Zero energy potential of a high-rise office building in a Mediterranean climate

(1)

Zero energy potential of a high-rise office building in a Mediterranean climate

Using multi-objective optimization to understand the impact of design decisions towards

zero-energy high-rise buildings

Giouri, Evangelia Despoina; Tenpierik, Martin; Turrin, Michela

DOI

10.1016/j.enbuild.2019.109666

Publication date

2020

Document Version

Final published version

Published in

Energy and Buildings

Citation (APA)

Giouri, E. D., Tenpierik, M., & Turrin, M. (2020). Zero energy potential of a high-rise office building in a

Mediterranean climate: Using multi-objective optimization to understand the impact of design decisions

towards zero-energy high-rise buildings. Energy and Buildings, 209, [109666].

https://doi.org/10.1016/j.enbuild.2019.109666

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)

Green Open Access added to TU Delft Institutional Repository

'You share, we take care!' - Taverne project

https://www.openaccess.nl/en/you-share-we-take-care

Otherwise as indicated in the copyright section: the publisher

is the copyright holder of this work and the author uses the

Dutch legislation to make this work public.

(3)

ContentslistsavailableatScienceDirect

Energy

&

Buildings

journalhomepage:www.elsevier.com/locate/enbuild

Zero

energy

potential

of

a

high-rise

oﬃce

building

in

a

Mediterranean

climate:

Using

multi-objective

optimization

to

understand

the

impact

of

design

decisions

towards

zero-energy

high-rise

buildings

Evangelia

Despoina

Giouri

∗

,

Martin

Tenpierik

,

Michela

Turrin

Faculty of Architecture and the Built Environment, Delft University of Technology, Julianalaan 134, 2628 BL, Delft, the Netherlands

a

r

t

i

c

l

e

i

n

f

o

Article history:

Received 25 August 2018 Revised 10 June 2019 Accepted 1 December 2019 Available online 2 December 2019

a

b

s

t

r

a

c

t

Currently40%ofEU’sfinalenergyconsumptionisattributedtobuildings.AchievingtheEU’sclimate tar-getswouldentailimprovedstrategiesindesigningnearlyZeroEnergyBuildings.Thisresearchaimedto createanintegrateddecision-makingstrategyindesigningZEBswiththeuseofmulti-objective optimiza-tionofbuildingdesign andconstructionparametersforminimizingenergydemand,whilemaximizing energyproductionand adaptivethermalcomfort.Goal istodefinewhichparametershavethe highest impactand potential forfurther optimizationand tooffer analternative tocurrentsteppedstrategies suchasthe NewSteppedStrategy.Theproposed integratedapproachisappliedonatypicalhigh-rise officebuilding inGreece. Energy simulationswithDesignBuilderareused as benchmarkforthe opti-mizationrunwithEnergyPlusthroughRhinoand Grasshoppersoftwareviatheplug-insHoneybeeand Ladybug,coupledwithmodeFRONTIER.Forthefirstoptimizationround,theinvestigatedparametersare: window-to-wallratio,wallU-value,glazingconstructionU-value,glazingg-value,air-tightnessofthe fa-cade, cooling set-pointof the mechanical cooling systemand PV facade surfacearea.For thesecond round,theparameters ofwindow-to-wall ratio,shadingareaand PVsurfaceareaareadaptedfor four facadeorientations.Theoptimizations resulted inabuildingwithan annualfinalenergyreductionof 33%.

1. Introduction

Within an urbanizing environment where 66% of the world’s populationisprojected tobeurban by2050[27],theneedto re-duce globalCO2 emissionsisbecomingapparent.Currentlyinthe EUnearly40%ofﬁnalenergyconsumptionand36%ofgreenhouse gas emissionsare attributed to buildings[2]. Inorder to achieve theEU’s 2020targetsintheEPBD Directive,butalsotomeetthe longer term objectives of the climatestrategy of the low carbon economy roadmap 2050, optimized strategiesin designing nearly ZeroEnergyBuildings (nZEBs)andhigh-risenZEBsneedtobe de-veloped. Azeroenergybuildingreferstoabuildingthatproduces asmuchenergyasitconsumesinadeﬁnedperiod.

The existing stepped strategies such as the Trias Energetica

[15]andtheNewSteppedStrategy[28] optimizedesignvariables andespeciallypassiveandactivedesignsystemsinastepped ap-proach.TheNewSteppedStrategydoessobyincludingprinciples

∗ _{Corresponding author: Faculty of Architecture and the Built Environment, Delft}

University of Technology, Julianalaan 134, 2628 BL, Delft, the Netherlands E-mail address: edgiouri@gmail.com (E.D. Giouri).

for closing cycles in the built environment. However, these ap-proachesandmanyothersarequalitativeinnature.Thedesignof a ZEB entailsparameters that have conflicting influence on vari-ousenergyloadsandthermalcomfortlevels. Moreover,some pa-rameterscanhaveminimalinfluenceonenergyloadscomparedto others.Inorder toinvestigatethe potential forimprovementand trade-off designsthat optimallysolveconflictingproblems,an ex-tensivequantitativedataanalysisofmultipledesignsandan inte-grated optimizationof various conflicting passive andactive sys-temsare needed. Thus,thisstudyintroduces the implementation ofan integratedstrategy through multi-objectiveoptimization of various building design parameters, that could resultin a highly energy-efficientandthermallycomfortablebuilding,through steer-ingresources in directionsthat havemorepotential for improve-ment and thus could also lead to more affordable ZEBs. As the briefliterature reviewinthenextsection willshow,several stud-ies argue for the need for better integrated quantitative strate-giesthatcanaidthedesignprocessofhighperformancebuildings. Some of these studies have already applied a basic optimization strategybasedonsingleormulti-objectiveoptimization.However, practically all of thesestudies focused on small relatively simple buildingsofmostlya few stories highand/ordid notinclude a https://doi.org/10.1016/j.enbuild.2019.109666

(4)

2 E.D. Giouri, M. Tenpierik and M. Turrin / Energy & Buildings 209 (2020) 109666

thorough analysis of the design space. Particularly such an analysisofthedesignspaceiscrucialforaquantitativeintegrated approach,because it on theone hand allows to check the valid-ityoftheresultsandontheotherhandallowsforshowingwhich parameters are impacting the energy useof the building andits indoorcomfortmostly.Particularlytheseparametersrequiremost attentionduringthedesignprocess.

The main objective of this study [8], is to propose an inte-grated strategy for the early design phase of a ZEB, which con-traryto existingstepped approaches, entailsthealgorithm aided, multi-objectiveoptimizationofpotentially conﬂictingpassive and active design parameters. The objectives refer to minimizing en-ergy demand, while maximizing energy production andadaptive thermalcomfortlevels.Theimplementationoftheproposed inte-gratedstrategyaimstohelp thedesignertakeinformeddecisions inshapinganearly-phasedesignstrategy,asasetofmeasures, to-wardsachievingaZEB.Theproposed integratedstrategy istested on a typical central-core, open plan, high-rise oﬃce building for the hot-summer Mediterranean climate (Csa) of Athens, Greece. Thereforethestudyaimstoanswerthefollowingquestions:

• Whatisthemosteffectivecombinationofparametersthatcan lead toa potentially zeroenergyhigh-riseoﬃce buildingina hot-dryclimate?

• Which parametershavethehighestimpact onthedesignofa ZEBhigh-risebuildinginahot-dryclimate?

• Howdodifferentparametersofthemostefficientdesign strat-egy influence different aspects of final energy and thermal comfortinthebuilding?

The proposed integratedstrategyandits computationalset-up isgeneric andcan be applied to various building typologies and climaticzones. Nonetheless,inthisresearch, theintegrated strat-egy isapplied on a building,that is representative ofa typology largelyappliedonhigh-riseoﬃcebuildings,that ofacentralcore open-planoﬃcebuilding.Thereforetheresultsofthisresearchcan beextrapolatedasmeasures towardsZEBsofthistypology inthe climaticzoneinvestigated.

The performance indicators for the optimizations refer to an-nual ﬁnal energy and adaptive thermal comfort levels. The ﬁrst entails annual energy demand for cooling, heating, lighting and equipmentandannualenergysavings asenergy productionfrom PVpanels. The latteris translatedas the percentageof time, for an annual period, at which comfortable conditions occur when the indoor temperature is within the comfort range determined bythe prevailingoutdoor temperature. Thisstudyfocusesonthe decision-makingdesign phaseof anew building,thus thechoice oftheparametersoptimized aimstoextract thetrendsthat indi-catetheelementsofabuildingthathavemorepotentialforfurther improvement,towardscreatingaZEB.Theaforementioned perfor-manceofabuildingregardingtheindicatorscanbeaffectedbythe shape and orientation of a building. Additionally, parameters af-fectingtheperformance canbethewindowtowallratio,thewall U-value,theglazingconstructionU-value,theglazingg-value,the air-tightness of the facade, the shaded area of the openings,the coolingset-pointofthemechanicalcoolingsystemandthesurface areaofphotovoltaicpanelsonthefacade.Thefollowingliterature review indicates building parameters optimized towards energy-savingsolutionsforbuildings.

2. Overviewofpreviousstudies

The optimal combination of parameters that lead to the de-sign of a nearly zero energy building can be attained through multi-objectiveoptimizationstowardsathermallycomfortableand energy-eﬃcientbuilding.Thefollowing studiesare relatedto this concept.

2.1. Reviewonoptimizationstudies

Xuetal.[29] aimed tominimize heatingandcooling loadsof anofficespaceinSeoulusingoptimizationdrivenbyNSGA-II.The followingparameterswereinvestigated:floorarea,building orien-tation,ceilingheight,aspectratio,plenumheight,window-to-wall ratio,wallinsulation,windowinsulation,solarheatgaincoefficient andairleakage. The influenceof HVACsystemswasalso investi-gatedanditwasconcludedthatdifferentheatingandcooling sys-temsledtoadifferentoptimumbuildingdesign.

AiminthestudyofYuetal.[31]wastofindtheoptimal solu-tionforaresidentialbuildinginChongqing,China,withregardsto energyconsumptionandindoorthermalcomfort.Theoptimization wasdrivenwithNSGA-IIandEnergyPluswasusedforenergy sim-ulations.Severaldesignvariableswereinvestigatedlikefloorarea, orientation,shape,wallandroofheattransfercoefficient,walland roof thermal inertia index, window heat transfer coefficient, and windowtowallratioforvariousorientations.

Hamdy et al. [9] used multi-stage optimization to develop a cost-optimalandnearly-zero-energybuilding.Theaimwastofind optimalcombinations ofdesign variables that influencethe ther-mal performance (heating, cooling, comfort) of the house: the building-envelope (insulationthicknessofexternalwall, roof,and floor,windowtype,andbuildingtightness)andtheheat-recovery unit.Asingle-familyhousewassimulatedwithMATLABand TRN-SYSsoftwareaidedbyavariantofthegeneticalgorithmNSGA-II.

Evinsetal.[6]exploredthetrade-off betweencostandcarbon emissions,forthedesignofamodularhotelunit.Anoptimization wasappliedforvariousclimatetypes.Investigatedvariables were envelope design parameters, different HVAC systems and energy generationfromPVandsolarthermalpanelsinstalledontheroof. TheneedofoptimizationforimprovingbuildingandHVAC sys-tem performance was underlined by Holst [10]. In his optimiza-tionforaschoolbuildinginTrondheim,Norwayheexploredboth passiveandactive designaspects.Thelaterentailednightsetback temperatures.

In a design optimization of insulation andspace conditioning load, Shi[24] aimed tofind abalance inthe objectives.The case studyisaone-storyofficebuildinginNanjing,China,with3 ther-malzones:aconferenceroomand2officespaces.modeFRONTIER was used as optimization platform coupled with EnergyPlus for simulationofthespaceconditioningload.

Loonen etal. [14] explored the potential of Climate Adaptive BuildingShells byusingbuildingperformancesimulationand op-timization.Theobjectivewastobalanceenergydemandand ther-mal comfort, by minimizing the sum ofheating andcooling en-ergy demand and the number of hours per year that tempera-tureexceeds25°C.Staticoﬃcebuildingshelldesignswere inves-tigated forashort-term periodandalong-termperiod.The non-dominated sorting genetic algorithm II (NSGA-II) drove the opti-mization.

Chantrelle et al. [4] used a genetic algorithm (NSGA-II) cou-pled withTRNSYS forthe renovation ofa school inthe southern French.Theoptimizationobjectiveswereannualenergy consump-tion (cooling, heating, lighting andventilation), thermalcomfort, cost and environmental impact. The investigated variables were various types of external wall, roof, ground ﬂoor, intermediate ﬂoor,internalwallandwindowtypes.

Caldas[3]usedGENE_ARCHtooptimizethebuildinggeometry inseveralapplications,dealingwithissuesofenergydemand, ma-terials,costs,andlightingbehavior.

Theneedformulti-objectiveoptimizationinearlydesignstages washighlightedbyNegendahlandNielsen[19].Theirresearch fo-cusedonthedesignofaspeciﬁcfacadepattern,basedonthe opti-mizationofenergyuse,capitalcost,daylightdistributionand ther-malloads.

(5)

Manzan [17] usedgeneticoptimization tofind ageometry for external, fixed,shadingdeviceswithlow energyandcostimpact. Theoptimizationmodifiedshadingdeviceheight,width,angle, dis-tance from the wall and various glazingproperties for an office roominTriesteandRome.Thermalloadsimulationswererunwith ESP-randilluminancesimulationswithDAYSIM.Theoptimization wasdrivenbymodeFRONTIER,withtheNSGA-IIgeneticalgorithm. The overview of optimization studies indicates the need for applying early-stage optimizationforvarious building design and construction parameters,aswellasforHVACsystemsandenergy generationfromrenewablesources inanintegratedmanner. Nev-ertheless,theoverviewindicates thatonly afew studiesexamine abroadrangeofvariables,whileincludingpassiveandactive mea-sures(Xu etal.[29],Evins etal.[6],Holst[10]). Thosewho focus on a limitedspectrum ofvariables, maylack in investigatingthe interrelationsandreciprocalinfluencesderivingfromanintegrated approach.

2.2. ReviewonZEBstrategies

In 1996 Lysen[15] presented a steppedenvironmental design approachforenergycalledtheTriasEnergica. The1ststepaimed toprevent theuseofenergy.The2nd stepreferstousing renew-ableenergysources aswidely aspossible.The laststeprelatesto theremainingenergydemandandentailsusingfossilfuelsas eﬃ-cientlyandcleanlyaspossible[11].

Another stepped strategy is The New Stepped Strategy that eliminatestheuseoffossilfuelsinexchangeoftheexploitationof waste ﬂows [13]. The 1ststepincludes passivestrategies such as shadingorimprovedinsulationofthebuildingenvelope.The2nd step includes reusing and recycling waste ﬂows, like exchanging heat betweendifferentbuildings orfunctions. The3rdstep refers to producing energy from active systems like PV panels orsolar collectors.

With regard to the Climate Responsive Design approach, ac-cordingto Looman[13],the designshould exploit naturalenergy sources likethe sun,earth,wind, sky,water,complemented with energy recovery from waste ﬂows. A combination of techniques willresultintoalow-energy,comfortablebuilding.

Thepassivehousestrategyisbasedontheprinciplesof reduc-ing losses andoptimizingpassive solargains, withoutthe useof activesystems[22].Thestrategyreferstooptimizingvariableslike theUvalueofexternalwalls,roofs,shadingsurfaces,windowarea, etc.individually,inasteppedapproach[21].Thisprocess may re-quireaminimizednumberofsimulations,butthe resultmaynot be optimal due to the conﬂicting inﬂuence that many variables can have on various energy loads. For example, large windows mayleadtoincreaseddaylightexploitation,buttheincreasedsolar gainscanleadtooverheating,especiallyinthesummer.

The active house strategy entails comfort, energy and envi-ronment considerations. Comfort amongothers includesdaylight, thermal comfort andair quality.The energy aspect refers to en-ergy demand reduction, using renewable energy and minimizing useofenergyfromfossilorigin[1].Thisstrategysuggests optimiz-ingvariableslikewindowsize,shadingandthermalmassand oth-ersformaximizingthermalcomfort.Optimizingtheenergyaspect includesoptimizingwallUvalue,buildingorientation,inﬁltration, using natural ventilation and increasing daylight availability [1]. Theapproachusedtoimplementthisstrategyisasteppedone,by optimizingaspectslikedaylightaccess,thermalcomfort,renewable energysourcesandenvelopedesignvariablesseparately.

Aquestionarisesastohowcanthedesignerﬁndthetrade-off point betweenmaximizing daylightexploitation, thermalcomfort andenergyproductionfromRES(renewableenergysources)while minimizingenergydemand.Forexample,largewindowsmean in-creased daylightexploitation, butminimize the opaque wall area

forintegration of PVpanels on the facade, thus reducing energy productionfromRES.Alsodependingontheclimaticzone,thermal comfortandthermalloadsareaffectedpositivelyornegativelyby largewindows.

From thereview of existing ZEBstrategies,it isapparent that designing a ZEB includes amongothers, objectives like maximiz-ing thermal comfort and daylight access and energy production fromRES, whileminimizingenergydemand.Theseobjectivescan bevariablyconﬂictingdependingontheclimaticconditionsofthe buildingsite. Thus asteppedapproach maynot leadto themost optimaldesignregardingalltheaforementionedobjectives.

2.3.Theadaptivethermalcomfortapproach

NicolandHumphreys[20] indicatedthat theadaptivecomfort model,whichisrelyingontheoccupants’tendencytoadapttothe building’soutdoorconditions,allowsthedesignertocalculate ther-malcomfortlevelsespeciallyinnaturallyventilatedbuildings.The authorsalsosuggestedtheuseoftheadaptivemodelfor mechani-callycooledorheatedbuildings,withtheaimofcalculatinga vari-ableset-pointforthemechanicalsystemsthatisrelatedtooutdoor temperature.Formechanicallyconditionedbuildings,thePMV/PPD modelisgenerallyused[20,25].

AccordingtoTeleghanietal.[25],theadaptivethermalcomfort levelsforanaturallyventilatedbuildingintheclimateofGreece, canbe estimatedbybothASHRAE 55andEN15251standards, al-though discrepancies, due to differences between the standards, mayoccur.

FortheclimateofAthens,Greece,itwasestimatedthatthe out-doorconditionswouldallowtheinvestigatedbuildingtobe natu-rallyventilatedfora largepartoftheyearandthus theadaptive approachwasselectedforthisresearch.

3. Methodology

Themethodologicalschemethatshowsthestepsofthisstudy is illustrated in Fig. 1. As a ﬁrst step, the ﬂoor plan shape fol-lowedbybuildingorientationoptimizationwasimplemented. De-sign Builder (Version4.7.0.027) was used for a small number of energydemandsimulationsthatservedalsoasbenchmarkforthe resultsofthefollowingoptimizationsthroughEnergy-Pluscoupled withmode-FRONTIER.

Forthe implementation ofthe integratedstrategy, two multi-objectiveoptimizationroundsofdesignandconstruction parame-tersthat could haveconﬂicting impact oncooling, lighting, heat-ing energy loads, energy production from PV panels and adap-tivethermalcomfortlevelswererun.Foreachoptimizationround, 1000differenthigh-risedesigns/constructionsweresimulated. Im-plementingtheintegratedstrategy intotwoseparate optimization roundsthatexaminedifferentvariableshelpedtoreducethe over-all optimization time by reducing the possible combinations of parameter values. Energy simulations were run with EnergyPlus through McNeel Rhinoceros/ Grasshopper software via the plug-ins Honeybee and Ladybug [23]. Daylight simulations were run through McNeel Rhinoceros/ Grasshopper software with Daysim via Honeybee. The optimization was driven by modeFRONTIER with the genetic algorithm NSGA-II (Non-dominated Sorting Ge-netic Algorithm) that has been widely used in studies of build-ingdesignoptimization[16].Theobjectivesweretominimize en-ergydemand,byminimizingcooling,heatingandartiﬁciallighting loads,tomaximizeenergygenerationfromPVpanels,whileatthe sametimetomaximizeadaptivethermalcomfortlevels,foran an-nualperiod.

Theparametersoptimizedfortheﬁrstoptimizationroundwere thewindow towall ratio,thewall U-value,the glazing construc-tion U-value,the glazing g-value, theair-tightness ofthe facade,

(6)

Fig. 1. Methodological scheme of the research.

the cooling set-point of the mechanical cooling system and the PVfacade surfacearea.For thisﬁrstoptimizationround, one op-timizationwas implemented without energy generation from PV panelsand one includingenergy generation, since thePV façade areaisinvestigatedthroughthewindowtowallratiovariable.For thesecond optimizationround,the parametersof thewindowto wallratio,shadingareaandPVsurfaceareaadaptedtoeachfacade orientation(North,South,West,East)wereoptimized.

Data analysis through chartsforthe various energyloads and adaptivethermalcomfortlevelsforthe1000buildingdesignswas implemented.Analyzingthegraphsandimplementingasensitivity analysisindicated theimpactofthevarious facadeparameterson theﬁnalenergyandadaptivethermalcomfortperformanceofthe building.

3.1.Referencebuilding

ThetowerofPiraeusisselectedasadesignreferenceand start-ingpointfortheoptimization.Thisbuildingisselectedbecauseit isatypicalcentral core,open plan,high-rise, officebuildingwith usableopen-officespaceintheperipheralfloorplanandrepeating floorplans(Fig. 2). Thisdesign starting-pointisrepresentative of alargerspectrum ofbuildingsthat belong totheaforementioned typology,whichislargelyappliedonhigh-riseofficebuildings. Lo-cated in the port of Piraeus in Athens, Greece, it is a 22-storey building of 84 m height and 45.42 × 27.26 m rectangular floor plan.Thefacadeismadeofsteelandglass.The bearingstructure ismadeoutofsteelreinforcedconcrete.

3.2.Climatedata

The building islocated atthe port ofPiraeus inAthens. With a hot-summer Mediterranean climate (Köppen–Geiger Csa), the dominant feature of Athens’ climate is alternation between pro-longed hot and dry summers and mild to cool winters with

moderate rainfall (414.1 mm yearly precipitation on average) [7]. In winter, temperatures by day reach 14.2 °C on average. At night the temperature falls to 7.7 °C. Spring temperatures reach 19.7°Cduringtheday.Duringsummertemperaturesvarybetween 21.8 °C and 30.5 °C. Highest mean direct normal radiation lev-els are recorded in the summer months, ranging from 6000 to 7000Wh/m2_per_day._Hourly_weather_data_for_Athens_from_the re-port“GRC_ATHENS_IWEC”[5]areusedforall thesimulated mod-els.

3.3. Simulatedbuildingmodels

For this research two model types were created, one for the Design Builder software and one for the McNeel Rhinoceros/ Grasshopper software. Transitioning from the case study design to thesimulation modelinDesignBuilder, severalsimplifications neededtobe implemented forreducing simulationtime. The ex-istingbuildinghas2cores,forthesimulation1closedcoreareais simulated.Themodelconsistsof31floorsof3.26mheight(Fig.3). The input data in Table 1 are used to run the simulations and are based on the current national standards of Greece for office buildings[18].Thesimulationperiodisannual.Internal floorsand partitionsofthecoreare adiabatic.Thesimulatedmodelsreferto buildingswithamixedmodeventilationsystemthatuseboth me-chanicalandnaturalventilation.

The Grasshopper model consists of5 different zones (Fig. 4): the core, and 4 zones. The longest zones have South and North orientationwhiletheshortesthaveWestandEastorientation.The dimensionsofthefloorplan(Fig.4)arethesameasthe rectangu-larDesignBuildermodel(Fig.3). Theinput dataforthe first op-timizationround inTable2rely onthenational Greekregulation foroffices [18]. Theresults ofthevariables optimizedin thefirst roundserveasinputforthesecond round.Thesimulationperiod isannual.Energy-PluswasusedviaHoneybee[23]forenergy per-formanceandenergygenerationsimulations.Daysimwasusedvia

(7)

Fig. 2. Case study building plans, from upper left corner clockwise: section, elevation and typical ﬂoor plan. Source: Source greekarchitects.gr/competition2010, 2010 [12] .

Table 1

Input details of building model for Design Builder.

Building parameter Unit Value

Occupancy density persons/m ² 0,1 [18]

Computers load W/m ² 15 [18]

Heating set-point °C 20 [18]

Cooling set-point °C 26 [18]

Natural ventilation set-point °C 24

Minimum fresh air m ³/h/person 30 [18]

Air-tightness ac/h 0,2 [18]

External wall U-value W/m ²K 0,35 [18]

External wall costrution _ (out to in) 100 mm brick/79.5 mm extruded polysterene/100 mm concrete/13 mm gypsum

Roof U-value W/m ²K 0,25 [18]

Roof construction _ 10 mm asphalt/144.5 mm glass wool/200 mm air gap/13 mm plaster

Floor construction _ 100 mm cast concrete (dense)

Ground ﬂoor U-value W/m ²K 0,25 [18]

Internal partition U-value W/m ²K 1,923

Window-to-wall ratio % 30

glazing type _ double glazing 6 mm/13 mm/6mm

Glazing U-Value W/m ²K 1,499 [18]

SHGC _ 0,564

Normalized power density for artiﬁcial lighting W/m ²−100lux 3,2 [18]

Heat recovery _ on

(8)

Fig. 3. Design Builder models and ﬂoor-plan dimensions.

Fig. 4. Grasshopper model.

Honeybee for annual daylight simulations. For the calculation of adaptivethermalcomfortlevels,theadaptivemodelasintegrated in Ladybug was used [23]. For the ﬁrst optimization round, the ASHRAE552013wasused asthe adaptivethermalcomfort

stan-dard,whereastheEN15251standardwasusedforthesecond op-timizationround.Themodelsrefertoachange-over,mixed-mode ventilationsystem, whenoneofthetwo options(natural ventila-tionormechanicalventilation)isactiveataspeciﬁctime.

3.4. modeFRONTIERplatform

In orderto implement a multi-objectiveoptimization and ex-tractdatafromalargenumberofsimulateddesigns,an optimiza-tion platformwasused. Theoptimization wasimplementedwith modeFRONTIER. ModeFRONTIER is an optimization platform that can drive an optimizationloop usingdifferent applications.It al-lowstoperformmulti-objectiveoptimizationsandgivestheoption toselectdifferentoptimizationalgorithms[17].Rhino/Grasshopper andmodeFRONTIERwereconnectedusingalinkdevelopedby ES-TECOwithTUDelft[30].Thenumberofsimulateddesignsforboth optimization rounds is 1000 each. The overall number of possi-bledesignsisrelatedtothenumberofvaluesofthevariables.All possiblecombinations ofdesigns ifsamplingwere appliedwould be 1.944 different designs for the ﬁrst optimization round and 160.000 for the second round. But, since an optimization is ap-plied,theoptimalsolutionisfoundearlierintheprocess.Thetime neededfortheevaluationof1000designs was44h and24min. The evaluation time of 1 design was approximately 2 40’. The optimizationrun oni7-5820 K,CPU at3,3GHz. Theworkﬂow of

(9)

Ta b le 2 In put de ta ils of building model fo r Ener gyPlus/Hone y b ee/Ladybug. Building par a me te r Un it Value fo r op timization Ro u n d 1 Value fo r op timization Ro u n d 2 Occupancy density persons/m ² 0,1 [18] 0,1 [18] Comput ers load W/m ² 15 [18] 15 [18] Heating se t-point °C 20 [18] 20 [18] Cooling se t-point °C (v ariable) 26 Min indoor t e m per atur e fo r natur a l v e ntilation °C 21 21 Max indoor t e m per atur e fo r natur a l v e ntilation °C 23 25 Minimum fr esh air m ³/h/person 30 [18] 30 [18] Air -tightness ac/h (v ariable) 0,1 Ext e rnal wa ll U-v alue W/m ²K (v ariable) 0,1 Floor cons truction _ acous tic tile/50 mm insulation boar d/200 mm hea vyw eight concr e te acous tic tile/50 mm insulation boar d/200 mm hea vyw eight concr e te Floor U-v alue W/m ²K 0,4726 0,4726 Int e rnal partition U-v alue W/m ²K 2,58 2,58 Windo w -t o-w a ll ra ti o % (v ariable) (v ariable per fa ca d e orient ation) Shading ar ea % 50 (v ariable per fa ca d e orient ation) Glazing U-Value W/m ²K (v ariable) 1,8 SHGC _ (v ariable) 0,3 N o rmalize d po w er density fo r artiﬁcial lighting W/m ²− 100lux 3,2 [18] 3,2 [18] Heat re co ve ry _ on on HVA C sy st em _ “Ideal Loads” “Ideal Loads” PV sy st em eﬃciency % 12 [18] 12 [18]

the optimization inthe interface of modeFRONTIER (Fig. 5) con-sistsoftheinputs,theoutputs,theconnectiontoGrasshopper,the DesignofExperimentscomponentandtheoptimizationalgorithm component.Analgorithmwidelyusedforenergyoptimizationsof buildingsisNSGA-II (non-dominated sortating geneticalgorithm). Inthis studyalsothis algorithmwas used.Uniform Latin Hyper-cubeisusedasspaceﬁllerwith25numbersofinitialdesignsand aRandomGeneratorSeedvalueof1.

4. Resultsanddiscussion

Thefollowing results referto shape andorientation optimiza-tions (Sections 4.1and 4.2). Additionally, theresults of the2 in-tegratedoptimizationprocessesreﬂecttheeffectofdifferent vari-ablesonannualenergydemand,energyproductionfromPVpanels andadaptivethermalcomfort levelsof thebuilding(Sections4.3

and4.4). The results on annualenergy demand refer to cooling, heating,artiﬁciallightingandcomputerequipmentloads.

4.1. Shapeoptimization

Thefirststepistheoptimizationofthefloor-planshape.Aimis todefinewhichshapeleadstoamorereducedenergydemandof ahigh-riseopenplanofficebuilding,withmixedmodeventilation systeminAthens,Greece.Thisisrealizedbykeepingthefloor-plan area constrained, while gradually changing the shape from more compacttomoreelongatedrectangular.Theconstraintsarethe to-talfloorplanarea(1034m2₎_and_the_area_of_the_service_cores (ap-proximately 21%of the total area forthe rectangular shapesand 23.4%forthecompactshapes).Thefloorplandepthforthe rectan-gular,squareandoctagonshapesisapproximately8.2m,whereas fortheelongatedrectangular,thefloorplandepthisminimizedto 6.3m, since the core should have realistic dimensions to fit the elevators,stairsandWC(Fig.3).

Figs.6–8 illustratethat coolingloadsandlightingloadsshare thebiggestpartofenergyconsumption,whereasheatingloadsare nearlyzeroduetotheclimaticcharacteristicsofAthens.For reduc-ing cooling loads, compact buildingsare favorable. In the graphs (Fig.6)forcoolingloadsrankingandsolargainsranking,astrong correlation is visible between the increase in cooling loads with theincreaseofsolargains.Morecompactshapeswithminimized facadesurfaceleadtominimizedsolargainsandthereforeto mini-mizedcoolingloads.Furthermore,thebuildingsareactivelycooled. So,theindoorairtemperatureisoftenbelowtheoutdoor temper-ature. The more compact the shape, the lower the heat transfer andthus thelower the coolingload. Lightingloads are relatively diminishedinthecaseoftheelongatedrectangularshapesinceit hasa reducedﬂoorplandepth andthus daylighthasaccessto a largerﬂoorareathantheotheroptions(Fig.8).

Thesquare building(Fig. 9) ismarginally the bestperforming buildingwith a value of2029 MWh/a andthe worst performing buildingistheonewiththerectangularlayoutwithapproximately 2088MWh/a.Itisevidentthattheeffectoftheaspectratioofthe layout ofa high-rise open plan oﬃce building,that hasa mixed mode ventilation system and within the speciﬁc climatic condi-tionsofAthens,isnegligible,thereforeadesignercouldbefreeto explorevariousoptions.

4.2.Orientationoptimization

Forcompactﬂoorplanschangingtheorientationwillhavelittle effectontheenergydemand.Alsotheelongatedrectangularshape has a shorter ﬂoor-plan than many reference high-rise buildings liketheCommerzbankhigh-riseinFrankfurt.Therefore,forthe ori-entationoptimization,the rectangularshapewaschosen. Forthis

(10)

Fig. 5. Workﬂow in modeFrontier interface for 1st optimization round.

Fig. 6. Annual solar gain and cooling loads rankings for different plan shapes.

step 4 different orientations were examined (North–South, East– WestandNorthwest–Southeast,Southwest–Northeast).

Thesolargainschart(Fig.10)indicatesthatgreaterexposureof thefacadetowardstheeastandwestdirectionsleadstoincreased solargains,sincetheincidenceangleissmall,thusthesolar radi-ationpenetratesthewholeﬂoorplan.Ontheotherhandexposure ofthe longside of thefacade towards the southleadsto

dimin-Fig. 7. Annual heating loads ranking for different plan shapes.

Fig. 8. Annual lighting loads ranking for different plan shapes.

ishedsolargainsduetothefactthat thesteepincidenceangleof thesolarradiation limitstheradiationfromreaching deepinthe ﬂoor-plan.

Thisstep(Fig.11)showsthattheorientationofabuildingwhen therearenosurroundinghigh-risebuildings,onlyaffectstoa mi-nor degree the energydemand within the climaticconditions of

(11)

Fig. 9. Annual total energy demand ranking for different plan shapes.

Fig. 10. Annual solar gain and cooling loads rankings for different orientations.

Fig. 11. Annual total energy demand ranking for different orientations.

Athens.Thus adesignerisnothighlyrestrictedtoadoptacertain orientation. Nevertheless, these optimizations were implemented on ahigh-risebuilding withnosurroundingbuildings. Surround-ings with high-rise buildings at close proximity or different cli-maticconditionscouldpossiblyleadtodifferentconclusions.

4.3. Integratedenvelope,HVACandenergygenerationmulti-objective optimization

At this stage, a multi-objective optimization is implemented. Apart from reducedenergy demand, also energy production and indoor thermal comfortof the occupantsplay an important role. The building envelope is the boundary between the indoor and outdoor environment. Therefore seven parameters were selected, ﬁve of which referring to the building envelope. The parameters examined were the window to wall ratio, the wall U-value, the glazingconstructionU-value,theglazingg-value,theair-tightness

of the facade, the cooling set-point of the HVAC systemand PV surfaceareainthefacade.

Toillustratethetrendsoftheimpactofeachparameter,within the spectrum of values tested, scatter plots are created.The fol-lowing plots illustrate various building designs depicted asdots, coloredaccording tothe value of thevariable withwhich itwas simulated.Eachdesignasdot islocated relativelyto thexandy axisaccordingtotheir simulatedperformanceon energydemand orﬁnalenergyandlevelsofthermalcomfortrespectively.The op-timalbuildingdesignsarelocatedontheParetofront[26](Fig.12) ofeachplot,amongwhichadesignercanoptforthemostenergy eﬃcientbuilding,themostthermallycomfortabledesignora de-signwithatrade-off performancebetweenthetwoobjectives.

4.3.1. Windowtowallratio(wwr)

Forthis variable, 6 different values were researched between 30%and80%.A30%wwr isexpectedtoreducecooling loads,but increaseelectriclightingloads,whereasan80%windowtowall ra-tioisexpectedtoincreasecoolingloadsanddecreaseelectric light-ingloads,sinceitreferstoalmostafullyglazedfacadethatallows moredaylightinthebuilding.Theglazingratioisthesameforall directionsforthisﬁrstoptimizationround.Thechart(Fig.12) illus-tratestheeffectofwwrontheenergydemandandcomfortlevels. Smallwindowshaveapositiveeffectonreducingenergydemand, butalsoincreasingcomfortlevels inabuilding,sincethey leadto reducedsolarheatgainsandthusreducedcoolingloads.

4.3.2. WallU-value

Forthe external wall U-value of the building,3 different val-ueswere simulated:0.1, 0.2and0.3W/m2_K._The_0.1_value _refers to well insulated buildings andthe 0.3 value to a less insulated building.Itisimportanttomentionthat0.3W/m2_K_is_even_better than thecurrent national standardsof Greecethat allow a value of0.5 W/m2_K _for_this _climatic_zone. _Reducing _the _wall _U_values doesnotdrasticallyimprovetheenergyuseorcomfortlevelsofa building(Figs. 15–17). Thereasonisthat thebuildingisnaturally ventilatedasaresultofwhichthethermalinsulationisonly effec-tiveduringthehourswhenthewindowsarenotopen.

4.3.3. Glazing/frameU-value

ThisU-valuereferstotheglazingandframeconstructionofthe openings.Fortheopenings’U-valueofthebuilding,3different val-uesweresimulated:0.6,1.2and1.8W/m2_K._The_0.6_W/m2_/K_refers totriple glazing,1.2W/m2_/K_to _high_performance _double_glazing and1.8W/m2_/K_to_double_glazing_with_a_worse_performance. _The buildingsimulatedisnaturallyventilatedandtheinsulationofthe openingsisusefulforthetime-span thatthewindowsareclosed. InFig.17,theweaknegativecorrelationindicatesthathigher glaz-ingU-valuesleadtoreductionofenergydemand(r=−0.376)and increaseofthermalcomfort(r=−0.213).Thiscouldbeinterpreted as the need for the building to have improved natural ventila-tion like night cooling. The simulated model has concrete ﬂoors withhighthermalmass.The heatis accumulatedduringtheday inthe concrete mass andis givenoff at night. Withlow glazing Uvalue, theheatistrappedwithinthebuildingwithclosed win-dows.Inthiscase,constructionswithaworst performingUvalue of1.8W/m2_/K _might_drive _the _heat _outside_the _building_quicker thanwellinsulatingconstructions.

4.3.4. Glazingg-value

Fortheg-value orsolarheat gain coeﬃcient (SHGC), 3 differ-entvalueswere explored:0.3, 0.55and0.8. The0.3 valueallows theleastamountofsolarheatgainsinthebuilding,whereasa g-value of0.8allows the mostamount ofsolar heatgains. Forthe climaticconditionsofAthens,withintensesolarradiation,smaller solarheatgainsareexpectedtoreducethecoolingloadsandthus

(12)

Fig. 12. Window to wall ratio effect regarding the objectives.

Fig. 13. Glazing g-value effect on the objectives.

reduce thetotal energydemand ofthe building.In Fig.13 lower g-valuestendtoreducetheenergyusageandimprovecomfort in-sidethebuilding.Lowerg-valuesmeanthat thebuildinghasless solarheatgainsthroughglazingandthereforelowercoolingloads.

4.3.5. Air-tightness

Air-tightnessofthefacade,orinfiltration,allowsforheat trans-ferbetweenindoorsandoutdoorsthroughtheenvelopeinan un-controlledway.Therateoftheinfiltrationwasinvestigatedinthis researchwithvaluesteps:0.1,0.3,0.5and0.7airchangesperhour. In Fig. 17, weak negative correlations show that low infiltration rates occur for designs with higher comfort levels (r = −0.387) andlow energy use(r₌ _−0.293),since airinfiltration is an un-controlledtypeofthermal/ventilationbridge.

4.3.6. Coolingset-point

Regarding the building services,this studyfocused foremostly onthe energydemandthus not includingthe effectsof different typesofsystems.Theonlybuildingservicesrelatedparameter in-vestigatedwasthecoolingset-point.Threevalueswereexamined:

24°C,26°Cand28°C.Thecooling set-pointrefersto theindoor airoroperativetemperatureabove whichthe mechanicalcooling systemwillstartworking.Avalueof24°Cisexpectedtoincrease the comfort levels of the building, but also increase the energy consumption.InFig.14,thecoolingset-pointseemstohavea dras-ticeffectonthebuilding’senergyuseandcomfortlevels.Cooling set-pointof28°Creducestheenergyusedrastically,butalsohas a negative effect on comfort levels. A cooling set-point of 26 °C seemstohaveabalancingeffectbetweencomfortandenergy us-age.

4.3.7. PVfacadesurfacearea

Thisoptimizationvariable islinked tothe windowtowall ra-tio. It refers to PV panelsmounted verticallyon all 4 facadesof the building where no transparent windows exist. It is obvious that forlower glazing ratios of 30%, higher amounts of electric-itywillbegeneratedandreversely,forhighwindowtowallratios of70%and80%, theleastamountof electricitywillbe produced. Nevertheless, for glazing ratios of 30%, the daylight that inﬁl-trates the building will be reduced compared to an 80% wwr.

(13)

Fig. 14. Cooling set-point effect regarding the objectives.

Fig. 15. Effect of variables on minimizing energy demand.

Fig. 16. Effect of variables on maximizing comfort.

Fig.18showshowthe presenceofenergygenerationsystems re-ducestheannualﬁnalenergyofthebuildingby17.5%.Small win-dows (30% wwr)in Fig. 19 havea positive effect on maximizing energyproduction,butalsoincreasingcomfortlevelsinabuilding.

4.3.8. Integrationofenvelope-HVAC-PVparameters

The correlationchart inFig.17indicatesthat theobjectivesof minimizing energydemand andmaximizing comfortare strongly correlatedwiththecoolingset-pointvariable.Thehigherthe cool-ing set-pointtemperature, thelower theenergyuse(r=−0.918)

andthelowerthethermalcomfort(r=−0.902).Inthesensitivity analysischart(Fig.15)createdwithmodeFRONTIER,itisapparent thatthecoolingset-pointhasthehighestinfluenceonminimizing theenergydemand,followedbyg-valueandwindowtowallratio. Regardingthewindowtowallratio,notethatthesensitivity anal-ysis(Figs.15–17),doesnotincludethePVareaoptimization. Nev-ertheless,it isvisible fromFig.18 that the energygeneration on thefacadewouldincreaseevenmorethealreadypositiveeffectof smallerwindowsinminimizingfinalenergy.GlazingU-value, infil-trationrateandwallU-valueseemtohaveminimizedinfluenceon

(14)

Fig. 17. Pearson correlation chart.

Fig. 18. Comparison of designs with energy generation optimization and without energy generation.

(15)

Fig. 20. East facade window to wall ratio effect regarding the objectives.

Fig. 21. West facade window to wall ratio effect regarding the objectives.

thisobjective.Inthesensitivityanalysischart(Fig.16)referringto theobjective ofmaximizingcomfort, coolingset-point isalsothe mostinﬂuential factorforcomfort, followed by g-valueand win-dowtowallratio.

4.4. Integratedenvelopeandenergygenerationmulti-objective optimization

To proceed to this optimization round, input data (Table 2) were used from the optimal design chosen (Fig. 18) in the previous optimization round. This building with final energy at 81.67 kWh/(m2_a)_is _chosen _as _the _optimal _{trade-off solution} be-tween minimizing final energyandmaximizing adaptivethermal comfort. This design’s parameters refer to: wwr ₌ 30%/ wall U-value = 0.1 W/m2_K _/ _glazing_U-value ₌ _1.8 _W/m2_K _/ _glazing g-value=0.3/infiltration=0.1ach/coolingsetpoint=26°C.

From the previous optimizationit becameclear that the win-dow to wall ratio and solar control were the most impor-tant facade-relatedvariablesto consider.The secondoptimization

thereforeexaminedthewindowtowallratiofor4different orien-tations(North,South,East,West),andshadedareaoftheopenings for4differentorientations(North,South,East,West).Thesedesign aspectsaredirectlyrelatedtotheamountofsolarheatgainsinthe building.

4.4.1. Windowtowallratio(wwr)

Forthisvariable,5different valueswere investigatedforeach of the 4 facade orientations between 20% and 60% window to wall ratios. A 20% wwr is expected to reduce cooling loads, but increase electric lighting loads, whereas a 60% wwr refers to an almost fully glazed facade and is expected to have the opposite effect.Figs.20–22illustratetheeffectofwwroftheeast,westand north facade on the final energy and comfort levels. Small win-dows (20% glazed area) have a positive effect on reducing final energyand increasing comfortlevels. They lead toreduced solar heatgainsandthusreducedcooling loads.Additionally, minimiz-ingfinalenergyisaidedwithincreasedenergyproductionfromPV panels.Although smaller windows mean increased electric light-ingloadsduetoreduceddaylightexploitation,forthespecific

(16)

cli-14 E.D. Giouri, M. Tenpierik and M. Turrin / Energy & Buildings 209 (2020) 109666

Fig. 22. North facade window to wall ratio effect regarding the objectives.

Fig. 23. South facade window to wall ratio effect regarding the objectives.

mateofAthens,thisnegativeeffectiscompensatedbythe afore-mentionedsolar heat gainsreduction and energygeneration. For thesouthfacade (Fig. 23) on thefinal energyandcomfortlevels, largerwindows (60%wwr) seem to lead to marginally improved comfortlevels than smaller windows (20% wwr). Solar radiation fromtheSouthentersthebuildingalmostverticallyanddoesnot reachdeepintothefloor-plan,sothewindowareaisnotso impor-tantfor thesolar heatgains. Furthermore,biggerwindows mean increasedflow ofnaturalventilationanddaylightadmission. This leadstoincreasedindoor thermalcomfortlevels andreduced de-mandforelectriclighting(Fig.24).

4.4.2. Shadingarea

For thisvariable,4differentinput values wereset foreach of the4facade orientations: 25%, 40%,55%, 70%shaded area of the openings.This investigation refers onlyto external shadings. 25% shaded glazedarea is expectedto increase cooling loads,but re-duceelectriclightingloads.A70% shadedglazedareaisexpected

to decreasecooling loadsand increase electriclightingloads.For the southorientation (Fig.25), shadingarea of 25%is marginally betterforenergydemandandthermalcomfortlevelsofthe build-ing.Fortheotherorientations,theoptimizationoftheshadingarea hasmuch lessinﬂuence ontheobjectives (Figs. 26–28) giventhe fact thatthe simulatedmodels referto abuilding withag-value of0.3thatblocksalargepartofsolarheatgainsfromenteringthe building,thushavingasimilareffecttoshading.Forthis optimiza-tion,visual comfort isnot takeninto account,so theconclusions derivedfromthisoptimizationareonlyreferring tothermal com-fortandﬁnalenergyobjectives.

4.4.3. Energygenerationfromphotovoltaicpanels

Thisvariableislinkedto thewindowtowallratio.Itrefers to PVpanelsmountedverticallyonthefacades,onthosepartswhere there are no windows. Forlower window to wall ratios of 20%, higheramounts ofelectricity willbe generated andreversely,for highwindow to wall ratiosof 50% and60%, theleast amountof

(17)

Fig. 24. South facade window to wall ratio effect on cooling and artiﬁcial lighting loads.

Fig. 25. South facade shading area effect regarding the objectives.

electricity will be produced. Nevertheless, glazing ratios of 20%, donot exploit daylightaswellasa 60%window towall ratio.In

Fig. 24, optimal daylight exploitation is achievedwith 60% wwr, thatreducestheartiﬁciallightingloadsbutthereverseistrue re-gardingthecoolingloadsthatarealsopredominantforthe inves-tigatedclimaticzone.

4.4.4. Integrationofenvelope-PVparameters

The correlation chart (Fig. 28) illustrates the positive correla-tion (r = 0.769) ofthe objectiveofminimizing ﬁnal energy with the southwindowto wallratio.The southfacade receiveshigher solarheatgainssosmallerwindowshaveagreatinﬂuenceon

min-imizingcoolingloadsandalsoproducingmoreenergythroughPV panels.Theobjectiveofmaximizing comfortisstronglycorrelated withthe north window to wall ratiowith a negativecorrelation (r = −0.791). As the window to wall ratio is reduced, the com-fort levels are increased. Since in the north, solar heat gains are reduced, morewall area helpsretain the existing indoor thermal comfortthroughmoreinsulationandsmalleropeningsthatresult toreducednaturalventilationrate,thusreducedheattransfer.The charts(Figs.26and27)createdwiththesensitivityanalysistoolof modeFRONTIERshow that thewindow towall ratioofallthe fa-cadeorientationsaremoreinﬂuentialthantheshadingarea opti-mizationbothfortheﬁnalenergyandthermalcomfortobjectives.

(18)

Fig. 26. Effect of variables on minimizing energy demand.

Fig. 27. Effect of variables on maximizing comfort.

Fig. 28. Pearson correlation chart.

5. Conclusion

By applyingthe integratedstrategy for theMediterranean cli-mateofAthens,it waspossibletoassessthe effectsofthe enve-lope,HVAC andenergy generation parameters on a central core, high-rise oﬃce building. Energy simulations were driven by an optimization strategy and derived data were analysed by sensi-tivity analysis that indicated the parameters with higher impact onannual energydemand, energy productionand adaptive ther-malcomfortlevels.Theseparametersare:cooling set-point,

natu-ralventilationstrategies,glazingg-value,window-to-wallratioand energyproductionwithPVpanels.Byapplyingtheproposed inte-grated strategy, the building’s energyperformance is reduced by 33% (from 109.12 kWh/(m2_a) _to _73.13 _kWh/(m2_a)) _and _the com-fort hours are increased by 18.2% (from 78.3% to 96.5%), from the starting point of the current regulations in Greece [18]. The startingbuilding refers to: window to wall ratio = 40%/ wall U-value = 0.5 W/m2_K _/ _glazing _U-value ₌ ₂ _W/m2_K _/ _glazing g-value = 0.5/infiltration = 0.2ach/cooling-set-point = 26°C. The finaloptimaldesigndepictedinFig.25,hasanannualfinalenergy

(19)

of73.13 kWh/(m2_a) _and_is_comfortable_for_the _96.49%_of _time_in anannualperiod,whentheoﬃcespacesareoccupied.

Anextensivenumberofsimulationsdriven byanoptimization algorithmwasneededinordertoinvestigatetheimpactofvarious design variablesonﬁnal energyandthermalcomfort. Theresults ofthisstudyindicatethefollowing:

• The measures that need to be taken in the early-phase of designing a nZEB central core, open plan, oﬃce building in a Csa (Köppen–Geiger) Mediterranean climate and derive from the optimal building design of this research are: win-dow to wall ratio = 20% for all facade orientations/ shading area of openings ₌ 25% for all facade orientations/ wall U-value = 0.1 W/m2_K/glazing _U-value ₌ _1.8 _W/m2_K/glazing g-value=0.3/inﬁltration=0.1ach/cooling-set-point=26°C. • The parameters with the highest impact on the objectives

of this research are the cooling set-point, natural ventilation strategies,theglazinggvalue,thewindow-to-wallratioand en-ergyproductionwithPVpanelsonthefacadesofthebuilding. • Theparameterswithlowerimpactontheobjectivesofthis re-searcharethewallU-value,theglazingU-value,theinﬁltration rate,shadingsystemsoftheopenings,ﬂoorplanshapeand ori-entationofthebuilding.

• FortheclimaticconditionsofAthens,adaptivedesignofthe fa-cade openings per orientation and adaptive shading area per orientation will notlead to signiﬁcantly reducedenergyloads inthepresenceofsmalleropeningsandenergyproduction sys-temsonthefacade.

• The presence of active systems has inﬂuenced passive design optimizations.

• The integrated optimization of window to wall ratio, energy generationonthefacadesandshadingareahasovershadowed theeffectsofadaptiveshadingareaperorientation.

• For cooling dominant climates with outdoor temperatures withintherangeofindoorcomfortforalargepartoftheyear, adopting natural ventilation strategies in combination with BMS (building management systems) has a high potential to-wardsdesigningazeroenergyhigh-risebuilding.

• Generating electricity fromPVpanels on the facadesof high-risebuildingscanalsogreatlyreducetheirenergyconsumption inclimatessimilartoAthens,Greece.

• Theproposedintegratedstrategyofconﬂictingpassiveand ac-tive systemsanditscomputational set-upare genericandcan beappliedtovariousbuildingtypologiesandclimaticzones. • The proposed integrated strategy that is driven by

algo-rithm aidedmulti-objectiveoptimizations,contraryto existing steppedstrategies,canenablethedesignertoattainsubstantial information,in a time-eﬃcient manner, through dataanalysis of large numberof different buildingdesigns. These data can helpthedesignertocomprehendtheimpactofconﬂicting de-signparametersregardingtheenergyandthermalcomfort per-formanceofabuilding.

Thederiveddesignmeasuresstemfromabuildingwithno sur-rounding buildings.Takingintoconsiderationadenseurban envi-ronmentcouldleavemoreroomforadaptivedesign,butwould de-mandmoreaccurateandtimeconsumingmodelsandsimulations. Simulation timecould alsodrastically increaseby includingmore detailed systems and parameters in the optimization procedure. Thisstudyfocusedonexaminingparametersrelatedtoearly deci-sionmakingphaseofdesigningahigh-riseofficebuilding.Possible nextstepscouldexploittheresultsfromthisdecisionmakingstep andperformfurtheroptimizations onmoredetailedandadaptive design of theparameters with higherpotential for improvement. Thisgradualoptimizationapproachwouldbebeneficialfortaking informeddesigndecisionsinatime-efficientmanner,towardsthe

transitionfromearly decision-makingconcepts todetailed execu-tion.

Authordeclaration

Wewishtoconfirmthat therearenoknownconflictsof inter-estassociatedwiththispublicationandtherehasbeenno signifi-cantfinancialsupportforthisworkthatcould haveinfluencedits outcome.

Weconfirmthatthemanuscripthasbeenreadandapprovedby allnamedauthorsandthattherearenoother personswho satis-fiedthecriteriaforauthorshipbutarenotlisted.Wefurther con-firm that the orderof authors listed inthe manuscripthas been approvedbyallofus.

Weconﬁrm that we havegivendueconsideration to the pro-tectionofintellectualpropertyassociatedwiththisworkandthat thereare no impediments topublication, includingthe timing of publication, withrespect to intellectual property. Inso doing we conﬁrmthat we havefollowedthe regulationsofourinstitutions concerningintellectualproperty.

WeunderstandthattheCorrespondingAuthoristhesole con-tactfortheEditorialprocess(including EditorialManagerand di-rect communications with the office). He/she is responsible for communicating with the other authors about progress, submis-sionsofrevisionsandfinalapprovalofproofs.Weconfirmthatwe have provided a current, correctemail address which is accessi-bleby the CorrespondingAuthor andwhich hasbeen configured toacceptemailfromEnergyandBuildings.

References

[1] Active House Alliance, in: ACTIVE HOUSE - the guidelines, Active House, Brux- elles, 2015, pp. 8–62 .

[2] U. Berardi , A cross-country comparison of the building energy consumptions and their trends, Resour. Conserv. Recycl. 123 (2017) 230–241 .

[3] L. Caldas , Generation of energy-eﬃcient architecture solutions applying GENE_ARCH: an evolution-based generative design system, Adv. Eng. Inform. 22 (1) (2008) 59–70 .

[4] F. Chantrelle , H. Lahmidi , W. Keilholz , M. Mankibi , P. Michel , Development of a multicriteria tool for optimizing the renovation of buildings, Appl. Energy 88 (4) (2011) 1386–1394 .

[5] Energyplus.net. (2017). Weather Data by Location, EnergyPlus. [Online] Avail- able at: https://energyplus.net/weather

[6] R. Evins , P. Pointer , S. Burgess , Multi-objective optimisation of a modular building for different climate types, in: Proceedings of the First Building Simulation and Optimization Conference, UK, 2012, pp. 173–180 .

[7] D. Founda , Evolution of the air temperature in Athens and evidence of climatic change: a review, Adv. Build. Energy Res. 5 (1) (2011) 7–41 .

[8] E.D. Giouri , Zero Energy Potential of a High Rise oﬃce Building in a Mediter- ranean Climate M.Sc. thesis., Delft University of Technology, 2017 .

[9] M. Hamdy , A. Hasan , K. Siren , A multi-stage optimization method for cost-optimal and nearly-zero-energy building solutions in line with the EPBD-recast 2010, Energy Build. 56 (2013) 189–203 .

[10] J. Holst , Using whole building simulation models and optimizing procedures to optimize building envelope design with respect to energy consumption and indoor environment, in: Proceedings of the Eighth International IBPSA Confer- ence, the Netherlands, 2003, pp. 507–514 .

[11] T. Konstantinou , Facade Refurbishment Toolbox, Supporting the Design of Res- idential Energy Upgrades Ph. D Thesis., Delft University of Technology, 2014 .

[12] Lianopoulos, M. (2010). [Online] Greekarchitects.gr. Available: http: //greekarchitects.gr/competition2010 [Accessed 2012]

[13] R. Looman , Climate-Responsive Design, A Framework for an Energy Concept Design-Decision Support Tool for Architects Using Principles of Climate-Re- sponsive Design Ph.D. Thesis., Delft University of Technology, 2017 .

[14] R. Loonen , M. Trka , J.L.M. Hensen , Exploring the potential of climate adaptive building shells., in: Proceedings of the 12th International Building Performance Simulation Association Conference, Sydney, 2011, pp. 2148–2155 .

[15] E.H. Lysen , The Trias Energica: solar energy strategies for developing countries, in: Proceedings of the Eurosun Conference, Freiburg, 1996, pp. 1–6. September 16-19 .

[16] V. Machairas , A. Tsangrassoulis , K. Axarli ,Algorithms for optimization of building design: a review, Renew. Sustain. Energy Rev. 31 (2014) 101–112 .

[17] M. Manzan , Genetic optimization of external ﬁxed shading devices, Energy Build. 72 (2014) 431–440 .

[18] Ministry of Environment Energy and Climatic Change, in: Detailed National Standards of Parameters for the Calculation of Energy Building Performance and Issuing of Certiﬁcates of Building Performance, Technical Chamber of Greece, Athens, 2012, pp. 20–143 .

(20)

18 E.D. Giouri, M. Tenpierik and M. Turrin / Energy & Buildings 209 (2020) 109666 [19] K. Negendahl , T. Nielsen , Building energy optimization in the early design

stages: a simpliﬁed method, Energy Build. 105 (2015) 88–99 .

[20] J.F. Nicol , M.A. Humphreys , Adaptive thermal comfort and sustainable thermal standards for buildings, Energy Build. 34 (6) (2002) 563–572 .

[21] R. Pﬂuger , in: Simulation Des Thermischen Gebäudeverhaltens eines Pas- sivhauses in Geschoßwohnungsbau- Typologie und Städtischer Bebauung, PHI Passivhaus-Institut, Darmstadt, 2007, pp. 3–31 .

[22] R. Pﬂuger , W. Feist , S. Ludwig , J. Otte , in: Nutzerhandbuch Für Den Geschoß- wohungsbau in Passivhaus-Standard, German Federal Oﬃce for Building and Regional Planning, 2007, p. 11. [ebook] .

[23] M.S. Roudsari , M. Pak , A. Smith , Ladybug: a parametric environmental plugin for grasshopper to help designers create an environmentally-conscious design, in: Proceedings of the 13th international IBPSA conference, Lyon, France, 2013, pp. 3128–3135. August .

[24] X. Shi , Design optimization of insulation usage and space conditioning load using energy simulation and genetic algorithm, Energy 36 (3) (2011) 1659–1667 .

[25] M. Taleghani , M. Tenpierik , S. Kurvers , A. van den Dobbelsteen , A review into thermal comfort in buildings, Renew. Sustain. Energy Rev. 26 (2013) 201–215 .

[26] T. Tusar , B. Filipic , Visualization of Pareto front approximations in evolutionary multiobjective optimization: a critical review and the prosection method, IEEE Trans. Evol. Comput. 19 (2) (2015) 225–245 .

[27] United Nations, Department of Economic and Social Affairs, in: World Urban- ization Prospects. The 2014 Revision Highlights, United Nations, New York, 2014, pp. 7–10 .

[28] A .A .J.F. Van den Dobbelsteen , Towards closed cycles - New strategy steps in- spired by the Cradle to Cradle approach, in: Proceedings of the 25th Confer- ence on Passive and Low Energy Architecture, Dublin, 2008, pp. 1–6. 22–24 October .

[29] J. Xu , J. Kim , H. Hong , J. Koo , A systematic approach for energy eﬃcient building design factors optimization, Energy Build. 89 (2015) 87–96 .

[30] D. Yang , S. Ren , M. Turrin , S. Sariyildiz , Y. Sun , Multi-disciplinary and multi-objective optimization problem re-formulation in computational design explo- ration: a case of conceptual sports building design, Autom. Constr. 92 (2018) 242–269 .

[31] W. Yu , B. Li , H. Jia , M. Zhang , D. Wang , Application of multi-objective genetic algorithm to optimize energy eﬃciency and thermal comfort in building design, Energy Build. 88 (2015) 135–143 .