A minimum equipment list entry is usually read as permission, and it is. What it rarely gets read as is an invoice. The technical half of a deferral is often the quick half: a circuit breaker collared, a placard fitted, a line written in the technical log, ten minutes of work by someone who will not fly the aeroplane. The operational half arrives later and lands on flight preparation, where it can cap the cruising level, add fuel, delete an approach, remove a destination or take the aircraft out of its extended-range approval for the whole repair interval. This is how a certified tolerance becomes a planning constraint, and why the small letter beside the item deserves more attention than the item itself.
Where the tolerance comes from, and why yours is narrower than the manufacturer’s
Nothing in a minimum equipment list is invented by the operator. The chain starts with the type certificate holder, which produces a master list for the type: a statement, backed by the certification analysis, of which items may be inoperative and under what conditions while the aeroplane still meets an acceptable level of safety. That document is evaluated and approved by the certifying authority through its operational evaluation process — in Europe it forms part of the operational suitability data for the type; in the American system it comes out of a flight operations evaluation board. Either way, it is a certification product, not an operational one.
The operator then derives its own list from it. The rule that governs the derivation is simple and often forgotten: the operator’s list may be more restrictive than the master, never less. In practice it is almost always more restrictive, for reasons that have nothing to do with caution in the abstract. It is tailored to the actual configuration of the fleet — an option not fitted cannot be deferred; to the kinds of operation approved — reduced vertical separation, performance-based navigation, low visibility operations, extended range; to the area of operations; and to what the operator’s own maintenance organisation can realistically support at each station it serves.
Two consequences follow on the line. First, two operators flying the same variant can hold genuinely different answers for the same failed item, and both are correct: a tolerance remembered from a previous employer is not evidence. Second, the list is closed. If an inoperative item is not in it, the aeroplane is not dispatchable with that failure — and the flight manual’s configuration deviation list does not fill the gap, because it covers missing secondary airframe parts rather than unserviceable equipment. Absence is a refusal, not a silence — and reading silence as latitude is the commonest misuse of the document.
The MEL is a safety document, not a flexibility tool. Airlines can only defer faults listed in their MEL, and their rules are stricter than the manufacturer’s. If it’s not in the list, you can’t fly with that fault.
Four columns and two letters
The layout is standardised enough to be read at speed, which is exactly why it gets read too fast. Each element does a distinct job, and the cost falls unevenly across them.
| Element | What it determines | Who pays for it |
|---|---|---|
| Item identification and system reference | Which failure is covered, at what level of granularity — one channel, one unit, or the whole function | Nobody, until the wrong item is matched to the observed symptom |
| Repair interval category | How long the aircraft may keep flying in that condition | Maintenance planning and the network |
| Number installed / number required for dispatch | Whether redundancy remains, and how much | Engineering, at the moment of the decision |
| Remarks or exceptions | The conditions: weather, airspace, aerodrome, kind of operation, combination limits | Flight preparation, on every subsequent sector |
| (M) maintenance procedure required | A specific action to be carried out before or during the deferral | The maintenance organisation, once |
| (O) operational procedure required | A specific action or limitation imposed on the operating crew and on planning | The crew and the flight plan, repeatedly, for the whole interval |
The asymmetry in the last column is where the money is. An (M) is a bounded job with a defined end. An (O) is a recurring obligation that reappears on every sector until the item is repaired — and it is almost never written out inside the entry itself. The entry says an operational procedure applies; the procedure lives in a separate supplement to the operations manual. Finding it is part of accepting the aircraft, and a crew that has read the entry without the procedure has read half the deferral.
The trap. An entry that grants dispatch with a condition is frequently summarised on the technical log line as the item name and the expiry date alone. The condition — level restriction, weather restriction, additional fuel, prohibited approach — does not travel with that line. If the only thing handed over at crew change is the log page, the constraint has been lost while the permission survives.
The repair interval is a calendar, not an intention
Repair intervals are expressed as categories, and the categories are widely harmonised between the major systems even where the approval routes differ.
| Category | Interval | Practical behaviour |
|---|---|---|
| A | As specified in the remarks column of the entry | May be expressed in flight cycles, flight hours or a number of days; read it every time, because no default exists and extensions are normally not available |
| B | Three consecutive calendar days | Effectively forces the aircraft towards a station with capability within a very short planning horizon |
| C | Ten consecutive calendar days | The common case; long enough to plan a repair, short enough that the operational penalty is paid many times over |
| D | One hundred and twenty consecutive calendar days | Reserved for the items whose loss has the least operational effect; over that length of time the risk is simply that the deferral becomes invisible |
The counting convention matters more than it looks. The day the fault is recorded does not count; the interval starts at the end of that day. That single convention is behind a recurring disagreement between operations control and maintenance control about when an aircraft actually becomes undispatchable, and it is worth confirming against the preamble to the operator’s own list rather than against memory.
Extensions exist. Many authorities allow a single extension, of a duration equal to the original interval, for categories B, C and D, under a procedure the operator has written down and had accepted: who may authorise it, on what justification, with what record. Category A is normally excluded. An extension does not extend the operational condition’s cost — it doubles it.
What an operational procedure actually does to a flight plan
Each operational condition is the visible end of a certification argument about what has been lost. They fall into a small number of families, and recognising the family tells you where the cost will appear.
Performance penalties change the numbers used for take-off and landing: a factored landing distance, a reduced take-off weight, a restriction on contaminated or short runways. They apply on every sector of the interval, and they must be re-applied whenever the runway, the weight or the configuration changes — including at the last minute.
Envelope limitations cap altitude or speed. A cap on cruising level is the quiet expensive one: it raises burn, which reduces payload, and it may push the aeroplane out of the airspace it was planned into.
Crew procedure changes add monitoring, cross-checking or a manual task a system used to perform. The cost is workload, and it lands where workload is already highest.
Operational restrictions remove conditions rather than capability: no flight into known icing, no night operations, no operation into a class of aerodrome. They do not reduce performance at all — they delete options from the plan.
Capability losses are the family that generates most of the disruption, and they deserve their own treatment.
Capability loss is the expensive part
An approval to enter a particular airspace or fly a particular procedure is not granted to a type in the abstract. It is granted on the basis that a defined set of equipment is serviceable, with defined redundancy. Defer a piece of that set and the approval lapses for that airframe until repair, whatever the aeroplane’s general health.
| Capability | Why equipment loss removes it | Where the cost lands |
|---|---|---|
| Reduced vertical separation | Depends on independent altimetry, altitude alerting, automatic altitude control and altitude reporting; the approval assumes a specific redundancy | Level capped below the relevant airspace: longer routing, higher burn, reduced payload, and a flight plan the system may reject |
| Performance-based navigation and required navigation performance approaches | The navigation specification names the sensors and the integrity monitoring that must be available | An approach removed from the plan, a different arrival, and sometimes a different alternate with different fuel |
| Low visibility operations | Autoflight, guidance and monitoring functions are part of the approval, often with dual-channel requirements | Higher minima for the whole interval, more alternates carried, more diversions in a bad season |
| Extended range operations | The approval rests on specific systems and on time-limited capacities; some items carry a separate extended-range restriction inside the entry | The threshold time is lost: reroute, tech stop, or the sector does not operate |
| Datalink-dependent airspace | Communication and surveillance performance is a condition of entry to certain oceanic and remote structures | Reroute onto a lower-priority track or out of the organised structure entirely |
| Surveillance and collision avoidance carriage | Carriage obligations attach in some cases to the aeroplane itself and in others to the airspace entered; either way they are conditions, not good practice | Airspace or routes closed to the airframe, with a consequence that planning must find before departure |
Read that table against the earlier one and the shape of the problem is clear. The maintenance action was finite. The capability loss recurs on every rotation, propagates into fuel, payload, alternates and route, and is paid by a different department from the one that made the decision.
The reflex. When a deferral appears, do not ask first how long it lasts. Ask which approvals the aircraft still holds with that item inoperative — vertical separation, navigation specification, low visibility, extended range, datalink — and check the planned route against the answer before anything else. The interval is a maintenance question; the approvals are the flight’s question.
A deferral can make an aircraft lose important approvals. This means longer flights, more fuel, or even cancelling the flight. Always check what capabilities are lost before planning.
An A320 loses one of its two independent altitude measurement systems. This removes its RVSM approval, since RVSM requires two of them. The aircraft can no longer fly between FL290 and FL410 in RVSM airspace, forcing a lower cruise and burning several hundred kilograms more fuel per flight. The airline must also plan for the extra fuel, at a cost that adds up every day until the fault is fixed.
The combinations nobody wrote down
Most entries are constructed on the assumption that they are the only deferral in play, or on the assumption of a stated combination that the analysis actually covered. Real aeroplanes accumulate items. Two deferrals that are individually acceptable can, together, remove the last independent path through a function, or stack two performance penalties that were each derived on the assumption that the other did not exist.
Lists generally carry a statement requiring that multiple deferrals be assessed for interaction, and that statement is doing a great deal of work: it transfers a piece of the certification analysis to the operator, and in practice to whoever is on shift. The assessment is not something a flight crew can improvise at the aircraft — it belongs to engineering, in advance, with the technical documentation available. What the crew can do is notice that more than one item is open, and ask whether the combination has been looked at, by name.
The moment the tolerance stops applying
The list governs dispatch. It answers whether a flight may be commenced with a known unserviceability, and stops answering once the aeroplane starts moving for the purpose of taking off. A failure after that point is not a deferral question: it is handled by the abnormal procedures, the flight manual limitations and the commander’s judgment. The list may still inform what has been lost, but it grants nothing at that stage.
The reverse boundary is worth stating just as plainly. The list is a maximum, not an instruction: it says what may be accepted, never what must be. A commander who judges a permitted deferral unwise for that particular flight — that runway, that weather, that route, that combination — is exercising authority the regulation explicitly reserves, and no well-written procedure treats the refusal as an anomaly to be explained away. It should produce a report, so that a tolerance which is legal and repeatedly refused becomes visible as data rather than as friction between two crew rooms.
How a deferral should reach the people who fly it
Everything above fails quietly if the transmission fails. The chain has four links, each with a characteristic failure.

- The technical log entry — the item reference, the category and the expiry. Failure: the condition is not carried on the line, only the permission.
- The placard — visible at the affected control or indicator. Failure: read as decoration, because placards outlive the deferrals that produced them if removal is not controlled.
- The operational flight plan and the performance figures — the penalty applied, in the numbers, not in a remark. Failure: applied at planning and not re-applied when the runway, the weight or the level changes on the day.
- The briefing — the condition stated out loud, with the specific consequence for this sector. Failure: stated as an item name rather than as a consequence, so the crew know what is broken and not what it removes.
A useful discipline is to brief the deferral as a sentence about the flight rather than about the aircraft. Not « the item is deferred under category C » but « we are capped, we carry the extra fuel, and this approach is not available to us today ». The first describes a document; the second describes the flight, and it is the only version a crew can act on. The interaction between deferrals, planning and the fuel that pays for them is worked through in the operations material at FL330 training.
A deferral has two halves and they are not the same size. The technical half is settled once, in a bounded piece of work with a date on it. The operational half is settled again on every sector, by different people, in fuel, payload, routing and approvals — and it decides whether the flight you planned is still the flight you can fly.

Leave a Reply
You must be logged in to post a comment.