supermarkets with high refrigeration loads Coupling night ventilative and active cooling to reduce energy use in Energy & Buildings

Zoi Mylona , Maria Kolokotroni, Savvas A. Tassou

RCUK National Centre for Sustainable Energy Use in Food Chains, Brunel University London, Kingston Lane, Uxbridge UB8 3PH, UK

a rt i c l e i nf o

Article history:

Received 24 January 2018 Revised 11 April 2018 Accepted 14 April 2018 Available online 18 April 2018 Keywords:

Supermarket Energy use HVAC Night ventilation EnergyPlus Frozen food

Environmental and energy monitoring

a b s t ra c t

Supermarketsareenergyintensivebuildingsandpresentauniquespaceconditioningchallengebecause oftheinteractionbetweentheHVAC systemand therefrigerateddisplaycabinets.HVAC systemisthe largest consumerof energyafter refrigeration dependingonsystemdesign, geographicallocation and controls.Nightventilationisusedextensivelyasalowenergystrategytocoolbuildingsinclimateswhere nighttemperaturesaresuitable.Thispaperpresentsastudyofcoolingbenefitsofnightventilationfor supermarketswithhighcoolingdemand.Energyandenvironmentaldatafromtwostoreswithhighper- centageoffrozenandchilledgoodsandwithdifferentHVACsystemsarepresented.Validatedmodelsin EnergyPlusaredevelopedforthetwostoresandtheirsystems.Aparametricstudyofthecoupledoper- ationofnightventilationandactivecoolingfortheclimaticconditionsofsoutheastEnglandiscarried outandoptimisationstrategiesaremodelled.Resultsindicatethateffectivenightventilationcanreduce thedurationofactivecoolingduringtradingtimesandachieve17%reductionincoolingannualenergy use,3.3%intotalannualenergyusewhilerefrigerationenergyuseisnotaffected.

© 2018TheAuthors.PublishedbyElsevierB.V.

ThisisanopenaccessarticleundertheCCBYlicense.(http://creativecommons.org/licenses/by/4.0/)

1. Introduction

Retail storesareamong themostenergy-intensive commercial buildings,consumingtwoorthreetimesasmuchenergyperunit floorareaasofficebuildings.IntheUS,supermarketsrepresent5%

ofthetotalcommercialbuildingprimaryenergyuse[1,2].Accord- ingtoUSEnergy InformationAdministration[3],commercialsec- torenergyuseaccountsfor18%ofthetotalenergyuseandisre- sponsiblefor16% ofthecarbondioxideemissionsofthecountry.

IntheAsian countriesforecastsshow asalesgrowthin theretail sector;the fastestinthe worldpresenting an averageof4.6% in- creasewithsalesof almost 7trillion USD in2014 [4].In theUK energyconsumptioninfoodsupermarketsisaround3.5%oftheto- talUKenergyconsumption[5].Currently,thereareover1million supermarkets in Europe. Thus, just a small percentage reduction inenergyusecanresulttosubstantialsavings.Estimating25%en- ergysavinginEuropeinsupermarketswillresultin31TWhofan- nualelectricitysavingswhichequatestocarbonreductionsof16.2 million tons[6]. The globalretail landscapeis evolvingand new trendssuchasinternetpurchasingandhomedeliveriesalongwith changes in consumers’ lifestyle have an impact on conventional foodretailstores. This tendencyhascreated a shifttowards new

∗ Corresponding author.

E-mail address: zoi.mylona@brunel.ac.uk (Z. Mylona).

relativelysmallconveniencefoodshopsinsteadofout-of-townhy- permarkets. Inthe UK, IGD [7]estimate that spending in conve- niencestoreswillriseandthatonlineandconveniencestores.

Heating,VentilationandAir-Conditioning(HVAC)systemscon- tribute to a considerable amount of the total energy use of su- permarkets.IntheUKsupermarkets, reportedenergyuseby sub- systemsassign35%refrigeration,26.8%toHVACand18.6%tolight- ing[8,9].Thetotalenergyuseofasupermarketdependsonbusi- nesspractices,storeformat,productfoodratio,equipmentusefor in store preservation anddisplay. This paperpresents a studyof the energy use and the potential for savings due to mechanical night ventilative cooling ofthe HVACsystems offrozen food su- permarkets and consequently of supermarkets with high cooling requirements. The paper focuses on small size supermarkets en- ergyusewhichisbasicallyfooddominant.Theyareusuallysmall insalesfloorarea(lessthan1000m²),andareclassifiedassmall supermarkets[10].

In supermarkets with high food goods refrigeration could be as high as 60% [11]. The smaller the store the highest the en- ergy usedueto thehigherrefrigeration equipmentused because ofthehigherratiooffoodandnon-foodproducts[12].Asthetotal salesarea increases,the refrigerationenergyuse shareinthe to- talenergyusereducesandthelightingbecomesmoresignificant.

Conveniencefoodstoresareusually definedasfoodstoreslocated in central urban areas, near stations and shopping malls with a https://doi.org/10.1016/j.enbuild.2018.04.021

0378-7788/© 2018 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license. ( http://creativecommons.org/licenses/by/4.0/ )

Abbreviations

AC Airconditioning ach airchangeperhour AHU AirHandlingUnit

BEM BuildingEnergyManagement BWM BoxWhiskerMean

CAV ConstantAirVolume CS1 Casestudy1 CS2 Casestudy2

CVRMSE Coefficient ofVariation of the Root MeanSquared Error

DX DirectExpansion HDD HeatingDegreeDays

HVAC Heating,VentilationandAirConditioning LED Lightingemittingdiode

LT LowTemperature MBE MeanBiasError MT MediumTemperature NC NightCooling ppm Partspermillion RH RelativeHumidity TRY TestReferenceYear Tsurface SurfaceTemperature VC VentilativeCooling VRF VariableRefrigerantFlow Symbols

Tout OutdoorTemperature T_in InsideTemperature

T_offset Differencebetweeninsidetemperatureandoutdoor temperature

T_surface SurfaceTemperature

foodsalesarea<400m²[13].Energyanalysisofsupermarketshas shownthatsmallerstoresaremoreenergyintensive[5]andthere- forein thedesire to capturecustomerssupermarkets mighthave overestimatedtheirprofitpotentialasconveniencestoresaremore expensivetobuildandoperate[14,15].

Foodretailmarketsarecomplexenvironmentsdesignedtohave highvisibilitydisplay ofgoodswithsufficientthermalcomfortto encouragelongerstayforthecustomers.Therefrigerationrequire- ments of the display goods andindoor environmentalconditions aresometimesinconflictbecauseofthesignificantheatexchanges between them. Thus, optimised control strategies for HVAC sys- tems are required in order to achieve acceptable environmental conditions for customers andgood operation of the refrigeration system.

One control strategy is Ventilative Cooling (VC) during non- operationalhours atnight.VChasbeenreceivingattentioninre- centyears becauseofthe energysaving potentialforall typesof buildings. However, most published work focuses on residential andrelativelysimpleinoperationcommercialbuildingssuchasof- ficesandschools[16].Asimplifiedtooltoevaluatethepotentialof ventilative cooling has been developed [17], focussing on similar typesofbuildings[18].However, thepotentialofventilativecool- ingisalsohighinmorecomplexbuildingssuchasshoppingmalls andsupermarketsbutveryfewstudiestodatereportsapplication tosuchbuildings[19,20,21].

First, an analysisof the measured energy use andindoor en- vironmentalconditions ofthe two casestudystoresrepresenting twodifferentHVACsystemsandassociatedcontrolsbutthesame displayed products and refrigeration system is presented. Differ- ences and similarities are discussed. It continues with the vali- dated baseline model by EnergyPlus which enables the coupling

approach of HVAC with the refrigeration system. This model is usedfor parametricanalysis ofthe nightventilative cooling con- trolstrategyappliedtoevaluate potentialofVCindifferentHVAC applicationsincomparisonwithactivecoolingduringnight.

2. SelectedHVACsystemsanddescriptionofcasestudy supermarkets

HVACsystems in supermarkets can be divided into two cate- gories:

1.coupled HVAC system where heating, ventilation and AC are providedbythesamesystem.

2.decoupled HVAC system where heating and AC is separated fromtheventilationsystem.

ThecoupledHVACsystemisthemostcommonandprovidesair through overhead distribution ductwork to different parts of the store.ThecoupledHVACsystemscanprovideuniformairdistribu- tionin largeareas withsimilar coolingrequirements such asthe retail shops and withthe potential to incorporate heat recovery can be a very efficient and trustworthy solution. The decoupled HVACsystem is a non-duct airconditioner where heat is trans- ferred to or fromthe space directly by circulating refrigerant to evaporators.Theyaremoresophisticatedmulti-splitsystemswith many evaporators and refrigerant management and control sys- tems. As they do not provide ventilation, a separate ventilation system is necessary. These systems are lightweight and modular anddo not requirebig and specific structure on the roof of the buildingssotheyareconvenientforretrofitinstallations.Condens- ing units are placed outside and asducts are not needed, apart fromventilationsystem,buildingcostsandspacearesaved.Energy efficiencyis alsoimproved dueto theelimination ofduct losses.

Moreover,compressors are variablespeed enablingthecontrol of therequiredload. Maintenancecosts includemainlythechanging offiltersandcleaningofcoils.

Twocasestudystores were selectedaccording theabove cat- egorisation. CS1 is a refurbished two storey building located in central west London using an all air constant air volume HVAC whichrepresentsthe coupledHVACcase. CS2 isa newpurposed builtstorelocatedinasuburbancommercialareainsouthernLon- don.Theheatingandcooling requirementsarefulfilledby avari- ablerefrigerant flow (VRF) systemandthus this store represents thedecoupledHVACcaseasthereisseparateductworkforextract mechanicalventilation. The refrigeration systems are stand-alone which resultto highinternal heat gains dueto heat released to the sales area from the condensers. For that reason, the cooling demand ishigher than heatingand mitigation actions ofcooling requirementsareessential[24].

Thetwocasestudystores(Fig.1)belonginthesamesupermar- ketchainwithsimilarproductsandsamerefrigerationequipment.

CS2hasbeenpresentedandanalysedindetailin[11] andinthis paperboth casestudy storesare analysed andcompared. CS1 is inacitylocation andsurroundedby commercialbuildings.It isa refurbishedtwostoreyheavy–weight[22] buildingwith thesales area(469 m²)onthegroundfloorwhilethesecond floorisused asastorage area. Thecoupled HVACsystemforthe salesarea is roof mounted AHU witha DX cooling coil (88kW) andan elec- tric heating coil (24kW). The set point temperatures have been set to 19.5°C forheating and 20.5°C for cooling during trading times while cooling for non-trading hours (night cooling, NC) is inoperation with16°C setpointtemperature. It isa ConstantAir Volume (CAV) system which provides sales area with 6 m³/s in trading hours through 11 four waydiffusers, 1 three-way and 3 twowayblowfixed bladediffusers.Thereisalsoanelectricdoor heaterratedat18kW.Ventilationratesfortheexhaustsystemdur- ingtradinghoursaresetto6achforsalesand1achforthestor-

Fig. 1. Sales areas floor plan and HVAC systems description of CS1 and CS2.

Fig. 2. a) LT lift-up lid frozen food cabinet, b) LT open top frozen food cabinet, c) MT open vertical cabinet.

agearea. There are also supplementary extract ducts only above theopen front multideckcabinets whosewarm airis eitherex- hausteddirectlytotheatmosphereorusedtoheatthestoragearea onthegroundfloorwhenheatingisrequired.Thelightingsystem istypicalT8typefluorescentforthesalesarea.Theyconsistoflu- minaireswith3lamps; 21 inthe tillsareaand63 inthedisplay area. LED stripsare installed in thenorth-east andback sides of thesalesareawhichoperate24h.

CS2, amedium-weightbuilding[22],isina typicalsmallout- of-townretail centre.It issinglestorey newly builtwith315 m² salesarea. ThedecoupledHVACsystemofthesalesarea isaVRF systemfor both heating andcooling. Two equally sized outdoor condensingunitsprovidetotalheatingoutputof113kWandcool- ing output 101kW delivered to salesarea onlythrough 7 ceiling cassettes and 1 door heater. The HVAC system is operated 24h with20–21°Csetpointtemperatureforbothcoolingandheating;

theheat pump workseither asa compressor or evaporator con- trolledby the BEM system. Extraction of the air from sales and staff area isby an extract fanoperated 24h. Ventilation rates for theexhaust system duringtradinghours havebeen set to 6ach forstaff areasand salesarea,10achfor restroomsandcloaksand 1achforthestoragearea.Duringnighttimetheexhaustfanisset tolowerspeed(3ach).ThelightingluminairesaretypicalT8type fluorescent for the sales area. They consist of luminaires with 3 lamps;23inthetillsareaand30inthedisplayarea.LEDstripsare installedinthenorth-eastandbacksidesofthesalesarea which operate24h.

For both case-studies, the refrigeration system consists of plugged-incabinets andfreezer and chillercoldrooms. The types ofcabinets(Fig.2)andtheloadsarepresentedinTable1. 3. Monitoringresultsanalysis

This section presents a comparison of the monitoring energy andindoor environmentalconditions results ofboth casestudies

stores. Some preliminary results for CS2 have been already pre- sented in [11] where the model development was presented. In this paperthe data are comparatively presented focusing on the impact ofdifferentHVACsystems andcontrolson energyperfor- manceandresultinginternalenvironmentalconditions.

3.1. Energyuse

Figs. 3 and 4 present an overview of hourly measured en- ergydatausingboxwhiskermean(BWM)plots.Themeanhourly energy use of the months is presented based on hourly data.

Fig. 3 shows 5 years data forCS1. Winter 2013 wascolder than theother winters (approximately535higherHDD);thereforeen- ergy use is higherduring thiswinter. The store has a consistent energydemandduringcoldmonthswithaveragetradinghoursen- ergy useataround 0.14kWh/m² sawithpeaks onwarm months 0.18kWh/m² sabeforefallingtothenon-tradinghoursenergyuse (75th percentile). For CS2 (Fig. 2) winter 2015 was colder than 2014(averageHDDforJune2013–May2014was1760andforJune 2014–May2015was1908)andthisresultedinhigherenergyuse.

CS2presentedanaverage0.14kWh/m²sa(25thpercentile)during tradingtimewithpeaksonwarmmonthsataround0.17kWh/m² sa.

AstheHVACoftheCS1isnot operatingduringnight(in com- parisonwithCS2whereHVACison24h)andonlyfreenightcool- ing isinoperation,there isadifference betweenthenon-trading time energy use betweenthe two stores. CS2 energy useduring non-tradingtimesobservedto bearound 0.10kWh/m² sa. Onthe other hand, energy use of CS1 during non-trading hours ranged from0.09to0.12kWh/m² sa.

Average annual energy use is 1103.3kWh/m² sa for CS1 and 1117.3kWh/m² saforCS2 whichare attheupperrangeofsuper- marketsandatthelowerrangeoftheconveniencestores[11]be- causeofthehigherrefrigerationload.

Fig.5presentsthecorrelationoftheirdailyenergyusewiththe outdoorairtemperature.Itisobservedthatforbothstoresthereis an outdoortemperaturewherethedailyenergyuseis atits low- estlevel. Thisisaround 9°CforCS1 andbetween8°Cand12°C forCS2.Abovethesetemperaturesthecoolingrequirementsofthe buildings increases and consequently the daily energy use; from 25% to 50% forCS1 and from19% to 42% forCS2 The maximum dailyenergyusemonitoredforwarmdaysisalmost thesamefor bothstoresbutslightlyhigherforCS1.

However, adifferentpatternemerges forcolddaysandthisis duetothedifferentcontrolstrategyoftheHVACsystems.CS1with thefreecoolingduringnightandnon24hHVACsystem,presented lowerdailyenergyuseduringcolddays.The24hHVACsystemin

Table 1

Refrigeration equipment and loads of CS1 and CS2.

Case Study Chilled food cabinets Frozen food cabinets Coldroom

Open front multi-deck cabinets Lift up lid cabinets Open top case frozen Freezer Chiller

CS1 10 70 3 60 m ² 12 m ²

Refrigeration Load 20.3 kW 30.7 kW 30 kW 5.2 kW

CS2 7 58 3 29 m ² 6 m ²

Refrigeration Load 10.4 kW 26.3 kW 8 kW 2.3 kW

Fig. 3. BWM plot of hourly measured energy use per sales area (July 11–June 16).

Fig. 4. BWM plot of hourly measured energy use per sales area (June 13–June 16).

CS2resultedinhigherheatingrequirementsandthushigherdaily energyuseduringcolddays.

3.2. Indoorenvironmentalconditions

Figs.6 and7presenttheresultsinBWMplots forairtemper- atures fortwomonths (July2014 andDecember2014),indicative for warm andcold periods respectively. In July andduring trad- inghours,airtemperaturevariedsignificantlybetweenthedaysof the month and ranged between 22°C to 24°C for the tills area and21°Cto23°Cforthedisplayarea.July2014wasthewarmest monthofthissummerandduringthedaysofthehighestoutside temperature thetemperature inside thestore (both tillsanddis- playarea)reached28°C.ForCS2,airtemperaturerangedbetween

22°Cand23.5°Cinthetillsareaandbetween19.5°Cand22°Cin displayarea.

Internal patternsoftemperature followexternal maximumair temperatures and the continuous opening of the door and heat gainsofthesingleglassed windowsinboth casestudystoresaf- fect significantlyinternal air temperature. Thisisthe reasonwhy theairtemperatureinthedisplayareadiffersfromtheairtemper- aturemeasuredinthetillsarea.Thisisobservedmoreremarkably intheCS2wherethetemperatureinthedisplayareafoundtobal- ance around thesetpoint temperature(21°C)whilethe tillsarea presented temperature 1–2°C higher or lower for warm months andcoldmonthsrespectively.CS1 seemedtopresentinsignificant fluctuations from the setpoint temperature (19.5°C) for Decem- berwhileabiggerdifferencewasobservedfortheJuly(1.5–2.5°C abovethesetpoint).

Fig. 5. Daily energy use per sales area according to different outdoor temperatures (left: CS1, right: CS2).

Fig. 6. BWM plots of measured air temperature in sales area (tills and display) for June 2014 during trading times (up: CS1, down: CS2).

Fig. 7. BWM plots of measured air temperature in sales area (tills and display) for December 2014 during trading times (up: CS1, down: CS2).

Accordingtomeasureddata,adifferenceof1–2°Cbetweentills and display area is observed. Thisis because the tillsarea is af- fectedbytheairinfiltrationofthecontinuousopeningofthemain entrance.Impactsoncomfortduringtheextremeoutsideweather conditionswere not analysedfurther ascustomers’visitsinstore doesnotexceedmorethananhour. Recommendations[23] men- tion that there is no discomfortand healthrisk below25 °C for summerperiodandabove19°C.Thesevalueswerenotmeasured morethan 2hinordertocreatediscomfort[24].Moreover, elec- tricdoorheatercurtainsareplacedabovethedoorsformaintain- ing acceptable environment in salesareas andhelping to reduce energylossesfromtheconditionedarea.

However, thetemperaturedifference betweentillsanddisplay area istakenintoconsiderationinthemodeldevelopmentwhere the salesarea is separated intotwo thermal zones, tillsanddis- play area based onthe observationofthe recordedtemperatures indifferentareasinsidethesalesarea.Forthisreasondifferentin- filtrationrateswereusedforthetillsareaandwereapproximately 80%higherthandisplayareaduringtradingtimes.

Duringnon-tradingtimes(Figs.8and9),differentcontrolstrat- egyisusedforthetwocasestudystoresasdescribedinSection2. For CS1 where the ventilation is coupled with the heating and coolingsystemandthefreecoolingisinoperation,theairtemper- atureofthetillsarearangedbetween20°Cand22°Cbutreached upto24°Cduringthewarmestdaysofthemonth.The sameap- pliesforthedisplay area. DuringDecemberthe effectofthefree night cooling is significant due to outside conditions permitting andthe airtemperatureof both tillsanddisplay area varied be- tween16°C(whichistheminimumsetpointtemperaturethatfree nightcoolingisactive)to18°C.

ForCS2, wherethe ventilationis decoupled fromheatingand cooling system and the HVAC is in operation 24h, during non- tradingtimesofJuly,averageairtemperatureinthetillsareafluc- tuatesbetween21°C and22°C,slightly higherthan the setpoint temperature(20–21°C).Inthedisplayareatheaverageairtemper- atureis1°C lowerthan thetillsarea.Theopposite wasobserved in December 2014;average air temperature inthe tills area was 1°Clowerthanthetemperatureinthedisplayarea.

Fig. 8. BWM plots of measured air temperature in sales area (tills and display) for July 2014 during non-trading times (up: CS1, down: CS2).

In bothcases,internal airtemperaturevariationsfollowexter- naldailyminimumtemperaturepattern.

RelativeHumidity(RH)doesnotpresentsignificantdifferences between tills and display areas; 40–75% for CS1 and 40–65%

forCS2 forwarm months and unremarkably lower in cold peri- ods.Carbondioxideconcentration measurements rangedbetween 400ppm during non-trading times and 650ppm during trading timesfor both casestudy storesindicationgood ventilation pro- vision[23,25].

4. EnergyPlusbaselinemodel 4.1.Modeldevelopmentandverification

The model developmentof the baseline models and theveri- fication methodology has been presented in [11] where the CS2 modeldevelopmentispresentedindetail.Thesamemethodology andprocedurewasfollowedforCS1.

Following two levels ofcalibration; level 1based on available designdatatocreatetheas-builtmodelandlevel2thatincluded the as-built and operating information, the final thermal model for CS1 with 14 thermal zones was validated against measured data for both energy use andtemperature conditions for a year.

The building’sannualenergy usefromJune 2014 to May2015 is 1103.6kWh/m² salesarea. Thefinalcalibratedmodelpredictionis 1098.5kWh/m² sales area (adeviation of −0.5%). Regarding CS2, the thermalmodel that wasdeveloped has 9thermal zonesand thefinal calibrated modelpredictionis 1104.3kWh/m² sales area whilethemeasured energyusefromJune 2014 toMay2015 was 1143.4kWh/m²(adeviationof3.4%).Figs.10and11enableaquick visual inspection ofmeasured andsimulated energyuse for two indicativeweeks.Fig.10referstoCS1whileFig.11toCS2.

ASHRAE Guideline 14-2002 defines the evaluation criteria to calibrateasimulationmodel.Monthlyandhourlydata,aswellas spot and short term measurements can be used for calibration.

Mean Bias Error (MBE) and Coefficient of Variation of the Root

Fig. 9. BWM plots of measured air temperature in sales area (tills and display) for July 2014 during non-trading times (up: CS1, down: CS2).

MeanSquaredError(CVRMSE)areusedtoevaluatethemodelun- certainties[26].

According theresults (Fig. 12), both casestudystoresEnergy- Plus models presented MBE and CVRMSE values within accept- ablelimitsforbothenergyuseandindoorairtemperature.Further analysisregardingthemodelvalidationisgivenin[11].After that a calibratedEnergyPlus modelwithverified energysystems (CAV

&VRFHVACsystemandrefrigerationsystem)isusedforthepara- metric analysisforoptimisationof NCcontrol strategy appliedin bothcoupledanddecoupledHVACsystems.403

4.2. DescriptionofNCoperationandcontrolstrategy

Mostsupermarkets useactivecoolingsystemtomeetbigcool- ing requirements duringtrading times whilesales area tempera- turerise after activecooling stops.When NC isin operation,the heatstoredatnightwillbereleasedtotheairduringthenextday todelaytheincreaseoftheroomtemperature.

Apreviousstudyfornightventilationimplementationtoasu- permarkethas concludedthat longer night cooling activation re- sults to fewer hours of AC system operation and higher energy savings [21].However, studiesforoffices andother non-domestic buildinghaveindicatedthatthreecontrolaspectsshouldbetaken into consideration [27]; duration, system initiation and system continuation in order to maximise energy savings. In this case study,thefollowingruleswereimplemented:

i) initiation:T_out<T_in,

ii)continuation:Tout<T_inandTout-T_in<T_offset iii) termination:continuationruleandT_in=T_min

The continuation rule ensures that the outsideair brought in iseffectivein cooling thebuilding.When thetemperature differ- encebetweeninsideandoutsideair(T_offset) is low,theincoming airwillhave littleeffecton coolingwhile theventilationfan en- ergyusewillincreasethetotalenergyuse.However,iftheoutside airtemperatureissignificantlylowerthantheinsideairtempera-

Fig. 10. Comparison between metered and simulated hourly energy use for an indicative warm and cold week, CS1.

Fig. 11. Comparison between metered and simulated hourly energy use for an indicative warm and cold week, CS2.

Fig. 12. MBE and CVRMSE analysis of the energy use based on hourly data.

Fig. 13. Heating, cooling and fans energy use for different air flow rates.

ture, T_min willbe achievedfastandtheduration ofnightventila- tionisdecreased[28].

Moreover, althoughNCcould increasethe totalenergysavings ofthestores, attentionshouldbe paidintheairconditions(tem- perature andRH) brought instore asit mayaffect the cold sur- faces of the cabinetsfromcondensation orit maybe harmful to theoperationoftherefrigerationsystemoritscontrols.Thestores’

LT cabinetsare glass liftup lid cabinets whichduring NCopera- tionremain closedso theevaporatorcoilsarenot affectedbythe ambient air(if hot orhumid) andthus crucial problems are not created in the evaporator coils operation. However, action might betakentopreventcondensationonthesurfaceoftheglass.Fog- gingandriskofcondensationontheexternalside oftheglassor the multi deckcabinets’ curtains might occur in humid climatic conditionswhile reducingthe ambienttemperature.Forthat rea- son, experimental results fromlaboratorytest in theCSEF centre facilities took place in order to evaluate the T_surface of the glass lidoftheLTcabinets(whichoccupythevastmajorityofthesales area).ThistemperaturegivesinsightsoftheRHlevelsthatmustbe maintained inthesalesarea inordertoprevent condensationon theglasslid.Accordingtomonitoringresults,thesurfacetempera- tureofthecabinet’sglasslidisnotaffectedsignificantlyfromthe temperatureinsidethefrozenfoodcabinetbutfromambientcon- ditions anddoes not fall to low enough temperatures wherethe highrelativehumiditycouldcreatecondensationproblems.Taken intoaccountthatat10°Cambientconditionsthesurfacetempera- tureoftheglasslidisaround7°C,RHshouldnotexceed85%[24]. TheparametricanalysiswascarriedoutwiththeTestReference Year(TRY)weatherfilefromCIBSE.

5. Resultsanddiscussion 5.1. CoupledHVACsystem

NCisalreadyinoperationinstore withcoupledHVACsystem duringnon-trading times.Thesystemis designedtoprovide free night cooling with 6 m³/s when the return air and outside air temperaturehave1°Cdifferenceanduntiltheinside temperature reaches16°C.

The parametric analysis was performed for different airflow ratesaccordingtofanspeed (1–6m³/s),T_offset (1–20°C)andT_min (10–17°C).Minimumtemperatureinsidethestorewaschosennot tofallbelow10°Cinordertoavoidcondensationontheglasscabi- nets.SettingtheT_mintothelowestlevels(10°C)glasssurfacetem-

Fig. 14. Cooling energy use for different T offsetand T min.

Fig. 15. Total energy use for different T offsetand T min.

peraturemonitoring wastakeninto accountto avoidrisk ofcon- densationontheglasslidasexplainedinSection4.2.

Fig. 13 presents fan, heating and cooling energy use for dif- ferent airflow rates,T_offset andT_min; the combinations are inte- grated andare presented asa range of energyuse inthe graph.

Fig.14presentsthecoolingenergyusefordifferentT_offsetandT_min. Theair flowrateduringnight coolingplays an importantrole as

the higher airflow increases the fans’ energy use. However, low airflow rates could have similar effect on cooling demand with areduction of heatingrequirements duringthe following day.In Fig.13,thefans’ annualenergyuserange isindicated asa result ofthedifferentT_offset.Higherairflowratehaswiderrangebecause thereductionofthe internaltemperaturetoT_min isachievedfast andthedurationoftheNCisdecreased.

For lower airflow ratesthere isa point wherethe maximum totalenergy usereduction occurred; energyuse startsincreasing until reaching the point where NC is not effective (total energy useequals thetotal energyuse when NCis off)(Fig.15).This is dueto theincrease ofthecooling energywhichafterwards leads toan increase of the totalenergy use(Fig.14). This pointis ob- served torange between5 and7°C. Refrigeration system energy usedecreaseswithlowerT_minbutafter5–7°CT_offsetstartsincreas- ingagainuntil therefrigeration energyuseobserved whenNC is notinoperation (Fig.16). Theoptimum combinationsofparame- tersleads to up to 3% of the total energyuse fromthe baseline model– thisequatestoenergyusereductionof35.3kWh/m²/year inthestorewhichismainlyduetothe10%reductionoftheHVAC energyuse.

Finally, Fig. 17 summarises the significance of the optimised control strategy of the HVAC system for NC operation with the optimum T_offset in order to achieve the biggest energy total en- ergyreductions.Itpresentsgatheredresultsforlowerairflowrates andsetpointtemperatures toindicatetheirsignificanceintheto- tal results. Lower air flow rates reduce the fans energy use and enablethebiggerdurationoftheNCandthusthetotalenergyuse reduction. Moreover, lower T_min inside the store reduces slightly the cooling demand of the store in comparison with the base-

erationsystemoperationandconsumptionbutwithsignificantde- creaseintheHVACduetoreductioninfansenergyuseandcooling requirements.

Without any change to the equipment of HVAC system, opti- misedcontrol strategyforexhaust NCresultedthat thelower air flow rates lead to bigger total energy use reduction due to re- ducedfansenergyconsumption(Fig.18).Higherairflowratespre- sentedtohavestrongestcorrelationwiththeT_offset asmentioned forcasestorewithcoupledHVACsystem;whileT_offset increases,a sharperreductionisoccurredandthisisbecausethecoldairthat is broughtinside has biggereffect on the inside airtemperature and T_min is achieved quickly andthus the duration ofthe NC is decreased.

It is also observed that for low air flow rates there is a spe- cific T_offset where the total energy use starts slightly increasing (T_offset>5°C). After that point, where the optimum total energy use reduction occurs, the cooling energy demand increases and withhigher T_offset the cooling energyuse increases moresignifi- cantlyastheNCisnotmoreeffective(Fig.19).Forhigherairflow ratesthisT_offset increases upto 7°C. Theoptimum combinations of theparameters lead to3.6% reduction in the total energyuse whichequalsto40.8kWh/m² peryear.Inotherwords,byreplac- ingtheactivecooling withNCduringnon tradingtimes,areduc- tionupto17%isachievedincoolingdemandandthesame17.5%

infansannualenergyuse.

Refrigerationenergyusewasfoundtofollowthesamepattern withwhatwasanalysedpreviously inSection5.1;aftera specific T_offsetrefrigerationenergyuseincreasestothelevelsthatNCisno moreeffective.

ForintakeNCtheresultsagreedwithwhathasbeendiscussed forexhaustNCcontrolstrategy. Theairflowrateisakey param- eter for the NC andthe lower air flow rates lead to lower total

Fig. 17. Total annual energy use and sub-systems energy use for different air flow rates and T min, coupled HVAC.

Fig. 18. Total energy use with different air flow rates for different T offsetand specific T min.

Fig. 19. Cooling, heating and fans energy use for different T offsetfor T min= 10 °C and with 1 ach air flow rate.

Fig. 20. Total annual energy use and sub-systems energy use for different air flow rates and T min, CS2.

energy use due to fans energy use decrease but with the same effect ofnight cooling duetothe fact that the nightcooling du- ration is bigger.However, asis proposed for storeswithcoupled HVAC systems, for lower air flow rates there is point that the cooling requirementsstartincreasingandNCisnomoreeffective (T_offset>7°C). With higher T_offset than 2°C, although the cooling

energydemandincreases,thefans energyusedropsmore signif- icantly andleadsto lower totalenergy use.The highesttotal re- duction observed for lower air flow rates. As the T_min increases, theduration oftheNCis decreasingandunremarkablereduction isobservedonthetotalenergyuse.Areductionofaround3.2%on

6. Conclusions

Twofrozenfoodsupermarketsweremonitored(energyuseand environmental conditions) and an EnergyPlus model was devel- opedandvalidated againstmeasurements. Nightventilative cool- ing, taking into account interaction with HVAC andrefrigeration systems,wasexploredasastrategytoreduceenergyuse.Apassive cooling through night ventilation coupled with active cooling is proposedasasolutionforsupermarketswithhighcoolingrequire- mentssuchasfrozenfoodsupermarketswithpluggedinrefrigera- tionequipment.ColdclimatessuchasLondonenablethisstrategy tocoolindoor environmentduringnon-tradingtimes.The poten- tialofnightventilativecoolingforsupermarketswithhighcooling requirements wasinvestigated for implementation coupling with most common active cooling strategies (coupled and decoupled HVACsystems).

ItwasproventhatNCincombinationwithhighbuildingmass hasa potentialtoreduce theworkinghoursandthusthecooling energyuseofthenext dayactive cooling.Parametricanalysisfor theoptimisationofNCcontrolstrategyresultedtothefollowings:

• Nightventilativecoolinghasgoodpotentialforthespecificcase studies asthey includehigh refrigerationloads whichare de- liveredwithplugged-incabinets.Coolingdemandissignificant higherthanheating.

• Control strategy for night ventilative cooling plays an impor- tant role as proved in CS1 where night ventilative cooling is alreadyinoperationbutwithbettercontrolsbiggerreductions areachieved.

• CS2 iscooled duringthe night to maintain the setpoint 24h.

Implementing free exhaust or intake night ventilative cooling leadstoareductioninthetotalenergyuse.

• Simulations indicate that longer period of night ventilative cooling operation leads to higher energy savings enabled by lower air flow rates which have a small impact on fans en- ergyusebutcooleffectivelyaslongerperiodisneededtoreach T_min.

• Inside-outside temperature is an important night ventilative coolingparameter. Parametricanalysisindicated thatoptimum savings occurred ifthe airinside thestores has5–7°Cdiffer- ence with the outside air. The higher the air flow rate, the higherthisdifferenceshouldbeforbetterchanges.

• Withnightventilativecooling,cooling demandduringtheday isdecreasedinbothstores.

• Refrigeration system energy use has an unremarkable reduc- tion; with T_offset higher than 5–7 °C, refrigeration energy use startsincreasinguntilitreachesenergyusewithoutNC.

• Although night ventilative cooling has good potential for to- talenergysavingsofthesupermarkets,condensationproblems mightariseontheglasssurfaceofthefrozenfoodcabinetsand careshouldbetakenintheselectionofthecontrolparameters.

References

[1] Statista, Number of supermarkets and grocery stores in the United States from 2011 to 2015, 2015. [Online]. Available: https://www.statista.com/statistics/

240892/number- of- us- supermarket- stores-by-format/ .

[2] J. Clark , Energy-Efficient Supermarket Heating, Ventilation, and Air condition- ing in Humid Climated in the United States, National Renewable Energy Labo- ratory, NREL, 2015 .

[3] EIA, Independent Statistics & Analysis, 2017. [Online]. Available: https://www.

eia.gov/todayinenergy/detail.php?id=30712 . [Accessed 2017].

[4] PWC, 2015-16 Outlook for retail and consumer products sector in Asia, 2015. [Online]. Available: http://www.pwc.com/kr/ko/publications/industry/

retail- in- asia- outlook _ 201516.pdf .

[5] S. Tassou , Y. Ge , A. Hadaway , D. Marriot , Energy consumption and conservation in food retailing, Appl. Therm. Eng. 31 (2-3) (2011) 147–156 .

[6] CREATIV, The future energy effective supermarket, 2014. 14 Febru- ary. [Online]. Available: http://www.sintef.no/projectweb/creativ/

the- future- energy- effective- supermarket/ .

[7] IGD, Growing UK convenience market, 2014. [Online]. Available: http://www.

igd.com/Research/Retail/Growing- UK- convenience- market/ .

[8] DECC, Non-domestic building energy use project phase I-pilot study of the food and mixed retails sector, 2013. [Online]. Available: https:

//www.gov.uk/government/uploads/system/uploads/attachment _ data/file/

207319/DECC _ Non _ - _ domestic _ building _ energy _ use _ project _ phase _ I.pdf . [9] S. Tassou , Y. Ge , Reduction of refrigeration energy consumption and environ-

mental impacts in food retailing, in: Handbook of Water and Energy Man- agement in Food Processing, Woodhead Publishing, Cambridge, UK, 2008, pp. 585–611 .

[10] Grocery Universe, “Results of the 54th inventory of retail grocery in Belgium,”

2016. [Online].

[11] Z. Mylona , M. Kolokotroni , S. Tassou , Frozen food retail: measuring and mod- elling energy use and space environmental systems in an operational super- market, Energy Build. 144 (2017) 129–143 .

[12] Annex 31, Advanced Modeling and Tools for Analysis of Energy Use in Super- market Systems, IEA Heat Pump Centre, Sweden, 2012 .

[13] Grocery Universe, Results from the 54th inventory of retail grocery in Belgium, 2016. [Online]. Available: http://www.nielsen.com/be/en/insights/reports/2016/

nielsen-grocery-universe-2016.html .

[14] G. Ruddick, Little convenience stores should become a big prob- lem for supermarkets, 2015. [Online]. Available: http://www.

telegraph.co.uk/finance/newsbysector/retailandconsumer/11652760/

Little- convenience- stores- could- become- a- big- problem- for- supermarkets.

html .

[15] V. Barford, The rise, fall and rise of the mini-supermarket, 2014. [Online]. Avail- able: http://www.bbc.co.uk/news/magazine-25762466 .

[16] IEA, EBC, Annex 62, ventilative cooling, state-of-the-art review, 2013.

[Online]. Available: http://venticool.eu/wp-content/uploads/2013/09/

SOTAR- Annex- 62- FINAL.pdf .

[17] A. Belleri , M. Avantaggiato , T. Psomas , P. Heiselberg , Evaluation tool of climate potential for ventilative cooling, Int. J. Ventil. (2017) .

[18] P. O’Sullivan , P.A. O’ Donovan , G. Zhang , G.C. da Graca , Design and perfor- mance of ventilative cooling: a review of principals, strategies and components from International case studies, 38th AIVC Conference, Nottingham, UK, 12-13 September 2017 .

[19] A. Belleri , M. Haase , S. Papantoniou , R. Lollini , Delivery and performance of ventilative cooling strategy: the demonstration case of a shopping centre in Trondheim, Norway, 38th AIVC Conference, Nottingham, UK, 12-13 September 2017 .

[20] M. Avantaggiato , A. Belleri , M. De Carli , R. Lollini , Mixed-mode ventilative cool- ing opportunity for an existing shopping mal retrofit, 38th AIVC Conference, Nottingham, UK, 12-13 September 2017 .

[21] Wu Li-Xia , et al. , Night ventilation and active cooling coupled operation for large supermarkets in cold climates, Energy Build. (2006) 1409–1416 . [22] BS EN ISO 13790:2008, “Energy performance of buildings-Calculation calcula-

tion of energy use for space heating and cooling”.

[23] CIBSE, Guide A: Environmental Design, Chapter 1 Table 1.5, 2015 .

[24] Z. Mylona , Experimental and Computational Study To Improve Energy Effi- ciency of Frozen Food Retail Stores PhD Thesis, Brunel University London, Lon- don, 2017 .

[25] ASGRAE, Standard 62.1: Ventilation for Indoor Air Quality, 2013 .

[26] ASHRAE, ASHRAE Guideline 14-2002, Measurement of Energy and Demand Savings, 2002 .

[27] M. Kolokotroni , Night ventilation for cooling: field tests and design tools, in:

Low-Energy Cooling Technologies For Buildings: Challenges and Opportunities for the Environmental Control of Buildings, Professional Engineering Publish- ing, London,UK, 10 March 1998, pp. 33–44 .

[28] H. Aria , H. Akbari , A predictive nighttime ventilation algorithm to reduce en- ergy use and peak demand in an office building, Energy Power Eng. 7 (2007) 1821–1830 .