Module D net (output) flow calculation using Ecoinvent EN 15804 GD

Question

I am interested in your take on the practical modelling of secondary material (SM) inputs and Material For Recycling (MFR) outputs when using the EN 15804 (GreenDelta implementation) system model of Ecoinvent.

The database (correctly) accounts for these indicators not only in the foreground system, but also in the background system. The result is that both indicators are likely not 0 even if my product contains no recycled material and is not sent for recycling at end-of-life.

My problem comes when modelling the net flow of materials leaving the product system (stages A-C) to be considered in module D. To account for processes in module D, the net flow (MFR minus SM, in kg) needs to be separated out for each material, but the indicator results from the life cycle inventory are aggregated. SM and MFR can thus both be a mix of e.g. iron scrap, aluminium scrap, waste cardboard, and there is no feasible way to model module D.

I cannot find any rules in the standard or PCR that would allow me to disregard from these flows just because they arise from the background system.

How have other LCA practitioners solved this? Are you tracking the use of "Recycled content cut-off" processes for each type of material and doing the net flow calculation manually? Are you manually changing names of elementary flows in the database to "SM (iron scrap"), etc., or are you simply disregarding the SM and MFR flows from the background system in your module D calculations?

I checked but this specific topic does not seem to have been addressed in the white paper Ciroth and Hamed (2025) Module D in Environmental Product Declarations (EPDs).

Answer 1

Dear Kristian,

You raise an important point that there is still no clear guidance if the input (SM/RSF/NRSF) and output inventory indicators (MFR/MER/CRU) should be calculated over the foreground only or over foreground+background. I have the feeling that the trend is going towards foreground only, even though this is not really a life cycle perspective. Of course, it has the advantage that results are easier to grasp and not mixed with the background, but it also has the disadvantage that background datasets with e.g. high recycled content will contribute with lower e.g. greenhouse gas emissions to the foreground, but all this is not reflected in the overall recycled content used in the whole product system.

If over the foreground only, as you said, module D inputs can be directly taken from MFR minus SM. There would be no contributions from the background database and all results are only manually inserted by the practitioner into the foreground.

If over the foreground+background: What do you think about the idea that we split all 6 indicators (SM/RSF/NRSF/MFR/MER/CRU) into the indicators from background and into the indicators that are manually inserted by the practitioner in the foreground only? For example, "Secondary materials" will be split into "Secondary materials (background)" and "Secondary materials (foreground)". The total result will be still the same as before, but practitioners can directly see how much stems from the background only and how much stems from the foreground only and how much stems from the sum of the two.

Best wishes,
Conrad

Comments

Hi Conrad,

I personally don't see EN 15804 or PCRs supporting the interpretation that these indicators are only a matter of foreground modelling. For example: most EPDs of construction products that consist of in whole or in part of steel will not have operational control all the way back to the SM input of iron scrap. For the indicator to make any sense, it thus needs to be included in background data at the point where the scrap enters the product system. Just the way that you have implemented it in EN 15804-GD.

With the current interpretation in PCR 2019-14 Construction products (which I don't personally agree with), the net output flow (MFR minus SM) that needs to be considered in module D can be both positive or negative. All product systems using more iron scrap than they generate (based on the chosen EoL scenario) will thus have to consider a "avoided processing after EoW state" and "avoided avoided burdens from replaced alternative production". To model this, it seems to me that the practitioner needs knowledge of what material types are actually crossing the system boundary both on the input side and the output side.

For this reason, I don't believe a split into SM (background) and SM (foreground) would solve the problem for me, but I am very interested in how others have interpreted the rules or have solved the problem.

---

What I meant was splitting all input and output indicators (SM, RSF, NRSF, MFR, MER, CRU). But we can do it anyways, since it will help verifiers to see the contributions from background database and from foreground, without changing the total result.

What I don't understand from your comment: Module D inputs cannot be MFR - SM over the whole supply chain. This net flow is only from the foreground, no? You can only have benefits and loads of the materials in your product, but not when there was a MFR somewhere deep in the supply chain, from an ecoinvent process.

PS: When you go to the contribution tree and you select the inventory indicator flow, you can see the contributions from each process. And when you go to the impact analysis tab, and you change the option to flow view, you can select the inventory indicator category and you will see which processes in the whole supply chain have e.g. SM or MFR inside, and you see their values and sorted from largest contributor to lowest. But maybe you still want to have the intermediate flow names in these processes that lead to SM or MFR?