Energy and Buildings

(1)

ContentslistsavailableatSciVerseScienceDirect

Energy

and

Buildings

jo u rn al h om epa g e :w w w . e l s e v i e r . c o m / l o c a t e / e n b u i l d

Life-cycle

assessment

of

a

house

with

alternative

exterior

walls:

Comparison

of

three

impact

assessment

methods

Helena

Monteiro,

Fausto

Freire

∗

ADAI-LAETA,DepartmentofMechanicalEngineering,UniversityofCoimbra,PóloIICampus,RuaLuísReisSantos,3030-788Coimbra,Portugal

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received7February2011

Receivedinrevisedform8December2011 Accepted22December2011 Keywords: Buildingenvelope Primaryenergy Environmentalimpacts Life-cycleassessment Single-familyhouse Cumulativeenergydemand Eco-indicator’99

CML2001

a

b

s

t

r

a

c

t

Alife-cycle(LC)modelhasbeenimplementedforaPortuguesesingle-familyhouse.Thefirstgoalisto characterizethemainLCprocesses(materialproductionandtransport,heating,cooling,maintenance) assessingsevenalternativeexteriorwallsforthesamehousetoidentifyenvironmentallypreferable solu-tions.Thesecondgoalistocomparetheresultsofthreelife-cycleimpactassessment(LCIA)methods– CED(cumulativeenergydemand),forprimaryenergyaccounting;CML2001(InstituteofEnvironmental SciencesofLeidenUniversity)andEI’99(Eco-indicator’99),formultipleenvironmentalimpacts– to deter-minetheextenttowhichtheresultsofalife-cycleassessmentareinfluencedbythemethodapplied.The resultsshowthatthemostsignificantLCprocessdependsontheoperationalpatternassumed. Regard-ingtheassessmentoftheexteriorwallalternatives,theresultsindicatethewood-wallisthepreferable solution.Non-renewableCEDshowsresultssimilartoabioticdepletion(CML2001)andresources(EI’99) categories,aswellassomecorrelationwithclimatechange/globalwarmingpotential(GWP),acidification andeutrophication.However,nocorrelationwasfoundwiththeremainingimpactcategories.Comparing CML2001andEI’99categories,GWP,ozonelayerdepletion,abioticdepletion,acidification,and eutroph-icationpresentrobustresultsthatpermitastraightforwardcomparisonbetweenthetwoLCIAmethods. Nevertheless,CMLresultspresentslightlyhigherimpactsfortheusephase,whileEI’99formaterial pro-duction.Inaddition,thetwoLCIAmethodscanpresentinconsistentresultsbetweensimilarcategories (differentrankingofalternatives),whichultimatelycaninfluencethechoiceamongsolutions.

1. Introduction

In the European Union (EU), the housing sector accounts for a substantial amount of energy consumption and envi-ronmental impacts [1]. To address this sector, EU regulations havebeendeveloped[1–3]focusingonreducingtheoperational energyconsumption ofdwellings,but disregardingthefull life-cycle environmental impacts. A comprehensive assessment of theenergyuseandenvironmentalimpactsofbuildingsrequires a life-cycle perspective to quantify the various impacts and to identify improvement opportunities towards more sustainable solutions. Life-cycle assessment (LCA) methodology, which has gainedincreasedinternationalacceptance, addressesthe poten-tialenvironmentalimpactsthroughoutaproductlife-cycle(LC). LCAcanbeappliedtodecision-making inordertoimprovethe environmentalperformanceofbuildings.

LCAstudiesperformedfor dwellingshavemainlyfocusedon energy requirements and greenhouse gas (GHG) emissions. A smallnumberofrecentLCAhavestudiedmultipleenvironmental

∗Correspondingauthor.Tel.:+35123979039;fax:+35135123979001. E-mailaddress:fausto.freire@dem.uc.pt(F.Freire).

impacts,oftenusingdifferinglife-cycleimpactassessment(LCIA) methods with non-comparable or contradictory conclusions

[4–10].ForbuildingLCApractitionersitisnotobviouswhichLCIA methodtochooseandneitherwhethertheinterpretationofresults fromdifferentmethodsleadstocoincidentorcontradictory conclu-sions[11].Sinceimpactcategories,characterizationindicatorsand characterizationfactorsvarybetweenLCIAmethods,ameaningful comparisonbetweenmethodsisdifficulttoperform.Furthermore, noneofthepublishedstudieshasusedmultipleLCIAmethodsto assessdwellingstodeterminewhethertheLCAfindingsare influ-encedbytheLCIAmethodapplied.

ThisarticlepresentsanLCAforaPortuguesesingle-familyhouse comparingsevenalternativeexteriorwallsolutionsaimedat iden-tifyingenvironmentallypreferablesolutions.Inthisresearch,the mainlife-cycleprocessesaffectedbythebuildingenvelope (mate-rialproductionandtransport,heating,cooling,maintenance)have beencharacterizedintermsofenergyandenvironmentalimpacts, consideringtwoalternativeoperationalpatterns.Notincludedis theenergyrequirementofotheroperation-phaseactivities (elec-tricappliances,lighting,cooking,domestichotwater)sinceitis notaffectedbytheexteriorwallsolution.Inthisresearch,three LCIAmethodshavebeenused:CED(cumulativeenergydemand), for life-cycleprimary energyaccounting; CML 2001(Instituteof 0378-7788/$–seefrontmatter©2011ElsevierB.V.Allrightsreserved.

(2)

EnvironmentalSciencesofLeidenUniversity,2001)andEI’99 (Eco-indicator’99),formultipleenvironmentalimpacts.Thesemethods havebeencomparedtodeterminetheextenttowhichtheresults ofanLCAareinﬂuencedbythemethodapplied.

Thispaperhassixsections,includingitsintroduction.Section

2presentsa review oflife-cyclestudies ofdwellings.Section3

describestheLCAmethodologyandtheimpactassessment meth-ods used.The life-cycle inventoryand model implemented are described in Section4. Resultsfor the exterior-wallhouse sce-narioanalysisarediscussedinSection5.Theﬁnalsectionisthe conclusion.

2. Literaturereview

Theﬁrstlife-cycle(LC)studiesforresidentialbuildingswere publishedabout15yearsago,e.g.[12–15].Forexample,Adalberth

[13]proposedamethodologytoaccountforLCenergyofbuildings fromcradletograveandappliedittothreeSwedishsingle-family housesprefabricatedinafactory[12].Amongotherresults, Adal-berth [12] concluded that thebuildinguse phase requires 85% of thetotal LCenergy, while construction materialsamountto 15%.Keoleian[14,15]evaluatedLCenergy,greenhousegas(GHG) emissions,andcostsofasingle-familyhouseinMichigantoﬁnd opportunitiesforconservingenergythroughoutpre-use,use,and demolitionphases.Sincethen,anumberofLCstudiesofdwellings

[4–9,12,16–22]aswellasafewreviewpapers[23–26]havebeen published.

The majority of studies (e.g. [5,6,12,13,17–21]) have been implementedfor cold climatehousesandhave shown thatthe operationalphase hasa preponderantweightin overall energy requirementandGHGemissions.Nonetheless,afewrecentstudies have pointed out that, for very-low-energy buildings,the non-operationalphase is the mostsigniﬁcant[4,5,27].Only a small numberofLCAstudiesofdwellingshaveaddressedawiderrange ofenvironmentalimpacts.Someofthosestudies[4–10]compared alternativebuildingscenariosandarereviewedbelow.Peuportier etal.[6]performedanLCAforthree differenthousesinFrance ((i)aconventionalhousewithconcreteblocks;(ii)ahousewitha solarheatingsystem;and(iii)ahigh-insulatedwoodhouse).The CML92LCIAmethodwasused,andresultsfortwelveimpact cat-egories werepresented comparingthethreehouses. Thewood high-insulatedhousehad generallythelowestimpacts,holding abouthalfoftheconcretehouseimpactsforninecategories.

CitherletandDefaux[5]comparedthreehouses:aSwiss stan-dardhouse;alow-energyhousethatcomplieswiththeMinergie standard;apassivehouse(withincreasedinsulation,aventilation heatrecovery system,solar collectorsand photovoltaicpanels). LCIAresultswere calculatedbased onthree CML 92 categories (globalwarmingpotential(GWP),acidification,andphotochemical oxidation(PO))aswellasnon-renewable energy(NRE) require-ments.ConsideringtheSwisselectricalmix,theresultsshowed thatNREwasreducedby33%intheMinergiehouseand66%in thepassivehouseincomparisontothestandardhouse. Concern-ingCML92impacts,thetwolow-energyhousespresentedasimilar performancewithsignificantreductions(e.g.62%forGWP,29%for PO,and10%foracidification)relativetotheSwissstandardhouse. BlenginiandDiCarlo[4]assessedalow-energyhousefor north-ernItalyandcompareditwithastandardhouseassuminga70-year lifetime.LCIAresultswerepresentedattwolevels:(i)mid-point indicators,which arerepresentativeofenvironmentalproblems andbasedoninternationalconventions(primaryenergy(PE),NRE, GWP,acidification,eutrophication,ozone layerdepletion(OLD), and PO);(ii)single scoreend-point indicators,which are more uncertain,butmightsimplifythedecisionprocess(EI’99,Ecological footprint,EPS2000).Thelow-energyhousepresentedareduction

ofimpacts relativetothestandard house;however,the magni-tudeofthereductionvariedwiththeend-pointindicatorused(e.g. reductionof60%forNRE,52%forEI’99,and38%forEPS2000). Fur-thermore,theresultsshowedthatmaterial-relatedimpactswere themostsigniﬁcantforthelow-energyhouse,whileforthe stan-dardhouse,heatingwasthemostimportantprocess.

Morerecently,acomprehensiveEuropeanstudy[8,9]assessed varioustypesofEUdwellings,addressingdifferentimprovement options. They have focused on single-family houses and two kindsofapartmentbuildings,consideringbothexistingandnew dwellings.TheLCIAresultswerepresentedforsixmid-pointCML impactcategories(PE,GWP,acidification,eutrophication,OLD,and PO);toxicitycategorieswerenotconsideredbecausetheystilllack scientificrobustness[8,9].Theresultsobtainedfornewdwellings showedthatthebuildingenvelopehadasignificantpartoftheLC environmentalimpacts.Theexteriorwallsandtheroofwerefound tobethemostimportantbuildingcomponents.

Somepublicationshavefocusedonexteriorwalls.Forinstance, Ortizetal.[7]assessedtheconstructionphaseofanapartment blockconsideringtypicalSpanishexteriorandinteriorwall sce-narios. TheCMLmethodwasusedtoassessacidiﬁcation, GWP, ionizingradiation,andOLDimpactsaswellasenergyandresource consumption. Thesteelelements, inparticular galvanized steel, wereresponsibleforhighenvironmentalimpacts.Ortizetal.[7]

concludedthattheLCAresultscansupporttheselectionofthe con-structivecombinationswiththelowestimpacts.Frenetteetal.[10]

comparedﬁvefactory-builtwood-frameexteriorwallsforahouse inQuebec(Canada)concerningenergyandthreeLCIAmethods: TRACI;Impact2002;andEI’99.Midpointresultswerecalculated forthementionedmethodsandendpointresultswerepresented forImpact2002andEI’99.ThegoalofFrenetteetal.[10]wasto supporttheselectionofanappropriateenvironmentalindextobe includedinamulti-criteriadecisionanalysisevaluation.Theresults showedthatwoodframinghadasmallcontributiontototal embod-iedimpacts.Slightlydifferentrankingsoftheexteriorwallswere observedfor somemidpointcategoryresultsobtainedwiththe threemethods, namely,respiratory organic,eutrophication,and mineralresults.Nevertheless,Frenetteetal.[10]concludedthat endpointresultsgenerallypresentedasimilarrankingof alterna-tives,andthatitseemedacceptabletouseclimatechangecategory asasingleenvironmentalindex.

3. LCAmethodologyandLCIAmethods

Life-cycleassessment(LCA)hasfourinterrelatedphases:goal andscopedefinition;life-cycleinventory(LCI);life-cycleimpact assessment(LCIA)andinterpretation[28,29].IntheLCIA,inventory dataisaggregatedintospecificenvironmentalimpactcategories accordingtoa method.DifferentLCIAmethodswillleadto dis-tinctresults(values,impactcategories,units).LCIAmethodscan besingle-category(e.g.primaryenergy,exergy,globalwarming potential) or multi-category, with specific sets of impact cate-gories.Multi-categoryLCIAmethodscanbeproblem-orientedor damage-oriented.Problem-orientedmethods(e.g.CML2001)have midpointimpactcategoriesandmodelproblemsatanearlystage inthecause-effectchain,allowingamoretransparentassessment andlimitingtheuncertainties.Damage-orientedmethods(e.g. Eco-indicator’99) model thecause–effect chainup totheendpoints (damagetohumansandecosystems),haveanarrowedsetof cate-gories,andhavehigheruncertaintythanmidpointmethods.

Inthisresearch,thevariabilitybetweenLCIAresultshasbeen assessedbyapplyingtothesamelife-cycleinventorythree meth-ods: a general single-issue method – the cumulative energy demand(CED)– toaccountforlife-cycleprimaryenergy require-ments,andtwowell-knownenvironmentalLCIAmethods–CML

(3)

Table1

CML2001andEco-indicador’99impactcategories.

CML2001category Unit Eco-indicator99category Unit

Abioticdepletion(AD) kgSbequiv. Fossilfuels MJsurplus

Minerals MJsurplus

Acidiﬁcation(Acid) kgSO2equiv. Acidiﬁc./eutrophi. PDF*m2_y

Eutrophication(Eut) kgPO4equiv.

Globalwarmingpotential(GWP) kgCO2equiv. Climatechange DALY

Ozonelayerdepletion(OLD) kgCFC-11equiv. Ozonelayer DALY

Photochemicaloxidation(PO) kgC2H4equiv. Resp.organics DALY

Freshwaterecotoxicity(Fecot) kg1,4-DBequiv. Ecotoxicity PDF*m2_y Marineecotoxicity(Mecot)

Terrestrialecotoxicity(Tecot)

Humantoxicity(Htox) kg1,4-DBequiv. Carcinogens DALY

Resp.inorganics DALY

Radiation DALY

Landuse PDF*m2_y

2001 and Eco-indicator’99 – to assess multiple environmental impacts.CEDquantiﬁesthetotal energyresourcedepletionand implicitlytakesintoaccountenergyquality,sinceprimaryenergy isthesumofﬁnalenergywithalltransformationlosses[30].CED (MJequiv.)iscalculatedbasedonthehigherheatingvalueand dis-tinguishesrenewable(Renew)andnon-renewable(Non-R)energy sources[31].Non-renewableCEDisveryimportanttoassessthe depletionoffossilenergyresourcesortherenewabilityofasystem

[30]anditisawidelyusedindicatortoassesstheenergyLC per-formanceofbuildings.TheCML2001andEco-indicator’99(EI’99) LCIAmethodsarebroadlyusedinLCAstudies,buthavenotbeen muchusedforbuildings.Table1showstheimpactcategoriesof CML2001andEI’99andtheirrespectiveimpactindicators[31]. Relatedcategoriesareplacedinthesameline.

DevelopedbytheInstituteofEnvironmentalSciences(CML)of LeidenUniversity,CML2001isaproblem-orientedmethodwith tenimpactcategoriesin thebaseline version(Table1).Toxicity categories normally are not addressed because theyhave high uncertainty and lack scientific robustness [8,32]. To perform a transparentcomparisonbetweenthetwoLCIAmethods,thefour toxicitycategorieshavebeenincludedinourcalculations; how-ever,thelackofscientificrobustnesshasbeentakenintoaccount intheinterpretationofresults.TheEco-indicator’99 characteriza-tionisfirstlymadeforelevenimpactcategories(Table1),which, accordingtotheirimpactindicator,canbegroupedinthreedamage categories:humanhealth;ecosystemquality;andresources[31]. Thedamagetohumanhealthisexpressedindisabilityadjusted lifeyears(DALY).Theecosystemqualitydamagesareexpressedin potentiallydisappearedfractionofspeciesinacertainareaovera periodoftime(PDF*m2_y)._The_damage_to_resources_is_expressed_in surplusenergytoextractmineralsorfossilfuels(MJsurplus)[31]. Moredetailedinformationontheassumptionsunderpinningthe LCIAmethods(characterizationfactorstoaggregatetheLCIinthe impactindicatorsoftheimpactcategories)canbefoundin[33]for CML2001,andin[34,35]forEI’99.

ThemagnitudeofLCIAresultscanbefurthercalculated rela-tivelytosomereferenceinformation(acommonscaletoallimpact categories, normally representing the background impact from society’stotalactivities).Thisiscallednormalizationanditsaimis betterunderstandtherelativesigniﬁcanceofeachindicatorresult inordertofacilitatetheinterpretationofresults.Inaddition, nor-malizationmaybealsohelpfulincheckingforinconsistencies.The CML methodhasdifferent sets of normalization: World (1990), WesternEurope(1995)andtheNetherlands(1997).Inthepresent research,theWesternEuropecontexthasbeenadopted.The nor-malizationcalculationconsistsofdividingtheLCIAresultsofeach impactcategoryper thereferencevalue (thetotal impactfrom emissions,extractions,radiation,andlanduse,perimpactcategory forWesternEuropeoverayear)[31,33].

ConcerningEI’99,normalizationisperformedondamage cate-gorylevel anditisdependentontheperspectivechosen.There arethreeperspectiveswithintheEI’99method:hierarchist; indi-vidualist and egalitarian.According to[34,35],theindividualist perspectivevaluesshorttermemissionswhose impactsare sci-entiﬁcallyproven,assumingthatlongtermimpactsareunrealistic becausetheywillbeavoidbyfuturetechnology. Theegalitarian perspectiveconsidersthatallfuturegenerationsareequally impor-tant tothe present population; therefore,it values allpossible effectsconsideringaverylong-termtimeframe.Thehierarchist isanintermediateperspectivebalancingshorttermandlong-term impactsthatarebasedonglobalconsensus.Thestandard perspec-tive,hierarchist,hasbeenassumedin this study.Thereference valueusedtocalculateEI’99normalizedresultsisfortheEuropean context(environmentalimpactperyearandpercapita, consider-ingapopulationof386millionforEurope,1993year,withsome updates[31]).

RecentdevelopmentsintheLCAmethodologyemphasizetwo types of approaches: attributional versus consequential. Attri-butionalLCA(ALCA) focuses ondescribing theenvironmentally relevant flows to and from a system,while the Consequential LCA(CLCA)aimstodescribehowenvironmentallyrelevantflows willchangeinresponsetopossibledecisions[36,37].ALCA gen-erally uses average data for the processes within the system studied and assumes that changes in those processes will not affectthemarket(backgroundand foregrounddata). CLCAuses economicandspecificmarketdata(marginaldata)tostudy pro-cesses within and outside the system that are affected by a changewithinthesystem.CLCAhasbeenusedtocalculate envi-ronmental impacts from indirect changes and rebound effects (e.g.landusechange)causedbyanincreased/decreaseddemand onproducts or services(minerals,biofuels, agriculture, energy, etc.).

InLCAstudiesofbuildings,theattributionalapproach(ALCA)is themostcommonlyused;onlyafewCLCAstudieshavebeen pub-lished,namelyVieiraandHorvath[38]andLesage[39,40].Lesage

[39,40]usedbothALCAandCLCAtoassessbrownfield rehabilita-tionforresidentialdevelopmentsandcompareittoanexposure minimizationscenario,which consistsincovering thesite with clean soil.AlthoughALCA resultsshowedno preferableoption, CLCAclearlysupportedrehabilitationofbrownfiedsforresidential reuseinordertoavoidthedevelopmentofsuburbansites.Withthe goalofdecreasinguncertaintyassociatedwithend-of-life techno-logicalforecasting,VieiraandHorvath[38]usedbothALCAand CLCAtoassesstheenvironmentalimpactsofconcreteusedinan officebuilding.They[38]concludedthatthereisnorelevant dif-ferencebetweenresultsfromALCAandCLCA,andthatthechoice betweentheuseofoneortheotherforbuildingsmaynotbea criticaldecision.

(4)

Fig.1. Architecturalmodelofthesingle-familyhouse.Axonometricdrawings:(a)westview,(b)eastview–groundfloorsectionand(c)eastview–firstfloorsection.

TheALCAapproachhasbeenadoptedinthisresearchfortwo reasons.Firstly,themainaimistoassesssevenalternative exte-riorwallsforthesamehouseinordertoidentifyenvironmentally preferablesolutions.Secondly,itwasassumedthattheoccurring changeswouldnotaffectthemarket.TheALCAmodeland inven-toryimplementedarepresentedinthenextsection.

4. Life-cyclemodelandinventoryforasingle-familyhouse

Alife-cyclemodel hasbeenimplementedfora single-family house in Portugal, addressing the most signiﬁcant buildingLC phasesand processesaffected bythesevenalternative exterior wallsolutionsassessed:constructionphase(materialproduction andtransportprocesses)andusephase(maintenance,heating,and coolingprocesses).AnALCAhasbeenimplemented,followingthe mainaimoftheresearchandbuildingonresearchpresentedin

[41–43].Thecase studyimplementedusingtheSimapro7 soft-ware(www.pre.nl)isbasedonabasicarchitecturemodel(Fig.1) representingatypicalsingle-familyhouselocatedinCoimbra,in thecenterofPortugal,withanexpectedlifespanof50yearsand 132m2_of_living_area._The_functional_unit_selected_is_the_building livingareaoverthebuildinglifespaninordertoanalyzethewhole buildingperformanceandthevariousexteriorwallhousesolutions. Ithasbeenassumedthatthehouseisoccupiedbya4-person fam-ily.Thetwo-ﬂoorbuildinghasacuboidalshapewitha7m×10m planbasisand2.7mbetweenslabs,resultinginatotalvolumeof 356m3_._In_order_to_focus_on_the_living_area_and_building_exterior walls,nobasement,garageandlofthavebeenconsideredinthe model.

4.1. Constructionphase

Thehouseconstructionphaseincludesmaterialproductionand transportation.Totracetheburdensassociated withthese pro-cesses,thebuildingmaterialshave beenaccountedin terms of massorvolumeforthevariousbuildingcomponents(structure, roof,groundfloor,firstfloor,walls,windowsanddoors)whichare describedinTable2.Constructionmaterialsandtechniquesused inPortugalhavebeenassumedtohaveastrongthermalinertia levelandhavebeendefinedtofulfillthePortuguesethermal build-ingregulationrequirements[2].Anadditional5%ofmaterialshave beenconsideredtoincludelossesonsiteduetocuttingandfitting processes.Theseelementshave beenmodeledbasedon Kellen-bergeretal.[44],whichpresentsaverageEuropeanLCIdataforthe productionofbuildingmaterials.

Concerningtheexteriorwalls,sevenalternativescenarioshave beenimplemented in themodel. Thedouble hollowbrick wall

solutionhasbeenadoptedasabase-case(scenarioH0),because itisasolutionwidelyusedinPortuguesebuildingpractice.Table3

details the sevenexterior wallscenarios. The walls considered havedifferentmaterialsintheircomposition(hollowbrick, fac-ing brick,concreteblocks and wood).Different insulation layer thicknesseshavebeenselectedinordertoobtainsimilarglobal thermal coefﬁcients for the various walls (U-values between 0.47and0.51W/m2◦_C)._{Consequently,}_identical_heating_and

cool-ing requirementshave been assumed for thevarious scenarios assessed,andthusheatingandcoolingenvironmentalimpactsdo notinﬂuencethecomparisonofalternatives.

Thetransportationoftheconstructionmaterialstobuildingsite hasbeenimplementedassuminglorrytransportation,with Euro-peanﬂeetaveragecharacteristics.Theinventorydataassociated withthisprocesswereobtainedfromSpielmannetal.[31]and[45]. Themainconstructionmaterialweightsandshippingdistancesfor thealternativeexteriorwallsandcommonbuildingcomponents arepresentedinTable4.

4.2. Usephase

The use phase includes heating, cooling and maintenance operationalrequirements.FollowingPortuguesebuildingthermal regulation(RCCTE) [2],heating/cooling set-pointsof 20◦C/25◦C have beenconsideredas wellas 4W/m2 _of _internal _heat_gains (from lights,electricalappliancesandoccupants),anda natural ventilationrateof0.6airchangesperhour.A10kWheatpump withacoefﬁcientofperformance(COP)of2.8forheatingand2.0 forcoolinghasbeenadoptedforthehouseheatingand cooling system.

Theannualheatingandcoolingrequirementshavebeen cal-culatedusingtheRCCTEcalculationmethod[2],whichisbased onaseasonal quasi-steadystatemethodproposedin the Euro-peanStandardENISO13790[46].TheRCCTEmethodassumesa permanentoccupancyofthebuildingwithinteriorseasonal set-points(20◦Cforheatingseason,25◦Cforcoolingseason).Forthe single-familyhouseofthisstudy,theannualheatingandcooling requirementshavebeenquantiﬁedas71.8kWh/(m2_year)_for heat-ingand3.8kWh/(m2_year)_for_cooling_(with_a_variation_of_3%_among theexteriorwallhousescenarios).However,itshouldbenoted thattheseareratherunrealisticHVAClevelsforPortuguese res-identialbuildingsbecausePortuguesehouseholdersarenotused toheating/coolingallroomssimultaneously,norcontinuously.In fact,thiscanbeconﬁrmedbyRCCTEprimaryenergycalculation, which,followingtheENISO13790(point13.2.2),usesa dimen-sionless reduction factorof 0.1 bothfor heating and coolingto takeintoaccounttheintermittentHVAC patternsofPortuguese

(5)

Table2

Descriptionofbase-casebuildingcomponents.

Buildingcomponents Area(m2₎ _Material_layers _Thickness_(m) _Volume_(m3_/m2₎* _Weight_(kg/m2₎* _Total_weight_(kg)

Exteriorwalls(H0) 222 Masonry:hollowbrick(30×20×11) 0.110 79.90 17737.8

-Cementmortar 22.00 4884.0

-Water 3.30 732.6

222 XPS–extrudedpolystyrene 0.400 1.20 266

201 Masonry:hollowbrick(30×20×15) 0.150 107.10 21527.1

-Cementmortar 30.00 6030.0

-Water 4.50 904.5

184 Baseplaster 0.040 66.00 12012.0

Interiorwalls 110 Masonry:hollowbrick(30×20×11) 0.110 79.90 8141.1

-Cementmortar 22.00 2406.8 -Water 3.30 361.0 Baseplaster 0.040 66.00 7220.4 Roof 74.4 Gravel 0.050 90.00 6696.0 FeltPP 0.002 0.13 9.7 XPS– extrudedpolystyrene 0.050 1.50 111.6 Bitumen 0.010 10.50 781.2 Anhydritescreed 0.050 50.00 3720.0 ConcreteC25/30 0.150 360.00 26784.0 Reinforcedsteel** _12.00 _892.8 Limemortar 0.02 20.00 1488.0

Firstﬂoorslab 76.4 Woodﬂoor(planks) 0.020 0.020 – 901.52

Woodensquarejoist 0.040 0.003 – 134.5

Anhydritescreed 0.030 30.00 2292.0

ConcreteC25/30 0.150 360.00 27504.0

Reinforcedsteel** _12.00 _916.8

Limemortar 0.020 20.00 1528.0

Groundﬂoor 80 Woodplanks 0.020 0.020 – 994.0

Woodensquarejoist 0.040 0.003 – 140.8

Lightweightanhydritescreed 0.050 50.00 4000.0

ConcreteC25/30 0.120 288.00 23040.0

Reinforcedsteel** _9.60 _768.0

Gravel 0.200 360.00 28800.0

Buildingcomponents Units Materiallayers Area/un(m2₎* _Weight/un_(kg/un) _Total_weight_(kg)

Structure -Beam(0.3×10×0.2) 6 ConcreteC25/30 1440.0 8640 Reinforcedsteel** _48.0 ₂₈₈ -Column(6×0.2×0.3) 9 ConcreteC25/30 864.0 7776 Reinforcedsteel** _28.8 ₂₅₉ -Foundation(2×0.3×0.4) 9 ConcreteC25/30 576.0 5184 Reinforcedsteel** _19.2 ₁₇₃

Windows(classIII,withself-controlled ventilationdevices)

11 Aluminiumframe 3.24

Doubleglazing(U=1,1) 7.95

Doors 1 Exteriorwoodendoor 2.00

8 Interiorwoodendoor 1.60

* _“m2_”_represents_“m2_of_building_{component”.} ** ₈₀_kg/m3_of_concrete.

householders.Furthermore, theaverageannual end-use energy consumptionofPortuguesehouseholds(2008)isabout2.7koe/m2 (31kWh/m2₎_[47]_,_the_lowest_level_in_European._This_value_is_less thanhalftheonecalculatedbyRCCTE.Therefore,inthisresearch twoalternativeoperationalpatterns(OP)havebeenconsidered, basedondifferentoccupancyandcomfortlevels:

-OP1,whichrepresentsanintensiveoccupancyofthehousewith amedium HVAClevel, assumes50%ofthecalculated heating andcoolingrequirements:35.9kWh/(m2_year)_for _heating_and 1.9kWh/(m2_year)_for_cooling;

-OP2,whichtypiﬁesamoderateoccupancyofthehousewitha modestHVAClevel,assumes10%ofthecalculatedannual heat-ingandcoolingrequirements:7.2kWh/(m2_year)_for_heating_and 0.4kWh/(m2_year)_for_cooling.

Amaintenanceactivityscheduleforeachbuildingcomponent hasbeenestablishedbasedondatafrom[14,44,48]andmaterial

producers.Themaintenanceactivitiesincludeinteriorand exte-riorpaintingofwalls,varnishingofwoodsurfaces,windowglazing replacement,andbitumenlayerﬁxing.Moredetailscanbefound in[43].

4.3. Modelsimpliﬁcations

SomesimpliﬁcationshavebeenemployedintheLCmodel. Fol-lowingtheapproachimplementedinotherLCstudieswithsimilar purposes[8,9,16,27],theenergyrequirementofoperation-phase activitiesthatarenotaffectedbytheexteriorwallsolution (elec-tric appliances,lighting,cooking, domestichot water)have not beenincluded.Concerningtheconstructionphase,theequipment usedand thetransportationof labourerstowork sitehave not beenincludedbecausetherelevanceoftheseprocessesisminor in residential buildings [9]. Theenergy and impacts associated withtheproduction andtransportationoffurniture,HVACheat distributionequipment,sanitaryceramicsandplumbingpipesfor

(6)

Table3

Exteriorwallhousescenarioswithsectiondetails.

Insulationlayer:XPS(extrudedpolystyrene);EPS(expandedpolystyrene).

waterconsumptionhavealsonotbeenconsidered.Theend-of-life phase(dismantlementandwastetreatmentscenarios)hasnotbeen includedbecauseitisconsideredtobeofminorimportancefor SouthEuropeansingle-familyhouses,anditisadifﬁcultphaseto foresee,particularlybecausebuildingsarelonglasting[6]. Accord-ingtoarecentEuropeanstudy[9],theend-of-lifephaseaccounts forlessthan3.2%oftheoverallenvironmentalimpactsofSouth Europeandwellings.

5. Resultsanddiscussion

Inthissection,life-cycleprimaryenergyrequirement(CED)and environmentalresults(CML2001andEco-indicator’99)are pre-sentedandcomparedinthreeparts.Firstly,Section5.1showsthe base-case houselife-cycle impacts fortwo operationalpatterns (OP1andOP2,describedinSection4.2).Fivelife-cycleprocessesare characterized:materialproductionandtransport;heating;cooling; andmaintenance.Secondly,Section5.2presentstheexteriorwall housescenarioanalysisfocusingonthethreeprocesses(material production,transport,maintenance)thathavedifferentimpactsin thevariousscenarios.Finally,Section5.3comparesresultsfromthe threeimpactassessmentmethods.

Table4

Houseconstructionmaterials:weightandtransportationdistances.

Constructionmaterials Mass(ton) Distance(km)

Exteriorwalls H0:Brick 39.26 65 H1:Brick 62.82 65 H2:Cblocks 40.86 65 H3:TCblocks 24.73 65 H4:AACblock 24.73 65 H5:Wood+brick 25.19 65 H6:Wood 8.50 65 Othercomponents Brick 8.74 65 Cementmortar 13.47 65 Anhydritescreed 14.54 65 Plaster 24.28 65 XPS/EPS 0.28 85 Gravel 42.19 170 PET 0.01 75 Bitumen 0.78 75 Concrete 84.23 40 Steel 2.81 130 Wood 2.12 65 5.1. Life-cycleimpacts

5.1.1. Cumulativeenergydemandresults

Fig.2presentstheCEDcalculatedforthebase-casedwellingfor twooperationalpatterns(OP1,OP2),andforanadditionalscenario, RCCTE.TotalCED,non-renewable(Non-R)andrenewable(Renew) energyrequirementsaredistinguishedforeachscenario.Itshould benotedthattheRCCTE(100%)scenarioassumesahypothetical permanentoccupancy,andconditioning,ofthebuildingwith inte-riorseasonalset-points(20◦Cforheatingseason,25◦Cforcooling season),withoutconsideringanyreductionfactortoaccountfor intermittentHVAClevels.Ontheotherhand,thetwooperational patterns,OP1(50%)andOP2(10%), assumereductionfactorsto representdifferentoccupancyand comfort levelsofPortuguese householders.

AscanbeseeninFig.2,thereisaninversionofthemost sig-niﬁcantLCphasefromoneoperationalpatterntotheother.The resultsobtainedunderOP1showthattheusephaseisthemost sig-niﬁcantduetoheating,whereasunderOP2,theconstructionphase hasthehighestimpactsowingtomaterialproduction.Concerning totalOP1results,heatingthehouserepresents56%ofthetotalLC

Fig.2. Life-cycleCEDresultsforbase-casehouse:twooperationalpatterns(OP1 andOP2)andRCCTE.

(7)

Fig.3.CML2001normalizedresultsforbase-casehouse:twooperationalpatterns (OP1andOP2).

primaryenergyrequirements(CED),whereasmaterialproduction holdsaround29%,andtheremainingprocessesrepresentabout5%. However,underOP2,materialproductionrepresents56%oftotal CED,whileheatingaccountsonlyfor22%;maintenanceand trans-portbecomemoresigniﬁcant(around10%)andcoolingdecreases itsimportance(2%).Itcanalsobeobservedthatthecontribution ofthedifferentLCprocessesisverysimilarfortotalCEDandfor non-renewableenergyfraction,sincethelatterhasamajorrole. Thesingle-familyhousetotalCEDamountsto1,751,000MJequiv. (265MJequiv./(m2_year))_under_OP1,_and_around_905,500_MJ_equiv. (137MJequiv./(m2_year))_considering_OP2._Concerning_the_RCCTE (100%)scenario,theprimaryenergyofheatingprocessrepresents around70%oftheLCimpacts.Moreover,undertheRCCTEscenario, theCEDofheatingishigherthanthetotalCEDinOP1.This compar-isonshowsthatforthePortuguesecontextitiscriticaltoconsider theintermittentHVACoperationalpatterns,especiallywhen heat-ingandcoolingloadsarecomparedwithotherLCprocesses.

5.1.2. CML2001results

CML2001environmentalimpactsarepresentedinTable5for thevariousLCprocessesofthebase-casehouseassumingtwo oper-ationalpatterns.Fig.3presentsCML2001normalizedresults.The inversionfoundinenergyresultsfromOP1toOP2,betweenthe constructionandusephases,isalsovalidforCMLresults. Consid-eringOP1,heatingisthemostsignificantprocessineightoutof tencategories(exceptions:OLDandfreshwaterecotoxicity), hold-ingaround50%to80%oftheenvironmentalimpacts.However,in OP2,materialproductionisthemostsignificantineightcategories (exceptions:acidificationandterrestrialecotoxicity),holdingfrom 48%to85%ofthehouselife-cycleimpacts.ConcerningOLDimpacts, materialproductionisthemostsignificantprocessunderthetwo operationalpatternsduetotheuseofXPSasbuildingenvelope insulation.

RegardingtotalnormalizedimpactsshowninFig.3,marine eco-toxicitypresentsthehighestimpacts,followedby,indescending order,abioticdepletion,terrestrialecotoxicity,acidification,GWP, andeutrophication.However,itshouldbenotedthatthetoxicity categoriesstilllackscientific robustness[8,32],which limitsthe significanceoftoxicityresults.

Fig.4.EI’99normalizedresultsforbase-casehouse:twooperationalpatterns(OP1 andOP2).

5.1.3. Eco-indicator’99results

TheEcoindicator’99(EI’99)resultsforthebase-casehouseare giveninTable6distinguishingthetwooperationalpatterns.Fig.4

shows the EI’99 normalized results. Assuming OP1, heating of thehouseisthemostsigniﬁcantprocessforsevenoutofeleven categories,representing46–69%oftheimpacts(exceptions: respi-ratoryorganics,radiation,OLD,andlanduse).Conversely,adopting OP2,materialproductionbecomesthemostimportantprocessin everycategory,beingresponsiblefor43–80%oftheenvironmental impactsofthehouse.ConcerningtotalEI’99normalizedimpacts, fossilfuelisthemostsigniﬁcantcategory,followedby(in descend-ingorder)respiratoryinorganicandclimatechangecategories. 5.2. Exteriorwallscenarioanalysis

5.2.1. CEDresults

Fig.5 presents the CED resultsfor theseven house scenar-ios,excludingheatingandcoolingbecausetheseremainconstant betweenscenarios.Ascanbeseen,materialproductionrequires signiﬁcantly more energy than transportation or maintenance. Concerningnon-renewableCED,thescenariowithlowestenergy requirementisthewoodwallhouse(H6).Incomparisonwiththe base-casescenario(H0),threescenariosreducethenon-renewable CED(H6,H5,H2)andonlyone(H1)requiresmoreenergy.Itcan alsobenotedthatthereisasigniﬁcantdifferenceinnon-renewable

(8)

Table5

LCIACML2011resultsforthebase-casehouse,includingtwoalternativeoperationalpatterns(OP1andOP2).

Abioticdepletion(AD);acidiﬁcation(Acid);eutrophication(Eut);ozonelayerdepletion(OLD);photochemicaloxidation(PO);humantoxicity(Htox);Freshwater ecotoxicity(Fecot.);marineecotoxicity(Mecot.);terrestrialecotoxicity(Tecot.).Thehighestvaluesforeachimpactcategoryhavetheirresultsunderlined(continuous lineforOP1,dashedlineforOP2).

CED results between the various house scenarios: H6 requires 176,350MJequiv.lessthanH1(areductionofabout26%).This vari-ationismostlyduetomaterialproductionCED.Ontheotherhand, thescenarioswithhigherwoodcontent(H6,H5)havemore renew-ableenergyincorporatedinthematerialproduction,sincewoodis consideredarenewablesourceofenergy.

5.2.2. CML2001results

CML2001LCIAresultsforthevarioushousescenariosare pre-sentedinTable7,andnormalizedLCIAresultsareshowninFig.6. Itcanbeobservedthatthescenariowiththelowestenvironmental impactisthewoodwallhouse(H6),innineoutoftencategories.On theotherhand,thescenarioswithhighestenvironmentalimpacts areH3(ﬁvecategories),H1(GWPandabioticdepletion)andH4 (twoecotoxicitycategories).

The comparison with the base-case house (H0) shows that threescenarios(H2,H5, andH6,indescendingorder)have sig-niﬁcantlylowerimpactsforGWP,abioticdepletion,acidiﬁcation andeutrophicationcategories.Concerningphotochemical oxida-tion,onlyH5andH6havelowerimpactsthanH0.RegardingOLD, H6hasthehighestimpact,whichisduetothetypeandthickness oftheinsulationmaterialassumed(XPS);thescenarioswithlower

impactsarethosewithEPSinsulationappliedwithinETICS(H2,H3 andH4).Lookingattheresultscategorybycategory,the follow-ingrelativevariationsbetweenthescenarioswiththelowestand highestimpactshavebeendeterminedbasedonLCdatapresented inFig.6:

•Abioticdepletion:H6has27%lessSbequiv.thanH1; •Acidiﬁcation:H6has38%lessSO2equiv.thanH3;

•Eutrophication:H6has21%lessPO4equiv.thanH3;

•GWP:H6has58%lessCO2equiv.thanH1;

•OLD:H2,H3andH4have58%lessCFC-11equiv.thanH6; •Photochemicaloxidation:H6has37%lessC2H4equiv.thanH3. 5.2.3. Eco-indicator’99results

EI’99 LCIA results for the seven scenarios are presented in

Table8.NormalizedLCIAresultsareshowninFig.7.Regardingthe scenariowiththelowestenvironmentalimpacts,similarresultsto theCMLapproachcanbeobserved:thewoodwallscenario(H6) performsbetterinnineoutofelevencategories.Thescenarioswith thehighestenvironmentalimpacts areH1(four categories),H3 (threecategories),H4andH6(twocategoriesforeachone).The comparisonwiththebase-casescenarioshowsthatthescenarios Table6

LCIAEI’99resultsforthebase-casehouse,includingtwooperationalpatterns(OP1andOP2).

Carcinogens (Carci); climate change (CC); ozone layer depletion (OLD); radiation; respiratory inorganics (R. inorg); respiratory organics (R. org.); acidiﬁcation–eutrophication(acid/eut);ecotoxicity(Ecot);landuse(L.use);fossilfuels(F.fuels).Thehighestvaluesforeachimpactcategoryhavetheirresults underlined(continuouslineforOP1,dashedlineforOP2).

(9)

withwoodintheircomposition(H5andH6)havelowerimpactsfor ninecategories(exceptlanduseandOLD).H2haslowerimpacts thanbase-caseforﬁvecategories,andH4forthreecategories(OLD, respiratoryorganicsandlanduse).Concerningthelanduse cate-gory,woodwallhouse(H6)hasthehighestimpactsbecauseofland usechangeduetologging.However,thishighimpactisdebatable ifloggingisbasedonasustainableforestmanagementstrategy.

Eco-indicator’99normalizedresultshavebeenfurthergrouped in three damage categories:resources; ecosystem quality; and humanhealth.H6isthescenariowithlowestimpactsinresources andhumanhealth butwithhighest impactsinecosystem qual-ity(duetolanduseimpactcategoryresults).ScenarioH4hasthe lowestimpactsinecosystemquality.AssessingtheEI’99results, categorybycategory,thefollowingvariationsbetweenthe scenar-ioswiththelowestandhighestimpactshavebeendetermined:

•Fossilfuels:H6has31%lessMJsurplusthanH1;

•Minerals:H6has14%lessMJsurplusthanH3;

•Acidiﬁcation/eut.:H6has27%lessPDFthanH1,H3;

•Climatechange:H6has59%lessDALYthanH1;

•Ozonelayer:H2has58%lessDALYthanH6;

•Resp.organics:H6has9%lessDALYthanH1;

•Carcinogens:H6has9%lessDALYthanH4;

•Radiation:H6has21%lessDALYthanH4;

•Resp.inorganics:H6has23%lessDALYthanH3;

•Ecotoxicity:H6has12%lessPDFthanH3;

•Landuse:H4has68%lessPDFthanH6;

•Resources:H6has30%lessMJsurplusthanH1;

•Ecosystemquality:H4has52%lessPDFthanH6;

•Humanhealth:H6has28%lessDALYthanH1.

5.3. Comparingthethreeimpactassessmentmethods

ThethreeLCIAmethods usedinthis studyhavevery differ-entapproaches.CEDisasingle-issuemethodaccountingonlyfor primaryenergyrequirements.Thetwoothersareenvironmental LCIAmethods:CML2001isproblem-orientedandEI’99is damage-oriented. In general, thethree methods indicatethat themost importantLCphaseandprocessdependontheoperationalpattern assumed.AspresentedinSection5.1,whenheatingandcooling loadsarereduced,theusephaseimpactisdecreasedandthe pre-useprocessesbecomerelativelymoreimportant,mainlydueto materialproduction.

However,amoredetailedassessment,attheindividual cate-gorylevel,showssomeimportantdifferences.Namely,CMLresults presenthigherimpactsfortheusephase,whilethoseofEI’99are higherformaterialproduction.Forexample,consideringOP1,the heatingisresponsibleformorethan50%oftheimpactsforeight CMLcategories,butonlyforsixEI’99categories.ConcerningOP2, materialproductionisthemostsignificantprocessforallEI’99 cate-gories,butonlyforeightoutoftenCMLcategories.Thecomparison ofCMLandEI’99normalizedresults(Figs.3and4)showsthatthe mostsignificantcategoryforEI’99isfossilfuels,whereasforCMLit ismarineecotoxicity.ThesecondmostsignificantEI’99categoryis respiratoryinorganics,whileforCMLitisabioticdepletion.Itcan beconcludedthatnormalizedEI’99andCMLresults,ingeneral,will leadtodifferentresultsintermsofthemostsignificantcategories. Concerning the exterior wall scenario analysis, the three methodsindicate thesame scenariowiththe best(energy and environmental)performance.Theresultsshowthatthewoodwall scenario (H6)has the lowest impacts for almost all categories (ozonelayerdepletionandlandusearetheexceptions).To deter-minein detail theextenttowhichtheresultsof ourstudyare influencedbytheLCIAmethodchosen,similarcategoriesofthe

threeLCIAmethodshavebeencomprehensivelyassessedforthe Table

7 LCIA CML 2001 scenario analysis results. House scenario AD (kg Sb equiv.) Acid (kg SO 2 equiv.) Eut (kg PO 4 equiv.) GWP (kg CO 2 equiv.) OLD (g CFC-11 equiv.) PO (kg C2 H4 ) Htox (kg 1,4-DB equiv.) Fecot (kg 1,4-DB equiv.) Mecot (g 1,4-DB equiv.) Tecot (kg 1,4-DB equiv.) H0 2.67E + 02 1.47E + 02 2.32E + 01 4.38E + 04 8.32E − 02 9.23E + 00 1.52E + 04 4.17E + 03 5.69E + 06 1.82E + 02 H1 2.98E ± 02 1.61E + 02 2.45E + 01 5.01E ± 04 8.37E − 02 1.03E + 01 1.58E + 04 4.08E + 03 6.08E + 06 1.85E + 02 H2 2.57E + 02 1.38E + 02 2.26E + 01 3.85E + 04 3.93E − 02 1.07E + 01 1.59E + 04 4.43E + 03 5.65E + 06 1.90E + 02 H3 2.91E + 02 1.94E ± 02 2.48E ± 01 4.32E + 04 4.00E − 02 1.26E ± 01 1.65E ± 04 4.52E + 03 6.14E ± 06 1.92E + 02 H4 2.76E + 02 1.44E + 02 2.33E + 01 4.30E + 04 3.97E − 02 1.06E + 01 1.65E ± 04 4.58E ± 03 5.97E + 06 1.97E ± 02 H5 2.45E + 02 1.34E + 02 2.15E + 01 3.25E + 04 8.29E − 02 8.71E + 00 1.49E + 04 4.15E + 03 5.52E + 06 1.78E + 02 H6 2.18E + 02 1.20E + 02 1.96E + 01 2.10E + 04 9.34E − 02 7.95E + 00 1.43E + 04 4.06E + 03 5.13E + 06 1.71E + 02 Abiotic depletion (AD); acidiﬁcation (Acid); eutrophication (Eut); ozone layer depletion (OLD); photochemical oxidation (PO); human toxicity (Htox); fresh water ecotoxicity (Fecot.); marine ecotoxicity (Mecot.); terrestrial ecotoxicity (Tecot.). For each impact category, the lowest result is written in bold and the highest is underlined.

(10)

Fig.6.CML2001normalizedresults:scenarioanalysis.

Table8

LCIAEI’99scenarioanalysisresults.

HouseHumanhealth(DALY) Ecosystemquality(PDF*m2_y) _Resources_(MJ) Carci CC OLD Radiation R.inorg. R.org. Acid/eut. Ecot. L.use F.fuels Minerals H0 2.07E−03 9.12E−03 8.80E−05 1.56E−04 2.78E−02 8.91E−05 8.23E+02 7.68E+02 2.21E+03 5.76E+04 1.75E+03 H1 2.04E−03 1.04E−02 8.85E−05 1.66E−04 3.01E−02 9.51E−05 9.18E±02 7.86E+02 2.05E+03 6.54E±04 1.81E+03 H2 2.05E−03 8.04E−03 4.19E−05 1.56E−04 2.75E−02 8.96E−05 7.75E+02 8.14E+02 2.22E+03 5.42E+04 1.90E+03 H3 2.12E−03 9.02E−03 4.26E−05 1.60E−04 3.13E−02 9.26E−05 8.88E+02 8.31E±02 1.99E+03 6.47E+04 1.94E±03 H4 2.13E−03 8.98E−03 4.24E−05 1.73E−04 2.79E−02 8.90E−05 7.96E+02 8.23E+02 1.90E+03 5.84E+04 1.93E+03 H5 1.99E−03 6.74E−03 8.77E−05 1.50E−04 2.62E−02 8.83E−05 7.53E+02 7.57E+02 4.01E+03 5.19E+04 1.74E+03

H6 1.94E−03 4.32E−03 9.88E−05 1.38E−04 2.42E−02 8.68E−05 6.71E+02 7.33E+02 5.87E±03 4.51E+04 1.68E+03

Carcinogens(Carci);climatechange(CC);ozonelayerdepletion(OLD);radiation;respiratoryinorganics(R.inorg.);respiratoryorganics(R.org.);acidiﬁcation–eutrophication (Acid/Eut);ecotoxicity(Ecot.);landuse(L.use);fossilfuels(F.fuels).Foreachimpactcategory,thelowestresultiswritteninboldandthehighestisunderlined.

scenarioanalysisperformed.Fig.8comparestheresultsofthe exte-riorwallhousescenariosforeachsetofsimilarimpactcategories. Eachgraphpresentstheimpactsofthevariousscenariosrelatively (asapercentage)tothescenariowiththehighestimpact(100%).

ItcanbeseeninFig.8thatnon-renewableCED,abiotic deple-tion(CML),andresources(EI’99)showsimilarresults.Furthermore, theGWPandOLDCMLcategorieshaveidenticalresultstotheir counterpart EI’99 categories (climate change and ozone layer).

AcidiﬁcationandeutrophicationaretwodistinctcategoriesinCML butjustoneinEI’99(seeTable1).Ingeneral,ananalogous order-ingofscenarioimpactscanbeobservedinCMLandEI’99,butEI’99 resultsaremoreclosetoCMLacidiﬁcationresults.Regarding eco-toxicity,photochemicaloxidation andhumantoxicity,H6 isthe scenariowithlowestimpactsinbothCMLandEI’99.However,the scenariowiththehighestimpactandtherankingoftheremaining scenariosdifferforthetwomethods.

(11)

Fig.8. RelativecomparisonofCED,CMLandEI’99similarcategories:scenarioanalysisresults.

Comparingnon-renewableCEDwiththetwootherLCIA meth-ods(multi-category),itcanbeseenthat,beyondbeingsimilarto abioticdepletionandresources,non-renewableCEDshowssome correlationwithGWPandClimateChangeaswellaswith acidifica-tion/eutrophication.However,nocorrelationcanbeseenbetween CEDandtheremainingCMLandEI’99impactcategories.Thiscanbe explainedbecauseGWP,acidificationandeutrophicationimpacts inthisresearcharemainlyassociatedwithfossilenergyuse,which is characterizedby CED non-renewable results. Concerning the comparisonbetweenthetwoLCIAmulti-categorymethods(CML andEI’99),GWPandOLDcategoriesshowthemostrobust compari-sonofresults,beingfollowedbyabioticdepletion,acidificationand eutrophication.Nevertheless,CMLandEI’99methodscanpresent inconsistentresultsbetweensimilarcategories(differentranking), whichultimatelycaninfluencethechoicebetweensolutions.

6. Conclusions

A life-cycle model hasbeen implemented for a Portuguese single-family house focusing on the life-cycle (LC) phases and processesaffected bythe buildingenvelopesolutions (material productionandtransport,heating,cooling,maintenance)to deter-minethevariousassociatedenergyandenvironmentalimpacts. Sevenexteriorwallhousescenarioswithdifferentmaterialsintheir composition(hollow brick,facingbrick, concreteblocks, wood) havebeencomparativelyassessedinordertosupportthe selec-tionofmoresustainableexteriorwalls.Twooperationalpatterns havebeenconsideredtorepresentdistinctoccupancypatternsand comfortlevelsofthehouse.ToassessthebuildingLCperformance andthesevenscenarios,threelife-cycleimpactassessment(LCIA) methodshavebeenused:(i)cumulativeenergydemand(CED);(ii) CML2001;(iii)Eco-indicator’99(EI’99).AcomparisonoftheLCIA methodshasbeenperformedtodeterminetowhatextenttheLCA resultsareinﬂuencedbythemethodapplied.

TheresultscalculatedforthethreeLCIAmethodsshowthatthe mostsigniﬁcantLCphaseandprocessarehighlyassociatedwiththe houseoperationalpatterns:forPortuguesehouseswithreduced

HVAClevels,materialproductionbecomesthemostimportant pro-cess.ThisﬁndingcontradictsmostofthepreviousLCAstudiesfor conventionalbuildings,becausethesestudieshavebeenperformed forcoldclimates,assumingpermanentoccupancyandhighcomfort levels.Moreover,forthePortuguesecontextitiscriticalto con-sidertheintermittentHVACoperationalpatterns,especiallywhen heatingandcoolingloadsarecomparedwithotherLCprocesses. Concerningtheexteriorwallscenarioanalysis,thethreeLCIA meth-odsindicatethatH6(woodwall)isthepreferablesolutionwiththe lowestimpactsformostcategories,whereasthealternativeswith higherimpactsareH1(doublewallwithfacingbrick),H3(thermal concreteblockwall)andH4 (autoclavedaeratedconcreteblock masonry).

RegardingthecomparisonofresultsfromthethreeLCIA meth-ods,ananalogousrankingofexteriorwallalternativesisobserved fornon-renewableCED,abioticdepletion(CML2001)andresources (EI’99).Non-renewableCEDresultsalsopresentsomecorrelation withCML2001andEI’99fortwosetsofhomologouscategories: GWP/climatechange;acidificationandeutrophication.However, nocorrelationcanbeseenwiththeremainingLCIAimpact cate-gories.ConcerningCML2001andEI’99environmentalimpacts,the resultsformostcategoriesarerobustandpermitastraightforward comparison. Nevertheless, these methods present inconsistent resultsfor humantoxicity,eco-toxicity and photochemical oxi-dation,namelyadifferentrankingofthealternativewalls,which ultimatelycaninfluencethechoiceamongsolutions.Inaddition, furtherdivergenceshavebeenobservedbetweenCML2001and EI’99,namelybetweenthemostsignificantnormalizedcategories (fossilfuelsforEI’99,marineecotoxicityforCML)andresults(CML presentsslightlyhigherimpactsfortheusephase,whileEI’99for materialproduction).

To sum up, although the wood wall has been pointed out asthepreferablesolutionbythethreemethods,theresultsfor theremainingexteriorwallscenariosareinﬂuencedbytheLCIA method applied, in particular for the toxicity categories. Our researchreinforcestheideathatdecision-makingbasedontheLCA ofdwellingsshouldincludemultipleenvironmentalimpacts,but shouldnotaddresstoxicitycategories.

(12)

Acknowledgements

Helena Monteiro is grateful for the ﬁnancial support pro-videdby FCT(Portuguese Scienceand TechnologyFoundation), undertheprogram MITPortugal –SustainableEnergySystems, through the doctoral degree grant SFRH/BD/33736/2009. Sup-port from the FCT in the scope of the following projects is alsoacknowledged:PTDC/TRA/72996/2006;MIT/SET/0014/2009; MIT/MCA/0066/2009;andPTDC/SEN-TRA/117251/2010.Wethank the anonymous reviewers for their comments, which helped improvethisarticle.WealsothankMITWriting and Communi-cationCenterforthehelponmattersofclarity.

References

[1]EPBD,Directive2002/91/ECoftheEuropeanParliamentandoftheCouncilof 16December2002ontheenergyperformanceofbuildings,OfﬁcialJournalof theEuropeanUnion(2002),L1/65-1/71.

[2] RCCTE,RegulamentodasCaracterísticasdeComportamentoTérmicodos Edifí-cios,CodeoftheBuildingsThermalBehaviourCharacteristics–DL80/2006, MinistryofPublicWorks,TransportationandCommunications,2468–2513, 2006.

[3] EPBD(recast),Directive2010/31/EUoftheEuropeanParliamentandofthe Councilof19May2010ontheenergyperformanceofbuildings(recast),2010, pp.L153/13–153/35.

[4]G.Blengini,T.DiCarlo,Thechangingroleoflifecyclephases,subsystemsand materialsintheLCAoflowenergybuildings,EnergyandBuildings42(6)(2010) 869–880.

[5]S.Citherlet,T.Defaux,Energyandenvironmentalcomparisonofthreevariants ofafamilyhouseduringitswholelifespan,BuildingandEnvironment42(2007) 591–598.

[6]B.Peuportier,Lifecycleassessmentappliedtothecomparativeevaluationof singlefamilyhousesintheFrenchcontext,EnergyandBuildings33(5)(2001) 443–450.

[7] O.Ortiz,J.C.Pasqualino,G.Díez,F.Castells,Theenvironmentalimpactofthe constructionphase:anapplicationtocompositewallsfromalifecycle per-spective,ResourcesConservationandRecycling54(2010)832–840. [8]F.Nemry,A.Uihlein,C.M.Colodel,C.Wetzel,A.Braune,B.Wittstock,I.Hasan,J.

Kreißig,N.Gallon,S.Niemeier,Optionstoreducetheenvironmentalimpactsof residentialbuildingsintheEuropeanUnion– potentialandcosts,Energyand Buildings42(7)(2010)976–984.

[9]F.Nemry,A.Uihlein,C.M.Colodel,B.Wittstock,A.Braune,C.Wetzel,I.Hasan, S.Niemeier,Y.Frech,EnvironmentalImprovementPotentialsofResidential Buildings(IMPRO-Building),ReportJRC:EUR23493EN,Luxembourg,2008. [10]C.D.Frenette,C.Bulle,R.Beauregard,A.Salenikovich,D.Derome,Usinglife

cycleassessmenttoderiveanenvironmentalindexforlight-framewoodwall assemblies,BuildingandEnvironment45(2010)2111–2122.

[11] L.C.Dreyer,A.L.Niemann,M.Hauschild,ComparisonofthreedifferentLCIA methods:EDIP97,CML2001andEco-indicator99,InternationalJournalLCA8 (4)(2003)191–200.

[12] K.Adalberth,Energyuseduringthelifecycleofsingle-unitdwellings: exam-ples,BuildingandEnvironment32(4)(1997)321–329.

[13]K.Adalberth,Energyuseduringthelifecycleofbuildings:amethod,Building andEnvironment32(4)(1997)317–320.

[14]S.Blanchard,P.Reppe,LifecycleanalysisofaresidentialhomeinMichigan, CenterforSustainableSystems,AnnArbor,MI,ReportNo.CSS98-05,University ofMichigan,1998.

[15]G.A.Keoleian,S.Blanchard,P.Reppe,Life-cycleenergycostsandstrategies forimprovingasingle-familyhouse,JournalofIndustrialEcology4(2001) 135–156.

[16]I.Z.Bribián,A.A.Usón,S.Scarpellini,Lifecycleassessmentinbuildings: state-of-the-artandsimpliﬁedLCAmethodologyasacomplementforbuilding certiﬁcation,BuildingandEnvironment44(2009)2510–2520.

[17]L.Gustavsson,A.Joelsson,Lifecycleprimaryenergyanalysisofresidential buildings,EnergyandBuildings42(2010)210–220.

[18]L.Gustavsson,A.Joelsson,R.Sathre,Lifecycleprimaryenergyuseand car-bonemissionofaneight-storeywood-framedapartmentbuilding,Energyand Buildings42(2010)230–242.

[19]J.N.Hacker,T.P.DeSaulles,A.J.Minson,M.J.Holmes,Embodiedandoperational carbondioxideemissionsfromhousing:acasestudyontheeffectsofthermal massandclimatechange,EnergyandBuildings40(2008)375–384. [20] A.Mahdavi,E.M.Doppelbauer,Aperformancecomparisonofpassiveand

low-energybuildings,EnergyandBuildings42(2010)1314–1319.

[21]C.Thormark,Alowenergybuildinginalifecycle,itsembodiedenergy,energy needforoperationandrecyclingpotential,BuildingandEnvironment37(2002) 429–435.

[22]A.Utama,S.Gheewala,LifecycleenergyofsinglelandedhousesinIndonesia, EnergyandBuildings40(2008)1911–1916.

[23]I.Sartori,A.G.Hestnes,Energyuseinthelifecycleofconventionaland low-energybuildings:areviewarticle,EnergyandBuildings39(3)(2007)249–257. [24] O.Ortiz,F.Castells,G.Sonnemann,Sustainabilityintheconstruction indus-try:areviewofrecentdevelopmentsbasedonLCA,ConstructionandBuilding Materials23(2009)28–39.

[25]T.Ramesh,R.Prakash,K.K.Shukla,Lifecycleenergyanalysisofbuildings:an overview,EnergyandBuildings42(10)(2010)1592–1600.

[26]A.Sharma,A.Saxena,M.Sethi,V.Shree,Lifecycleassessmentofbuildings:a review,RenewableandSustainableEnergyReviews15(2011)871–875. [27] G.Verbeeck,H.Hens,Lifecycleinventoryofbuildings:acontributionanalysis,

BuildingandEnvironment45(2010)964–967.

[28]ISO-14040,ISO14040,EnvironmentalManagement–LifeCycleAssessment– PrinciplesandFramework,2006.

[29]ISO-14044,ISO14044,EnvironmentalManagement–LifeCycleAssessment– RequirementsandGuidelines,2006.

[30]J.Malc¸a,F.Freire,Renewabilityandlife-cycleenergyefﬁciencyofbioethanol andbio-ethyltertiarybutylether(bioETBE):assessingtheimplicationsof allo-cation,Energy31(15)(2006)3362–3380.

[31] R.Frischknecht,N.Jungbluth,H.J.Althaus,G.Doka,R.Dones,R.Hischier,S. Hellweg,S.Humbert,M.Margni,T.Nemecek,M.Spielmann,Implementation oflifecycleimpactassessmentmethods:datav2.0,EcoinventreportNo.3, Dübendorf,CH,2007.

[32] L.Barnthouse,J.Fava,K.Humphreys,R.Hunt,L.Laibson,S.Noesen,G.Norris,J. Owens,J.Todd,B.Vigon,K.Weitz,J.Young,Lifecycleimpactassessment:the state-of-the-art,SETACPress,FL,1998.

[33]J.B.Guinée,M.Gorrée,R.Heijungs,G.Huppes,R.Kleijn,A.deKoning,L.van Oers,A.WegenerSleeswijk,S.Suh,H.A.UdodeHaes,H.deBruijn,R.van Duin,M.A.J.Huijbregts,HandbookonLifeCycleAssessment–Operational GuidetotheISOStandards.I.LCAinPerspective;IIa:Guide;IIb:Operational AnnexIII:Scientiﬁc Background,KluwerAcademicPublishers, Dordrecht, 2002.

[34]M.Goedkoop,R.Spriensma,TheEco-indicator99:adamageorientedmethod forlifecycleimpactassessment,Methodologyreport,2001.

[35]M.Goedkoop,R.Spriensma,TheEco-indicator99:adamageorientedmethod forLifeCycleImpactAssessment,MethodologyAnnex,2001.

[36] G.Finnveden,etal.,Recentdevelopmentsinlifecycleassessment,Journalof EnvironmentalManagement91(1)(2009)1–21.

[37]T.Ekvall,B.P.Weidema,Systemboundariesandinputdatainconsequentiallife cycleinventoryanalysis,TheInternationalJournalofLifeCycleAssessment9 (3)(2004)161–171.

[38]P.S.Vieira,A.Horvath,Assessingtheend-of-lifeimpactsofbuildings, Environ-mentalScience&Technology42(13)(2008)4663–4669.

[39] P.Lesage,T.Ekvall,L.Deschênes,R.Samson,Environmentalassessmentof brownﬁeldrehabilitationusingtwodifferentlifecycleinventorymodels.Part 1.Methodologicalapproach,InternationalJournalofLifeCycleAssessment12 (6)(2007)391–398.

[40] P.Lesage,T.Ekvall,L.Deschênes,R.Samson,Environmentalassessmentof brownﬁeldrehabilitationusingtwodifferentlifecycleinventorymodels.Part 2.Casestudy,InternationalJournalofLifeCycleAssessment12(7)(2007) 497–513.

[41]H.Monteiro,F.Freire,Lifecycleenergyandenvironmentalassessmentof alter-nativeexteriorwallsystems,in:L.Braganc¸a,M.Pinheiro,R.Mateus,etal.(Eds.), PortugalSB10SustainableBuildingAffordabletoAll–LowCostSustainable SolutionsProceedings,Vilamoura,Portugal,2010,pp.777–784.

[42]H.Monteiro,F.Freire,Environmentallife-cycleassessmentofasingle-family houseinPortugal:assessingalternativeexteriorwallswithtwomethods,Gazi UniversityJournalofScience24(3)(2011)527–534.

[43]H.Monteiro,LifecycleassessmentofaPortuguesehousewithdifferent exte-riorwallsolutionsandalternativeheatingsystems,MScthesisinEnergyfor Sustainability,UniversityofCoimbra,2010.

[44]D.Kellenberger,H.J.Althaus,N.Jungbluth,T.Künniger,M.Lehmann,P. Thal-mann,Lifecycleinventoriesofbuildingproducts,FinalreportecoinventData v2.0No.7,Dübendorf,CH,2007.

[45]M.Spielmann,R.Dones,C.Bauer,M.Tuchschmid,Lifecycleinventoriesof transportservices,EcoinventreportN◦14,v2.0,Dübendorf,CH,2007. [46]ENISO13790,Energyperformanceofbuildings–calculationofenergyusefor

spaceheatingandcooling,ISO,2008.

[47]ODYSSEE Energy Efﬁciency Database. http://www.odyssee-indicators.org/ registred/online-indicators/(accessedSeptember,2011).

[48]H.Künzel,H.M.Künzel,K.Sedlbauer,Long-termperformanceofexternal ther-malinsulationsystems(ETICS),ArchitecturaZeitschriftFürGeschichteDer Baukunst5(2006)11–24.