On a peak Friday at a major hub, a twin-engine jet spends forty minutes between pushback and takeoff, engines at idle in the queue. Its emissions sheet, however, was established on a cycle that allocates 26 minutes of taxiing at 7% thrust—a convention set over forty years ago, never designed to describe a day of operations. What the reference cycle measures, how it measures it, and why the gap with the real ground is a misunderstanding rather than a scandal: that’s what needs to be understood before making certified figures say what they never measured.
What the test bench actually measures
Emissions certification applies to an engine, not a flight. The test engine conforms to the production definition, new or nearly so, without bleed air extraction, without degradation, with fuel of specified properties. Emission indices—grams of pollutant per kilogram of fuel burned—are measured for nitrogen oxides, carbon monoxide, unburned hydrocarbons, smoke, and, under the most recent standards, non-volatile particles, at four conventional thrust settings.
Raw measurements are then corrected to a reference day: standard atmosphere, reference humidity for NOx, whose formation chemistry depends directly on the water content of the intake air. This isn’t an engineer’s refinement: without these corrections, the same engine measured on a dry morning and a humid afternoon would produce two different sheets. One simple thing to remember: the certified figure is, by design, a figure that exists in no real atmosphere. It has never claimed otherwise.
The regulatory value isn’t the index itself. It’s a mass per cycle relative to thrust: the total mass of each pollutant emitted over the full reference cycle, divided by the nominal certification thrust. This quantity is compared to a limit line that depends on the engine’s overall compression ratio—higher pressure at the combustor means higher flame temperature, harder to control NOx, and the standard accounts for it. The requirement thresholds tighten with each working cycle of the ICAO committee responsible for environmental protection: that’s where the « CAEP » designations followed by a number come from, found in the sheets and databases.
The certification process measures emissions under very specific, controlled conditions that don’t match real-world operations. It’s designed to compare engines fairly, not to predict real-world pollution. The numbers are useful for comparisons, but they don’t tell you exactly what an airplane emits during a flight.
An Airbus A320neo with CFM LEAP-1A engines is certified with a certain NOx emission index at full thrust. In reality, most takeoffs are at reduced thrust, which means the actual NOx emissions are lower than the certified value. Meanwhile, taxiing times at busy airports often exceed the 26-minute standard, leading to higher CO and unburned hydrocarbon emissions than the certification predicts.
Thirty-three minutes, including twenty-six at idle
The landing-takeoff cycle—the LTO cycle—models only what happens below approximately 900 meters above ground, the conventional mixing height below which what the engine emits affects local air quality. Four modes, four thrust-duration pairs, all fixed once and for all. Takeoff at full thrust for 0.7 minutes, climb at 85% of nominal thrust for 2.2 minutes, approach at 30% for 4 minutes, taxi and ground idle at 7% for 26 minutes.
The total is 32.9 minutes, including 26 at idle: the certification cycle is, four-fifths of the time, a taxiing cycle. It wasn’t a mistake. When the standard was written over forty years ago, taxiing times at major airports justified an envelope value, and a single convention was needed to compare two engines on identical bases. The key word is: the cycle is a comparison tool between engines, not a flight reconstruction. Everything that follows stems from this distinction.
Two immediate consequences. First: the certified mass of CO and unburned hydrocarbons depends overwhelmingly on the 7% point alone, since it occupies most of the duration—the choice of this conventional regime weighs more in the result than the quality of combustion at full thrust. Second: everything that happens above the mixing height is out of scope. The climb to cruise level, cruise, non-CO2 altitude effects are not part of this certification, which is a local air quality tool and nothing else. CO2, meanwhile, is subject to separate certification, established at the full aircraft level on a cruise metric—precisely because the LTO cycle can’t provide useful information about it.
Real taxiing doesn’t match the convention
On the ground, the convention diverges at both ends. Duration, first. Real taxiing time depends on the platform’s geometry, the runway in use, the parking stand, and the time of day: a regional airport wraps it up in a few minutes all day, while a very large hub sees its peak departures exceed the cycle’s envelope value—queues, crossings, and sequencing included. The same aircraft, the same day, may taxi several times longer in the evening than in the morning.

Thrust setting, next. The ground idle of a modern turbofan typically sets below the conventional 7% of nominal thrust. Yet it’s precisely at very low thrust that combustion degrades: low combustor temperature, poorly distributed local richness, quenching of reactions near walls—the CO and unburned hydrocarbons rise quickly as thrust decreases. The certified index at idle thus underestimates the real idle index, while the real fuel flow, for its part, is lower than that of the test point. Two opposing discrepancies that don’t compensate reliably: the net result depends on the engine, the duration, and what is being calculated.
Add everything the convention ignores: taxiing on one engine, where the live engine runs higher than idle while the other is shut down and the APU sometimes runs; successive power-ups in a queue, where the thrust setting oscillates above idle with each taxi restart; the startup itself; engine running deicing. The APU, in particular, is not covered by any comparable emissions standard—its operating hours at the gate are entirely outside the cycle’s framework.
| Mode | Certified cycle | Exploitation courante | Direction of the deviation |
|---|---|---|---|
| Take-off | 0.7 min at 100% | Reduced thrust on most departures | Actual NOx for this mode well below the certified point |
| Initial climb | 2.2 min at 85% | Reduced climb thrust, variable profiles | Moderate deviation, in the same direction as take-off |
| Approche | 4.0 min at 30% | Variable power settings, continuous descents | Rough representation, of limited magnitude |
| Taxi and idle | 26 min at 7% | Duration highly variable by airport and time of day; actual idle below 7%; single-engine taxi | Higher actual CO/HC index, lower flow — deviations not offsetting; duration over- or underestimated depending on the airport |
The trap: crossing a real column and a certified column. The classic error in « in-house » balances is to multiply the platform’s real taxiing time by the emission index and fuel flow of the certified 7% point. The resulting figure looks precise and is wrong twice: the real idle flow is lower than the test point’s, the real CO/HC index is higher. Depending on the engine and duration, the two errors add up instead of canceling out—and the balance shifts from an excess to a deficit without any warning.
Each pollutant lives at one end of the cycle
Each pollutant behaves differently across the thrust range. Nitrogen oxides are a product of temperature: their main formation pathway, called thermal, runs away exponentially with flame temperature, and thus with pressure and thrust. CO and unburned hydrocarbons are the opposite: markers of incomplete combustion, maximal at idle, negligible at full thrust. Non-volatile particles—soot—depend more on the combustor architecture: classic rich primary zone designs produce them mostly at high thrust, lean-burn designs shift the trade-off.
This pollutant geography explains a point the certified sheet doesn’t show: most real takeoffs are at reduced thrust, when runway length and limitations allow. The combustor temperature then drops moderately—but since the thermal formation of NOx is exponential, the drop in NOx emitted during the takeoff mode is, itself, very marked. The certified point at full nominal thrust describes a case that operations avoid as much as possible.
Remain the particles measured around airports and not found in any sheet. Certification counts non-volatile particles at the nozzle exit plane. In the plume, a few dozen meters further on, a second population forms: volatile particles, condensed from the fuel’s sulfur compounds and unburned organics, often ultrafine, very numerous. Measurement campaigns on the platform periphery count them; the certified database, by design, does not contain them. Anyone comparing ambient counts to certified values compares two quantities that do not measure the same thing.
The right reflex: pair each term, mode by mode. Before any calculation, set up four columns—duration, thrust setting, fuel flow, emission index—and forbid mixing an observed column and a certified column in the same product. This is the logic of the advanced methods in the airport air quality manual: real times, real flows when they exist, indices adjusted for consistency. If a column is missing, revert to the full convention rather than mixing.
What the certified figures are actually used for
The certified figures serve three key purposes. The first is comparison: same test bench, same reference day, same durations for all engines in the world—the certified sheet is the only place where a recently designed combustor and its predecessor are measured on strictly equal terms. It’s the tool for fleet choices: between two engine variants of the same aircraft, the sheet difference is solid information, because the convention’s biases apply to both.
The second is the fee. Several major European platforms modulate their fees according to the certified NOx values of the exact engine, grouped by emission classes. The basis is the sheet, not the flight: taxiing less doesn’t change the billed line, choosing a better-rated engine variant does. It’s a signal aimed at fleet decisions, and it’s intentional.
The third is the inventory. Air quality studies of a platform start with certified indices—there’s no better public basis—but serious methods replace conventional durations with observed times mode by mode, and adjust thrust settings and flows when operational data exists. The convention provides the raw material; the inventory provides realism.
What you should verify on your side
Three verifications are worth it. Operations: what taxiing time assumption feeds your fuel calculations and ground consumption statistics, and when was it last compared to observed times on your main platforms. Fees: on which platforms your bills contain an emissions line, and in which class your exact engine falls—two variants of the same engine, on the same aircraft type, can change class. Communication: if your operator publishes a local emissions balance, what method underpins it—raw certified values with cycle durations, or real times substituted.
The gray area, however, doesn’t close: volatile particles absent from the sheets, APU outside the framework, non-CO2 altitude effects outside the scope of local air quality. For those preparing a qualification or technical interview—the subject comes up regularly in qualification preparations—knowing how to trace the exact boundary of what certification covers, and calmly stating what it doesn’t cover, is better than reciting the four regimes of the cycle.
A final obvious point: the mechanisms described above describe general mechanisms. The applicable texts, the certification data sheet of the engine actually installed, and the operator’s approved documentation take precedence over this article for any operational or compliance decision.

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