MEL Dispatch

The airborne accounts are settled (articles 08–10); this closing article settles the ground ones. It reflects some operators' MEL practice — MELs differ by operator and authority, so treat everything here as a worked example of the reasoning, not as your company's numbers. Read any ATA-36 MEL entry through three questions: does this failure touch ETOPS (APU bleed is the extended-range backup air source)? does it touch high-elevation airfields (where performance takeoffs lean on APU bleed and pressurisation margins are thin — article 07)? and is the remaining detection/protection redundancy still adequate (leak loops, the OPV, the BMCs)? Every condition below answers one of those three.

[!warning]- Scope note This article paraphrases operator-MEL logic — no manual text is quoted verbatim, and specific repair intervals, item numbers and conditions vary between operators. The mechanisms invoked are those of articles 01–10.

1. The alert-to-dispatch map — and four hard stops

Typical operators' MELs route each ECAM alert of this chapter to a dispatch item — with four exceptions that end the conversation:

The fourth line is the teaching moment. Three of the four are active failures; the last is only lost monitoring — treated as severely as a live leak. The logic (article 06): the wing loops guard the anti-ice ducting along the leading edge and the fuel-tank neighbourhood, and there is no substitute monitoring and no way to render that ducting cold — wing anti-ice and the packs still need the wing ducts hot. Dispatching blind would mean flying a sector in which a real leak goes undetected. Contrast the APU loop below (section 5), whose duct can be made cold — and which therefore has dispatch paths. A useful transferable principle: a detector's dispatch treatment mirrors whether the thing it guards can be de-energised.

At the other extreme, one alert typically has no dispatch item at all: BLEED HI TEMP. Article 09 explains why — it is a demand-side event, not a broken part. Perform the procedure; there is nothing to defer.

2. One engine bleed system inoperative — the six-condition template

The chapter's central dispatch case. In some operators' MELs it comes in two flavours — a non-ETOPS version at a standard repair interval, and an ETOPS-restricted version tightened to roughly a single sector — sharing a conditions list of about six items:

1. ETOPS restricted (either not flown at all, or capped — e.g. not beyond 180 minutes);
2. the affected ENG BLEED pushbutton selected OFF — the failed side is entombed (article 09's end state);
3. the crossbleed selected OPEN — the healthy side feeds the whole network for the entire flight (article 09's borrowing configuration, made permanent);
4. APU and APU bleed operative — the backup behind the backup: if the surviving bleed also fails, APU air is the only ticket to the FL 220 regime of article 10;
5. no dispatch to high-elevation destinations (thin pressurisation and performance margins — article 07);
6. must be operative when departing a maintenance base — fix it where fixing is possible.

Each condition is an article of this chapter compressed into a line. The accompanying operational procedure adds the sharpest sentence in the whole ATA-36 MEL:

[!warning]- Wing anti-ice above 31 000 ft can kill the other bleed too Some operators' MEL procedures warn that with one bleed system already inoperative, using wing anti-ice above 31 000 ft may cause the opposite (surviving) engine bleed system to fail. Mechanism: one bleed is feeding both wings' anti-ice through the crossbleed; at altitude the demand spike drives the survivor toward its overtemperature limits (article 09's HI TEMP threshold set includes an anti-ice-above-FL-320 condition — same physics). Route-planning in icing season starts with "can we stay below 31 000?" The same procedure notes the forward-cargo cooling loss and — most usefully — that if the opposite side then fails in flight, the crew performs the corresponding ECAM and then the AIR ENG 1+2 BLEED FAULT procedure, resetting the placarded pushbutton as commanded even though it is placarded inoperative. A placard never outranks an ECAM in the air.

3. Components: one default, two elegant exceptions

The default — "component out = side out". The bleed valve (secured closed), the fan air valve, the precooler and (in one variant) the IP check valve are each typically dispatchable one-side-inoperative by treating the whole engine bleed system as inoperative — which routes straight back to section 2's six conditions. Article 03's verdict made contractual: a bleed that cannot regulate its temperature or pressure does not enter the network.

Exception one — the OPV can be waived on its history. Some operators' MELs allow both overpressure valves inoperative, on one condition: no overpressure-caused BLEED FAULT occurred on the preceding flight. Why so generous with a safety valve? Articles 02/05: the OPV is the third tier of overpressure defence, behind the PRV's own regulation and the BMC's 60-psig electrical closure. With the first two tiers demonstrably healthy (a clean recent history), the backstop may be absent for a while. A backstop's dispatchability is priced on the health of the primary defences. The alternative variant — treat the side as inoperative — covers the case where that history is not clean.

Exception two — the HPV family and the "low-power corridor". A stuck-closed or deliberately secured-closed HPV does not kill the side; it only empties article 02's Cases 1–2. Typical conditions: the affected bleed not used at low power settings, and the opposite system operative. The matching operational procedure is a small masterpiece of profile flying:

• Taxi (engines at idle): affected ENG BLEED — OFF; X-BLEED — OPEN; perform the resulting ABNORM BLEED CONFIG procedure (article 09).
• Before takeoff thrust, until descent: affected ENG BLEED — ON; X-BLEED — AUTO; normal configuration restored (at high power, IP feeds everything — the HPV was going to be closed anyway).
• Before descent, until landing: OFF and OPEN again — idle thrust is Case-2 territory and this side has no Case 2.

Fly the profile in your head once: the crossbleed covers exactly the segments the HPV would have covered. When the valve cannot work its shift, the crew rosters the crossbleed instead. A stuck-open IP check valve gets the mirror treatment (secure the HPV closed, then the same corridor), since an IPC that cannot block means HP air must never be admitted behind it.

The quietest entry — one BMC. Typically: one may be inoperative, standard interval, no further conditions. The shortest line in the chapter's MEL, and the entire justification is article 05's architecture: the opposite computer keeps four alerts alive, the FAULT-light loss is covered by its own indicator logic, and the lost auto-closure has a procedural human backup (article 09's BMC FAULT branch). Architectural redundancy converts directly into dispatch freedom — the same trade seen across Airbus MELs.

Indication items follow the same grammar: BLEED-page symbols and pushbutton lights are individually deferrable (the ENG BLEED FAULT light typically requires the related BMC operative — a dead bulb is acceptable only while the brain behind it works), with one instructive chain: losing the APU bleed-valve indication drags the crossbleed AUTO control down with it, because the AUTO logic's opening condition is "APU valve fully open" (article 04) — blind the input and the logic is declared inoperative, which routes to the manual-control procedure below.

4. The APU side: ETOPS and high fields, twice over

APU bleed system, APU bleed valve, APU check valve — the three carry a shared conditions core in some operators' MELs: ETOPS capped (typically at 180 minutes), the relevant control off or the component secured, and no high-elevation-destination dispatch. The reasoning is article 10 read backwards: without APU bleed, a subsequent single-bleed failure has no FL 220 regime — and article 07's high-field performance takeoffs lean on APU air. A valve stuck open adds "APU not used in flight" (an uncommanded supply path must not meet the flight envelope), plus ground rules to keep the pushbutton matched to APU state so the position-disagree alert stays quiet, and APU off before takeoff.

Their operational procedures share two house rules: no APU bleed for main-engine starts (use the air-start-unit procedure of article 07) and no single-engine taxi — the delayed second start depends on APU air or a crossbleed start, neither guaranteed healthy under these placards.

Crossbleed valve controls — two motors insuring each other. Article 04's independent AUTO and MANUAL motors become two mirror MEL items: automatic control inoperative is dispatchable if the manual control is checked operative, and vice versa — each check performed on the BLEED page during cockpit preparation (select, watch the symbol move, restore). With AUTO dead, the crew becomes the relay for the whole day: X-BLEED OPEN whenever APU bleed is on or an engine start is in progress, CLOSE otherwise — a human implementation of the AUTO law, and the standing version of article 09's X BLEED FAULT technique.

5. The APU leak loop: three roads, three prices

The counterpoint to section 1's wing-loop hard stop. With both APU loops failed (the DET FAULT of article 08), some operators' MELs offer three dispatch paths, each an implementation of one idea — if the nerves are blind, make the duct cold:

1. Remove the APU check valve and blank the duct — the backflow path of the FED-BY-ENG case (article 06) is physically deleted; APU bleed treated inoperative (standard interval).
2. Verify the check valve operative before every flight — a short-life variant (days, not weeks); APU bleed still treated inoperative.
3. Entomb the left corridor without touching hardware — no ETOPS; no APU-leak alert on the preceding sector; and a cockpit-prep configuration of ENG 1 BLEED OFF, PACK 1 OFF, X-BLEED CLOSED, APU BLEED OFF; no flight in known or forecast icing (the surviving right side alone cannot warrant anti-ice for both wings).

Path 3's operational day is worth walking: for engine start the crossbleed opens and APU bleed is permitted — the start window again, mirroring article 06's MES override — then everything closes back down after start; in flight, if the one working bleed fails, the crew goes straight to the AIR ENG 1+2 BLEED FAULT procedure of article 10. Related single-warning items (the pylon-leak warning, the FED-BY-ENG warning) follow the same recipe: deactivate the loop, confirm no actual leak, and physically address the check valve.

A closing note on the smallest print: a single failed loop generates no ECAM at all (article 06's AND logic) — it surfaces as a maintenance-level status message, and the MEL's interest begins only when the pair is gone.

6. The chapter's three faces — a closing map

Draw article 01's H-network one last time and annotate each component with its dispatch face: the wing corridors are sacred (their nerves included), the crossbar is insured twice, the APU stub trades against ETOPS, and the brains are cheap because there are two. One diagram now carries the system, its failures and its paperwork — the chapter closed.

Self-test

[!note]- Q1. Name the four hard stops and explain why one of them is "only" a detection failure.

Bleed leak, dual engine bleed failure, wing leak — and wing-leak detection failure. The wing loops guard anti-ice ducting and fuel-tank surroundings that must stay hot in service and have no alternative monitoring; blind dispatch would mean an undetectable real leak. (Contrast the APU loop, whose duct can be made cold.)

[!note]- Q2. Recite the six-condition template for one bleed inoperative and tie each condition to its mechanism.

ETOPS restricted (APU bleed is the extended-range backup); affected pushbutton OFF (entomb the failure); crossbleed OPEN (one side feeds all — article 09's configuration); APU and APU bleed operative (the FL 220 ticket if the survivor fails — article 10); no high-elevation destinations (thin margins — article 07); operative out of a maintenance base (fix where fixable).

[!note]- Q3. Why can wing anti-ice above 31 000 ft endanger the surviving bleed?

One bleed is heating both wings via the crossbleed; at altitude that demand drives the survivor toward its overtemperature limits (the HI TEMP trigger set already flags anti-ice above FL 320). The de-icing load can kill the last air source — plan icing-season routes below 31 000 ft.

[!note]- Q4. Describe the HPV "low-power corridor" and state why the system needn't be declared dead.

Taxi and descent-to-landing (idle segments): affected bleed OFF, crossbleed OPEN — borrow air exactly where the HPV would have worked. Takeoff through cruise: bleed ON, crossbleed AUTO — IP supplies everything at high power. The HPV only matters in Cases 1–2; the crossbleed covers its shift.

[!note]- Q5. On what logic can BOTH overpressure valves be inoperative, and what single fact revokes it?

The OPV is the third defence tier behind PRV regulation and the BMC's 60-psig closure; with those proven healthy, the backstop may be absent. An overpressure-caused BLEED FAULT on the preceding flight revokes the waiver — the primary defences are no longer demonstrably sound.

[!note]- Q6. Why does one dead BMC carry no conditions at all?

Article 05's architecture: the opposite BMC maintains four alerts, and the lost auto-closure has a trained human replacement (the BMC FAULT branch). Redundancy priced directly into dispatch freedom.

[!note]- Q7. The APU-loop "path 3" entombs the left corridor. Which four switches, which two prohibitions, and which exception window?

ENG 1 BLEED OFF, PACK 1 OFF, X-BLEED CLOSED, APU BLEED OFF; no ETOPS and no known/forecast icing. Exception: for engine start the crossbleed opens and APU bleed is permitted — the start window, then back to the entombed configuration.

Key takeaways

References

This article is a synthesis of some operators' MEL practice for ATA-36 (dispatch entries for the chapter's ECAM alerts; the engine-bleed-system, bleed-valve, OPV, FAV, precooler, IP-check-valve, HPV, BMC, indication, APU-bleed-system/valve/check-valve, crossbleed-control and APU-leak-loop items with their operational procedures), deliberately paraphrased rather than quoted: item numbering, repair intervals and exact conditions differ between operators, and the current operator MEL is always controlling. The mechanisms invoked are those documented in FCOM DSC-36 / PRO-ABN-AIR, AMM 36-11-00 / 36-12-00 / 36-21-00 / 36-22-00 and FCTM references as developed across articles 01–10 (crossbleed geography and motors, closure inventories, defence tiers, detection logic, the FL 220/100 regimes). The "three faces" map and the corridor framing are integrative syntheses.

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.