The GHG Protocol Corporate Standard gives companies two broad pathways for calculating emissions: spend-based estimation and activity-based calculation. Spend-based is faster to set up and requires less granular data. Activity-based takes more work but produces numbers that hold up under verification. Understanding the difference matters because many companies start with spend-based estimates and then discover, when a third-party verifier reviews the methodology, that the numbers are not traceable to source.
This article explains how the activity-based approach works in practice for a manufacturing context: what data it requires, where that data comes from, what you do with it, and where the boundaries of the approach lie.
The core formula
Activity-based GHG calculation follows a consistent structure across all emission sources:
Emissions = Activity Data x Emission Factor x Global Warming Potential
Activity data is the physical quantity of the emission-generating activity. Liters of diesel burned. Kilowatt-hours of electricity consumed. Tonnes of natural gas purchased. Kilograms of raw material processed.
The emission factor converts that physical quantity into a mass of greenhouse gas. For diesel combustion, the factor is in kilograms of CO2 per liter of diesel. For electricity, it is in kilograms of CO2 per kilowatt-hour consumed from the relevant grid.
The global warming potential (GWP) converts non-CO2 greenhouse gases to a CO2-equivalent (CO2e) basis, using the IPCC's assessed equivalence values. The IPCC AR6 values are the current standard reference.
The result is a figure in tonnes of CO2e, which is the unit of account for corporate GHG reporting under all major disclosure frameworks.
What makes it "activity-based" rather than "spend-based"
The distinction is in what you measure as the starting point. Spend-based methods use total financial expenditure in a procurement category multiplied by an economic-intensity factor expressed in kgCO2e per dollar of spend. The number this produces is statistically reasonable at a sector level but imprecise at the company level because it cannot distinguish between different suppliers, processes, or materials within the same spend category.
Activity-based methods measure a physical quantity tied to the actual emission event. The emission comes from burning fuel, not from the act of paying for it. The activity data is the quantity of fuel burned. This is traceable to source documents (fuel delivery invoices, utility bills, meter readings) and verifiable by an auditor in a way that a spend estimate is not.
The reason activity-based produces more defensible numbers is also the reason it demands more from your data infrastructure. You need physical quantities, not just financial totals. And you need them at a level of granularity that maps to appropriate emission factors.
Scope 1: combustion sources in a manufacturing context
For Scope 1 (direct emissions from company-owned or controlled sources), the activity data sources in a manufacturing operation typically include:
- Fuel delivery receipts for diesel, LPG, and fuel oil used in on-site equipment and generators
- Natural gas utility bills showing metered consumption in MJ or therms
- Internal fuel dispensing records for company vehicles on-site
- Boiler maintenance logs if utility bills do not capture full consumption
The key data point from each of these is the physical quantity in a standard unit: liters, therms, gigajoules, cubic meters. The supplier or utility typically provides this on the document, though the unit varies by country and energy type.
Matching these quantities to the correct emission factors requires knowing the fuel type precisely. Natural gas has a different emission factor from LPG. Diesel has a different factor from biodiesel blends. A generic "combustion" factor applied to all fuels will introduce error. The activity-based approach requires you to maintain fuel type records at the source document level.
Scope 2: purchased electricity
Scope 2 is methodologically simpler than Scope 1 but has its own wrinkle: the GHG Protocol requires disclosure under both the location-based method (using the average grid emission factor for your region) and the market-based method (using supplier-specific or contractual factors when applicable). If you purchase renewable energy certificates, the market-based factor may be substantially different from the location-based factor.
Activity data for Scope 2 is electricity consumption in kilowatt-hours from utility bills. For a multi-facility manufacturer, this means aggregating across every utility account for every facility in scope. The emission factor for the location-based method in Singapore is the national grid emission factor published by the Energy Market Authority; it is updated periodically and the reporting year factor applies to the relevant billing period.
The challenge here is not the calculation. It is ensuring complete coverage: no facility accounts missed, correct factor vintages applied to each billing period, and consistent treatment of any renewable energy purchases across the two reporting methods.
Scope 3: activity data requirements by category
Scope 3 is where the gap between spend-based and activity-based approaches has the most practical impact. For Category 1 (purchased goods), activity data is the physical quantity of each material purchased: tonnes of steel, liters of chemicals, kilograms of packaging. This comes from purchase order line items and supplier invoices, provided the invoices state quantities rather than only cost.
For Category 4 (upstream transportation), activity data is tonne-kilometers: total weight of goods transported multiplied by distance. Freight invoices sometimes contain this directly; sometimes you need to reconstruct it from weight and origin-destination data in logistics records.
For Category 5 (waste), activity data is waste mass by disposal route: tonnes to landfill, tonnes incinerated, tonnes recycled. Waste contractor invoices typically state this in weight, though the quality varies.
Each category requires different source documents and different data extraction steps. The activity-based approach for Scope 3 is not a single process; it is a collection of separate extraction workflows, one per relevant category, each with its own data source and factor matching logic.
Where estimates are still necessary
Even within an activity-based framework, not every data point is directly measurable. The GHG Protocol acknowledges that some activity data will need to be estimated using engineering assumptions or proxy data. If a small generator on-site has no fuel meter, you estimate consumption from hours of operation and the rated fuel consumption of the unit. If a Category 1 supplier invoice states value but not mass, you may estimate mass from known typical material costs per unit weight.
These estimates are acceptable under the standard provided they are documented, their uncertainty range is characterized, and they do not dominate the total where direct measurement is available. The verifier will ask which sources are measured and which are estimated, and will scrutinize the estimation methodology for reasonableness.
This is the honest limit of the activity-based approach: it produces better numbers than spend-based estimation, but it still involves judgment calls and estimation for sources where direct measurement is impractical. The goal is to maximize the proportion of directly measured activity data and document everything else with enough transparency that an external reviewer can assess the quality.
The data infrastructure implication
Running the activity-based approach year after year requires a consistent process for collecting source documents, extracting physical quantities, applying factors, and retaining the audit trail. The work is repeatable but not trivial, and the infrastructure has to be maintained from one reporting cycle to the next.
The biggest risk we see in practice is not calculation error. It is gaps: a utility account added mid-year that was not captured in the Scope 2 calculation, a new fuel type in the production process that does not match any existing factor mapping, a Scope 3 category that was scoped in one year and silently dropped the next. Systematic coverage checks are as important as calculation accuracy.
What Zevero does is build the data structure that connects source documents (invoices and meter reads) to the activity-based calculation chain, including factor matching and CO2e summation. The activity-based approach is methodologically correct for disclosure under GHG Protocol and any framework that endorses it. Getting there requires the document extraction and classification work to happen reliably, not once but every reporting period.