Impact on Renewable Design Requirements of Net-Zero Carbon Buildings Under Potential Future Climate Scenarios

Article Impact on Renewable Design Requirements of Net-Zero Carbon Buildings under Potential Future Climate Scenarios

Dongsu Kim 1, Heejin Cho 2,* , Pedro J. Mago 3, Jongho Yoon 1 and Hyomun Lee 1

1 Department of Architectural Engineering, Hanbat National University, Daejeon 34158, Korea; [email protected] (D.K.); [email protected] (J.Y.); [email protected] (H.L.) 2 Department of Mechanical Engineering, Mississippi State University, 210 Carpenter Engineering Building, P.O. Box 9552, Mississippi State, MS 39762, USA 3 Department of Mechanical and Aerospace Engineering, West Virginia University, Morgantown, WV 26506, USA; [email protected] * Correspondence: [email protected]; Tel.: +1-662-325-1959

Abstract: This paper presents an analysis to foresee renewable design requirement changes of net- zero carbon buildings (NZCBs) under different scenarios of potential future climate scenarios in the U.S. Northeast and Midwest regions. A climate change model is developed in this study using the Gaussian random distribution method with monthly temperature changes over the whole Northeast and Midwest regions, which are predicted based on a high greenhouse gas (GHG) emission scenario (i.e., the representative concentration pathways (RCP) 8.5). To reﬂect the adoption of NZCBs potential in future, this study also considers two representative future climate scenarios in the 2050s and 2080s of climate change years in the U.S. Northeast and Midwest regions. An ofﬁce prototype building model integrates with an on-site photovoltaics (PV) power generation system to evaluate NZCB performance under the climate change scenarios with an assumption of a net-metering electricity

purchase agreement. Appropriate capacities of the on-site PV system needed to reach NZCB balances are determined based on the building energy consumption impacted by the simulated climate scenarios. Results from this study demonstrated the emission by electricity consumption increases Citation: Kim, D.; Cho, H.; Mago, P.J.; Yoon, J.; Lee, H. Impact on as moving toward the future scenarios of up to about 25 tons of CO2-eq (i.e., about 14% of the total Renewable Design Requirements of CO2-eq produced by the electricity energy source) and the PV installation capacity to offset the Net-Zero Carbon Buildings under emission account for the electricity consumption increases significantly up to about 40 kWp (i.e., up Potential Future Climate Scenarios. to more than 10% of total PV installation capacities) as the different climate scenarios are applied. Climate 2021, 9, 17. https://doi.org/ It is concluded that the cooling energy consumption of office building models would significantly 10.3390/cli9010017 impact GHG emission as future climate scenarios are considered. Consequently, designers of NZCBs should consider high performance cooling energy systems in their designs to reduce the renewable Received: 27 December 2020 energy generation system capacity to achieve net-zero carbon emission goals. Accepted: 17 January 2021 Published: 19 January 2021 Keywords: net-zero carbon building; climate change; EnergyPlus; building energy modeling; office building; photovoltaics (PV) Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affil- iations. 1. Introduction The building sector accounts for approximately 40% of the total end-use energy consumption (i.e., primary and secondary energy usages) in the United States [1], including about 21% and 18% for residential and commercial buildings, respectively. In the commer- Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. cial building sector, about 74% of the energy consumption was captured for the electric This article is an open access article power sector to generate and supply electricity retails. The U.S. Energy Information Ad- distributed under the terms and ministration (EIA) also estimated that electricity use in the commercial building sector was conditions of the Creative Commons expected to increase by about 14% by 2050 compared to levels in 2019 [2]. This growth has Attribution (CC BY) license (https:// led to concerns regarding the future increased energy consumption and human emission creativecommons.org/licenses/by/ of carbon dioxide (CO2) from the building sector. According to the Congressional Research 4.0/). Service reports [3,4], it was pointed out that the electric power sector contributed the

Climate 2021, 9, 17. https://doi.org/10.3390/cli9010017 https://www.mdpi.com/journal/climate Climate 2021, 9, 17 2 of 20

second-largest percentage, 35% of U.S. CO2 emission contributions based on the 2017 U.S. CO2 emission contribution data by sectors. Within the building sector, the commercial sector presented more than 37% of the electric power sector’s CO2 contributions. Although the amount of electricity power generation had remained flatted after 2010, CO2 emissions continued a general trend of reduction after 2007 mainly because of increased renewable sources and natural gas use as well as reduced coal use in electricity generation, resulting in a 20% reduction of CO2 emissions when compared to the CO2 emissions levels of 2005 [3,4]. However, there are still concerns regarding greenhouse gas (GHG) emission and global warming issues as the global growth rate of GHG emission from human activities has still increased [5–7]. As an effort to reduce GHG emission, the concept of net-zero carbon emission buildings (NZCBs) has been developed, and the idea has been gaining significance globally to achieve feasible long-term goals of GHG emission reductions for the coming decades [8]. To appropriately implement the NZCB concept, high energy-efficient appliances and systems must be encouraged [9], and an appropriate selection and design of on-site renewable energy systems must be incorporated to effectively reduce both on-site and central generation of GHG emission [10,11]. Louwen et al. [12] presented a good overview of the impact of GHG emission by PV-based renewable energy generation. They reviewed over 40 years of PV development and analyzed energy demand and GHG emission impacts in terms of PV production. Their results illustrated that when the doubling of installed PV capacity occurred, energy consumption and GHG emission could be decreased by about 13% and 17–24%, respectively. Pinel et al. [13] evaluated the impact of allowing to buy CO2 compensation to enable the zero emission neighborhoods design. They pointed out that there were large differences in energy system design and GHG emission generation depending on variations in weather conditions and the price of external compensation. Researchers have conducted numerous studies on climate change impacts on building energy consumption. Wan et al. [14] investigated building energy use variations based on climate change scenarios in subtropical climates. They demonstrated that improving efficiencies in building components, such as lighting and chiller, can provide potential mitigation for increased future energy use due to climate changes in the late 21st century. Nik and Kalagasidis [15] assessed possible climate changes and uncertainties in future energy performance of the residential building stock in Stockholm. Four uncertainty factors (i.e., global climate models, regional climate models, emissions scenarios, and initial conditions) were considered in their study, and they pointed out that all their climate change scenarios up to 2100 showed that the future heating demands would be decreased by about 30% compared to the date before 2011 while the cooling demand would increase mainly because of increased outdoor air temperatures. Kikumoto et al. [16] constructed future standard weather data for use in building design and energy analysis for 2030s (i.e., 2031–2035). Based on their study, it was observed that there would be a substantial impact of climate change on the energy performance of a detached house, including 15% increases in sensible heat loads predicted under their study condition. Shen [17] evaluated the energy consumption variation of residential and commercial buildings by climate change scenarios in four representative cities in the U.S. using EnergyPlus. They showed that climate changes could have potential impacts on residential and office building energy usages during the years 2040–2069, including −1.64% to 14.07% and −3.27 to −0.12% of annual energy changes for residential and office buildings, respectively. Wang et al. [18] presented energy simulation-based analysis to investigate impacts of climate change on building energy consumptions in U.S. climate zones. The EnergyPlus medium office prototype model was used to predict energy consumption with two different climate change models, including Hadley Centre Coupled Model version 3 and NCAR Community Earth System Model version 1. With their generated future weather files, they examined that climate change mitigation measures related to heating, ventilation, and air conditioning (HVAC) operations such as adjustment of thermostat setpoints and reduced HVAC operation hours. In addition, Summa et al. [19] compared energy consumption of a Climate 2021, 9, 17 3 of 20

current residential NZEB in Rome, Italy with that expected in 2050 under Representative Concentration Pathway (RCP) 4.5 and 8.5 scenarios. Their results showed that about 50% of increases in cooling energy use were presented, while heating energy use showed significant decreases under the future climate changes. Although there have been several pieces of research investigating building energy use due to climate changes, there is still a lack of studies understanding the adoption of NZCBs associated with future potential climate changes in a quantitative manner in the literature [20]. To fill this research gap, this paper investigates the impact of climate changes on commercial building energy consumption and carries out an analysis to foresee renewable design requirements of NZCBs under different scenarios of future climate changes in the U.S. Northeast and Midwest regions. The medium-sized office prototype building model, which complies with a widely and globally accepted commercial building energy standard (i.e., ASHRAE Standard 90.1-2019 [21]), is subjected in this study to evaluate NZCB design requirements under different climate change scenarios. The focus of this paper is to provide useful insights into variations in building energy consumption and renewable design requirements when office buildings are subjected to a net zero carbon design under the current typical meteorological year (TMY) data and future climate scenarios. It is important to note that this paper does not investigate different strategies for enabling net zero carbon operations to commercial buildings, e.g., advanced building envelope and HVAC technologies to reduce building energy consumption.

2. Methodology 2.1. Description of Climate Change Model This section describes the methodology to generate climate change data for building energy simulations. Various weather conditions, including dry-bulb temperature, wet-bulb temperature, solar radiation, relative humidity, and wind speed/direction, can impact the building thermal load. Such factors can be considered in comprehensive building energy modeling and prediction to estimate energy consumption in a building. Typical meteorological year 3 (TMY3) weather data [22], which is statistically derived annual climate data, is used in the building energy simulations of this study. Our previous study [23] revealed that the ambient temperature is a dominant contributing factor to the thermal load during both heating and cooling seasons among all weather variables. Therefore, predicted changes in the dry-bulb temperature of outdoor air are assumed to be a single dominant variable for the climate change model in this study. A stochastic method using the normal distribution with a random sampling technique is applied to generate future weather proﬁles for building energy simulation. The probability density function of the normal distribution is deﬁned as:

2 1 − (x−µ) p(x) = √ e 2σ2 (1) 2πσ2 where µ is the mean value and σ is the standard deviation, and the standard deviation is assumed to be one (1) in this study. As the value of σ increases, the range of potential random errors also increases, therefore, this study considers appropriate distribution patterns by assuming the standard deviation of the predicted dry-bulb temperature to be 1 ◦C to ensure that the adjacent hourly temperature does not deviate more than 2 ◦C. The mean values of the predicted dry-bulb temperature are determined based on the study conducted by Byun et al. [24]. The Byun’s model [24] was developed based on hydrologic modeling experiments over 20 Midwestern watersheds using the Variable Inﬁltration Capacity (VIC) model forced by historical observed datasets and future projections from the statistically downscaled Global Climate model (GCM). Their model provides monthly projected changes in hydrometeorolgical variables over the whole Midwest and Northeast regions, including precipitation, outdoor air temperature, evapotranspiration, and soil moisture. The projected monthly temperature changes from the Byun’s model are used to generate future Climate 2021, 9, x FOR PEER REVIEW 4 of 20

Climate 2021, 9, 17 4 of 20 changes in hydrometeorolgical variables over the whole Midwest and Northeast regions, including precipitation, outdoor air temperature, evapotranspiration, and soil moisture. The projected monthly temperature changes from the Byun’s model are used to generate futureweather weather ﬁles for files all for selected all selected locations locations in this in this study. study. Figure Figure1 presents 1 presents the the temperature temperaturechanges changes of the of the Byun’s Byun’s model model for for two two representative representative scenariosscenarios of future future study study periods periods (i.e., 2050s and and 2080s). 2080s). The The future future weather weather pr projectionsojections in in Byun’s Byun’s study study are are based based on on two two GHG concentration concentration scenarios, scenarios, including including Repr Representativeesentative Concentration Concentration Pathway Pathway (RCP) (RCP) 4.5 4.5 and 8.5 for for medium medium and and high high cases, cases, respectively. respectively. The The two two scenarios scenarios are are based based on on 30-year 30-year windows centered on on the the 2050s 2050s (2041–2070) (2041–2070) and and 2080s 2080s (2071–2100) (2071–2100)..

Figure 1. 1. MonthlyMonthly projected projected changes changes in the in theoutdoor outdoor air temperature air temperature [24]. RCP: [24]. Representative RCP: Representative concentration pathways. pathways.

TheThe RCP RCP 8.5 8.5 model’s model’s projection projection shown shown in in Figure Figure 11 isis selected selected in in this this study study to to generate generate random distributions distributions of of temper temperatureature changes changes with with the the scenarios scenarios with with an anassumption assumption of a of high-emissiona high-emission scenario. scenario. Monthly Monthly mean mean values values of outdoor of outdoor air temperature air temperature changes changes are de- are termineddetermined from from the theByun’s Byun’s RCP RCP 8.5 model 8.5 model shown shown in Table inTable 1. These1. These mean meanvalues values are used are inused the in probability the probability density density function function in Equation in Equation (1) to (1) generate to generate random random distributions distributions of hourlyof hourly temperature temperature variations variations for each for eachclimat climatee change change scenario scenario in different in different climate climate locations.locations. Table Table 1 summarizes1 summarizes monthly monthly mean mean valu valueses of outdoor of outdoor air temperature air temperature changes changes ex- tractedextracted from from the the Byun's Byun’s RCP RCP 8.5 8.5model model..

Table 1. MonthlyTable mean 1. Monthly values mean (μ) of values outdoor (µ )air of temperature outdoor air temperaturechanges obtained changes from obtained [24]. from [24].

Jan. Feb. Mar. Jan.Apr. Feb.May Mar. Jun. Apr. MayJul. Jun. Aug. Jul.Sep. Aug.Oct. Sep. Nov. Oct. Nov.Dec. Dec. 2050s 3.9 3.1 2050s2.2 3.92.4 3.13.2 2.23.1 2.4 3.5 3.2 3.14.2 3.54.0 4.24.0 4.03.2 4.0 3.24.8 4.8 2080s 6.0 5.8 2080s4.1 6.04.0 5.85.0 4.15.8 4.0 6.1 5.0 5.87.5 6.17.2 7.56.3 7.25.3 6.3 5.36.3 6.3 Three locations, including New York, NY, Buffalo, NY, and Rochester, MN, in the U.S. NortheastThree locations, and Midwest including regions New are York, chosen NY, Buffalo, to evaluate NY, the and impact Rochester, of climate MN, in change the U.S. onNortheast the building’s and Midwest energy use. regions Those are locations chosen are to evaluateselected because the impact they of can climate represent change three on differentthe building’s climate energy zones use. in the Those U.S. locationsNortheast are and selected Midwest because regions they defined can represent by ASHRAE three 90.1-2019,different climatei.e., climate zones zones in the 4A, U.S. 5A, Northeastand 6A, which and can Midwest show variations regions deﬁned in net zero by ASHRAE carbon requirement90.1-2019, i.e., and climate design zoneswithin 4A,the 5A,same and regi 6A,on. Hourly which cantemperature show variations changes of in the net sce- zero nariocarbon in requirementthe 2050s and and 2080s design are randomly within the gene samerated region. separately Hourly for temperature each climate changes location of usingthe scenario the mean in thevalues 2050s in Table and 2080s 1 and are Equation randomly (1). Figures generated 2 and separately 3 show monthly for each distri- climate butionslocation of using randomly the mean generated values hourly in Table temper1 and Equationature changes (1). Figures for the2 scenario and3 show in the monthly 2050s anddistributions 2080s, respectively, of randomly in climate generated location hourly 4A temperatureNew York, NY. changes Note that for the dotted scenario line in in the each2050s plot and indicate 2080s, respectively, the mean value in climate of each location month, 4Aand New each York, bin size NY. in Note the thathistogram the dotted is determinedline in each plotbased indicate on the thenumber mean of value hourly of eachtemperature month, andchange each data bin sizein each in the month histogram and scenario.is determined Figure based 4 shows on the climate number change of hourly patte temperaturerns of hourly change temperatures data in corresponding each month and toscenario. each scenario Figure 4in shows the selected climate locations. change patterns of hourly temperatures corresponding to each scenario in the selected locations.

Climate 2021,, 9,, xx FORFOR PEERPEER REVIEWREVIEW 5 of 20 Climate 2021, 9, 17 5 of 20

FigureFigure 2. 2.Monthly Monthly distributions distributions of of hourly hourly temperature temperature changes changes for for scenario scenario the the 2050s 2050s in in 4A 4A New New York, York, NY. NY.

Figure 3. Monthly distributions of hourly temperature changes for scenario the 2080s in 4A New York, NY. Figure 3. Monthly distributions of hourly temperature changes for scenario the 2080s in 4A New York, NY.

Climate 2021, 9, 17 6 of 20 Climate 2021, 9, x FOR PEER REVIEW 6 of 20

FigureFigure 4.4.Hourly Hourly changeschanges inin generatedgenerated temperature based on the two two different different scenarios. scenarios.

ThisThis studystudy usesuses typicaltypical meteorologicalmeteorological year version 3 3 (TMY3) (TMY3) weather weather data data [25] [25 ]for for thethe hourlyhourly buildingbuilding energyenergy simulations.simulations. The TMY3 weather dataset dataset was was created created using using datadata collectedcollected overover aa 1991–20051991–2005 periodperiod basedbased on the procedures developed by by National National RenewableRenewable NationalNational Laboratories,Laboratories, and its data datasetset covers more more than than 1000 1000 locations locations in in the the U.S.U.S. [ 25[25].]. This This study study assumes assumes that that the TMY3the TM weatherY3 weather data woulddata woul representd represent typical weathertypical conditionsweather conditions in the present in the time present in the time selected in the climateselected locations, climate locations, and it is and used it asis aused baseline as a weatherbaseline scenario weather for scenario climate for change climate analysis. change To analysis. generate To the generate weather the scenario weather in thescenario 2050s andin the 2080s, 2050s TMY3 and 2080s, weather TMY3 datasets weather are datasets modified are to modified reflect the to reflect climate the changes climate shown changes in Figureshown4 .in More Figure specifically, 4. More specifically, the hourly the outdoor hourly airoutdoor dry-bulb air dry-bulb temperatures temperatures in the TMY3in the weatherTMY3 weather data file data in eachfile in location each location are modified are modified to determine to determine the weather the weather scenario scenario in the 2050sin the and 2050s 2080s. and The 2080s. modified The modifi hourlyed dry-bulb hourly dry-bulb temperatures temperatures (TOA,modified,i (TOA,modified,i) for the) weatherfor the scenarioweather inscenario the 2050s in the and 2050s 2080s and can 2080s be determined can be determined by: by:

,, =,, +, (2) TOA,modi f ied,i = TOA,original,i + Tchange,i (2)

where ,, is the hourly outdoor air dry-bulb temperature of current TMY3 where TOA,original,i is the hourly outdoor air dry-bulb temperature of current TMY3 weather weather files and , is the hourly temperature changes, presented in Figure 4. files and Tchange,i is the hourly temperature changes, presented in Figure4. 2.2. Description of Building Model 2.2. Description of Building Model The Department of Energy (DOE)’s flagship building energy modeling software, En- The Department of Energy (DOE)’s flagship building energy modeling software, Ener- ergyPlus version 9.0, is used to model office buildings in the selected climate locations gyPlus version 9.0, is used to model office buildings in the selected climate locations [22]. [22]. In this study, the DOE’s medium-sized office prototype building model is used to In this study, the DOE’s medium-sized office prototype building model is used to rep- represent all office building types. The office building type is chosen in this study because resent all office building types. The office building type is chosen in this study because the building model has the highest construction weights among all other building types the building model has the highest construction weights among all other building types (i.e., construction weight of 14.98% based on 2003–2007 surveyed data in the U.S.) and (i.e., construction weight of 14.98% based on 2003–2007 surveyed data in the U.S.) and complies with the minimum energy code requirements prescribed in ASHRAE Standard complies with the minimum energy code requirements prescribed in ASHRAE Standard 90.1-2019 [25,26]. Figure 5 shows a three-dimensional (3D) perspective view and a floor 90.1-2019 [25,26]. Figure5 shows a three-dimensional (3D) perspective view and a floor plan of the DOE’s medium-sized office prototype model. plan of the DOE’s medium-sized office prototype model.

Climate 2021, 9, 17 7 of 20 Climate 2021, 9, x FOR PEER REVIEW 7 of 20

Figure 5. MediumFigure 5.officeMedium building ofﬁce model: building (a) 3D model: view ( aand) 3D (b view) floor and plan (b) [8]. ﬂoor plan [8].

The prototypeThe office prototype building office is configured building is as configured a three-story as a three-storybuilding with building a rectan- with a rectangular gular floor plan,floor as shown plan, as in shown Figure in5. FigureEach floor5. Each has floorfive (5) has thermal five (5) zones, thermal including zones, includingfour four (4) 2 (4) perimeter zonesperimeter and zonesone (1) and core one zone, (1) corewith zone,a total with floor a totalarea of floor 4982 area m². of Its 4982 construc- m . Its construction tion consists ofconsists steel-framed of steel-framed exterior exteriorwalls, an walls, insulation an insulation layer entirely layer entirely above abovedeck for deck a for a flat roof, flat roof, and andslab-on-grade slab-on-grade floors. floors. All Allthe the construction construction details details and and internal internal heat heat gains gains comply with comply with thethe minimum minimum energy energy code code requirem requirements,ents, i.e., i.e., ASHRAE ASHRAE 90.1-2019 90.1-2019 [21]. [21]. Table Table 2 summaries 2 summaries representativerepresentative input input parameter parameter and and its its characteristics characteristics of ofthe the office office model. model. In In terms of a terms of a HVACHVAC system, system, the the original original prot prototypeotype office office model model employs a rooftoprooftop unitunit (RTU) system (RTU) systemwith with variable variable air air volume volume (VAV) (VAV) and and electric electric reheat reheat box tobox meet to requiredmeet required heating and cooling heating and coolingloads inloads each in zone each of zone a building, of a building, as shown as shown in Figure in 5Figure. With 5. a RTU-VAVWith a RTU- system, natural VAV system, gasnatural and gas electricity and electricity are used are for aused central for heatinga central coil heating and a reheatcoil and coil a reheat of a VAV reheat box, respectively. For cooling, a direct expansion (DX) cooling coil with two-speed fan operation coil of a VAV reheat box, respectively. For cooling, a direct expansion (DX) cooling coil is used, and the cooling coil requires two sets of performance data to calculate electricity with two-speed fan operation is used, and the cooling coil requires two sets of perfor- consumption based on the fan speed. All the same performance curves that the original mance data to calculate electricity consumption based on the fan speed. All the same per- prototype model provides are used in this study. Table3 summarizes detailed information formance curves that the original prototype model provides are used in this study. Table of the HVAC systems used in the prototype building model. 3 summarizes detailed information of the HVAC systems used in the prototype building

model.Table 2. Representative input parameter of the medium-sized ofﬁce model [22,27].

ParameterTable 2. Representative input parameter of the medium-sized office Characteristicsmodel [22,27]. LocationParameter 4A New York, NYCharacteristics 5A Buffalo, NY 6A Rochester, MN Total conditionedLocation floor area 4A New York, NY 5A Buffalo, NY4982 m 6A2 Rochester, MN Floor-to-ceiling heightTotal conditioned floor area 2.74 m 4982 (1.22 m² m above-ceiling plenum) Floor-to-ceiling height 2.74m (1.22m above-ceiling plenum) Window-wall-ratio 33% Window-wall-ratio 33% 2 2 2 U-factorU-factor 0.360.36 W/m²-K W/m -K 0.36 W/m²-K 0.36 W/m -K 0.34 W/m²-K 0.34 W/m -K Window Window SHGCSHGC 0.36 0.36 0.38 0.38 0.38 0.38 Exterior wallType Type Steel-frame Steel-frame walls walls Exterior wall construction constructionU-factor U-factor 0.3630.363 W/m²-K W/m 2-K 0.312 W/m²-K 0.312 W/m 2-K 0.278 W/m²-K 0.278 W/m2-K Built-up roof: roof membrane, insulation, and metal Roof TypeType Built-up roof: roof membrane, insulation, and metal decking Roof construction decking construction 2 2 2 U-factorU-factor 0.1800.180 W/m²-K W/m -K 0.180 W/m²-K 0.180 W/m -K 0.180 W/m²-K 0.180 W/m -K Floor TypeType Slab-on-grade Slab-on-grade floors floors Floor construction constructionU-factor U-factor 0.2300.230 W/m²-K W/m 2-K 0.230 W/m²-K 0.230 W/m 2-K 0.226 W/m²-K 0.226 W/m2-K Occupancy densityOccupancy density 18.57 m²/person18.57 m2/person Lighting power density Lighting power density (LPD) 6.89 W/m²6.89 W/m2 (LPD) 2 Equipment powerEquipment density (EPD) power density 8.1 W/m 8.1 W/m² Domestic hot water(EPD) (DHW) system Natural gas water heater with 0.81 of thermal efficiency and constant pump Domestic hot water (DHW) Natural gas water heater with 0.81 of thermal efficiency system and constant pump

Climate 2021, 9, x FOR PEER REVIEW 8 of 20

Climate 2021, 9, 17 8 of 20

Table 3. HVAC input parameter of the medium-sized office model [22,27].

TableParameter 3. HVAC input parameter of theCharacteristics medium-sized office model [22,27]. HVAC type Rooftop unit (RTU) with variable air volume (VAV) fan Parameter Characteristics and reheat box HVAC typeHVAC heating Rooftop coil unit (RTU) Natural with variable gas furnace air volume with (VAV) 0.81 of fan efficiency and reheat box HVAC heating coilHVAC VAV box Natural gas furnace Electrical with 0.81 reheat of efficiency coil with 1.0 of efficiency HVAC VAV boxHVAC cooling Electrical coil reheat coil Direct with expansion—two 1.0 of efficiency speed cooling coil with 3.4 of coefficient of performance (COP) HVAC cooling coil Direct expansion—two speed cooling coil with 3.4 of coefficient of performance (COP) Capacity of cooling and Auto-sized based on the design day Capacity of cooling and heating equipmentheating equipment Auto-sized based on the design day HVAC fanHVAC fan Variable air volume Variable fan with air 0.6 volume of efficiency fan with 0.6 of efficiency Maximum fan flow rateMaximum fan Auto-sized flow rate based onAuto-sized the design based day on the design day Zone thermostat set-pointZone thermostat Cooling: set-point 24 ◦C/Heating: Cooling: 21 24°C/Heating:◦C 21°C Zone thermostat set-back Cooling: 26.7°C/Heating: 15.6°C Zone thermostat set-back Cooling: 26.7 ◦C/Heating: 15.6 ◦C

Modeling the energy impacts of the building’s internal heat gains in EnergyPlus re- quiresModeling assumptions the energyabout the impacts internal of theheat building’s gain intensity internal and heat operation gains inschedules. EnergyPlus All requiresinput values assumptions and schedules about regarding the internal internal heat gain heat intensity gains are and directly operation obtained schedules. from Allthe inputoriginal values version and of schedules a prototype regarding office model internal without heat gains modifications. are directly Figure obtained 6 shows from frac- the originaltion schedules version of of representative a prototype office internal model heat without gainsmodifications. [26,27] (i.e., lighting, Figure6 plug, shows and fraction occu- schedulespancy), which of representative are used to determine internal heat total gains internal [26, 27heat] (i.e., gain lighting, values plug,inside and by simply occupancy), mul- whichtiplying are the used internal to determine heat gain total density internal times heat its faction gain values value inside in the by time simply horizon. multiplying In addi- thetion internalto the internal heat gain heat density gains, ASHRAE times its faction Standard value 90.1-2019 in the time[21] requires horizon. additional In addition plug to theloads internal consumed heat by gains, the office ASHRAE building, Standard such 90.1-2019as exterior [21 lighting] requires and additional electricity plugused loadsby an consumed by the office building, such as exterior lighting and electricity used by an elevator. elevator. All the parts of building energy consumption are considered for NZCB design All the parts of building energy consumption are considered for NZCB design in this study. in this study.

FigureFigure 6. Representative 6. Representative input input schedules schedules of the of simulationthe simulation model model [26, 27[26,27].].

2.3. Net-Zero Carbon Building (NZCB) Design Net-zero GHG emission building criteria can be achieved when the annual site GHG emission is equal or less than zero. The net GHG emission (EMnet) can be calculated by: emission is equal or less than zero. The net GHG emission (EMnet) can be calculated by:

= , +, − , EM = EM + EM − EM (3)(3) net ∑ elec−used,i gas−used,i ∑ elec−gen,i k k where , and , indicate CO emission values of the whole build- whereing electricityEMelec− andused, inationaland EM gasgas− end-use,used,i indicate respectively,CO2 emission and i valuesvaries from of the 1 wholeto 8760 building [hr] for electricity and national gas end-use, respectively, and i varies from 1 to 8760 [hr] for an annual simulation. , is the saved CO emission value due to the on-site anelectricity annual generation. simulation. EachEM elecemission−gen,i is value the savedcan beCO calculated2 emission by valuemultiplying due to the the building on-site electricity and natural gas end-uses times each conversion factor as follows:

Climate 2021, 9, 17 9 of 20

electricity generation. Each emission value can be calculated by multiplying the building electricity and natural gas end-uses times each conversion factor as follows:

EMelec−used = Eelec−used × CFelec (4)

EMgas−used = Egas−used × CFgas (5)

EMelec−gen = EPV−gen × CFelec (6)

where Eelec−used and Egas−used are total building electricity and natural gas end-uses, respectively. CFelec and CFgas are emission conversion factors for electricity and natural gas, respectively, in each state. Table4 shows the U.S. CO2 emission conversion factors by fuel used for this study.

Table 4. U.S. Emission conversion factors by fuel [28].

Emissions Conversion Factor Electricity, CFelec Natural Gas, CFgas State of New York 189.91 (kg/MWh) 181.33 (kg/MWh) State of Minnesota 454.99 (kg/MWh)

A solar photovoltaic (PV) system is one of the most promising distributed power technologies and can be used on building footprints and building site areas for on-site electricity generation. Although there are many ways to integrate PV systems with a building, this study only considers open area and roof mount types for PV installation with a ﬁxed tilt angle because this comparative study focuses only on climate changes and GHG emission reductions. To enable a NZCB balance, both roof and parking areas are used for solar PV installation locations. This study uses the PVWatts module in EnergyPlus to calculate on-site electricity generation. The PVWatts model is originally developed by the National Renewable Energy Laboratory (NREL), and is a web application and module in NREL’s system advisor model software that estimates the electricity generation of a grid-connected PV system for buildings based on a few simple inputs. EnergyPlus uses the same calculation algorithm without any additional needs to call the web application during the calculation process. The hourly available PV production is deﬁned [29] as:

I P = tr,i P 1 + γ T − T (7) dc,i 1000 dc0 cell,i re f

where Pdc,i and Pdc0 indicate hourly generated DC power from the plane-of-array and specified nameplate DC rate, Itr,i is the transmitted irradiance on a cell surface, γ is the temperature coefficient, and Tcell,i and Tre f are the computed cell temperature at each time step using a first-principles heat transfer energy balance model developed by Fuentes [30] and the reference cell temperature, respectively. The reference cell temperature is 25 ◦C, and reference irradiance is 1000 W/m2 [29,31]. The input values of the PVWatts module used in this study are listed in Table5.

Table 5. Input parameters of the PVWatts module in EnergyPlus [29,31].

Field Used Value System size Determined DC powers (kWp) (presented in Table6) Module type Standard System losses 14% Array type Fixed roof mount Tilt angle 37◦ Azimuth angle South facing (0◦) Inverter DC/AC ratio 1.1 Inverter efﬁciency 96% Climate 2021, 9, 17 10 of 20

Table 6. Maximum power capacity of PV installations to enable NZCBs.

Location 4A: New York 5A: Buffalo 6A: Rochester Climate 2021, 9, x FOR PEER REVIEW 10 of 20 Original TMY3 345.4 kWp 376.1 kWp 340.1 kWp 2050s 359.6 kWp 386.7 kWp 396.4 kWp Scenarios 2080s 354.2 kWp 395.8 kWp 388.9 kWp This study chooses the “standard” option because this option represents typical ploy- or mono-crystalline silicon modules, with efficiencies in the range of 14–17%, which is commonlyThis study used chooses for roof the mounted “standard” PV optioninstallati becauseon. Losses this optionin the representsPV system typical include ploy- the orimpacts mono-crystalline of soiling, shading, silicon modules, snow cover, with mismatch, efﬁciencies wiring, in the rangeconnections, of 14–17%, light-induced which is commonlydegradation, used nameplate for roof rating, mounted system PV age, installation. and operational Losses inavailability the PV system [31,32]. include This study the impactsuses 14% of of soiling, the default shading, total snowsystem cover, loss, mismatch,which is provided wiring, by connections, the EnergyPlus light-induced PVWatts degradation, nameplate rating, system age, and operational availability [31,32]. This study model. The inverter model is based on an analysis of California Energy Commission uses 14% of the default total system loss, which is provided by the EnergyPlus PVWatts (CEC) inverter performance data [33]. Hourly AC power outputs can be determined from model. The inverter model is based on an analysis of California Energy Commission (CEC) the DC outputs [29] as: inverter performance data [33]. Hourly AC power outputs can be determined from the DC outputs [29] as: , :0,  , = :,< < (8)  ηPdc,i : 0 Pdc,i Pdc0 : =0 Pac,i = Pac0 0 : Pdc,i ≥ Pdc0 (8)  0 : P = 0 where is the inverter efficiency, , is the hourlydc0 generated DC power outputs from

wherethe plane-of-array,η is the inverter efficiency, is the specifiedPdc,i is nameplate the hourly DC generated rate, and DC power is the AC outputs nameplate from therating plane-of-array, determined fromPdc0 is the the DC specified rating of nameplate the system DC and rate, the and DC-to-ACPac0 is the ratio. AC nameplate can be ratingcalculated determined by simply from multiplying the DC rating of times the system 96% and of thetheDC-to-AC default nominal ratio. P acefficiency0 can be calculatedused for the by calculation simply multiplying in this study.Pdc0 times η 96% of the default nominal efficiency used for theThe calculation electric load in this center study. transformer in EnergyPlus allows for feeding the surplus powerThe from electric on-site load power center generators transformer to the in electrical EnergyPlus grid. allows The load for center feeding object the matches surplus powerthe voltage from between on-site power on-site generators power generation to the electrical and total grid. electricity The load used center by object a building matches fa- thecility. voltage All surplus between electricity on-site produced power generation by the PV and power total generators electricity can used be by fed a back building into facility.the electrical All surplus grid through electricity the produced main panel by op theeration PV power that generatorsis connected can to be both fed the back utility into thegrid electrical and on-site grid throughpower generators the main panel of the operation NZCB. thatFor isthe connected transformer to both performance, the utility grid the andnominal on-site efficiency power generatorsis set to 97.7%, of the which NZCB. is Fordirectly the transformer obtained from performance, EnergyPlus the prototype nominal efficiencybuilding’s is default set to 97.7%,value which[32]. Figure is directly 7 depict obtaineds the on-site from EnergyPlus PV power prototypegenerators building’s with net- defaultmetering value operation [32]. Figurefor the7 NZCB depicts model the on-site. PV power generators with net-metering operation for the NZCB model.

Figure 7.7. A simulation modelmodel withwith anan on-siteon-site photovoltaicsphotovoltaics (PV) (PV) system system with with net-metering net-metering to to enable ena- net-zeroble net-zero carbon carbon buildings buildings (NZCBs) (NZCBs) [10]. [10].

2.4. Simulation Setting Description Based on thethe ScenariosScenarios The detaileddetailed stepssteps ofof the the simulation simulation setting setting based based on on the the two two scenarios scenarios are are depicted depicted in Figurein Figure8. The 8. The original original EnergyPlus EnergyPlus prototype prototype office office models models of each of each climate climate location location are first are simulatedfirst simulated with with modified modified weather weather files offiles the of different the different scenarios scenarios to evaluate to evaluate its electricity its electricity and national gas consumption. CO2 emission values are then calculated based on each building’s electricity- and natural gas-based emissions.

Climate 2021, 9, 17 11 of 20

Climate 2021, 9, x FOR PEER REVIEWand national gas consumption. CO2 emission values are then calculated based on11 each of 20

building’s electricity- and natural gas-based emissions.

Figure 8. FlowchartFlowchart of a simulation procedure.

The total maximum PV power outputs instal installedled on both the roof and parking areas of the simulated office office model are are presented in in Table Table 66 toto enableenable NZCBNZCB performance.performance. About 83% of the entire roof area (1661 mm²)2) is is used for for the roof-mounted PV PV installation. installation. Additional PV PV panels, panels, which which are are about about 3986.4 3986.4 m², m are2, areinstalled installed on parking on parking areas areas until untileach simulatedeach simulated model model shows shows satisfactory satisfactory annual annual NZCB NZCB performance performance corresponding corresponding to each to scenario.each scenario. It should should also also be be noted noted that that only only on-site on-site PV PVsystems systems are areapplied applied to the to existing the existing pro- totypeprototype office office building building model model with with no technological no technological advance advance changes changes to enable to enable NZCBs NZCBs be- causebecause the the objective objective of ofthis this study study is isonly only to to investigate investigate the the future future climate climate change’s change’s impacts on electricity electricity and and natural natural gas gas energy energy consumption consumption and and thus thus the the potential potential need need of PV of PVca- pacitycapacity for for NZCB NZCB performance. performance. This This simplified simplified way way can can give give precise precise insight insight into into how the current office office building model itself reacts to the future climate changes for the potential enabling of a zero carbon emission building with the scenarios over the coming decades. enabling of a zero carbon emission building with the scenarios over the coming decades. 3. Results and Discussion Table 6. Maximum power capacity of PV installations to enable NZCBs. This section presents and discusses results obtained from building energy simulations of NZCBs underLocation the different climate4A: change New scenarios York in5A: the Buffalo U.S. Northeast 6A: and Rochester Midwest regions.Original There are two climate TMY3 change scenarios 345.4 kWp updated 376.1 from kWp TMY3 weather 340.1 data kWp based on future changes obtained2050s from the proposed 359.6 kWp climate change 386.7 kWp model. Those 396.4 scenarios kWp are appliedScenarios in a conventional office building model (i.e., DOE’ medium office prototype 2080s 354.2 kWp 395.8 kWp 388.9 kWp building model) to estimate energy consumption and GHG emission as a baseline. Then the baseline office building model is designed to be the NZCB with the PV-based renewable 3. Results and Discussion energy generation system. A comparative analysis is conducted to determine the capacity of theThis on-site section PV systempresents to and enable discusses net zero results carbon obtained balance. Thefrom following building sectionsenergy simula- present tions of NZCBs under the different climate change scenarios in the U.S. Northeast and Midwest regions. There are two climate change scenarios updated from TMY3 weather data based on future changes obtained from the proposed climate change model. Those scenarios are applied in a conventional office building model (i.e., DOE’ medium office

Climate 2021, 9, x FOR PEER REVIEW 12 of 20

prototype building model) to estimate energy consumption and GHG emission as a base- Climate 2021, 9, 17 line. Then the baseline office building model is designed to be the NZCB with the12 PV- of 20 based renewable energy generation system. A comparative analysis is conducted to determine the capacity of the on-site PV system to enable net zero carbon balance. The following sections present the climate data obtained from the proposed climate change themodel, climate energy data performance, obtained from GHG the proposedemission climateresults for change climate model, change energy scenarios performance, deter- GHGmined emission from the results prototype for climate office building change scenarios models, and determined the potential from need the prototype for on-site office PV buildingsystem capacities models, andto enable the potential the NZCBs need with for the on-site climate PV systemchange capacitiesscenarios over to enable a yearly the NZCBscycle. with the climate change scenarios over a yearly cycle. 3.1. Climate Change Data 3.1. Climate Change Data InIn thisthis section,section, thethe profilesprofiles of generated hourly hourly dry-bulb dry-bulb temperatures temperatures are are presented. presented. EachEach climate climate location location has has three three different different temperature temperature profiles, profiles, including including the original the original TMY3. TheTMY3. hourly The temperaturehourly temperature profiles profiles of the two of the scenarios two scenarios are developed are developed based based on Equation on Equa- (2), Figuretion (2),9 depicts Figure the9 depicts hourly the profiles hourly of profile new generateds of new generated outdoor airoutd dry-bulboor air dry-bulb temperatures tem- forperatures two scenarios for two and scenarios three climate and three locations. climate The locations. temperature The temperature rises in hourly rises temperature in hourly profilestemperature throughout profiles all throughout scenarios and all climatescenarios locations and climate can be locations recognized can becausebe recognized of the risesbe- incause monthly of the projected rises in monthly outdoor airprojected temperatures, outdoor as air shown temperatures, in Figure as1. shown Those temperaturein Figure 1. profiles,Those temperature as shown in profiles, Figure9, as generated shown in by Figure the random 9, generated distribution by the are random used to distribution create new TMY3are used weather to create files new and TMY3 the weather weather flies files areand then the weather applied flies to building are then energy applied simulation to build- foring the energy comparative simulation analysis for the of comparative NZCBs. Note analysis that the of impactNZCBs. of Note other that parameters, the impact such of asother solar parameters, radiation andsuch wind as solar speed, radiation in the and TMY wind weather speed, file in isthe not TMY reflected weather in thisfile is study not whenreflected creating in this new study weather when filescreating because new theweather try bulb files temperature because the istry the bulb most temperature influential parameteris the most factor influential to evaluate parameter building factor energy to evaluate consumption, building especially energy consumption, when HVAC energyespe- usagescially when are focused. HVAC energy usages are focused.

FigureFigure 9. 9.Profiles Profiles of of generated generated outdoor outdoor air air temperatures temperatures (OATs) (OATs) for thefor scenariosthe scenarios in each in climateeach climate location. location. Based on the dry-bulb temperatures shown in Figure9, the design day weather data were derived by determining the maximum and minimum temperatures of the scenario in the 2050s and 2080s in design-day months. The design day information is used for sizing the HVAC systems in Section 2.2. The design day information for the baseline and scenario in the 2050s and 2080s for the selected locations are provided in Table7. Climate 2021, 9, 17 13 of 20

Table 7. Climate zones and its design-day weather information.

Design Day Dry-Bulb Temperature Climate Zone (CZ) City, State Scenario Heating (◦C) Cooing (◦C) TMY3 −10.7 32.1 4A New York, NY 2050s −6.8 35.6 2080s −4.7 38.2 TMY3 −16.3 30.3 5A Buffalo, NY 2050s −12.4 33.8 2080s −10.3 36.4 TMY3 −26.2 31.2 6A Rochester, NY 2050s −22.3 34.7 2080s −20.2 37.3

3.2. Energy Use Intensity (EUI) of Baseline Model This section compares the annual whole building EUIs of conventional baseline building models for the original weather file versus the two scenarios in the three U.S. climate locations. There are two energy source types in the prototype office building model, including electricity and natural gas. Space heating and domestic hot water (DHW) systems use natural gas as an energy source, while all other parts use electricity in the model. Figure 10 shows the detailed breakdown of annual building EUIs with the unit of kWh/m2-year. As expected, EUIs from HVAC systems (i.e., heating and cooling) vary significantly depending on the climate changes while EUIs for the other part remains almost the same. For example, in the 5A climate condition, increases of the cooling EUIs are 64% and 122% for the scenario in the 2050s and 2080s, respectively, whereas the heating EUIs show 31% and 47% of reductions corresponding to each scenario when compared to the original weather condition. The results from the other two climate locations (i.e., 4A and 6A) also indicate that similar trends are followed. The results indicate that there are significant impacts of climate changes on the HVAC energy consumption, thus it could be observed that GHG emission would vary primarily depending on HVAC energy usages. However, it should be noted that the total annual building EUIs do not show large differences when compared to the original case as shown in Figure 10 because the cooling EUI increases as the heating EUI decreases. Also note that the heating EUI consists of two fuel types, including electricity and natural gas, from VAV reheat coils and central heating coils, respectively. Since it is expected that changes in GHG emission generated by buildings primarily depend on heating and cooling energy usages, GHG emission variation by energy source types is also analyzed in this study. The annual EUIs of baseline models by the energy source type are presented in Figure 11 with the unit of kWh/m2-year. The results of all the climate locations indicate that the annual electricity usages tend to increase, as expected, while the annual gas usages show a decreased trend corresponding to climate change scenarios. Climate 2021, 9, 17 14 of 20 Climate 2021, 9, x FOR PEER REVIEW 14 of 20 Climate 2021, 9, x FOR PEER REVIEW 14 of 20

FigureFigure 10.10. AnnualAnnual wholewhole building Energy Use Intensity (EUI) of of baseline baseline models. models. Figure 10. Annual whole building Energy Use Intensity (EUI) of baseline models.

Figure 11. Annual electricity and natural gas energy use intensity of baseline models. FigureFigure 11.11.Annual Annual electricityelectricity andand natural gas energyenergy use intensity of of baseline baseline models. models. 3.3. GHG Emission of Baseline Model by Energy Source Type 3.3. GHG Emission of Baseline Model by Energy Source Type 3.3. GHGAnnual Emission GHG ofemission Baseline values Model byof Energyeach energy Source source Type type for the scenarios and cli- Annual GHG emission values of each energy source type for the scenarios and cli- mateAnnual locations GHG in the emission building values are ofdetermined each energy by source Equation type (4) for and the Equation scenarios (5), and respec- climate mate locations in the building are determined by Equation (4) and Equation (5), respec- locationstively. Figures in the 12 building and 13 present are determined annual GHG by Equationemission values (4) and of Equation baseline building (5), respectively. models tively. Figures 12 and 13 present annual GHG emission values of baseline building models Figuresby energy 12 sourceand 13 types present and annual HVAC GHG parts. emission In Figure values12, the ofGHG baseline emission building values models from the by by energy source types and HVAC parts. In Figure 12, the GHG emission values from the energyelectricity source consumption types and show HVAC increased parts. In patterns Figure 12and, the that GHG the natural emission gas values consumption from the electricity consumption show increased patterns and that the natural gas consumption electricityshows reduced consumption values as showthe scenario increased of climate patterns changes and that moves the further natural towards gas consumption the 2050s shows reduced values as the scenario of climate changes moves further towards the 2050s showsand 2080s, reduced whereas values the as total the scenariobuilding ofEUIs climate do not changes vary significantly. moves further For towards example, the in 2050s cli- and 2080s, whereas the total building EUIs do not vary significantly. For example, in cli- andmate 2080s, 4A, increases whereas in the the total GHG building emission EUIs values do from not varythe electricity signiﬁcantly. are 6.7 For and example, 13.0 tons in mate 4A, increases in the GHG emission values from the electricity are 6.7 and 13.0 tons climateof CO2-eq 4A, for increases the 2050s in theand GHG 2080s emission scenario, values respectively, from the when electricity compared are 6.7 to andthe 13.0original tons of CO2-eq for the 2050s and 2080s scenario, respectively, when compared to the original ofTMY3 CO -eqweather for the condition. 2050s and Reductions 2080s scenario, in thos respectively,e from the natural when gas compared are 6.8 toand the 9.92 original tons TMY32 weather condition. Reductions in those from the natural gas are 6.8 and 9.92 tons TMY3of CO2 weather-eq for the condition. 2050s and Reductions 2080s scenarios, in those resp fromectively, the natural when compared gas are 6.8 to and the 9.92original tons of CO2-eq for the 2050s and 2080s scenarios, respectively, when compared to the original ofweather CO -eq condition. for the 2050s The results and 2080s from scenarios, other two respectively, climate locations when (i.e., compared 5A and 6A) to the also original show weather2 condition. The results from other two climate locations (i.e., 5A and 6A) also show weathersimilar patterns condition. of Theincreases results and from decreases other two in climate GHG emission locations values, (i.e., 5A but and the 6A) amount also show of similar patterns of increases and decreases in GHG emission values, but the amount of similarGHG emission patterns values of increases of the 6A and location decreases is mu inch GHG higher emission than the values, others but due the to amountthe emis- of GHG emission values of the 6A location is much higher than the others due to the emis- GHGsion conversion emission values factor of of the electricity. 6A location is much higher than the others due to the emission sion conversion factor of electricity. conversionFigure factor 13 depicts of electricity. annual GHG emission values of baseline models by HVAC compo- nentsFigure for each 13 scenariodepicts annual and location GHG emissionwith the unitvalues of tonof baseline CO2-eq. models From this by figure,HVAC changes compo- innents the forGHG each emission scenario by and HVAC location fan with energy the usunitages of areton shownCO2-eq. to From be relatively this figure, low changes when comparedin the GHG to emission the original by HVACTMY3. However,fan energy the us ageschanges are shownof GHG to emission be relatively in HVAC low whenheat- ingcompared and cooling to the energy original usages TMY3. are However, significant. the Inchanges climate of 5A, GHG increases emission in thein HVAC GHG emis-heat- sioning and values cooling from energy cooling usages energy are use significant. are 5.3 and In 10.0climate tons 5A, of increasesCO2-eq for in scenarios the GHG inemis- the sion values from cooling energy use are 5.3 and 10.0 tons of CO2-eq for scenarios in the

Climate 2021, 9, x FOR PEER REVIEW 15 of 20

2050s and 2080s, respectively, when compared to the original case. In terms of GHG emis- 2050ssion from and 2080s,HVAC respectively, fan energy use when in 5A,compared it indicates to the that original there case. are 0.3 In termsand 0.5 of tons GHG of emis- CO2- Climate 2021, 9, 17 sioneq for from scenarios HVAC in fan the energy 2050s anduse in2080s, 5A, itrespectively, indicates that when there compared are 0.3 and to the 0.5 originaltons of 15 COcase. of2 20- eqNote for that scenarios the GHG in the emission 2050s andvalues 2080s, of electric respectively,ity and nationalwhen compared gas energy to theusages original presented case. Notein this that section the GHG are used emission as the values baseline of electric to furtherity and develop national to NZCBs. gas energy usages presented in this section are used as the baseline to further develop to NZCBs.

FigureFigure 12.12. AnnualAnnual embodied embodied greenhouse greenhouse gas gas(GHG) (GHG) emissi emissionon of baseline of baseline models models by energy by source energy type. sourceFigure type.12. Annual embodied greenhouse gas (GHG) emission of baseline models by energy source type.

Figure 13. Annual GHG emission of baseline models by HVAC components. Figure 13. Annual GHG emission of baseline models by HVAC components. Figure 13. Annual GHG emission of baseline models by HVAC components. 3.4. GHG Emission and PV Installation for NZCB 3.4. GHGFigureThis Emissionsection 13 depicts discusses and annualPV Installation variations GHG emission for of NZCB GHG values em ission of baseline values when models the by prototype HVAC compo- office nentsbuildingThis for each modelssection scenario discussesachieve and the variations location performance with of GHG the of unitNZCBs em ofission ton corresponding COvalues2-eq. when From tothe thiseach prototype figure, scenario changes office and inbuildingclimate the GHG location. models emission achieveThe PV by HVACtheinstallation performance fan energycapaciti of usagesesNZCBs needed are corresponding to shown enable to NZCB be to relatively each goals scenario is low also when andpre- comparedclimatesented inlocation. tothis the section. originalThe PV Figure TMY3.installation 14 However,depicts capaciti th thee scatteres changes needed plots ofto of GHGenable each emission NZCB performancegoals in HVAC is also heating pre-cor- andsentedresponding cooling in this energyto section. each usagesscenario Figure are and14 significant. depicts location th ewith Inscatter climate the plotsunit 5A, of increases toneach CO NZCB2-eq. in the performanceThe GHG annual emission GHG cor- valuesrespondingemission from from to cooling eachthe baseline scenario energy building use and are location 5.3energy and with 10.0end-use the tons unit is of compared COof ton2-eq CO for against2-eq. scenarios The that annual from in the on-site GHG 2050s andemissionpower 2080s, generation from respectively, the baselineto investigate when building compared carbon energy emission to end-use the original balances is compared case. for each In against terms scenario ofthat GHG andfrom location. emission on-site frompowerThe ratio HVAC generation of fany-axis energy toto investigatex-axis use should in 5A, carbon itbe indicates one emission or greater that balances there to reach are for 0.3 zero each and carbon scenario 0.5 tons emission and of CO location. perfor-2-eq for scenariosThemance, ratio as of in shown y-axis the 2050s into Figurex-axis and 2080s,should 14. It is respectively, be observed one or greater that when all to the compared reach cases zero of to carbonthe the building original emission model case. perfor- Notepre- thatmance,sent the acceptable as GHG shown emission net in zero Figure values carbon 14. ofIt balances. is electricity observed Th andethat GHG national all emissions the cases gas energyof NZCBsthe building usages are in presentedmodel the ranges pre- in thissentof 80–120 section acceptable tons are for usednet climate zero as the carbon zones baseline balances. 4A toand further 5A Th ande GHG develop 220–260 emissions to tons NZCBs. forof NZCBsclimate arezone in 6A,the respec-ranges oftively. 80–120 Since tons the for GHGclimate emission zones 4A of andthe 5Awhol ande building’s220–260 tons energy for climate end-use zone is determined6A, respec- 3.4.tively. GHG Since Emission the GHG andPV emission Installation of the for whol NZCBe building’s energy end-use is determined This section discusses variations of GHG emission values when the prototype office building models achieve the performance of NZCBs corresponding to each scenario and climate location. The PV installation capacities needed to enable NZCB goals is also presented in this section. Figure 14 depicts the scatter plots of each NZCB performance corresponding to each scenario and location with the unit of ton CO2-eq. The annual Climate 2021, 9, 17 16 of 20

GHG emission from the baseline building energy end-use is compared against that from on-site power generation to investigate carbon emission balances for each scenario and location. The ratio of y-axis to x-axis should be one or greater to reach zero carbon emission performance, as shown in Figure 14. It is observed that all the cases of the building model Climate 2021, 9, x FORClimate PEER 2021 REVIEW, 9, x FOR PEER REVIEW 16 of 20 16 of 20 present acceptable net zero carbon balances. The GHG emissions of NZCBs are in the ranges of 80–120 tons for climate zones 4A and 5A and 220–260 tons for climate zone 6A, respectively. Since the GHG emission of the whole building’s energy end-use is determined based on electricitybased and onon electricitynaturalelectricity gas andand usages, natural natural both gas gas electricity usages, usages, both both and electricity gaselectricity consumed and and gas ingas consumeda consumedbuild- in in a building a building need to be consideredneeding need to be to considered tobe calculate considered to accept calculate to calculateable acceptableNZCB accept balances.able NZCB NZCB Figure balances. balances. 14 also Figure presentsFigure 14 also 14 also presents presents the the near NZCBnearthe balances near NZCB NZCB that balances considers balances that onlythat considers considerselectricity only onlyconsumption. electricity electricity consumption. consumption.

Figure 14. GHG Figureemission 14. balance GHG emission for NZCB balance designs for in NZCB climate designs zones in(CZ). climate zones (CZ).

Equation (3) is Equationused to determine (3) is used the to determinetotadeterminel PV installation thethe totaltotal PVPV capacities installationinstallation needed capacitiescapacities for the needed needed forfor thethe NZCB goals byNZCB offsetting goalsgoals all byby offsettingGHG offsetting emissions all all GHG GHG from emissions emissions the electricity from from the andthe electricity electricitynatural and gas and natural con- natural gas consump- gas consumption of thetionsumption building of the ofmodel. building the building Figure model. 15model. Figuredepict Figures 15 the depicts capacities 15 depict thes capacitiesof the on capacities-site PV of on-site installation of on PV-site installation PV installation and and its GHG emissionitsand GHG its GHGintensities emission emission to intensities enable intensities the to zero enableto enable carbon the the emission zero zero carbon carbon balances. emission emission This balances. figurebalances. ThisThis ﬁgurefigure shows that the showscapacity that of thePV capacityinstallation of PVto offset installation the GHG to offset emission the GHGfrom emissionemissionthe electricity from thethe electricityelectricity consumption isconsumption significantly ishigher is significantly signiﬁcantly than that higher higher from thanthe than natural that that from fromgas theconsumption the natural natural gas gasin consumption all consumption cli- in all in cli- all mate locations.climatemate locations. locations.

Figure Figure15. PV installation 15.FigurePV installation 15. capacity PV installation capacity and its andGHGcapacity its emission GHG and emissionits intensity GHG intensityemission to enable tointensity enableNZCBs. to NZCBs. enable NZCBs.

As expected, theAs PV expected, capacities the of PV covering capacities the of electricity’s covering the GHG electricity’s emission GHGtend toemission in- tend to increase graduallycrease according gradually to the according climate change to the climate scenarios change mainly scenarios due to risesmainly in outdoordue to rises in outdoor air temperatures.air Fortemperatures. example, in For Buffalo example, (climate in Buffalo zone 4A), (climate 314, 327,zone and 4A), 341 314, kWp 327, of and PV 341 kWp of PV installation capacitiesinstallation are showncapacities for arethe shownclimate forscenarios the climate of original scenarios TMY3, of original 2050s, andTMY3, 2050s, and 2080s, respectively.2080s, Conversely, respectively. for Conversely, the PV capacities for the toPV cover capacities GHG toemission cover GHG from emissionthe from the national gas usage,national it is gas shown usage, that it isthe shown 2080s thatscenarios the 2 080spresent scenarios the lowest present capacities, the lowest capacities,

Climate 2021, 9, 17 17 of 20

As expected, the PV capacities of covering the electricity’s GHG emission tend to increase gradually according to the climate change scenarios mainly due to rises in outdoor air temperatures. For example, in Buffalo (climate zone 4A), 314, 327, and 341 kWp of PV installation capacities are shown for the climate scenarios of original TMY3, 2050s, and 2080s, respectively. Conversely, for the PV capacities to cover GHG emission from the national gas usage, it is shown that the 2080s scenarios present the lowest capacities, which are relatively higher reductions when compared to other cases. For example, in the climate 4A case, 46 kWp of PV installation capacities is needed to offset the GHG emission from natural gas, whereas about 60 kWp of PV installation capacity is needed for the other scenarios (i.e., TMY3 and 2050s) in the location. The results from the other two climate locations (i.e., 5A and 6A) also show similar patterns of increases and decreases in the PV installation capacities to offset GHG emission values. In addition, Figure 15 presents GHG emission generated from PV installations. A total of 50 g CO2-eq/kWh of GHG intensity obtained from [34,35] is used to calculate the GHG emission of PV installation in this study. These results indicate that the emission of PV installation is about one order of magnitude lower than the emission produced by the building energy consumption.

4. Conclusions This study investigated the impacts of potential future climate changes on annual building energy use intensities and GHG emission in the U.S. Northeast and Midwest climate locations. It also analyzed renewable design requirements to enable NZCBs under future climate scenarios using an on-site PV system. The Gaussian random distribution method was used to develop climate change scenarios with monthly temperature change values over the whole Northeast and Midwest regions, which were developed based on a high GHG emission scenario (i.e., the representative concentration pathways (RCP) 8.5). A simulation-based study of a prototype office building was conducted to evaluate the annual electricity and natural gas usages of the building models in different climate locations in the U.S. Northeast and Midwest regions. The PV-based on-site power generation system was considered and applied to the simulated building model to enable NZCB performance by assuming net-metering operation under future climate scenarios. Appropriate capacities of the on-site PV power system for reaching NZCB balances were required to be determined according to the energy source types of building energy consumption. The amount of GHG emission from the electricity or natural gas usages of the office building model could vary significantly depending on future climate scenarios. The capacities of the on-site PV systems for NZCBs could also vary based on a combined effect of annual building energy usages in each location and climate scenario, the emission conversion factors, and solar irradiation availability in each climate location. Key findings from this study are as follows: • The simulated office building models that comply with the minimum energy code requirements of ASHRAE Standard 90.1-2019 presented around 79–84 kWh/m2-year and 13–26 kWh/m2-year for the annual electricity and natural gas energy usages, respectively, in U.S. Northeast and Midwest climate locations. As the future climate scenarios of 2050s and 2080s were applied, the amount of annual electricity consumption tended to increase gradually from the baseline year climate conditions, whereas that of annual natural gas use decreased gradually toward the future years. This is primarily because of increases in outdoor air temperatures of climate change data used in the building simulations, which led to increases in the cooling energy consumption in the building, decreasing in heating energy usages; • The GHG emission trends in each climate followed the same pattern as the trends of the energy source type energy consumption (i.e., electricity and natural gas) changes toward the future climate scenarios. The emission by electricity consumption increased by, moving toward the future scenarios, up to about 25 tons of CO2-eq (about 14% of the total CO2 produced by the electricity energy source) while the emission produced by the natural gas consumption in the building decreased by up to about 20 tons of CO2-eq (about 30% of the total CO2-eq produced by the natural gas energy source). Climate 2021, 9, 17 18 of 20

The total GHG emission of the building in each climate location remained relatively even in all climate scenarios, although each climate location showed a different level of the emission rate. Depending on the climate conditions and electricity emission conversion factors in different climate locations, the same building can generate more emissions as the climate scenarios toward 2050s and 2080s as shown in the results of the New York case (climate zones 4A and 5A) in Figures 13 and 14; • The capacities of on-site PV systems required to enable NZCBs can differ depending on the local climate conditions (i.e., temperature and irradiation) and climate change scenarios as shown in shown in Figure 15 although the total PV capacity responsible by all energy source types did not change significantly (i.e., less than about a 15 kWp increase of the total PV capacity to offset the GHG emissions in the different climate scenarios). The building models of climate zone 4A used around 345–360 kWp for the zero-carbon emission target, while the other two climates (i.e., 5A and 6A) showed similar trends of PV capacities, including 376–387 kWp and 380–396 kWp for climate 5A and 6A, respectively. The results in Figure 15 also showed that the PV installation capacity to offset the emission account for the electricity consumption increases significantly up to about 40 kWp (i.e., up to more than 10% of total PV installation capacities) as the different climate change scenarios were applied. Such results from this study could provide useful insights into not only GHG emission generated from the electricity and national gas consumption of office buildings under different future climate scenarios in the selected climate locations, but also renewable design requirements (i.e., the capacity of on-site PV systems) when the office buildings were considered for the NZCB design. From the results, it can be concluded that the cooling energy consumption would significantly impact GHG emissions as future potential climate scenarios are considered. Consequently, the designers of NZCBs should consider highly efficient cooling energy systems in their designs to reduce the renewable energy generation system capacity to achieve net-zero carbon emission goals. It should be noted that there are limitations in this study, including that (a) only outdoor air dry-bulb temperature was used as a weather data parameter for developing the climate change model, and (b) the studied climate regions were limited within the U.S. Northeastern and Midwestern areas mainly due to the limited information of future climate change data in other regions of the U.S. Based on this study, future work can be extended to include various climate parameters in the whole building simulation analysis to reflect climate change scenarios in higher resolutions and improve the estimation of building energy consumptions and renewable energy requirements for NZCBs. Furthermore, all U.S. climate locations could be considered for a national scale analysis of NZCBs.

Author Contributions: Conceptualization, H.C. and D.K.; methodology, H.C., P.J.M. and D.K.; software and simulation, H.C., D.K. and H.L.; formal analysis, D.K.; investigation, H.C. and P.J.M.; resource, H.C., J.Y.; writing—original draft, D.K.; writing—review and editing, H.C., P.J.M., and J.Y.; visualization, H.C. and D.K. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Data Availability Statement: The data presented in this study are available within this article. Acknowledgments: The work was supported by the Korea Institute of Energy Technology Evaluation and planning (KETEP) and the Ministry of Trade, Industry & Energy (MOTIE) of the Republic of Korea (No. 20204030200080), the Department of Mechanical Engineering at Mississippi State University (MSU), and the Department of Mechanical and Aerospace Engineering at West Virginia University (WVU). Conﬂicts of Interest: The authors declare no conﬂict of interest. Climate 2021, 9, 17 19 of 20

References 1. U.S. Energy Information Administration. Monthly Energy Review. 2020; Volume 0035. Available online: https://www.eia.gov/ totalenergy/data/monthly/pdf/mer.pdf (accessed on 19 September 2020). 2. U.S. Energy Information Administration. EIA Energy Outlook 2020 with Projections to 2050. 2020. Available online: https: //www.eia.gov/outlooks/aeo/pdf/AEO2020%20Full%20Report.pdf (accessed on 19 September 2020). 3. CSR. U.S. Carbon Dioxide Emissions in the Electricity Sector: Factors, Trends, and Projections. 2019. Available online: https: //crsreports.congress.gov (accessed on 19 September 2020). 4. De Chalendar, J.A.; Taggart, J.; Benson, S.M. Tracking emissions in the US electricity system. Proc. Natl. Acad. Sci. USA 2019, 116, 25497–25502. [CrossRef][PubMed] 5. U.S. Global Change Research Program. Climate Science Special Report: Fourth National Climate Assessment; Global Change Research Program: Washington, DC, USA, 2018; Volume 1, p. 470. [CrossRef] 6. Jäger-Waldau, A.; Kougias, I.; Taylor, N.; Thiel, C. How photovoltaics can contribute to GHG emission reductions of 55% in the EU by 2030. Renew. Sustain. Energy Rev. 2020, 126.[CrossRef] 7. Collins, M.; Knutti, R.; Arblaster, J.; Dufresne, J.-L.; Fichefet, T.; Friedlingstein, P.; Gao, X.; Gutowski, W.J.; Tim, J.; Krinner, G.; et al. Chapter 12: Long-Term Climate Change: Projections, Commitments and Irreversibility; Cambridge University Press: Cambridge, UK; New York, NY, USA, 2014; Available online: https://www.ipcc.ch/report/ar5/wg1/long-term-climate-change-projections- commitments-and-irreversibility/ (accessed on 9 August 2020). 8. Kim, D.; Cho, H.; Koh, J.; Im, P. Net-zero energy building design and life-cycle cost analysis with air-source variable refrigerant flow and distributed photovoltaic systems. Renew. Sustain. Energy Rev. 2020, 118, 109508. [CrossRef] 9. Pylsy, P.; Lylykangas, K.; Kurnitski, J. Buildings’ energy efficiency measures effect on CO2 emissions in combined heating, cooling and electricity production. Renew. Sustain. Energy Rev. 2020, 134, 110299. [CrossRef] 10. Kim, D.; Cho, H.; Luck, R. Potential impacts of net-zero energy buildings with distributed photovoltaic power generation on the U.S. electrical grid. J. Energy Resour. Technol. ASME 2019, 141, 1–15. [CrossRef] 11. Reich, N.H.; Alsema, E.A.; van Sark, W.G.J.H.M.; Nieuwlaar, E. CO2 Emissions of PV in the perspective of a Renewable Energy Economy. In Proceedings of the 22nd European Photovoltaic Solar Energy Conference, Milan, Italy, 3–7 September 2017; pp. 3538–3542. 12. Louwen, A.; van Sark, W.G.J.H.M.; Faaij, A.P.C.; Schropp, R.E.I. Re-assessment of net energy production and greenhouse gas emissions avoidance after 40 years of photovoltaics development. Nat. Commun. 2016, 7, 1–9. [CrossRef][PubMed] 13. Pinel, D.; Korpås, M.; Lindberg, K.B. Impact of the CO2 factor of electricity and the external CO2 compensation price on zero emission neighborhoods’ energy system design. Build. Environ. 2021, 187.[CrossRef] 14. Wan, K.K.W.; Li, D.H.W.; Lam, J.C. Assessment of climate change impact on building energy use and mitigation measures in subtropical climates. Energy 2011, 36, 1404–1414. [CrossRef] 15. Nik, V.M.; Kalagasidis, A.S. Impact study of the climate change on the energy performance of the building stock in Stockholm considering four climate uncertainties. Build. Environ. 2013, 60, 291–304. [CrossRef] 16. Kikumoto, H.; Ooka, R.; Arima, Y.; Yamanaka, T. Study on the future weather data considering the global and local climate change for building energy simulation. Sustain. Cities Soc. 2015, 14, 404–413. [CrossRef] 17. Shen, P. Impacts of climate change on U.S. building energy use by using downscaled hourly future weather data. Energy Build. 2017, 134, 61–70. [CrossRef] 18. Wang, L.; Liu, X.; Brown, H. Prediction of the impacts of climate change on energy consumption for a medium-size office building with two climate models. Energy Build. 2017, 157, 218–226. [CrossRef] 19. Summa, S.; Tarabelli, L.; Ulpiani, G.; Perna, C. Impact of Climate Change on the Energy and Comfort Performance of nZEB: A Case Study in Italy. Climate 2020, 8, 125. [CrossRef] 20. Good, C.; Kristjansdottír, T.; Wiberg, A.H.; Georges, L.; Hestnes, A.G. Influence of PV technology and system design on the emission balance of a net zero emission building concept. Sol. Energy 2016, 130, 89–100. [CrossRef] 21. ASHRAE. Energy Standard for Buildings Except Low-Rise Residential Buildings (I-P); ANSI/ASHRAE/IES Standard No. 90.1-2019; ASHRAE: Atlanta, GA, USA, 2019. 22. US DOE. Commercial Prototype Building Models|Building Energy Codes Program; U.S. Department of Energy: Washington, DC, USA, 2013. Available online: https://www.energycodes.gov/development/commercial/prototype_models (accessed on 4 March 2020). 23. Yun, K.; Luck, R.; Mago, P.J.; Cho, H. Building hourly thermal load prediction using an indexed ARX model. Energy Build. 2012, 54, 225–233. [CrossRef] 24. Byun, K.; Chiu, C.M.; Hamlet, A.F. Effects of 21st century climate change on seasonal flow regimes and hydrologic extremes over the Midwest and Great Lakes region of the US. Sci. Total Environ. 2019, 650, 1261–1277. [CrossRef] 25. Wilcox, S.; Marion, W. Users Manual for TMY3 Data Sets. 2008. Available online: https://www.nrel.gov/docs/fy08osti/43156.pdf (accessed on 4 March 2020). 26. Goel, S.; Rosenberg, M.; Athalye, R.; Xie, Y. Enhancements to ASHRAE Standard 90.1 Prototype Building Models; Pacific Northwest National Laboratory (PNNL): Richland, WA, USA, 2014. Available online: http://www.energycodes.gov/sites/default/files/ documents/PrototypeModelEnhancements_2014_0.pdf (accessed on 18 August 2019). Climate 2021, 9, 17 20 of 20

27. Thornton, B.; Wang, W.; Xie, Y.; Cho, H.; Liu, B.; Zhang, J. Achieving the 30% Goal: Energy and Cost Savings Analysis of ASHRAE Standard; Technical Report: PNNL-20405; Paciﬁc Northwest National Laboratory (PNNL): Richland, WA, USA, 2011. 28. U.S. Environmental Protection Agency (EPA). Emission & Generation Resource Integrated Database (eGRID): EGRID Summary Tables. Available online: https://www.epa.gov/egrid/egrid-summary-tables (accessed on 19 December 2020). 29. US DOE. EnergyPlus Version 9.0 Documentation: EnergyPlus Engineering Reference; U.S. Department of Energy: Washington, DC, USA, 2019. 30. Fuentes, M.K. A Simpliﬁed Thermal Model for Flat-Plate Photovoltaic Arrays; Technical Report: SAND85-0330; Sandia National Laboratories: Albuquerque, NM, USA, 1987. Available online: https://prod-ng.sandia.gov/techlib-noauth/access-control.cgi/ 1985/850330.pdf (accessed on 18 August 2019). 31. Dobos, A.P. PVWatts Version 5 Manual; Technical Report: NREL-TP-6A20-62641; National Renewable Energy Laboratory (NREL): Denver, CO, USA, 2014. Available online: http://www.nrel.gov/docs/fy14osti/62641.pdf (accessed on 18 August 2019). 32. US DOE. EnergyPlus Version 9.0 Documentation: EnergyPlus Input and Output Reference; U.S. Department of Energy: Washington, DC, USA, 2019. 33. CCE. California Public Utilities Commission Rule. Available online: http://www.gosolarcalifornia.org (accessed on 18 Au- gust 2020). 34. Nugent, D.; Sovacool, B.K. Assessing the lifecycle greenhouse gas emissions from solar PV and wind energy: A critical meta- survey. Energy Policy 2014, 65, 229–244. [CrossRef] 35. NREL. Life Cycle Greenhouse Gas Emissions from Solar Photovoltaics. Renew. Sustain. Energy Rev. 2012, 50, 80. Available online: https://www.nrel.gov/docs/fy13osti/56487.pdf (accessed on 19 October 2020).