MEL Dispatch

The airborne stories are told (articles 11–12); this closing article settles the morning after. It reflects some operators' MEL practice — MELs differ by operator and authority, so read everything here as a worked example of the reasoning, never as your company's numbers. Three structural rules carry almost the whole chapter, and once you hold them the clauses stop needing memorisation: paired probes ground the aircraft; almost every ice-protection dispatch drags a weather fence — no flight into known or forecast icing — so the price of dispatch is a fence around your route; and everything to do with seeing out reads the PF's side, not the captain's.

[!warning]- Scope note This article paraphrases operator-MEL logic — no manual text is quoted verbatim, and item numbering, repair intervals and exact conditions vary between operators (several conditions noted below, such as high-elevation-airfield restrictions, are operator-specific additions). The mechanisms invoked are those of articles 01–12.

1. Three layers, one reading order

A typical ATA-30 MEL stands in three layers: an alert table (each of the chapter's ECAM alerts mapped to a dispatch verdict), the items (equipment entries with repair interval, installed/required numbers, placards and their (o)/(m) conditions), and the operational procedures (what the crew actually does after dispatching). The reading order is fixed:

ECAM alert → alert table
  ├─ "no dispatch"      → fix it, then fly (eight entries — §2)
  ├─ "not an MEL matter" → two entries (§2) — weather reports, not failures
  └─ "see item 30-xx"   → item conditions (interval / installed / required / (o)(m))
        └─ any (o) → operational procedure: preparation limits (CAT capability,
                     performance accounts) + phase-by-phase actions

Repair intervals follow the standard alphabet: A carries no standard interval — the clock is written into the item's own conditions; B gives three consecutive calendar days, excluding the day of discovery; C gives ten; D gives one hundred and twenty. Where each ATA-30 item lands on that alphabet is itself teaching material — the chapter's only B and only A items are both worth a section of their own (§5, §6).

2. Eight hard stops and two non-items

Of the chapter's alert-table entries, eight end the conversation — no dispatch:

The rule in one line: paired = grounded. This is article 11's "coherent but incorrect" mechanism mirrored onto the ground: two same-type probes without heat are, before the wheels even roll, set up to ice together and identically — the precondition for two agreeing liars outvoting the honest source. No source-switching scheme can protect the vote, so the ground answers with a veto. Both windshields is the same grammar for eyes: with two iced panes there is no "PF's side is still good" to retreat to. Meanwhile every single-point failure — one pitot, one AOA, one TAT, one static port, one windshield, a side window, one probe-heat computer — routes to a conditional item: architectural redundancy buys maintenance time; when the redundancy is spent, the aircraft stops.

The two "not an MEL matter" entries are ICE DETECTED and SEVERE ICE DETECTED. They report weather, not equipment (article 06); dispatch cares about the detectors themselves (§7), not about what they said yesterday.

3. Wing anti-ice valves — two locked positions, two lives

Article 12's disobedient valves can be manually locked for dispatch, and which position they are locked in writes two entirely different placards (four valves installed in either case).

Locked closed — required: zero; ten-day class. Typical conditions: no ETOPS beyond 180 minutes; no flight into known or forecast icing; and, in some operators' MELs, all valves must work for high-elevation destinations. A closed valve means an unprotected leading-edge section, so icing becomes forbidden terrain — the fence clause in its purest form — while long-range ETOPS wants the full anti-ice inventory behind it (article 10's icing-fuel machinery presumes the system works).

Locked open — required: two, one serviceable valve per side; with typical conditions that engine starts happen only at OAT at or below 19 °C, and all engine starts use APU bleed only (plus the operator-specific high-elevation addition).

[!warning]- Locked open demands more serviceable valves — and worries about the ground, not the sky Counterintuitive twice. A valve locked open still protects — that slat section is on permanent anti-ice, article 12's free gift — so flight capability survives and each side keeps one good valve to regulate with. The exposure has moved to the ground: during an engine start the manifold is pressurised, and a stuck-open valve feeds hot air onto a slat with no ram cooling (article 02's 30-second logic). The two start conditions read as insurance on exactly that leak — cold ambient air keeps the escaping heat harmless, and APU-only starts keep the manifold at APU regulation rather than the harder-working crossbleed-start regime, so the leak flow stays modest. (That reading is interpretation; the conditions themselves are the operator's.) The performance bill for flying this way is §8's first account.

4. Engine anti-ice valves — the strictest closed, the cheapest open

Locked closed (two installed, one required) collects the chapter's harshest condition set: typically no ETOPS at all, no flight into known or forecast icing, and in some operators' MELs further route restrictions on top (no high-elevation destinations among them). Note the gradient against the wing item: WAI locked closed banned ETOPS beyond 180 minutes; EAI locked closed bans ETOPS entirely. On a single-engine driftdown into an icing layer, the intake anti-ice of the surviving engine is that engine's bodyguard — and an engine eating its own intake ice is a structural risk, which is why the valve was designed to fail open in the first place (article 03, read backwards).

Locked open — one or both, and no weather fence at all: the anti-ice capability is intact, so what remains is arithmetic. The item sends you to a performance-loss procedure (in the operator's performance application), the fuel line of which is §8's second account — and the cockpit-preparation line is the mine-disarming move article 12 promised: select the affected engine's ANTI ICE pushbutton ON, so the FADEC's thrust datum matches the valve's reality and no ENG THRUST LOSS ambush waits at takeoff power.

5. The probe family — the visible-moisture fence, layer by layer

The family's items share a two-condition template, here in the probe-heat-computer version (three computers installed, two required): the ADR, heaters and fault warnings associated with the working components must themselves work (the voting floor of article 11 must be intact), and the aircraft does not fly in visible moisture or in known or forecast icing.

[!warning]- "Visible moisture" is a stricter fence than "icing conditions" Article 01's icing definition needs two ingredients — moisture and near-freezing temperature. The probe items fence on moisture alone: an unheated probe flying through cloud can be troubled by water without any classical icing forecast in sight. Ice protection dispatch generally trades against icing; probe dispatch trades against wetness.

On that template the family stacks its add-ons:

Two of those rows repay a pause. Pitot is the chapter's only B interval — three days against everyone else's ten. Pitot pressure is airspeed; the family's heart gets the shortest leash: you may fly, but don't linger. And the standby items' contaminated-runway clause has a concrete picture behind it (interpretation): on a slush-covered runway the nosewheel spray breaks against the lower nose skin — exactly the standby probes' neighbourhood — and an unheated probe taking spray at near-freezing temperature can be frozen shut by the time of rotation; OAT above 5 °C keeps the spray liquid.

Two more probe-family mechanics complete the picture. Some items offer a clean-cut variant: instead of placarding the heater, declare the associated ADR inoperative and dispatch under the ATA-34 item — when the heat can't be trusted, retire the whole data channel in one stroke. And the operational procedures pre-position the flight: air-data switching to the third source at cockpit preparation, and — on meeting icing conditions in flight — the affected ADR selected OFF and kept OFF for the remainder of the flight. No second chances for a probe that iced once while dispatched heaterless: the rule forbids the "it recovered, switch it back" oscillation. If, dispatched one AOA heater down, another AOA heat fault appears in flight, the procedures map the situation onto the corresponding pair alert of article 11 — the dispatched probe plus the new casualty are treated as the pair they now are.

6. Windshields and wipers — everything reads the PF's side

Window heat (one of two computers may be inoperative, C interval): no known or forecast icing; the working side's window heaters and warnings intact; and — the load-bearing lines — no CAT 2 and no CAT 3 if the inoperative side is the PF's. The operational procedure states the same thing as capability: PF's side inoperative → best landing capability CAT 1; PM's side inoperative → capability unaffected, still CAT 3 DUAL. The system authority behind both sentences is the QRH's CAT 2/3 equipment table with its "one for PF" rows (article 05).

[!warning]- Same failure, CAT 1 or CAT 3 DUAL — depending on who sits where The low-visibility visual segment belongs to the PF alone; a misted pane on the monitoring side degrades nothing the minima care about. The operational consequence: crew rostering can rescue a dispatch — which seat is PF is a hidden variable of this MEL item. Worth saying aloud in a briefing: this is an equipment placard that a seating decision can neutralise.

Side-window heat is the family's gentlest entry — typically four installed, none required, no conditions: they only demist (article 05). The PROBE/WINDOW HEAT pushbutton's AUTO function dispatches with a human replacing the logic: manually select the system ON before engine start, every sector — the crew standing in for the air/ground and engine-running automatics of articles 04/05.

Wipers are a three-step gradient — and contain the chapter's only A-interval item: the PM-side wiper may be inoperative for three sectors. Generous use (no capability lost, no weather fence) on a tight clock (three legs, then fix it) — the A category's "no standard interval" filled in by the item's own condition. One or both wipers inoperative: dispatch only if no precipitation is expected in the departure and arrival areas, and if the PF's side has lost both wiper and rain repellent, CAT 2/3 are off the table. A wiper running uncontrollably is dispatched by disconnecting it, with approach minima that don't require wipers. A failed PARK function is dispatchable if the resting blade doesn't obstruct vision — otherwise the arm comes off. (Distinguish article 05's in-flight verdict that a stuck arm imposes no speed limit: that was continuation of flight; this is next-morning paperwork.) Rain repellent closes the family at D interval, none required — with the note that the item doesn't apply where the system has been deactivated fleet-wide, a reminder that some operators have retired it entirely (article 05).

7. Water, detection, and the indication small print

Drain-mast heating (configuration-dependent): dispatch by not using the associated galleys and lavatories and shutting off the related water supply — article 07's logic inverted into a placard: no water, no ice.

The ice detection system (D interval, none required) dispatches with a one-line crew substitute: periodically check the windshield and wiper areas for ice accretion. That is article 06's "you are the ice detector now", with the lookout points officially named. The visual ice indicator itself isn't an MEL matter at all — it routes to the configuration-deviation list. And its light, failed, gets the bluntest substitute in the chapter: at night, in icing conditions — engine and wing anti-ice ON. Can't see the accretion probe? Assume it's accreting: article 10's may/must ladder collapsed to its dark-sky floor.

The panel-and-indication items run on one grammar — an indication may die if the function it reports has another window. The ENG FAULT light routes to whichever locked-valve item applies. The WING FAULT light requires the BLEED-page anti-ice arrows to work, and its procedure turns article 02's ground self-test into a lamp: at cockpit preparation, WING ANTI ICE ON, watch the arrows show amber and clear as the 30-second test closes the valves — ignoring the momentary LO PR alert while pressure is still building (the same reason the FAULT light blinks on selection). The ON light, the BLEED-page ANTI ICE label and the arrows themselves: no conditions. Wiper speeds: FAST inoperative requires SLOW to work; SLOW inoperative carries no conditions — FAST remains.

8. The two stuck-open accounts

Wing valve locked open (§3's second life) files the chapter's biggest performance return:

Why does altitude raise the fine? The stuck valve bleeds continuously, and at high level the engines already work near their ceiling — the same extraction is a larger slice of what's available (interpretation). The operational lines: perform the WING OPEN drill on the ground; with only the right side affected and APU bleed feeding the air conditioning, keep the crossbleed CLOSED — don't let APU air across to feed the stuck right valve (§3's start-window insurance, airborne edition); obey the recall above 1500 ft; a bleed failure in flight puts the affected pack off; expect the ground drill again at landing.

Engine valve locked open is the light account: +0.5 % fuel per stuck valve, driftdown ceiling −0/−500 ft on the same temperature split, performance per the operator's application — and the pushbutton-ON preparation line of §4. Cross-check with article 12's emergency-electrics row: both engine valves failed open cost about 1.5 % there; one valve at 0.5 % is the same physics on a smaller invoice. Three numbers on one string: 0.5, 1.5, 5.5 — one engine valve, both engine valves, a wing valve at cruise altitude.

9. Five questions at the aircraft

Meeting an ATA-30 placard at the door, run five questions: (1) does today's route or destination carry known or forecast icing — the fence clauses can kill the flight before anything else is read; (2) any high-elevation airfield involved — some operators' items tighten there; (3) is it an ETOPS sector — the wing item caps it, the engine item bans it, some AOA variants exclude it; (4) who will be PF — the windshield and wiper items' CAT capability turns on the answer; (5) are the performance accounts in the computation — the two stuck-open items each carry fuel and driftdown corrections that must reach the flight plan and the performance application.

And one boundary worth teaching (interpretation): the fence clause is a joint dispatcher-and-captain judgment made against the forecast. Weather that turns after takeoff doesn't retroactively void the dispatch — but flying dispatched-with-a-fence into unforecast icing means handling it by articles 11/12 and knowing the dispatch premise has broken: land, report, let the operation re-decide. An MEL is not a talisman; it is a conditional pass.

That closes the chapter: thirteen articles from unrestricted operation (article 01) to the conditional pass — one system, two heat sources, and a failure philosophy that always prefers symmetric honesty to asymmetric hope.

Self-test

[!note]- Q1. Which failures admit no dispatch at all, and what single rule generates the list?

All pitot heat; any pitot pair; any AOA pair; both windshields — paired = grounded. Two same-type unheated probes are pre-configured to ice together and identically, which is exactly the coherent-lie scenario no source-switching can defend against; two unheated windshields leave no PF side to retreat to. Single-point failures all dispatch conditionally: redundancy buys time, spent redundancy stops the aircraft.

[!note]- Q2. Why do ICE DETECTED and SEVERE ICE DETECTED not appear as dispatch items?

They are weather reports, not equipment failures. The MEL's interest is the detector hardware — a separate item with its own crew-lookout substitute.

[!note]- Q3. Wing valves locked closed versus locked open — required numbers and the shape of each condition set?

Locked closed: none required; ETOPS capped at 180 minutes, no flight into known/forecast icing (fence clause — the section is unprotected). Locked open: two required, one per side; conditions live on the ground — starts at OAT ≤ 19 °C and on APU bleed only, insuring the start-window leak onto an uncooled slat. Capability survives (the open valve protects continuously); the cost moves to the fuel account.

[!note]- Q4. State the ETOPS gradient between wing and engine valves locked closed, and its logic.

Wing: no ETOPS beyond 180 minutes. Engine: no ETOPS at all — on a single-engine driftdown into icing, the surviving engine's intake anti-ice is its bodyguard, and intake ice is a structural threat to the engine.

[!note]- Q5. The probe family's two-condition template, its strictest word, and the only B-interval item?

Working components' ADR/heaters/warnings intact (the voting floor), and no flight in visible moisture or known/forecast icing — moisture alone fences the probes, stricter than the two-ingredient icing definition. Pitot heaters carry the B interval: three days, because pitot pressure is airspeed.

[!note]- Q6. One windshield heat inoperative: the two capability sentences, and the hidden variable?

PF's side inoperative → best capability CAT 1; PM's side → unaffected, CAT 3 DUAL. The hidden variable is rostering: assign the serviceable side to the PF and the placard costs nothing the minima care about.

[!note]- Q7. After dispatching one probe heater out, icing is met in flight. What is the standing rule, and what if a second same-type heat fault then appears?

Affected ADR OFF — and it stays OFF for the remainder of the flight; no re-engaging a source that iced once while dispatched heaterless. A second fault of the same probe type is handled as the corresponding pair alert: the dispatched probe plus the new one are now the pair they resemble.

Key takeaways

References

This article is a synthesis of some operators' MEL practice for ATA-30 (the alert-to-dispatch table; the panel/indication, wing, engine-intake, probe, windshield-heat, rain-protection, waste-water and ice-detection items with their operational procedures; the repair-interval definitions of the MEL preamble), deliberately paraphrased rather than quoted: item numbering, intervals and exact conditions differ between operators, several conditions noted are operator-specific additions, and the current operator MEL is always controlling. The CAT 2/3 equipment basis is the QRH table discussed in article 05; the underlying mechanisms are those of FCOM DSC-30 and the abnormal procedures as developed across articles 01–12. The start-window insurance reading, the nosewheel-spray picture, the altitude-fine reasoning and the five-question checklist are integrative interpretations and are flagged as such in the text.

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.