Scholarlink

<p>Life Cycle Assessment (LCA) models are key for supporting decision-makers in reducing the environmental impact of products, but their results are inevitably affected by uncertainty. This study compares the performance of two background databases, namely, exiobase and ecoinvent, in terms of uncertainty of LCA results. Data uncertainty in exiobase was estimated using two approaches: assuming the same uncertainty across the database and using the pedigree matrix method. The model uncertainty arising when foreground activities are matched with background datasets was as well estimated using the pedigree matrix method. An emerging wastewater treatment technology is used as a case study. Ad hoc simulations were designed in the Brightway python environment to perform stochastic error propagation via Monte Carlo simulation on both databases and obtain climate impact scores. Results confirm a proportional relationship between input and output uncertainty in exiobase. The LCA results and uncertainties calculated for the wastewater treatment technology via exiobase (median 0.146 million tonnes CO<sub>2</sub>-eq, 5th–95th percentile: 0.140–0.150), are comparable to those calculated via ecoinvent (0.148 million tonnes CO<sub>2</sub>-eq.&#xa0;(5th–95th percentile: 0.143–0.152). It was thus excluded that, when using the pedigree matrix approach to estimate uncertainty, using one database leads to more uncertain results than using the other.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Uncertainty propagation of input–output and process-based life cycle inventories at different aggregation levels

  • 期刊论文

  • 发表日期：

    2026年7月4日

查看全文
  • Elisabetta Pigni,
  • Ning An,
  • Serena Righi,
  • Diego Marazza,
  • Enrico Balugani,
  • Massimo Pizzol

摘要

Life Cycle Assessment (LCA) models are key for supporting decision-makers in reducing the environmental impact of products, but their results are inevitably affected by uncertainty. This study compares the performance of two background databases, namely, exiobase and ecoinvent, in terms of uncertainty of LCA results. Data uncertainty in exiobase was estimated using two approaches: assuming the same uncertainty across the database and using the pedigree matrix method. The model uncertainty arising when foreground activities are matched with background datasets was as well estimated using the pedigree matrix method. An emerging wastewater treatment technology is used as a case study. Ad hoc simulations were designed in the Brightway python environment to perform stochastic error propagation via Monte Carlo simulation on both databases and obtain climate impact scores. Results confirm a proportional relationship between input and output uncertainty in exiobase. The LCA results and uncertainties calculated for the wastewater treatment technology via exiobase (median 0.146 million tonnes CO2-eq, 5th–95th percentile: 0.140–0.150), are comparable to those calculated via ecoinvent (0.148 million tonnes CO2-eq. (5th–95th percentile: 0.143–0.152). It was thus excluded that, when using the pedigree matrix approach to estimate uncertainty, using one database leads to more uncertain results than using the other.