Occurrence reporting deadline: from when the operator is deemed aware of the occurrence

A maintenance crew inspects a deformed wheel rim on an airliner in a dimly lit hangar at dusk.

A crew writes a line in the journey log on a Friday at 21:40: abnormal contact at the end of the taxi run, no visible damage. On Monday, the workshop finds a deformed wheel rim. The reporting file is opened on Tuesday, and the operator counts its seventy-two hours from that point. The authority, for its part, will read the timestamp in the log. The whole question lies in that gap: the deadline does not run from the moment the flight safety department learns of the occurrence, but from the moment the organisation, through any one of its own people, became aware of it.

What “becoming aware” actually covers

The European regulation on the reporting of occurrences in civil aviation sets transmission to the competent authority within seventy-two hours of becoming aware of the occurrence. The wording looks innocuous; in reality it carries the whole difficulty, because the subject of the verb is a legal entity. An operator perceives nothing by itself: it knows through the people who act on its behalf.

Three conditions are enough, and they can be checked one by one. First, a human being attached to the organisation — crew, operations officer, engineer, duty manager, instructor — must receive the information. Second, they must receive it in the exercise of their function, and not at a hotel bar counter. Finally, the information received must be sufficient to characterise a reportable occurrence: a corridor rumour triggers nothing, while reading a log entry describing a ground contact triggers everything.

Nothing in this mechanism refers to the flight safety department, to the nominated post holder or to the reporting tool. Those are processing organs, not organs of perception. The organisation knows as soon as one of its own knows, and the rest is merely the transport of information inside the house — a transport paid for in hours taken from the same deadline.

Two distinct obligations intersect here, and they are often confused. The one that bears on the individual consists in reporting to their organisation. The one that bears on the organisation consists in transmitting to the authority. The prudent reading is that they do not add up into a double deadline: as soon as the person holding the information is one whose perception is attributed to the operator, the second clock is already running while the first one runs. Counting the deadline twice amounts to betting on a favourable interpretation that no approved manual sets out.

The 72-hour deadline starts when any employee of the company learns about the issue while working, not when the safety department is notified. The company is considered aware as soon as one of its staff knows, even if that information takes days to reach the safety team.

Two possible starting points for counting the seventy-two hoursHorizontal time axis graduated at the event, at twenty-four, forty-eight and seventy-two hours. A full bar starts at the event and covers the whole deadline. A second, shorter bar starts at the opening of the file and covers only the remaining portion. Beyond seventy-two hours, a zone marks the overrun.full deadline, counted from the eventwhat is left after the relaysevent24 h48 h72 hpast deadlinefile opened
The deadline has already started before the file exists. The upper bar starts at the event, the lower bar at the opening of the file: the gap between the two is time already consumed, invisible in the safety tool but readable in the original timestamps.

An A320 crew notes a hard landing at 22:30 on a Saturday. The maintenance team finds a cracked landing gear leg during the next check on Monday at 09:00. The 72-hour deadline actually started at 22:30 Saturday, not when the maintenance report was written. If the safety report is only submitted by Tuesday at 22:30, it’s 3 hours late – because the company was aware of the issue since Saturday night.

One event, several points of entry

One and the same occurrence can enter the organisation by very different paths, and each one places the starting instant elsewhere. The useful exercise, in a manual, is not to restate the deadline — it is always there — but to state, channel by channel, which timestamp will prevail the day someone works back up the chain.

Two pilots in a cockpit reviewing a document and taking notes.
Documenting occurrences promptly ensures compliance with reporting deadlines.
What fixes the instant of awareness according to the point of entry
Channel Instant of awareness Timestamp that prevails Main risk
Crew report entered in the tool Submission of the form Server log Draft written but never submitted
Entry in the journey log or the maintenance log Writing of the entry Time written by the signatory Delayed reading by the management organisation
Call from air traffic control, the aerodrome or a third party The agent taking the call Operations day log, if one is kept No written trace of the moment
Workshop finding or scheduled check Opening of the discrepancy record Maintenance management system Treated as a mere technical discrepancy
Flight data analysis Identification by the analyst Timestamp of the readout session Readout cadence taken as the start of the deadline
Report from a passenger, an insurer, the press Receipt by the first staff member concerned Mailbox or departmental register Route not foreseen by the procedure

Two lines deserve elaboration. Flight data analysis works in batches: a weekly or monthly readout does not push the start of the deadline back to whatever date is convenient, but neither does it move it back to the date of the flight, since no one held the relevant element at that time. The start lies at the instant when the analyst identifies an exceedance that characterises a reportable occurrence. The practical consequence is clear: a slow cadence is not an infringement, but it becomes indefensible as soon as it appears to serve to delay awareness.

The third-party line is the most slippery. A contracted ground handling agent, another operator sharing an apron, a de-icing provider: their awareness is not the operator’s, but the call they make becomes so the second a staff member takes it. That is why subcontracting contracts should carry an information escalation clause, and why the day log must record incoming calls with their time — it is the only document that will later prove that the organisation knew at that moment and not before.

Single point of entry versus chain of relaysTwo routes compared on the same time scale. At the top, the information goes through a single point of entry and the remaining margin occupies almost the whole deadline. At the bottom, it crosses four successive relays and the margin shrinks to a short segment before the deadline.single timestamped point of entryalmost the entire marginchain of successive relaysresidual margincrewstationon-callsafetyevent24 h48 h72 htime elapsed since the event
Each relay is a levy on the deadline, never on the obligation. The number of intermediaries does not change the due date: it only changes the time left to work the case when the file finally reaches the right desk.

The weekend suspends nothing

The deadline is counted in calendar hours. An occurrence known on a Friday at 21:40 must be transmitted before Monday at 21:40, whatever the working days and the department’s organisation. No provision suspends the count at night, at the weekend or on public holidays, and that is precisely where files get lost.

The mechanism of the loss is almost always the same. The procedure provides that the duty manager, alerted at two in the morning, sends a message to the safety department’s mailbox. The mailbox is checked at the start of the next working day. On paper, everyone has done their job; in fact, the organisation had known since two in the morning on Saturday and the clock was running while no one was watching. When the file is opened on Monday at nine, a handful of hours remain to gather the evidence, obtain a technical opinion and write — and nothing at all remains if the occurrence fell on the Friday of a long weekend.

Crossing of the channel with the time of receiptTwo-entry table. In columns, timestamped channel and oral channel. In rows, working hours and on-call period. Three cells give a proven start or a trace to be created within the day. The cell crossing the oral channel with the on-call period is the only one where the deadline is consumed without the organisation knowing it.timestamped channeloral channelworking hourson-callproven startby the timestamptrace to createwithin the dayproven startmonitoring to ensuredeadline lostwithout knowing itonly one cell in four costs hours
The risk is not spread out: it concentrates on one cell. A call taken at night by someone on call is the only case where the time of awareness exists nowhere; it is also the only one the procedure must address by name.

The trap. An automatic acknowledgement of receipt is not a timestamp of awareness. Many organisations date the start of the deadline from the opening of the ticket in the safety tool, while the journey log, the day log, the call register or the workshop discrepancy record carry a time several days earlier. Those timestamps exist, they are retained under other obligations, and they are the ones an auditor will compare with the date of transmission.

When the reportable character only appears later

Delayed discovery is the most discussed case and the least treated in manuals. A bird strike with no apparent effect, a deceleration judged normal, a technical defect closed with no follow-up: the event is known, but nothing at that point indicates that it falls within the scope of mandatory reporting. The next check reveals a deformation, an incipient crack, wear beyond tolerance. From what date does the deadline run?

The reasoning that holds is that of characterisation. Awareness bears on a reportable occurrence, not on a raw fact: as long as no available element makes it possible to attach it to a category calling for a report, the clock cannot start. It starts at the instant when the new finding makes that attachment possible. But this reasoning only protects if it is written, dated and verifiable, and it collapses as soon as it is shown that a sufficient element existed earlier and was not exploited.

Hence the only robust practice: put both dates in the report, that of the event and that of the characterisation, and explain in one sentence what changed between the two. An organisation that writes “contact during taxi on the 12th, deformation found on the 15th at the transit check, characterised as reportable on the 15th” has nothing to fear. One that shows only the date of the 15th suggests it kept silent for three days, and will have to refute that with documents it has not retained.

The sequence that starts the clockThree blocks linked by arrows. First block, the information reaches a person acting for the operator. Second block, the element is sufficient to characterise a reportable occurrence. Third block, the clock starts at that instant. A downward branch indicates that if characterisation is not yet possible, the clock starts when it becomes so, both dates being retained.the information reachesa person actingfor the operatorthe element is enoughto characterise areportable occurrencethe clock startsat that instantyesnothe clock starts when the reportable characterbecomes identifiable, and both dates are retained
The same question arises at every channel, in the same order. It is not the event that triggers the count, it is the moment when an element held by the organisation makes it possible to qualify it — and the lower branch must be documented, not invoked after the fact.

The right reflex. Transmit incomplete rather than late. The initial report does not have to contain the analysis: it carries what is needed to identify the flight or the aircraft, the date, the place, the nature of the occurrence and the origin of the report, the detail of the expected fields being that of the authority’s form. The rest belongs to the follow-up milestones. Waiting for the manufacturer’s opinion or the workshop’s feedback before sending the first message is the most frequent cause of overrun, and the simplest to eliminate.

After the seventy-two hours, two further milestones

The initial transmission does not extinguish the obligation, it opens it. The same regulation then requires the organisation to analyse the occurrence, to draw the necessary measures from it and to report on that analysis according to two successive deadlines: preliminary results of the order of a month after notification, final results of the order of a quarter. These orders of magnitude matter less than the principle, and the exact values are to be read in the applicable text and in the manual: what must be remembered is that a notified occurrence remains open, and that a file closed without documented follow-up is just as visible as a file transmitted late.

This architecture explains why the seventy-two hours are not an investigation deadline but an alerting deadline. They serve to ensure that the authority, and through it the European collection scheme, has an exploitable signal early and can bring together occurrences coming from several operators. For that system, a late but complete report is worth less than three lines sent on time and completed afterwards.

One case escapes this rhythm entirely. Accidents and serious incidents fall under another regime, that of immediate notification to the investigation authority, without waiting for anything. The occurrence report exempts from nothing and replaces nothing; confusing the two channels loses the one thing that can never be recovered, the perishable data and the state of the scene.

The three milestones that follow the eventTimeline on a compressed scale. A first milestone marks the initial report at seventy-two hours, a second the preliminary results of the analysis of the order of a month later, a third the final results of the order of three months after the notification. The heights increase from left to right.event72 hinitial reportabout one monthpreliminary resultsabout three monthsfinal resultsnon-linear scale
The first milestone alerts, the next two investigate. Trying to fit the content of the second milestone into the deadline of the first is the confusion that produces delays; the exact due dates are to be verified in the applicable text.

What the deadline does to just culture

A short deadline exerts pressure, and that pressure moves downwards as soon as the organisation fails to absorb it. One then sees crews chased up at outstations, curt messages demanding a detailed account before the end of the day, sometimes a remark about the time taken to report. The result is mechanical: reports become shorter, more cautious, less useful, and spontaneous reporting dries up exactly where losing it costs the most.

The rule of allocation is nonetheless simple. The regulatory deadline bears on the organisation, not on the author of the report. What is expected of the crew or the engineer is a dated and honest account; what is expected of the operator is a channel that carries that account to the authority without distorting it and without devouring the deadline. Making a late report an individual matter amounts to trading one file for all the following ones.

Protection of the source, provided for by the same regulation, is moreover worth something only if the information circulates: it reserves separate treatment for wilful misconduct and for manifest disregard of an obvious risk, but in principle it forbids a report from being used to the detriment of the person who wrote it. Field feedback on these situations is worth more than an internal note, provided it stays anonymised — which is what exchanges between professionals make possible when they are held outside the company setting.

What to go and check in your own manual

The useful control point is not the sentence recalling the seventy-two hours: it appears in every manual. It is the six elements that decide whether that deadline is holdable on a Friday evening.

  • A written definition of awareness by the organisation, naming the functions whose perception binds the operator.
  • A single, timestamped point of entry, reachable from outside the internal network, and a clear instruction for everything that arrives by another route.
  • An on-call arrangement that can transmit, and not merely alert: the person on duty must hold access to the tool and the report template.
  • A count in calendar hours written in black and white, with the Friday evening case treated as an example in the procedure itself.
  • Retention of the original timestamps — journey log, day log, call register, workshop record — and not only that of the safety tool.
  • A reclassification procedure that requires both dates and the sentence explaining what changed.

There remains a grey area that no manual closes completely: information held by a contractual third party. A provider that knows and says nothing does not make the operator knowledgeable, but it does not exonerate the operator either if the contract provided for escalation and no one ever demanded or tested it. For anyone preparing to take up a flight safety post, an organisation audit or a selection interview, this is the question that separates a copied procedure from a considered one — and the one to ask first in order to understand how an organisation really handles its occurrences.

This text describes a mechanism, it replaces nothing. The applicable regulation, the decisions of the competent authority and the operator’s approved documentation prevail over this article in all circumstances, including where their wording differs from the orders of magnitude given here.

Questions fréquentes

Is the seventy-two hour deadline counted in working hours?

No, it is counted in calendar hours. An occurrence known on a Friday at 21:40 must be transmitted before Monday at 21:40, weekends and public holidays included. That is why an on-call arrangement that alerts without being able to transmit lets the deadline be consumed without anyone noticing.

From what date does the deadline run when the reportable character only appears at the next check?

The deadline runs from the instant when an element held by the organisation makes it possible to attach the fact to a category calling for a report, hence from the new finding. This reasoning only holds if it is documented: the report must carry the date of the event and that of the characterisation, with a sentence explaining what changed between the two. Without that trace, the gap looks like deliberate silence.

Does a ground handling provider's awareness bind the operator?

The provider’s awareness is not in itself the operator’s, since it is not part of the organisation. On the other hand, the call it makes brings the occurrence to the operator’s knowledge as soon as one of its staff takes it, and the subcontracting contract should impose that escalation. A clause never tested in operation protects no one on the day the discrepancy comes to light.

Should one wait for the technical analysis before sending the initial report?

No, and this is the leading cause of overruns. The initial report carries what is needed to identify the flight or the aircraft, the date, the place, the nature of the occurrence and the origin of the report; the analysis belongs to the follow-up milestones, whose deadlines are counted rather in weeks and months. Transmitting partial content on time is worth more than a complete file out of time.

Does an occurrence detected by flight data analysis start the deadline at the date of the flight?

No: on the date of the flight, no one in the organisation held the element that made it possible to qualify the occurrence. The deadline runs from the moment when the analyst identifies an exceedance that characterises a reportable occurrence. A slow readout cadence is not an infringement in itself, but it becomes hard to defend if it appears to serve to delay awareness.

Can the author of a report be sanctioned for having reported late?

The deadline for transmission to the authority bears on the organisation, not on the author of the report, and the regulation in principle forbids a report from being used to the detriment of the person who wrote it, save for wilful misconduct and manifest disregard of an obvious risk. Sanctioning an individual delay amounts to obtaining one file at the price of all subsequent reports. Handling a recurring delay is a matter of training and of the design of the channel, not of discipline.

Commentaires

Leave a Reply

Découvrez l’écosystème FL330

FL330, c’est toute l’aviation professionnelle — explorez nos autres espaces.