How Carbon Accounting Works | CO₂e, Scope 1, Scope 2, Scope 3 and MRV

A carbon number can look extraordinarily precise. 12,438 tonnes CO₂e. It can sit in a sustainability report, a procurement tender, an airline fuel certificate or the dashboard of a factory. It can be compared year to year and translated into targets, prices and investment decisions.

But the number is not measured directly in the same way as a thermometer measures temperature. It is assembled.

Fuel burned, electricity purchased, kilometres travelled, tonnes of steel bought, refrigerant leaked, methane released, fertiliser applied, products sold and carbon stored are translated through emission factors, physical models, allocation rules and accounting boundaries into a common greenhouse-gas inventory.

Carbon accounting is the system used to quantify, classify, aggregate, report and verify greenhouse-gas emissions and removals across a defined organisation, product, activity, project or value chain.

Its job is not to make carbon physical. Carbon is already physical. Its job is to make unlike greenhouse-gas flows comparable, traceable and decision-useful.

Wait, what? Carbon accounting is not the same thing as measuring carbon

Some quantities can be measured directly. A gas analyser can measure carbon-dioxide concentration. A flow meter can measure fuel volume. A utility meter can measure electricity. A mass-flow meter can measure injected CO₂.

But a corporate greenhouse-gas inventory asks a larger question:

How much climate-relevant greenhouse-gas impact should be attributed to this organisation, product or activity during this reporting period?

Answering that requires accounting rules:

Which facilities are inside the organisational boundary?
Which emissions belong in Scope 1, Scope 2 or Scope 3?
Which electricity factor should be used?
How should renewable-energy contracts be represented?
How should methane and nitrous oxide be converted into CO₂-equivalent?
Which supplier data are sufficiently reliable?
How are removals reported?
What happens when a company acquires another company halfway through the year?
How are estimates revised when better data appear?

The inventory is therefore a model of physical emissions under explicit rules.

The direct answer

Most greenhouse-gas accounting follows a recurring architecture:

define the reporting entity and boundary;
identify emission and removal sources;
collect activity data;
apply emission factors or physical measurement;
convert greenhouse gases into a common CO₂-equivalent basis where appropriate;
classify results into scopes, lifecycle stages or other reporting categories;
aggregate and reconcile the inventory;
document assumptions and uncertainty;
compare with a base year or target pathway;
subject important claims to quality control and, where required, independent assurance.

The result is useful only if the boundary, methods and data remain visible. A carbon number without provenance is just arithmetic.

Why CO₂e exists

Carbon dioxide is not the only greenhouse gas.

Greenhouse-gas inventories can include methane, nitrous oxide, hydrofluorocarbons and other gases covered by the chosen reporting standard.

Different gases affect climate differently and persist for different lengths of time. Accounting therefore converts them into a common unit called carbon dioxide equivalent, written CO₂e.

A simplified calculation is:

mass of greenhouse gas × selected global-warming-potential factor = CO₂e.

The selected factor and time horizon matter. A methane tonne is not physically transformed into carbon dioxide; CO₂e is a comparison unit.

Activity data × emission factor

The most familiar carbon-accounting equation is simple:

Activity data × emission factor = estimated emissions.

Examples:

litres of diesel × kg CO₂e per litre;
kilowatt-hours of electricity × kg CO₂e per kWh;
tonnes of cement purchased × kg CO₂e per tonne;
passenger-kilometres flown × emissions factor per passenger-kilometre;
kilograms of refrigerant leaked × greenhouse-gas factor.

This equation is powerful because it turns operational records into an emissions inventory.

It is also dangerous when users forget that the result inherits the quality of both terms.

Primary data versus secondary data

Primary data come directly from the activity being accounted for.

metered electricity;
fuel purchase records;
factory process measurements;
supplier-specific product data;
measured refrigerant top-ups;
verified CO₂ injection meters.

Secondary data come from databases, averages, industry studies, national statistics or generic emission factors.

Primary data can improve specificity, but “primary” does not automatically mean “good”. A badly calibrated meter can be worse than a robust industry average. Data quality depends on completeness, representativeness, measurement control and method consistency.

Emission factors

An emission factor translates an activity into expected greenhouse-gas emissions.

Some factors are rooted closely in chemistry. Burning a known mass of carbon creates a predictable mass of carbon dioxide if combustion is complete.

Others are statistical averages. An electricity-grid factor can represent the average emissions associated with generating a kilowatt-hour across a geographic system during a particular period.

Still others are lifecycle factors that include upstream extraction, manufacturing or transport.

The phrase “emission factor” therefore hides many different boundaries.

A factor must match the activity boundary

Suppose a company records 1,000 litres of diesel.

One factor might include only carbon dioxide released at combustion.

Another might include upstream oil extraction, refining and fuel distribution.

Multiplying the same 1,000 litres by these different factors answers different questions.

A credible inventory must know which question it is answering.

Organisational boundary

Before counting emissions, a company must decide what the company means for accounting purposes.

Large organisations can contain subsidiaries, joint ventures, franchises, leased buildings and minority investments.

Different consolidation approaches can attribute emissions according to equity share or control, depending on the reporting framework and purpose.

The boundary matters because two companies can participate in the same asset without both counting 100% of the same direct emissions in the same inventory category.

Operational boundary

Once the organisation is defined, emissions are classified according to their relationship to the organisation.

The Greenhouse Gas Protocol’s corporate framework is widely known for three scopes:

Scope 1: direct emissions from sources owned or controlled by the reporting company under the chosen consolidation approach;
Scope 2: indirect emissions associated with purchased or acquired electricity, steam, heat or cooling;
Scope 3: other indirect value-chain emissions upstream and downstream.

The scopes are accounting categories, not physical types of molecule.

Scope 1

Scope 1 covers direct emissions from controlled sources.

Examples can include:

fuel burned in company boilers;
fuel burned in company vehicles;
process CO₂ released from cement calcination;
methane released from operations;
refrigerants leaking from company-owned cooling systems.

Scope 1 is often the easiest scope to connect to physical plant operations because the reporting entity controls the emitting source.

Scope 2

Scope 2 covers emissions associated with purchased or acquired energy such as electricity, steam, heating and cooling.

The reporting company may not operate the power station, but its energy consumption creates demand in the energy system.

Electricity accounting is difficult because electrons mix on a grid while commercial contracts can identify different generation attributes.

That tension produces two major approaches in current GHG Protocol practice: location-based and market-based reporting where the latter applies.

Location-based electricity accounting

Location-based accounting uses an emissions factor representing the electricity system serving the location, commonly a grid-average or other geographically defined factor.

The question is:

What is the emissions intensity of electricity physically supplied in this grid region?

This approach reflects the local generation mix but may not reflect a company’s contractual electricity purchases.

Market-based electricity accounting

Market-based accounting uses qualifying contractual instruments and supplier information under the applicable rules to represent purchased electricity attributes.

The question becomes:

What generation attributes has this company contractually acquired under the accounting system?

This can involve power-purchase agreements, energy-attribute certificates, supplier products and residual-mix calculations depending on the market.

Market-based accounting is not permission to treat accounting attributes as literal electron tracing. It is an accounting representation of contractual energy attributes.

Why Scope 2 is being revised

Electricity systems have changed dramatically since the original Scope 2 Guidance was published. Renewable procurement has expanded, storage has grown, grids have become more granular, and stakeholders increasingly ask whether annual certificate matching reflects the time and place in which electricity is consumed.

As of September 2026, GHG Protocol has published feedback from its Scope 2 public consultation and continues work on updated corporate accounting rules. It is also coordinating Scope 2 with its Actions and Market Instruments workstream and the broader harmonisation programme with ISO.

That matters because electricity accounting is no longer a side calculation. For many companies, Scope 2 is one of the main ways procurement decisions become climate claims.

Residual mix

If some electricity users claim specific renewable attributes, the remaining grid mix should not allow the same attributes to be claimed again by everyone else.

Residual-mix accounting attempts to remove explicitly claimed attributes from the pool used for untracked electricity.

The principle is simple:

one environmental attribute should not support two incompatible claims.

Temporal matching

Annual electricity accounting can match a company’s annual consumption with renewable generation even when generation and consumption occurred at different hours.

Tighter temporal matching asks whether low-carbon generation occurred at the same time as consumption.

Hourly matching can improve physical representativeness, but it increases data requirements and can raise procurement cost.

This is a recurring carbon-accounting theme: greater precision can improve decision quality while increasing complexity.

Geographic matching

Electricity cannot move without limit across congested grids.

A renewable certificate produced far away does not necessarily describe the marginal generation serving a constrained load centre.

Geographic rules therefore attempt to keep contractual claims connected to a physically meaningful market or grid region.

Scope 3

Scope 3 includes other indirect emissions across the value chain.

This can include:

purchased goods and services;
capital goods;
fuel- and energy-related activities not already included in Scopes 1 or 2;
upstream transport;
waste;
business travel;
employee commuting;
leased assets;
downstream transport;
processing and use of sold products;
end-of-life treatment;
franchises;
investments.

For many companies, Scope 3 is much larger than Scopes 1 and 2 because most physical activity occurs in suppliers and customers rather than inside company-owned facilities.

Scope 3 is difficult because control becomes weaker as the boundary expands

A company can meter its own boiler closely.

It may know far less about the factory making a purchased component three tiers up the supply chain.

Scope 3 inventories therefore rely on a hierarchy of data:

supplier-specific measured data;
supplier-specific lifecycle factors;
industry-average physical factors;
spend-based economic factors;
modelled estimates.

The wider the value chain, the more important data-quality scoring becomes.

Spend-based accounting

Spend-based methods estimate emissions from money spent multiplied by an economic emissions factor.

They are useful for screening because financial data are usually available.

They are weaker for operational improvement because spending can change due to inflation or price without physical emissions changing proportionally.

A company can appear to reduce emissions simply because a product becomes cheaper.

Spend methods are therefore often a starting point, not an ideal end state for high-impact categories.

Supplier-specific data

Supplier-specific data can reveal real differences between two products that generic industry averages hide.

But the buyer must ask:

What boundary did the supplier use?
Which year?
Which allocation method?
Which electricity accounting method?
Was the data independently assured?
Does it describe the exact product or the whole factory?

More specific data are only better if they are comparable.

Double counting across scopes is not automatically an error

The same physical emission can appear in different companies’ inventories under different scopes.

A power station counts combustion emissions in its Scope 1. A customer can report associated electricity emissions in Scope 2. A product manufacturer may include the same electricity contribution in a product footprint sold downstream.

This is intentional because the inventories answer different responsibility questions.

The mistake would be adding all companies’ scopes together as though they formed one non-overlapping national inventory.

Corporate inventory versus national inventory

National greenhouse-gas inventories generally organise emissions territorially and by economic sector under different accounting rules.

Corporate inventories organise emissions around organisational and value-chain relationships.

Both can be valid while producing different totals and classifications.

Accounting systems should therefore never be compared without checking their boundaries.

Product carbon footprint

A product footprint asks how much greenhouse-gas impact is associated with one product across a defined lifecycle.

The GHG Protocol Product Standard supports lifecycle accounting across stages such as:

raw materials;
manufacturing;
transport;
use;
end-of-life.

As of 2026, GHG Protocol and ISO are also working toward harmonised product-level greenhouse-gas accounting, reflecting growing demand for comparable product carbon data across supply chains.

Cradle-to-gate

A cradle-to-gate footprint follows a product from raw-material extraction through production until it leaves the factory gate.

It excludes downstream use and disposal unless the chosen method says otherwise.

This boundary is useful for intermediate materials such as steel, cement or chemicals that become inputs to other products.

Cradle-to-grave

A cradle-to-grave footprint extends through product use and end-of-life.

This is important when the use phase dominates.

A washing machine’s operating electricity can matter more than manufacturing. A combustion vehicle’s fuel use can dominate its lifecycle. A passive building material may have most emissions upfront.

The boundary should match the decision.

Functional unit

Product comparisons need a functional unit.

Comparing “one kilogram of insulation” can be misleading if different materials provide different thermal performance.

A better functional unit might be:

insulation required to deliver a specified thermal resistance over a specified area for a specified life.

The functional unit converts product comparison into service comparison.

Allocation

Many industrial processes produce more than one product.

A refinery produces petrol, diesel, jet fuel and other products. A dairy system produces milk and meat. A chemical plant can produce co-products.

The total emissions must somehow be allocated among outputs unless the method uses system expansion or another approach.

Possible allocation bases include:

mass;
energy content;
economic value;
physical causality.

Different valid allocation rules can produce different product footprints. The rule must therefore be disclosed.

Recycling allocation

Recycling creates a boundary problem.

If aluminium is recycled, who receives the benefit—the product supplying scrap or the product using recycled aluminium?

Different methods distribute burdens and credits differently.

This is another reason product footprints should not be compared casually unless methodology is aligned.

Carbon intensity

Absolute emissions tell us how large a total footprint is.

Carbon intensity divides emissions by a unit of output or service.

Examples:

kg CO₂e per kWh;
kg CO₂e per tonne of steel;
g CO₂e per passenger-kilometre;
kg CO₂e per square metre of floor area;
tonnes CO₂e per million dollars of revenue.

Intensity is useful for efficiency and benchmarking. It can also hide growth.

A company can reduce emissions per product while selling so many more products that total emissions rise.

Absolute emissions and intensity answer different questions

Absolute emissions ask:

How much greenhouse gas entered the accounting boundary?

Intensity asks:

How much greenhouse gas was associated with each unit of output or service?

A serious climate strategy often needs both.

Base year

Targets need a reference point.

A base year establishes the historical inventory against which future performance is measured.

But organisations change. They acquire companies, sell divisions, outsource activities and improve data.

Accounting standards therefore define circumstances in which historical inventories should be recalculated so performance reflects real change rather than structural accounting change.

Structural changes

Suppose Company A buys Company B.

Its emissions rise immediately even if neither factory becomes less efficient.

Without base-year recalculation, the acquisition can look like climate deterioration.

Similarly, selling a high-emission factory can make emissions fall even if the factory continues operating under another owner.

Accounting must distinguish operational decarbonisation from changes in organisational perimeter.

Inventory reduction versus avoided emissions

This is one of the most important distinctions in carbon accounting.

An inventory reduction occurs when emissions inside the defined inventory fall.

Avoided emissions estimate the emissions difference between a real activity and a hypothetical alternative.

Example:

A company sells efficient air conditioners.

The company’s Scope 3 inventory can still be large because customers use electricity. At the same time, those products may avoid emissions relative to less efficient competitors.

Both statements can be true.

The avoided-emissions claim should not simply be subtracted from the corporate inventory unless the applicable accounting framework explicitly permits that treatment.

Consequential versus attributional accounting

Attributional accounting allocates a share of existing emissions to a product, company or activity.

Consequential analysis asks how total system emissions change because of a decision.

These approaches can produce different answers because they ask different questions.

Corporate inventories are primarily attributional. Policy analysis, marginal electricity studies and avoided-emissions claims can be consequential.

Confusion begins when one type of answer is presented as though it were the other.

Marginal electricity emissions

An average grid factor describes the average emissions intensity of generation over a boundary and period.

A marginal factor estimates how emissions change when electricity demand changes.

For operational decisions such as flexible electrolysis, battery dispatch or load shifting, marginal emissions can reveal effects that annual average accounting cannot.

But marginal analysis should not be substituted silently into a corporate inventory that requires another method.

Lifecycle emissions

Lifecycle accounting follows emissions beyond the point of direct use.

For a fuel, this can include:

resource extraction;
processing;
fuel production;
transport;
storage;
combustion or end use.

This is why a fuel can have zero carbon dioxide at the exhaust and still have high lifecycle emissions upstream.

Hydrogen, ammonia and synthetic fuels all require lifecycle accounting to distinguish genuinely low-emission pathways from relabelled upstream emissions.

Well-to-tank, tank-to-wheel and well-to-wheel

Transport fuels often use nested lifecycle boundaries.

well-to-tank: upstream production and delivery;
tank-to-wheel: vehicle-use emissions;
well-to-wheel: the combined chain.

Shipping commonly uses well-to-wake. Aviation can use well-to-wing.

The language changes by sector, but the accounting purpose is the same: prevent upstream and downstream stages from disappearing outside the boundary.

Embodied carbon

Embodied carbon generally refers to greenhouse-gas emissions associated with producing materials, products and construction before or apart from their operational use.

In buildings, operational energy can fall as grids decarbonise, making embodied emissions from steel, cement, glass and construction increasingly important.

Whole-life accounting prevents a low-energy building from appearing carbon-free if enormous emissions were created during construction.

Biogenic carbon

Carbon from biomass needs special treatment because it was recently part of the atmospheric-biological cycle rather than geological fossil storage.

But “biogenic” does not mean impact-free.

Accounting can need to consider:

land-use change;
forest-carbon change;
harvest;
regrowth;
soil carbon;
processing emissions;
storage in products;
timing of uptake and release.

The 2026 GHG Protocol Land Sector and Removals Standard was created partly because companies needed more consistent treatment of these land and removal flows.

Removals in corporate inventories

Carbon dioxide removals are not simply negative emissions typed into a spreadsheet.

A credible removal claim needs evidence that atmospheric CO₂ was removed, stored, attributable to the reporting entity under the applicable method, and reported with the required safeguards.

GHG Protocol’s Land Sector and Removals Standard, published in January 2026 and effective from 1 January 2027, establishes requirements for land emissions, removals and certain technological removals and geologic storage in corporate inventories.

The physical removal owner remains How Carbon Dioxide Removal Works. This article owns how such flows enter an accounting system.

Captured CO₂ with geologic storage

Accounting needs to distinguish several quantities:

CO₂ captured;
CO₂ transported;
CO₂ injected;
CO₂ verified as stored;
residual source emissions;
capture-energy emissions;
transport emissions;
storage-system emissions.

The physical owners remain Carbon Capture, CO₂ Transport and Geological Carbon Storage.

Carbon accounting asks how those measured flows become reportable inventory entries without double counting or losing lifecycle emissions.

Gross capture versus avoided emissions

If one million tonnes of CO₂ are captured but extra energy creates 150,000 tonnes CO₂e elsewhere, the project has not necessarily avoided one million tonnes.

Gross capture is a physical flow.

Avoided emissions are a net comparison against a defined baseline.

Accounting should preserve the distinction.

Carbon intensity of hydrogen

Two kilograms of hydrogen can be chemically identical and have very different greenhouse-gas intensities.

One may be produced from low-carbon electrolysis. Another may come from natural gas with methane leakage and no capture. Another may use fossil gas with high capture but residual combustion emissions.

The molecule cannot reveal its production history.

Carbon accounting provides the missing provenance.

Carbon intensity of electricity

Electricity can be described using several emissions intensities:

annual grid average;
hourly grid average;
supplier-specific factor;
contractual market-based factor;
marginal emissions factor;
lifecycle generation factor.

These are not interchangeable. A decision model should choose the factor that answers the question being asked.

Carbon intensity of fuels

Fuel carbon intensity can be expressed per unit energy, mass, volume or final service.

For transport policy, g CO₂e per megajoule can compare fuel pathways. For vehicle performance, g CO₂e per kilometre may be more useful. For aviation, emissions per passenger-kilometre or tonne-kilometre may better reflect service.

The denominator changes the meaning.

Book-and-claim

Some low-carbon commodities are scarce and unevenly distributed.

A book-and-claim system can separate the environmental attribute from the physical molecule or unit of fuel under defined registry rules.

This can help early markets scale without requiring every buyer to receive the exact low-carbon molecule physically.

The accounting risk is double claiming. A credible system needs unique issuance, tracking and retirement of attributes.

Mass balance

Mass-balance systems allow certified and conventional materials to mix physically while environmental attributes are allocated according to defined rules.

This is common in complex chemical and fuel supply chains where dedicated pipes and tanks for every certified molecule would be impractical.

The physical mass balance and the attribute ledger must reconcile.

Chain of custody

Chain of custody records how a product or environmental attribute moves from producer to buyer.

A strong chain-of-custody system answers:

who created the attribute;
what methodology was used;
what quantity was issued;
who transferred it;
whether it was split or combined;
who ultimately retired it;
whether another party is claiming the same benefit.

Accounting integrity increasingly depends on digital evidence because physical molecules can mix while claims remain distinct.

MRV

MRV usually means measurement, reporting and verification, though specific programmes may define terms more precisely.

Measurement establishes the physical or estimated quantity.

Reporting applies the required classification and disclosure rules.

Verification or assurance tests whether the reported information is materially accurate under the stated criteria.

MRV turns an internal spreadsheet into an accountable information system.

Measurement hierarchy

Where practical, accounting can move from rough estimates toward better evidence:

spend estimates;
industry-average physical factors;
supplier-specific factors;
metered activity data;
continuous emissions measurement or verified process data.

The correct endpoint depends on materiality and decision need. It is inefficient to install laboratory-grade measurement on an immaterial category while leaving the largest source as a rough estimate.

Materiality

Materiality helps organisations focus assurance and data-improvement effort on errors large enough to affect decisions.

Materiality does not mean small sources are automatically omitted. The applicable standard defines completeness expectations.

It means the audit process recognises that not every decimal place carries equal decision importance.

Uncertainty

Every inventory contains uncertainty.

Sources include:

meter accuracy;
sampling error;
emission-factor uncertainty;
supplier estimates;
model assumptions;
allocation choices;
missing data;
currency-to-spend-factor conversion;
global-warming-potential selection.

Uncertainty should not be used as an excuse to avoid accounting. It should be made visible and reduced where it matters.

Precision versus accuracy

A spreadsheet can report 12,438.27 tonnes CO₂e even when the underlying emission factor has ±20% uncertainty.

That is numerical precision without equivalent physical accuracy.

Good reporting avoids creating false confidence through unnecessary decimal places.

Data lineage

Data lineage records where each number came from and how it was transformed.

A strong inventory can trace:

source invoice or meter → activity record → unit conversion → emission factor → calculation → scope/category → reported total.

This makes corrections possible. Without lineage, a company may know the final number but not how to repair it.

Controls

Carbon accounting needs internal controls similar to financial and operational reporting.

approved factor libraries;
version control;
unit validation;
duplicate detection;
reasonableness tests;
segregation of duties;
documented recalculation policies;
review and sign-off.

The largest carbon-accounting errors are often mundane: wrong units, duplicated invoices, missing months or obsolete factors.

Assurance

Independent assurance evaluates whether reported greenhouse-gas information conforms to stated criteria and is free from material misstatement at the relevant assurance level.

Assurance can examine:

boundaries;
source completeness;
data controls;
calculation methods;
emission factors;
evidence;
recalculations;
claims.

Assurance does not make a chosen accounting rule physically perfect. It tests whether the organisation followed the rule and whether the resulting statement is materially reliable.

The ISO–GHG Protocol harmonisation

Greenhouse-gas accounting has historically developed through overlapping standards and programmes.

In July 2026, GHG Protocol announced that it and the International Organization for Standardization would combine their corporate carbon-accounting standards into a single harmonised global corporate standard.

GHG Protocol says the consolidated work is intended to harmonise its Corporate, Scope 1, Scope 2, Scope 3 and Actions and Market Instruments workstreams with ISO 14064-1.

This matters because companies increasingly report into multiple regulatory and voluntary systems. Harmonisation can reduce contradictory methodologies and improve interoperability—although detailed technical decisions still need to be developed through the standards process.

Standards evolve because the energy system evolves

A carbon-accounting rule created for a fossil-heavy grid may struggle with hourly renewable procurement, batteries, green hydrogen, carbon removals and digital certificate systems.

Standards therefore need periodic revision.

The challenge is stability versus relevance:

stable rules improve comparability across years;
updated rules improve physical and market relevance.

Good governance manages both.

Worked example 1: a tuition centre

A small education business rents classrooms and operates no industrial equipment.

Its direct Scope 1 emissions may be very small.

Its inventory can still include:

purchased electricity in Scope 2;
paper and devices in Scope 3;
business travel;
employee commuting if included under the relevant category;
cloud and digital services through purchased-services categories;
leased-asset treatment depending on organisational boundary.

The example shows that carbon accounting is not only for smokestacks. Every organisation participates in physical supply chains.

Worked example 2: manufacturing company buys renewable electricity

A factory consumes 100 GWh of electricity.

Its location-based Scope 2 reflects the relevant grid factor.

It also acquires qualifying renewable-energy attributes under the applicable accounting rules, producing a different market-based result.

The company should report the methods required by the applicable standard rather than choosing only the number that looks better.

The two results answer different questions.

Worked example 3: steel buyer

A construction company purchases 10,000 tonnes of steel.

At first it uses an industry-average factor.

Later, suppliers provide verified product-specific carbon data.

The company can now distinguish high- and low-emission steel rather than treating every tonne identically.

Accounting becomes a procurement signal.

Worked example 4: hydrogen

Two suppliers deliver chemically identical hydrogen.

Supplier A uses low-carbon electrolysis.

Supplier B uses natural gas, captures part of process CO₂, but has upstream methane leakage and residual combustion emissions.

A kilogram-only procurement specification cannot distinguish them.

A verified kg CO₂e per kg H₂ metric can.

Worked example 5: synthetic aviation fuel

An airline buys synthetic aviation fuel made with low-carbon hydrogen and captured carbon dioxide.

The fuel’s combustion still releases CO₂.

The lifecycle claim depends on carbon origin, electricity, hydrogen production, synthesis, transport and chain of custody.

The physical owner remains How Synthetic Fuels Work. Carbon accounting provides the evidence that lets the fuel’s production history travel with the commercial product.

Worked example 6: carbon capture plant

A cement plant captures 900,000 tonnes of CO₂.

Compression electricity, solvent replacement and upstream energy create 70,000 tonnes CO₂e. The plant still vents 100,000 tonnes directly.

Reporting only “900,000 tonnes captured” does not describe the full climate result.

A complete inventory records gross capture, residual emissions, energy emissions, transport/storage flows and the appropriate baseline comparison.

Worked example 7: Scope 3 data improvement

A retailer initially estimates purchased-goods emissions using spend-based factors.

The category dominates its footprint.

It therefore asks its largest suppliers for product-level physical data and verified factors.

Within three years, most of the high-impact category moves from spend estimates to supplier-specific information.

The inventory becomes more useful not because the total necessarily falls, but because the signal becomes more actionable.

Failure mode: treating every carbon number as comparable

One company reports location-based electricity emissions. Another reports market-based electricity emissions. Their totals are compared directly without noting the method.

Repair: compare like boundary with like boundary and disclose accounting method.

Failure mode: mixing inventory reductions with avoided emissions

A company calculates the emissions customers might avoid by using its product and subtracts that estimate from Scope 1–3 emissions.

Repair: keep inventory accounting and comparative impact claims separate unless a standard explicitly defines the combination.

Failure mode: spend-based Scope 3 becomes permanent

A company screens its supply chain using spend factors but never improves data for the largest categories.

Repair: use screening to prioritise supplier-specific and physical data improvement.

Failure mode: false precision

An inventory reports emissions to two decimal places even though major categories use ±30% estimates.

Repair: align displayed precision with data quality and disclose uncertainty.

Failure mode: old emission factors

A company uses an electricity factor from five years ago even though the grid has changed substantially.

Repair: control factor versions and update them on a defined schedule.

Failure mode: duplicate energy attributes

A renewable certificate is claimed by a buyer while the same renewable attribute remains embedded in another supplier’s factor.

Repair: use residual-mix and certificate-retirement systems that prevent incompatible double claims.

Failure mode: a supplier-specific number with an unknown boundary

A supplier says its product emits 0.8 kg CO₂e per unit but does not disclose whether transport, capital goods or use-phase emissions are included.

Repair: require methodology, functional unit, boundary, year and verification information.

Failure mode: acquisitions create fake emissions growth

A company buys a large factory and reports emissions growth as operational deterioration.

Repair: apply the relevant base-year recalculation policy.

Failure mode: removals used as an eraser

A company combines gross emissions and low-durability removals into one small “net” number while hiding both components.

Repair: disclose gross emissions, gross removals, storage type, durability and netting rules separately.

Failure mode: one Scope 3 factor creates the strategy

An uncertain industry-average factor makes one supplier category look dominant, so the company reorganises procurement without validating the data.

Repair: improve data quality before making high-cost decisions when rankings are uncertainty-sensitive.

Failure mode: accounting replaces physics

A contractual renewable claim leads managers to ignore when and where electricity is physically consumed.

Repair: use inventory accounting for reporting and additional operational metrics for physical decarbonisation decisions.

Common misconceptions

CO₂e is a comparison unit, not a physical gas.
Scope 1, 2 and 3 are accounting categories, not different kinds of emissions.
The same physical emission can legitimately appear in different companies’ scopes.
Corporate inventories should not be summed casually into a national total.
Location-based and market-based electricity numbers answer different questions.
Scope 3 estimates are not automatically less useful, but data quality varies widely.
Spend-based factors are useful for screening but weak for detailed operational decisions.
Product carbon footprints depend on functional unit, boundary and allocation rules.
Absolute emissions can rise while carbon intensity falls.
Avoided emissions are not the same as inventory reductions.
Gross carbon capture is not the same as net avoided emissions.
Carbon removals require specific accounting safeguards and cannot simply be typed as negative emissions.
Chain-of-custody systems matter because low-carbon attributes can be invisible in chemically identical products.
Assurance tests conformity and evidence; it does not make every accounting rule physically perfect.
More decimal places do not create more accuracy.

A universal carbon-accounting audit

Define the purpose of the inventory.
Choose the governing standard or methodology.
Define organisational boundary.
Define operational boundary.
Identify all material emission and removal sources.
Map each source to the required scope or lifecycle stage.
Collect activity data with documented lineage.
Choose emission factors whose boundaries match the activity.
Use current global-warming-potential factors required by the methodology.
Document units and conversions.
Separate measured from estimated data.
Score data quality.
Use supplier-specific data where material and reliable.
Apply Scope 2 method requirements consistently.
Prevent duplicate environmental-attribute claims.
Define product functional units before comparison.
Document allocation and recycling rules.
Separate inventory reductions from avoided-emissions claims.
Separate gross capture from net avoided emissions.
Report removals with storage and durability information.
Define base year and recalculation policy.
Quantify important uncertainty.
Run reasonableness and mass-balance checks.
Maintain factor and methodology version control.
Preserve evidence for assurance.
Report limitations transparently.
Use the inventory to improve decisions, not merely disclosure.

The deepest carbon-accounting principle

Carbon accounting can feel like an administrative layer sitting above real engineering.

It is more useful to see it as a translation layer.

Factories speak in tonnes, litres, kilowatt-hours, kilometres, pressures and temperatures.

Supply chains speak in invoices, products, contracts and shipments.

Climate strategy needs those physical and commercial activities translated into a common language without losing where the numbers came from.

That is what good carbon accounting does.

Its purpose is not to make a company look clean.

Its purpose is to make the carbon system visible enough that managers, engineers, buyers, regulators and investors can distinguish real change from accounting noise.

A carbon inventory is valuable when every important number can be traced back to a physical activity, a declared rule and evidence strong enough to survive scrutiny.

How Carbon Accounting fits the Energy series

How Carbon Capture Works owns source-side CO₂ separation. How Carbon Dioxide Transport Works owns CO₂ movement. How Geological Carbon Storage Works owns subsurface storage. How Carbon Dioxide Removal Works owns physical removal and durability. How Synthetic Fuels Work owns manufactured fuels and lifecycle pathways. This article owns the measurement and accounting layer that translates those physical systems into inventories, intensities, scopes, lifecycle footprints and auditable climate claims.

Carbon Markets remain a separate economic owner. Financial disclosure remains a Finance owner. Carbon accounting provides the measured substrate both need without re-owning their economic or reporting-governance jobs.

Current standards and further reading

The final lesson is simple: we cannot manage carbon intelligently if we cannot say what was counted, why it was counted, where the data came from and what the number actually means.

How Energy Works | Main Series

How Carbon Accounting Works | CO₂e, Scope 1, Scope 2, Scope 3, Lifecycle Emissions, Carbon Intensity and MRV