1、 Stephen Ray is an Assistant Professor in the Department of Physics and Engineering at North Park University in Chicago, IL and Technical Advisor at Skidmore, Owings & Merrill LLP in Chicago, IL. Luke Leung is Director of Sustainable Engineering at Skidmore, Owings & Merrill LLP in Chicago, IL Do Ta
2、ller Buildings Require More Energy? Stephen Ray, PhD, PE Luke Leung, PE, PEng Member ASHRAE Member ASHRAE ABSTRACT Do tall buildings consume more energy per unit area than typical buildings? According to the Council on Tall Buildings and Urban Habitat, more buildings exceeding 200 m (656 ft) in heig
3、ht were constructed in 2014 than ever before. Is this increase in height accompanied by an increased energy usage intensity (EUI)? This paper reviews sustainability reports from several major U.S. cities to explore these questions using measured EUIs around the country. Energy benchmarking data from
4、 New York and Philadelphia is also used to approach the question from a more detailed empirical perspective. Although “large” and “tall” buildings are sometimes used interchangeably, this paper demonstrates the important difference between them, particularly when discussing energy consumption trends
5、. . INTRODUCTION As more cities publically disclose building energy benchmarking data, additional opportunities to compare actual energy consumption across building types arises. As of July 2015, thirteen U.S. cities had passed legislation requiring commercial buildings to report building energy dat
6、a. While not all of them have collected and reported the data to date, the list includes: Austin, Atlanta, Boston, Cambridge, Chicago, Kansas City, Minneapolis, New York City, Philadelphia, Portland, San Francisco, Seattle, Washington, D.C. (Institute for Market 2015). This paper reviews city report
7、s for information on tall building energy consumption before more closely analyzing the data from New York City and Philadelphia to investigate how much energy tall buildings consume. The data from only New York City and Philadelphia are analyzed due to lack of public availability of energy usage da
8、ta, ease of access of energy usage data, availability of building height data, and space restrictions of this publication. CITY ENERGY BENCHMARKING REPORTS Although thirteen cities have passed legislation requiring certain commercial buildings to disclose energy consumption, some have not yet collec
9、ted data and others do not publically release it. Six benchmarking reports have been reviewed from the following cities: Chicago, Minneapolis, New York City, Philadelphia, Seattle, Washington, D.C. Only office buildings are considered if they are separately analyzed in each report. None of these cit
10、ies reports a direct measure of building height, though four of them plot energy consumption versus floor area, a proxy sometimes used for height. This proxy seems justified, especially in denser urban environments where buildings must balance the cost of land against the cost of building upwards. H
11、owever, it will be further investigated later in the article. Seattle and Minneapolis do not report any direct comparison between building size and energy usage intensity (EUI) (City of Seattle 2014)(City of Minneapolis 2013). Building height data in both Seattle and Minneapolis was not easily acces
12、sible and the page limitation of the current publication precluded the detailed data analysis of these cities. Chicago concludes “little correlation between building size and site or source energy use intensity” upon reviewing weather normalized site and source EUI plotted against building area (Cit
13、y of Chicago 2014). Energy usage data was not publicly released at the time of publication for Chicago, which is why it is not considered in the more detailed analysis. Results from Washington, D.C. provide minimal insight into the energy consumption of tall buildings due to citys height restriction
14、s. Nevertheless, the 2014 report plots site EUI versus square footage in 50,000 ft2 (4,600 m2) bins and concludes “In the metro Washington market, due to the height limits, energy intensity tracks more closely to building size, with the largest properties being the most efficient and smaller propert
15、ies being the least efficient. This generally holds true, except for the smallest properties in the analysis” (ULI 2014 2014). The height restrictions in Washington, D.C. reduce the relevance of its energy usage data to this paper, which is why it is not further analyzed. The Philadelphia report inv
16、estigates the energy impact of building size in two ways. The primary comparison is the Energy Star score plotted against square footage (City of Philadelphia 2014). The city reports, “ little correlation is apparent between building size and Energy Star score, particularly for smaller buildings” (C
17、ity of Philadelphia 2014). The report goes on to say “the largest buildings tended to report higher-than-average scores, a reflection of high performance among Philadelphias commercial high-rises” and that “larger buildings tended to perform above the national median Energy Star score of 50” (City o
18、f Philadelphia 2014). Later in the document, the city offers another comparison and plots site EUI against a coarser measure of building size, providing three bins for small, medium, and large buildings (City of Philadelphia 2014). No written discussion is provided and a clear trend is not immediate
19、ly present from visual inspection. The latest New York City energy benchmarking report analyzing 2014 data reports source EUI against floor area. A figure in the report plots offices in bins of 100,000 ft2 (9,300 m2) and shows a steady increase in EUI with floor area (New York 2014). The report conc
20、ludes “When comparing median EUI of groupings of properties by size, there is a direct relationship between size and energy use intensity” (New York 2014). Summarizing the latest publically available energy benchmarking reports from U.S. cities, Chicago concludes little correlation between square fo
21、otage and EUI, the Washington, D.C. metro area and Philadelphia report a slight correlation between increased floor area and decreased EUI, and New York City reports a direct relationship between floor area and increased EUI (City of Philadelphia 2014)(City of Chicago 2014)(ULI 2014 2014)(New York 2
22、014). New York City accounts for over half of the gross square footage in the U.S. required to report building energy usage and thus its observed trend of increased energy usage with building floor area based on the largest number of office buildings among cities considered in this paper (New York 2
23、014). All these trends are based on gross floor area, which often serves as a proxy for building height. In New Yorks dense urban fabric and high-priced real estate, one might expect floor area to better approximate building height than in less dense cities. Nevertheless, some of the physical driver
24、s for energy consumption in tall buildings (elevators, pump energy, reduced air temperature, increased view factor to sky, etc) do not affect large low or mid-rise buildings. Further discussion of these drivers is provided in previous studies or the new ASHRAE Tall Building Design Guide (Leung and R
25、ay 2013)(ASHRAE 2015). The following sections of this paper more closely investigate the data from New York City and Philadelphia to confirm the reported trends between EUI and square footage, and more importantly explore the impact of building height on EUI. The total number of floors is used as an
26、 indicator of height rather than building elevation because some of the elevation includes unoccupiable areas of the building that do not impact energy usage as directly as occupied floors. NEW YORK AND PHILADELPHIA BENCHMARKING DATA The urban fabric and climates of New York and Philadelphia differ,
27、 so although both data sets are from 2013, a direct comparison of EUI between cities should only be made after controlling for all primary factors impacting EUI, which is outside the scope of this paper. Instead, comparisons across buildings in each city are made before the trends within each city a
28、re compared. New York Data In 2009 New York City was the first city in the United States to require large buildings, with gross areas greater than 50,000 ft2 (4,600 m2), to publically release energy consumption data by passing New York City Local Law 84. The data analyzed in this paper was collected
29、 by the City of New York in 2012 and released in September of 2013 (The City of New York 2013). Unfortunately, the city does not release any direct measure of building height in their energy benchmarking data. However, the New York City Department of City Planning collects immense data on all buildi
30、ngs and lots in New York through the Primary Land Use Tax Lot Output (PLUTO), including the number of floors of each building (NYCDCP 2013). The PLUTO data used in this study was collected between February and May 2013 (NYCDCP 2013). The data has been cleaned to remove any building with incomplete d
31、ata or large outliers, which likely were the result of poor data. Some care is taken to control extraneous factors. Since all of the data is collected in New York over the same year, the same weather conditions applied to all the buildings. Only office buildings are considered, which are defined as
32、buildings with at least 80% of their total area used as office space. The total number of buildings that meet all these criteria and are thus used in this study is 706. Energy usage is measured using energy usage intensity (EUI) in the units of kBtu/ft2/yr (and alternatively kWh/m2/yr), which is cal
33、culated by dividing the entire building site energy in one year by the total floor area. Although source energy more accurately accounts for the carbon footprint of a building, site energy provides a better comparison between buildings, which is of higher interest in this study. Philadelphia Data As
34、 with the New York City data, the 2013 Philadelphia data used in this study only considers office buildings. It is similarly cleaned to ensure all buildings have EUI and floor area properly recorded, resulting in 194 office buildings. The number of floors is once again not reported in the energy ben
35、chmarking data and had to be obtained elsewhere, which was only done for buildings larger than 250,000 ft2 (23,000 m2), totally 67 buildings. Most of the building height data is from Emporium with fewer than 10% approximated from Google Maps Street View (Emporium)(Google). NEW YORK AND PHILADELPHIA
36、RESULTS New York Results Before plotting EUI versus height, EUI is plotted against square footage to confirm the reported trend of increased energy consumption with size from the City of New Yorks report and to investigate the office-only EUI versus size in Philadelphia. Among office buildings in Ne
37、w York City, the average site EUI of a given building size increases with gross floor area, as shown in Figure 1. These findings, which are from the 2013 data, are consistent with published results from the city using source energy from the 2014 data, as published in its 2014 benchmarking report (NY
38、C 2014 report). When the same data is plotted versus the number of floors, a different trend emerges, as shown in Figure 2. Figure 1 Average site EUI plotted against gross floor area of 706 office buildings as documented in the 2013 New York City Energy Benchmarking data. Figure 2 Average site EUI p
39、lotted against number of floors for 706 office buildings as documented in the 2013 New York City Energy Benchmarking data. Although the EUI of buildings from 0-29 floors steadily increases with number of floors, a distinct jump is observed at 30-39 floors, after which a plateau is reached for taller
40、 buildings. The average EUI of buildings with 10-19 floors increases by 33% as compared to buildings with 30-39 floors. An increase in twenty floors leads to 33% more energy consumption. With the same increase of twenty floors from 30-39 to 50+ floors, the average EUI actually slightly decreases by
41、3% after reaching the plateau at 30-39 floors. This plateau is consistent with the internal framework for tall building energy usage. Above 30-40 floors, the lost chilled-water efficiency from introducing a heat exchanger in a pressure break is already incurred. Sky view factors will also typically
42、not increase with elevation past 30-39 floors, since most buildings of that height are already exposed to a large sky. Wind speeds also do not increase as rapidly at higher elevations. Aerosol levels and moisture content similarly vary less at higher elevations, as they are further away from the pol
43、lutant and moisture sources at ground level. While these factors alone do not fully account for the observed trend, they very likely contribute to an explanation. Another factor that may lead to higher EUIs in taller buildings is the fact that many financial institutions with energy dense trading fl
44、oors and even onsite datacenters typically occupy tall buildings. Additional insight into these two trends can be obtained by plotting floor area against number of floors, which is done in Figure 3. Figure 3 (left) Gross floor area plotted against number of floors for 706 office buildings as documen
45、ted in the 2013 New York City Energy Benchmarking data. (right) Zoomed view of gross floor area versus number of floors for buildings ranging from 200,000 to 1,000,000 ft2 (19,000 to 93,000 m2). As expected, the number of floors in New York office buildings tends to increase with gross floor area. W
46、hen all 706 office buildings are considered, an R2 of 0.56 is observed. However, when the smallest and largest buildings are excluded, and only buildings between 200,000 to 1,000,000 ft2 (19,000 to 93,000 m2) are considered, a much weaker correlation between gross floor area and number of floors exi
47、sts, an R2 of 0.12. This tighter range of buildings includes most of those captured in the critical 30-39 floor bin from Figure 2 and shows the wide variation in floor area and number of floors, allowing for the two differing trends between EUI and floor area and EUI and number of floors. Furthermor
48、e, the increasing number of very slender buildings skews the common assumption that the number of floors increases with floor area. Philadelphia Results Recall the City of Philadelphia most directly considered the relationship between floor area and energy consumption by using the Energy Star score.
49、 They concluded larger buildings tended to have better Energy Star scores than smaller buildings. Numerous concerns have been raised about the meaningfulness of Energy Star scores for large buildings due to the limited number of large buildings in CBECS the Commercial Buildings Energy Consumption Survey (CBECS), on which the Energy Star score is heavily based (Leung and Ray 2013). When site EUI is plotted against building size for all 194 office buildings in Philadelphia, little correlation appears between EUI and building size. The largest buildings reported a very modest decrease i
