PEF Database 3.1 in openLCA shows only elementary flows. How to access or replace intermediate inputs?

Question

Hello,
 

I downloaded the PEF database v3.1 via openLCA Nexus and imported it into openLCA. However, when I inspect the processes, I can only see elementary flows in the inputs and outputs. I do not see any intermediate (product) flows such as electricity, transport, or other background processes.

For my LCA modelling, I need to:

1. Identify the intermediate inputs embedded in these PEF processes, and

2. Replace some default background inputs (e.g. electricity mix, transport) with company-specific primary data, while keeping the rest of the dataset unchanged.

Without visible intermediate flows, it is unclear how to technically substitute parts of the process with primary data, as required by PEFCRs.
 

Could someone please clarify:

1. Whether the absence of intermediate flows in PEF 3.1 is by design, and if so, why?

2. What is the recommended workflow in openLCA for using primary data (e.g. electricity, transport) when working with aggregated PEF datasets?

Any guidance, documentation references, or practical examples would be greatly appreciated.

Answer 1

The EF database from European Commission is only provided as aggregated data sets (system processes). The reason is due to licensing/confidentiality of data providers. You will not find a fully disaggregated  EF database anywhere. Some processes in the database are partly disaggregated so that you can change and adapt transport or energy inputs.

Also for PEF studies and using PEFCR, you should not change the background data. You can create your own foreground processes with primary data (e.g. how much electricity for your own processes and which electricity source) and link it to the background data sets. Which PEFCR you are referring to? Some PEFCRs might be old and based on EF 2.0 where more datasets where partly disaggregated so that more datasets could be adjusted with fixed parameters in the processes. As far as I know, most of these parameters are gone in EF 3.1.

Comments

Thank you for the clarification regarding the aggregated structure of EF 3.1 datasets.

I am referring specifically to the Apparel and Footwear PEFCR based on EF 3.1.

I understand that when full primary data are available for a process, the recommended approach is to create a new foreground process and link each input (e.g. electricity, fuels, materials, transport) to the corresponding EF background datasets.

However, I'm still confused about situations where only partial primary data are available.

For example:
I may know the electricity consumption and transport distances for a process,
But I may not have complete data on auxiliary materials, emissions, waste flows, etc.

In databases like ecoinvent, one could keep the default intermediate inputs and modify only specific exchanges (e.g. replace electricity mix). With EF 3.1 being mostly aggregated, this does not seem possible.

In such cases is it correct that the only modelling options are:
(a) Use the EF process fully aggregated without modification, or
(b) Build an entirely new foreground process with all inputs and outputs defined?

How can I handle situations where collecting 100% of foreground data is not feasible, but some key parameters (e.g. electricity source) are known?

---

Hi,
From my perspective what you are stating at the end of your question is completely true. You do have this problem that you cannot implement "partially company-specific datasets", where you only exchange the electricity mix.

I have only one addition to the discussion: From my pov, you do not really have the option to choose between option a) and b) freely, since the demand for company-specific datasets derives from the Data Needs Matrix. So either you run the process yourself, then you must create a company-specific dataset (if the process is among the most relevant ones) OR it is not run by your company, in that case you are free to choose what you do.
Hope my thoughts help :)

---

Please also have a look in the PEFCR supplementary documents. There is an excel file described which data to choose and which not (A&FW_PEFCR_v3.1_AnnexVII - Inventory modelling and default datasets) and also the example for the representative product (A&FW_PEFCR_v3.1_RP_study_Annex I-LCI). Hope this helps

Answer 2

Hi,

Everything that Conrad said, plus we have EF 3.1 Lite and Developer version. The Developer version holds unit processes and product systems. So, there you will find sometimes the same process as a unit and a system process. If you use the unit processes (partly disaggregated), you have to make sure that you set the providers correctly. However, you are not supposed to modify those processes you use in the background, as Conrad stated.
Hope this helps