Purpose A review of LCA process datasets is an important element of quality assurance for databases and for other systems to provide LCA datasets. Somewhat surprisingly, a broadly accepted and applicable set of criteria for a review of LCA process datasets was lacking so far. Different LCA databases and frameworks are proposing and using different criteria for reviewing datasets. To close this gap, a set of criteria for reviewing LCA dataset has been developed within the Life Cycle Initiative. Methods Previous contributions to LCA dataset review have been analysed for a start, from ISO and various LCA databases. To avoid somewhat arbitrary review criteria, four basic rules are proposed which are to be fulfilled by any dataset. Further, concepts for assessing representativeness and relevance are introduced into the criteria set from established practices in statistics and materiality. To better structure the criteria and to ease their application, they are grouped into clusters. A first version of the developed review criteria was presented in two workshops with database providers and users on different levels of experience, and draft versions of the criteria were shared within the initiative. The current version of the criteria reflects feedback received from various stakeholders and has been applied and tested in a review for newly developed datasets in Brazil, Malaysia and Thailand. Results and discussion Overall, 14 criteria are proposed, which are organised in clusters. The clusters are goal, model, value, relevance and procedure. For several criteria, a more science-based definition and evaluation is proposed in comparison to ‘traditional’ LCA. While most of the criteria depend on the goal and scope of dataset development, a core set of criteria are seen as essential and independent from specific LCA modelling. For all the criteria, value scales are developed, typically using an ordinal scale, following the pedigree approach. Conclusions Review criteria for LCI datasets are now defined based on a stringent approach. They aim to be globally acceptable, considering also database interoperability and database management aspects, as well as feedback received from various stakeholders, and thus close an important gap in LCA dataset quality assurance. The criteria take many elements of already existing criteria but are the first to fully reflect the implications of the ISO data quality definition, and add new concepts for representativeness and relevance with the idea to better reflect scientific practice outside of the LCA domain. A first application in a review showed to be feasible, with a level of effort similar to applying other review criteria. Aspects not addressed yet are the review procedure and the mutual recognition of dataset reviews, and their application for a very high number of datasets.

中文翻译：

---

生命周期清单数据集审查标准——新提案

目的 LCA 过程数据集的审查是数据库和其他提供 LCA 数据集的系统质量保证的重要元素。有点令人惊讶的是，迄今为止缺乏一套被广泛接受和适用的 LCA 过程数据集审查标准。不同的 LCA 数据库和框架提出并使用不同的标准来审查数据集。为了弥补这一差距，生命周期计划中制定了一套审查 LCA 数据集的标准。方法 从 ISO 和各种 LCA 数据库开始分析以前对 LCA 数据集审查的贡献。为了避免有些武断的审查标准，提出了任何数据集都必须满足的四个基本规则。更多，评估代表性和相关性的概念被引入根据统计和重要性的既定做法设定的标准中。为了更好地构建标准并简化它们的应用，它们被分组到集群中。已制定的审查标准的第一个版本在两个研讨会上与具有不同经验水平的数据库提供商和用户进行了介绍，并在该倡议内共享了标准的草案版本。当前版本的标准反映了从各个利益相关者收到的反馈，并已在巴西、马来西亚和泰国新开发的数据集的审查中得到应用和测试。结果和讨论 总体而言，提出了 14 条标准，这些标准按集群组织。集群是目标、模型、价值、相关性和程序。对于几个标准，与“传统” LCA 相比，提出了更基于科学的定义和评估。虽然大多数标准取决于数据集开发的目标和范围，但核心标准集被视为必不可少且独立于特定 LCA 建模。对于所有标准，按照谱系方法开发价值量表，通常使用序数量表。结论 LCI 数据集的审查标准现在是基于严格的方法定义的。他们的目标是全球可接受，同时考虑数据库互操作性和数据库管理方面，以及从各个利益相关者收到的反馈，从而缩小 LCA 数据集质量保证方面的重要差距。该标准采用了现有标准的许多元素，但首先充分反映了 ISO 数据质量定义的含义，并添加具有代表性和相关性的新概念，以更好地反映 LCA 领域之外的科学实践。审查中的第一个申请被证明是可行的，其努力程度类似于应用其他审查标准。尚未解决的方面是审查程序和数据集审查的相互认可，以及它们对大量数据集的应用。