Evaluation of Factors Impacting EUI from High Performing Building Case Studies
The Diamond Building is the headquarters of Malaysia’s Energy Commission.
Lin Ho. Courtesy of Putra Perdana Development Sdn Bhd.
ASHRAE began publishing High Performing Buildings (HPB) magazine in the winter of 2008 with the intent of providing key details of actual high performing buildings including: architectural features, energy and water consumption values, lighting and daylighting strategies, HVAC system technologies, renewable energy, etc. Grumman and Hinge1 summarized the energy performance data and the common features found among eight selected buildings from a total set of 60 HPB magazine case studies. Teichman et al.2 reviewed 100 HPB case studies and focused on identifying characteristics that promote good indoor air quality (IAQ). This study builds upon these two previous studies and focuses on identifying any correlations between building performance and factors such as building type, climate zone, LEED certification level, and HVAC system technologies. In this study, “building performance” refers to energy consumption where energy use intensity (EUI) is the key performance indicator available. Certainly, although out of the scope of this study, “high performance” may also refer to other important issues (e.g., water consumption, recycling waste, IAQ, or other sustainability-related features).
Scope and Methodology
This study investigated 90 case studies from two sources (45 from each source): ASHRAE High Performing Buildings Magazine (HPB) and ASHRAE Technology Awards (from 2010 to 2015). The case studies chosen were limited to buildings that are >15,000 ft2. Limited by the available data in the articles, key information was extracted from each case study and entered into a database:
Application or Building Type
(e.g., Office, School, etc.)
LEED (Year, Version, Certification Level)
ASHRAE Award Information
HVAC System(s) Description
Breakdown of Main HVAC System(s)
(Geothermal, VAV, DOAS, etc.)
An interactive data visualization tool was then used to graphically explore the database and various relationships:
HVAC vs. Building Type and Climate Zone
EUI vs. Climate
EUI vs. LEED Certification Level
EUI vs. Building Type
EUI vs. HVAC Type
Only site EUI (subsequently referred to as EUI), and not source EUI, was evaluated in this study. In most case studies the site EUI was directly provided. In others, it was derived from the energy consumption data and building square footage. Of the 90 case studies, 10 did not provide enough information to calculate the EUI. In addition, although the database captures information about renewables, no distinction was made between renewable and nonrenewable energy in the EUI values. The intent of this study is to focus on how much energy the building consumes and not necessarily the source of that energy.
Locations and Climate Zones
Figure 1 and Table 1 show that the majority of the case studies are located in North America (87), while two are in China and one is in Germany. Each case study was further classified by the climate zones designated in ANSI/ASHRAE Standard 169-2013.3
The climate zones are not equally represented within the case studies. The cool and dry climates of 5A and 6A alone represent 40% of the case studies, with Quebec being the most frequent location. Figure 2 is an enlarged view of the North American case study locations.
Fifty-eight percent of the case studies were located in the cool/cold climate zones, and 26% were in the mixed climate zones. Only 17% were located in the warm, hot, or very hot (both dry and humid) climate zones.
Based on these 90 case studies, therefore, one may conclude that achieving high performance is easier in the cool/cold and mild climates and more challenging in hot and humid climates. For example, free cooling by economizers, natural ventilation, etc., is less applicable in the hot and humid climates.
It would be beneficial to collect more case studies from the hot and humid climate zones.
In understanding high performing buildings, the types of HVAC systems/technologies that are used are an important consideration. The system categories (mainly secondary, airside systems) that were captured from the case studies are:
Air/Air = Air-to-Air Energy Recovery
ASE = Airside Economizer
CB = Chilled Beams
DCV = Demand-Control Ventilation
DOAS = Dedicated Outdoor Air System
DV = Displacement Ventilation
Evap. = Evaporative Cooling
GSHP = Geothermal/Ground-Source Heat Pump
Rad. = Radiant Heating/Cooling
UFAD = Underfloor Air Distribution
VAV = Variable Air Volume
VRF = Variable Refrigerant Flow
WSE = Waterside Economizer
Figure 3 shows the number of case studies that have each HVAC system type (filtered by the two data sources—ASHRAE Technology Awards and ASHRAE HPB) and a pie chart showing the relative proportion of case studies that have a certain number of HVAC technologies.
Several systems and sub-systems were frequently found in these high performance buildings including: airside economizer, air-to-air energy recovery, dedicated outdoor air systems, ground-source heat pumps, and variable air volume. The frequency of the air-side economizer seems to be consistent with the prevalence of all-air VAV systems.
The number of case studies with systems such as chilled beams (CB), radiant cooling/heating (Rad.), displacement ventilation (DV), underfloor air distribution (UFAD), and variable refrigerant flow (VRF) is relatively low. This may not necessarily be due to energy performance, but rather could be due to factors such as lack of industry experience with these system types.
Observe the high percentage of case studies that can be considered to have “hybrid” HVAC system designs. “Hybrid” in this case refers to multiple HVAC system technologies used within the same building. This excludes the HVAC system combinations that normally go together as a design solution (e.g., VAV and ASE, GSHP and DOAS, etc.). As shown, 74 case studies (or 82%) are using several HVAC systems.
On average, approximately four HVAC technologies are applied per case study. This is a good indication that high performance often involves integrated design with multiple HVAC technologies working together to achieve high performance.
HVAC System vs. Building Type and Climate Zone
From the 90 case studies, the following building types are represented: hospital, industrial, K–12 school, laboratory, mixed-use, office, retail, and university. The mixed-use buildings include some combination of office, educational, laboratory, retail, etc. Figure 4 shows that K–12 school and office are the two most represented individual building types (i.e., 44% of all case studies). Mixed-use buildings are also very common.
Figure 5 shows the number of case studies that have each HVAC system type filtered by building type. Many of the HVAC system types are being used in most, if not all, the building types. For example, VAV was found in seven of the eight building types and was most common in office and mixed-use buildings. Similarly, DOAS was found in seven of the eight building types.
GSHP was found in all eight building types, with significant application in K-12 schools, offices, and mixed-use buildings. VRF, on the other hand, was not well represented in the case studies and was only found in two building types, office and mixed-use.
EUI vs. Climate Zone
The case study EUIs were first compared to climate zone to see if there is any strong correlation. Looking at the average building EUI values vs. climate zone can be misleading due to the fact that not all climate zones are well represented and the fact that the building’s EUI is influenced by more than just the weather. As a result, box and whisker plots are used throughout this study to show EUI distributions as a function of various factors.
Figure 6 shows the distribution of EUI values within each climate zone. The horizontal lines on the box represent the 25th, 50th, and 75th percentiles. The triangle represents the mean or average value. The whiskers extend to the maximum and minimum points within 1.5 times the interquartile range (75th to 25th percentiles). The red dots are any “outliers” or any points that fall outside the whiskers.
It is important to view this plot along with Figure 1, which shows how many case studies are in each climate zone. Clearly, not all climate zones are represented enough to have a healthy sample.
EUI vs. LEED Certification Level
Figure 7 shows the number of case studies by LEED certification level. Thirty-nine case studies did not mention any certification. One
case study was stated as being LEED certified, but the level was not provided. Fifty, or 56%, of the case study buildings are LEED certified with the different levels shown in Figure 7. Figure 8 shows that the EUI distributions are relatively similar for the various LEED certification levels and for the non-certified case studies. The non-certified dataset does include some lab/hospital “outliers” that are impacting the average.
Within the case studies with LEED certification, the EUI performance only slightly (based on averages) follows the certification level (i.e., Platinum  < Gold  < Silver ). There are only four case studies with the lowest level of “certified.”
Not enough information on each case study is available to determine whether or not the non-certified buildings could become LEED certified (i.e., assuming they simply did not participate in the program).
EUI vs. Building Type
Figure 9 shows the relationship between EUI and building type. The hospital and laboratory average EUIs appear to be reasonable estimates for high performance cases, while the university buildings’ average seems to be low. Figure 9 shows there are some wide distributions for hospital and laboratory EUI, but relatively narrow distributions for K–12 schools and office buildings. This is reflective of the fact that hospitals and laboratories are more unique and can vary significantly with respect to plug/equipment loads, hours of operation, different outdoor air requirements, etc.
EUI vs. HVAC System Type
Figure 10 shows the number of case studies that use each system type, and Figure 11 shows the relationship between EUI and HVAC system type. Most of the case studies had multiple HVAC system types. Of the 90 case studies, there were 68 unique HVAC technology/system combinations or configurations.
Each building EUI value includes the energy consumed by all HVAC systems within that building. This means that system-specific energy consumption could not be disaggregated. For example, the VAV average EUI (101 kBtu/ft2·yr) is the average of all the case studies that have a VAV system. However, those same buildings may have other systems as well. In addition, not enough information is provided in the case study articles to determine what percentage of each building is served by any one system type.
The EUI distributions provide more information than the average EUI values. For example, the chilled beam system had the highest average EUI (123 kBtu/ft2·yr), but this could be due to the fact that only nine case studies had chilled beams, and these case studies may have been applications with high plug/process loads.
GSHP shows relatively low EUI, which may be a result of the building type where it is used or because other measures help reduce energy use (e.g., GSHP is often combined with DOAS, which handles the outdoor air load). Keep in mind the number of case studies with each system type (Figure 10) since not all of the system types are well represented in this data set.
Due to the large number of system configurations within the case studies, being able to isolate the performance of one system type vs. another will require a simulation-based analysis where all other variables can be controlled.
Results and Discussion
Based on the 90 high performance building case study projects reviewed in this study, we agree with Grumman and Hinge1 that the energy consumption of high performing buildings can vary widely depending on many factors. In other words, a high EUI does not necessarily mean the building is performing poorly. In particular, building type or application and building location also need to be considered when evaluating EUI.
In addition, there are many ways in which a building can become a high performer. In terms of HVAC systems, for example, there is no clear “winner,” and there are many HVAC system designs that can result in a high performance building. Berger, et al.,4 describe the dataset comparison tools (e.g., graphical comparisons of distributions or multivariate regression analysis) available in the DOE Building Performance Database, which are intended to help one make more informed conclusions about why the average EUI from one dataset is different from that of another dataset.
We similarly recognized that EUI can be impacted by many factors, which is why we explored the relationships between EUI and climate zone, building type, LEED certification, and HVAC system types. Looking at any one relationship in isolation can be misleading. We believe the use of an interactive data visualization tool can help avoid the pitfalls of empirical data analysis (e.g., help identify confounding factors). The overall conclusions from this study can be summarized as follows:
Sample Size: Similar to the Building Performance Database,4 the case studies reviewed in this study do not constitute a statistically representative sample of the national building stock. Not all building types, climate zones, or HVAC system types were equally represented. In addition, since this study reviewed only high performing buildings, the EUI values are not representative of normal or average buildings.
Location and Climate Zones: Most of the geographical locations of the case studies are in the cold and mild climates.
HVAC: High performance buildings are more likely to apply several HVAC technologies, i.e. “hybrid” designs. This may be a result of more careful integrated design practices. Frequently used systems in these case studies included: variable air volume, dedicated outdoor air systems, ground-source heat pumps, and air-to-air energy recovery. Systems such as chilled beams, radiant cooling/heating, displacement ventilation, underfloor air distribution, and variable refrigerant flow appeared less frequently.
HVAC vs. Building Type: K–12 schools and office buildings are the two most represented individual building types among the case studies. Mixed-use buildings are also common. Several of the HVAC system types (e.g., VAV, DOAS, GSHP, and air/air) were found in most of the building types indicating no favored system type for any application.
EUI vs. Climate: There appears to be no strong correlation between EUI and climate zone.
EUI vs. LEED: Many of the high performance building case studies are LEED-certified buildings. However, the EUI distributions do not show major differences between the non-certified and certified case studies. Not enough information was available to indicate whether or not the non-certified case studies could be certified if they participated in the LEED program.
EUI vs. Building Type: As expected, the hospital and laboratory EUIs were higher than the office and K–12 school EUIs. The range of EUIs for hospitals and labs is also larger than the other applications. This is likely due to the variations in processes, equipment loads, etc. In all building types, the average EUIs appear to be reasonable estimates of “high performance” for those applications.
EUI vs. HVAC System Type: It is difficult to conclude if there is a correlation between EUI and HVAC system type. This relationship is complicated by the fact that the majority of the buildings used multiple HVAC system types. This observation may also imply that building energy performance is more driven by the loads in the building and not necessarily the HVAC systems chosen to meet those loads. Simply observing the average EUI for each system type can be misleading; the EUI distributions show that no system is clearly outperforming any other system. To make more objective comparisons between HVAC system types, a simulation-based approach is needed. With simulation, the same building geometry/architecture, lighting, plug loads, schedules, weather conditions, etc., can be analyzed with different HVAC system types or configurations. In this way, the performance of the various HVAC system types can be isolated and analyzed.
1. Grumman, D.L., A.W. Hinge. 2012. “What makes buildings high performing.” High Performing Buildings (Spring):47 – 53.
2. Teichman et al. 2013. “Wealth of intent, dearth of data.” High Performing Buildings (Fall):35 – 41.
3. ANSI/ASHRAE Standard 169-2013, Climatic Data for Building Design Standards.
4. Berger et al. 2016. “Big data analytics in the building industry.” ASHRAE Journal 58(7).
About the Authors
Itzhak Maor is director of Smart Building Technologies for Johnson Controls, Inc.
Steven C. Snyder is an energy services engineer for Johnson Controls, Inc., and a Ph.D. candidate at Drexel University, Philadelphia, PA.