Kent Academic RepositoryFull text document (pdf)

Copyright & reuse

Content in the Kent Academic Repository is made available for research purposes. Unless otherwise stated all

content is protected by copyright and in the absence of an open licence (eg Creative Commons), permissions

for further reuse of content should be sought from the publisher, author or other copyright holder.

Versions of research

The version in the Kent Academic Repository may differ from the final published version.

Users are advised to check http://kar.kent.ac.uk for the status of the paper. Users should always cite the

published version of record.

Enquiries

For any further enquiries regarding the licence status of this document, please contact:

researchsupport@kent.ac.uk

If you believe this document infringes copyright then please contact the KAR admin team with the take-down

information provided at http://kar.kent.ac.uk/contact.html

Citation for published version

Nikolopoulou, Marialena and Kotopouleas, Alkis and Lykoudis, Spyridon (2018) From indoorsto outdoors and in-transition; thermal comfort across different operation contexts. In: 10th WindsorConference: Rethinking Comfort, 12-15 April 2018, Windsor.

DOI

Link to record in KAR

http://kar.kent.ac.uk/66790/

Document Version

Author's Accepted Manuscript

Fromindoorstooutdoorsandin-transition;thermalcomfortacrossdifferent

operationcontexts

MarialenaNikolopoulou

1,AlkisKotopouleas

1andSpyridonLykoudis

2

1CentreforArchitectureandSustainableEnvironment(CASE),KentSchoolofArchitecture,

UniversityofKent,CanterburyCT27NZ,M.Nikolopoulou@kent.ac.uk,A.G.Kotopouleas@kent.ac.uk

2 IndependentResearcher,Athens,Greece,slykoud@yahoo.com

Abstract: This paper focuses on the investigation of thermal comfort conditions in three very differentoperationalcontextsusingmeta-analysisofdifferentstudieswithinasimilarclimaticcontextintheUK.Thisincludesextensive surveys indoors fromoffices,outdoors fromurbanareas, aswell as indoors fromairportterminals.Recentresearchinairportterminalbuildingshashighlightedthatthereareverydifferentusergroups,withdiverserequirementsforthermalcomfortinsuchfacilities.Thepaperinvestigatesthehypothesisthatstaffworkinginthedifferentareashaveneedsmoresimilartothoseofstaffworkinginoffices,whilepassengersusethebuildingasatransitionareawithverydifferentrequirementsandhenceclosertotheoutdoorenvironment.Analysingandcomparing the thermalcomfortconditions fromthedifferentcontexts, itexplores theroleofadaptationforthermalcomfortattainmentandsatisfactionwiththeenvironmentandthesimilaritiesofverydifferentoperationalcontextsintermsoftheirthermalcomfortcharacteristics.Finally,thepaperhighlightedtechniquesforthepotentialtransformationofthermalcomfortscales,whichcanenablecomparisonbetweendifferenttypesofsurveysandinformthewiderthermalcomfortdebate.

Keywords:meta-analysis,adaptation,scaletransformation,surveys

1.! Introduction

In the last 20 years, the field of thermal comfort has witnessed a significant increase inthermal comfort surveys in different operational contexts, which has provided a broaderperspectivefromwhichtoviewcomfort inurbanenvironments. Ithasalsoenabledustounderstand adaptation processes more closely and evaluate the subtle ways which theypresentthemselvesandtheirimportanceinachievingthermalcomfortindifferentcontexts.

This paper focuses on thermal comfort in three very different contexts; in offices,outdoorurbanspacesandairportterminals,usingmeta-analysisofdifferentstudieswithinasimilar climatic context in the UK. Recent research in airport terminal buildings hashighlightedthatthereareverydifferentusergroups,withdiverserequirementsforthermalcomfortinsuchfacilities(KotopouleasandNikolopoulou,2016).Thepaperinvestigatesthehypothesisthatstaffworkinginthedifferentareashaveneedsmoresimilartothoseofstaffworkinginoffices,whilepassengersusethebuildingasatransitionareawithverydifferentrequirementsandhenceclosertotheoutdoorenvironment.Analysingandcomparingthethermal comfort conditions from the different contexts, the paper explores the role ofadaptationforthermalcomfortattainmentandsatisfactionwiththeenvironmentandthesimilarities of very different operational contexts in terms of their thermal comfortcharacteristics.

2.! ResearchFramework

Beforeproceedingwithexplainingthedatasourcesandmethodologyemployedforthestudy, it is worth discussing the development of the hypothesis and the reason for thecomparisonofthedifferentoperationalcontexts.RecentresearchfundedbytheEPSRCtominimisethecarbonfootprintofairportterminalbuildings,identifiedtheoccurrenceoftwodistinct user groups with consistent differences in thermal comfort requirements(Nikolopoulou and Kotopouleas, 2016). Despite the identical environmental operationcontext, the analysis highlighted the difference in the way the terminal is perceived astransitionvs.indoorworkspaceforpassengersandstaffrespectively.

Such differences, which could only be justified by personal and cognitive factorsdiscussedintheframeworkofpsychologicaladaptation(NikolopoulouandSteemers,2003),ledoneoftheauthorstoputforwardthehypothesisthatadaptiveopportunityshouldinfactbetreatedasacontinuum(Nikolopoulou,1998,2004).Nikolopoulouarguedthatononeendofthespectrum,conditionswerefullycontrolledwithnoadaptationpossible,e.g.inclimatechambers,whileon theotherend, conditionswere totallyuncontrolledandvariable, e.g.outdoorswithadaptationdevelopingfullybothphysicallyandpsychologically(Fig.1). Shespeculated that buildingsoccupied various points in between, according to thedegreeofadaptation they allowed for. Fully controlled HVAC buildings not allowing interactionbetweentheoccupantsandthesystemwouldbecloser to theclimatechamber,whereasfree-runningbuildingsclosertotheoutdoorsituation.

Figure1:Schematicdiagramoftheadaptiveopportunitycontinuum(Nikolopoulou,1998)

Followingthiscontinuum,applicationoftheoreticalcomfortmodelscouldthenbecomparedwithoccupants’ thermal comfort conditions. For example, as comfortmodels havebeendevelopedfromsurveysinclimatechambers(e.g.Fanger1970),itwouldbeexpectedthetwotobeidenticalattherespectiveendofthespectrum.Movingtowardstheotherend,thebiggest difference would be expected for outdoor spaces, where research has indeedhighlighted large discrepancies between theoretical models and actual outdoor thermalcomfortconditions(Nikolopoulouetal.,2001;NikolopoulouandLykoudis,2006).Withthebuilt environment falling inbetween these twoextremes,where thebuildingenvelope issealedandtheindoorconditionsarefullycontrolledbyacentralHVACsystem,itwouldbeexpectedthattheoreticalmodelsareveryclosetoactualthermalcomfortconditions,asaresult ofminimal adaptive opportunity. Indeed, thiswas corroborated by de Dear et al.(1997), who demonstrated that the PMV model (ISO 7730) describes well the thermalsensationsforcloselycontrolledbuildings.However,infree-runningbuildingsthedifferencebetweenthetwoincreasessignificantly(deDearetal.,1997).Thisbehaviourcouldbearguedtobeduetothehigherdegreeofadaptationwhereoccupantsinteractwiththebuildingsforenvironmentalcontrol.

Althoughtheabovemodelissimplified,itisreasonabletoassumethatdifferencesinthedegreeof adaptation still exist evenwithineachof thesegenericgroups, althoughofsmaller magnitude. For example, in free-running buildings, the degree of adaptive

Adaptive Opportunity Continuum

+

Climate chambers

Outdoor spaces

Controlled HVAC builings

Free-running buildings

-

opportunitycanvarybetweenacellularandanopenplanroom.Similarly,inoutdoorareas,thereisavarietyofspaces,allowingaccesstosunandshade,etc.

With the recent field surveys from a different building typology, namely airportterminals,wheredistinctthermalcomfortconditionswererevealedfordifferentusergroupseven within the same environment, the speculative model of the adaptive continuum isrevisitedtoevaluatethepossibleroleofadaptiveopportunityandidentifysimilaritieswithotherphysicalcontexts.

3.! Datasources

Thestudycomprisesareviewandmeta-analysisofthreeextensivethermalcomfortdatasets,fromdifferentoperational contexts includingoffices,airport terminalsandoutdoorurbansettings.TheseincludetheASHRAERP-884databasethatwasusedforthedevelopmentofthefirstadaptivethermalcomfortstandardforindoors(ANSI/ASHRAE,2004),theEU-fundedRUROS database for outdoors (Nikolopoulou and Lykoudis, 2006) and the data from theEPSRC-fundedprojectonairportterminals(KotopouleasandNikolopoulou,2016,2018).TheASHRAERP-884andRUROSdatabases includeresultsfromcomfortsurveysfromdifferentcountries around theworld. To enable comparison, between indoors/outdoors aswell asairports,acommongeographicalgroundneededtoselected.HencethefocuswasontheUK.

Officeswereselectedfortheindoorenvironment,toenableabettercomparisonwithworkingconditionsofairportstaff.ThestudiesselectedwerebyNicoletal.inOxford(1996)andbyWilliamsinLiverpool,StHelensandChester(1995).Fortheoutdoorenvironment,theRUROSstudiesfortheUKincludedthesurveysbySteemersetal.inCambridge(2001-02)andKang et al. in Sheffield (2001-02). Finally, for the airport terminals, the surveys byNikolopoulouandKotopouleasfromManchesterTerminals1and2andLondonCityAirport(2013-14)wereemployed. The studies included summerandwinter surveysexcept fromNicol et al. in Oxford (1996) which was carried out in summer only. In some ways, thecomparisonwaslimitedtodatasetsavailableforthespecificcriteriaintheclimaticcontextinvestigated,andtheseweretheonlyonesavailabletotheauthors,i.e.throughthepubliclyavailabledatasetsforindoorsandoutdoorsandthemorerecentworkonairportterminals.AcomparisonoftherelevantstudiesfortheanalysisisshownonTable1.

Overall,thereare1374participantsintheoffices,3087intheairportsand1957intheoutdoor surveys. The environmental parameters monitored are similar, including airtemperature,globetemperature(TglobewasnotcollectedfortheWilliamsstudy;alsoablackglobewasusedfor indoorsandgreyglobeoutdoorsforRUROS),relativehumidityandairmovement. Basedonthesemeasurements itwasthenpossibletocalculatemeanradiantandoperativetemperatures.

Itshouldbehighlightedthattheconditionsincludedamixtureofmixedmode(somebuildings inwinter inWilliams’ study) and free-running case studies (Nicol’s studywas innaturally-ventilatedbuildings,aswassomeofWilliams’buildings,whileRUROSbydefinitionwasinthenaturallyoccurringoutdoorthermalenvironment).Ontheotherhand,theairportswereinfullHVACmodeacrossbothseasons.

Subjectivedata fromtheparticipants included thermal sensation,and inmostcasesinformationongender,clothingandmetabolicratewasalsoavailable.ThermalpreferencedatawerenotavailablefortheRUROSstudy;hencethisparameterwasnotincludedintheanalysis.AmajordifferencebetweenthestudiesindoorsandoutdoorswasthattheRUROSprojectemployeda5-pointthermalsensationscale,asopposedtotheASHRAE7-pointscale,whichhadbeenintroducedtoaidtheinterviewingprocessofindividualsafterapilotstudy

outdoors,inwhatsometimescouldberegardedasunfavourableconditions(Nikolopoulouetal.,2001).

This was an important obstacle for potential comparison; hence it was critical totranspose the5-pointRUROS thermal sensation scale intoa7-point scalewhich couldbedirectlycomparablewiththerestofthestudies.

Table1:Summarydataofthecomfortsurveysemployedforindoors,outdoorsandairportterminals

Nicol

Summer

NV(i)

Williams

Summer

NV

Williams

WinterNV

Williams

Winter

Mixed

Airports

HVAC

Summer&

Winter

RUROS

Summer&

Winter

GeneralLocation Oxford Liverpool,StHelensandChester

London&Manchester

Cambridge&Sheffield

Environment Indoors Indoors Indoors Outdoors

Casestudies3office

buildings8officebuildings

3airportterminals

4urbanlocations

Participants Sample 877 167 209 121 3087 1957

Gender 絃 x x x 絃 絃

Clothingins. 絃 絃 絃 絃 絃 絃

Thermalsensation

絃 Missing19 絃 絃 絃 絃(5-point)

ThermalPref. 絃 x x x 絃 x

Indoorconditions

Tair 絃(at0.6m) 絃(at0.6m) 絃(at0.6m) 絃(at0.6m) 絃(at1.7m) n/a

Tg 絃(at0.6m) x x x 絃(at1.7m) n/a

Tmr Missing2 絃 絃 絃 絃(at1.7m) n/a

Top Missing2 絃 絃 絃 絃(at1.7m) n/a

RH% 絃 絃 絃 絃 絃(at1.7m) n/a

Airmovement

missing215(at0.6m)

絃

(at0.6m)

絃

(at0.6m)

絃

(at0.6m)絃(at1.7m) n/a

Outdoorconditions

Tair 絃(ii) 絃

(ii) 絃

(ii) 絃

(ii) 絃(meteo) 絃

Tg(iii)

x x x x x 絃

Tmr x x x x x 絃

RH% 絃(ii) 絃

(ii) 絃

(ii) 絃

(ii) 絃(meteo) 絃

Windspeed x x x x x 絃

iNaturallyVentilatediiAvailabledataformin(at6am)andmax(at3pm)iiiTglobewasmeasuredwithagreyglobeoutdoors(asopposedtoablackglobeusedindoors)

3.1.!TransformationofRUROS5-pointtoASHRAE7-pointthermalsensationscale

Scale transformation has been investigated in other disciplines, particularly psychology,wheretheuseofLikertscales,i.e.scalesallowingindividualstoexpresstheirdis/agreementinaparticularstatement, iscommonlyfound. Previousstudiesthat lookedat5-and7-ptscaletransformationhaveproposedtwoinverseequationsfortheestimationofequivalencesbetweenthetwoscaleformats(ColmanandNorris,1997),anddatagatheredfroma5-pointformat canbe readily transferred to 7-point equivalencyusing a simple rescalingmethod(Dawes,2008)producingthesamemeanscore.Inthefieldofthermalcomfort,probitandsimpleregressionhavebeenshowntohavetwoimportantequivalences(Nicoletal.,2012).

The rescaling process of the thermal sensation scale involved a two-step approach.Firstly, theextremeandmiddle categoriesof the5-point scalewere corresponded to theextremesandmiddleofthe7-pointscalesothatpoints±2become±3and0remains0.Thesecond step was to rescale points ±1. A simplified transformation would be the

correspondence to points ±1.5 on the 7-point scale. This approach, however, assumeslinearitybetween thermal sensationand thecontrol variable (temperature)which - ifnotsatisfied,e.g.due tomeasurementerrororadaptation -mayresult inmisleading findings(Nicoletal.,2012).

Therefore, to rescale points ±1, the scale’s interval propertywas investigated as toidentifytherelevantthermaldistancesbetweencategories-2and-1,-1and0,0and+1,+1and+2,whichinthelinearapproachwouldbeequalto1.Forthispurpose,logisticregression(withcategory+2setasthereferencecategory)andprobitanalysiswereemployedusingairtemperature (Tair), mean radiant temperature (Tmr) and globe temperature (Tglobe) ascontrolvariables.To enable comparability, it was important to select indices available for all the studies.Correlation analysis of the RUROS data demonstrated that thermal sensation is bettercorrelated with Tglobe (r=0.68, p<0.01) than with Tair (r=0.63, p<0.01) and Tmr (r=0.62,p<0.01).Globetemperaturedatahoweverwereavailableforonlysomeoftheindoorstudiesreviewed(Table1).Asaresult,anoperativetemperatureindexwascalculatedfortheRUROSdata which could be tested as a control variable. The index was determined using theformula:

Top=[Tair*(10*Vair)0.5+Tmr]/[1+(10*Vair)0.5](Humphreysetal.,2015)

Where: Topistheoperativetemperature,Tairrepresentsairtemperature,

Tmristhemeanradianttemperature(°C)andVairthewindvelocity(m/s).

TheresultsoftherescalingprocessforthedifferentindicesarepresentedinFigure2wherethe intersectionbetweenthesigmoid linesandthe0.5 linedenotethe logit/probitcut-offpoints,summarisedinTable2.Thesepointscorrespondtoa50%percentprobabilityofavotechange to the next category. Subsequently, the transformed scores of the (5-point scale)categories -1 and +1 were calculated from -3*(Τ[0] -Τ[-1])/(Τ[0] -Τ[-2]) and 3*(Τ[+1] -Τ[0])/(Τ[+1] -Τ[-1]) respectively, where Τ[-1] is the temperature cut-off point for “cool”sensation, T[0] for “neither cool nor warm”, etc. Interestingly, the results revealed highconsistencybetweenthecut-offpoints(Table2)aswellasbetweenthetransformedpoints(Table3)determinedfromthedifferentcontrolvariables,andparticularlybetweenTopandTglobe,whichinstilledfurtherconfidencefortheselectionofTopasthethermalindexfortheevaluationofcomforttemperatures.

Table2:Cut-offpointsforTop,Tair,TglobeandTmrderivedfromlogisticregressionandprobitanalysis.

Logisticregression Probitanalysis

ThermalsensationTop50%

Tair50%

Tglobe50%

Tmr50%

Top50%

Tair50%

Tglobe50%

Tmr50%

Verycold 6.1 5.1 6.0 5.2 4.9 3.1 4.8 -0.1

Cool 13.6 11.3 13.3 16.0 14.0 11.8 13.8 15.9

Neithercoolnorwarm 21.1 18.6 20.8 28.2 21.1 18.6 20.8 28.6

Warm 33.0 29.8 32.7 59.0 34.8 32.0 34.4 54.3

Table3.Transformationof5-pointscale±1categoriesto7-pointscale.

Method 5-pointscale

7-pointscale

Top Tair Tglobe Tmr

Logisticregression

-1 -1.50 -1.62 -1.52 -1.59

+1 1.84 1.82 1.84 2.15

Forcedprobit

-1 -1.31 -1.32 -1.31 -1.33

+1 1.98 1.99 1.98 2.01

0.0

0.5

1.0

-15-10 -5 0 5 10 15 20 25 30 35 40 45

Pro

bab

ility

Temperature(°C)

Top-logistic

0.0

0.5

1.0

-15-10 -5 0 5 10 15 20 25 30 35 40 45

Pro

bab

ility

Temperature(°C)

Top-probit

0.0

0.5

1.0

-15-10 -5 0 5 10 15 20 25 30 35 40 45

Pro

bab

ility

Temperature(°C)

Tair-logistic

0.0

0.5

1.0

-15-10 -5 0 5 10 15 20 25 30 35 40 45

Pro

bab

ility

Temperature(°C)

Tair-probit

0.0

0.5

1.0

-15-10 -5 0 5 10 15 20 25 30 35 40 45

Pro

bab

ility

Temperature(°C)

Tglobe-logistic

0.0

0.5

1.0

-15-10 -5 0 5 10 15 20 25 30 35 40 45

Pro

bab

ility

Temperature(°C)

Tglobe-probit

Figure2:Logisticregression(leftcolumn)andprobitanalysis(rightcolumn)usingTop,Tair,TglobeandTmrascontrolvariables.

AsshowninFigure2,alltheanalysisforthetransformationofthescaleswasdonewithbothprobitand logistic regression. Thiswasduetothe fact that the formermethodhasbeentraditionallyassociatedwiththeinterpretationofdatafromfieldsurveysinthermalcomfortstudies,whilethelatter,isbeingincreasinglyusedfortheanalysisofthermalcomfortsurveys.Theeaseofuseoflogisticregressionwithmodernstatisticalpackages(asalsohighlightedbyNicoletal.,2012),themoreintuitiveinterpretationofitsresults,andthefactthatthetwomethodsprovidedverysimilarresultsledtotheadoptionoflogisticregressionresultsforthemeta-analysis.

4.! Meta-analysis

As theaimwas toevaluatewhether staff inairport terminalshavecomfort requirementsclosertostaff inoffices,whilepassengersusetheterminalasatransitionareawithmoresimilaritiestopeoplefoundoutdoors,itwasnecessarytoseparatetheairportstudyintwodistinctusergroups,passengersandstaff.ThesummaryTable4,showsthatalthoughtheratioofstafftopassengers intheairports isroughly1:5 inbothseasons,neverthelessthesampleislargeenoughtoallowstatisticalanalysisandcomparablewiththerestofthesurveypopulations.

ThedatawereanalysedbymeansoftheStatisticalPackageforSocialSciences(SPSS)and were initially subjected to quality checks to ensure high fidelity of the developeddatabasetotalling6100people.

Table4:Cleanedupdataonthesampleofthepopulationanalysedforthedifferentcontexts.

AirportsStaff,HVAC

AirportsPassengers,HVAC

Nicol,NV

Williams,NV

Williams,mixed

RUROS Total

Summer 236 1188 875 148 n/a 1264 3711Winter 229 1145 n/a 209 121 685 2389

Total 465 2333 875 357 121 1949 6100

Asummaryoftheoperativetemperaturesforthedifferentstudiesatthedifferentseasonsispresented inTable5. With theexceptionof theoutdoor temperatures forRUROS,whichdemonstrate a large range and standard deviation, as would be expected for externalconditions,therestofthemeanoperativetemperaturespresentafairlyuniformprofilewith

0.0

0.5

1.0

-15-10 -5 0 5 10 15 20 25 30 35 40 45

Pro

bab

ility

Temperature(°C)

Tmr-logistic

0.0

0.5

1.0

-15-10 -5 0 5 10 15 20 25 30 35 40 45

Pro

bab

ility

Temperature(°C)

Tmr-probit

awiderrangeofminimumandmaximumtemperaturesfornaturallyventilatedbuildingsinthesummer.

Table5:Summarydatafortheoperativetemperatureinthedifferentstudies

Study Season N Top_min Top_max Top_mean Std.Deviation

AirportsStaff summer 236 19.1 25.8 22.9 1.3

NicolNV summer 877 14.3 30.2 21.8 2

WilliamsNV summer 167 16.6 25.9 21.9 1.7

AirportsPassengers summer 1188 19.4 26.3 22.8 1.3

RUROS summer 1264 10.7 36.2 23.2 5.4

AirportsStaff winter 229 16.7 24.3 22.1 1.4

WilliamsNV winter 209 18.6 25.9 21.9 1.5

WilliamsMixed winter 121 18.7 25.9 23.4 1.5

AirportsPassengers winter 1145 16.2 25.6 21.9 1.6

RUROS winter 685 2.3 27.4 13.3 4.8

Following the transformation of the 5-point scale, analysis focused on understandingdifferencesinthermalsensationandidentifyingtheevidenceofpotentialadaptivebehaviour.

4.1.!Clothing

(a)

(b)Figure3:Clothinginsulationasafunctionofoutdoorairtemperatureforthedifferentstudiesatdifferentseasons,(a)forthestaff-indoorgroupand(b)forthetransitionandoutdoorgroup.

y=-0.023x+1.1173

R²=0.06112

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

8 9 10 11 12 13 14 15 16 17 18 19 20 21

Clo

thin

gin

sula

tion(cl

o)

Outdoorairtemperature (ォ)

NicolSummer

y=-0.0133x+0.8727

R²=0.227760

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

-5 -2 1 4 7 10 13 16 19 22 25 28

Clo

thin

gin

sula

tion(cl

o)

Outdoorairtemperature (ォ)

AirportsStaff(Summer&Winter)

y=-0.0322x+1.3974

R²=0.45449

00.20.40.60.8

11.21.41.61.8

22.22.42.6

0 3 6 9 12 15 18 21 24 27 30 33

Clo

thin

gin

sula

tion(cl

o)

Outdoorairtemperature (ォ)

RUROS(Summer&Winter)

y=-0.029x+1.1069

R²=0.36517

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

-6 -3 0 3 6 9 12 15 18 21 24 27

Clo

thin

gin

sula

tion(cl

o)

Outdoorairtemperature (ォ)

Airportspassengers(Summer&Winter)

Considering clothing as a potential adaptive mechanism (Humphreys, 1977, 1979;NikolopoulouandLykoudis,2016),clothinginsulationlevelswereevaluatedagainstoutdoorairtemperature.Fromtheoffices,onlytheNicolstudycouldbeused.Thedataonexternalair temperatureforWilliamsconsistedofa fixedvalueperday,providingtwoexternalairtemperatures for summer and another three forwinter,which did not provide sufficientvariationtoinformtheanalysis.

Theregressionanalysisofclothinginsulationasafunctionoroutdoorairtemperatureis presented in Figure 3. It is noticeable that passengerswear awider range of clothinginsulation, which is more comparable to clothing levels outdoors, with 37% and 46% ofclothing varying along with external temperature at the two seasons. For airport staff,however, the clothing range worn is narrower, indicative of the set uniform for indoorconditionsrequiredinairports,morecomparabletoanofficeenvironment.

4.2.!Neutraltemperature

Neutral temperature, i.e. the temperature yielding a sensation of neither cold nor hot(Humphreys, 1976), was determined by means of weighted linear regressions using half-degree(°C)incrementsofoperativetemperature(deDearetal.,1997).ThemeanTSscorewas calculated for each bin and regression models were fitted between mean TS andoperative temperature. Thermal neutrality was subsequently derived from solving theregressionequationsforTS = 0.Theregressionmodelswerealsousedfortheevaluationoftheoperativetemperaturerangesinwhich80%and90%ofpeoplewouldfindthethermalconditionsacceptable,inaccordancetothestatisticalassumptionsunderlyingthePMV/PPDheat-balancemodel(ISO7730,2005).Alltheparametersinthemodels,presentedinFigure4,achievedastatisticalsignificancelevelof99%orbetter.

y=0.3142x- 6.6884

R²=0.61192

-3

-2

-1

0

1

2

3

18 19 20 21 22 23 24 25 26 27

Mea

nther

mal

sen

sasi

on

Operativetemperature(虻C)

AirportsSummerStaff

y=0.3371x- 7.3079

R²=0.51594-3

-2

-1

0

1

2

3

16 17 18 19 20 21 22 23 24 25

Mea

nther

mal

sen

sasi

on

Operativetemperature(虻C)

AirportsWinterStaff

Figure4:Meanthermalsensationasafunctionofoperativetemperature(°C)forthedifferentstudies

y=0.1991x- 4.4049

R²=0.86273

-3

-2

-1

0

1

2

3

13 15 17 19 21 23 25 27 29 31

Mea

nther

mal

sen

sation

Operativetemperature(虻C)

NicolSummerNV

y=0.4071x- 8.1983

R²=0.61499

-3

-2

-1

0

1

2

3

18 19 20 21 22 23 24 25 26 27

Mea

nther

mal

sen

sasi

on

Operativetemperature(虻C)

WilliamsWinterMixed

y=0.4362x- 9.2847

R²=0.76644

-3

-2

-1

0

1

2

3

19 20 21 22 23 24 25 26 27

Mea

nther

mal

sen

sasi

on

Operativetemperature(虻C)

WilliamsSummerNV

y=0.5286x- 10.88

R²=0.78908

-3

-2

-1

0

1

2

3

18 19 20 21 22 23 24 25 26 27

Mea

nther

mal

sen

sasi

on

Operativetemperature(虻C)

WilliamsWinterNV

y=0.25x- 5.13

R²=0.85325

-3

-2

-1

0

1

2

3

18 19 20 21 22 23 24 25 26 27

Mea

nther

mal

sen

sasi

on

Operativetemperature(虻C)

AirportsSummerPassengers

y=0.2316x- 4.3547

R²=0.78843

-3

-2

-1

0

1

2

3

15 16 17 18 19 20 21 22 23 24 25 26 27

Mea

nther

mal

sen

sasi

on

Operativetemperature(虻C)

AirportsWinterPassengers

y=0.1384x- 2.213

R²=0.89583

-3

-2

-1

0

1

2

3

10 12 14 16 18 20 22 24 26 28 30 32 34 36 38

Mea

nther

mal

sen

sasi

on

Operativetemperature(虻C)

RUROSSummer

y=0.1753x- 3.1257

R²=0.78767-3

-2

-1

0

1

2

3

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30

Mea

nther

mal

sen

sasi

on

Operativetemperature(虻C)

RUROSWinter

Theanalysishighlightsanumberofissues.Examiningtheslopeoftheequationasameasureof thermal sensitivity, it becomes apparent that in both winter and summer, airportpassengersarelesssensitivethanthestaffandmoresimilartotheoutdoorsetting.Aunitincreaseofpassengers’TSwouldrequireatemperatureriseof4.0°Cinsummerand4.3°Cinwinter,particularlycomparablewith5.7°Cforoutdoorsinwinter.

Airportstaff,however,aremoresensitiveandclosertoofficestaff.Thetemperaturechangerequiredtoalterairportstaff’sTSby1unitisnearly3.0°Cforbothseasons,similarlytoofficebuildingswhereTSwouldnotbealteredwithtemperaturechangesbelow3.7°Cinsummerand2.2°Cinwinter.

Lookingatneutraltemperatures(Table6),itbecomesevidentthatTnforairportstaffisdirectlycomparabletostaffinofficesinbothseasons.Forpassengers,directcomparisonwiththeoutdoorsismoredifficultasairportsarefullyair-conditioned,andyetitisnoticeablethatinthesummerTnforpassengersislowerthanallofficeworkersandairportsstaff,whileinwinterpassenger’sTnisveryclosetoTnfortheoutdoorsetting(at18.8°Cand17.8°Crespectively). In addition, the winter results derived from the evaluation of the 80%acceptabilitytemperaturerangesdemonstrateaconsiderablesimilaritybetweenthecomfortzoneforpassengersandoutdoorsettingsandparticulartolerancetocolderconditions(Table6).Table6:Summarydatafortheneutraltemperatureinthedifferentstudies

Study Season Buildingtype N Slope R2

Tneutral

(°C)

80%

accept.

(°C)

90%

accept.

(°C)

AirportsStaff summer HVAC 236 0.31 0.61 21.3 18.6-24.0 19.7-22.9

Nicol summer NV 875 0.20 0.86 22.1 17.9-26.4 19.6-24.6

Williams summer NV 148 0.44 0.77 21.3 19.3-23.2 20.1-22.4

AirportsPassengers summer HVAC 1188 0.25 0.85 20.5 17.1-23.9 18.5-22.5

RUROS summer n/a 1264 0.14 0.90 16 9.8-22.1 12.4-19.6

AirportsStaff winter HVAC 229 0.34 0.52 21.7 19.2-24.2 20.2-23.2

Williams winter NV 209 0.53 0.79 20.6 19.0-22.2 19.6-21.5

Williams winter Mixed 121 0.41 0.62 20.1 18.1-22.2 18.9-21.4

AirportsPassengers winter HVAC 1145 0.23 0.79 18.8 15.1-22.5 16.6-21.0

RUROS winter n/a 685 0.18 0.79 17.8 13.0-22.7 15.0-20.7

5.! Conclusions

Astheresultshighlight,thereisconsiderabledifferenceintheadaptivecapacitybetweenthedifferentgroupsanalysed.Thecomforttemperaturesforallemployees,intheterminalsandoffices,areclosertothemeanoperativetemperature(Tables5-6),reflectingtheirlong-termacclimatisation to theworking thermalenvironment. On theotherhand,passengersandpeopleoutdoorsdemonstratewideradaptationcapacitywithabiggerdifferencebetweentheir mean operative and comfort temperature, while being less sensitive to thesedifferences, asdemonstratedby the lowgradientof the respectiveequations for thermalsensationandneutraltemperature.

Inthatrespect,thepapersucceededinprovingthehypothesisthatthethermalcomfortrequirementsofairportstaffarecloselycomparedtothoseofstaffworkinginoffices,asalsofoundbythesimilarneutraltemperaturesforthetwogroups.However,themajorityofthe

population inairport terminals ispassengers,who inhabit thespaceasa transitionspace,morecloselyrelatedtothecomfortrequirementsofpeopleusingoutdoorurbanspaces,astherespectiveneutraltemperatureshighlighted.Onceagain,thisbringstotheforefronttheimportant roleofadaptation,bothphysicalaswellasbehaviouralandpsychological,withexperiencesandexpectationsenablingthelattergroupstoachievewiderthermalcomfortzones.

In fact, beyond the broad categories of different physical environments, it is thepsychological adaptation that enablesmoving along the adaptive opportunity continuum,presented inFigure1,basedon thepotential foradaptivecapacityatapersonal level,asmanifestedwiththedifferentgroupsatairportterminals.Furtherworkindifferentclimaticcontextsandemployingadditionaldifferentdatabasescouldshedfurtherlightontheabove,eliminatinganyimplicitbiaswhichmaybeinherenttospecificdatasets.

Suchfindingshaveimportantimplicationsforenergyuseinbuildingsandparticularlythehighenergy-consumingsectorofairportterminals.Fromintroductionofsoftpoliciestoaddressflexibilityinclothingforstaffuniforms,tothedesignoflocalisedbuildingservicesforstaffratherthantreatinglargevolumesofairinterminals,itbecomesapparentthatthermalcomfort surveys continue to play an important role not only for research but also forunderstanding human behaviour and ultimately improvements to the design of the builtenvironment.

Finally,theworkhasshedsomelightonthetechniqueofpotentialtransformationofthermalcomfortscales.Thelast15yearshavewitnessedanincreasedamountofoutdoorthermalcomfortsurveys,manyofwhichhaveusedavarietyofthermalsensationscalesfromfive-point (Nikolopoulou et al., 2001; Nikolopoulou and Lykoudis, 2006; Aljawabra andNikolopoulou,2010;Nikolopoulouetal.,2011)tonine-point(Kántoretal.,2016).Thepaperidentified possibilities for eventual comparison of suchwork from different geographical,climaticand socio-cultural contexts thatwill inform thewiderdebateon thermal comfortfurther.

6.! ReferencesAljawabra,F.andNikolopoulou,M.,2010.Theinfluenceofhotaridclimateontheuseofoutdoorurbanspaces

andthermalcomfort:doculturalandsocialbackgroundsmatter?,IntelligentBuildingsInternational,Vol.2,No.3,pp.198-217.

ANSI/ASHRAEStandard55-2004.ThermalEnvironmentalConditionsforHumanOccupancy.Colman,A.M.andNorris,C.E.,1997.Comparingratingscalesofdifferentlengths:Equivalenceofscoresfrom5-

pointand7-pointscales,PsychologicalReports,Vol.80,pp.355-362.Dawes,J.,2008.Dodatacharacteristicschangeaccordingtothenumberofscalepointsused?Anexperiment

using5-point,7-pointand10-pointscales,InternationalJournalofMarketResearch,Vol.50,No.1,pp.61–104.

de Dear, R.J., Brager, G.S. and Cooper, D., 1997. Developing an Adaptive Model of Thermal Comfort and

Preference,FinalReportonASHRAERP-884.Fanger,P.O.,1970.ThermalComfort.,DanishTechnicalPress,Copenhagen.Humphreys,M.A.,1979.The influenceofseasonandambienttemperatureonhumanclothingbehaviour. In

(eds.)P.O.FangerandO.Valbjorn IndoorClimate:Proc.of the first InternationalClimateSymposium,Copenhagen,1978,DanishBuildingResearchInstitution,Copenhagen.

Humphreys,M.A.,1977.Clothingandtheoutdoormicroclimateinsummer,BuildingandEnvironment,Vol.12,pp.137-142.

Humphreys,M.A.,1976.Fieldstudiesofthermalcomfortcomparedandapplied,J.Inst.Heat.&Vent.Eng,44,pp.5-27.

Humphreys,M.,Nicol,F.andNikolopoulou,M.,2015.Environmentalcriteriafordesign,GuideA:Environmental

Design,CharteredInstituteofBuildingServicesEngineers,CIBSE,London.

ISO 7730, 2005. Ergonomics of the Thermal Environment - Analytical Determination and Interpretation of

ThermalComfortUsingCalculationofthePMVandPPDIndicesandLocalThermalComfortCriteria.Kántor,N.,Kovács,A.,Takács,Á.,2016.Seasonaldifferencesinthesubjectiveassessmentofoutdoorthermal

conditions and the impact of analysis techniques on the obtained results, International Journal ofBiometeoroly,Vol.60,No.11,pp1615–1635.

Kotopouleas, A. and Nikolopoulou, M., 2016. Thermal comfort conditions in airport terminals: Indoor ortransitionspaces?BuildingandEnvironment,Vol.99,pp.184-199.

Kotopouleas,A. andNikolopoulou,M., 2018. Evaluationof comfort conditions in airport terminal buildings,BuildingandEnvironment,Vol.130,pp.162-178.

Nicol,F.,Humphreys,M.andRoaf,S.,2012.AdaptiveThermalComfort:PrinciplesandPractice,Routledge.Nikolopoulou, M., 2004. Outdoor Thermal Comfort. In (eds.) K. Steemers & M.A. Steane,Architecture and

Variety:EnvironmentalPerspectives,SponPress,pp.101-119.Nikolopoulou, M., 1998. Thermal Comfort in Outdoor Urban Spaces, PhD thesis, University of Cambridge

(unpublished).Nikolopoulou,M., Baker,N. and Steemers, K., 2001. Thermal comfort in outdoor urban spaces: the human

parameter,SolarEnergy,Vol.70,No.3,pp.227-235.Nikolopoulou,M.,Kleissl,J.,Linden,P.F.andLykoudis,S.,2011.Pedestrians'perceptionofenvironmentalstimuli

throughfieldsurveys:Focusonparticulatepollution.ScienceofTheTotalEnvironment,Vol.409,No.13,pp.2493-2502.

Nikolopoulou,M.andLykoudis,S.,2006.Thermalcomfort inoutdoorurbanspaces:AnalysisacrossdifferentEuropeancountries.BuildingandEnvironment,Vol.41,pp.1455-1470.

Nikolopoulou,M.andSteemersK.,2003.Thermalcomfortandpsychologicaladaptationasaguidefordesigningurbanspaces,EnergyandBuildings,Vol.35,No.1,95-101.

Spagnolo,S.anddeDear,R.,2003.Afieldstudyofthermalcomfortinoutdoorandsemi-outdoorenvironmentsinsubtropicalSydneyAustralia,BuildingandEnvironment,Vol.38,pp.721-738.

Acknowledgements Theavailabilityofthedatasetshasbeenpivotalforenablingsuchanalysis.Theseinclude:[1]IndoorOffices:ASHRAERP-884AdaptiveModelProject-DataDownloaderhttp://sydney.edu.au/architecture/staff/homepage/richard_de_dear/ashrae_rp-884.shtml[2]Outdoorurbanspaces:EUFP5projectRUROS(RediscoveringtheUrbanrealmandOpenSpaces)databasehttp://alpha.cres.gr/ruros,EVK4-CT2001-00032.[3]Airportterminals:EPSRCproject“Integrationofactiveandpassiveindoorthermalenvironmentcontrolsystemstominimizethecarbonfootprintofairportterminalbuildings”,EP/H004181/1.