kent academic repositorykar.kent.ac.uk/66790/1/windsor paper-kent_final.pdf · from indoors to...
TRANSCRIPT
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:
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,[email protected],[email protected]
2 IndependentResearcher,Athens,Greece,[email protected]
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.