Why there is no single number for a kilowatt-hour
A gallon of gasoline always releases 8.887 kg of CO2, because the number is set by the fuel's chemistry. Electricity has no such property. The kilowatt-hour that runs your dishwasher is physically indistinguishable from any other, but the emissions attributed to it depend entirely on what was generating at the time and place you drew it.
That is why the emission factor is an input to this calculator rather than a constant. EPA's eGRID database divides the US power system into subregions and publishes an emission rate for each in pounds of CO2e per megawatt-hour. The spread is enormous: hydro- and nuclear-dominated regions come in at a small fraction of coal-heavy ones. Using a national average where a regional figure is available is the largest avoidable error you can make in this calculation, and it can be wrong by a factor of several in either direction.
Three further refinements separate a rough estimate from a number you can put in a disclosure. Intensity — pounds or grams per kWh — lets you compare grids and set targets independently of volume. Line losses capture the fact that more electricity was generated than arrived at your meter. And market-based accounting reflects the contracts you actually signed rather than the physics of the wires, which is how a company on a fully renewable tariff reports differently from its neighbour on the default supply.
Location-based and market-based: you report both
The GHG Protocol Scope 2 Guidance requires dual reporting. Neither figure replaces the other, and quoting only the flattering one is the most common reporting failure in corporate disclosures.
The location-based figure multiplies your consumption by the average emission rate of the grid you are physically connected to. It answers the question "what did the power system emit to serve loads like mine?" It ignores contracts entirely. It is the figure that moves when the grid decarbonises, and it is the one a policy analyst cares about.
The market-based figure reflects contractual instruments: renewable energy certificates, guarantees of origin, power purchase agreements, and supplier-specific rates. kWh covered by retired certificates are valued at that instrument's rate — typically zero for a renewable REC. kWh not covered are valued at the residual mix, which is the grid pool after everyone else's claimed clean generation has been removed from it. That is why the residual factor is higher than the grid average, and why a company holding no certificates at all can report market-based emissions above its location-based ones.
Certificates cannot create negative emissions. Retiring more certificates than you consumed does not push your figure below zero; the surplus is simply unused. This calculator enforces that floor.
Line losses are neither. Electricity lost in transmission and distribution — around 5 percent on the US grid as a whole — is generated but never delivered. Under the GHG Protocol those losses fall in Scope 3 category 3 for the end consumer, not in Scope 2. The calculator shows the grossed-up figure so you can report it as its own line, and you should not fold it into your Scope 2 total.
Worked example: a small office using 120,000 kWh a year
An office consumes 120,000 kWh in the reporting year. Its eGRID subregion publishes 950 lb CO2e/MWh. Grid losses are 4.7 percent. The company bought and retired 40,000 kWh of RECs, and the residual mix factor for its region is 1,050 lb CO2e/MWh.
- Convert the grid factor. 950 ÷ 1,000 = 0.95 lb per kWh, and 0.95 × 0.45359237 = 0.430913 kg per kWh.
- Location-based emissions. 120,000 × 0.430913 = 51,709.5 kg, or 51.71 tonnes CO2e. This is the Scope 2 location-based figure.
- Electricity generated on its behalf. 120,000 ÷ (1 − 0.047) = 125,918.2 kWh.
- Emissions including losses. 125,918.2 × 0.430913 = 54,259.7 kg. The 2,550 kg difference is the Scope 3 category 3 line, reported separately.
- kWh not covered by certificates. 120,000 − 40,000 = 80,000 kWh.
- Residual mix intensity. 1,050 ÷ 1,000 × 0.45359237 = 0.476272 kg per kWh.
- Market-based emissions. 80,000 × 0.476272 = 38,101.8 kg, or 38.10 tonnes.
The disclosure therefore reads: Scope 2 location-based 51.71 tCO2e, Scope 2 market-based 38.10 tCO2e, Scope 3 (T&D losses) 2.55 tCO2e. Three numbers, three different questions answered, none of them substitutable for another.
Carbon intensity conversions and the emissions of 1,000 kWh
| lb CO₂e/MWh | kg CO₂e/MWh | g CO₂e/kWh | kg CO₂e per 1,000 kWh | Typical generation mix |
|---|---|---|---|---|
| 100 | 45.4 | 45.4 | 45.4 | Overwhelmingly hydro and nuclear |
| 200 | 90.7 | 90.7 | 90.7 | Very low-carbon grid |
| 400 | 181.4 | 181.4 | 181.4 | Substantial nuclear or renewables |
| 600 | 272.2 | 272.2 | 272.2 | Mixed, gas-leaning |
| 800 | 362.9 | 362.9 | 362.9 | Near the US national average |
| 1,000 | 453.6 | 453.6 | 453.6 | Predominantly gas |
| 1,200 | 544.3 | 544.3 | 544.3 | Gas with significant coal |
| 1,600 | 725.7 | 725.7 | 725.7 | Coal-dominated |
Kilograms per MWh and grams per kWh are numerically identical, which is a useful shortcut. Multiply the lb/MWh figure by 0.45359237 to reach either. The mix descriptions are indicative of what produces that intensity, not a claim about any specific region.
How to read the result
Intensity is the number to track over time. Total emissions fall when you use less electricity and when the grid cleans up, and those two causes have completely different management implications. Reporting kWh, intensity and total emissions separately lets a reader see which one moved.
Check the gap between your two Scope 2 figures. If market-based sits far below location-based, that gap is carried entirely by contractual instruments, and a reader will reasonably ask what those instruments are and whether they caused any new generation to exist. Unbundled certificates from existing plants are the weakest version of that claim; a long-term power purchase agreement that financed a new project is the strongest.
Grossed-up emissions belong in their own line. The T&D figure is useful and often overlooked, but folding it into Scope 2 makes your total incomparable with everyone else's. Report it as Scope 3 category 3.
Use the gasoline equivalence only for communication. Dividing by 8.887 kg per gallon converts a tonnage into something people can picture. It is a presentation device, not an accounting one, and it should never appear in the inventory itself. The driving CO2 calculator is where that factor properly belongs.
Mistakes that show up in real inventories
- Using a national average when a subregional factor exists. The single largest error available in this calculation, and the easiest to fix.
- Confusing units. lb/MWh, kg/MWh and g/kWh all appear in published sources, and kg/MWh and g/kWh are numerically equal while lb/MWh is 2.2 times larger. A factor of 2.2 error is common and invisible.
- Reporting only the market-based figure. The GHG Protocol requires both. Publishing only the lower one is a disclosure failure, not a rounding choice.
- Claiming negative emissions from surplus certificates. The floor is zero. Surplus instruments are unused, not a credit.
- Applying the previous year's grid factor. eGRID is revised as the generation mix changes, and factors have moved substantially in recent years. State the vintage of the factor you used.
- Mixing a CO2-only factor with a CO2e total. eGRID publishes both. The difference is small for electricity but it is not zero, and consistency matters more than the magnitude.
- Double-counting on-site generation. Solar you generate and consume behind the meter never appears on your bill, so it is already excluded from your consumption. Do not also subtract it.
Key terms
- Scope 2
- Indirect emissions from purchased electricity, steam, heating and cooling consumed by the reporting organisation.
- Location-based method
- Values electricity at the average emission intensity of the grid where consumption occurs, ignoring contracts.
- Market-based method
- Values electricity according to the contractual instruments the consumer has purchased and retired, with uncovered volume valued at the residual mix.
- Residual mix
- The generation pool remaining after all contractually claimed clean generation has been removed. Always at least as carbon-intensive as the grid average.
- REC
- Renewable Energy Certificate — a tradable instrument representing the environmental attributes of one MWh of renewable generation, which must be retired to be claimed.
- eGRID
- EPA's Emissions & Generation Resource Integrated Database, the standard US source for subregional electricity emission rates.
Where this calculation sits
For a household, this is one line of the total — usually the largest or second largest — and it feeds directly into the household carbon footprint calculator, which uses exactly the same grid factor and conversion. For an organisation, it is the whole of Scope 2 for electricity, sitting alongside Scope 1 fuel combustion and the many categories of Scope 3.
Two adjacent calculations are worth doing at the same time. First, convert your electricity into site energy in MMBtu with the fuel BTU content converter so you can add it to gas and oil for a single energy-use intensity figure. Second, if you are considering electrifying a fossil-fuel load, model the emissions on both sides before and after: moving a heating or hot-water load onto electricity, as the water heating cost calculator compares on price, also moves it onto a factor that falls every year as the grid changes, while a therm of gas will emit 5.30 kg forever.
One caveat about averages. Every factor discussed here is an annual average. The real carbon intensity of a grid varies hour to hour, sometimes by more than a factor of two between a windy night and a still evening peak. Marginal emissions — what an extra kilowatt-hour actually causes to be generated right now — can differ substantially from the average and are what matter for load-shifting decisions such as when to charge a vehicle or run a battery. Average factors are the right tool for inventory accounting and the wrong tool for scheduling.