# How do I spot a wrong emission factor or unit error in a carbon report before it goes to a buyer or auditor?

Canonical: https://senseible.earth/answers/how-to-spot-a-wrong-emission-factor

Check four things on every large line: the unit of the activity matches the unit of the factor, the factor sits near its published value, the factor edition and gas basis are stated, and the result makes sense against production. Most serious errors are 1,000x unit slips or a factor for the wrong unit.

## The unit errors that cause the biggest mistakes

- kWh entered as MWh, or MWh as kWh: a 1,000x error either way.
- Litres against a per-kg factor, or kg against a per-litre factor: LPG is usually per kg, diesel per litre. Diesel weighs about 0.84 kg per litre, so the error is about 16% to 19%, small enough to slip through.
- Kilolitres read as litres on bulk fuel invoices: another 1,000x.
- kVAh units treated as kWh: overstates by roughly the gap between power factor and 1, about 5% at a power factor of 0.95.
- Gross versus net calorific value when a factor is per GJ: about 5% to 10% for most fuels.

## A quick range check for common factors

Keep a short list of published values and compare every factor in the report against it. Using the IPCC 2006 defaults: diesel about 2.68 kg CO2 per litre, petrol 2.31, LPG 2.98 kg CO2 per kg, coal 2.42 kg CO2 per kg. For grid power in India, the CEA CO2 Baseline Database version 21.0 gives 0.710 kg CO2 per kWh. A factor more than about 20% away from these needs a stated reason, such as a supplier-specific value or a different fuel grade.

## Edition, gas basis and scope errors

- Old edition: a grid factor from an earlier CEA database edition used for the current year. The report should name the version and fiscal year.
- CO2 versus CO2e: fuel and grid factors above are CO2 only. Adding CO2e factors for refrigerants or product footprints is fine, but label the total as CO2e and do not mix the two silently.
- GWP set: refrigerant CO2e changes between IPCC AR5 and AR6 values. One set should be used throughout.
- Market-based factor without a certificate: a zero factor for electricity needs a REC, green tariff or PPA document behind it.

## Sense checks against the business

1. Divide emissions by output, such as tonnes CO2 per tonne of product, and compare month to month. A sudden 10x jump is almost always an error.
2. Compare electricity kWh in the report with the total on the discom bills for the year.
3. Compare fuel litres with the purchase ledger in rupees, divided by a typical price per litre.
4. Search for lines with a quantity but no factor, or a factor but no unit.

## Worked example: a 1,000x unit slip

- Bill shows 42,000 kWh for the month
- Correct: 42,000 kWh x 0.710 kg CO2 per kWh = 29,820 kg, about 29.8 tonnes CO2
- Entered as 42,000 MWh: 42,000,000 kWh x 0.710 = 29,820,000 kg, about 29,820 tonnes CO2
- Intensity check: if monthly output is 500 tonnes of product, the error shows as 59.6 tonnes CO2 per tonne instead of 0.06, which no reviewer should accept

## What Senseible does here

Senseible runs rule checks on every line, with penalties for a missing unit (15), a missing factor (10) and an abnormal quantity (25), applies no grid factor to kVAh lines, and recomputes each figure within max(0.01, 0.1%). It cannot catch an error in a figure you typed in by hand from outside its documents.

- [Choosing emission factors](https://senseible.earth/climate-intelligence/emission-factors-selection-guidance)
- [Emission factor reference](https://senseible.earth/emission-factors)
- [Invoice in litres, factor in kg](https://senseible.earth/answers/invoice-unit-mismatch-litres-kg)
- [AR5 or AR6 GWP values](https://senseible.earth/answers/ar5-vs-ar6-gwp-which-to-use)

Sources: [IPCC 2006 Guidelines, Volume 2 Energy](https://www.ipcc-nggip.iges.or.jp/public/2006gl/vol2.html), [CEA CO2 Baseline Database](https://cea.nic.in/cdm-co2-baseline-database/?lang=en), [ISO 14064-1:2018, accuracy and consistency principles](https://www.iso.org)

Last reviewed 2026-10-15.