THE SCIENCE BEHIND THE NUMBER

How your car’s carbon history is calculated

Distance becomes fuel or electricity. Materials become a car. Replacement parts mark the years. Here is how those pieces become the number you see.

What goes into the total?

Stage What we count What you should know
Production Vehicle materials, assembly and a traction battery where applicable Historical process estimates, not factory records for your car
Driving Fuel combustion, fuel supply and electricity supply Depends on distance and consumption; supply factors are historical proxies
Servicing Replacement parts and fluids when a scheduled event occurs Generic intervals and assumed quantities, not your service invoices
End of life Treatment when you choose recycling Ending the simulation alone does not scrap the car
Recycling benefit Potentially avoided primary steel and aluminium production Separate from gross emissions; no atmospheric removal is implied

From litres to tonnes

Fuel used = kilometres × consumption per 100 km ÷ 100. Multiply that quantity by the appropriate CO₂ factor. Divide kilograms by 1,000 to display tonnes.

Fuel Unit Combustion kg CO₂ / unit Fuel supply kg CO₂ / unit
petrol L 2.33955 0.43883
diesel L 2.62818 0.48887
LPG L 1.55491 0.18851
CNG kg 2.50268 0.39994

The combustion column comes from the separate CO₂ component of the UK government’s 2026 conversion factors. These mineral-fuel values approximate fuel chemistry, not every country’s historical blend. Supply includes extraction, processing and delivery before combustion. Its physical-CO₂ proxies come from the TU Wien study, Table 25; petrol and diesel supply use its US GREET entries as geographical proxies.

For example, 8 L/100 km over 100,000 km uses 8,000 litres, producing 18.7 tonnes of direct CO₂ under the petrol factor. Supply, production and servicing are additional stages.

After a gas conversion

Pre-conversion kilometres retain their petrol emissions. After conversion, gas consumption and any additional petrol you enter apply. Installation adds assumed equipment: 30 kg of steel for LPG or 50 kg for CNG.

LPG starts at 20% more litres than petrol: 8 becomes 9.6 L/100 km. This is an adjustable ATLAS assumption, not a universal rule. DOE’s propane explanation supports lower energy per litre, not this exact percentage for every LPG blend or engine. Use actual consumption when available. CNG uses kilograms and a separate DOE energy-equivalence starting point.

Electricity: where the estimate is less specific

Supply proxy g CO₂ per kWh
European Union 428.80
Germany 558.34
Austria 207.55
United States 514.06

These are fixed historical supply factors from the 2017 TU Wien study, Table 25. Other countries use the EU proxy. They are neither current national measurements nor forecasts and cannot reproduce a country’s changing electricity supply.

Wall or charger readings are used directly. Battery readings are divided by 0.90 for an assumed 10% charging loss: 18 kWh at the battery requires 20 kWh at the wall. The adjustment is not applied twice.

A plug-in hybrid uses a simplified electric-distance share and separate consumption for each mode. Do not enter an already blended fuel figure as the fuel-mode value. Real PHEVs can combine both energy sources; DOE explains their operation.

Production happens once

The model estimates a body material mix from vehicle mass, then adds assembly and the traction battery. Profiles interpolate between historical snapshots for 1985, 2001, 2010 and 2018, holding the nearest endpoint outside that period. ORNL table 7.15 (PDF, page 154) and National Academies table 7.6 describe broad North American averages, not an exact European model.

Missing mass is estimated by class: 1,200, 1,500, 1,900 or 2,200 kg. Traction-pack mass is estimated at 6 kg per kWh for BEVs/PHEVs or 35 kg for hybrids, and subtracted before calculating body materials. BEV/PHEV battery capacity must be entered or confirmed.

Production factors and assumptions
Material kg CO₂ per kg
steel 2.92
iron 0.82
aluminium 6.23
rubber 3.16
plastics 3.80
glass 0.58
copper 3.97
fluids 3.64

Assembly adds 683.64 kg CO₂. Historical battery-material factors are 4.27 kg CO₂/kg for BEVs, 4.70 for PHEVs and 4.98 for hybrids; lithium-ion assembly adds 49.13224 kg per pack. These older process proxies do not represent each modern chemistry or factory. See TU Wien Tables 10, 21, 23 and 24. Aluminium is an assumed equal wrought/cast mix; electronics and other metals use copper, and miscellaneous material uses plastic. These substitutions add uncertainty.

The car’s history and your ownership share

Vehicle history places production at manufacture. My ownership allocates it over the first ten registration years using this ATLAS convention:

Registration year of life Production share
1 25.6%
2 19.3%
3 14.6%
4 11.1%
5 8.5%
6 6.5%
7 5.0%
8 3.9%
9 3.1%
10 2.4%

The shares total 100%. They are not atmospheric decay, a scientific depreciation law or removal. After ten years no production share remains to allocate, but historical emissions still exist. Early retirement assigns the unallocated remainder to the last owner.

Buying at 100,000 km and driving another 200,000 km means 200,000 km of your travel and 300,000 km on the car. Unknown earlier travel is distributed evenly and appears only in vehicle history.

Years, scenarios and averages

A 15-year simulation starting in 2021 covers 2021–2035, ending at the start of 2036. The current year and later years are projections. Annual sliders adjust individual years between 50% and 150% of the starting average; they change the total rather than redistribute kilometres.

Changing cars takes effect at the start of the current year. Both scenarios share past ownership and identical future travel. A sold car may keep emitting elsewhere. This is an ownership-allocation comparison, not a forecast of the global consequences of a purchase.

Annual bars show each year’s emissions, split by stage. The large total accumulates through the selected year. Average g CO₂/km is gross accumulated CO₂ divided by the applicable distance; it is undefined at zero kilometres. The separate million-kilometre explorer counts full production once and assumes vehicle survival, without predicting future failures.

Servicing creates individual events

Event Modelled interval
Tyres - set of four 20,000 km or 10 years
Engine oil & filter 15,000 km or 1 year
Brake pads - set 40,000 km
Brake discs - set 80,000 km
Shock absorbers - set 80,000 km
Brake fluid 2 years
Coolant 5 years
12 V battery 5 years

Each event adds assumed replacement materials at its distance or age threshold. Quantities scale by the square root of vehicle mass divided by 1,500 kg, an ATLAS assumption. Earlier service timing is reconstructed from estimated prior travel, not known invoices.

The tyre distance is a scenario choice; Michelin supports the ten-year precaution. Volkswagen gives 9,300 miles (about 15,000 km) or one year as a fixed-service example. Monroe recommends shock replacement at 80,000 km. Brembo’s general brake-fluid guidance says 2–3 years; the model chooses two. Other intervals and quantities are assumptions, not maintenance instructions.

Recycling: a separate potential benefit

Only an explicit recycling choice activates treatment and a recovery credit. Treatment adds 176.01 kg CO₂. The final owner receives the potential credit, even with a small production allocation.

Gross emissions − potential recycling benefit = net allocated balance. Gross emissions remain visible; a negative balance would not mean a carbon-negative car.

How the material credit is calculated

Credit = maximum of zero and (material mass × recovery × substitution − material mass × recycled-input share), multiplied by (primary-production factor − secondary-production factor).

Material Primary / secondary kg CO₂ per kg Recovery Substitution Recycled-input share
Steel 2.26 / 1.21 90% 90% 26.4%
Aluminium 10.08 / 0.27 90% 80% 50%

Factors come from TU Wien Appendix B.3.1–B.3.2. Recovery, substitution and the mixed-aluminium recycled-input share are assumptions. Subtracting previously credited recycled input avoids double counting. Secondary processing is already deducted. Copper recovery is listed without a compatible credit; battery-recycling benefits and residual incineration are not quantified.

EPA’s WARM explanation describes waste-management scenarios. Its aggregate CO₂e factors are not used as physical CO₂ here. Avoiding future material production differs from removing past emissions.

CO₂ is not the same as CO₂e

CO₂ is a mass of carbon dioxide. CO₂e expresses the warming effects of several greenhouse gases in a common unit. Trees take up CO₂; they do not directly capture the methane represented in a CO₂e total. No universal conversion ratio is applied. Other gases remain important even though they are outside this headline.

What this estimate cannot tell you

It cannot reconstruct exact factory emissions, historical fuel blends, every country’s annual grid or your full service record. Workshop energy, refrigerant leakage, roads, major unplanned failures and traction-battery replacement are not comprehensively modelled. No precise confidence interval or verified inventory is claimed. The purpose is to make the included stages and assumptions understandable.

Explore how biological carbon removal works →