G-24 Collecting foreground and background data for product LCA

Aspect G-24 Collecting foreground and background data for product LCA
Different data need to be collected for a product LCA study. The most important data collection is usually the production stage especially for highly complex manufacturing processes. The data can be based on industry or companies involved in the project else on secondary sources e.g. background databases providing average datasets.

related study objective

stand-alone LCA comparative assertion

related study phase

goal and scope definition inventory analysis (LCI) impact assessment (LCIA) interpretation reporting

relevant for

new buildings existing buildings construction products screening LCA simplified LCA complete LCA
Provisions The practitioner should follow ILCD Handbook provisions as well as EN 15804 if more relevant and specific for the goal and scope of the study.The provisions of this aspect is linked to the “definition of system boundaries for product LCA”.
Rules from:

EN 15804

6.3.6 Selection of data

CEN/TR 15941,

Sustainability of construction works — Environmental product declarations — Methodology for selection and use of generic data


The ILCD Handbook provides very useful guidance on this topic. See below the references of the rules (and the quotation of the rules where relevant). Note that the rules from ILCD concerning the definition of system boundaries for products are reported in the corresponding aspect.
Provisions 6.8.2 Technological representativeness

“I) SHALL – Good technological representativeness (…)

III) SHALL – Different technologies for attributional and consequential modelling
          III.a) Attributional modelling

          III.a.i) Foreground system: (…) technology-specific primary data. (…) Secondary data of the actual suppliers/downstream actors should be preferred to other (third party) secondary data.

          III.a.i) Background system: average technology as market consumption mix data should be used (…)

          III.c) Using not fully representative data: (…) use justifiable only if this is not relevantly changing the overall LCIA results compared to using fully representative data”

Provisions 6.8.3 Geographical representativeness

“I) SHALL – Good geographical representativeness (…)

II) SHALL – Different technologies for attributional and consequential modelling

          II.a) Attributional modelling

          II.a.i) Foreground system: (…) site or producer/provider specific data  (…) supplier-specific data for the products that connect the foreground with the background system. Generic data of geographical mixes can be used also in parts of the foreground system if for the given case justified as being more accurate, precise, and complete than available specific data (especially for processes operated at suppliers).

          II.a.ii) Background system: average market consumption mix data (…)”

Provisions 6.8.4 Time-related representativeness

“I) SHALL – Good geographical representativeness (…)

II) SHALL – Specific seasonal or diurnal situation?

III) SHOULD – Time-related representativeness of future processes: (…) fully time-representative future data should be used (…). If this is not possible:

          III.a) Best Available Technology (BAT) and recent data (…)”

Provisions 6.9 Types, quality and sources of required data and information

“I) MAY – Overview of the principles types of data and information

II) SHOULD – General requirements on data and data set quality

III) SHOULD – Potential sources for the required data, data sets and information

          III.a) Well documented data

          III.b) Pre-verified data”

Provisions for the foreground system of an analysed system

Provisions 7.3 Planning data collection

“I) SHALL – Identify newly required, study-specific unit processes

II) SHALL – Average and generic data

III) MAY – Identify data and information sources

IV) MAY – SI units

V) SHOULD – Multi-annual or generic data to be preferred?

VI) MAY – Relevance-steered data collection:

(…) Building on existing experience that sufficiently reflects the analysed process or system and that is of high quality is an essential guide. (…) PCR can represent this experience (…).”


Provisions Types of input and output flows to collect

“I) SHALL – Types of input and output flows

          I.a) Input of “consumed” products

          I.b) Input of wastes

          I.c) Input of resources from nature

          I.b) Emissions to air, water, soil

          I.b) Other input and output side interventions with the ecosphere

          I.b) Output of wastes

          I.b) Output of valuable goods and services provided by the  process, as product flows”

Provisions Reference amount of the reference flow

Provisions Representativeness regarding operation conditions

“I) SHALL – Full operational cycle of the process, if required

II) SHOULD – One full year as data basis:

for measured data of operated process

III) SHOULD – For parameterised processes”

Provisions Interim quality control

Provisions Emission of particles to air

“I) SHALL – Inventory only poorly water soluble compounds as particles

II) SHOULD –  Differentiate particles size classes

III) SHALL – Inventory particles additionally as the substances they are composed of”

Provisions 7.6 Selecting secondary LCI data sets

“I) SHALL – Use consistent secondary data sets

II) SHOULD – Quality-oriented selection of secondary data sets

III) MAY – Prefer pre-verified data sets

IV) MAY – Prefer well-documented data sets”

Provisions 7.8 Modelling the system


II) SHALL – Complete system model (…)

V) SHALL – Use consistent data to fill gaps


ISO 14044 Data quality requirements
ILCD guidance should be followed. It needs to be precised that the data collection is a very time consuming step in the LCA. However, this step is highly linked to the overall quality of the study. The practitioner may face different challenges for collecting foreground or background data in practice.Generally speaking, the foreground data collection at the manufacturer plant should follow consistent rules (e.g. see ILCD provisions Checking legal limits; provisions Interim quality control).
General guidance for product LCA studies
The data collection should be conducted depending on the goal and scope of the study.Specific guidance for EPD studies

Depending on the EPD program, some already pre-verified background data may be made available for the practitioner. If so, there is no need to comprehensively review the background data as they are supposed to meet the quality guidelines and goal and scope of the PCR (e.g. EN 15804).

It is very important to comply as much as possible to the specific rules given in the PCR (EN 15804 but also national implementation). One key issue is to know where we need to stop collecting data from the manufacturer. Usually, he controls only the gate-to-gate processes. However, under some circumstances, he may also be responsible of upstream process e.g. extraction of clay. In that context, it may be relevant to collect primary data for this process instead of using secondary data available in background databases.

For the gate-to-gate stage (under the control of the manufacturer), the data collection is function of the type of EPD.

For single manufacturer’s EPD, data can be taken from 1 production site or averaged across more than one the production sites.

For group of manufacturer’s EPD, data are taken from different companies (their data may be already averaged internally if they have more than one production site). They need to be averaged (horizontal averaging according to the ILCD). In that case, it is very important to make sure that the different companies have similar environmental impacts for the production of the product. PCR (in every national context) may provide more specific rules than EN 15804 to calculate these group of manufacturer’s EPD. This issue being to make sure that the declaration is representative of the companies with a limited variability. By requesting a low variability on the final results (e.g. 10% for GWP, 20% for acidification) it leads to differentiate different group of companies that have different technologies and related environmental impacts. Otherwise, the issue is to have a company that has higher impacting production process being hidden under the “average” value thanks to the more efficient/less impacting production processes.

Guidance for E2B EI research projects

Within E2B EI research projects, some confidentiality reasons may lead the use only of generic data for both foreground and background system. In some cases, where the company is keen on giving the gate to gate information for the LCA study, the modelling will be more representative of the product.

The limits encountered during the data collection step should be part of the LCA report. The interpretation of results should be made according to the quality of data collected during the inventory analysis.

Comments are closed.