| As urbanization accelerates,energy consumption and greenhouse gas emissions are also significantly increasing.Building operational energy use accounted for 21%of the total energy consumption in China,building energy efficiency is crucial for sustainable urban development.The government formulated the 14th Five-Year Plan for Building Energy Conservation and Green Building Development in 2022,vigorously promoting the energy-efficient renovation of existing buildings and the application of solar photovoltaic(PV)to develop green and low-carbon cities.Therefore,in order to understand urban-scale building energy demands in the current stage,and quantify the impact of different energy conservation measures(ECMs)measures on reducing urban building energy consumption,it is necessary to establish physics-based urban building energy models.Numerous foreign scholars have conducted research on urban building energy consumption and developed urban building energy modeling(UBEM)tools to simulate energy use and evaluate energy-saving potential.However,most studies are limited to a small number of buildings at a neighborhood scale or single buildings due to the absence of rich urban building datasets in China.Therefore,this paper starts from data-algorithm-tool and finally develops an UBEM tool suitable for Chinese building characteristics.Taking Changsha City as a case study,the tool automatically generates urban building energy models based on geographic information system(GIS)data,calculates dynamic energy demand under standard operating conditions,and evaluates the energy-saving potential of different ECMs.The main contents of this paper are as follows:(1)The method of obtaining GIS data for urban buildings was investigated.Firstly,68,966 vector building footprints from five districts of Changsha City were collected from Amap,including the information of number of floors.To solve the problem of difficulty in obtaining direct data on urban-scale building type and year built in China,this paper proposed the method of using unsupervised clustering and supervised classification to identify building types,based on GIS data such as building footprint,point of interest(POI),and community boundary.The results showed that 95.6%of building types were successfully identified,and the total residential floor area differed by only 6.9%compared to data from the Changsha Statistical Yearbook.Additionally,a method based on historical satellite image data was proposed to automatically extract building footprints from different eras using convolutional neural network(CNN)algorithms,with an average precision of 80%.Then,the year built of each building was inferred using intersection analysis.(2)Archetype buildings were developed based on GIS data.According to the identified building type and year built,22 building types in three construction periods are defined as archetype buildings,including residential,office,hotel,commercial mixed-use buildings,etc.,representing 87.4%of the total floor area.Then,referring to national building energy standards,parameters such as envelope,internal loads,and heating,ventilation and air conditioning(HVAC)systems were determined to establish an archetype building database.Open Studio-Standards was used to automatically generate Energy Plus models for 66 archetype buildings.Based on the energy use intensities of the archetype buildings,the annual energy use of 59,332 buildings was quickly calculated to get 13,864 GWh for electricity use and 23.6×10~6 GJ for natural gas use.The results were within a reasonable range compared to statistical data.(3)An UBEM tool Auto BPS was developed.This tool was based on Ruby language programming and Open Studio SDK,using Energy Plus as the simulation engine,then automatically generated urban building energy models based on GIS dataset,considering the shading effect by surrounding buildings.A downtown district of 3633 buildings in Changsha was selected as a case study to calculate urban building energy use and analyze energy retrofit and rooftop PV potential.The results showed that the annual total energy use was 1793 GWh,which differed by only 5.6%compared to the results of the archetype aggregation method.The median percentage difference in energy use was 15%at the individual building level.Three individual ECMs and packages were adopted.When considering a single ECM,window upgrade was most energy-efficient for residential buildings,while lighting upgrade was that one for commercial buildings.When combining all ECMs,annual energy savings ranged from2.7%to 30.7%at the individual building level,with a median of 17.2%.Regarding the total annual energy reduction,in the three individual ECMs,the cooling system upgrade involved the smallest annual energy use reduction at 4.6%,while lighting upgrade had the largest at 9%.Combining all three ECMs,the total building energy saving percentage was 18.5%,which could reach a maximum of 38.6%after the rooftop PV system was installed.(4)The UBEM calibration method was studied,and the impact of climate change on the energy performance of calibrated models was evaluated.Due to the lack of publicly available metered energy consumption data for individual buildings in China,483 residential buildings with annual metered data in Geneva,Switzerland,were selected for modeling.A fast and automatic calibration method was proposed,which used Latin Hypercube Sampling to obtain 1000 combined parameter samples and conducted 1000 Energy Plus simulations based on archetype buildings.By learning from the energy performance of archetype buildings,it took less than ten simulation runs per building to calibrate a building model,with less than a 5%error.Four future weather files were generated using an open-source tool to simulate building energy use under the latest Shared Socioeconomic Pathways(SSP)scenarios.The results showed that in this temperate climate zone,changes in cooling demand were much greater than those in heating demand under all four scenarios.In Neighborhood 1(NBH 1)and Neighborhood 2(NBH 2),there was a decrease of 22-31%and 21-29%for heating demand,an increase of 113-173%and 95-144%for cooling demand by 2050.The heating and cooling demand change were greater for individual buildings in the dense urban area(NBH 1)than sparse suburban area(NBH 2).The average annual heating intensity dropped from 81 k Wh/m~2 in the current typical climate to 57 k Wh/m~2 in the SSP5-8.5,while the cooling intensity rose from 12 k Wh/m~2 to 32 k Wh/m~2.In NBH 1,the peak heating demand of the coldest winter day was reduced by 11-18%,and the peak cooling demand of the hottest summer day increased to 24-34%under four scenarios.At last,the potential of ECMs in mitigating the impact of climate change on building energy use was examined.The results suggested that window upgrade had the greatest potential to reduce cooling demand as a single ECM,with an average reduction of 22%across all four SSP scenarios.Meanwhile,external wall insulation was found to be the most effective measure for reducing heating demand,with an average reduction of 28.6%.Considering the limited impact of roof insulation on energy saving,a combined ECM(window upgrade+external wall insulation)may be more cost-effective to reduce the total heating and cooling demand. |
PDF Full Text Request