The carbon footprint is the photograph, at a given moment, of all of an organization's greenhouse gas (GHG) emissions.
And it is in order to be able to measure this carbon footprint that several methodologies have been developed, like GHG Protocol or ISO 14069 standard.
All of these methodologies rely on the use of models to make the link between business activity data and GHG emissions. These are calculation standards that have been developed in order to measure the carbon footprint.
It is therefore an estimate of the emissions associated with the company's activities. And who says estimate inevitably says uncertainty.
Where does uncertainty come from in calculating a company's carbon footprint? How is it measured? How is it integrated into our Sami platform? And how do you interpret uncertainty results? We explain everything to you in this article.
1. Uncertainty in the measurement of GHG emissions
1.1 What is uncertainty?
Carbon footprint uncertainty refers to the margin of error or imprecision in estimating the carbon footprint compared to the real value.
In general, uncertainty is given as a single value (for example 15%) or as an interval (-10% to +15%), for a given confidence level, often 95%. That is, the probability of finding the real value in the given interval is 95%.
- The lower the uncertainty, the smaller this interval
- The greater the uncertainty, the longer the interval
1.2 Sources of uncertainty when calculating a carbon footprint
A company's carbon footprint is obtained by aggregating all the direct and indirect greenhouse gas emissions induced by the activity of this company.
And these emissions are calculated using known and collected company activity data (surface area, electricity consumption, etc.) to which an emission factor is associated in order to estimate the emissions induced by this activity data.
The uncertainty of each emission item therefore occurs both in terms of activity data (for example, electricity consumption from a bill will be more accurate than if it is estimated from the surface area) and at the level of the associated emission factor (the emissions induced by the combustion of one liter of gasoline are fairly reliable while those induced by the production of a complex manufactured product will depend on a number of hypotheses).
For each activity data and each emission factor, the sources of uncertainty are multiple:
- Lack of comprehensiveness : measurements or other data are not available, either because the process is not yet recognized, or because a measurement method does not yet exist, or simply because the data could not be collected.
- Lack of reliability : the data is not measured (certain hypotheses are taken to estimate them) or are not verified (causing measurement, reporting, transmission errors).
- Lack of temporal representativeness : the data or the associated emissions calculation are dated and may introduce a bias.
- Lack of geographic representativeness : there are emission data or factors but they are not specific to the study area.
- Lack of technological representativeness : emissions linked to a specific process cannot be measured and the latter is modelled by another process, more or less similar.
This is in detail how uncertainty is treated for both sources, the activity data and the emission factor.
1.2.1 Uncertainty around activity data
In order to facilitate the processing of the balance sheet, default uncertainty values have been defined for all activity data collected via the Sami collectors.
Activity data is often a combination of different user inputs as well as conversion factors. A default uncertainty has been assigned to each entry and each conversion factor. These entries can be grouped into 4 main categories:
- 0% uncertainty:
- Known fixed or contractual parameters: amortization period, paid working time, number of months of occupancy of a premises, etc.
- Extensive tracking data: data from the FEC for example.
- Data affecting only the categorization of emissions: share of use of a vehicle for each use, share of trips to each site, etc.
- Between 5% (measured, complete, up-to-date data, etc.) and 30% (slight extrapolation, estimation, adaptation) uncertainty :
- Data consolidated by the company: number of equipment, unit weight, liters of fuel in the company fleet, electricity bill, etc.
- Reliable declarative data: distance between home and work, number of publications, proportion of teleworking, etc.
- 30-50% uncertainty:
- Moderately reliable declarative data: vehicle weight, type of lunch regime, etc.
- Average hypotheses: site consultation time, parking area based on the number of spaces, etc.
- > 60% uncertainty:
- Strong hypotheses/high data variability: refurbished device, annual air conditioning consumption, consumption based on the degree of insulation, etc.
- Unreliable declarative data
1.2.2 Uncertainty about EFs
The emission factors we use come from several databases.
Unfortunately, apart from Ademe's Empreinte® database, most of these databases do not communicate the uncertainty associated with emission factors.
For the integration of uncertainty measurement on the Sami platform, we distinguish three sources of EF:
EF from the Empreinte® database with uncertainty
We have taken up the uncertainty associated with it, with a few differences. Some uncertainties have been slightly inflated due to their age (lack of temporal representativeness): this is in particular the case of all monetary ratios and electricity operating expenses whose methodology used (average over 4 years) introduces significant uncertainty compared to the reporting year.
EF interveners in Sami collectors
These emission factors (around 1000) are used for a large part of the assessments carried out on the platform. The uncertainties of these EFs have been determined as accurately as possible:
- based on data quality indicators if available
- by observing the variability of similar EFs in other databases or articles or by looking through the sensitivity analysis of the study from which the EF is extracted
- drawing on the uncertainties of the Empreinte® database for categories of products/services similar to other databases
- as a last resort, based on internal expert opinions based on industry knowledge
Other EF
For the other emission factors present on our consolidated basis, uncertainties are gradually added.
- from the program category using inspiration from The Matrix Pedigree
- based on the specificity of the EF (for example individual or collective declaration, by default on INIES)
2. How do you measure uncertainty?
Uncertainty is most generally described by a normal or log-normal distribution distribution.
We decided to implement a normal law in the application that is more intuitive and easier to handle and manipulate. In addition, the Empreinte® base, one of the only bases for specifying the uncertainty of its emission factors, provides them in the form of a normal distribution.
2.1 Uncertainty aggregation formulas
In the Sami application, the uncertainty of an emission line is calculated using the uncertainties in the activity data and the EF. To obtain the total uncertainty of the balance sheet, the uncertainties of all emission lines must then be aggregated.
The uncertainty in a data resulting from a sum or a product is calculated using two mathematical formulas:
- For each emission line, the activity data is multiplied by the associated emission factor. The uncertainty of an emission line is therefore given by the following formula:
- For the final balance, the emissions of each line are summed up. The overall uncertainty of the balance sheet is obtained with the following formula:
2.2 Practical application: an example of calculating uncertainty on emission lines
For example, consider this group of four emission lines:
Thus, the uncertainty of the first emission line (product of the activity data and the EF) is given by the following formula:
The same goes for calculating the uncertainty of the other emission lines. Then, the aggregate uncertainty of these 4 emission lines (sum of the 4 emissions) is given by the following formula:
3. The integration of uncertainty calculation in the Sami platform
3.1 Intermediate uncertainties at different collection points
In form collectors Sami, the uncertainties in the emission factor and in the activity data are automatically applied according to the rules explained above. The details of the uncertainties applied are available in each collector and can be modified during the consolidation.
3.2 Overall balance sheet uncertainty in the application
We calculate global balance sheet uncertainty only if all significant emissions are covered by uncertainty, but we do not impose total coverage. Indeed, the overall uncertainty of the balance sheet will be very strongly dependent on the most representative emission items on the balance sheet and, conversely, very little dependent on the low emission items. It is thus possible to assess the overall uncertainty even if the low emission positions have no specified uncertainty.
The total uncertainty is presented in the summary of the carbon balance results.
4. How to interpret the uncertainty of a carbon footprint and how to reduce it?
Different indicators can be looked at for the interpretation of uncertainty.
- The global uncertainty of the balance sheet is a first indicator of the quality of the carbon footprint
- Low (< 15%): The real impact of the company is close to the assessed estimate.
- Moderate (between 15% and 30%): The carbon footprint estimate gives a good order of magnitude of the real impact of the company but it is necessary to identify which emission items increase uncertainty and how to obtain more accurate data on these items
- Strong (> 30%): The carbon footprint results should be taken with a grain of salt. The company must make efforts to better know, monitor and consolidate its activity data internally. While the impact comes mainly from suppliers, it is important to engage in discussions with them to improve traceability and product composition.
- Another indicator to consider is the type of data and emission factors used.
- The percentage of physical and monetary data.
- The percentage of specific, semi-specific or generic data
To reduce uncertainty, the company can set goals on a minimum percentage of physical or specific data in the coming years and conduct LCAs on its best-selling products in order to refine the accuracy of measuring the environmental footprint.
{{newsletter-blog-3}}
Other resources around the carbon footprint:
- The General Carbon Plan
- Understand everything about the carbon footprint
- The carbon footprint, a tool for economic performance
- Analytical carbon accounting: making carbon accounting a strategic tool
- Understand everything about emission factors
Mission Décarbonation
Each month, a description of the business climate news and our advice to help you decarbonize, followed by more than 5000 CSR managers.
Never miss the latest climate news and anticipate new regulations!
Your carbon footprint with Sami
Sami supports you in measuring and reducing your GHG emissions
.png)
Les commentaires