks on what the numbers do:
◆ One inner falls ≥500 kg in 30 min more than the other → engine still suspect:
THR LEVER (leak side) ........ IDLE
ENG MASTER (leak side) ........ OFF
FUEL LEAK ................. MONITOR
├─ leak STOPS → ENGINE LEAK CONFIRMED:
│ restore transfers (CTR pumps ON, T TANK FEED AUTO),
│ X FEED as required, DO NOT RESTART
└─ leak CONTINUES → WING LEAK SUSPECTED:
ENGINE RESTART .......... CONSIDER ← engine exonerated
FUEL LOSS REDUCTION ..... CONSIDER (Section 3)
⚠ DO NOT APPLY THE FUEL IMBALANCE PROCEDURE
(one tank full, the other empty is acceptable
for approach and landing)
◆ Both inners fall at the same rate → the leak is in the centre/trim/APU plumbing:
L + R CTR PUMP ................. ON ← wing books proved clean
├─ fuel smell in cabin:
│ APU ..................... OFF
│ T TANK FEED .... MAINTAIN ISOL ← tail line stays prime suspect
└─ no smell → CENTER/TRIM TANK LEAK SUSPECTED:
FUEL LOSS REDUCTION ..... CONSIDER
Landing, all cases: DO NOT USE REVERSERS — even if available; compute landing performance without them. Reversers throw air (and leaked fuel) forward around a hot nacelle.
[!warning]- The 500 kg / 30 min figure is a floor, not a definition It is the smallest difference the FQI's accuracy can certify in that time. A smaller but steady or growing asymmetry under full isolation is still a leak — treat it as one (FCTM). The number protects the procedure from gauge noise, not you from small leaks.
3. Fuel loss reduction — saving what can be saved
When the leak procedure recommends it (wing leak, or centre/trim leak):
◆ centre or trim tank not empty (six-tank rows):
L + R CTR PUMP .............. ON
CTR TK XFR ................. MAN ← manual, symmetric forward feed —
T TANK FEED (not empty) ... AUTO bypasses the automatic "top up the
T TANK MODE (not empty) .... FWD emptiest (= leaking) tank" bias
▸ trim empty: T TANK MODE AUTO · T TANK FEED ISOL ← lock the tail line
▸ centre empty: CTR TK XFR AUTO
The point of the manual settings: automatic transfer logic prefers the emptiest tank — which, in a leak, is the leaking tank. Manual symmetric transfer gets the recoverable fuel moving toward the engines without volunteering more to the hole. Note trim's end state is ISOL, not AUTO — in any leak story the tail line stays locked once drained.
4. Feedline burst — the leak with its own alert
A rupture of an engine's feed line is detected and announced: FUEL ENG 1(2) FEEDLINE BURST (on earlier-standard aircraft the same failure presents through the WING PUMPS LO PR rupture branch — pump failures).
LAND ASAP
WING X FEED ............ DO NOT OPEN ← the hard prohibition
GRAVITY FEED CEILING:
FL200 — airborne > 30 min
FL150 — airborne < 30 min
JP4/JET B: FL100 (GE engines) / FL70 (RR engines)
ENG START SEL .................. IGN
THR LEVER (affected) .......... IDLE
PUMPS (affected side) .......... OFF ← stop pressurising the break
STBY PUMP (affected) ........... OFF
▸ no relight after 30 s: ENG MASTER (affected) OFF
▸ at the gravity ceiling: ENG RELIGHT ... CONSIDER
The logic in one line: the pumps are feeding the break, so the pumps go off — which starves that engine to flame-out; the descent to the gravity ceiling then restores unpumped feed, and the engine (which was never damaged, only starved) becomes a relight candidate. STATUS carries the bill: gravity ceilings as altitude limits, the affected side's pumps inoperative, that inner tank partly unusable (the gravity-feed 2 t rule), and — six-tank — aft transfer inoperative once the centre is empty.
The crossfeed prohibition outranks everything: even a perfect crossfeed valve must stay shut, or the healthy side's fuel pumps overboard through the same break (crossfeed).
Self-test
[!note]- Q1. Why must the imbalance procedure never run during an unlocated leak? Balancing feeds the healthy side's fuel toward the leak (and an open crossfeed destroys the per-side bookkeeping that locates it). The leak drill explicitly forbids it — a lopsided landing is acceptable; an empty aircraft is not.
[!note]- Q2. What does the isolation box (X FEED closed, CTR pumps off, trim ISOL) achieve? It splits the system into two closed accounts so the falling account identifies the leaking side — the procedure's entire diagnostic engine.
[!note]- Q3. After shutting the suspect engine down, the loss continues. What have you learned, and what may you do? The engine is innocent — the wing side is leaking. Consider restarting the engine (you need it), consider fuel loss reduction, and do not balance.
[!note]- Q4. Both inners fall together under isolation, and there is a fuel smell in the cabin. Configuration? Centre pumps back ON (wings are clean), APU OFF, trim line kept isolated — the centre/trim/APU plumbing is the suspect, and the smell points at the tail line.
[!note]- Q5. In a feedline burst, why are the affected side's pumps switched off when that starves the engine? Pressurised pumps feed the rupture overboard. Stopping them trades the engine (temporarily) for the fuel; below the gravity ceiling the engine can be relit on unpumped feed.
Key takeaways
| Point | Value |
|---|---|
| First reflex | conservation check; any anomaly = leak until cleared; LAND ASAP once suspected |
| Method | isolate (X FEED / CTR pumps / trim ISOL) → watch per-side books |
| Fork | ≥500 kg/30 min one side → engine test; same rate both → centre/trim plumbing |
| Hard bans | no imbalance procedure; no reversers at landing; no crossfeed after feedline burst |
| Loss reduction | manual symmetric transfers; trim line ends ISOL |
| Feedline burst | pumps off = stop the leak; gravity ceilings; relight candidate at ceiling |
References
- QRH fuel leak / fuel loss reduction procedures + FCOM PRO-ABN-FUEL expanded versions (tree, both branches, configuration rows).
- FCOM PRO-ABN-FUEL FEEDLINE BURST (procedure and STATUS); FCTM (leak-first discipline, 500 kg floor reading).
- The "two closed accounts" framing is integrative synthesis of the isolation logic.
Independent study material, not an Airbus publication and not endorsed by the manufacturer. Always defer to the current operator FCOM, FCTM, and QRH for operational use.