4.2 - Method for collecting activity data

How to collect the activity data necessary for accounting?

Source: Midjourney

This subsection details what an activity data is, how to collect them and how the different types of activity data can be used.

As a reminder, the emissions from each emission source of the organization are estimated as follows:

Emission from a source = Activity data x emission factor = result ± uncertainty

Activity data

Activity data are the data that account for the different physical flows of the organization. A mass of raw material, the number of kilometers traveled by the various people involved, the quantities of energy consumed by the organization are examples of activity data.

For a given emission source, the activity data and the denominator of the chosen emission factor must be expressed in the same unit. In general, it is advised to collect the data that allows using the most accurate emission factor, and convert it if necessary. If only one type of data is available, then the appropriate emission factor should be chosen.

Activity data (km) x emission factor (kgCO₂e/km) = result (kgCO₂e) ± uncertainty (kgCO₂e)

Collection of activity data

Collecting these activity data is often the most time-consuming phase for an Initial level approach. Indeed, for each emission source it is necessary to identify whether and in what form the data exists within the organization, and who holds it. If some data are not available, process improvement actions will need to be part of the transition plan, in order to build monitoring of these data in anticipation of the next Bilan Carbone®. If data collection was planned upstream and improved after the first assessment, it can prove faster (notably for a Standard Assessment). An organization carrying out an Advanced Assessment will have put in place a rolling data collection dashboard.

To optimize present and future data collections, the organization must:

• Create a data collection matrix, which defines precisely the data to be collected, relying on theidentification of emission sources, the boundaries developed previously and the data already present within the organization.

• Assign a person in charge for each emission category, or for each type of activity (depending on the company structure), who is responsible for collecting the data for that category from the different holders.

• Document and keep all collected data, including in particular their source, the units and theuncertainty associated, related documents (for example invoices, delivery notes, internal reports) and any other information deemed relevant, to ensure that these data are traceable and comparable from year to year.

• Balance collection efforts: Collecting very precise data for a non-significant emission source is often time-consuming. It is important to ensure careful collection for significant emission sources. Focusing on action and reducing significant emissions is the core of the Bilan Carbone®.

In the long term, it is recommended that the organization equips itself with a system that monitors data collection, which will allow it to raise certain indicators, carry out future Bilan Carbone® approaches more quickly, and ensure monitoring of its transition plan. However, note, as developed below, that automated collection of accounting activity data will not allow achieving these three objectives, because these data will call upon spend-based emission factors.

🔎 To express the Bilan Carbone® with a so-called "analytical" reading, consistent with analytical carbon accounting, activity data must be collected by analytical dimension, and therefore broken down by responsibility (suppliers, customers, sites, etc.). Accounting codes make it possible to be exhaustive on the emission sources supported by the organization. The associated data sources are thus linked to each accounting code (for example an export of purchases for accounting code 601321), then to the analytical dimensions (for example an export of purchases broken down by suppliers).

Data sources must integrate the analytical dimensions identified as relevant for the analysis of results. The absence of analytical dimensions in a data source necessary to split emissions according to that analytical dimension can be a justification for abandoning that analytical dimension. Adding the analytical dimension in the data source will be a process improvement action of the approach.

Data collection matrix

For each emission source, this matrix must indicate the Bilan Carbone® category and subcategory concerned, the label, the value, the unit, the type, the source and the uncertainty (with a single characteristic) of the activity data.

🔎 In analytical carbon accounting, the matrix indicates for each activity data, the accounting code and the associated analytical dimension.

This matrix will also be used to list the emission factors that will be attached to these activity data, indicating the desired emission factor label, the label of the emission factor actually used, its value, its unit, its uncertainty (all characteristics), its type and its source.

⏳[WIP] An example of a data collection matrix will be made available in September 2024.

The different types of activity data

The four main types of activity data are detailed below. They are described from the most reliable (real data) to the least reliable (approximated data).

Real activity data

Real activity data can represent a physical or accounting.

🔎 These real activity data are called primary activity data by the Regulatory GHG Assessment.

Physical activity data

These are physical data available to the organization in raw form and with a high degree of accuracy, either because they precisely track the physical flows involved, or because they can be traced retrospectively. For example, the mass of raw material used (example: 200 tonnes of new stainless steel per year) or the number of kWh of electricity consumed (shown on the invoice), are real data that are easily collected by and for organizations.

Physical activity data are the most reliable data for carrying out a Bilan Carbone®. It is recommended to use this type of activity data as much as possible.

However, the organization must maintain a critical view of the statistic used to qualify the associated uncertainty.

Accounting activity data

These are data available to the organization with a very high degree of accuracy, because financial accounting is highly regulated. For example, the cost of electricity for the organization is information that will always be available. These data are therefore associated with very low uncertainty.

However, it is strongly discouraged to generalize the use of these data, for two reasons:

• These accounting activity data are associated with spend-based emission factors, which are themselves very uncertain

• These activity data do not allow a physical analysis of what is happening within the organization, which prevents the development and implementation of a relevant transition plan. Indeed, the only action that can be associated with accounting activity data is to reduce the organization's expenditures, which is limited and sometimes even counterproductive (buying better often means buying more expensive)

Extrapolated activity data

These are data derived from an extrapolation from other physical, statistical or approximate data. This type of data is regularly used in Bilan Carbone® approaches, notably for travel and catering categories, via sending a questionnaire to the organization's employees. A certain number of people respond to the questionnaire, and the results are extrapolated to all employees. The organization can also use extrapolations when some data for the current year are not yet available (for example using an electricity bill from a prior year), or when it uses activity data from sectors, sites or geographic locations that are not representative of its situation, applying a correction factor. Note, if unrepresentative data are used as-is and uncorrected, those data are not extrapolated data but approximated data.

These data are generally quite reliable, as they are specifically adapted to the organization. However, it is preferable to use real data.

The assumption used to extrapolate must be documented within the data collection matrix.

Data that would come from computer simulations or Artificial Intelligence (AI) models are also considered extrapolations. The organization must then judge the acceptability level of these simulations or models.

Extrapolations can be of several types. For example, a temporal extrapolation (using a bill from year N-2) will not have the same consequences as a geographic extrapolation (using data valid in rural areas for sites in urban areas), or as an extrapolation to a population (40% of people respond to a questionnaire, and the results are applied to all employees). The organization must judge the acceptability level of the extrapolation, as some incongruous extrapolations can be much less precise than using statistical data.

Once the extrapolation has been made, the organization must take a critical view to qualify the associated uncertainty. The organization can thus balance its collection efforts: if the extrapolation represents a significant share of the Bilan Carbone®, either the extrapolation must be precise, or the organization must endeavor to collect real activity data. Conversely, if the extrapolation represents an extremely small share of the Bilan Carbone®, a moderately precise extrapolation may be acceptable.

🔎 These activity data are also called extrapolated activity data by the Regulatory GHG Assessment.

Statistical activity data

These are data that come from more or less targeted statistics, for example for average home-to-work travel distances: the average may concern France, or the tertiary sector in France, or the scale of a certain municipality.

These data are generally associated with significant uncertainties, although this varies depending on the dataset (France, municipality, sector) on which the statistic is performed, the data source, and the date of the statistics used. It is recommended to minimize the use of this type of data, and to use them only for emission sources where data would be completely missing, or for non-significant emission sources. For example, for a supplier or partner who does not respond to the organization's requests.

The organization must have a critical view of the statistic used to qualify the associated uncertainty.

🔎 These statistical activity data are called secondary activity data by the Regulatory GHG Assessment.

Approximated activity data

These are data that are not representative of the situation within the organization, but which are nonetheless used in cases where the organization is not able to propose an extrapolation adapted to the organization or access physical or statistical data.

These data are associated with very large uncertainties. It is recommended to minimize the use of this type of data, and to use them only for emission sources where data would be completely missing, or for non-significant emission sources.

🔎 These activity data are also called approximated activity data by the Regulatory GHG Assessment.

Summary of the different types of activity data

Type of AD	Example	Associated uncertainty
Real	kWh of electricity consumed on site, shown on the invoice	Very low
Extrapolated	kWh of electricity consumed on site, extrapolated from invoices for the first six months of the year	Variable depending on the quality of the extrapolation
Statistic	kWh of electricity consumed on site, obtained from the French average of electricity consumption per m² in the tertiary sector	Average
Approximated	kWh of electricity consumed by another organization in the same sector, uncorrected	High

Requirements related to data collection

Here are different requirements to be met in terms of collection of activity data for each of the 3 maturity levels.

Initial Level: criterion K1

All methods of data collection are accepted, provided that the most precise data are always prioritized.

The organization must build a solid documentation process that will improve data collection for future Bilan Carbone® exercises. If done accurately, the data collection matrix can be an integral part of this documentation.

The organization must precisely identify the data that would not be sufficiently accurate or accessible, and make them reliable before the next Bilan Carbone® exercise.

Standard or Advanced Level: criteria K2 and K3

Any extrapolated, statistical or approximated data must be framed and be the subject of extensive documentation justifying:

• The necessity of its use instead of physical data (specifying why the physical data is inaccessible or unsatisfactory)

• The associated assumptions and the methodology used to obtain the data

• If relevant, the tools used to obtain the data

In the case of an Advanced Assessment, it is strongly recommended that the organization attempts as much as possible to exclude the use of any extrapolated, statistical or approximated data.

The organization must build a solid documentation process that will improve data collection for future Bilan Carbone® exercises. If done accurately, the data collection matrix can be an integral part of this documentation.

In addition, the organization must put in place actions that will improve data quality, facilitate access to real data, for example by supporting its suppliers to make the data transmitted by them more reliable.

The organization must also implement data monitoring systems within itself, including notably a dashboard of significant emissions. These monitoring systems can be automated and periodic (monthly, yearly, etc.), depending on the difficulty of collection and the volumes of GHGs associated with these data. Data associated with highly emitting activities should be covered first by these monitoring systems. Less emitting activities will be covered gradually as the organization's maturity on carbon issues increases.

Such processes allow the organization to develop a "climate culture" internally, which promotes the implementation of reduction actions and the monitoring of their implementation.

---

Do you have a comprehension question? Consult the FAQ. The method is living and therefore likely to evolve (clarifications, additions): find the change log here.