In-situ and real time measurements of thermal comfort and its determinants in thirty residential dwellings in the Netherlands

(1)

Ioannou, Taso; Itard, Laure

DOI

10.1016/j.enbuild.2017.01.050

Publication date

2017

Document Version

Final published version

Published in

Energy and Buildings

Citation (APA)

Ioannou, T., & Itard, L. (2017). In-situ and real time measurements of thermal comfort and its determinants

in thirty residential dwellings in the Netherlands. Energy and Buildings, 139, 487-505.

https://doi.org/10.1016/j.enbuild.2017.01.050

Important note

To cite this publication, please use the final published version (if applicable).

Please check the document version above.

Copyright

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons. Takedown policy

Please contact us and provide details if you believe this document breaches copyrights. We will remove access to the work immediately and investigate your claim.

This work is downloaded from Delft University of Technology.

(2)

Contents lists available atScienceDirect

Energy and Buildings

j o u r n a l h o m e p a 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

In-situ and real time measurements of thermal comfort and its

determinants in thirty residential dwellings in the Netherlands

Anastasios Ioannou

∗

, Laure Itard

OTB—Research for the Built Environment, Delft University of Technology, Julianalaan 134, 2628BL Delft, Netherlands

a r t i c l e

i n f o

Article history: Received 7 October 2016 Received in revised form 21 December 2016 Accepted 14 January 2017 Available online 20 January 2017

Keywords: In-situ measurement PMV Thermal comfort Clothing Metabolic activity Thermal sensation Occupancy behaviour Energy consumption Residential dwellings Wireless monitoring.

a b s t r a c t

Reducing energy consumption in the residential sector is an imperative EU goal until 2020. An important boundary condition in buildings is that energy savings shouldn’t be achieved at the expense of thermal comfort. There is, however, little known about comfort perception in residential buildings and its relation to the PMV theory. In this research an in-situ method for real time measurements of the quantitative and qualitative parameters that affect thermal comfort as well as the reported thermal comfort perception was developed and applied in 30 residential dwellings in the Netherlands. Quantitative data (air temperature, relative humidity, presence) have been wirelessly gathered with 5 min interval for 6 months. The thermal sensation was gathered wirelessly as well, using a battery powered comfort dial. Other qualitative data (metabolic activity, clothing, actions related to thermal comfort) were collected twice a day using a diary. The data analysis showed that while the neutral temperatures are well predicted by the PMV method, the cold and warm sensations are not. It seems that people reported (on a statistically signiﬁcant way) comfortable sensation while the PMV method doesn’t predict it, indicating a certain level of psychological adaptation to expectations. Additionally it was found that, although clothing and metabolic activities were similar among tenants of houses with different thermal quality, the neutral temperature was different: in houses with a good energy rating, the neutral temperature was higher than in houses with a poor rating. © 2017 Elsevier B.V. All rights reserved.

1. Introduction

The built environment is responsible for about 40% of total energy use in Europe. Of this 40%, 63% is related to residential energy consumption[1]. European and national regulations like the Energy Performance of Buildings Directive EPBD and speciﬁc parts of national building codes aim to reducing the energy consump-tion of buildings in order to achieve the goals set for emissions and resource consumption by 2020.

The prediction and assessment of the energy consumption of residential dwellings is an important means to this end. Building performance simulation is a widely accepted method for this pur-pose. Buildings are highly complex systems in their own right. Both new buildings and renovated ones that are equipped with new heating and ventilation systems have high performance require-ments that are closely related to EU sustainability goals for 2020. Increasing the reliability of building performance simulations can

∗ Corresponding author.

E-mail addresses:a.ioannou@tudelft.nl(A. Ioannou),L.C.M.Itard@tudelft.nl (L. Itard).

make an important contribution to reduction of the energy con-sumption of residential building stock.

The need for increased reliability of building simulations is also closely related to the discrepancy between actual and predicted energy use in the residential building sector. Researchers in the Netherlands and elsewhere have found a substantial gap between actual and predicted energy use in residential dwellings, with the worst dwellings (those with an energy rating of F or G) consum-ing signiﬁcantly less energy than expected while dwellconsum-ings with a higher energy rating consume more[2]. One reason for this dis-crepancy could be limited information on the building’s thermal envelope and installations (more obvious in older dwellings where no records are available on the materials used). Another important reason is related to a misunderstanding or underestimation of the role of the occupant’s behaviour[3,4,5]. Simulation software in its current form has very limited capabilities for taking the energy-related behaviour of the occupant into account. There is a clear need to take this behaviour into account during the design phase of new residential buildings or the renovation phase of older ones

[3,4,6,7].

An important requirement both for new dwellings and for the refurbishment of older ones is that thermal comfort should be maintained or improved. Many commercially available simulation

(3)

thermal responses ofoccupants of residential and ofﬁce build-ingsrecordedinvariouscountriesdifferfromthepredictedvalues

[10,11,12,13,14,15]thoughHumphreysshowed,inaworld-wide

datasetof16,762caseswithvarioussettings,thattheperceived thermalcomfortagreedquitewellwiththemodel’spredictions

[15].Thismeansthatitisverydifficulttodrawgeneralconclusions forspecificlocalsettings,despitethemodel’sstrongphysicalbasis. Residential dwellings, unlike office buildings, include zones withvariablethermalcomfortrequirements,arecharacterisedby lesspredictableactivitiesandprovidemorewaysforthetenant toadapttohisthermalenvironmentinordertoreachthedesired comfortlevel[16].Theseconditionsintheseresidentialsettings differgreatlyfromthoseapplyingintheclimatechamberFanger usedtodevelopthePMVthermalcomfortindex.

Temperaturelevelsand proﬁlesindwellingsareexpectedto haveanimportanteffectontheenergyconsumptionforheating andtenants’thermalcomfort[17,18,19].Furthermore,the opera-tivetemperatureisacriticalcomponentofthePMVcomfortindex. Variousstudieshavederived indoortemperatureproﬁlesfor theresidentialbuiltenvironmentbuttheydifferinthemethods used,thelengthofthemonitoringperiodandtheseasonwhen measurementsweremade.In many cases,temperaturesensors withdatarecordingintervalsof15,30,45or60minwereused

[20–30].Thedurationofthemeasurementcampaignvariedfrom

1to4weeks[25,31]insomestudies,whileinothersitcoveredthe wholeheatingperiod(DecembertoAprilinonenorthernEuropean country(Belgium) [31]; astudy inonesouthern Mediterranean country(Greece)[24]alsocoveredthewholeheatingperiod−one thatismuch shorter thannorthernEuropeancountrieslikethe NetherlandsorBelgium.Inonestudythetenantsweregiventhe temperaturesensortogetherwiththeoperatingmanualandwere invitedtoinstallitthemselves[26],whichcouldlowertheaccuracy ofthemeasureddata.Inallthesestudiesthedatawerecollected locallyindataloggersandhadtoberetrievedmanually.Other stud-iesusedquestionnairesordiariesforrecordingthetemperatures wherethetenantshadtoﬁllintherequiredinformation[32,33]. Thisprobablyledtolargeuncertainties,asnomeasurementswere performed.

Theaimofthepresentpaperistoprovideinformationonakitfor in-situreal-timemeasurementofthequantitativeandqualitative parametersthataffectthermalcomfortonthereportedtenant’s thermalsensationandﬁnallytopresenttheresultinganalysisof energy-relatedoccupantbehaviour(inparticulartheparameters thataffectthePMVcomfortindex).Thisisimportantbecause ther-malcomfortmayaffectlargelyoccupantbehaviour,whichrelates toenergyconsumptionandwhichinturnisanimportantfactorfor thediscrepancybetweenactualandtheoreticalenergy consump-tionintheresidentialdwellings.

The results presented here are taken from the Ecommon (EnergyandComfortMonitoring)campaignwhichtookplacein theNetherlands as part of the Monicair [34], SusLab [35] and Installaties 2020[36] projects. Thirty-two residentialdwellings (classiﬁedbyenergyratingandtypesofheatingandventilation system)weremonitoredfora6-monthperiod,fromOctober2014 toApril2015,whichistheheatingseasonfornorthWesternEurope. Quantitativedata(airtemperature,relativehumidity,CO2leveland

atmosphericconditions(temperatureT,relativehumidityRHand CO2level),whichcouldimprovethereliabilityofthePMV calcu-lations(seeSection2.3.1).Alldata(quantitativeandqualitative) wereavailableforinspectionandanalysisinrealtimethroughout thewholecampaignviaaremotedesktopapplication.

Thenextchapterdescribestheresearchquestions,thedesignof thisstudy,thewaythecampaignwassetup,thedataacquisition equipmentandthedatamanagementsystem.Theresultsfollow inchapter3whichﬁrstpresentstheneutraloperative tempera-tures,perroomtype,derivedfromthePMVcalculationsandthe recordedthermalsensationofthetenants.Furtheron,the relation-shipbetweenthereportedthermalsensationandthecalculated PMVisexploredinordertofurthervalidatetheabilityofthePMV indextopredictthetenant’srealthermalsensation.Thenexttwo sections(3.4and3.5)describetheclothingandmetabolicactivity ofthetenantsduringthemeasurementcampaignagainstthe oper-ativetemperatureandthermalsensation.Further,thecloandmet valuesthatcorrespondtotheneutralthermalsensationofthe ten-antswerecalculatedandtheeffectoftheinaccuracyofthesevalues wasresearched.Finally,asectionwithdiscussion,conclusionsand recommendationsconcludethepresentstudy.

2. Studydesign

Comforthasseldombeenresearchedonsiteinactualconditions, andevenmorerarelyhasbeenmeasuredinotherwaysthanusing surveys.Themainresearchquestionsinthispaperaimto deter-minewhetheritispossibletomakesuchmeasurementsandhow theresultsofthesemeasurementscompetewithalreadyexisting insightsfromPMVtheory.

2.1. Researchquestions Thegoalsofthisstudyare:

1)Toperformin-situreal-timemeasurementofquantitativeand qualitativedataoncomfortandoccupantbehaviourandtheir underlyingparametersinaneasy,unobtrusiveway,ina resi-dentialenvironment.

2)Todeterminethetenants’temperatureperceptioninrelationto theenergyratingandtheventilationandheatingsystemsused inthedwellings.

3)Todeterminethetypeofclothingwornbythetenantsandtheir activitylevelsinrelationtothethermalsensationofthe occu-pants.

4)Todeterminetheneutraltemperaturelevelscalculatedbythe PMVmethodandtocomparethemtotheneutraltemperatures derivedfromthemeasurementsthermalsensation.

5)TodeterminetowhatextentthePMVcomfortindexagreeswith thethermalsensationreportedbythetenants.

6)Todetermineifthereisarelationshipbetweenthetypeof cloth-ingandmetabolicactivitywiththermalsensationandtheindoor operativetemperature.

(4)

2.2. Ecommoncampaignset-up

Theoriginaldesignofthestudywastohavestratiﬁedrandom sampling.Thedwellingsweregroupedaccordingtothevarious heatingsystems,totheirenergylabelandtheirventilationsystem. Howeverforpracticalreasonswedeviatedfromthat.Thisisalso whywedonotclaim universalityinourresultsbutweinstead showthemethodsthatcanbeappliedinordertomeasureinsitu thequalitativeandquantitativeparametersofthePMV.

ThesampleusedintheEcommonmonitoringcampaignwas restrictedtosocial housing,inordertomatchthis studywitha previousoneinwhichmostdatawerecollectedforsocialhousing

[37].SocialhousingintheNetherlandsrepresentsapproximately one-thirdofthetotalresidentialhousingstockandisquite rep-resentativeoftheresidentialhousingstockasawhole[2,38,39]. Furthermore,housingassociationshave theenergyrating ofall theirhousingstock determined, which is not thecase withall individualowners.ThesamplehadtobedividedintoA-ratedand F-rateddwellings,inordertoaddressissuesofcurrentenergy rat-ingmodels.Infact,A-ratedandB-rateddwellingswereselectedat oneextremeandF-rateddwellingsattheother.F-rateddwellings wereselectedinpreferencetoG-ratedones,sincepreviousstudies

[2,37]hadshownthattherearefewdwellingsintheNetherlands

withaGenergyrating.

Themethodusedtocalculatetheenergy ratingis described in DutchbuildingcodeISSO82.3[40].The energysurvey used asabasisfortheenergyperformancecertiﬁcate(EPC)rateseach dwellingonascalefrom‘A++’(themostefﬁcient)to‘G’.The cat-egoriesaredeterminedwithreferencetotheenergyindex,which iscalculatedonthebasisofthetotalprimaryenergydemand(Q

total);thisrepresentstheprimaryenergyconsumedforheating,hot

water,pumps/ventilatorsandlighting,aftersubtractingtheenergy gainsfromPVcellsand/orcogeneration.

Wesentalettertomorethan2,000addresses,invitingthemto participateinthestudyandtheresponseratewas8.6%.Surveysthat areintendedforexternalaudiencesusuallyhaveareturnrateof

5–10%.Consideringthelonglengthofthemeasurementcampaign, theamountofequipmentthathadtobeplacedineachdwelling, thefrequentintrusionofTUDelftpersonnelintothetenants’ pri-vacy(installingtheequipment,handingover andretrieving the comfortdial,callingtenantstorestartthedatagatheringminipc, retrievingtheequipment)andﬁnallythefactthatthedata gath-eredcouldcompromisethetenant’sprivacyandpotentiallytheir security(tenantswerenotiﬁedforalltheseissuesintheinitialletter theyreceived),thereturnrateof8.6%isconsideredverysuccessful. Furthermorecompensationwasofferedtotheparticipantsforthe electricitycostsoftheequipmentfortheperiodofthesixmonths, twogiftcardsof20euroseachwasofferedtothemandthefeedback wereceivedforthispresentwasverypositive.

Acarefulselectionhadtobemadefromamongthehouseholds willingtoparticipateinordertomaximisetheamountofdatathat couldbecollected.WeusedtheSHAEREdatabasedevelopedby Aedes[41],thefederationofDutchhousingassociations,toselect respondentsonthebasisoftheirenergyratingandheatingsystem. Atotalof58dwellingswereselected.Finally,duetolimitations inthemonitoringequipmentused,32dwellingsweremonitored overa6-monthperiod,fromOctober2014toApril2015.Thefinal samplemaybeseeninTable1.TheA-ratedandB-rateddwellings weredividedintothosewithanelectricalheatpumpcoupledwith lowhydronicfloorheatingandthosewithefficientcondensinggas boilers.TheF-rateddwellingsallhadtheiroldinefficientboilers replacedbynewcondensinggasboilers,apartfromthreethatwere stillequippedwitholdgasstovesconnectedtotheradiatorsinthe variousroomstoprovideacentralheatingsystem.

Thedwellingswerealsoclassiﬁedonthebasisoftheir venti-lationsystems.Eighthadbalancedventilation,10hadcompletely naturalventilation(supplyandexhaust)and14hadnaturalair sup-plyandmechanicalexhaust(usuallyinwetroomsandkitchens). Dwellings9and30havebeenexcludedfromtheanalysisdueto unavailability of data.Technical reasonsrelated tothewireless transmissionofthetemperature,humidityand CO2,resultedin

Table1

DwellingsparticipatingintheEcommoncampaign.

No. Energyrating Heatingsystem Ventilationsystem No.ofrooms No.ofoccupants Averageage

W001 F Condensinggasboiler NaturalsupplyMech.Exhaust 6 1 67

W003 A Heatpump BalancedVent. 4 2 73

W005 A Condensinggasboiler BalancedVent. 4 1 92

W006 A Condensinggasboiler BalancedVent. 3 2 77

W010 A Condensinggasboiler NaturalsupplyMech.Exhaust 7 2 29

W011 A Condensinggasboiler NaturalsupplyMech.Exhaust 7 2 69

W012 F Condensinggasboiler NaturalVent. 5 4 40.5

W013 F Condensinggasboiler NaturalVent. 5 3 53

W014 F Gasstove NaturalVent. 5 1 83

W015 B Condensinggasboiler NaturalsupplyMech.Exhaust 3 2 25

(5)

Fig.1.T,CO2,RHbox(left)andmovementsensor(right)asusedduringtheEcommonmeasurementcampaign.

Table2

Types,modelsandaccuracyofsensorsusedduringtheEcommonmeasurement campaign.

Sensortype Model Accuracy

CO2 GETelaire 400–1250ppm:3%of

reading

1250−2000ppm:5%of reading

RelativeHumidity HoneywellHiH5031 +/−3%

Temperature KTThermistor 1%per◦C

Movement HoneywellIR8M 11×12m(rangeat2.3m mountingheight)

completelossofdataforthesetwodwellings.Detailsofthe venti-lationsystemsofthevariousdwellingsarealsogiveninTable1.

2.3. Dataacquisitionandequipment

2.3.1. Honeywellequipmentusedtocollectindoorclimatedata Thesystemusedtocollecttemperature(T),relativehumidity (RH),CO2levelandpresencedatawasacustom-builtcombination

ofsensorsdevelopedbyHoneywell.CO2datawerenotrequired

forthescopeofthepresentpaper,andarethereforenotreported. Thetemperature,humidityandCO2sensorswereallmountedina

singleboxthatwasinstalledinuptofourhabitablerooms(living room,bedrooms,studyandkitchen) ineach houseparticipating inthemeasuringcampaign.Thetype,modelandaccuracyofthe sensorsareshowninTable2.TheT,CO2andRHsensorswerenot

batterypoweredandthereforehadtobepluggedintoawallsocket. ThePIRmovementsensor,ontheotherhand,wasbatterypowered.

Fig.1givesanimpressionofthearrangementofthesensors. Themeasuringfrequencyofallsensorswas5min.Thevalue recordedforeach5-minintervalwastheaverageofthereadings duringthatinterval.Temperaturesweremeasuredin◦C,relative humidityin%andCO2levelsinppm(partspermillion).The

tem-perature sensor is fully compliant with theISO 7726standard fortypeC,measurementscarriedoutinmoderateenvironments approachingcomfortconditions(comfortstandard)speciﬁcations andmethods.Thehumiditydataweredisplayedasrelative humid-ity(%)whichwasderivedbythevoltageoutputofthesecapacitive sensorsandintermsofaccuracycompliesfullywiththeISO7726

[56].

ThePIRsensordatawereinbinaryform(0and1),0meansthat nomovementwasdetectedduringthe5-minintervalinquestion while1 meansthat movementwasdetectedatleast once dur-ingtheinterval.ThePIRsensorhad11m×12mmdetectionrange whichwasenoughforalltheroomstheywereinstalledin.Theyhad selectablepetimmunity(0.18–36kg)apatentedlookdownmirror inordertodetectmovementexactlybelowthesensor,frontand reartampersandoperativetemperaturerangebetween₋₁₀◦Cand 55◦C.Thebatterylifewas4.5yearswhichwasexceedingbyfarthe timeframeofthisprojectandwasensuringthatthedatawould

Fig.2. ComfortDialusedtocaptureperceivedcomfortlevelsoftenantsduringthe Ecommonmeasurementcampaign.

besafelystoredincaseofwirelesstransmissionproblems.Finally theywerecompliantwiththeNENstandardforalarmsystems[55]. 2.3.2. Qualitativedata:comfortdialandlogbook

TheEcommonmeasurementcampaigncollectedqualitativeas wellasquantitativedata.Dataonperceivedcomfortlevelswere collectedwiththeaidofadevicedevelopedbyDelftUniversityof Technology’sDepartmentofIndustrialDesignundertheumbrella oftheEuropeanInterregprojectSustainableLaboratoriesNorth WestEurope(SusLab)[35].Thiswirelessdevice,called“comfort

dial”(Fig.2),allowedthetenantstodigitallyrecordtheirperceived thermalcomfortlevelatanytimeofthedayona7-pointscale,from −3(cold)via0(neutral)to+3(hot).

Thecomfort dial isportable andrelatively smallinsize and therefore tenants could carry it with them anywhere in the dwelling.Thatiswhythedataofthecomfortdialhadtobe cou-pledtothePIRsensordatainordertodeterminethelocationofthe tenantthatparticularmoment.

Tenantsalsoreceiveda paperlogbook,showninFig.3.This logbook,likethecomfortdial,wasdevelopedbyDelftUniversity ofTechnology’sDepartmentofIndustrial Design.It wasinitially intendedtobeinonlineformatsothatpeoplecouldlogontotheir computer,smart-phoneortabletandﬁllinvariousqualitativedata suchas:

• Perceivedcomfortlevelontheabove-mentioned7-pointscale. • Theroomtheyareoccupyingwhen ﬁllinginthelog(kitchen,

livingroom,bedroometc.)

• Clothingcombinationworn:achoiceofsixcombinationsfrom verylighttoverywarmclothingisavailable;seeFig.3andTable4. • Actionstakenduringthepasthalfhourrelatingtocomfortand energyconsumption,suchasopeningorclosingthewindows, drinkingacoldorhotdrink,takingclothesofforputtingthem on,raisingorloweringthethermostatsettingandhavingahot orcoldshower.

• Activity level:lying/sleeping,relaxed sitting, doing light desk work,walking, jogging, running.These activities canthen be relatedtothemetabolicrate.

(6)

Fig.3.Paperlogbookforentryofqualitativedata.

However,weﬁnallyusedapaperversionofthelogbookdueto acombinationofﬁnanciallimitations(notenoughtabletsavailable toprovidealloccupantsofthe32dwellingswithone)andthefact thatmanyparticipantswereelderlyandnotwellacquaintedwith digitaltechnology.

Theoccupantsofthehousesweregiventhecomfortdialand comfortlogbookfora2-weekperiodinMarchandearlyApril2015. Thelogbookwasgiventothemin 45copies,3perdayforthe periodofthetwoweeksthatthetenantshadtousethecomfort dial.Theyhadbeeninstructedtouseitatleast3timesperday(it wasequippedwithatimeline,seeFig.3)togetherwiththecomfort dial.Thecomfortdialontheotherhandcouldbeusedasoftenas theywantedthroughoutthewholeday.

Thedatafromthecomfortdialwerewirelesslyandinrealtime recordedtoourdatabasewhilethedatafromthecomfortlogbook aswellastheequipment(comfortdial)wereretrievedintheend ofthe2weeksperiod.Inthatwaywemanagedtoobtainthermal sensationdata(comfortdial),qualitativedatarelatedtothePMV (clothingandmetabolicactivity),andquantitativedatarelatedto thePMV(temperatureandhumidity)alluniversallytimestamped. ThisenabledustomakecalculationsonthePMVwithprecisionof 5minwhichwastheintervalofthesensorquantitativedata.

Themainrespondent(onlyonepersonperhouseholdwasasked tousethecomfortdialandlogbook)wasaskedtouseitasoftenas heorshewanted,butatleastthreetimesaday(preferablyinthe morning,middayandevening).Theyalsohadtoﬁllinthepaper log,atleastwhentheywereusingthecomfortdial.

Furthermore,tenantshadtoﬁllinaquestionnaireduringthe installationofthemonitoringequipment,andalldwellings par-ticipatinginthestudywereinspectedatthesametime.Thesetwo measuresprovidedextradatainhouseholdcharacteristics,heating andventilationpatternsandperceivedcomfortlevels.

2.3.3. Datastorageandmanagement

ThedatacollectedbytheHoneywellsensorsweremanagedby softwaredevelopedbyHoneywell.Thissoftwaremadeitpossible toselectmeasurementfrequencyof1,5,10oranyothernumber ofminutesatanymoment.Ameasurementfrequencyof5minwas chosenforthisproject.

Allthedatawerewirelesslytransmittedfromthesensorsto alocallyinstalledmini-PConwhichtheHoneywellsoftwarewas installed.Thedatawereregularlycopiedfromthismini-PCtoour SQLdatabaseatDelftUniversityofTechnology.Thisset-upallowed

thedatatobestoredbothlocally,ontheharddriveofthemini-PC, andcentrallyinthedatabaseatDelft.

AnotherpointworthmentioningisthateachHoneywell sen-sorbox(containingthetemperature,relative humidityandCO2

sensors)alsoactedasawirelesstransmitterfortheadjacent sen-sorbox,sothatonemini-PCcouldcollectdatafromneighbouring dwellings.Thisreduced overallequipmentcostsfortheproject. DatafromthecomfortdialweretransmittedtothedatabaseatDelft UniversityofTechnologyviaaconnectportandthelocalinternet connectionora3Gnetwork,ifavailable.

2.3.4. Occupantsurveyandinspectionlist

Occupantswereaskedtoﬁllinaquestionnaireduring installa-tionofthesensorsintheirhome.Thequestionsaskedfellintothree categories:1)generalinformationontheparticipatinghouseholds, suchashouseholdcomposition,income,age,educationlevel;2)the occupants’heating,showeringandventilationhabits;and3) over-allperceptionofthecomfortofthedwellings.Thequestionnaire wastakenfromanexistingtemplatethathasbeenusedinpast projects,withdifferentscopes,priortoEcommon[57].

Furthermore,eachdwelling wasinspected duringthe instal-lationofthemonitoringequipment.Theinspectioncovered the followingitemsthatwererelevanttothepresentstudy:thetype ofspaceheatingsystemused,thetypeofglazing,thetypesof venti-lationpresentinthedwelling(extractionpointinthekitchen,other mechanicalventilationusuallypresentinthekitchenorbathroom andbalancedventilation)andinformationonthethermostat:type ofthermostat,settingsandcontrolprogramme.

3. Results

3.1. Perceiveddwellingtemperatureinrelationtotheenergy ratingandventilationsystem

Thissectionpresentstheresultsofthisstudystartingwiththe tenant’soverallperceptionofthedwellingtemperature.The fol-lowingpart(3.2.3)presentsthecalculationoftheneutraloperative temperature,perroomtypeandenergyrating,accordingtothe cal-culatedPMVandthethermalsensationrecordedbythetenants.In thetwosectionsthatfollow(3.4and3.5)thecloandmetvalues aredisplayedversustherecordedthermalsensationofthetenants andtheoperativetemperature,forthelivingroom,andastatistical analysisfollowsinordertodeterminetheextentofpossiblebiasin

(7)

Fig.4.Temperatureperceptioninthewinterperenergyrating.

thecalculationsfrompotentialmistakesinthegatheringoftheclo andmetdata.

Fig.4showstheanswerstothequestion”Howdoyoufeelabout thetemperatureofthedwellingduringthewinter?”asafunction oftheenergyratingofthedwellingandthetypeof ventilation systemused.Itwillbeseenthattheproportionofoccupantswho regardthedwellingasbeingtoocoldincreasesaswemovefrom energy-efﬁcientclassAdwellingstoclassFdwellings,whichhave apoorenergyperformance.Thisﬁndingisinagreementwiththe resultsreportedbyMajcenetal.[38],andisprobablyrelatedtothe insulationlevelandair-tightnessofthedwellings.

Thetenantsofdwellingswithbalancedventilationhadthe high-estpercentage (85.7%) ofresponsesindictatingthattheindoor temperatureduringthewinterwasallright.Itshouldbenoted thatallthesedwellingshadenergyratingAorB.Inthatsense,these resultscouldbeexpectedandrelatemoretotheenergyratingthan totheventilationsystem.

AsmaybeseenfromTable1,somedwellingswith

mechani-calexhaustventilationhadenergyratingA/B,whileotherswere F-rated.Fig.4showsthattheproportionof”toocold”responses increasesfromA/B-rateddwellingstoF-ratedones.Occupantsof dwellingswithcompletelynaturalventilationwerelesslikelyto ﬁndtheindoortemperatureacceptable(55.6%).Alldwellingswith naturalventilationhadenergyratingF.Itisnoteworthythatthis groupincludedthreedwellingswithanoldgasstove.The occu-pantsofallthreestatedthattheyfoundtheindoortemperatureto beacceptable.

Itmightbeexpectedthattemperatureperceptionduringthe winterismorecloselyrelatedtotheenergyratingthantothetype ofventilation.Thiswasnothoweverfoundtobethecaseinall dwellingswithnaturalventilationandmechanicalexhaust.Some occupantsofenergy-efficientdwellingsinthiscategorystatedthat theyfelt too cold in the winter, while someoccupants of less energy-efficientdwellingsweresatisfiedwiththeindoor temper-ature.Furtherinvestigationoftheactualenergyconsumptionin thesedwellingsisrequiredtodeterminewhethertheseresponses arerelatedtoexcessiveenergyuseindwellingswithlowenergy efficiencyorverylowconsumptioninthemoreenergy-efficient dwellings.

3.2. NeutraltemperaturesinrelationtoPMVandreported thermalsensation

Fanger’smethod[14,42]forcalculationofthepredictedmean vote(PMV)isusedworldwidetoestimatethethermalcomfort lev-elsthancanbeachievedundervarioushydro-thermalconditions.

Thismethodusesthefollowingparameters:airtemperature(Tair),

meanradianttemperature(Tmrt),airvelocity(v),relativehumidity

(RH)andtwoparametersrelatedtothethermalresistanceof occu-pants’clothing[clo]andtheirmetabolicactivity[met].Duringthe presentstudy,dataformostoftheabove-mentionedparameters werecollectedwiththeaidofthesensors,thecomfortdialandthe logbook.Theparametersforwhichnodirectdatahadbeen gath-eredwerethemeanradianttemperatureTmrtandtheairspeed;

thelatterinparticularisaverydifﬁcultparametertorecordsince ithasaverystrongtopicaleffectanditsvaluemayvary signiﬁ-cantlyfromplacetoplaceinagivenroom.EnergyPlussimulations asdescribedbelowwereperformedinordertoestimateTmrt,

sen-sitivityanalysisforTmrtandairvelocityhasbeenincludedinall

furtheranalysesinthispaper.

3.2.1. Estimationofmeanradianttemperature(Tmrt),indoorair

speed,clovaluesandmetabolicactivityrates

Areferencedwellingwitha surfaceareaof75m2 _divided_in

twozones (livingroomandbedroom) wassimulatedusingthe weatherdataforTheHague,theNetherlands,forthewholemonth ofMarch 2015,whichwasthe monthwhen tenantswere pro-videdwithcomfortdialsinordertorecordtheirthermalsensations, clothingvalues,actionsaimedatmodifyingthermalsensationand metabolicactivity.The sizeand characteristicsof thereference dwellingweresimilartothetypesofdwellingsthatwerefound inthesampleoftheEcommoncampaign.Thedwellingwas simu-latedinEnergyPlusin3differentways:asanA-rateddwellingwith acondensinggasboilerfortheheatgenerationandradiatorsfor heatdistributionintherooms,asanArateddwellingwitha water-to-waterheatpump,a groundheatexchangerandgroundfloor heating,andfinallyasanF-rateddwellingwithcondensingboiler andradiators.Thesethreeconfigurationscoverallthedwellings usedintheEcommonmeasurementcampaign.

Occupancyschedules,commonlyavailableinsimulation soft-warelibraries and adjusted toDutchhabits, wereused forthe simulationsofthelivingroom(presenceearlyinthemorning,and from5pmtillmidnight)andbedroom(presence/sleepingduring thenighthours).Thenumberofpeopleoccupyingthereference dwellingwassetto2andthethermostatsettingswere18◦Cduring daytimeoccupancyand12◦Catnight.Thethermaltransmittance (U)valuesusedforA-rateddwellingswere0.251W/m2_-K_for_the

externalwalls,0.346W/m2_-K_for_the_roof_and_0.232_W/m2_-K_for

thegroundﬂoor.ThecorrespondingvaluesforF-rateddwellings (whichwerevery poorlyinsulated)were 2.071W/m2_-K _for_the

externalwalls,1.54W/m2_-K_for_the_roof_and_3.11_W/m2_-K_for_the

(8)

Table3

EnergyPlussimulationresultsforMarch2015,hourlyaverageindoorair,radiantandoperativetemperatures.

A-rated–boiler F-rated–boiler A-rated–Heatpump

Average St.dev Average St.dev Average St.dev

AirTemperature(◦C) 20.45 1.05 20.12 0.15 20.98 1.08

RadiantTemperature(◦_C) _20.09 _2.16 _16.21 _1.48 _22.20 _1.46

OperativeTemperature(◦_C) _20.27 _1.54 _18.17 _0.77 _21.59 _1.22

Table4

Rangeofclothingandmetabolicactivitiesavailableforselectioninconnectionwith entriesinthecomfortlogbookduringtheEcommonmeasurementcampaign,and thevaluesusedtocalculatetheirthermaleffects.

Clothingensemble Clovalue Metabolicactivity Metvalue Verylight(Sleeveless

T-shirt,iconinFig.3)

0.5 Lying/sleeping 0.7

Light(NormalT-shirt,icon inFig.3)

0.55 Sittingrelaxed 1 Normal(Knitsportshirt,

iconinFig.3)

0.57 Lightdeskwork 1.1

Ratherwarm

(Long-sleevedshirt,iconin Fig.3)

0.61 Walking 2

Warm(Long-sleevedshirt plusjacket,iconinFig.3)

0.91 Jogging 3.8

Verywarm(Outdoor clothing,iconFig.3)

1.30 Running 4.2

doubleglazingwith6mmglassthicknessand13mmairﬁllingwith aUvalueof2.7W/m2_-K_set_in_wooden_window_frames_with_a_U

valueof3.3W/m2_-K.

Thereasonwhythesamedoubleglazingwasusedforboth A-ratedandF-rateddwellingsis thatourinspectionrevealedthat allF-rateddwellingshadtheiroutsideglazingupgradedto dou-ble.Similarly,allthesimulationsmadeuseofthesamecondensing boiler(variableflow,nominalthermalefficiency0.89,maximum looptemperature100◦C)andradiatorswithaconstantwater tem-peratureof80◦C,sincenearlyalltheF-rateddwellingshadnew condensingboilersinstalled.Inbothcasestheinfiltrationwasset to0.5airchangesperhourwhiletheac/hduetowindownatural ventilationwassetto3.Thewindowscovered30%ofthewalland thelightinggainsweresetto5W/m2_-per₁₀₀_lx.

Table3presentstheaveragesofthehourlysimulationresults

forMarch2015,themonthwhentenantsusedthecomfortdials torecordcomfort-relateddata.Itwillbeseenthatthedifference betweentheradiantandairtemperaturesinA-rateddwellingswith aboilerwasonlyabout0.3◦C,appreciablylessthanthe respec-tivestandarddeviations.Itwasthereforedecidedthattheradiant temperatureforthesedwellingscouldbesetequaltotheair tem-peraturerecordedbythesensors.

Table3furthershowedthatthedifferencebetweentheaverage

radiantandairtemperaturesinF-rateddwellingswithcondensing boilerswasabout4◦C.Finally,thesimulatedradianttemperature forA-rateddwellingswithheatpumpsandunderﬂoorheatingwas about1.2◦Chigherthantheairtemperature,duetotheradiant heatingeffectofthehydronicﬂoorheatingsystem.The instanta-neousvalueofTmrtforthesedwellingswasthereforecalculatedas

Tair−4◦CandTair+1.2◦Crespectively.

Thus,theEnergyPlussimulationsmadeitpossibletoestimate theradianttemperatureonthebasisofthesensorreadingsofair temperature.

Furthermore,twovaluesoftheindoorairspeedwerechosen forthePMVcalculations,alowoneof0.1m/secandahigherone of0.3m/sec[8,40].

Table4presentsthevaluesusedtocalculatetheeffectsof

cloth-ingandmetabolicactivity,takenfromthemanualoftheAmerican SocietyofHeating,RefrigerationandAirConditioningEngineers,

(ASHRAE)[43].Tenantswereaskedtonotetheclothestheywere wearingandthemetabolicactivitiestheyperformedinthelogbook atregularintervals.Allclothingensemblesincludeshoes,socks andbriefsorpanties.Theinsulatingeffectofchair(0.15clo)was neglected.

3.2.2. PMVandreportedthermalsensationasfunctionsofthe operativetemperature

Asmentionedabove,tenantswereaskedtoﬁllinthecomfort logbookatleast3timesadaytoprovideinformationabouttheir clothingandthemetabolicactivitiestheyperformed.Theyalsohad torecordhowhotorcoldtheyfeltatthesametime.Allthis infor-mationwastimestampedandtimecoupledwiththequantitative datacollectedbythesensorsat5-minintervals.Thisintervalis assumedtobelargeenoughtoensurethatthecomfortlevelisnot relatedtopriorcomfortlevelsandconditions:anadaptiontimeof approximately4minwhenpeoplearesubmittedtotemperature stepchangeswasreportedinthestudiesofZhangetal.(2004)and Xiuyuanetal.(2014)[44,45],whichimpliesthatthecomfort sen-sationmaybeassumedtohavereachedasteadystateafter4min underthesameconditions.

ThePMVwascalculatedforeachroominthedwellingforall 5-minintervalsforwhichacompletesetofdatawasavailable. Fur-theranalysisofthedatapoints(metabolicactivity,clothing,actions, quantitativedataetc.)wasonlyperformedifmotionwasdetected intheroominquestionatanygiventime.Thisselectionprocedure resultedinatotalof194datapointsforthe2-weekperiodinwhich thetenantswereprovidedwiththecomfortdialandthelogbook. Theradianttemperaturewasassumedtobethatderivedfromthe EnergyPlussimulations(seeSection3.4.1),while,calculationswere performedfortwoairspeeds,0.1m/secand0.3m/sec.The calcu-latedPMVvaluesandthereportedthermalsensationwereplotted againsttheoperative temperatures,andregressionanalysiswas usedtodeterminethedatatrendline.

Asmostdatawereavailableforthelivingroom,Figs.5and6

showthescatterplotsoftheoperativetemperatureversusthePMV (calculatedforanairspeedof0.1m/sec)andthereportedthermal sensationforthelivingroomsofA/B-ratedandF-rated.Thesamples usedfordeterminationofthePMVandforthereportedthermal sensationareofdifferentsizesbecausemorerecordsofquantitative parametersfromthesensorswereavailablethanrecordsofthermal sensationmadewiththeaidofthecomfortdial.Furthermorethe numberofcasesfor“Alldwellings”isslightlydifferentthanthesum ofcasesforA/BandFdwellings;thisisbecauseintheregressions forthedifferentroomsandenergylabels,differentoutliershadto beexcludedeachtimeandbecausefortheA/Bdwellingskitchen andlivingroomdatawereputtogetherinthesameregression.

Regressionanalysisshowedsigniﬁcantcorrelationbetweenthe operativetemperatureand thePMVorreportedthermal sensa-tion(RTS) inboth A/B-ratedand F-rateddwellings.Signiﬁcance levels of p=0.01 and p=0.04 respectively were found in A/B-rateddwellings,andp=0.02andp=0.001respectivelyinF-rated dwellings.Itmaybenotedthatthekitchenandlivingroomwere treatedasa singleroomforthepurposesofregression analysis onA/B-rateddwellings,sincethekitchenandlivingroominthese dwellingswereinonecontinuousspacewithnodoorsorwalls

(9)

sep-Fig.5.OperativetemperatureversusPMVandRTS(reportedthermalsensation)scatterplotandregressionanalysistrendlineforthekitchen/livingroomsofA/Bdwellings atanairspeedof0.1m/sec.

Fig.6. OperativetemperatureversusPMVandRTSscatterplotandregressionanalysistrendlineforthelivingroomsofFdwellingsatanairspeedof0.1m/sec.

aratingthem.Thebasicstatisticaldataforallregressionlinesare givenforeachroominTables5and6.

Asexpected, both PMV and thereported thermal sensation increasewhentheoperativetemperatureincreases.Thesametrend wasobservedwhenthePMVcalculationwascarriedoutwithanair speedof0.3m/sec,bothforlabelA/B-ratedandF-rateddwellings.It isnoteworthy,however,thatthefullrangeofbothPMVvaluesand reportedthermalsensations(from−4to+3)isobservedin A/B-rateddwellingsattemperaturesbetween20◦Cand26◦Candin F-rateddwellingsattemperaturesbetween14◦Cand24◦C.PMV andreportedthermal sensationseemtobeclosertoeachother intheFdwellingsthanin theA/B dwellings.TheR2 _values _are

low(12.6%and10.9%),meaningthattheoperativetemperature explainsonly12.6and10.9%ofthevarianceinPMVorRTS.

Inordertoexploreiftherearesigniﬁcantdifferencesbetween theneutraltemperaturesforthelivingroombetweenthelabelA/B andFdwellingsananalysisofvariancewasperformed.The opera-tivetemperatures(perroomtype)oftheA/BandFdwellingswhile thetenant’srecordedneutralthermalsensationweregatheredand anANOVAwasperformed.Theresultswerehighlysigniﬁcant:for thelivingrooms p=4.66E-10,F=61.87and Fcrit=4.05 while for

thebedroomsp=7.22E-06,F=56.25andFcrit=4.74andtheyare

displayedinFig.7andshowthattherearesigniﬁcantdifferences betweentheneutraltemperaturesofthelivingroomsofA.BandF rateddwellings.

Fig.7.ANOVAsinglefactorfortheoperativetemperaturesthatcorrespondtothe neutralthermalsensationsofthetenants.

3.2.3. Neutraloperativetemperature(To)accordingtoPMVand

reportedthermalsensation

Theneutraltemperature,thetemperatureatwhichoccupants feelneitherhotnorcold,canbeestimatedbysolvingtheregression equationsofSection3.2.2forneutralthermalsensation.Solution oftheequationsinFigs. 5and6for PMV=0or forRTS=0thus permitscomparisonoftheneutraloperativetemperaturesbased onreportedthermalsensationandonPMVindex.

Only the signiﬁcant regression lines (as indicated in

Tables 5 and 6) were taken into account. Two of the

(10)

Table5

Basicstatisticaldatafortheregressionsbetweenoperativetemperature(OT)andPMV(signiﬁcantresultsinbold),andcalculatedneutraloperativetemperature(seesection 3.2.3).

0.1m/secairspeed

Room NeutralOT–all

dwellings pvalue Number ofcases R2 _Neutral OT–A/B-rated dwellings pvalue Number ofcases R2 _Neutral OT—F-rated dwellings pvalue Number ofcases R2 Kitchen 19.47 0.010 34 0.189 23.08 0.025 37 0.149 18.78 0.04 23 0.19 LivingRoom 21.67 0.003 79 0.105 20.3 0.02 48 0.086 Bedroom1 – 0.280 32 0.007 23.11 0.005 10 0.655 – 0.88 18 0.001 Bedroom2 18.61 0.003 21 0.223 – – – – 18.29 0.02 19 0.265

0.3m/secairspeed

Kitchen 19.61 0.008 32 0.211 23.4 0.038 37 0.117 18.99 0.01 21 0.302

LivingRoom 21.81 0.020 78 0.068 20.78 0.04 45 0.094

Bedroom1 – 0.655 26 0.008 – – – – – 0.68 16 0.003

Bedroom2 18.77 0.031 21 0.221 – – – – 18.4 0.02 19 0.265

Table6

Basicstatisticaldatafortheregressionbetweenoperativetemperature(OT)andreportedthermalsensation(RTS)(signiﬁcantresultsinbold),andcalculatedneutraloperative temperature(seesection3.2.3).

Room NeutralOT–all

dwellings pvalue Number ofcases R2 _Neutral_OT–A/B dwellings pvalue Number ofcases R2 _Neutral_OT–F dwellings pvalue Number ofcases R2 Kitchen 19.1 0.040 40 0.106 22.5 0.04 34 0.125 18.2 0.03 27 0.169 LivingRoom 23.2 0.001 89 0.121 20.4 0.001 57 0.175 Bedroom1 18.1 0.006 39 0.188 22.5 0.04 10 0.429 16.3 0.01 25 0.136 Bedroom2 – 0.578 24 0.014 – 0.30 3 0.797 – 0.92 21 0.000

Fig.8.NeutraloperativetemperaturescalculatedfromRTSandPMVregressionsforallroomtypesandenergyratings.

signiﬁcant,becauseof,theverysmallamountofdatapoints(only three)involvedinbothcase.

Fig.8shows theneutraloperativetemperaturesforallroom typesand energy ratingsderived fromthe calculatedPMV and thethermalsensationreportedbythetenants.Despitethe uncer-taintiesintheparametersneededtocalculatethePMV(airspeed andoperativetemperature)whichweredeterminedindirectlyon thebasisof assumptions andsimulations,theneutral tempera-ture(To)in both A/B and F dwellings is wellpredicted bythe

PMVmodelandcloselymatchestheneutraltemperaturesobtained usingthereportedthermalsensationoftenantsindifferentrooms of dwellings with different energy ratings. However, when all dwellingsareconsideredtogether,theneutraltemperatureisless wellpredictedbythePMVmodel,especiallyforthelivingroom. A/BandFdwellingsgivenoticeablydifferentresultshere.The aver-ageneutraltemperatureforthekitchenandbedroom2calculated foralldwellingsisquitesimilartothatcalculatedforFdwellings only(theregressionsforA/Bdwellingswerefoundnottobe

signif-icantinthiscase,asexplainedabove).Ontheotherhand,thereare markeddifferencesbetweenaverageneutraltemperaturesinthe kitchen,livingroomandbedroomofA/BandFdwellingsatboth airspeeds.

The regression predicts a neutral temperature for theliving roomsofA/Bdwellingsthatisabout3◦Chigherthanthatforthe liv-ingroomsofFdwellings.Thedifferenceisevenbiggerforbedroom 1,about4◦C.

ThelowerneutraltemperaturesinFdwellingscouldindicate thatair velocitiesarelowerin thesedwellings(thisis possible, becausethebalancedandmechanicalventilationsystemsusedin A/Bdwellingsareknowntogivehigherairvelocities).Other possi-bleexplanationsarethatpeopleinFdwellingsmaywearwarmer clothesorhavehighermetabolicactivity.Finally,thisdifference couldbe attributedtodifferent thermal expectationsor ageor genderdifferencesbetweenthetenantsofA/BandF dwellings. Thelast-mentionedexplanationseemsunlikely,however,sincethe averageageofthetenantsoftheA/BandFdwellingsis56and57

(11)

Fig.9. PlotsofreportedthermalsensationagainstPMVforA/BandFdwellings,atairspeedsof0.1m/secand0.3m/sec(bluelineTS=PMV,redline=regressionline).(For interpretationofthereferencestocolourinthisﬁgurelegend,thereaderisreferredtothewebversionofthisarticle.).

yearsrespectively,andmenandwomenwereequallydistributed betweenthetwodwellingtypes.

3.3. RelationshipbetweenreportedthermalsensationandPMV TofurthervalidatethePMVindexanditsabilitytopredict ten-ants’realthermalsensation,allthermalsensationvaluescollected duringthecampaignwerecomparedwiththecalculatedvaluesof thePMV.ThePMVvaluesforallenergyratings,typesofroomsand airspeedscenariosweregroupedinsub-setsaroundeachinteger valueofPMV.Forexample,thesub-setaroundaPMVof−1includes allPMVvaluesbetween−1.5and−0.5.Thereasonforthiswasthat tenantswereaskedtorecordtheirthermalsensationonascaleof integernumbersfrom−3to+3.ThePMVcalculations,ontheother hand,leadtonon-integernumbers.Furthermore,eachPMVvalue between−0.5and+0.5isconsideredtobeneutral.Valuesbetween −1.5and−0.5correspondtoarathercoolthermalsensation,and soon.Fig.9showtheplotsofreportedthermalsensationagainst PMVforallA/BandFdwellings,andforairspeedsof0.1m/secand 0.3m/sec.ThelineonwhichRTSequalsPMVseparatesthethermal sensationpointsthatarewarmerthanthePMVpoints(abovethe line)fromthosethatarecooler(belowtheline).Thebestﬁtlines areshowninred.

ThepredictionsuccessofthePMVmodelneverexceeds30%. WhenthePMVfailstopredictthethermal sensationcorrectly, itusuallyunderestimatesitespeciallyathigherairspeeds.These ﬁndingsareinagreementwithotherstudiesfromvarious

coun-tries[9,46,47]andaresimilarforeachtypeofroom(seeFig.10for abreakdownoftheresultsbyroom).However,thePMVmethod neverclaimedtogiveaccuratepredictionsonacasebycaselevel, butonlyatastatisticallevel.TheR2_values_given_in_Fig.₉_show_that

onlylessthan1.7%ofthevariationsinthereportedthermal sen-sationcanbeexplainedbythePMV;itfollows,therefore,thatthe PMVcannotbeconsideredasanaccuratepredictoroftheactual thermalsensationandthatotherparametersmustplayarole.

However,thebestﬁtlinesinallfourgraphscrosstheRTS=PMV linearoundtheneutrallevel,whichshowsthatneutralityiswell predicted.Furthermore,thebestﬁtlineforA/Bdwellings,iswithin thecomfortband(correspondingtoPMVvaluesbetween−0.5and +0.5)atalltimes,whileitissomewhatlowerinFdwellings.This showseitherthatthePMVdoesnotperformwelloutsidethe cli-matechamber,orthatpeopleadapttocoolerconditionsandtake actiontoimprovetheirthermalcomfort.Anotherpossibilitythat thecloandmetabolicactivityvaluesusedinourcalculationswere notaccurateenough,dueeithertoincorrectassumptions(wrong valuesattributedtoqualitativelyrecorded clovalues and activ-itylevelsfromASHRAEtables),ortoinaccuraterecordingbythe tenants.ThesepossibilitiesareexploredinSections3.6and3.7.

3.4. Clothingandreportedthermalsensation

Fig. 11 shows the clothing types worn by tenants in A/B dwellingsforeachreportedthermalsensation,whileFig.12gives thecorrespondingresultsforF dwellings.Thedifferenttypesof

(12)

Fig.10. ThermalsensationcomparedtoPMVforalltypesofrooms,energyratingsandwindspeedscenarios.

Fig.11.ClothingtypeswornatallthermalsensationlevelsinAandBdwellings(n=94).

Fig.12.ClothingtypeswornatallthermalsensationlevelsinFdwellings(n=155).

clothingarecolour-coded,whilethenumbersineachsegment rep-resentthenumberoftimesthetypeofclothinginquestionsisworn (totaln=94forA/Bdwellingsandn=155forFdwellings).

ThesestackedgraphsshowﬁrstofallthatnotenantsinA/B dwellingsreportedfeeling“cold”(inagreementwiththethermal sensationgraphsofFig.10),while8tenantsinFdwellingsmade thisobservation.Notenantsfromeithertypeofdwellingreported feeling“hot”.Themostpreferredclothingensembleforbothtypes ofdwellingsisthewarmensemble,asdeﬁnedin,Table4.When tenantsfeelwarmer,theyreplacethewarmensemblebylighter

ensembles.Theonlyinstanceswhentenantsreportwearingthe outdoorwarmensemblewereinA/Bdwellings,generallywhen theyhadjustcomeinfromoutsideandimmediatelyﬁlledinthe comfortapp/logbook.Theyusuallyreportedfeelingratherwarm orwarmin thesecases,probably becauseof thelower outdoor temperature.

Theclovaluecorrespondingtoneutralthermalsensationcanbe determinedbyplottingtheclovalueagainstthereportedthermal sensationandapplyingregressionanalysistotheresultinggraph.

(13)

calcula-Fig.13.ClovalueversusthermalsensationscatterplotandregressionanalysisforthelivingroomsofA/BandFdwellings.

tion,and Fig.13 showsthescatterplots andtrendlinesforthe livingroomsofA/BandFdwellings.Bothregressionswere signif-icantwithp=0.02andthetotalnumberofcaseswas31and62 respectively.Theregressionsforbedroom1ofA/Bdwellingsand bedroom2ofFdwellingswerefoundnottobesigniﬁcant.

Although the spread of the data is large, especially in A/B dwellings,theclovaluewasfoundto decreasewithincreasing thermalsensationinbothcases.Thisconﬁrmsthatclothingisan adaptivebehaviouralfeatureexercisedinordertofeelmore com-fortable.Accordingtotheregressionanalysis,15.7%ofthevariance inclorelatestothethermalsensation.Weseeafallinclovaluefrom alittleabove0.7(warmensemble)tosomewhatbelow0.5(light ensemble)in A/Bdwellingsasthethermal sensationrisesfrom −2(cool)to+2(warm.AsimilareffectisobservedinFdwellings, thoughthedropinclovalueongoingfromathermalsensationof −2(cool)to+2(warm)isslightlysmaller.

Thedatacollectedinthismeasurementcampaignindicatethat thetenantsofbothA/BandFdwellingsseemtowearmuchthe sametypeofclothing,whichmeansthatclothingdoesnotseem tobethereason forthelowerneutral temperaturesfoundin F dwellings(seeSection3.2.3).Thesametrendwasfoundforthe othertypesofrooms(kitchen,bedroom1and2)asthelivingroom.

Table7displaysthecalculatedclovaluescorrespondingto

neu-tralthermalsensation(zeroonthehorizontalaxisofFig.13)for eachtypeofroom.Identicalvalueswerefoundforthelivingroom (theroomforwhichmostdatawererecorded)inbothA/BandF dwellings.

Fig.14showstheclovalueplottedagainsttheoperative temper-atureforthelivingroomsofA/BandFdwellings.Bothregressions weresigniﬁcant,withp=0.0009and p=0.047respectively. The trendlinefortheA/Bdwellingsisslightlyascendingwhileforthe Fdwellingsitisslightlydescending.However,acloserlookatthe resultsfortemperaturesbetween20◦Cand24◦Cshowsthatthe clovalueforA/Bdwellingsstartsaround0.5(verylightclothing) andendsaround0.6(ratherwarmclothing).InFdwellings,theclo valueisalready0.6at20◦Candendsupslightlybelow0.6at24◦C. Inotherwords,peopleinA/Bdwellingsactuallytendtowear

some-what warmer clothingas theoperative temperaturerises from 20◦Cto24◦C,whilepeopleinFdwellingswearlighterclothing; theclovaluesconvergeatatemperatureof24◦C.Inbothcases,the slopeofthetrendlineisveryshallowandthevalueofR2_is_small.

Atoperativetemperaturebelow23◦C,theoccupantsofFdwellings seemtobewearingwarmerclothesthantheircounterpartsinA/B dwellings.TherisingtrendforA/Bdwellingsiscounterintuitive, butcouldberelatedtothehigherairspeedofthebalanced venti-lationsystem.Intuitivelythiscouldmeanthatwhentenantsturn uptheventilationinsuchcasestodealwithtemperaturerises,the higherairspeedsmaycausethentowearwarmerclothing.

Thefollowingprocedurewasusedtogainaninsightintothe effectoftheinaccuracyinclovaluesonthePMV:ThereportedRTS valuesandthecalculatedPMVvalueswerecollectedandsplitinto twogroups,oneforA/BdwellingsandtheotherforFdwellings. ThedifferencePMV-RTS,whichisthemostlogicalindicatorofthe qualityofthePMVcalculation,wasthencalculatedandassignedto 5groupsbyclovalue.(Sincenodatawererecordedforverywarm clothing,theclovalue1.30giveninTable4wasomitted).Aone wayanalysisofvariancewasthenusedtocalculatethe95% con-fidenceintervalofthedifferencePMV-RTSwithinthevariousclo categories.Ifthe95%confidenceintervalsoftwocategoriesoverlap, thismeansthatthequalityoftheprediction(PMV-RTS)cannotbe assumedtodiffersignificantlybetweenthetwoclocategories.Ifthe 95%confidenceintervalsdonotoverlap,thisindicatessignificant differencesinthequalityofprediction;inotherwords,thereare goodreasonstosuspectabiasrelatingtoclovalueinthebehaviour

ofthePMV[15].Figs.15and16displaythemeandifference

PMV-RTSandthe95%conﬁdenceintervalfor eachclovaluecategory theclosertothezeroline,themoreaccuratethepredictionofthe thermalsensation.

Theconﬁdenceintervalsof(PMV-TS)forA/Bdwellingsoverlap inthecategoriesclo=0.5,0.57and0.61,meaningthatthequalityof theTSpredictionbythePMVisprobablynotdifferentintheseclo categories.Theresultsforclo=0.91dohoweverdiffersigniﬁcantly fromthoseforothercategories.

(14)

Fig.14.ClovalueplottedagainstoperativetemperatureforthelivingroomsofA/BandFdwellings.

Fig.15.Predictivebias(PMV-TS)oftheclovalueagainstthePMVforA/Bdwellings.

Fig.16.Predictivebias(PMV-TS)oftheclothingvalueagainstthePMVforFdwellings.

ThereseemtobetwogroupsofclocategoriesforFdwellings withnodifferenceinthequalityofprediction.Oneisthegroupfor clo=0.5andclo=0.55andtheotherforclo≥0.57.Thequalityofthe predictionisworseinthelowerclocategoriesthaninthehigher.It mightbethoughatﬁrstsightthatthisisbecausethelowclovalues werenotaccuratelydetermined.Previousstudiesindicatethatitis difﬁculttodetermineclovaluespreciselyinsitu[48,54].

However, closer examination of the above graphs does not revealanyevidencethattheproblemliesintheclovalue.Inorder toreducethepossiblebiasatlowclovaluesinFig.16,theaverage PMV-RTSvalueforthelowerclocategorywouldhavetomove ver-ticallyupwardstowardsthezeroline.SinceRTShasaﬁxedvalue

reportedbythetenants,thismeansthatPMV(andhencetheclo value)wouldhavetoincrease:forexample,thecategoryclo=0.5 mightmoveupto0.61forA/Bdwellingsand0.57forFdwellingsif theclovaluesweremeasuredaccurately.Thisisunlikely,however, sinceitwouldhavetheresultofmovingallclocategoriescloser togethersothatit would beimpossibletodistinguishbetween them.

Alternatively,theproblemmaynotlieinthePMVcalculation andthepoordeterminationoftheclovaluebutinthereported ther-malsensation.Weusedthewidelyaccepted7-pointscale,butthis scalemaybetoodetailedfortherangeofoperativetemperature foundinthebuildingsthatweremonitored.Peopleareaccustomed

(15)

betweenA/BandFrateddwellings.TheAnovawasperformedfor theclothinglevelthatcorrespondedtotheneutralvotesof ther-malsensationofthetenants.Theresultwashighlyinsigniﬁcant withp=0.993andF=6.23E-05andFcrit=3.94whichmeansthat

wecannotrejectthenullhypothesisthattheclovaluesinthe liv-ingroomforneutralthermalsensationbetweenA/BandFrated dwellingsareequal(Fig.17).

3.5. Metabolicactivityandthermalsensation

Fig.18displaysthemetabolicactivityforeachthermal sensa-tionlevelrecordedbytenantsofA/Bdwellingswiththeaidofthe comfortdialandthecomfortlogbook,whileFig.19givesthe cor-respondingresultsforFdwellings.Themetabolicactivityshown hereistheaverageactivitylevelasdeﬁnedinTable4reportedfor thehalfhourbeforeuseofthecomfortdial.Theactivitylevelsare colour-coded,whilethesuperimposednumbersrepresentthe fre-quencyofreportingeachtypeofmetabolicactivity(intotaln=147 forA/Bdwellingsandn=206forFdwellings).

Themetabolic activitymost oftenreported in both A/B and F dwellings was ”relaxed sitting”. This was followed by “light

Fig.17.ANOVAsinglefactorfortheclovaluesinthelivingroomsforneutralthermal sensationsofthetenants.

deskwork”andthen“walking”inA/Bdwellings.”Walking”was recordedthan”lightdeskwork”inFdwellings.

”Lying/sleeping”wasthefourthmetabolicactivitylevelforboth typesofdwellings.Themetabolicactivityofthetenantscanbe calculatedasafunctionofthereportedthermalsensation,inmuch thesamewayaswasdonefortheclovalueabove.Fig.20showsthe scatterplotsandtrendlinesforthemetabolicactivityvalueplotted againstreportedthermalsensationsforthelivingroomsoftheA/B andFdwellings.Bothregressionsweresigniﬁcantwithp=0.008 andp=0.04respectively,andthetotalnumberofcaseswas56and 82respectively.TheRTSexplains12%ofthevarianceofmetabolic activityinA/Bdwellings,butonly5%inFdwellings.Thestatistical signiﬁcancevaluesforeachregressionaregiveninTable8.

Theregressionforbedroom2wasonlysigniﬁcantinFdwellings. The regressionsfor all other types of roomwere signiﬁcantat p≤0.01.ThemetabolicactivityinthekitchenofA/Bdwellingsis

Fig.18.MetabolicactivityreportedatvariouscomfortlevelsinA/Bdwellings(n=147).

(16)

Fig.20.MetabolicactivityversusreportedthermalsensationscatterplotandregressionanalysistrendlineforthelivingroomsofA/BandFdwellings.

Table8

BasicstatisticaldatafortheregressionsbetweenTSandmetvalues(signiﬁcantresultsinbold),andcalculatedmetvaluesforneutralthermalsensation.

Room Averagemetvalueall

dwellings

pvalue Averagemetvalue A/B-rateddwellings

pvalue Averagemetvalue F-rateddwellings pvalue Kitchen 1.53 0.002 1.88 0.01 1.38 0.01 LivingRoom 1.41 0.039 1.44 0.008 1.32 0.043 Bedroom1 1.46 0.048 1.28 0.050 1.90 0.040 Bedroom2 0.286 0.069 1.45 0.048

appreciablyhigherthaninthelivingroomandbedroom1,which istobeexpectedsincethekitchenis wheredinnerisprepared andwherepeopleusuallyhavebreakfastinthemorningbefore theyleavehome.Boththosecommonactivitiesforkitchensare associatedwithhighermetabolicactivitylevels.Furthermore,A/B dwellingsallhadtheirkitchensandlivingroomscombinedina sin-glelargespace.Thisislikelytomakeformorefrequentmovement betweenthetwohalvesofthespaceforexample;breakfastmaybe preparedinthekitchenandeatenatthetableintheadjacentliving area,unlikethecasewithseparatekitchenscontainingabreakfast TableSimilarconsiderationsapplytothemetabolicactivitylevels inthekitchensandlivingroomsoftheFdwellings.Themetabolic activityishigherinthekitchenthaninthelivingroom,butalot lessthaninA/Bdwellings.

AlltheFdwellingsinthisstudyhadseparatekitchens,andthe conﬁnedspacecouldleadtolowermetabolicactivity.Thehighest metabolicactivityforneutralthermalsensationwasobservedin thebedroom1ofFdwellings.ThedatapointsforA/Bdwellingsin thiscasewerefor3dwellings;twoofthosebelongedtoelderly peo-plewhousedthebedroomonlyforsleepingwhilethethirdhouse belongedtoayoungcouplewhoalsousedthebedroomonlyfor sleepingsincetheyhadasecondbedroomthattheyusedasastudy. TheFdwellingsontheotherhandprovidedenoughdatapointsfor accuratecalculationoftheregressions;thesehouseholdsallhad youngfamilymembers(fromsmallchildrenuptoteenagers)who usedtheroomsactivelyduringthedaytime,notjustforsleeping.

Apartfromthespecialcasesanalysedinthepreviousparagraph, similarlevelsofmetabolicactivitywerefoundinthelivingroom inbothtypesofdwellings;thistypeofroomwasusedinthesame wayinbothA/BandFdwellings,andalsoprovidedmostofthedata pointsfortheregressionanalysis.ThisisalsoevidentfromFig.20, wherethereportedthermalsensationrangesfrom−3to+2inboth casesandthemetabolicactivityusuallyvariesfrom0.75to1.5.

Fig.21displaysthemetabolicactivityasafunctionofthe oper-ativetemperatureforthelivingroomsofA/BandFdwellings.As inthecaseoftheclovaluediscussedinSection3.6,thetrendline isrisingforA/BdwellingsandfallingforFdwellings,convergingto thesamelevelsofmetabolicactivityasthetemperaturerisesfrom 18◦Cto24◦C.Furthermore,theslopeofthetrendlinesisvery

shal-lowandtheR2_values_are_even_lower_than_for_the_clo_trend_lines._The

increaseinthemetabolicactivityofthetenantsinA/Bdwellingsas theoperativetemperaturerisesmaybeduetothedesignofthese dwellings.Mostofthemhavethekitchensandlivingrooms com-binedinonecontinuousspace.Cookingcausesthetemperatureof thespaceandthelevelofmetabolicactivitytorise,sinceitrequires moreactivitythantypicallyfoundinthelivingroom,whichis nor-mallyassociatedwithmorerelaxedactivitiessuchaswatchingTV, readingabookorlisteningtomusic.Peoplewhowererecording theirmetabolicactivityinthelivingroomweremorelikelytobe inarelaxedstate,sittingonacouchorinachair,whilepeople recordingtheirmetabolicactivityinthekitchenwouldbemore active(cooking,usingthedishwasheretc.).

Asintheprevioussection,weexploredtheeffectthat inaccu-racyindeterminationofthevaluesofmetabolicactivitymighthave onthecalculatedPMV.ThedifferencePMV-RTSwasonceagain determined,groupedbytheenergy ratingofthedwellingsand categorisedbymetabolicactivityvalueinto7groups asdeﬁned

inTable4.Onewayanalysisofvariancewasagainusedtotest

whetherthedifferentmeandiscrepanciesforthevariousgroups couldbeattributedtochance.Figs.22and23displaythemean discrepancy(predictivebias)plottedagainstthemetvalue(met valueof1.5appearsinthegraphdespiteitsabsenceinTable4; thatisbecausetenantsmanytimesrecordedmorethanonetype ofmetabolicactivityforthepasthalfhourandsoanaveragemet valueofthoseactivitieswasused),togetherwiththe95% conﬁ-denceintervalforeachcategory.IfthePMVwerefreefrombias relatingtothemetvalue,theconﬁdenceintervalsofallcategories wouldoverlap.

It wasfoundthatthediscrepancies werenot attributableto chance and were highly signiﬁcant at p <0.001. A/B dwellings showed substantial bias for met=0.7 (lying/sitting), met=1 (relaxedsitting)andmet=4.2(running),thoughthebiasismuch smallerinthelasttwocategories.ThePMVishoweverfreefrom seriousbiasformetvaluesof1.1(lightdeskwork),1.5and2 (walk-ing).

The discrepancies in F dwellingswere also not attributable tochance and were highly signiﬁcantat p <0.001.The biasin thesedwellingswasmoresubstantialthaninA/Bdwellings.All

(17)

Fig.21.Metabolicactivity(metvalue)plottedagainstoperativetemperatureforthelivingroomsofA/BandFdwellings.

Fig.22.PredictivebiasofthemetvalueagainstthePMVforA/Bdwellings.

Fig.23.PredictivebiasofthemetvalueagainstthePMVforFdwellings.

categoriesofmetabolicactivityshowedmarkedbias,apartfrom met=1.5andmet=2.

Anova:singlefactorwasusedtodetermineifthereareany sig-niﬁcantdifferencesbetweenthemetabolicactivityvaluebetween A/B and F rated dwellings. The Anova was performed for the metabolicactivitylevelforthelivingroomsthatcorrespondedto theneutralvotesofthermalsensationofthetenantsforbothA.B andFdwellings.Theresultwashighlyinsigniﬁcantwithp=0.488 andF=0.483andFcrit=3.91whichmeansthatwecannotrejectthe

nullhypothesisthatthemetabolicactivityvaluesinthelivingroom

forneutralthermalsensationbetweenA/BandFrateddwellings areequal(Fig.24).

4. Discussion

Despitelimitationsonmaterialsandequipment,theEcommon measurementcampaignsuccessfullycollectedadequate quantita-tiveandqualitativedataoncomfortandoccupantbehaviourina relativelyeasyandunobtrusivewayintheresidentialenvironment. Thetenantswereveryinterestedinthecomfortdial,andusedit

(18)

Fig. 24. ANOVA single factor for the metabolic activity values in the living rooms for neutral thermal sensations of the tenants.

much more often than the requested minimum three times a day. The high frequency (every 5 min) of the sensor measurements of quantitative parameters, the unobtrusive wireless method used to collect thermal sensation data and the remote management of the entire sensor system ensured minimal data loss over the whole six months of the measurement campaign.

Furthermore, the reported thermal sensation data used for the comfort calculations were collected electronically for the first time with a time stamp linked to the quantitative sensors; this approach compares favourably with the questionnaire tenants had to fill in by hand in previous monitoring campaigns. The precision of data collection is much higher in this approach: tenants no longer had to write down the exact time they filled in the comfort log book, and the 5-min interval used for quantitative data collection ensured that the quantitative data entered in the comfort log book could be easily and precisely linked with the qualitative data. At the same time, the motion sensors helped to identify where the tenants were when they were filling in the comfort log book, thus allowing the appropriate room type to be linked with the corresponding data entry.

One of the issues that arose during the analysis of the campaign data was the possible effect of direct solar radiation on tenants’ thermal preferences. EnergyPlus accounts fully for the effects of direct and diffused solar radiation in the interior of a building when simulating air, radiant and operative temperatures[49]. However, these simulations were based on a reference building (described in Section3.2.1) which may differ in architecture (placement, size and orientation of the windows) from the real buildings dealt with in the campaign. Furthermore, while the average hourly radiant temperature in each ﬂat was approximated in detail in EnergyPlus simulations, we have no way of knowing whether tenants were sitting in front of a window while they recorded their thermal sen-sation. The Netherlands may not be the sunniest country in the world and monitoring did take place during the winter, but direct solar radiation could still have played a role in determining tenants’ thermal sensation. Besides, the radiant temperature at a given time may differ from the average hourly value obtained from Energy-Plus simulations. However,Table 3shows that the highest standard deviation found for the air temperature was 1.08◦C while that for the radiant temperature was 2.16◦C. In order to estimate the effect of temperature variations, the PMV equation was subjected to sen-sitivity analysis with reference values of 20◦C for air and radiant temperature. The maximum effect on PMV produced when the air and radiant temperatures were varied in 0.5◦C steps from 18◦C to 22◦C (in order to cover the entire possible range of twice the standard deviation) was 0.7. It follows that possible deviations of the radiant temperature from the average at a given time shouldn’t have a dramatic effect on the PMV.

Another point of discussion is related to the 7-point scale used for the PMV. This scale was developed in climate chamber experiments where subjects were exposed to a variety of climatic

conditions. It was validated by determining the regression between the calculated PMV values and people’s reported thermal sensa-tions. There is however no guarantee that a thermal comfort level of−3 reported by a Dutch subject corresponds to −3 on the PMV scale. Greater robustness could be achieved by collecting large scale data sets for a wide variety of subjects and areas in the Netherlands and using these data to deﬁne the PMV scale for the Netherlands together with the thermal sensation scale for Dutch subjects. It is claimed that the PMV model can be applied irrespective of climate and social convention, way of life and kind of clothing, though some distinction needs to be made between winter and summer[13]. In contrast with this, previous thermal comfort studies found that subjects’ thermal sensations varied from individual to individual and were dependent on race, climate, habits and customs[50,51]. Furthermore, the thermal sensations recorded by the tenants in the present study ranged mainly between−2 and +2. Comfort lev-els of−3 (cold) were recorded very infrequently (only 9 cases out of 192, all in F dwellings), while comfort levels of +3 (hot) were never recorded. Most reported comfort levels were between−1 and +1. As discussed in section 3.6, the PMV shows little bias for clo and met values that are close to those for neutral comfort levels. These facts reﬂect the possible effect of psychological adaptation on the tenants in the present study. Thermal adaptation can cause peo-ple to perceive, and react to, sensory information differently on the basis of past experience and expectations[52]. Personal comfort set points are far from thermostatic, and expectations may be relaxed in a way that resembles the habituation found in psychophysics

[53,54]where repeated exposure to a constant stimulus leads to

a diminishing evoked response[52]. The tenants who participated in the Ecommon campaign might not even have a clear feeling of what a thermal sensation of−3 means. They are always in their own personal space, which they always try to keep as comfort-able as possible, and this feeling of comfort is what they know and what they associate with their home. It follows that their response are more accurate around the neutral comfort level and less accu-rate at more extreme comfort levels approaching−3 or +3, which correspond to thermal sensations to which they are much less accustomed in their own homes. Similarly, our analysis of the bias in PMV due clo and met values showed that bias was low around the neutral point, but could be substantial at lower and higher clo and met vales.

5. Conclusions and proposals for further research

The PMV model predicts neutral temperatures for the various room types well, in line with those derived from the thermal sen-sations reported by tenants.

The thermal sensation reported by tenants ranged from −3 (cold) to +2 (warm), while the PMV calculations showed ther-mal comfort levels ranging from−8 to +3. This means that people feel more comfortable than indicated by the predictions. The PMV model underestimates the thermal comfort of the tenants in residential dwellings. Furthermore, people seem to have better per-ception of thermal comfort around neutrality. This could indicate a certain level of psychological adaptation and expectation since each person’s home is associated with comfort, relaxation and rest, in contrast to ofﬁce buildings for example that are associated with work and higher levels of stress, effort and fatigue.

Tenants of A/B and F dwellings seem to show no differences in clothing and metabolic activity patterns, even though, F-rated dwellings had lower neutral temperatures. Age and gender also seem to have no effect on neutral temperature levels, which leaves the indoor air speeds and psychological adaptation and expecta-tions as possible explanatory factors for the difference in neutral temperatures between A/B and F dwellings.