2009 年 5 巻 4 号 p. 529-541
Objective. To attain higher efficiency in expediting LCA researches, it is inevitable to utilize background data wherever environmental load is considered small. For this reason, development of background database should be an issue of utmost urgency; but in fact, much inventory data collection still has to be done. Presently, background database are often found built up on the ground of inventory data that are inconsistent in data origins; which may affect reliability of thus-developed database. If not enough careful consideration is given to this fact, LCA studies can most likely result in serious misinterpretation. Data quality management of background data is, therefore, indispensable for every LCA researcher when self-evaluation or verification of his/her LCA study is sought after. Objective of our study was to develop a practical method that is useful for evaluation/verification of the quality of various exiting background data. Coping with the limited availability of information on data quality studies, priority was given to developing a practical method that could help an easy and general evaluation of background data quality.
Results and Discussion.
1) Factors required for more reliable background data were selected referring to the data quality requirements of ISO for inventory data. Then problems in existing data quality were discussed. ISO’s requirements of inventory data are coverage, completeness and representativeness of the data, consistency and reproducibility of the method, sources of the data, uncertainty of the information and so on. In contrast, many factors of existing data like consistency, reproducibility, uncertainty were unknown for lack of data information.
2) The method of data quality management for background data was proposed. In this method, background data from each data source is evaluated in quality from comprehensiveness of input data and range of coverage. The former is classified into 5 levels, the latter is classified into 4.
3) More than 900 existing data were evaluated using the method, and the overall quality of existing data was profiled.
4) Future assignments in management of background data quality were discussed based on the study above. Efforts are to be made to record information related to data quality, to make a set of rules to improve data quality and so on.
Conclusions. The method of data quality evaluation for management of existing background data was developed, so that can evaluate the quality broadly and easily. This will help select most suitable data and make more accurate LCA data analyses.