developed a method to create Evolving Models for the Environmental Evaluation of
Complex Systems (EMEECS). EMEECS are supported by Hierarchical
Agglomerative Clustering (HAC) [17] for processing vehicles’ available LCA
results followed by an algorithm for creating optimal clusters relative to a fixed
uncertainty threshold.
For building EMEECS, available results of previous LCA are collected over a
fixed number of life cycle stages and of environmental indicators. The life cycle
stages are:
• Manufacturing: including raw materials extraction, manufacturing processes;
excluding upstream and downstream logistics; the assembly plants are considered separately;
• Use: including well-to-wheels fuel consumption, tank-to-wheels emissions of
CO 2 and pollutants (CO, NOx, HC), maintenance;
• End of life: no credit has been taken into account because of recycled material
provision.
Four environmental indicators are included in the study. These indicators are
generally used by car manufacturers to communicate, or by scientific papers in the
automotive sector. They are calculated using the CML 2001 method:
• Global warming potential, GWP, [kg CO 2 -Equivalent];
• Eutrophication potential, EP, [kg Phosphate-Equivalent];
• Photochemical ozone creation potential, POCP, [kg Ethen-Equivalent];
• Abiotic depletion potential, ADP, [kg Sb-Equivalent].
LCAs are processed with a HAC in order to classify the vehicles’ environmental
indicators values in dendrograms per life cycle stage and per environmental impact
category. HAC is chosen because it is one of the most common methods to make
clusters from a statistical population. As a result, each dendrogram is processed for
extracting the “optimal” clusters, i.e. clusters whose error relative to real values of
vehicles within the cluster is inferior to a fixed threshold. Each cluster has a value
(the average value of vehicles’ environmental indicators that form the cluster) and
an uncertainty (the standard deviation of those same values). The EMEECS are
environmental impact archetypes, calculated on several life cycle stages, of complex system clusters, i.e. the user of the proposed method would directly use the
environmental impacts values of the initial and innovative solutions, and of the
global system.
5.2 Case Studies: Micro-Hybridization Systems
TEEPI was tested on two micro-hybridization systems: fuel-electricity (Inno_1) and
fuel-compressed air (Inno_2). The reduction in fuel consumption is achieved
through supplying energy, complementary to traditional fossil energy to move the
376
J. Garcia et al.
Complex Systems (EMEECS). EMEECS are supported by Hierarchical
Agglomerative Clustering (HAC) [17] for processing vehicles’ available LCA
results followed by an algorithm for creating optimal clusters relative to a fixed
uncertainty threshold.
For building EMEECS, available results of previous LCA are collected over a
fixed number of life cycle stages and of environmental indicators. The life cycle
stages are:
• Manufacturing: including raw materials extraction, manufacturing processes;
excluding upstream and downstream logistics; the assembly plants are considered separately;
• Use: including well-to-wheels fuel consumption, tank-to-wheels emissions of
CO 2 and pollutants (CO, NOx, HC), maintenance;
• End of life: no credit has been taken into account because of recycled material
provision.
Four environmental indicators are included in the study. These indicators are
generally used by car manufacturers to communicate, or by scientific papers in the
automotive sector. They are calculated using the CML 2001 method:
• Global warming potential, GWP, [kg CO 2 -Equivalent];
• Eutrophication potential, EP, [kg Phosphate-Equivalent];
• Photochemical ozone creation potential, POCP, [kg Ethen-Equivalent];
• Abiotic depletion potential, ADP, [kg Sb-Equivalent].
LCAs are processed with a HAC in order to classify the vehicles’ environmental
indicators values in dendrograms per life cycle stage and per environmental impact
category. HAC is chosen because it is one of the most common methods to make
clusters from a statistical population. As a result, each dendrogram is processed for
extracting the “optimal” clusters, i.e. clusters whose error relative to real values of
vehicles within the cluster is inferior to a fixed threshold. Each cluster has a value
(the average value of vehicles’ environmental indicators that form the cluster) and
an uncertainty (the standard deviation of those same values). The EMEECS are
environmental impact archetypes, calculated on several life cycle stages, of complex system clusters, i.e. the user of the proposed method would directly use the
environmental impacts values of the initial and innovative solutions, and of the
global system.
5.2 Case Studies: Micro-Hybridization Systems
TEEPI was tested on two micro-hybridization systems: fuel-electricity (Inno_1) and
fuel-compressed air (Inno_2). The reduction in fuel consumption is achieved
through supplying energy, complementary to traditional fossil energy to move the
376
J. Garcia et al.