Goal, Scope and BackgroundThe main aim of this paper is to present some methodological considerations concerning existing methods used to assess quality of the LCA study. It relates mainly to the quality of data and the uncertainty of the LCA results. The first paper is strictly devoted to methodological aspects whereas, the second is presented in a separate article (Part II) and devoted mainly to a case study.MethodsThe presented analysis is based on two well-known concepts: the Data Quality Indicators (DQIs) and the Pedigree Matrix. In the first phase, the Sensitivity Indicators are created on the basis of the sensitivity analysis and then linked with the DQIs and the Quality Classes. These parameters indicate the relative importance of input data and their theoretical quality levels. Next, the Weidema’s Pedigree Matrix (slightly modified) is used to establish the values of the new parameter called the Data Quality Distance (DQD) and to link them with the DQIs and Quality Classes. This way the information about the “real” quality levels is provided. Further analysis is performed using the probabilistic distributions and Monte Carlo simulations.Results and DiscussionThanks to this approach it is possible to make a comparison between two types of the quality factors. On the one hand, the sensitivity analysis allows one to check the importance of input data and to determine their required quality. It is done according to the following relation: the higher the sensitivity indicator, the higher the importance of input data and the higher quality should be demanded. On the other hand the data have a certain real quality, not always in accord with the demanded one. To make possible a comparison between these two types of quality, it is necessary to find and develop a common denominator for them. Here, for this purpose the DQIs and Quality Classes are used.ConclusionsIn the further stage of the assessment the DQIs are used to perform the uncertainty analysis of the LCA results. The results could be additionally analysed by using other techniques of interpretation: the sensitivity-, the contribution-, the comparative-, the discernability- and the uncertainty analysis.Recommendations and OutlookThe presented approach is put into practice to conduct the comparative LCA study for the industrial pumps by using the Ecoindicator99 method. Thanks to this, complex analysis of the credibility of the results is carried out. As a consequence, uncertainty ranges for the LCA results of every product system can be determined [1].
[1]
David Vose,et al.
Quantitative Risk Analysis: A Guide to Monte Carlo Simulation Modelling
,
1996
.
[2]
Henri Lecouls,et al.
ISO 14043: Environmental management · life cycle assessment · life cycle interpretation
,
1999
.
[3]
David Vose,et al.
Monte Carlo Risk Analysis Modeling
,
1996
.
[4]
Raul Carlson,et al.
Guideline for collection, treatment and quality documentation of LCA data
,
2004
.
[5]
Anna Björklund,et al.
Survey of approaches to improve reliability in lca
,
2002
.
[6]
Mark A. J. Huijbregts,et al.
Framework for modelling data uncertainty in life cycle inventories
,
2001
.
[7]
Gjalt Huppes,et al.
Quality assessment for LCA
,
2000
.
[8]
Bo Pedersen Weidema,et al.
Data quality management for life cycle inventories—an example of using data quality indicators☆
,
1996
.
[9]
Mark A. J. Huijbregts,et al.
Application of uncertainty and variability in LCA
,
1998
.
[10]
R. Kleijn,et al.
Numerical approaches towards life cycle interpretation five examples
,
2001
.
[11]
Douglas C. Montgomery,et al.
Screening stochastic Life Cycle assessment inventory models
,
2002
.