Carbon Calculator Methodology: Formulas, Sources and Limits
Version 4.2 · methodology updated 2 November 2026
Choose the method that matches your tool
The seven-step personal calculator and the bill/device comparison tool are different tools. Neither is a measured footprint, complete lifecycle inventory, utility bill or research certification. Teaching values are labelled; replace them with your own.
The seven-step calculator: scope
Version 4.2 combines entered home electricity, cooking fuel, generator fuel, your own road travel and entered flight passenger-kilometres, plus broad diet-study context. Electricity and diet use lifecycle CO2e; fossil fuel and aviation factors are combustion CO2 only and omit additional gases/upstream stages. The displayed combined estimate therefore has mixed and incomplete coverage. Goods, services, much infrastructure and other sources are not included. A low result does not prove a complete lifestyle meets a climate budget.
Grid electricity
One unit equals one kWh. Annual personal allocation = monthly grid units × 12 × national electricity factor ÷ household members. The embedded Ember snapshot was downloaded 24 October 2026 (CC BY 4.0), covers 209 countries/economies, and selects a usable observation per country. Displayed years differ, including older observations for incomplete series. Factors are rounded in this tool, and are not live measurements.
National generation-weighted lifecycle intensity is a proxy, not your supplier's import-adjusted delivered-electricity or marginal factor. g CO2e/kWh is divided by 1,000 to get kg CO2e/kWh. The country article explains observation years and the site's screening rule. Do not substitute these factors into a CO2-only calculator.
Households and travel sharing
Home electricity, cooking and generator emissions are allocated equally across the positive whole-number household size. Travel fields describe journeys you personally take. Annual vehicle distance = your weekly distance per vehicle × 52 × entered vehicle count. Petrol/diesel litres = annual km ÷ entered km/litre. Divide vehicle emissions and costs by the entered people sharing those journeys, independently of household size.
Petrol uses approximately 2.31943 kg combustion CO2/litre and diesel 2.69720; the formulas below retain full conversion precision. An electric car uses 18 kWh/100 km × the selected grid factor ÷ occupants; this is an illustrative efficiency, not a vehicle-specific measurement. Exclude separately counted EV charging from home electricity to avoid double counting. Public transport, embodied vehicle emissions and fuel upstream emissions are not modelled here.
Cooking, generator, flights and food
All fuel and flight factors below estimate combustion CO2 only. They omit upstream fuel emissions, CH4 and N2O. Electricity and diet use lifecycle CO2e, so the combined partial estimate has mixed coverage.
| Input | Factor and derivation | Source / limits |
|---|---|---|
| Petrol / diesel | 8.78 / 3.785411784 = 2.3194306; 10.21 / 3.785411784 = 2.6971967 kg CO2/litre. | EPA Hub, 15 January 2025, table 2. US gallons; combustion CO2. |
| LPG | 47.3 TJ/Gg × 63,100 kg CO2/TJ / 1,000,000 kg/Gg = 2.98463 kg CO2/kg. | IPCC 2006, volume 2, chapter 1, tables 1.2 and 1.4, LPG rows. Generic defaults, not a measured local LPG blend; no density conversion is required. Enter net fuel mass per cylinder; 11.8 kg is an editable teaching input. |
| Natural gas | 0.05444 kg CO2/standard cubic foot / 0.028316846592 m3/ft3 = 1.9225305 kg CO2/standard m3. | EPA Hub table 1, Natural Gas. A US standard-volume/composition proxy. The volume must use the same reference basis; do not treat uncorrected local meter volume, therms or MMBtu as interchangeable. The historical tool accepts energy units when these match the bill. |
| Coal | EPA kg CO2/short ton / 907.18474 kg/short ton: bituminous 2325 → 2.5628738; anthracite 2602 → 2.8682140; sub-bituminous 1676 → 1.8474738; lignite 1389 → 1.5311104 kg CO2/kg. | EPA Hub table 1, named coal rows. Select the supplier's actual coal type. These defaults are not interchangeable with charcoal or coke; unknown coal classification is a limitation. |
| Wood / other cooking fuel | Not estimated; excluded, not zero. | The unsupported 0.4 kg/kg factor has been withdrawn. No universal net biomass factor is substituted. Results, category labels and the running total flag this exclusion. |
| Flights | 0.207 / 1.609344 = 0.1286238; 0.129 / 1.609344 = 0.0801569; 0.163 / 1.609344 = 0.1012835 kg CO2/passenger-km. | EPA Hub table 10. Short: leg under 300 miles (482.8032 km); medium: 300 to under 2,300 miles (482.8032 to under 3,701.4912 km); long: at least 2,300 miles (3,701.4912 km). CO2 only, no cabin-class adjustment or other aviation warming. |
Cooking annual allocation = cylinders/month × entered net kg/cylinder × 12 × LPG factor ÷ household members, or standard gas m3/month × 12 × gas factor ÷ household members, or named coal kg/week × 52 × its factor ÷ household members. Wood/other cooking emissions are not included. Generator litres/week × 52 × the selected petrol/diesel factor are also allocated across the household. One main cooking fuel is modelled; mixed-fuel homes need a separate documented inventory.
Enter your annual passenger-kilometres separately for each flight-leg distance band. Annual flight CO2 = short-band km × short rate + medium-band km × medium rate + long-band km × long rate. Classify each connecting leg by its own length; add all return legs. A 600 km leg belongs in the medium band. No fixed 600/2,000/7,000 km distance or time-band conversion remains. The 4,000 km initial medium-band input is a labelled teaching example, not inferred travel.
Source correction, 2 November 2026, version 4.2: LPG 2.94 and natural gas 1.93 defaults were replaced with the explicit derivations above; coal 2.42 was replaced by named coal-type factors; the unsupported wood factor was withdrawn. Flight counts and assumed distances were replaced by entered passenger-kilometres with EPA distance bands. Petrol/diesel now retain the full conversion precision. Recalculate earlier fuel/flight results; input fields and tool version have changed. The historical bill/device tool and the CO2-only US worked example retain their separate methods.
Diet context uses Scarborough et al. (2023) daily GWP100 CO2e medians: vegan 2.47, vegetarian 4.16, fish/no meat 4.74, low meat 5.37, medium 7.04, high 10.24 kg/day × 365. These are UK adult group comparisons standardised to 2,000 kcal/day, not a global meal inventory or nutrition prescription.
Diet-factor correction, 2 November 2026: version 4.1.1 replaces four values that did not match Table 3 of Scarborough et al. Vegetarian, fish, low-meat and medium-meat groups now use 4.16, 4.74, 5.37 and 7.04 kg CO2e/day respectively. Vegan and high-meat values are unchanged. For the medium-meat teaching input, the annual food component changes from 2,054.95 to 2,569.60 kg CO2e. Earlier results using these four groups should be recalculated. This changes a modelled comparison, not a measurement of a person's actual diet.
Appliances: preserve the entered numbers
Monthly units = electrical input watts ÷ 1,000 × device quantity × running hours/day × 30. Library watts are editable teaching estimates. Use measured average input for cycling equipment; nameplate watts × connected hours can overstate use. The list is never rescaled to fit a bill. If device totals exceed grid units plus entered solar self-use, a warning is shown and device-saving scenarios are withheld. An unexplained difference is not evidence of waste.
Solar and backup power
Panel capacity, battery choice or export eligibility alone does not generate estimated output, income or emissions credits. Only entered measured monthly solar self-use is added to device reconciliation; exports are recorded separately. Grid bill emissions are not reduced again for existing solar. UPS use is already in the bill and no fixed battery-loss assumption is added. Solar manufacturing is not quantified in this partial result.
A future-solar comparison accepts the monthly grid units the visitor expects to replace. It is an explicit what-if assumption, not an installer forecast. No export credit, system cost, payback period or guaranteed marginal emissions saving is calculated.
Money and combined changes
Prices start blank. Enter your currency code and applicable avoidable flat electricity energy charge or fuel price. Unknown prices do not produce a money estimate. Flat-rate estimates exclude fixed fees, taxes, changing eligibility, time-of-use, export revenue and capital costs. National price averages are descriptive context, not your marginal tariff. Use the separate bill comparison tool for manually verified progressive blocks; neither tool automatically verifies your utility's bill.
The switch-off comparison uses at most min(1, entered daily runtime) fewer hours, so it cannot remove more running time than exists. Reduce only unused, non-essential operation. Preserve necessary cooling, food storage, water and medical equipment.
Device tips use stated illustrative reduction percentages rather than claiming a measured effect of the advice. For the combined scenario: apply device-unit reductions up to grid imports, then apply prospective solar replacement to remaining imports. Remaining imports cannot be negative; the same electricity charge is not saved twice. Your allocated travel reduction is reported separately. Emissions changes hold factors/efficiency constant and are accounting comparisons, not marginal grid predictions.
Context and reproducibility
The 21 displayed country averages and world comparator were checked against Our World in Data, CO2 emissions per capita, 2024, using Global Carbon Budget (2025) and Population based on various sources (2024); indicator updated 13 November 2025, archive downloaded 2 November 2026. Country values are rounded to one or two decimals (trailing zeroes may be omitted); world 4.729075 is shown as 4.7 tonnes/person. The chart's CSV, metadata and original values are retained in the verification archive. These are territorial fossil-fuel and industrial CO2 excluding land-use change; international aviation/shipping is counted in the world total only. National/world fossil CO2 averages have different coverage from this partial estimate. They are labelled context, not a like-for-like ranking. The seven-step tool does not provide the older tool's downloadable versioned research record or factor-year selection. Record your inputs, tool version, assumptions and sources independently; do not describe output as measured data or a validated study. No numerical accuracy guarantee is made.
Sources
- Ember Yearly Electricity Data and methodology: national generation lifecycle intensity.
- EPA January 2025 Emission Factors Hub: tables 1 (gas/coal), 2 (petrol/diesel) and 10 (flights); retain units and scope.
- IPCC 2006 Guidelines, volume 2, chapter 1: LPG defaults, tables 1.2 and 1.4.
- Scarborough et al., Nature Food, 2023: diet group context, table 3.
- Our World in Data CO2 per capita dataset: 2024 territorial fossil/industrial CO2 context, indicator updated 13 November 2025, verified 2 November 2026.
Sources do not endorse this site. Contact the publisher with a specific source, units and proposed correction.
Historical tool methods and the 3.47 t US example
Historical global-grid and CO2-only tools
Version 3.0-global-grid reports electricity greenhouse gases (CO2e), combustion CO2 and diet study context separately. It is an educational calculation of entered activities, not a measured or complete footprint or certified inventory. The historical calculator has no assumed country. The standalone electricity page and historical US worked example retain their explicitly labelled CO2-only methods.
Global electricity factors
The historical calculator embeds a snapshot of Ember Yearly Electricity Data, retrieved 24 October 2026 under CC BY 4.0. It covers 209 countries and economies, with 2023–2025 records where available and the latest earlier record otherwise. Each country selector shows actual available years; no missing year is filled with a US or world average.
National generation-weighted intensity reflects each country's reported electricity mix and lifecycle greenhouse gases. Divide the published g CO2e/kWh by 1,000 to obtain kg CO2e/kWh. Electricity units × monthly/annual multiplier × factor ÷ household size gives the personal allocation. EV electricity uses the same selected factor. Household electricity must exclude separately counted EV charging and self-generated solar.
This is a generation-based proxy, not an import-adjusted consumption or utility-specific delivered-electricity factor. It does not add distribution losses or model marginal changes. For a local study, a matching published utility factor may be more suitable; custom inputs must identify the source, year and scope. See Ember's methodology and country notes.
Data quality and reproducibility
The snapshot contains 495 country-year records. A conservative screening rule disables records with negative intensity or generation shares outside 0–100%. Eleven records from 2025 are flagged; raw values remain in exports with a usability flag. Each affected country has an earlier usable record. This is our screening rule, not a correction or endorsement by Ember. The source data may be revised.
Download the exact factor snapshot and JSON calculation record from the historical calculator. They identify the data retrieval date, original file SHA-256 fingerprint, calculator version, inputs, factors, scopes and results. CSV results support analysis, and activity-year mismatches are flagged. Cite the underlying sources, document how you collected activity data, and assess uncertainty and sampling limitations. Calculator output is modelled data, not new measured emissions evidence.
Historical CO2-only electricity and optional fuel
Annual electricity CO₂ = entered kWh × period multiplier × kg CO₂/kWh. The multiplier is 12 for monthly input and 1 for annual input. Divide by household members for an equal per-person allocation. The historical US CO2-only reference is explicitly frozen to the EPA Hub dated 15 January 2025, table 6, US Average, total-output CO2 column (underlying eGRID2023 January release), not the later revised eGRID summary table. It is 771.5 lb CO₂/MWh × 0.45359237 kg/lb ÷ 1,000 kWh/MWh = 0.349946513455 kg CO₂/kWh.
Fuel fields appear only when the user selects that the home also burns fuel. Unchecking the option excludes that fuel even if an earlier amount remains in the field. Fuel uses consumption × period multiplier × selected fuel factor ÷ household members. Natural gas is 5.306 kg CO₂/therm, derived from 53.06 kg CO₂/MMBtu and 0.1 MMBtu/therm. Heating oil uses 10.21 kg CO₂/US gallon; LPG uses 5.68. Divide gallon-based factors by 3.785411784 for litres. Use the bill's actual unit; do not enter cubic metres as therms.
Road travel and electric cars
Fuel use = annual km × litres/100 km ÷ 100. Petrol uses 8.78 kg CO₂/US gallon and diesel 10.21, each converted to litres. These use the EPA inventory fuel factors, which can differ from other EPA consumer calculators. Divide road emissions by the number of people sharing those journeys. Electric vehicles use annual km × kWh/100 km ÷ 100 × the chosen electricity factor ÷ occupants. Keep vehicle charging out of the household electricity input to avoid double counting.
Flights
Annual passenger-km are entered separately for short, medium and long flight legs. The tool uses EPA table 10 CO₂ factors of 0.207, 0.129 and 0.163 kg/passenger-mile respectively, divided by 1.609344 for kilometres. Original boundaries are under 300 miles, 300 to under 2,300 miles, and 2,300 miles or more; the precise kilometre boundaries are 482.8032 and 3,701.4912 km (historical tool labels are rounded). Include connecting legs and returns. Factors are class-average, not cabin-specific. Upstream fuel emissions, non-CO₂ gases and aviation's other warming effects are not included.
Separate diet comparison
Diet context uses daily GWP100 CO₂e medians from table 3 of Scarborough et al. (2023), multiplied by 365. The research standardizes UK adult diets to 2,000 kcal/day. Its six groups are broad comparisons, not meal-by-meal or country-specific personal inventories. Diet includes lifecycle greenhouse gases and is displayed separately; adding it to combustion-only CO₂ would blur the calculation boundaries.
Uncertainty and scenarios
Input quality, regional supply, fuel composition and behavior all affect accuracy. No numerical accuracy percentage is claimed. The scenario reduces selected activity proportionally while holding factors and efficiency constant. It is not an estimate of marginal power-system savings. The total omits goods, services and sources not entered. Blank required fields, negative amounts and invalid household sizes are rejected.
Sources and maintenance
- Ember Yearly Electricity Data: global generation mixes and lifecycle CO2e intensity, snapshot retrieved 24 October 2026, CC BY 4.0.
- US EPA Emission Factors Hub, January 2025: tables 1, 2, 6 and 10.
- EPA eGRID2023 summary data: electricity generation averages.
- Scarborough et al., Nature Food, 2023: dietary impacts, table 3.
Dataset years are stated rather than represented as live values. Contact the publisher with a specific source and correction; changes should identify the factor, units and affected results. Source organizations do not endorse this site.
Updated 2 November 2026 · Climate Calculator. Calculation version 4.2; source dataset years remain as stated.
No comments:
Post a Comment