How private jet CO₂ is calculated
If you've ever read a headline saying a private jet emitted "X tonnes of CO₂ on a single flight" and wondered where that number came from, this article is for you.
There is no standardised, regulated way to calculate emissions from a private jet flight. There's no fuel meter feeding a public database. Operators don't file CO₂ figures.
So every number you see – from a tracker, an NGO, a journalist or an offset calculator – is a model. The question is what kind of model, what assumptions sit behind it, and what's been left in or stripped out.
This guide explains how we do it at The Aircraft Ledger: the science, the trade-offs and the limitations. It also traces a single Gulfstream G650 flight from raw flight tracking data to a final CO₂ figure, so you can see exactly where every digit comes from.
If you're a journalist citing our numbers, an ESG analyst using our data, or a researcher trying to understand how private aviation emissions are quantified at all, this is the foundation document.
Why this matters
Private aviation emissions are among the most-cited and least-understood numbers in the climate conversation.
A Gulfstream G650 flight from London to New York might be reported as 25 tonnes of CO₂ in one outlet and 60 tonnes in another. Both can be technically defensible. They're just measuring different things.
- Some calculators include the fuel's entire lifecycle, from extraction at the wellhead to combustion in the engine.
- Others add radiative forcing multipliers to account for contrails.
- Some apply load factors to spread emissions across an assumed number of passengers.
- Some use scheduled airline-style block times rather than actual flight duration.
None of these approaches is wrong, but they aren't comparable. And when a journalist puts a figure next to a name, the methodology behind it has to hold up.
Our position is simple. We publish a single, narrow, defensible number – tank-to-wake mission CO₂ – calculated using the EMEP/EEA Air Pollutant Emission Inventory Guidebook 2023, published by the European Environment Agency, whose aviation methodology draws on EUROCONTROL data.
- We do not include lifecycle emissions.
- We do not apply contrail multipliers.
- We do not assume passenger loads.
- We do not include reserve fuel, APU burn or ground power.
What we publish is the carbon dioxide that came out of the engines on the flight itself, from taxi-out to taxi-in.
The four inputs we need for any flight
Every emissions figure on The Aircraft Ledger is built from four data inputs. Get any one wrong and the final number drifts.
- Aircraft identity. We need to know exactly which aircraft flew. Not "a Gulfstream" but the specific tail number, which tells us the airframe, engine type and engine count. A G650 with Rolls-Royce BR725 engines burns fuel at a different rate to a G550 with BR710s. We track 22,000+ private jets worldwide by registration, each mapped to a manufacturer-specified fuel burn profile.
- Origin and destination airports. Our flight tracking data gives us the flight path, and we anchor each flight to its origin and destination ICAO codes. Those resolve to coordinates, which give us the great-circle distance used to identify and sanity-check the flight.
- Flight duration. This comes from our flight tracking data: the time between the first and last airborne position we receive for the aircraft on that flight. Duration is what drives cruise emissions.
- Engine and aircraft specifications. For each type we hold engine count, LTO-cycle fuel flow per mode (idle/taxi, approach, climb, takeoff) and a manufacturer-derived cruise fuel burn rate.
That's it. Four inputs. Everything else in the calculation is derived.
Our flight history runs from April 2023 onwards, worldwide.
The two phases of a flight: LTO and cruise
The EMEP/EEA methodology splits every flight into two emissions phases, because engines burn fuel very differently depending on what they're doing.
- Taxi-out
- Takeoff
- Climb to 3,000 ft
- Climb, cruise and descent
- Approach
- Taxi-in
Phase 1: the landing and takeoff cycle (LTO)
The LTO cycle covers everything below 3,000 feet: taxi-out, takeoff roll, climb to 3,000 ft, approach from 3,000 ft and taxi-in. ICAO standardises the time spent in each of four modes:
| Mode | Time in mode | Engine power | | --- | --- | --- | | Taxi/idle (out and in combined) | 26 minutes | 7% | | Takeoff | 0.7 minutes | 100% | | Climb out | 2.2 minutes | 85% | | Approach | 4.0 minutes | 30% |
Engines are tested and certified at these power settings, and each mode's fuel flow (in kilograms per second, per engine) sits in the ICAO Aircraft Engine Emissions Databank. We pull those figures for every aircraft type we track.
LTO fuel burn for a single flight is therefore:
LTO fuel (kg) = sum over the four modes of (mode duration in seconds × fuel flow per engine in kg/s × number of engines)
Multiply that fuel figure by 3.16 kg of CO₂ per kg of jet fuel and you have LTO CO₂.
The 3.16 emission factor comes from the chemistry of combustion. Jet-A1 is roughly 86% carbon by mass, and burning one kilogram produces around 3.16 kilograms of CO₂. It's a fixed industry constant used by the IPCC, ICAO and EUROCONTROL alike.
Phase 2: cruise
Cruise is everything above 3,000 ft: the long, fuel-efficient middle of the flight. Fuel burn per second drops sharply at altitude. A G650 at FL450 burns roughly half what it does on the climb out.
We model cruise as a flat hourly rate per aircraft type:
Cruise CO₂ = cruise hours × CO₂ per flight hour
The CO₂-per-hour figure comes from manufacturer fuel burn data, typically the long-range cruise figure in the Aircraft Flight Manual or performance tables, multiplied by 3.16.
Cruise hours are the tracked flight duration minus the airborne part of the LTO cycle: takeoff, climb out and approach, 6.9 minutes in total. Taxi happens on the ground, outside the tracked flight time, so it isn't taken out of it.
The headline formula
Put together, every flight figure on The Aircraft Ledger comes from one formula:
Total CO₂ = LTO CO₂ + (cruise hours × CO₂ per flight hour)
Two terms. One for the noisy, fuel-hungry bit at each end, one for the long, steady middle. Add them together and that's the flight.
Worked example: a Gulfstream G650 from London to New York
Let's take one G650 flight and trace it from raw tracking data to a final published number. It's a real flight from our database: N331AB, a Gulfstream G650, from London Luton (EGGW) to Teterboro (KTEB) on 28 April 2026.
Step 1: the aircraft
The tracking data carries the aircraft's Mode-S transponder code, which we cross-reference to the tail number in our registry. That resolves the airframe to:
- Type: Gulfstream G650
- Engines: 2 × Rolls-Royce BR725 (BR700-725A1-12)
- Cruise fuel burn: derived from manufacturer data. Gulfstream documentation puts a G650 at Mach 0.85 and FL470 at around 1,100 kg/hr at low cruise weight, but a fully fuelled westbound transatlantic burns materially more. We use a cruise CO₂ rate calibrated to typical operating conditions for the type.
- LTO fuel flow per engine: from the ICAO Aircraft Engine Emissions Databank entry for the BR725, which holds certified fuel flow at the four standard power settings.
These figures are loaded once per aircraft type and reused for every flight that type makes.
Step 2: the route
The flight's origin (EGGW) and destination (KTEB) resolve to coordinates:
- EGGW: 51.8747° N, 0.3683° W
- KTEB: 40.8501° N, 74.0608° W
The great-circle distance between them is 2,989 nautical miles (5,536 km): the shortest path over the surface of a spherical Earth.
It is not the distance actually flown. Real flights vector around weather, follow published routes and chase jet streams, typically adding 3–8% on a transatlantic crossing. That extra routing doesn't distort our figure, because cruise emissions are driven by measured flight time, not distance. A longer path simply shows up as a longer flight.
Step 3: the duration
The first airborne position for N331AB was received just after takeoff from Luton, and the last just before landing at Teterboro. Time between them:
24,056 seconds = 6 hours 40 minutes 56 seconds = 6.682 hours
A note on accuracy: measured duration can be clipped slightly if coverage near either airport is thin. We accept this as known noise and don't "correct" it with assumed times – we use what the data gives us.
Step 4: the LTO calculation
For each of the four ICAO modes, fuel flow per engine for the BR725 is multiplied by mode duration in seconds and by two engines, then summed.
Under the EMEP/EEA Guidebook 2023 methodology (see the aviation chapter and LTO calculator), the BR725-powered G650 uses approximately 390 kg of fuel across the full cycle (both engines, all four modes).
LTO CO₂ = 390 kg fuel × 3.16 = 1,232 kg CO₂ ≈ 1.23 tonnes
This figure is fixed per aircraft type. Every G650 flight gets the same LTO emissions because the cycle is standardised. What varies between flights is cruise.
Step 5: the cruise calculation
Taxi happens on the ground, so only the airborne LTO modes come out of the tracked time: takeoff (0.7 min), climb out (2.2 min) and approach (4.0 min). That's 6.9 minutes, or 0.115 hours:
Cruise hours = 6.682 − 0.115 = 6.567 hours
Our G650 cruise rate is approximately 4,720 kg of CO₂ per hour, equivalent to about 1,494 kg of fuel per hour. That's calibrated for representative operating conditions rather than the light, optimal-altitude figure quoted in marketing materials. A heavy westbound transatlantic with full fuel burns notably more per hour than a G650 near the end of cruise at FL510.
Cruise CO₂ = 6.567 hours × 4,720 kg/hr = 30,996 kg CO₂ ≈ 31.00 tonnes
Step 6: the total
Total CO₂ = 1.23 + 31.00 = 32.23 tonnes CO₂
That's the published figure: approximately 32.2 tonnes of mission CO₂ for one G650 crossing from London Luton to Teterboro on 28 April 2026.
The LTO cycle is a fixed 1.23 tonnes for every G650 flight, so on a long crossing almost everything comes down to measured time in the air.
A sanity check
A G650 operator has publicly quoted about 16,000 lb (around 7,260 kg) of fuel for a Teterboro–Le Bourget crossing at Mach 0.85, a similar distance in the opposite direction. That's about 22.9 tonnes of CO₂ eastbound.
- This flight, westbound (our method)32.2 t
- Operator fuel quote, eastbound22.9 t
Our westbound figure of 32.2 tonnes is materially higher. Part of that gap is direction: westbound transatlantic flights face headwinds, take longer and burn more fuel per mile. The rest reflects our cruise rate, which is calibrated to representative operating conditions across the type rather than a single operator's light, well-planned crossing. Both effects push the same way, so for long westbound legs our figure sits towards the higher end of what an individual operator might report.
What that figure represents, and what it doesn't
| It represents | It does not represent | | --- | --- | | Tank-to-wake CO₂ from the aircraft's engines, from taxi-out to taxi-in | Upstream lifecycle emissions from extracting, refining and transporting the fuel | | Manufacturer-published fuel burn data | APU burn on the ground before pushback | | Actual tracked flight duration | Reserve fuel that was loaded but not burned | | The specific aircraft type and engines | Non-CO₂ effects: NOx, water vapour, contrails, soot | | The whole flight | Per-passenger allocation, or ground operations such as towing, ground power and catering |
Every item on the right is a legitimate thing to measure. We just don't measure it in this number.
What other trackers and calculators do differently
This is where reasonable people make different methodological choices. Understanding them is essential if you're comparing figures across sources. Here's how the same G650 flight changes depending on what each method chooses to count:
- Tank-to-wake (The Aircraft Ledger)32.2 t
- Well-to-wake, lifecycle +15 to 25%37–40 t
- With radiative forcing, 1.7x to 3x55–97 t
Every bar starts from the same flight. The spread comes entirely from what each method chooses to count.
Lifecycle (well-to-wake) vs tank-to-wake
Some calculators publish lifecycle emissions, often called well-to-wake. That covers everything from extracting crude oil and refining it into Jet-A1 to distributing it to airports and finally burning it.
Lifecycle multipliers typically add 15–25% on top of tank-to-wake. A G650 transatlantic we publish at about 32.2 tonnes tank-to-wake might be about 37–40 tonnes well-to-wake.
Both are real; they answer different questions. Tank-to-wake asks "what came out of this aircraft's engines?" Well-to-wake asks "what was the total carbon cost of this flight, including the fuel supply chain?" We publish tank-to-wake because it isolates what's attributable to the aircraft and its operation, and stays comparable across flights and types without fuel-supply variables.
Contrails and non-CO₂ radiative forcing
Aviation's climate impact isn't only CO₂. Contrails, water vapour and NOx at altitude trap heat in ways that are scientifically real but work differently from CO₂'s long-term warming.
Some calculators apply a radiative forcing multiplier, typically 1.7× to 3×, to capture total climate impact. We don't, for two reasons:
- The science on the right multiplier is genuinely unsettled, and a wrong multiplier introduces more error than leaving it out.
- Mixing CO₂-equivalent units with raw CO₂ across sources creates apples-to-oranges comparisons.
If you want to apply a multiplier to our figures for a specific analysis, you can.
Passenger load factor
Commercial aviation calculators often divide emissions across an assumed number of passengers. That makes sense when comparing modes of transport.
It makes less sense for private aviation, where occupancy is highly variable, often confidential and frequently very low. A G650 might leave London with eight passengers, two, or none on a repositioning leg. We publish the whole-flight figure and let analysts decide whether and how to allocate it.
Block time vs ADS-B time
Some estimators use scheduled block time, gate departure to gate arrival. That works for airlines, whose schedules are public. Private aviation has no schedule, just a flight.
We use measured flight duration from our tracking data. It's the cleanest signal available for unscheduled traffic, and it reflects when the aircraft was actually flying.
Fixed averages vs aircraft-specific data
A surprising number of public calculators use one CO₂-per-hour figure for "private jets" as a category, sometimes applied to everything from a Citation Mustang to a Global 7500. The fuel burn difference between those two is roughly 6×.
We resolve every flight to its specific aircraft type and apply that type's fuel burn profile. That matters enormously when comparing fleets, operators or routes.
Limitations and known sources of error
We're open about what the model gets wrong. Anyone publishing emissions data without acknowledging uncertainty is overselling.
Tracked duration covers the airborne flight only. Taxi emissions therefore come entirely from the ICAO standard cycle, not from what happened on the ground that day.
Tracked duration may be clipped at either end. Where receiver coverage near an airport is thin, the first or last signal can come after takeoff or before landing, shortening the measured flight slightly.
LTO uses ICAO default cycle values. Real taxi times at busy airports can be far longer than 26 minutes, and much shorter at quiet ones. We use the standard ICAO cycle for consistency. Airport-specific delays would need ground movement data we don't currently hold.
Cruise burn is modelled linearly. We use one cruise rate per aircraft type. In reality fuel burn changes with altitude, payload, headwind and stage of cruise (early cruise is heavier, late cruise lighter). The error is typically within ±5% on a long flight but can be larger on short legs.
Signal gaps and detection edge cases. Tracking coverage is excellent over land in Europe, North America and most populated regions, and patchy mid-ocean. Our pipeline handles gaps without inventing data, but occasionally a flight is only partially detected. We flag these where we can.
Taken together, our figures are best read as directional estimates of mission CO₂, accurate to within roughly ±10% on most flights.
Why we publish this
Transparent methodology is the only thing that makes the numbers useful. A CO₂ figure from a black box is just a claim.
A figure with a published method, a verifiable emission factor, a duration you can cross-check and a worked example you can replicate is different. A journalist can stand behind it, an analyst can cite it, and an operator can rebut it on specifics.
We welcome corrections. If you operate an aircraft we track and have better fuel burn data than the manufacturer figures we use, we want to hear from you. If a flight's duration looks off, tell us. The methodology gets stronger when it's challenged.
In summary
How is private jet CO₂ calculated? The short version:
- Identify the aircraft from its tail number and load its engine and fuel burn profile.
- Resolve the origin and destination airports to coordinates.
- Take the airborne flight duration directly from tracking data.
- Calculate LTO emissions using the ICAO standard cycle and per-engine fuel flow.
- Calculate cruise emissions as cruise hours × cruise fuel burn × 3.16 kg CO₂ per kg of fuel.
- Add the two together and publish the total.
No hidden multipliers. No assumed loads. No lifecycle. Just the carbon dioxide that came out of the engines, calculated with a public, reproducible method drawn from the EMEP/EEA Guidebook.
Sources
- The Aircraft Ledger flight data. Methodology
- Aircraft type: Gulfstream G650/G650ER
First published .