Every flight leaves the ground with two descriptions of itself. One is the operational flight plan, the operator’s document, which computes the fuel and establishes that the flight is legal. The other is the flight plan filed with air traffic services, which tells the system how the flight will be handled. They are usually produced by the same software minutes apart, and they still disagree. The disagreements are not errors: they are the residue of the filters the filing has to pass. A few of them are paid for in fuel from the top of climb onwards, on a flight whose fuel was computed for a profile it was never to be given.
Two documents that answer two different questions
The operational flight plan answers an internal question: given this aircraft, this payload, this forecast and this route, how much fuel is required, is the flight within its performance and mass limits, and does it comply with the operator’s fuel policy. It is a computation, the crew’s fuel reference in flight, and a record kept afterwards.
The filed flight plan answers an external question: what the air traffic system needs in order to provide service, separate this aircraft and manage the flow. It is a message in a fixed format, validated by a system indifferent to the operator’s economics, and it is the basis on which the flight is allocated a route, a level and, where flow management applies, a departure time.
They therefore contain different things. The operational plan carries a fuel breakdown, a wind and temperature profile, waypoint estimates with fuel remaining, alternate computations and often a step climb schedule. The filed plan carries a route string, one requested level, one speed, elapsed times, destination alternates and coded indicators. Only a subset of each appears in the other, and that subset is where the trouble lives.
| Operational flight plan | Filed flight plan | |
|---|---|---|
| Question it answers | Is the flight fuelled, legal and within limits | How will the air traffic system handle this flight |
| Audience | Crew, dispatch, the authority on audit | Air traffic services and flow management |
| Fixes | Fuel figures, mass and balance assumptions, alternates, profile | Route, requested level, speed, times, equipment and capability |
| Validated against | Operator fuel policy, performance data, aircraft limitations | Route availability rules, airspace restrictions, format validation |
| Effect of an error | The fuel figure is wrong and nobody is told | The plan is rejected, amended or suspended, and somebody is told |
| Changes after filing | Only if dispatch regenerates it | Routinely, by amendment message or network rerouting |
Why they diverge when nobody has made a mistake
The filing process does not go straight into the air traffic system. In a dense network it passes validation against published route availability rules, airspace closures, capability requirements and format. The cheapest route the optimiser found may not be filable, so the filing engine substitutes one that is. The requested level may be constrained by a flow measure capping levels through a congested sector. And the system may return a rerouting proposal hours after acceptance.
Each of those is a legitimate action by a system doing its job. What makes them expensive is that they change the filed plan and do not, on their own, change the fuel. The operational flight plan sitting in the briefing pack was computed against the route and the level the optimiser wanted. Unless the change triggers a regeneration, the crew departs with fuel figures for one profile and a clearance for another.
The flight plan you file isn’t always the one you get. Air traffic systems make changes to routes and levels for safety and efficiency, but these changes don’t automatically update the fuel calculations. So you might be flying with fuel numbers based on a different plan than what you’re actually flying.
An A320 flying from Paris to Barcelona is filed for FL330 but gets a clearance for FL310 due to traffic congestion. The fuel was calculated for FL330, which is more efficient. At takeoff mass, the aircraft burns 200kg more fuel per hour at FL310 than it would have at FL330. Over a 2-hour flight, that’s 400kg of extra fuel burned – money wasted because the crew didn’t know to request the higher level.
The cruise level: the divergence that starts costing at top of climb
This is the costliest structural difference and the least visible. The operational plan is computed on a profile that starts at or near the optimum level for the departure mass and steps up as fuel burns off. The filed plan carries a single requested cruising level for the route, with level changes expressible at points but frequently not used. If the level actually cleared is below the level the fuel was computed at, the penalty is incurred immediately and continuously.
The physics is worth stating, because it explains why the penalty never improves on its own. Specific range, the distance flown per unit of fuel, peaks near an optimum altitude that falls as mass rises. At a heavy departure mass the optimum is relatively low, so the gap between it and a capped level is at its narrowest; but fuel flow is at its highest at that moment, so even a modest percentage loss is already a real figure in kilogrammes per hour. As fuel burns off the optimum climbs and the cap does not, so the gap widens with every hour flown. Taken at either end, a cap that persists costs more than the crew expects, and nothing about it gets better with time.
The correct response is not resignation. A constraint from a flow measure often disappears once the aircraft is clear of the constrained sector, and a level request made early in cruise, before the sector fills, is far likelier to succeed than the same request an hour later. The crew that knows which level the fuel was computed at can ask for it; the crew that does not cannot.
A cap absorbed by contingency is a cap that nobody investigates. A level penalty of a few hundred kilogrammes disappears silently into contingency fuel, so the flight lands without event and the route is never flagged. Repeated on the same pair, week after week, it becomes a systematic bias in the operator’s own consumption record, which contaminates any statistical scheme built on it. The fix is a report, not extra fuel.
Speed: the number the controller holds is not the number you fly
The filed plan carries a single cruising speed, expressed as a Mach number or a true airspeed. The operational plan is computed at whatever speed the cost index produces, which varies with mass and with wind and changes across the flight. The two are rarely the same, and the difference costs something other than fuel.
The filed speed is what the air traffic system uses to build its estimates over successive points and boundaries. Those estimates drive coordination between units and feed the calculations that produce departure times. An aircraft consistently slower than filed arrives later than the system expects; a faster one arrives early into a sector that planned otherwise. Neither is unsafe, and both degrade the data the network runs on.
It becomes a fuel matter at second order: a network holding poor estimates produces cruder flow measures, which produce more level caps and more rerouting. An operator filing a fixed nominal speed while its crews fly a low cost index is contributing to a problem it then pays for.
Route: the rerouting that changes the filing and not the fuel
Rerouting after filing is routine: a network proposal accepted by the operator, a restriction published after filing, or a tactical amendment. In each case the filed route changes. Whether the operational flight plan changes depends on whether the operator’s process regenerates it, and that regeneration is often not automatic.
The failure mode worth naming is a briefing pack whose fuel figures and waypoint estimates belong to a route the flight will not fly. A reroute that lengthens the track meaningfully normally triggers a new plan. It is the small ones that survive, a shortened departure routing here, a different arrival transition there, and they show up as a systematic mismatch between the plan’s waypoint fuel and the fuel actually remaining, which is exactly the comparison in-flight fuel management depends on. A crew below plan at every point cannot tell a burn problem from a route mismatch.
Alternates: the ones air traffic knows and the ones only you do
The filed plan carries destination alternates in its designated field and can carry en-route alternates among the coded indicators. The operational plan carries far more: the take-off alternate when the departure aerodrome is below landing minima, the en-route alternates underpinning an extended range operation, the fuel to reach each of them, and the driftdown or depressurisation case that justified the routing.

Two practical points follow. The take-off alternate has a place in the filing only among the optional coded indicators and is frequently not filed at all, so an aircraft returning shortly after departure may be heading somewhere the air traffic system has no advance knowledge of. And on an extended range operation the en-route alternates are the basis of the routing itself: one that has gone below minima or closed has removed part of the justification for the plan while the aircraft is airborne, and nothing in the filing will say so.
| Element | Why they differ | When the penalty appears | What to do before pushback |
|---|---|---|---|
| Cruising level | Optimiser profile with steps against a single filed level, then flow level capping | Immediately, at the heaviest mass, and it grows if the cap persists | Compare the cleared level with the plan level; request the plan level early in cruise |
| Cruising speed | Cost index speed against a fixed filed value | Indirectly, through degraded network estimates and the measures they produce | Know what was filed; report a persistent mismatch to dispatch |
| Route | Availability rules, validation, network rerouting after filing | Throughout, and it corrupts waypoint fuel comparison | Compare the cleared route with the plan route point by point; demand a regeneration if it differs |
| Alternates | Take-off and en-route alternates optional in the filing and often omitted | Only when needed, and then completely | Confirm the take-off alternate is usable; confirm en-route alternates for extended range |
| Off-block time | Wind and temperature computed for a departure time that has moved | In cruise, as a wind forecast that no longer verifies | Check the plan validity; ask for a new plan if the delay is significant |
| Zero fuel mass | Estimated payload in the plan against the actual load sheet | At top of climb, as a level that cannot be reached | Compare load sheet mass with plan mass before accepting the figures |
| Equipment and navigation capability | Coded capability in the filing against what the aircraft can actually do today | On the day a deferred defect removes a capability that was filed | Check the technical log against what has been filed for the flight |
Time: the slot, the amendment and the validity of the wind
The off-block time in the filing is a planning input for the whole network, not a statement of intent. Where flow management applies it is the basis on which a calculated take-off time is issued, and that time carries a tolerance window that is short and asymmetric. Missing it returns the flight to the allocation process, which can produce a worse time than the one just missed.
The rule that catches operators is the amendment obligation. If the off-block time will be exceeded by more than a set margin, the filed plan must be amended by message. That margin is an exact published figure, tighter in a flow-managed network than the ICAO baseline, so the value to know is the one published for the airspace you depart from. The usual consequence of failing to amend is suspension of the plan, and a flight that discovers this at the holding point has lost far more time than the amendment would have cost.
There is a quieter cost. The plan’s wind and temperature came from a forecast valid for a window, and a departure slipping well outside it is flying against weather that has been superseded. In stable conditions this is trivial; across a shifting jet or on a long oceanic sector it is not, and the answer to a long delay is a new plan rather than an old plan and optimism.
Weight: the load sheet that removes the level the plan assumed
The plan is computed on an estimated zero fuel mass; the load sheet supplies the actual one. When the actual mass is materially higher, two things follow. The fuel figure is understated, because the plan burned a lighter aeroplane around the same route. And the initial cruise level may simply not be achievable, because the maximum level at a given mass and temperature is a performance limit, not a preference.
That second effect is the surprise crews meet at top of climb. The plan says one level, the aircraft will not hold it, and the flight levels off lower with a fuel figure computed for the higher one. It is a level cap like any other, except that no controller imposed it and no report will be filed. Comparing load sheet mass against plan mass takes seconds and is the only place this is catchable on the ground.
Cross-check three documents, not two. The plan, the clearance and the flight management system route must agree on route, initial level and mass. Most crews compare the plan with the system and treat the clearance as an afterthought, which is backwards: the clearance is the only one of the three that reflects what air traffic has actually decided.
The cross-check that catches most of it
- Compare the cleared route against the plan route to the first significant point and at every airway change
- Compare the cleared initial level against the plan level
- Compare the load sheet zero fuel mass against the plan mass
- Check the plan’s validity time against the actual off-block time
- Confirm the take-off alternate, and the en-route alternates if the operation depends on them
Where a difference is found the decision is binary: either it is small enough to be absorbed knowingly, with the crew aware of which way it pushes the fuel, or a regenerated plan is requested. What should not happen is the common third option, where the difference is noticed, filed mentally as probably fine, and never quantified.
The honest summary is that the two documents were never meant to match. One is an economic computation and the other is a service request, and small differences between them are normal. The discipline is not to eliminate the divergence but to know which ones cost fuel from top of climb, to catch those before the door closes, and to report the ones that keep recurring on the same pair, because those are the ones quietly rewriting the operator’s own consumption record. Flight planning and network topics are discussed further in the operations community at the FL330 forums.

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