中國典型與極端天氣下制冷季辦公建筑空調系統(tǒng)短期管理柔性用能潛力

Abstract： To meet the challenge mismatches between power supply dem，modern buildings must schedule flexible energy loads inorder to improvethe efficiency power grids.Furthermore，it is essential to underst the effectiveness flexibility management strategies under diferent climate conditions extreme weather events. Using both typical extreme weather data from cities in five major climate zones ， this study investigates the energy flexibility potential an ce building under three short-term HVAC management strategies in the context diferent climates.The results show that the peak load flexibility overall energy permance thethree short-term strategies were affected by the surrounding climate conditions. The peak load reduction rate the pre-cooling zone temperature reset strategies declined linearlyas outdoor temperature increased.Under extreme climate conditions，the daily peak-load time was found to be over two hours earlier than under typical conditions， the intensive solar radiation found in the extreme conditions can weaken the correlation between peak load reduction outdoor temperature，risking the ability a building's HVAC system to maintain a comtable indoor environment.

Keywords： energy flexibility；dem-side management; extreme weather;HVAC systems；thermal require ments

1 Introduction

The simultaneous growth in global population urbanization has resulted in a marked increasein urban energy dem over the past several decades. Furthermore， many societies must now confront issues sustainability other problems caused by reliance on fossil fuels，such as energy shortages， greenhouse gas emissions health problems[1]. Mitigating greenhouse gas emissions reducing fossil fuel usage have become priorities many countries， more than 454 cities 23 regions have committed to achieving zero carbon emissions by the 2lst century[2]. aims to reach its“carbon peak” by 2O3O carbon neutrality by 2060. As a result，renewable energy sources are being developed rapidly to replace conventional energy ones.However，the nature many renewable energy sources can lead to undesirable changes in energy supply structures[3].Due to its intrinsic variability， renewable energy ten exacerbates the discrepancy between energy supply dem aggravates the fluctuations in net system load4]，causing unexpected curtailment rates[5]. To meet the challenge thispotential mismatching between supply dem，flexible load scheduling isrequired in order to improve the efficiency power grids[6].

Buildings represent a large portion the world' s energy consumption associated CO₂ emissions， consuming around 40% total energy usage in developed countries[7]. Since the energy dem a building can be shifted in time，a large potential energy flexibility can be achieved by low-carbon design energy-eficient operation in buildings， which can in turn provide better power grid manage ment. The term“building energy flexibility”is defined by Annex 67 as“the ability to manage a building's dem power generation according to local climate conditions，user needs， energy network requirements\"[3]. Lu et al. [8] have summarized the general approach to achieving energy flexibility，including dem-side management，dem response， the flexible control resources.

Under the building energy flexibility approach，Shan et al.[9]highlight four types flexible management strategies that can be used in buildings’ demresponseprocesses， includingdem limiting， dem shedding，dem shifting on-site gener ation.Identifying building power usage characteris ticslis one the most critical steps revealing a building’senergy flexibility.To thisend，with their large proportion energy utilization schedulable operation control，the ability HVAC systems to shifton-peak electricity dem alleviate shortterm dem-supply mismatch events has attracted increasing research attention[1o].

HVAC operation is regarded as the most influential energy consumer in buildings. In fice buildings alone HVAC operation can create a -15% to 70% variation in building energy consumption annuallyll. Thus，applying different control strategies to building HVAC systems can lead to considerable benefits in dem side management， this flexibility potential can be further enhanced by integrating HVAC systems with occupant behaviour control strategies， building thermal mass thermal energy storage systems[12]. For example，Malik et al.[13] found that more than half the total air-conditioners monitored in Australia had shown intensive energy usage during on-peak periods，indicating a considerable dem response potential. Although the ability HVAC systems to cut down peak-time power loads has been postulated，occupant thermal comt （\"userneeds\"）should not be adversely compromised under building energy flexibility management. Asa result，precooling/preheating zone temperature resets are typical methods used in HVAC operation control that enable energy flexibility ensure that room temperatures are within a comtable range.

Existing studies have already reported on the benefits precooling/preheating zone temperature reset. As a natural thermal storage source， building thermal mass is ten an integral part energyflexibility strategies. As the most straight ward approach to achieving dem shedding，directly switching f HVAC cooling/heating units commonlyresponds quickly to a shedding signal without affecting occupants’ short-term thermal comt[14]. However，unbalanced cooling distribution issues the instability zone/space thermal environments should be considered in how to maintain spaces within the comt range when HVAC systems is partly f [15]

To minimize the risk unacceptable thermal environmentsduring the dem-responseperiod， researchers have looked into the effectiveness zone temperature resets. Chen et al.[12] quantified the energy flexibility a maximum peak power reduction by up to 25% the maximum cooling power dem. Furthermore，by simultaneously resetting zone air temperature， supply air temperature，chilled water temperature condenser water temperature，the short-term curtailment approach can reduce building cooling dem by 23%-47% ，depending on occupant comt limits[16].Wang et al.[17] investigated the spinning reserve capacity a rising room air temperature setpoint （up to 3°C ） in Hong Kong reported a link between reserve capacity temperature rise that contributed more than 68% to the spinning reserve capacity required by the power grid in Hong Kong. In addition to HVAC systems building thermal mass，outdoor weather is also a factor that influences a building’s energy flexibility. For example，Yin et al.[18]simulated the hourly dem response potential indoor air temperature setpoint resets in both commercial residential buildings， the results highlighted the decreased dem response potential in the presence extreme outdoor air temperatures.

Additionally，one typical precooling/preheating strategy utilizes the thermal charging capacity building thermal mass to shed/store heat bee the occupied period or dem response period， this ismore commonly applied in commercial buildings with larger building thermal mass more dynamic incentives to lower/shift peak dem[19]. Xu et al. [20] reported that the precooling strategies can reduce chiller power by 80% during on-peak hours in fice buildings without thermal comt complaints. Jiang et al.[21] likewise developed a deep reincement learning framework to reduce HVAC electricity cost peak dem a single-zoneficebuildingthat was able to save 6%-8% monthly electricity costs， Hu et al. [22 developed a thermal-based self-learning model to examine the dem response potential residential air conditioners.These latter resultsrevealedthat thecombination room temperature reset precooling was able to reduce electricity consumption by 26% during the peak-time period a typical summer day without sacrificing indoorthermal comt.However，theeffectiveness precooling strategies on building load shifting is also influenced by contextual factors.

For example，Stopps et al. [23] examined the load-shifting capacity a pre-conditioning strategy in high-rise residential buildings using HVAC runtime data found that the overall impact on load der m reduction was not effective.However，the load-shifting potential the pre-conditioning strate gy varied in suites with different orientations.Similarly，Turner et al.[24] evaluated the electricity loadshifting potential integrating precooling buildingthermal mass.Theyfound thatalthoughthe pre cooling could effectively shift on-peak cooling load， its effectiveness was largely dependent on climate zone local outdoor weather.

Growing numbers extreme weather climate events have been reported[25]，especially high daily outdoor temperatures[26]. Dai et al.[27] compared the differences in AC usage behavior in between the extremely hot summers normal sum mers found differences between both the frequen cy use the duration operation by air conditioner users in each type summer.Considering the theory building thermal dynamics，HVAC energy consumption indoor thermal conditions can be largely influenced by the outdoor thermal environment，which can result in the effectiveness building energy flexibility that relies on precooling/pre heating zone temperature reset strategies.

Previous studies relevant to building energy flex ibility dem response mainly focus on the flexibility potential a given strategy its corresponding effects on indoor thermal comt. Although the effects certain energy flexibility strategies could be various due to the variation climate building structure，the permance applicabilityenergyflexibility strategies under different climate conditions are nonetheless understudied.Hence，the novelty thisarticle is to evaluatebuilding energy flexibility potential under different local outdoor climates local building regulations on building envelops in specifically，including one extreme weather condition. Three flexibility strategies are studied，in cluding pre-cooling，zone-temperature reset， par tial shutdown. The energy flexibility under five typical climate zones one extreme climate condition are compared.The following section first describes the details simulation modeling data input. Section 3 then analyzes the energy flexibility three typical flexibility strategies under various local climates extreme weather conditions， followed by a discussion regarding the effect outdoor climate on building energy flexibility in Section 4.

2 Methods

To evaluate the energy flexibility fice buildings with typical HVAC flexibility management strategies in the different climate conditions ，we simulated the HVAC power load shifting shedding potential in the cooling season （July August） a real prototype fice building in five major climate conditions one extreme climate condition， considering three previously mentioned typical shortterm HVAC flexibility management strategies.

2.1 Energyflexibilityindicators

Methods assessing energy flexibility have been established by existing studies to quantify building energy flexibility.However，the energy flexibilityindicators used in these quantification methods vary depending on building control objectives. Lu et al.[8] have summarized the main energy flexibility indicators evaluating flexibility capacity by considering building power/energy reduction[22]， financial cost[28]， CO₂ emission[29]，building thermal mas[30]， renewable energy generation[31]. In this study we use three indicators to assess the energy flexibility HVAC flexibility-management strategies under different outdoor climate conditions，including peak load reduction rate during the on-peak period，energy consumption reduction rate， the ratio difference peak load during the on-peak period. The parameters adopted to build the flexibility indicators are

shown in Fig. 1.

Peak load reduction rate during the on-peak peri od （ F_P is adopted to indicate the load shedding ability a given flexibility control strategy （Eq. （1））， energy consumption reduction rate（ F_E ）illustrates the difference in the overall HVAC energy consump tion bee after applying flexibility management strategies （Eq. （2））.

where， F_P is peak load reduction rate during the on peak period to 6：00PM asidentified in this study） （%） ， P_peak.base P_peak⊥fle are the peak loads the baseline case the flexibility case （kW），respectively.

where， F_E is the rate reduction energy consump tion the HVAC system （%） ， E_base E_fle are the energy consumption the HVAC system the baseline case the flexibility case （kW?h） ，respec tively.

Apart from the potential lowering peak load overall energy consumption，a system’s loadshifting abilityisalso an important indicator assessing the effectiveness a flexibility strategy. For this purpose we use “the reduction rate peak load during the on-peak period\" to present the energy flexibility the peak load shifting the baseline case the flexibility case （Eq. （3））.

where， F_R isthe difference in the ratio peak load during the on-peak period （2：00PM to 6：00PM as identified in this study） （%） ， R_on-peakbase R_on-peakfle are the ratios peak load during the on-peak period the baseline case the flexibility case，respectively.

2.2 Localweatherconditions

spans across several climate zones. Its national code Code Thermal Design Building（GB 50176—2016） divides into five major climate zones，the severe cold（SC）zone，the cold zone，the hot summer cold winter （HSCW） zone，the hot summer warm winter （HSWW） zone， the temperate zone，which are classified by the average temperature the coldest month （January） the hottest month（July）historically. To evaluate the energy flexibility fice buildings in different climate zones under different flexibility control strategies，we used the average annual weather data five typical cities in the five climate zones： Harbin the SC zone，Beijing the cold zone， the HSCW zone，Kunming the temperate zone， Guangzhou the HSWW zone.

The Chinese Stard Weather Data （CSWD） was used simulations typical weather conditionsas it has already been used extensively in building energy simulation studies in . Although CSWD data represent the long-term means regarding weathervariables，therecertainlyis real-world variance from year to year[32]. In addition，in order to assessthe impact extreme weather on building energy flexibility，we used the measured weather data collected in the extremely hot summer （July Au gust） 2O22 in ， （HSCW zone）， during which the consecutive days with a maximal outdoor temperature higher than 35°C numbered 32. Researchershave yet to reach a consensus as to what constitutes the extremely hot period.However，the most common one adopted is the period at least three consecutive days with a daily maximal outdoor temperature higher than 35 ° [27， 33].

2.3 Buildingsimulationdescription

A prototype building model was established in EnergyPlus stware based on an actual fice building. This building has 27 stories with an occupied area （HVAC operating） 40068m² .The model floor layout are shown in Fig. 2.

To simplify the simulation process，we divided the building layout into eight occupied zones depending on four building orientations. The envelope structure thermal permance this building prototype is based on the regulationsbuildings that are subject to trade-f judgement in General code energy efficiency renewable energy application in buildings （GB 55015—2021）[34]. The simulated building's shape coefficient window-wall ratio are O.1 0.6，respectively，which satisfies the requirements in GB 55015—202l. Since the building design code isdifferenteachclimatezone，thethermalcharacteristics the prototype building in each climate zone arelisted in Table1， they all satisfy the local building energy efficiency stards in GB 55015— 2021.

The internal disturbance schedule this model also followed the regulations in GB 55015— 2021. The occupant density was set to 10m²/per^- son， the power density lighting equipment are 9W/m² 8W/m² respectively. The occupancy，lighting， electrical equipment usage prile canbe found in Fig.3. Finally，the chiller ventilation systems the HVAC system were set to operate from O7：00 to 19：00 with the indoor temperature setpoint at 26^°C on working days.

2.4Energy flexibility strategies

Asmentioned above，rescheduling an HVAC system’s operation prile is an effective approach to achieve flexibility in load shifting shedding，par ticularly when integrated with a passive thermal storage source such as building thermal mass[35]. To assess the effect outdoor climate on the energy flexibility potential short-term HVAC management in fice buildings，this study concentrates on three typical energy flexibility strategies under different local summer climate conditions，including pre-cooling， zone temperature reset， partial shutdown. The requirement indoor thermal comt during theoccupied period isconsidered in theanalysis the flexibility strategy application.

Under the baseline case，the fice building is occupied from O8：00 to 18：00 from Monday to Friday. The chiller is operated from O7： OO to 18：00 onworkdays， the ventilation system is shut down one hour later （O7：00 to 19：0O） to secure indoor air quality. The indoor temperature setpoint is set to be 26^°C occupied areas on working days during the cooling season （July August）.

The basic chiller ventilation system opera tion schedules under the three flexibility strategies arethe sameasthe baseline case.However，each flexibility case adopts one HVAC setting adjustment scheme，asshownin Table2.

Current research mostly uses the time range prior to operating hours or the valley period bee peak dem as the pre-cooling period[4，20，24]． For fice buildings[2o]， the pre-cooling period is typically set to be from O5：OO to 14：OO. Theree，under the pre-cooling case，the indoor temperature setpoint in occupied areas was set to be 24^°C until 13：00，bee the peak hours，then to 28°C from 13：00 to 18：00 on workdays.

Under the zone-temperature reset case，the indoor temperature setpoint in occupied areas was automaticallyreset from 26^°C to 28^°C 5hours from 13：00.

Under the partial shutdown case，the simulated building was closed to deal with a simulated emergencyevent.Considering that the effect solar heat gain building thermal mass discharging on building flexibility can be impacted by building orientation local solar radiation，we simulated the strategy partially shutting down the HVAC systems indoor zones facing west occupancy in these zones was set to unoccupied. The shutdown schedule was from 13：00 until the end the day， the HVAC system operation schedule occupancy in the remain ing zones was unchanged relative to the baseline case.

2.5 Typical extremeclimateconditions

The hourly average outdoor dry bulb tempera ture five major climate zones one extreme hot condition in July August is summarized shown in Fig.4. Based on this，Kunminginthe tem perate climate Harbin in the SC climate present the lowest outdoor temperatures， ， Guangzhou， Beijing cities present the highest outdoor temperatures during the cooling season. As expected，the hourly average outdoor dry bulb temperature the extreme condition was the highest among all conditions. Fig.5 demonstrates the hourly outdoor dry bulb temperature ratio all conditions. Those above 35°C account nearly 40% total summer hours in the extreme summer condition.On the contrary，the out doordry bulb temperatures in the five typical climate conditions are mostly lower than 30°C

Fig.6 compares the hourly average total solar radiation all climate conditions in July August， we can see that the distribution total solar radiation five cities in typical climate conditions is similar.Forthe extreme condition case，the overall total solar radiation is obviously higher than other typical climate conditions; the peak total solar radiation at noon is over 200W/m² higher than that other conditions. Again， the extreme summer condition presents a proportion total solar radiation higher than 550W/m² than other conditions，as shown in Fig. 7.

To quantify the cooling dem different climate conditions， the monthly cooling degree days during the cooling season （July August） based on （20 26°C （CDD26） are shown in Table 3. With the continuously hot outdoor condition， the cooling degree day in extreme climate condition was much higherthan the other conditions，which indicatesa stronger cooling dem.

Aspreviously mentioned，the outdoor thermal conditions five cities in different climate zones present discrepancies in amplitude temporal scale due to geographic location，which may potentially result in a difference in load shifting/shedding effectiveness when applying the same energy flexibility strategies. Furthermore，a considerable discrepancy was found in the outdoor temperature solar radiation between typical extreme summer conditions.Thus，this intensive outdoor thermal condition may potentially impact the final permance energy flexibility strategies.

3 Results

In this section，we compare the overall energy consumption characteristics in different climate zones in order to reveal the load shifting/shedding potential the three flexibility management strategies （as in Fig.3）. The prototype building’s baseline HVAC operation power loads in the five cities are shown in Fig. 8. The patterns the HVAC power loads were similar，experiencing a surge at the beginning reaching stable level during the occupied period. Since this study mainly investigates flexibility during the on-peak period，the focusis on the HVAC opera tion management during the occupied period （08： 00 to 18：00）.

In the baseline cases，the peak power load HVAC systems mainly appeared in afternoons （the on-peak period），though a slight decrease in HVAC power load was observed at noon due to lower occupancy.Fig. 9 compares the differences in peak power load time，peak load， energy consumption the 6 total climate conditions. For Guangzhou，Beijing， Harbin，peak power loads mostly appeared between 12：O0 to 17：O0，with the median timelingaround15：OO. Thepeak loadsin the extreme condition occurred slightly earlier， that Kunming was later.The daily peak power loads the HVAC systems were about 1000kW in Guangzhou Beijing， which were higher than those Harbin Kun ming. The extreme condition presented thehighestpeakloadsat 1157.21kW .Thetrend daily HVAC energy consumption in the five cities was similar to the trend peak power load in the cooling season.

3.1The flexibility pre-cooling strategy

Thepre-coolingstrategysetsthe indoor temperature setpoint 2^°C loweruntil13：00，then 2^°C higher from 13：00 to 18：00 on working days. Fig. 10 shows the corresponding seasonal average HVAC power load difference between the pre-cooling baselinecases inthesixclimateconditions.

Based on loads inFig. 1O，with increased power load，the pre-cooling strategy reduced the HVAC power load during the on-peak period （l4：OO—18：00） in all climate conditions，but the flexibility capacity pre-cooling strategy varied from city to city.

Based on Equations （l）-（3），the results three energy flexibility indicators （peak load reduction rate， energyconsumptionreductionrate，theratiodifference peak load during the on-peak period） are shown in Table 4. The Pre-cooling strategy largely reduced the peak load during the on-peak period without adversely increasing HVAC energy consumption.Additionally，the pre-cooling strategy’s aver age peak load reduction was 15% to 23% ，which was lower when the outdoor climate was warmer. With lower outdoor temperatures during the cooling season，the pre-cooling strategy in Kunming Har bin cities showed higher energy flexibility with better load shedding capacity，though they also had higher energy consumption. Although the HVAC power load in the extreme condition was higherthan that the typical conditions，the average dai ly power load reduction rate the energy consump tion reduction rate showed no significant difference.

Although the pre-cooling strategy effectively reduced shifted the peak load，the proportion indoor PMV in the comtable range （-0.5lt; PMVlt;0.5） dropped significantly due to the cooler outdoor climate （Fig. 1l），Kunming experienced an increased proportion PMV being lower than -0.5 ，but an increased proportion （about 40% ） “slightly warm” sensations were observed in other conditions. However，an increase in the possibility anunacceptable thermalenvironment was observed in the extreme condition that was 12% higher than that the typical condition.

3.2 Theflexibilitythezonetemperaturereset strategy

The zone temperature reset strategy reset the indoor temperature setpoint in occupied areas from 26^°C to 28°C 5 hours from 13：OO on workdays. Fig.12 shows the seasonal average HVAC power load difference between the zone-temperature reset baseline cases in the six climate conditions.We can see that the power load decreased after 13：00 when the zone temperature was reset to 28^°C ·

Unlike the pre-cooling strategy，the zone tempera ture reset strategy was able to provide energy savings alongside peak power load-shedding capacity.

Similarly，although the zone temperature reset strategy presented a higher energy conservation po tential，ithadalowerpeakload reduction rate.The average energy consumption reduction rate peak load reduction rate the zone temperature reset strategy were about 4.46% 12.13% the baseline energy consumption peak load.Compared to the typical condition case， the energy saving potential in the extreme condition was slightly lower but slightly higher in terms peak load reduction rate.

The three energy flexibility indicators the zonetemperature reset strategy are shown in Table 5.The zone temperature reset strategy reduced peak load by 11% to 19% reduced peak load during the on-peak period by about 70% to 96% ．The reduction rate peak load during the on-peak period was generally positively related to the outdoor climate conditions cooling degree days （CDD）， except Kunming，which had a cooler climate. For this reason，the highest reduction rate in energy peak load was found in Kunming. The zone temperature reset strategy also showed higher energy conservation potential with its warmer indoor temper atures 28°C ）. Again，the warmer the outdoor climate condition，the less potential load shifting shedding.

The zone temperature reset strategy kept the occupied zone's thermal environment mostly within the acceptable PMV range -1

3.3The flexibility the partial shutdown strategy

The partial shutdown strategy partially closed some occupied building zones to compensate an emergency power shortage. In particular，we investigated the load shifting shedding potential HVAC systems by shutting down the occupied zones facing west （zones 7 8 in Fig.2）. Fig. 14 pres ents the seasonal average HVAC power load differ ence between the partial shutdown baseline cases in the six climate conditions.

The power load HVAC systems was reduced after 13： OO when the partial shutdown started. The average amount power load reduction was generally higher in GuangzhouBeijing， extreme condition due to their warmer climate conditions.

The three energy flexibility indicators the partial shutdown case are shown in Table 6. Asa local control strategy only，the peak load shedding load shiftingabilitywere lowerthan in the other two strategies. The average peak load reduction rate the six conditions was around 10% ， the reduction rates peak load occurring during the on peak period were around 50% Guangzhou， Beijing.

Theindoor thermal environmentin the occupied areas remained nearly unchanged under the partial shutdown case， benefitingfrom theunchanged indoor temperature setpoint in the occupied areas duringthe on-peak period.As shown in Fig.15，the PMV value was within the comtable range （204 （-0.5

4 Discussion

Building energyflexibilityisimportant regional power managers to balance the power load supply dem during the on-peak period the cooling season. As the major energy consumers power load sources during this time，HVAC systems in fice buildings have been recognized their positive impact on load shifting shedding potential under various flexibility management strategies. However，the flexibility effectiveness HVAC systems can be influenced by the outdoor climate. Forthis reason，in this studywe used typical weather data five climates in extreme weather data one them to simulate the energy flexibility an fice building under three short-term HVAC management strategiesinorderto evaluatethemwith respect to these various different climates specifically.

4.1 Loadflexibilitycapacityclimateconditions

Forthe typical climate conditions，although the cooling dem（determinedby CDD）varied from city to city，all three short-term HVAC management strategieswere effective to somedegree in all climate conditions. Taking advantage a building’s thermal mass，the pre-cooling strategy had the highest peakload reduction shifting potential. However，due to a lower temperature setpoint during the f-peak period，the pre-cooling strategy can lead to increased overall energy consumption during the occupied period in all climate conditions. The zone temperature reset partial shutdown strategies hadhigher energy-saving potential but slightly lower peak-load reduction rates. The lowest load-shifting was found in the partial shutdown strategy.

Generally， the peak-load reduction rate was found to be highest in Kunming，which had cooler outdoor conditions lower in Guangzhou， Chongq ing， Beijing，which had warmer conditions. Fig.l6 shows the relationship between outdoor drybulb temperature the peak-load reduction rate in five typical climate conditions. Although variation wasobserved in the correlation，the peak load reduction rate the pre-cooling zone temperature reset strategies generally presented a negative linear relationship with outdoor temperature. Interestingly， morevariations in the peak-load reduction rate were observed in the cooler outdoor environment，whereas the peak load reduction capacity was more convergent in warmer conditions. The peak-load shedding potential the partial shutdown strategy was less correlated with the outdoor temperature，but it also presented a larger variation in the cooler climate condition compared to thewarmerconditions.

Based on the analysis in section 3.3，the daily peakload time in the extreme condition was found to be earlier than that the typical conditions. The average peak-load time median peak-load time were 12：52 12：OO in the extreme condition， which was about 2 hours 4 hours earlier than thetypical climate condition. This early peak-load time resulted in a considerably higher peak load reduction rate during the on-peak period all three flexibility management strategies （Table 4 5）. Fig. 17 shows the correlation between outdoor dry-bulb tem perature peak-load reduction rate the three flexibility management strategies in the extreme condition. Since the outdoor temperature in the extreme condition was mostly between 30^°C to 38^°C ，the peak-load reduction potential showed relatively less variation. One the reasons could be the higher so lar radiation less cloud cover in the extreme condition dramatically increasing the solar heat gain. This may indicatea difference in the effectiveness building energy flexibility management strategies under extreme normal weather conditions，which emphasizes the importance using real weekly monthly weather ecast data （not only the outdoor temperature but also solar radiation） into considerationwhen assessing energy flexibility.

4.2 Indoor thermal comt under flexibility management

The primary aim HVAC systems is to pro vide occupants with a comtable，or at least acceptable indoor thermal environment. Asa hy approach to managing building flexibility， typical flexibility management strategies ten integrate HVAC control with a building's passive thermal storage capacity （building envelope indoor furniture）in order to adjust the indoor temperature setpoints to the upper limit the acceptable temperature during the on-peak period. However， this could potentially result in an unpleasant indoor thermal environment.

This study evaluated indoor thermal comt by simulating the hourly PMV valueunder three flexibility management strategies. In typical climate conditions，applying pre-cooling zone temperature reset strategies kept the indoor thermal environment within the acceptable range，increasing the possibility feeling“warm”in the occupied zones by 30% 40% .With the occupied zones being barely influenced，the partial shutdown strategy permed betterin keeping the occupied zone comtable，espe cially in the extreme condition.

In the extreme condition，although the pre-cooling zone temperature reset strategies presented better potential peak load overall energy reduction，a large risk wasobserved in creatingan un comtable indoor environment （Fig. 1l 13）. Again，the main reason could be the strong solar radiation，which led to the setpoint setback （ 28^°C barely being able to stabilize the indoor environment within the expected range. However，taking advantage not influencing the HVAC operation in the occupied zones，the partial shutdown strategy simultaneously provided a reasonable peak-load shifting shedding capacity sacrificed less thermal comt in the occupied zone，which might be a useful flexibility management approach buildings in the extremely hotareas.

As a result， the indoor temperature setpoint was ten set to be the upper limit to decrease building cooling dem boost energy flexibility，especiallyon the hot summer days when regional power dem was higher. As reported by this study，the peak-load reduction capacity was about 10%-25% in different climate conditions，but this reduction rate would become lower if we considered thermal comt，especially in extreme conditions in which buildingenergy flexibility management is in high dem. Theree，although load shifting shedding capacity were regarded as the main indicators to assess building energy flexibility，itis important to include a specific indicator or assessment on indoor thermal comt to figure out actual usable load flexibility.

4.3 Limitationsrecommendationsfuture work

This study has several limitations.First，the building modeling data was obtained from a real fice building property， solwe recommend that future work compare simulated peak load flexibility with actual measurement data to validate our results the reported results.Second，since the indoor ther mal comt was beyond the acceptable range under passive management strategies in extreme weather conditions，we recommend that future work focus on developing comt-determined building flexibility management strategies.

5 Conclusion

To evaluate the energy flexibility three typical short-term HVAC flexibility management strategies （pre-cooling，zone temperature reset， partial shutdown） in different climate conditions ， this study simulated the HVAC power load shifting shedding potential during the cooling season （July August） an actual prototype fice building in five major climate conditions one extreme climate condition. Three building energy flexibility indicators were used to assess the peakload shifting/shedding capacity overall energy permance. Indoor thermal comt wasalso evaluated each flexibility management strategy underdifferentclimate conditions.The main conclusions are as follows.

The building peak load flexibility overall energypermance the three short-term HVAC flexibility management strategies were impacted by outdoor climate conditions. The peak-load reduction ratewas foundto be highestin Kunming，which had the coolest outdoor conditions lower in Guangzhou Beijing，which had warmer outdoor conditions.

All three short-term HVAC management strategieswere effectiveto somedegree in allclimate conditions. The pre-cooling presented strategy the best peak-load shifting shedding capacity between 15% 30% ， but there was a risk more energy consumption. The partial shutdown strategy had the lowest loadshifting shedding ability， which was between 8% 13% ，but itwasable to maintain a comtable thermal environment in the occupied zones.

The peak load reduction rate pre-cooling zone temperature reset strategies generally declined linearly with increases in outdoor temperature but weremore convergent inwarmer conditions.

In the extreme summer condition， the daily peakload timewas found to be over two hoursearlier than that the typical conditions. The correlation between peak load reduction outdoor tempera ture was weak in the extreme condition as well， which may have resulted from intensive solar radiation.

The short-term HVAC flexibility management strategies that integrate temperature setpoint adjust ment with passive thermal storage may lead to a risk maintaininga comtable indoor thermal environ ment，especially in the extreme summer condition with intensive solar heat gain，in which the possibili ty uncomtable thermal environments could be up to 81% ·

Thisstudy demonstrated the effectiveness short-term HVAC flexibility management strategies under different climate conditions， future work could focus on developing comt-determined building flexibility management strategies validating loadshifting/shedding capacity against measured building operation data.

References

[1]CURTINJ，MCINERNEYC，GALLACHOIRBO， et al.Quantifying string risk fossil fuel assets implications renewableenergy investment：A review the literature [J].Renewable Sustainable Energy Reviews，2019，116：109402.

[2]Race to zero campaign [EB/OL].[2023-03-2O]. https：// www.climatechampions.net/campaigns/race-to-zero/.

[3]JENSEN S O，MARSZAL-POMIANOWSKA A， LOLLINIR，et al.IEA EBC annex 67energy flexible buildings [J]. Energy Buildings，2O17，155： 25-34.

[4］KEENEYKR，BRAUNJE.Application building precooling to reduce peak cooling requirements [C]// American Society Heating，Refrigerating AirConditioning Engineers （ASHRAE）Winter Meeting， Philadelphia，PA（UnitedStates），24-28Feb.，1997.

[5]NUYTTENT，CLAESSENSB，PAREDISK，et al. Flexibility a combined heat power system with thermal energy storage district heating[J].Applied Energy，2013，104：583-591.

[6]LUNDPD，LINDGRENJ，MIKKOLAJ，etal.Re view energy system flexibilitymeasures to enable high levelsvariable renewable electricity [J].Renewable Sustainable Energy Reviews，20l5，45：785-807.

[7]AMASYALI K，EL-GOHARY NM.A review datadriven building energy consumption prediction studies [J].Renewable Sustainable Energy Reviews，2018， 81： 1192-1205.

[8]LU F，YU Z Y，ZOU Y，et al.Energy flexibility assessment a zero-energy ice building with building thermal mass in short-term dem-side management [J]. Journal Building Engineering，2O22，5O： 104214.

[9]SHAN K，WANG S W，YAN C C，et al. Building dem response control methods smart grids： A review [J]. Technology the Built Environment，2016，22（6）： 692-704.

[10] ADUDA K O，LABEODAN T，ZEILER W，et al. Demside flexibilitycoordination in fice buildings： A framework case study application [J]. Sustainable Cities Society， 2017， 29： 139-158.

[11] WANG LP，MATHEWP，PANG XF.Uncertainties in energy consumption introduced by building operations weather a medium-sizeice building[J]. Energy Buildings，2012，53：152-158.

[12] CHEN Y B，CHEN Z，XU P，et al. Quantification electricity flexibility in dem response： Ofice building case study [J].Energy，2019，188：116054.

[13] MALIK A， HAGHDADI N， MACGILL I， et al. Appliance level data analysis summer dem reduction potential from residential air conditioner control [J].Ap plied Energy，2019，235： 776-785.

[14] BODE J. Measuring short-term air conditioner dem reductions operations settlement [R/OL]. [2023- 03-23]. https：//escholarship.org/uc/item/50v9d6xc.

[15] WANG S W， TANG Supply-based feedback control strategy air-conditioning systems direct load control buildings responding to urgent requests smart grids [J]. Applied Energy，2017，2Ol： 419-432.

[16] OLIVIERI S J，HENZE G P，CORBIN C D，et al. Evaluation commercial building dem response potential using optimal short-term curtailment heating， ventilation， air-conditioning loads [J]. Journal Building Permance Simulation，2014，7（2）： 100-118.

[17] WANG HL，WANG S W， TANG Development grid-responsive buildings： Opportunities，challenges，capabilities applications HVAC systems in non-residential buildings in providing ancillary services by fast dem responses to smart grids [J]. Applied Energy， 2019，250：697-712.

[18] YIN R X，KARA EC，LI Y P，et al. Quantifying flexibility commercial residential loads dem response using setpoint changes [J]. Applied Energy， 2016，177： 149-164.

[19] KISHORE R A，BIANCHI M V A，BOOTENC，et al. Modulating thermal load through lightweight residential building walls using thermal energy storage con trolled precooling strategy [J]. Applied Thermal Engi neering，2020，180：115870.

[20] XU P，HAVES P，PIETTE MA，et al. Peak dem reduction from pre-cooling with zone temperature reset in an fice building [C]// 2004 ACEEE Summer Study on Energy Efficiency in Buildings，Pacific Grove，CA （US），23-27 August， 2004.

[21] JIANGZH，RISBECKMJ，RAMAMURTIV，et al. Building HVAC control with reincement learning reduction energy cost dem charge [J].Energy Buildings，2021，239：110833.

[22] HU M M，XIAO F，WANG L S. Investigation dem response potentials residential air conditioners in smart grids using grey-box room thermal model [J]. Applied Energy，2017，207： 324-335.

[23] STOPPS H，TOUCHIE M F.Load shifting energy conservation using smart thermostats in contemporary high-rise residential buildings：Estimation runtime changes using field data [J]. Energy Buildings， 2022，255：111644.

[24] TURNER WJN，WALKERIS，ROUXJ.Peak load reductions：Electric load shifting with mechanical precooling residential buildings with low thermal mass [J].Energy，2015，82： 1057-1067.

[25]UMMENHOFER C C，MEEHL G A.Extreme weather climate events with ecological relevance：A review[J].Philosophical Transactions the Royal Society B： Biological s，2017，372（1723）： 20160135.

[26] STOTT How climate change affects extreme weather events [J]. ，2016，352（6293）： 1517-1518.

[27]DAI L K，LI ZQ，CHEN X Y，et al. Usage behavior characteristics household air-conditioners during the extremely hot summer： A case study [J]. Building Environment，2023，234： 110160.

[28]MAJDALANI N，AELENEI D，LOPESR A，et al. The potential energy flexibility space heating cooling in Portugal[J].Utilities Policy，2O20，66： 101086.

[29] JUNKER R G，AZAR A G，LOPES R A，et al. Characterizing the energy flexibility buildings districts[J]. Appied Energy，2018，225： 175-182.

[30] LIU M Z，HEISELBERG Energy flexibility a nearly zero-energy building with weather predictive control on a convective building energy system evaluated with diferent metrics [J].AppliedEnergy， 2019，233/234： 764-775.

[31] VANHOUDT D，GEYSEN D，CLAESSENS B，et al. An actively controlled residential heat pump： potential on peak shaving maximization self-consumption renewable energy [J].Renewable Energy，2Ol4，63： 531-543.

[32]CUIY，YAND，HONG TZ，et al.Comparison typical year multiyear building simulations usinga 55-year actual weather data set from [J].Applied Energy，2017，195：890-904.

[33]THOMPSONR，HORNIGOLDR，PAGEL，etal. Associations between high ambient temperatures heat waves with mental health outcomes：A systematic review[J].PublicHealth，2018，161：171-191.

[34] General code energy efficiency renewable energy application in buildings：GB 55015—2021 [S].Beijing： Architectureamp;.BuildingPress，2021.

[35]TANG H，WANG S W，LI H X.Flexibility categorization，sources，capabilities technologies energy-flexible grid-responsive buildings： State-the-art future perspective [J].Energy，2O21，219： 119598.

（編輯胡英奎）