Frequently asked questions and answers for potential data suppliers

This document lists common questions from potential data suppliers to the ecoinvent database. If you are considering submitting a dataset to ecoinvent, the answers listed below will help you understand the conditions and requirements ecoinvent has for submitted datasets (you can find the following text as a pdf-File here !).

 

When will version 3 be released?

The release of ecoinvent v3 is scheduled for June 2012.

 

How long does the review process take?

There are several editors involved in the review: a main editor responsible for the industrial sector the dataset is in, and several cross-cutting editors that look at certain issues such as emissions, land use or conditions in specific countries. The minimum time between submission of a dataset and its inclusion in the working database is around one month, but a more reliable estimation should be more conservative. It is advisable to consider an period of 18 weeks for datasets without major flaws. Naturally, flawed datasets will need revisions, increasing the length of the review.

 

When do I have to submit my datasets to be included in v3.0?

If you want your datasets to be published in the first release of ecoinvent version 3, you should consider submitting your data at least 18 weeks before the release. The exact length of the review process depends on the availability of our editors and the quality of your submission. If your datasets are accepted by the editors, but not on time for being included in the first release of ecoinvent v3, they will be included in further release of the database.

 

How long does it take to create a dataset?

We cannot say how long data collection will take as this varies based on the complexity of the dataset, the experience of the author and the quality of source materials. However, based on the experience of our long-standing data providers, one week per dataset is often used as a rough estimate. This can vary with the scope of the dataset and possible synergies in large data collection projects. Data submission via the free software ecoEditor should not take more than a few hours for most datasets.

 

Who should create the dataset?

The dataset for submission, including documentation, needs to be prepared by the author of the dataset. For this purpose, the ecoinvent centre provides the ecoEditor software free of charge, available on the ecoinvent website. The author holds the copyright for the dataset, but does not have to be the person who collected the data – datasets can be based on published information from third parties.

 

Can I ask the ecoinvent centre to convert my raw data into ecoinvent datasets?

No, the ecoinvent centre does not accept raw, unformatted data - all submissions to ecoinvent must follow the ecoinvent quality guidelines; this should be strived for by the dataset author. If you are unwilling to do this work yourself, ecoinvent can recommend qualified experts who may be willing to perform this task for you if you can cover the costs.

 

Can the review take place while the data are gathered?

The review begins only when the dataset is submitted to ecoinvent. Editors check not just the consistency of the data but also the quality of the documentation and preparation of the dataset according to the data quality guidelines, thus the dataset needs to be finalized before the review can begin.

 

Do I have to pay for the review? Or submitting the dataset?

No, submitting to ecoinvent version 3 is free of charge and the review is financed completely by the ecoinvent centre.

 

Who will decide whether the data are published or not?

All submissions are reviewed by at least 3 members of the ecoinvent editorial board. Only if all agree that the submission is acceptable will the dataset be published in the ecoinvent database.

 

What is a unit process?

Ecoinvent is based on unit processes. A unit process represents a specific human activity, as opposed to a system process, which describes the whole product system from cradle to gate. Examples of unit processes are the assembly of a car, transporting goods for a certain distance or incinerating waste. Datasets for submission to the ecoinvent database should be on a level of detail where each unit process describes the transformation of one or more marketable products into others. Further disaggregation is not necessary (i.e. it is not necessary to divide a process into singe datasets for each individual operation).

For more information on this concept, please consult chapter 5.1 of the data quality guidelines.

 

Can I provide aggregated datasets?

As far as possible, unit processes should be aggregated neither horizontally (over different production sites) nor vertically (over several production steps). However, some level of aggregation may be acceptable if more detailed data are not available or confidential.

 

Can datasets be confidential?

Although transparency is a central element of the ecoinvent database, confidentiality can be maintained by e.g. combining production data from several companies through industry associations or through limited vertical aggregation. In special cases where the content would significantly improve the quality of the ecoinvent database, there is an option for the unit process to be kept confidential. Such datasets are still reviewed and used on a unit process level for calculation of system datasets but database users will only be able to see and use the LCI results of the aggregated system. Such datasets may be replaced by transparent data at a later time if such data become available.

 

Can the dataset include the name of my company or product brand?

Most ecoinvent datasets are for generic products, not for specific brands. It is possible to submit branded datasets, but this requires additional review steps which need to be financed by the authors. If you are interested in branded datasets, please see chapter 11.3 of the data quality guidelines.

 

How complete should the dataset be?

The dataset should cover all known material, energy and product inputs as well as all known outputs of products, recyclable materials, wastes and emissions. The level of detail is important – at least all flows that are suspected to be environmentally relevant need to be included. In many cases, existing ecoinvent datasets on similar products can serve as an example on the level of detail expected. There are also editors for different sectors who can provide information on what information datasets in a specific field should contain. Check the ecoinvent documentation for more guidance on this.

 

What should the documentation contain?

The dataset should contain documentation explaining the activity described, the products and the quality of the data. The origins of all data points should be clearly declared and need to be provided with references where applicable. The data format provides comment fields for all data points and for general comments for this purpose. Please see the data quality guidelines chapter 7 for more information on good practice in documentation.

 

Do I have to provide primary sources, e.g. measurement data or production documents?

No, you do not. The review checks the consistency and plausibility of the dataset and the correct use of sources, but actual measurement data or primary sources are not audited. Only branded datasets are audited for data accuracy.

 

Does ecoinvent want copyright to my data?

No, you keep the copyright to your data. Upon submission, ecoinvent asks you to confirm that you are the copyright holder, and asks you to grant ecoinvent the non-exclusive rights to use the dataset. This includes the rights to publish,sell, update and distribute the data, as required for the operation of the ecoinvent database. .

 

Can I update my dataset later?

Yes, and in fact this is welcomed. Updates are subjected to the standard review process to maintain data quality.

 

Can I completely remove the dataset later if I no longer feel it is appropriate?

As datasets are linked within the ecoinvent database, removing existing datasets is difficult from a technical standpoint. It is therefore not possible to simply remove the dataset. Rather, an inappropriate dataset should be updated or replaced by a more appropriate version. Withdrawing a dataset completely without update or replacement is only possible if the replacement costs for the ecoinvent centre are covered.

 

Will my dataset change over time?

The unit process datasets are not changed automatically. Only justified updates that have passed the ecoinvent review will change unit processes. As long as you want to be responsible for a dataset, you will be able to update it manually. If you do not maintain your submissions, it is possible that other experts provide improved or more recent data that could replace your dataset.

 

Will the LCA results for my dataset change over time?

Since almost all processes are indirectly linked to a large amount of other processes, updating any dataset will slightly affect the results for all other datasets. For this reason, LCA results for all products slightly change with each new release of the ecoinvent database as new and updated datasets are included. This is a normal consequence of the growing availability of inventory data world-wide. This also means that LCA results for your submissions will always use the most current background data available. 

 

If I submit datasets to ecoinvent, do I get a free license?

All dataset providers receive a free ecoinvent license for each 25 original datasets (i.e. datasets not largely extrapolated from datasets already existing in the ecoinvent database) accepted for publication.