Data Requirements In The PEF Approach
For an LCA study to be compliant with the ISO 14040 and 14044 standards, it needs to include the data quality requirements and a data quality analysis. A lot of LCA practitioners who provide these required elements don’t derive actual value from them, though. They often do it to tick all the boxes. But why is this? And how is the PEF initiative changing this?
It’s obvious that high-quality data is crucial if you want to get sound results from your assessment of product life cycles and organization value chains. LCA practitioners are doing the data quality analysis and setting the requirements anyway, so why don’t they get anything out of it? The answer again is simple: there are no clear and quantifiable data quality requirements to be met. This article is focused on the what is most novel and important with regards to data quality criteria in the PEF initiative.
Materiality Approach for Data Quality Requirements In The PEF Initiative
In LCA, there is traditionally a distinction between the foreground and background system of a product life cycle. The foreground refers to the activities under the operational control of the company conducting the LCA whereas background refers to upstream and downstream activities.
The PEF initiative is providing concrete instructions for dealing with data quality and data collection. Here, the principle of materiality is applied, namely focus where it matters. It names two relevant elements for determining which data you need to collect and what level of quality it needs to have:
- Level of influence: how much influence does the company performing the study have on them?
Any company in a product’s value chain can perform an EF study of the product and, depending where they are, they will have control of different activities. The manufacturing company of the product has control of the manufacturing activities whereas retail has control of the retail activities.
- Impact relevance: what are the most relevant processes driving the environmental profile of the product?
If some of the activities controlled by a company cause little environmental impact, there is no point in investing to collect company-specific data. So, primary data is not required for the complete foreground system but it is for environmental hospots.
Further, there are six data quality criteria, scored in five levels from very good (score 1) to very poor (score 5). The data quality score of a dataset is the average of the scores of all six data quality criteria. Completeness, parameter uncertainty and methodological appropriateness and consistency are generic criteria. Technological, geographical and time-related representativeness are context-specific criteria. The generic data quality criteria are fully described in the PEF and OEF methods; their data quality tables are applicable to all EF studies. The other three data quality criteria depend on context. Therefore, PEFCRs must provide further guidance on data quality assessment, explaining which data quality levels and ratings are to be used to calculate context-specific data quality ratings (DQRs).
In addition, three more aspects are included in the quality assessment: review, documentation compliance with ILCD and nomenclature compliance with ILCD.
Required Elements Of Product Environmental Footprint Category Rules (PEFCRs)
When developing PEFCRs for a specific product group, some specific elements must be included.
- The minimum list of processes to be covered by company-specific data. For instance, gate-to-gate activities and processes. This helps avoid that a PEF profile will be calculated solely on the basis of default data
- The list of the activity data that the applicant should declare. This is data about activities that are likely to be under the operational control of the company. For instance, x kWh of electricity used in the company’s manufacturing site.
- Context-specific data quality tables. This is where the six data quality criteria come into play. The data quality tables for the generic criteria are already set. The PEFCR must determine the data quality tables for context-specific criteria that need to be provided by the applicant. For instance, that data of a period of 3 to 5 years prior to the period under study have a quality level of 2 out of 5, so good quality.
- Default datasets to be used for all processes except those that are under your operational control and have been identified as the most relevant (see situation 1 below). Providing these defaults datasets means that the only differences in PEF profiles that follow their PEFCR result from differences in company-specific primary data. This is what makes it possible and even crucial for PEF profiles to meet the comparability objective set by the PEF initiative. The European Commission is now purchasing data to ensure that default secondary datasets can consistently be used by all that would like to calculate PEF or OEF profiles for free.
In itself, this is already fantastic news for businesses that so far have had to purchase their own commercial life cycle inventory (LCI) data. But this revolutionary action from the Commission may also completely change the landscape of LCI data provision in the future! We’ll just have to wait and see.
Implications For EF Studies And Beyond
If your company is implementing a PEFCR, you need to check for each process in your value chain if it falls into the ‘most relevant’ category, and how much influence your company has over it. The decision tree below can help.
Once it is clear in which situation the process falls and whether it’s in the ‘most relevant’ category, you can follow the rules from the data needs matrix (DNM) below.
Source: European Commission. 'Guidance for the implementation of the EU Product Environmental Footprint (PEF) during the Environmental Footprint (EF) pilot phase. Version 5.2' – February 2016.
There are a total of 4 options that summarize the data needs matrix:
- Situation 1. Primary non-aggregated (with some transparency) data is required and the data quality score must be high (at least 1.6).
- Situation 2. Let’s imagine, for example, that you purchase a material from your supplier and this has been identified as a most relevant process in your PEFCR. If your supplier is not willing to provide a PEF-compliant dataset with some transparency for the material you purchase, you are asked to replace from the default secondary dataset using the country specific electricity mix of the material’s manufacturing site as well as replace the dataset used for the transport and the distance it travels from your supplier to your manufacturing site.
- Situation 3. You are required to use the default secondary dataset (provided in the PEFCR) with a DQR of at least 3.0.
- Situation 4. For processes that are not considered the most relevant and are not run by your company, the data quality requirements are less strict (DQR ≤4.0) and the default secondary dataset (provided in the PEFCR) must also be used.
Then, the LCA applicant shall re-calculate the DQR for all the datasets used for the most relevant processes, the new ones created, and other processes in situation 1 (so all except processes which have not been identified as most relevant and that are in situations 2 or 3).
Data Requirements Consistency – Another Milestone For The Comparability Objective
Traditional product category rules (PCRs) are already very much aligned in scope, with the purpose of setting similar rules for all products within a product group. However, PCRs are not very prescriptive in their data quality requirements. PCRs do not have a list of default secondary datasets to be used, nor do they specify which default activity data a company needs to use. These two requirements, however, are in my opinion absolutely essential if we want to achieve comparability between PEF profiles that were not assessed in the same study.
Because the PEF initiative is now only in its pilot phase, we will still have to wait to see how this evolves. I believe that the data quality requirements defined in the PEF initiative are a very important - or even defining - step in standardizing and guaranteeing the comparability of PEF profiles. One thing is certain: people following the EF data requirements and analysis in a PEF study won’t just be doing it to tick all the boxes.
Learn More About PEF
If you want to learn more about our role in the PEF initiative, please drop me an e-mail. See other episodes of this series: