lable at ScienceDirect

Renewable Energy 74 (2015) 237e246

Contents lists avai

Renewable Energy

journal homepage: www.elsevier .com/locate/renene

Comparative evaluation of life cycle impact assessment software toolsthrough a wind turbine case study

E. Martínez a, *, J. Blanco a, 1, E. Jim�enez b, 2, J.C. Saenz-Díez b, 3, F. Sanz a, 1

a Department of Mechanical Engineering, University of la Rioja, Edificio Departamental, C/Luis de Ulloa, 20, 26004 Logro~no, La Rioja, Spainb Department of Electrical Engineering, University of la Rioja, Edificio Departamental, C/Luis de Ulloa, 20, 26004 Logro~no, La Rioja, Spain

a r t i c l e i n f o

Article history:Received 20 January 2014Accepted 1 August 2014Available online

Keywords:Wind energyLCALCIARenewable energy

* Corresponding author. Tel.: þ34 941 299 524.E-mail addresses: e.camara@eolicas.net (E. Martí

(J. Blanco), emilio.jimenez@unirioja.es (E. Jim�enunirioja.es (J.C. Saenz-Díez), felix.sanz@unirioja.es (F.

1 Tel.: þ34 941 299 524.2 Tel.: þ34 941 299 502.3 Tel.: þ34 941 299 489.

http://dx.doi.org/10.1016/j.renene.2014.08.0040960-1481/© 2014 Elsevier Ltd. All rights reserved.

a b s t r a c t

This paper seeks to analyse the differences between environmental impact assessment software tools byexamining the results that they give when applied to a multi-megawatt wind turbine.

Seven different life cycle impact assessment software tools are compared: CML 2001, Eco-indicator 99,Ecopoints 97, EDIP, EPS 2000, IMPACT2002 and TRACI.

In Acidification and Eutrophication two groups are found: one includes the results provided by CML,Ecopoints 97, EDIP, EPS and TRACI and the other those of Eco-indicator 99 and Impact2002. In AbioticDepletion all the results are similar except those of the EPS method, which gives negative figures.Likewise in Ozone Layer Depletion the results provided by Ecopoints 97 differ from the rest. In HumanToxicity and Ecotoxicity markedly different results are obtained by each of the LCIAs studied.

In some categories major differences are found between the results provided by the 7 LCIAs examined.Which of the impact assessment software tools currently available in LCA software is chosen is thereforea critical issue. The results provided by the different software tools are not always similar, and this needsto be realised and taken into account when using the resulting data in decision-making processes.

© 2014 Elsevier Ltd. All rights reserved.

1. Introduction

From an environmental viewpoint it is becoming increasinglyimportant to analyse the potential impact of products and actions[1e3]. Life Cycle Assessment (LCA) has become a highly importanttool for providing in-depth analyses of this kind, for instance instudies concerned with the replacement of fossil fuels by renew-ables in electricity production, and a significant option in the pro-cess of transition towards a low-emission production economy[4,5]. In the field of energy there are a great many LCA studiesthat deal with the production and storage of energy. Some of thesestudies are listed in Table 1. Each study listed uses just one of thevarious Life Cycle Impact Assessment (LCIA) software toolscurrently available, but there is seldom any discussion or

nez), julio.blanco@unirioja.esez), juan-carlos.saenz-diez@Sanz).

comparison with other software tools. There are several articles onthis topic (see Table 2), but they tend to concentrate on specific casestudies and very few of them deal with wind energy.

For example, Brent and Hietkamp [27] evaluate and comparethe applicability of the five LCIA procedures that have been usedin the South African manufacturing sector for decision and designpurposes, on a qualitative and quantitative basis. Cavalett et al.[29] expand the discussion about how, and how much, the envi-ronmental performance is affected by the use of different LCIA,illustrated by the case study of the comparison between gasolineand ethanol produced from sugarcane in Brazil. And Dreyer et al.[31] develop a quantitative comparison of the CML 2001 andEDIP97 performed on the characterised indicator scores and onthe normalised scores, while some more qualitative pointsregarding the use of midpoint and endpoint software tools areillustrated by a comparison of the weighted Eco-indicator 99 andEDIP97 results.

Non-specialists and persons with only a passing knowledge ofLCA often ask why different results may be obtained when differentLife-Cycle Impact Assessment (LCIA) software tools are used, andwhether the degree of uncertainty in results is high enough [41] tocast doubt on the scientific arguments put forward.

Table 1Some LCA studies about energy production and storing.

Authors Journal Methodology Field

Barba-Guti�errez, Y. et al. [6] Environmental Modeling and Assessment Ecopoint 97 Eco-efficiencyBravo, Y. et al. [7] Solar Energy CML/Eco-indicator 99 Power generationChester, M. et al. [8] Environmental Research Letters TRACI TransportationCoventry, Z.A. et al. [9] International Conference on Thermal Treatment Technologies

and Hazardous Waste Combustors 2012TRACI Waste to energy technologies

Desideri, U. et al. [10] Applied Energy Eco-indicator 99 Solar powerDufour, J. et al. [11] International Journal of Hydrogen Energy Eco-indicator 95 Hydrogen productionEsteban, B. et al. [12] Biomass and Bioenergy CML/EDIP Bio fuelsGebreslassie. B.H. et al. [13] Computers & Chemical Engineering Eco-indicator 99 Bio fuelsGonz�alez-García, S. et al. [14] Science of The Total Environment CML BiomassHajjaji, N. et al. [15] Journal of Cleaner Production CML/Eco-indicator 99 Hydrogen productionMartínez-Gonz�alez, A. et al. [16] CTyF e Ciencia, Tecnologia y Futuro IMPACT2002 FuelsMartinez, E. et al. [17] International Journal of Life Cycle Assessment CML Wind powerMartinez, E. et al. [18] Renewable Energy Eco-indicator 99 Wind powerMiller, V.B. et al. [19] Renewable Energy TRACI Hydrokinetic energyModahl, I.S. el al. [20] International Journal of Life Cycle Assessment EPS Fossil gas powerNishimura, A. et al. [21] Applied Energy NETS Solar powerPawelzik, P.F. and Zhang, Q. [22] Biomass and Bioenergy Eco-indicator 95 Bio fuelsP�erez-Fortes, M. et al. [23] Computer Aided Chemical Engineering IMPACT2002 Hydrogen and biomassTalens Peir�o, L. et al. [24] Energy CML Bio fuelsTorres, C.M. et al. [25] Fuel Eco-indicator 99 Bio fuelsWang, B. et al. [26] Computers & Chemical Engineering Eco-indicator 99 Gasification

E. Martínez et al. / Renewable Energy 74 (2015) 237e246238

The research described here seeks to examine the differencesbetween various environmental impact assessment software toolsbased on the results obtained when they are applied to a multi-megawatt wind turbine. These environmental impact assessmentsoftware tools applied on renewable energies are commonly usedto analyze, compare, and make decisions about the merits of suchenergy sources over the conventional ones. For this reason it isimportant, from the renewable energy point of view, to know inmore detail the possible effect that can be obtained on the finalresult of the analysis the use of the different environmental impactassessment software tools. In this paper seven different LCIA soft-ware tools are compared: CML 2001 [42], developed by the Insti-tute of Environmental Sciences of Leiden University; Eco-indicator99 (EI99), the first endpoint impact assessment software tool whichallowed the environmental load of a product to be expressed in asingle score [43]; Ecopoints 97, developed by BUWAL (the SwissAgency for the Environment), one of the earliest systems for impactassessment with a single score [44]; EDIP, developed by the Centrefor the Environmental Design of Industrial Products (EDIP) in theNetherlands as an upgrade and improvement on CML 92 in severalrespects [45]; EPS 2000 (Environmental Priority Strategies), whichis oriented towards sustainable product development [46];

Table 2Comparative studies of LCIA.

Authors Journal

Brent, A.C., Hietkamp, S. [27] International Journal of Life Cycle AssessmentBovea, M.D. and Gallardo, A. [28] Materials & DesignCavalett, O. et al. [29] International Journal of Life Cycle AssessmentCarvalho, I.S. et al. [30] Ships and Offshore Structures

Dreyer, L.C. et al. [31] International Journal of Life Cycle AssessmentFantozzi, F. and Buratti, C. [32] Biomass and BioenergyKnoeri, C. et al. [33] International Journal of Life Cycle AssessmentMonteiro, H. and Freire, F. [34] Energy and BuildingsPant, R. et al. [35] Journal of Life Cycle AssessmentRenou, S. et al. [36] Journal of Cleaner ProductionSim~oes, C.L. et al. [37] Waste Management and ResearchValderrama, C. et al. [38] Journal of Cleaner ProductionVan Caneghem, J. et al. [39] Journal of Hazardous MaterialsVan Caneghem, J. et al. [40] Resources, Conservation and Recycling

IMPACT2002, which proposes a feasible implementation of acombined midpoint/damage approach, linking all types of life cycleinventory results via 14 midpoint categories [47]; and TRACI (Toolfor the Reduction and Assessment of Chemical and other environ-mental Impacts) for environmental impact factors, developed bythe Environmental Protection Agency [48]. Among the differentexisting LCIA software tools these seven have been chosen as arepresentative sample of the most used software tools in the sci-entific literature to date (Tables 1 and 2). Table 3 shows the numberof LCA studies using the different software tools in those references.Based on these results, the software tools that appear more thanonce have been selected, with the exception of Eco-indicator 95,which is not included in the study due to the selection of Eco-indicator 99.

2. Method

A brief outline is given below of each of the LCIA software toolsexamined, the impact categories included in the study, the envi-ronmental impact and effects of recycling wind turbines at the endof their useful lifetimes and a sensitivity analysis for the case in handto determine the impact of the variations in the parameters used.

Methodology Field

CML/Eco-indicator 99/EPS ProductionEDIP/CML/EPS/Eco-indicator 99/Eco-indicator 95 Eco-designCML/IMPACT2002/EDIP/Eco-indicator 99/TRACI Bio fuelsCML/IMPACT2002/Eco-indicator 99 Dismantling activity

and recycling.EDIP/CML/Eco-indicator 99 ProductionEPS/EDIP/Eco-indicator 99 BiomassEcoindicator 99/Ecological scarcity 2006 RecycledCML/Eco-indicator 99 BuildingCML/IMPACT2002/EDIP ProductionCML/Eco-Indicator 99/EDIP/EPS/Ecopoints 97 Waste water treatmentEco-indicator 99/CML/EPS/, Eco-indicator 95/EDIP RecycledCML/Eco-indicator 99 Sewage sludge valorisationCML/Eco-indicator 99/EPS/EDIP and USEtox Emitted to airCML/CExD/EPS/Eco-indicator 99 Production

Table 3Number of LCA studies using the different software tools.

Software tool Number

Eco-indicator 99 20CML 18EDIP 9EPS 7IMPACT2002 5TRACI 4Eco-indicator 95 4Ecopoints 97 2Ecological scarcity 2006 1USEtox 1CExD 1NETS 1

E. Martínez et al. / Renewable Energy 74 (2015) 237e246 239

2.1. LCA of a wind turbine

This research analyses a doubly-fed induction generator (DFIG)wind turbine; specifically the unit in question is a model G8XGamesa onshore wind turbine, with 2 MW rated power and thefollowing general dimensions: rotor blade 80 m, swept area5,027 m2 and height 70 m. The wind turbine model under study isconsidered representative because Gamesa is one of the largestmanufacturers of wind turbines worldwide. Furthermore, the windturbines of similar size from other major manufacturers, such asVestas and GE Wind, present similar technology and generalcharacteristics to the G8X Gamesa models.

The work focuses on compiling the LCI data on importantcomponents such as the base, the tower, the nacelle and the rotor.The materials and energy used in the various components wereincorporated into the model using data provided by Gamesa andthe commercial databases of SimaPro [18]. During the operationalphase, all the maintenance operations have been taken into ac-count. These maintenance operations are performed by theowner company of the wind farm and they are recorded in itsenvironmental management system according to the ISO 14001standard [18].

Table 4 shows a summary of the components, materials, andenergy considered for the LCA.

Table 4Key parameters of the life cycle inventory [18].

Component Materials Energy

Rotor 11.7 t resin 20,15 MWh7.8 t fibre glass 12 MWh14 t cast iron 0.95 MWh0.124 t fibre glass 0.4 MWh0.186 t resin 17,000 MJ

Foundation 700 t concrete 170,000 MJ25 t iron 9 MWh15 t steel 5.3 MWh

Tower 143 t steel 200,000 MJ

Nacelle 10.5 t iron 265,000 MJ6.1 t steel 495,000 MJ0.149 t silica 6.2 MWh1.5 t copper 20,15 MWh3.3 t steel 12 MWh0.195 t silica 0.95 MWh2 t copper 0.4 MWh4.29 t steel 17,000 MJ8 t iron 170,000 MJ8 t steel 9 MWh0.8 t fibre glass 5.3 MWh1.2 t resin 200,000 MJ

Because of cost limitations, the LCA was conducted with thefollowing assumptions:

� A cut-off point based on component weight: between them allthe components excluded account for no more than 5% of thebase, 5% of the tower and 15% of the nacelle and rotor.

� Data on electricity were taken from the SimaPro database[49,50].

� The recycling rate assigned for the turbine LCA was calculatedon the basis of the wind farm dismantling projects handled bythe company GER (Grupo E�olicas Riojanas).

� The turbine's useful lifetime was taken as 20 years.� The annual output of the turbine was taken as 4 GWh. Theestimation of 2000 equivalent hours of production per year hasbeen selected because this is the value established by promotersand developers to determine the economic viability of a windfarm. From this data it is possible to obtain directly the result ofthe environmental impact of energy production per kWh, usefulto comparisons with other technologies or studies.

� One generator replacement in the course of the 20-year usefullifetime of the turbine was factored into the calculations.

2.2. Sensitivity analysis

Due to the uncertainty involved in various parameters of theLCA study, a sensitivity analysis was conducted to gauge the effectsof potential variations in environmental impact results for theturbine depending on the type of LCA used. The scenarios consid-ered were the following:

� Alternative scenario AS1: Increase in corrective maintenanceoperations. This scenario includes an additional replacement ofthe set of blades and the gearbox. The use to which these newcomponents are put at the end of their useful lifetime is alsospecified. Two possibilities for end-of-life treatment areproposed:

B Option AS12: The criteria applied are the same as for the restof the wind turbine, i.e. the metal parts are recycled and thefibre glass blades are landfilled.

B Option AS12: 50% of the material obtained from thereplacement is reused and the other 50% is recycled.

� Alternative scenario AS2: Increased consumption. Two possi-bilities are proposed:

B Option AS21: 10% increase in consumption of materials.

B Option AS22: 10% increase in consumption of energy.� Alternative scenario AS3: 50% reduction in recycling of materialscompared to the baseline scenario.

� Alternative scenario AS4: recycling of 80% of the compositematerial used in the blades.

2.3. LCIA software tools

7 different LCIA software tools were examined:

� CML 2 baseline 2000 V2.03/World, 1990� Eco-indicator 99 (E) V2.03/Europe EI 99 E/E� Ecopoints 97 (CH) V2.03/Ecopoints� EDIP/UMIP 97 V2.03/EDIP World/Dk� EPS 2000 V2.02/EPS� IMPACT2002þ V2.02/IMPACT2002þ� TRACI V2.00

E. Martínez et al. / Renewable Energy 74 (2015) 237e246240

SimaPro 7.0 software was used to apply the various LCIA soft-ware tools to the Life Cycle Inventory (LCI).

The impact categories involving the ecosystem and toxicitywere taken as the reference points for comparing the results pro-vided by each software tool. The former comprise acidification,nutrient enrichment (eutrophication), Global Warming (climatechange), Abiotic Depletion and the ozone layer; the latter compriseHuman Toxicity and Ecotoxicity.

As might be expected, not all the impact assessment softwaretools examined used these exact categories, or indeed others thatwere directly comparable. In these cases an effort was made toachieve as close an approximation as possible within the differentimpact categories available in each software tool. If it provedimpossible to find one or more comparable impact categories, theLCIA in question was ruled out of the comparative analysis.

Table 5Names of impact categories, abbreviations used and units of measurement.

Name Abbreviation Units

Abiotic Depletion AD kg Sb eqGlobal Warming (GWP100) GW kg CO2 eqOzone Layer Depletion (ODP) OLD kg CFC-11 eqHuman Toxicity HT kg 1,4-DB eqFresh water aquatic ecotox. FWAE kg 1,4-DB eqMarine aquatic ecotoxicity MAE kg 1,4-DB eqTerrestrial ecotoxicity TE kg 1,4-DB eqPhotochemical oxidation PO kg C2H4

Acidification AC kg SO2 eqEutrophication EU kg PO4 eqCarcinogens CA DALYRespiratory organics RO DALYRespiratory inorganics RI DALYClimate change CC DALYRadiation RA DALYOzone layer OL DALYEcotoxicity ET PAF*m2 yrAcidification/Eutrophication AE PDF*m2 yrLand use LU PDF*m2 yrMinerals M MJ surplusFossil fuels FF MJ surplusNOx NO gSOx SO g SO2 eq.NMVOC NMVOC gNH3 NH gDust PM10 PM gCO2 CO g CO2 eq.Ozone layer OL g CFC-11Pb (air) PB gCd (air) CD gZn (air) ZN gHg (air) HG gCOD COD gP P gN N gCr (water) CR gZn (water) ZNW gCu (water) CU gCd (water) CDW gHg (water) HGW gPb (water) PBW gNi (water) NI gAOX (water) AOX g Cl-Nitrate (soil) NIT gMetals (soil) MET g Cd eqPesticide soil PS g act. subst.Waste WA gWaste (special) WAS gLMRAD LMRAD cm3

HRAD HRAD cm3

Energy EN MJ LHVGlobal Warming (GWP 100) GW g CO2

Ozone depletion OD g CFC11Acidification AC g SO2

Eutrophication EU g NO3

For the sake of readability, this paper uses abbreviations torefer to the impact categories examined. Table 5 shows the fullnames of the various impact categories along with the abbrevia-tion used for each one and the units in which they are measured.To facilitate the presentation of the data, the units of measurementfor the various impact categories are omitted in the rest of thetables in the paper.

3. Results

The results provided by each impact assessment software toolexamined are shown below, starting with the results for the envi-ronment impact of the multi-megawatt wind turbine used as a casestudy. Then the environmental improvements involved in thedifferent recycling processes in the end-of-life decommissioning

Name Abbreviation Units

Photochemical smog PS g etheneEcotoxicity water chronic EWC m3

Ecotoxicity water acute EWA m3

Ecotoxicity soil chronic ESC m3

Human toxicity air HTA m3

Human toxicity water HTW m3

Human toxicity soil HTS m3

Bulk waste BW kgHazardous waste HW kgRadioactive waste RW kgSlags/ashes SA kgResources (all) RES kgLife Expectancy LE PersonYrSevere Morbidity SM PersonYrMorbidity MO PersonYrSevere Nuisance SN PersonYrNuisance NUI PersonYrCrop Growth Capacity CGC kgWood Growth Capacity WGC kgFish and Meat production FMP kgSoil Acidification SA Hþ eq.Prod. Cap. Irrigation Water PCIW kgProd. Cap. Drinking water PCDW kgDepletion of reserves DR ELUSpecies Extinction SE NEXCarcinogens CA kg C2H3ClNon-Carcinogens NCA kg C2H3ClRespiratory inorganics RI kg PM2.5

Ionising radiation IR Bq C-14Ozone Layer Depletion OLD kg CFC-11Respiratory organics RO kg ethyleneAquatic ecotoxicity AE kg TEG waterTerrestrial ecotoxicity TE kg TEG soilTerrestrial acid/nitri TAN kg SO2

Land occupation LO M2org.arableAquatic acidification AA kg SO2

Aquatic eutrophication AE kg PO4 P-limGlobal Warming GW kg CO2

Non-renewable energy NRE MJ primaryMineral extraction ME MJ surplusGlobal Warming GW CO2 eq.Acidification AC Hþ moles eq.HH Cancer HHC Benzene eq.HH Cancer Ground-Surface HHCGS Benzene eq.HH Cancer Root-Zone HHCRZ Benzene eq.HH Noncancer HHNC Toluene eq.HH Noncancer Ground-Surface HHNCGS Toluene eq.HH Noncancer Root-Zone HHNCRZ Toluene eq.HH Criteria Air-Point Source HHCAPS PM2.5 eq.HH Criteria Air-Mobile HHCAM PM2.5 eq.Eutrophication EU N eq.Ozone Depletion OD CFC-11 eq.Ecotoxicity ET 2,4-D eq.Smog SM NOx eq.

Table 7Reduction of environmental impact due to recycling according to the TRACI LCIA.

Impact category Tower Foundation Rotor Nacelle

GW �3.04Eþ02 �7.79Eþ01 �1.67Eþ01 �2.68Eþ02NRE �3.53Eþ04 �9.61Eþ03 �3.18Eþ03 �1.72Eþ04ME �1.51Eþ03 �4.18Eþ02 �1.46Eþ02 �1.14Eþ03GW �2.59E-01 �6.01E-02 �4.03E-03 �1.15E-01AC �2.58E-01 �6.00E-02 �4.03E-03 �1.15E-01HHC �3.10Eþ07 �8.61Eþ06 �3.01Eþ06 �9.66Eþ06HHCGS �6.31Eþ02 �8.28Eþ01 1.32Eþ02 �3.99Eþ02HHCRZ �1.50Eþ03 �2.20Eþ02 2.64Eþ02 �9.09Eþ02HHNC �7.09Eþ02 �1.96Eþ02 �6.78Eþ01 �2.53Eþ02HHNCGS �7.12Eþ02 �1.97Eþ02 �6.80Eþ01 �2.55Eþ02HHNCRZ �2.17Eþ02 �6.00Eþ01 �2.06Eþ01 �6.45Eþ01HHCAPS �1.95E-03 �4.77E-04 �7.83E-05 �8.44E-04HHCAM �2.94Eþ05 �8.15Eþ04 �2.84Eþ04 �2.34Eþ05EU �3.50Eþ02 �9.32Eþ01 �2.90Eþ01 �1.40Eþ02

E. Martínez et al. / Renewable Energy 74 (2015) 237e246 241

and disposal process of the turbine are described. Finally, the re-sults of the sensitivity analysis are presented for each LCIA.

3.1. Environmental impact

Table 6 compares the overall environmental impact results forthe multi-megawatt wind turbine provided by each different LCIA.These are the results from the characterisation phase of each LCIA.No results are shown for the various normalisation and weightingphases because not all the LCIAs examined include such phases.

3.2. Recycling

In regard to data on the reduction of environmental impact dueto recycling for each of the main components of thewind turbine, itwas decided to include here only the results for the TRACI LCIA(Table 7). These can be taken as an example of the results obtainedfrom all the LCIAs examined.

3.3. Sensitivity analysis

Here also, the sensitivity analysis results for only one LCIA areincluded (Table 8), as an example. In this case the software toolselected is Eco-indicator 99.

4. Interpretation and assessment of results

4.1. Environmental impact

The results for 7 categories of environmental impact arecompared. Table 9 shows the categories selected in each LCIA forthis comparison.

As in all the other comparisons in this paper, these are based onthe relative percentages for the main components obtained with

Table 6Environmental impact of a multi-megawatt wind turbine according to the different LCIA

CML 2 Eco-indicator 99 Ecopoints 97 EDIP

AD 2.94Eþ03 CA 5.86E-02 NO 1.76Eþ06 GW 5.23EGW 5.18Eþ05 RO 7.07E-04 SO 1.81Eþ06 OD 4.12EOLD 4.11E-02 RI 3.61E-01 NMVOC 5.50Eþ05 AC 3.07EHT 4.99Eþ05 CC 1.09E-01 NH 1.25Eþ04 EU 4.25EFWAE 2.14Eþ05 RA 3.09E-03 PM 1.50Eþ05 PS 2.49EMAE 3.24Eþ08 OL 4.33E-05 CO 5.20Eþ08 EWC 4.88ETE 9.55Eþ03 ET 5.91Eþ05 OL 4.23Eþ02 EWA 4.01EPO 1.22Eþ02 AE 1.21Eþ04 PB 1.71Eþ03 ESC �2.5AC 3.05Eþ03 LU 6.60Eþ03 CD 2.28Eþ02 HTA 3.30EEU 4.19Eþ02 M �1.82Eþ03 ZN 3.69Eþ03 HTW 3.47E

FF 4.61Eþ05 HG 2.66Eþ02 HTS 2.52ECOD �5.91Eþ05 BW 1.30EP 1.86Eþ03 HW 4.99EN 6.91Eþ04 RW 2.11ECR 4.14Eþ04 SA 1.19EZNW 1.36Eþ03 RES 2.10ECU 1.61Eþ04CDW �2.61Eþ02HGW 3.64Eþ01PBW 2.80Eþ02NI 2.81Eþ04AOX 6.41Eþ00NIT 8.29Eþ02MET 5.20Eþ00PS 1.73Eþ01WA 1.31Eþ08WAS 0.00Eþ00LMRAD 5.48Eþ03HRAD 1.37Eþ03EN 8.06Eþ06

each impact software tool. This avoids the differences in criteriaused in each software tool to assess the environmental impact foreach category. The reference units used to establish impact tend tovary from one software tool to another, which makes it muchharder to compare their results directly. It was decided that themost advisable method was to compare the relative percentages ofimpact associated with each of the main components of the turbinetaken as our case study.

4.1.1. AcidificationA comparison of the results for this category provided by the

different LCIAs examined can be seen in Fig. 1. As can be observed,they are very similar, though slight differences can be appreciatedbetween the results from the CML, Ecopoints 97, EDIP, EPS andTRACI on the one hand and those from the Eco-indicator 99 andIMPACT2002 on the other. This difference can be explained by thefollowing factors:

s.

EPS IMPACT2002 TRACI

þ08 LE 6.74E-01 CA 3.99Eþ04 GW 1.94Eþ04þ01 SM 1.76E-01 NCA 1.65Eþ04 AC 1.64Eþ05þ06 MO 3.69E-01 RI 5.17Eþ02 HHC 5.91Eþ03þ06 SN 4.98E-01 IR 1.54Eþ07 HHCGS 4.90Eþ00þ05 NUI 1.68Eþ01 OLD 4.52E-02 HHCRZ 4.90Eþ00þ08 CGC 1.76Eþ03 RO 3.24Eþ02 HHNC 7.80Eþ06þ07 WGC �2.47Eþ04 AE �1.96Eþ07 HHNCGS 4.11Eþ049Eþ07 FMP �6.67Eþ01 TE 1.52Eþ07 HHNCRZ 8.75Eþ04þ11 SA 4.85Eþ03 TAN 1.16Eþ04 HHCAPS 6.53Eþ02þ07 PCIW 0.00Eþ00 LO 3.42Eþ03 HHCAM 6.70Eþ02þ05 PCDW 0.00Eþ00 AA 3.07Eþ03 EU 5.23Eþ02þ05 DR 1.91Eþ05 AE 8.58Eþ00 OD 4.52E-02þ02 SE 8.98E-09 GW 4.96Eþ05 ET 9.69Eþ05þ01 NRE 7.88Eþ06 SM 1.77Eþ03þ03 ME 3.52Eþ04þ02

Table 8Sensitivity analysis results according to the Eco-indicator 99 LCIA.

Impact category AS11 AS12 AS21 AS22 AS3 AS4

CA 6.50E-02 6.18E-02 1.67E-01 1.81E-01 1.98E-01 5.80E-02RO 9.93E-04 8.50E-04 7.68E-04 8.08E-04 7.76E-04 6.59E-04RI 5.63E-01 4.62E-01 5.39E-01 5.57E-01 6.69E-01 3.24E-01CC 1.55E-01 1.32E-01 1.15E-01 1.20E-01 1.33E-01 1.03E-01RA 4.06E-03 3.57E-03 3.45E-03 3.37E-03 3.29E-03 3.18E-03OL 6.02E-05 5.18E-05 4.50E-05 4.77E-05 4.59E-05 4.34E-05ET 6.20Eþ05 6.06Eþ05 1.04Eþ06 1.14Eþ06 6.70Eþ05 5.90Eþ05AE 1.80Eþ04 1.51Eþ04 1.52Eþ04 1.58Eþ04 1.48Eþ04 1.11Eþ04LU 8.04Eþ03 7.32Eþ03 1.43Eþ04 1.53Eþ04 8.49Eþ03 6.62Eþ03M 7.91Eþ02 �5.14Eþ02 9.28Eþ04 1.02Eþ05 8.45Eþ04 �1.85Eþ03FF 6.90Eþ05 5.75Eþ05 4.87Eþ05 5.08Eþ05 6.24Eþ05 4.05Eþ05

Table 9Environmental impact categories compared from each LCIA.

Impact category CML Eco-indicator 99 Ecopoints 97 EDIP EPS IMPACT2002 TRACI

Acidification AC AE e AC SA TAN ACEutrophication EU AE e EU e TAN EUGlobal warning GW CC CO GW e GW GWAbiotic Depletion AD M þ FF e RES DR NRE þ ME e

Ozone Layer Depletion OLD OL OL OD e OLD ODHuman Toxicity HT CA þ RO þ RI e HTA þ HTW þ HTS e RI HHCAMEcotoxicity FWAE þ MAE þ TE ET e EWC þ EWA þ ESC e TE ET

E. Martínez et al. / Renewable Energy 74 (2015) 237e246242

- Eco-indicator 99 uses a single combined impact category thatcovers the effects of both acidification and eutrophication.

- In IMPACT2002 the characterisation factors for the TAN categoryare taken directly from Eco-indicator 99 [43].

- Eco-indicator 99 and IMPACT2002 allocate greater weight to theeffect of nitrogen oxides in acidification than the other softwaretools analysed.

A detailed examination of the various characterisations of ni-trogen oxides reveals a difference between CML and IMPACT2002(whose impact categories use the same units of measurements, i.e.kg SO2 eq.) of 997.59%: the total given by CML is 879 kg SO2 eq.,while IMPACT2002 gives 9645 kg SO2 eq.

4.1.2. EutrophicationThe results for this category can be seen in Fig. 2. Three different

groups of results are clearly obtained depending on the LCIA used:CML and EDIP give one type of result, Eco-indicator 99 andIMPACT2002 give another and TRACI gives a third. It is true, how-ever, that in determining which components of the wind turbinehave the greatest impact in this category, all the software tools

Fig. 1. Comparison of results for the category of Acidification.

except TRACI list the rotor first, followed by the tower and thefoundation.

As expected, the differences in the results provided by Eco-in-dicator 99 and IMPACT2002 are due basically to the same factorsindicated under Acidification in Section 4.1.1, i.e. their treatingacidification and eutrophication as a single category means thattheir results differ considerably from those of the other softwaretools. In this case, the main reason is the consideration of sulphurdioxide in this category.

A comparison of the results provided by the TRACI with those ofthe other software tools that use a separate eutrophication impactcategory shows that the divergence stems from the differences inthe way in which they rate phosphate. This is the substance withthe greatest weight in the TRACI results (four times greater than theeffects of nitrogen oxides), while in the case of CML, for instance,nitrogen oxides occupy the number one spot, with an effect that isalmost double that of phosphate.

4.1.3. Global WarmingAs can be seen in Fig. 3, the results for the Global Warming

category provided by CML, EDIP 96, IMPACT2002 and Ecopoints 97

Fig. 2. Comparison of results for the category of Eutrophication.

Fig. 3. Comparison of results for the category of Global Warming. Fig. 5. Comparison of results for the category of Ozone Layer Depletion.

E. Martínez et al. / Renewable Energy 74 (2015) 237e246 243

are very similar. All these software tools use the internationallyaccepted IPCC coefficients for calculating Global Warming.

Almost all the software tools cite the rotor as the componentwith the greatest impact, followed by the foundation and the tower.The only exception in this category is TRACI: this is because it doesnot take into account the effects of fossil carbon dioxide and fossilmethane (two of themost important substances in this category forsoftware tools such as CML). The result provided is therefore clearlydifferent from those of the other software tools even though TRACIis also based on IPCC coefficients.

4.1.4. Abiotic DepletionAs can be seen in Fig. 4, the results obtained in this category can

be placed in three groups: first those provided by CML, Eco-indi-cator 99 and IMPACT2002, which find the rotor to be the compo-nent with the greatest impact; second those of EDI, which find thetower to be the highest-impact component; and finally those ofEPS, which also make the tower the highest-impact component butshownegative impacts, i.e. environmental benefits, for componentssuch as the nacelle and for maintenance.

Particularly noteworthy in the case of EPS is the positive impactof the recycling of copper. On the other hand, an examination ofEDIP shows that the impact category of Resources considersessentially the impact of nickel and copper. Lower weights areallocated to components which other LCIAs see as more important,such as natural gas and crude oil.

4.1.5. Ozone Layer DepletionThe results for Ozone Layer Depletion are compared in Fig. 5. As

can be seen, most of the LCIAs studied provide similar final results:the only one that differs substantially from the rest is Ecopoints 97,

Fig. 4. Comparison of results for the category of Abiotic Depletion.

which particularly emphasises the effects of the rotor, followed atsome distance by the nacelle. The remaining LCIAs also consider therotor to be the highest-impact component, but with figures muchcloser to those of the second-place component (the tower).

A detailed examination of the causes of this discrepancy revealsthat Ecopoints 97 is the only software tool that considers theimpact of chlorinated hydrocarbons, which originate from thematerials used to manufacture the prepreg (Pre-impregnatedComposite Fibres) that goes to make up the blades of the windturbine and the nacelle cover. For example a comparison of theresults for this category provided by Ecopoints 97 and CML (whichuse the same unit of reference: g CFC-11) shows that the totalimpact according to CML is 41.1 g CFC-11, while Ecopoints 97 puts itat 423 g CFC-11. 386 g out of that total of 423 g CFC-11 correspondsto chlorinated hydrocarbons, and 37 to all the other componentsconsidered. So Ecopoints 97 is not just the only software tool thatassesses the effects of this compound in this category but it alsorates them as far greater than the effects of the other componentsfound in the case study.

4.1.6. Human ToxicityAs can be seen in Fig. 6, markedly different results are provided

in this category by each LCIA. Only Eco-indicator 99 and TRACIshow the same tendency: the others are all totally different. Forinstance CML highlights the impact of the nacelle, followed by thetower and maintenance work. EDIP also puts the nacelle in thenumber one spot, but places maintenancework in second place andthe tower third. IMPACT2002 lists the tower first, followed by thenacelle and the rotor. These substantial discrepancies are analysedin detail below.

Eco-indicator 99 assigns much higher weights to componentssuch as nitrogen, oxides and particulates <2.5 mm than to arsenic

Fig. 6. Comparison of results for the category of Human Toxicity.

Table 10Main components in the category of Ecotoxicity according to each LCIA.

LCIA 1� 2� 3� 4�

CML Hydrogenfluoride

Vanadium, ion Nickel, ion Beryllium

Ecoindicator99 Nickel Chromium Zinc LeadEDIP Copper, ion Strontium Iron, ion Chromium VIIMPACT2002 Zinc Copper Nickel ChromiumTRACI Aluminium Copper Zinc Nickel

E. Martínez et al. / Renewable Energy 74 (2015) 237e246244

and nickel, which are the main components highlighted by CML.The components highlighted by TRACI are similar to those cited byEco-indicator 99, which is why the final results provided by thesesoftware tools are similar.

For EDIP the greatest impact in this category is that of lead, acomponent that is all but irrelevant according to other softwaretools. For instance in CML it is 25th on the list, with an impact fiveorders of magnitude lower than that of the top-ranked component.

IMPACT2002 emphasises mainly aromatic hydrocarbons,arsenic and dioxins.

4.1.7. EcotoxicityFig. 7 shows the results for the category of Ecotoxicity. As in Hu-

man Toxicity, the different software tools provide widely differingresults. They do all coincide more or less in pointing to the tower asthe componentwith thehighest impact in this category, but theyvaryin regard to which components are assigned the second-highestimpact: for instance CML and TRACI point to the rotor, while Eco-indicator 99 and IMPACT2002 assign second place to the nacelle.

A more detailed examination of each LCIA studied reveals thatCML is the only software tool that considers hydrogen fluoride as themain cause of Ecotoxicity, followedby vanadium ions, nickel ions andberyllium. Eco-indicator 99, EDIP, IMPACT2002 and TRACI all attachmore importance to different metals, such as nickel, chromium, zinc,aluminium, etc., and do not allocate any weight at all to hydrogenfluoride. Byand large the LCIAs coincide in themain components thattheyfind to havemost effect in this category, but theweight assignedto each one of the software tools varies (see Table 10).

4.2. Recycling

This section analyses how the positive effects of recycling asindicated by LCAs differ depending on which LCIA is used. As maybe expected, the connections found in the specific analysis of eachcategory also appear when recycling is examined in isolation.However it is worth highlighting the results obtained in the impactcategory of Global Warming, where the various LCIAs are almostunanimous in emphasising the importance of recycling the metalmaterials from the wind turbine tower: almost all of them showreductions in environmental impact of around 25% in this category.The exception is TRACI (see Fig. 8), which shows a reduction of lessthan 2% for tower recycling. This is mainly because this softwaretool does not take into account the potential benefits of recyclingmetals such as iron, copper, etc.

4.3. Sensitivity analysis

An in-depth examination of the results of the sensitivity analysisconducted in each alternative scenario reveals that the use of

Fig. 7. Comparison of results for the category of Ecotoxicity.

different LCIAs results in the scenarios being ranked differently. Ingeneral, there is at least one LCIA in each impact category studiedthat differs from the rest in regard to which alternative is consid-ered as entailing the biggest increase over the baseline scenario. Forinstance, a detailed examination of the category of Acidification(see Fig. 9) shows that Eco-Indicator 99 and IMPACT2002 point toAS11 as the alternative scenario with the greatest impact, whileCML, EDIP, EPS and TRACI point to AS22 and AS21. This same sit-uation is repeated more or less clearly in all the impact categoriesanalysed, showing how choosing one software tool or another canlead to totally different decisions being made when products areanalysed from an environmental viewpoint.

5. Conclusions

It is often possible for users lacking in experience or with only apassing knowledge of LCA to choose an LCIA software tool (orsimply to accept the default option preselected by the softwareused) and generate an environmental impact report. This may wellbe an attractive option, since it is fast and entails relatively littleinvestment. However it is important to realise that different LCIAsand differences in the impact categories selected for examinationmay lead to different results. This is evidenced by the results ob-tained in this paper for a case study involving a multi-megawattwind turbine and seven different LCIA software tools.

Major differences in the results for some categories are founddepending on which of the 7 LCIAs is used. In Acidification andEutrophication two groups are found: one includes the resultsprovided by CML, Ecopoints, EDIP, EPS and TRACI and the otherthose of Eco-indicator 99 and IMPACT2002. In Abiotic Depletion allthe results are similar except those of the EPS, which gives negativefigures. Likewise in Ozone Layer Depletion the results provided byEcopoints 97 differ from the rest. In Human Toxicity and Ecotoxicitymarkedly different results are obtained by each of the LCIAsstudied.

Which of the impact assessment software tools available in theLCA software used is selected is therefore a crucial issue. Not all

Fig. 8. Comparison of results in regard to recycling for the category of Global Warming.

Fig. 9. Comparison of results as regards sensitivity analysis for the category ofAcidification.

E. Martínez et al. / Renewable Energy 74 (2015) 237e246 245

software tools provide similar results, and this needs to be realisedwhen the data obtained are used for decision-making, especially infields such as wind energy and indeed renewables in general,where the weight of environmental considerations compared tofinancial considerations may be especially high.

In summary, CML and Eco-indicator 99 impact assessmentsoftware tools, present results that are robust and permit astraightforward comparison for most categories. Therefore, theiruse would be reasonable, but taking into account that these soft-ware tools inconsistent results for Human Toxicity and eco-toxicitycategories.

References

[1] Lemoine DM, Fuss S, Szolgayova J, Obersteiner M, Kammen DM. The influenceof negative emission technologies and technology policies on the optimalclimate mitigation portfolio. Clim Change 2012;113(2):141e62.

[2] Kunkel CM, Kammen DM. Design and implementation of carbon cap anddividend policies. Energy Policy 2011;39(1):477e86.

[3] Sanderson BM, O'Neill BC, Kiehl JT, Meehl GA, Knutti R, Washington WM. Theresponse of the climate system to very high greenhouse gas emission sce-narios. Environ Res Lett 2011;6(3). art. no. 034005.

[4] Ferguson G. Subsurface energy footprints. Environ Res Lett 2013;8(1). art. no.014037.

[5] Cherubini F, Bright RM, Strømman AH. Global climate impacts of forest bio-energy: what, when and how to measure? Environ Res Lett 2013;8(1). art. no.014049.

[6] Barba-Guti�errez Y, Adenso-Díaz B, Lozano S. Eco-efficiency of electric andelectronic appliances: a data envelopment analysis (DEA). Environ ModelAssess 2009;14(4):439e47.

[7] Bravo Y, Carvalho M, Serra LM, Monn�e C, Alonso S, Moreno F, et al. Environ-mental evaluation of dish-Stirling technology for power generation. Sol En-ergy 2012;86(9):2811e25.

[8] Chester M, Pincetl S, Elizabeth Z, Eisenstein W, Matute J. Infrastructure andautomobile shifts: positioning transit to reduce life-cycle environmental im-pacts for urban sustainability goals. Environ Res Lett 2013;8(1). art. no.015041.

[9] Coventry ZA, Tize R, Karunanithi A. Environmental life-cycle assessment ofwaste-to-energy technologies. In: Air and Waste Management Association eInternational Conference on Thermal Treatment Technologies and HazardousWaste Combustors; 2012. p. 159e74.

[10] Desideri U, Zepparelli F, Morettini V, Garroni E. Comparative analysis ofconcentrating solar power and photovoltaic technologies: technical andenvironmental evaluations. Appl Energy 2013;102:765e84.

[11] Dufour J, Serrano DP, G�alvez JL, Moreno J, García C. Life cycle assessment ofprocesses for hydrogen production. Environmental feasibility and reduction ofgreenhouse gases emissions. Int J Hydrogen Energy 2009;34(3):1370e6.

[12] Esteban B, Baquero G, Puig R, Riba JR, Rius A. Is it environmentally advanta-geous to use vegetable oil directly as biofuel instead of converting it to bio-diesel? Biomass Bioenergy 2011;35(3):1317e28.

[13] Gebreslassie BH, Slivinsky M, Wang B, You F. Life cycle optimization for sus-tainable design and operations of hydrocarbon biorefinery via fast pyrolysis,hydrotreating and hydrocracking. Original Research Article Comput Chem Eng2013;50:71e91.

[14] Gonz�alez-García S, Mola-Yudego B, Dimitriou I, Aronsson P, Murphy R.Environmental assessment of energy production based on long term

commercial willow plantations in Sweden. Sci Total Environ 2012;421e422:210e9.

[15] Hajjaji N, Pons MN, Renaudin V, Houas A. Comparative life cycle assessment ofeight alternatives for hydrogen production from renewable and fossil feed-stock. J Clean Prod 2013;44:177e89.

[16] Martínez-Gonz�alez A, Casas-Leuro OM, Acero-Reyes JR, Castillo-Monroy EF.Comparison of potential environmental impacts on the production and use ofhigh and low sulfur regular diesel by life cycle assessment. CTyF Cienc TecnolFuturo 2011;4(4):123e36.

[17] Martínez E, Sanz F, Pellegrini S, Jim�enez E, Blanco J. Life-cycle assessment of a2-MW rated power wind turbine: CML method. Int J Life Cycle Assess2009;14(1):52e63.

[18] Martínez E, Sanz F, Pellegrini S, Jim�enez E, Blanco J. Life cycle assessment of amulti-megawatt wind turbine. Renew Energy 2009;34(3):667e73.

[19] Miller VB, Landis AE, Schaefer LA. A benchmark for life cycle air emissions andlife cycle impact assessment of hydrokinetic energy extraction using life cycleassessment. Renew Energy 2011;36(3):1040e6.

[20] Modahl IS, Askham C, Lyng KA, Brekke A. Weighting of environmental trade-offs in CCS e an LCA case study of electricity from a fossil gas power plantwith post-combustion CO2 capture, transport and storage. Int J Life CycleAssess 2012;17(7):932e43.

[21] Nishimura A, Hayashi Y, Tanaka K, Hirota M, Kato S, Ito M, et al. Life cycleassessment and evaluation of energy payback time on high-concentrationphotovoltaic power generation system. Appl Energy 2010;87(9):2797e807.

[22] Pawelzik PF, Zhang Q. Evaluation of environmental impacts of cellulosicethanol using life cycle assessment with technological advances over time.Biomass Bioenergy 2012;40:162e73.

[23] P�erez-Fortes M, Bojarski AD, Puigjaner L. Co-production of electricity andhydrogen from coal and biomass gasification. Chem Eng Trans 2011;25:507e12.

[24] Talens Peir�o L, Lombardi L, Villalba M�endez G, Gabarrell i Durany X. Life cycleassessment (LCA) and exergetic life cycle assessment (ELCA) of the productionof biodiesel from used cooking oil (UCO). Energy 2010;35(2):889e93.

[25] Torres CM, Ríos SD, Torras C, Salvad�o J, Mateo-Sanz JM, Jim�enez L. Sustain-ability analysis of biodiesel production from Cynara Cardunculus crop. Fuel2013;111:535e42.

[26] Wang B, Gebreslassie BH, You F. Sustainable design and synthesis of hydro-carbon biorefinery via gasification pathway: integrated life cycle assessmentand technoeconomic analysis with multiobjective superstructure optimiza-tion. Comput Chem Eng 2013;52:55e76.

[27] Brent AC, Hietkamp S. Comparative evaluation of life cycle impact assessmentmethods with a South African case study. Int J Life Cycle Assess 2003;8(1):27e38.

[28] Bovea MD, Gallardo A. The influence of impact assessment methods on ma-terials selection for eco-design. Mater Des 2006;27(3):209e15.

[29] Cavalett O, Chagas MF, Seabra JEA, Bonomi A. Comparative LCA of ethanolversus gasoline in Brazil using different LCIA methods. Int J Life Cycle Assess2012;18:647e58.

[30] Carvalho IS, Ant~ao P, Soares CG. Modelling of environmental impacts of shipdismantling. Ships Offshore Struct 2011;6(1e2):161e73.

[31] Dreyer LC, Niemann AL, Hauschild MZ. Comparison of three different LCIAmethods: EDIP97, CML2001 and eco-indicator 99: does it matter which oneyou choose. Int J Life Cycle Assess 2003;8(4):191e200.

[32] Fantozzi F, Buratti C. Life cycle assessment of biomass chains: wood pelletfrom short rotation coppice using data measured on a real plant. BiomassBioenergy 2010;34(12):1796e804.

[33] Knoeri C, Sany�e-Mengual E, Althaus HJ. Comparative LCA of recycled andconventional concrete for structural applications. Int J Life Cycle Assess2013;18:909e18.

[34] Monteiro H, Freire F. Life-cycle assessment of a house with alternative exte-rior walls: comparison of three impact assessment methods. Energy Build2012;47:572e83.

[35] Pant R, Van Hoof G, Schowanek D, Feijtel TCJ, De Koning A, Hauschild M, et al.Comparison between three different LCIA methods for aquatic ecotoxicity anda product environmental risk assessment: insights from a detergent casestudy within OMNIITOX. Int J Life Cycle Assess 2004;9(5):295e306.

[36] Renou S, Thomas JS, Aoustin E, Pons MN. Influence of impact assessmentmethods in wastewater treatment LCA. J Clean Prod 2008;16(10):1098e105.

[37] Sim~oes CL, Xar�a SM, Bernardo CA. Influence of the impact assessment methodon the conclusions of a LCA study. Application to the case of a part made withvirgin and recycled HDP. Waste Manage Res 2011;29(10):1018e26.

[38] Valderrama C, Granados R, Cortina JL, Gasol CM, Guillem M, Josa A. Compar-ative LCA of sewage sludge valorisation as both fuel and raw material sub-stitute in clinker production. J Clean Prod 2013;51:205e13.

[39] Van Caneghem J, Block C, Vandecasteele C. Assessment of the impact on hu-man health of industrial emissions to air: does the result depend on theapplied method? J Hazard Mater 2010;184(1e3):788e97.

[40] Van Caneghem J, Vermeulen I, Block C, Cramm P, Mortier R, Vandecasteele C.Abiotic depletion due to resource consumption in a steelwork assessed by fivedifferent methods. Resour Conserv Recycl 2010;54(12):1067e73.

[41] Davidsson S, H€o€ok M, Wall G. A review of life cycle assessments on windenergy systems. Int J Life Cycle Assess 2012;17(6):729e42.

[42] Guin�ee JB, editor. Life cycle assessment: an operational guide to the ISOStandards; LCA in perspective; guide; operational annex to guide. TheNetherlands: Centre for Environmental Science, Leiden University; 2001.

E. Martínez et al. / Renewable Energy 74 (2015) 237e246246

[43] Goedkoop M, Effting S, Collignon M. The eco-indicator 99 e a damage ori-ented method for life cycle impact assessment. Manual for designers. 17-4-2000. 2nd ed. Amersfoort, The Netherlands: PR�e Consultants B.V.; 2000.

[44] Brand G, Braunschweig A, Sche idegger A, Schwank O. Ecopoints 97 (CH)V2.03/Ecopoints (Bewertung in Oekobilanzen mit der Methode der€okologischen Kn appheit Oekofaktoren 1997). BUWAL Schriftenreihe UmweltNr. 297. Bern: BUWAL; 1998.

[45] Wenzel H, Hauschild M, Airing L. EDIP/UMIP 97 V2.03/EDIP World/Dk: envi-ronmental assessment of products. Methodology, tools and case studies inproduct development. 1st ed., vol. 1. United Kingdom: Chapman & Hall; 1997.Kluwer Academic Publishers; Hingham, MA, USA.

[46] Steen B. EPS a systematic approach to environmental priority strategies inproduct development (EPS). Version 2000 e general system characteristics.

CPM Report 1999:4. Gothenburg, Sweden: Chalmers University of Technol-ogy; 1999.

[47] Jolliet O, Margni M, Charles R, Humbert S, Payet J, Rebitzer G, et al. IMPACT2002þ: a new life cycle impact assessment methodology. Int J Life CycleAssess 2003;8(6):324e30.

[48] Bare JC, Norris GA, Pennington DW, McKane T. TRACI: the tool for thereduction and assessment of chemical and other environmental impacts. J IndEcol 2003;6(3e4):49e78.

[49] Frischknecht R, Rebitzer G. The Ecoinvent database system: a comprehensiveweb-based LCA database. J Clean Prod 2005;13(13e14):1337e43.

[50] Frischknecht R, Jungbluth N, Althaus H-J, Doka G, Dones R, Heck T, et al. TheEcoinvent database: overview and methodological framework. Int J Life CycleAssess 2005;10(1):3e9.