Estimating Missing Unit Process Data in Life Cycle Assessment Using a Similarity-Based Approach.

Area of Science:

• Environmental Science
• Computational Science
• Industrial Ecology

Background:

• Life cycle assessment (LCA) relies on unit process data, which is often expensive and time-consuming to collect empirically.
• Existing methods for compiling LCA data face challenges due to data gaps and collection costs.

Purpose of the Study:

• To develop and validate a novel computational approach for estimating missing unit process data in LCA.
• To leverage similarity-based link prediction for efficient LCA data compilation.

Main Methods:

• A similarity-based link prediction method was employed to estimate missing unit process data.
• The ecoinvent 3.1 dataset was utilized, split into training and testing sets.
• Missing data points were estimated using similarity-weighted means from the training set and compared against original values.

Main Results:

• The proposed method accurately estimates missing unit process data when less than 5% of data is missing within a process.
• Estimation accuracy decreases as the proportion of missing data increases.
• The computational approach demonstrates effectiveness in addressing data gaps in LCA.

Conclusions:

• A novel computational method offers a viable solution for estimating missing LCA unit process data.
• Similarity-based link prediction shows significant potential for improving the efficiency and accuracy of LCA data compilation.
• This approach can reduce the cost and time associated with empirical data collection in LCA studies.