Methodology: flight emissions

Oncarbon’s flight emissions calculation method takes into account a wide variety of a flight itinerary’s properties: plane models, engine models, the number of seats in the aircraft, an estimation of the amount of cargo on the flight, airport busyness factors, and the great-circle distances between the origin, destination, and any stops on the way.

In addition to direct CO₂, NOx, and other emissions, flying at high altitudes also results in atmospheric warming effects via some complex processes like the formation of contrails and cirrus cloudiness. Our numbers also account for these effects, and the total warming effect is communicated in terms of kilograms of carbon dioxide equivalent (CO₂e). What’s more, we account for the emissions that happen not only as fuel is burned in the engine of the aircraft, but also when the raw materials for that fuel are sourced and the fuel is produced.

Once our model has arrived at the warming effect of one entire flight, we attribute those emissions to each seat on that flight. When we do this, we take into account an estimation of the amount of cargo on the flight and the number of seats installed on that specific aircraft. As a result we deliver a pretty accurate estimation of the warming effect of one flight seat, communicated as kilograms of CO₂e per seat. Here’s all of that explained in more detail:

Step 1: calculate fuel burn

To calculate the fuel burn of a given flight we use four main parameters: plane model,engine models installed on that plane, great-circle distance of the flight, andairport busyness of the origin and destination airports. If an itinerary (i.e. a combination of flights) has more than one flight, we use all of the above parameters for each of those flights.

The fuel burn calculation method is inspired by the work done by the European Environment Agency. ¹ Airport busyness and the great-circle distance have an effect on the length of different flight phases (taxi out and taxi in, take off, climb, cruise, etc.). Theplane model and the engine models are then used to determine the engine thrust used in those phases.

As a result of these calculations, we arrive at an estimate of how much fuel is burned in each of the flight’s phases. If the itinerary includes several flights, we repeat all of the above steps for each of those flights.

In some cases, we are not able to estimate fuel burn for every flight phase. When that happens, we break down the emissions to only two phases or present just one emission number for the entire flight.

Step 2: convert fuel burn to emissions and to the warming effect

As fuel is combusted in the engine of an aircraft, the exothermic chemical reaction produces several products, one of which is carbon dioxide (CO₂). The ratio at which the combustion of one kilogram of fuel produces CO₂ is a physiological constant of 3.157 ², and therefore the conversion of fuel burn to CO₂ emissions is a straightforward one.

CO₂ is a major greenhouse gas and also a big contributor to the total warming effect of flight emissions. Unfortunately, the emissions from aviation heat the atmosphere in other ways as well. Other important warming effects result from the release of nitrogen oxides (NOx), water vapor, and contrails (increased cloudiness).

The warming effect of these non-CO₂ terms is traditionally accounted for with something called the Radiative Forcing (RF) index. The RF index describes the rate at which flight-time emissions at high altitudes are warming the atmosphere compared to the associated CO₂ emissions alone.

A recent study ³ concluded that a realistic RF index for flights that happen today is 3.0, whereas many earlier calculations have used indices ranging from 1.5 to 2.0. Although, as any good study into this topic will admit, there are uncertainties associated with any RF index, our model uses the latest research into the topic, and hence we use the RF index of3.0.

Step 3: add emissions that result from fuel production

Emissions that are directly related to flying happen not only as fuel burns in the engine of an aircraft, but also as that fuel is produced and the raw materials for it are sourced.

To account for those emissions, we use a constant of 0.617 kg of CO₂ per each kilogram of fuel produced (based on the energy density of jet fuel and an assessment of its life-cycle emissions⁴), which we add to the warming effect we arrived at in the previous step.

Step 4: attribute the warming effect to one seat of a flight itinerary

Now, we are at a point where we have calculated a warming effect for one entire flight.

To get to CO₂e emissions per seat, we first attribute a portion of the emissions of the entire flight to the cargo on that flight. To do this, we use an estimation of the amount of cargo on the flight, which is based on information on the movements of global air cargo between geographical regions. ⁵

After we’ve attributed part of the emissions to the cargo, we distribute the remainder of the emissions between the number of seats installed on the plane taking into an account the seating class factor. Seating class factors vary depending on the aircraft body type (wide or narrow) and follow IATA recommended practice .⁶

Again, for a flight itinerary with more than one flight, we repeat this step individually for each of those flights.

We are always happy to discuss and receive feedback on our calculation methodology: send feedback here

Comparison to other flights on the same route

To make comparison between different flight options easier, we also show how a specific itinerary compares to other options on the same route. This comparison is done with a distribution graph, which plots the emissions of different flight options based on global scheduled data. To build this graph, we look at all available flight options on that route that depart up to three days before and up to three days after the selected itinerary (i.e. within one week).

In addition to direct flights, we include plausible connecting flight options into the comparison. What qualifies as a plausible connecting flight is based on the transfer times at the connecting airport (a minimum of 1 hour and a maximum of 4 hours) and the duration of the total flight time. For example, where the shortest possible flight time is 4 hours, the longest plausible flight time that we consider is 5.6 hours, based on the following logic. When shortest flight time available on this route = Tminand the longest plausible flight time = Tmax:

if T_min < 3 h, then T_max = 1.5 * T_min
if 3 h ≤ T_min < 6 h, then T_max = 1.4 * T_min
if T_min ≥ 6 h, then T_max = 1.3 * T_min

This way, we ensure that our comparison doesn’t include non-sensible connecting flight options, like flying from London to New York with a change of planes in Dubai.

We then calculate emissions for all those alternative flight options and divide them into buckets based on their total emissions. The tallness of the bucket shows how many flight options there are available with similar total emissions.