Research

A Strategy for Energy Performance Analysis

at the Early Design Stage: Predicted vs. Actual
Building Energy Performance
Ajla Aksamija, PhD, LEED AP BD+C, CDT1

abstract

Downloaded from http://meridian.allenpress.com/jgb/article-pdf/10/3/161/1768560/jgb_10_3_161.pdf by Moldova, Republic of user on 23 December 2020

Developments in information technology are providing methods to improve
current design practices, where uncertainties about various design elements can be
simulated and studied from the design inception. Energy and thermal simulations,
improved design representations and enhanced collaboration using digital media
are increasingly being used. With the expanding interest in energy-efficient build-
ing design, whole building energy simulation programs are increasingly employed
in the design process to help architects and engineers determine which design strat-
egies save energy and improve building performance. The purpose of this research
was to investigate the potential of these programs to perform whole building
energy analysis during the early stages of architectural design, and compare the
results with the actual building energy performance. The research was conducted
by simulating energy usage of a fully functional research laboratory building using
two different simulation tools that are aimed for early schematic design. The results
were compared with utility data of the building to identify the degree of close-
ness with which simulation results match the actual energy usage of the build-
ing. Results indicate that modeled energy data from one of the software programs
was significantly higher than the measured, actual energy usage data, while the
results from the second application were comparable, but did not correctly predict
monthly energy loads for the building. This suggests that significant deviations may
exist between modeled and actual energy consumption for buildings, and more
importantly between different simulation software programs. Understanding the
limitations and suitability of specific simulation programs is crucial for successful
integration of performance simulations with the design process.

Keywords
building performance analysis, simulated vs. actual energy usage in buildings,
architectural design

INTRODUCTION
Building performance simulation tools are increasingly used for analysis of the energy per-
formance of buildings (Augenbroe et al. 2004, Aksamija 2009, Aksamija 2010, Wetter 2011,
Aksamija 2012). Building energy simulation is a powerful method for studying the energy

1. Assistant Professor, Department of Architecture, University of Massachusetts Amherst, 151 Presidents Dr., Amherst, MA
01003, 413-545-7150, [email protected].

Journal of Green Building161

performance of buildings and for evaluating architectural design decisions, as well as choices
for construction materials and methods. Complicated design issues can be examined, and
performance of different design strategies can be quantified and evaluated. Currently, there are
many building performance simulation programs with different user interfaces and different
simulation engines that are capable of these analyses (Crawley et al. 2008).
Simulation and energy analysis are essential to designers in developing effective forms
and components for their buildings. Building energy simulation is an analysis of the dynamic
energy performance of a building using computer modeling and simulation techniques. There
is a wide range of simulation tools available that help predict various aspects of building

Downloaded from http://meridian.allenpress.com/jgb/article-pdf/10/3/161/1768560/jgb_10_3_161.pdf by Moldova, Republic of user on 23 December 2020

behavior, such as energy performance, acoustical performance, fire movement, structural per-
formance, life-cycle assessment, etc.
Energy performance simulation tools allow designers to:
• Predict thermal behavior of buildings in relation to the outdoor environment.
• Simulate the impact of daylight and artificial light inside buildings.
• Estimate the size/capacity of equipment required for thermal and visual comfort, and
the associated energy usage for their implementation.
• Calculate the effect of various building components on each other and predict
resulting conditions.
• Check for compliance with energy codes.
• Consider the building as a single integrated system.

However, past research on the utilization of simulation tools during the architectural
design process indicates that despite the increase in number of available tools in the last decade,
some architects and designers find it difficult to use these tools, since they are not compatible
with the working methods and needs, or the tools are judged as complex and cumbersome
(Gratia and de Herde 2002, Punjabi and Miranda 2005). To remain competitive, design pro-
fessionals must weigh the value of information gained through simulation tools against the
invested time and resources, and against the value of comparable information that might be
gained through other means.
In order to evaluate and optimize building performance, different analysis cycles should
be part of an integrated design process. Tools and applications that support integrated
design and analysis from the earliest stages of the design can aid the decision-making process
(Aksamija 2012). Figure 1 shows the impact of design decisions on actual building perfor-
mance and relationships to project stages, as well as uncertainties about building components
that are present during the different stages of the design.
As early as the conceptual phase, the analysis should focus on design aspects such as
climate information, orientation, passive strategies, programming and building massing. Then
at the schematic stage, the analysis should explore shading methods, solar access, and build-
ing envelope design options. During the design development stage, optimization of shading
devices, daylight and glare studies, detailed energy performance studies, thermal analysis and
optimization should take place. It is important to note that these types of studies have the
greatest impact on building performance if they are conducted early in the design process
(conceptual, schematic and design development phases). They should not be performed
during the construction documentation phase, since their impact on building performance
is too small to justify analysis at this stage, and the cost of design changes is typically pro-
hibitive. Therefore, simulation tools and applications that can be used in early stages of the

162 Volume 10, Number 3

Figure 1: Design decisions and effects on building performance uncertainties in relation to
different stages of the architectural design process.

Downloaded from http://meridian.allenpress.com/jgb/article-pdf/10/3/161/1768560/jgb_10_3_161.pdf by Moldova, Republic of user on 23 December 2020

design process are important for energy-efficient design, decision-making and investigations
of design strategies that can be used to improve the overall building performance.
Figure 2 shows several building performance analysis tools and applications, and their
applicability to different design stages. For example, during the early stages of the design
process (conceptual and schematic design), tools such as Green Building Studio, Energy 10,
Sefaira, Design Builder and Ecotect are applicable. During the schematic and initial design
development stages, tools such as eQuest, IES VE, and EnergyPlus with OpenStudio plug-in
for SketchUp are applicable. During the design development stage, EnergyPlus is the most

Figure 2: Energy and environmental analysis software applications in relation to design stages.

Journal of Green Building163

useful energy modeling tool, since its robustness allows for detailed calculations and optimiza-
tion of building systems.
But, how are these different applications used in architectural design? A recently con-
ducted survey investigated utilization of building performance tools in architectural practice,
particularly comparing these following simulation programs: ECOTECT, HEED, Energy 10,
Design Builder, eQuest, DOE-2, Green Building Studio, IES VE, Energy Plus and Energy
Plus-SketchUp plug-in (OpenStudio) (Attia et al. 2009). With 249 responses, the survey
ranked the utilization of tools and investigated requirements for potential future improve-
ments of these applications. Findings indicated that Ecotect is currently most widely used

Downloaded from http://meridian.allenpress.com/jgb/article-pdf/10/3/161/1768560/jgb_10_3_161.pdf by Moldova, Republic of user on 23 December 2020

(156 users), followed by eQuest (123 users), EnergyPlus-SketchUp (81 users) and EnergyPlus
(81 users), IES VE (60 users), Energy 10 (57 users), Design Builder (54 users), DOE-2 (48
users), HEED (45 users) and Green Building Studio (27 users).

1.1 Research Purpose and Objectives

The intent of this work was to conduct a comparative analysis research on conceptual Whole
Building Energy Analysis (WBEA) programs, which are applicable to early design stages.
The purpose is to inform designers and engineers about the potential for integrating simu-
lation programs with the design, which would yield accurate predictions about the build-
ing performance from the earliest stages of the design process. This is an essential aspect in
the design of high-performance buildings, and improvement of the design decision-making
process (Aksamija and Abdullah 2013). Given the significant variety of such simulation tools,
it is crucial to understand limitations of the tools and the complexity of simulations. Prior to
conducting this research, our objective was to find an efficient and beneficial method of seam-
lessly integrating WBEA into the design process.
The notion of calculating building’s energy usage as a “whole” is not a new concept—
there are existing simulation tools that have been around for the last two decades, as discussed
in the previous section. However, integration with the design process and BIM technologies
are newer concepts, which are still being investigated (Aksamija and Mallasi 2010, Aksamija
2012). In essence, WBEA is the process of analyzing a building’s energy performance by cal-
culating how well the building’s form, systems, and envelope perform under the surrounding
environmental conditions. Software tools that integrate graphical results with context-sen-
sitive guidance are likely to have the most appeal for architects and designers. In contrast,
engineers need software tools that can be used in both the conceptual design stage, when little
is known about the building, as well as in the later design stages, when majority of the project
details have been finalized.
When a building is modeled for the same climate in different simulation programs, the
outputs of simulation runs are expected to be similar. However, different software programs may
exhibit a significant difference in output for the projected energy usage of a building (Agami
2006, Maille et al. 2007). This is a major issue for successful integration of performance simu-
lations with the design process, especially for early stages of the design. In the conceptual and
schematic design phases, design decisions can have a significant impact on the overall building
performance and its energy usage (such as building massing, geometry, orientation, window-to-
wall ratio, shading strategies, etc.). Therefore, the objectives of this study were:
• To conduct literature review and identify current research efforts relating to
integration of building performance simulation programs with design process, and
relationships between simulated and measured energy performance in buildings.

164 Volume 10, Number 3

 • To investigate properties and applicability of two different software programs, which
are suitable for early stages of architectural design process (Vasari/Green Building
Studio and Sefaira).
• To model a building similarly in these software programs by closely mapping the
input parameters.
• To compare the results of simulations with measured utility data and identify
discrepancies.
• To document the findings of the study, and identify which simulation tools are better
suited for early stages of the design.

Downloaded from http://meridian.allenpress.com/jgb/article-pdf/10/3/161/1768560/jgb_10_3_161.pdf by Moldova, Republic of user on 23 December 2020

1.2 Methodology
The “conceptual” aspect of WBEA is the attempt to integrate WBEA into earlier phases of
design in order to allow all parties working on the project to make the best informed design
decisions prior to more detailed design and additional months of labor. This can be achieved
by integrating BIM technologies with energy simulations, however, most BIM-based WBEA
software are new or still in beta versions. Therefore, this study was conducted to test two dif-
ferent software programs by modeling the energy usage of an existing building, and compar-
ing to the actual building performance data.
Different simulation programs may have different software architecture, different algo-
rithms to model building and energy systems, and require different user inputs even to
describe the same building envelope or HVAC system components. For this study, the research
methodology was to identify a recently designed and constructed existing high-performance
building, and to model the identical inputs for building systems, environmental conditions,
control strategies, and material components in different software programs. Simulations were
completed using Vasari/Green Building Studio (GBS) and Sefaira, both of which have been
specifically developed for early conceptual design.

2. Literature Review
There are existing studies that compare modeling capabilities of different building perfor-
mance simulation programs (Maile et al. 2007, Crawley et al. 2008). Different tools have
different modeling features and methods for describing the building and its systems, as well
as calculation methods. But, regardless of the type of simulation tool, it is necessary to val-
idate results and understand the capabilities of modeling software programs. For example,
the Building Energy Simulation Tests (BESTEST) were developed to standardize methods
for testing building energy analysis computer software (NREL 1995). The methodology pre-
scribed by BESTEST has been adopted by the ANSI/ASHRAE standard method for testing
and evaluating building performance analysis tools (ASHRAE 2001). Existing research exists
that reviews validation of several building energy simulations programs using this method,
comparing results from the studied software programs to analytical solutions developed for
test cases (Neymark 2002).
One of the primary methods for investigating the validity of energy modeling and build-
ing performance simulation results is by comparing the modeled results to actual building
performance data (Ryana and Sanquist 2014, de Wilde 2014, Fumo 2014). A recently pub-
lished literature review by de Wilde suggests that there is often significant differences between
predicted energy performance of buildings and actual measured energy use (2014). It suggests

Journal of Green Building165

that there are three main types of performance gaps: 1) between initial predictions and mea-
surements, 2) between computational models and measurements, and 3) between predic-
tions and performance certificates in legislation. It also states that the current literature on
the energy performance gap suggests various causes for the mismatch between predicted and
measured energy usage in buildings, which can be grouped into three categories:
• Causes related to the design stages (such as modeling inputs not closely resembling
the real conditions, building occupancy schedule, equipment loads, etc.)
• Causes related to construction stage (such as building elements and systems not being
constructed as intended)

Downloaded from http://meridian.allenpress.com/jgb/article-pdf/10/3/161/1768560/jgb_10_3_161.pdf by Moldova, Republic of user on 23 December 2020

• Causes related to operational stage (such as building systems not being properly com-
missioned).

But how big are these discrepancies? Previous studies that investigated discrepancies
between simulated and actual energy usage in buildings indicate that these gaps can be sub-
stantial, and in the range from 10 to 30% (Diamond et al. 2006, Fowler and Rauh 2008,
Turner 2008, Turner and Frankel 2008, Newsham et al. 2009, Widener 2009, Scofield 2009,
Stoppel and Leite 2013). The common performance measure that was used in these studies is
the Energy Use Intensity (EUI) parameter (kBtu/ft2 or kWh/m2). It is calculated by adding
energy usage for all building systems on an annual basis (for heating, cooling, ventilation,
lighting, hot water needs, equipment), and normalizing by the building’s gross area. This allows
comparison among different buildings of similar types, and is commonly used to compare pre-
dicted to actual energy consumption. But, it does contain limitations since it does not distin-
guish between buildings with differing occupant density, usage patterns or process loads.
Results of studies that investigated predicted vs. actual energy performance of large
data sets of buildings suggest that significant variation among individual buildings may exist
(Diamond et al. 2006, Turner 2008, Turner and Frankel 2008, Widener 2009). For example,
Turner and Frankel conducted a study of 121 LEED certified buildings, and their findings
indicate that the average predicted annual energy savings of 25% (compared to energy code
baseline buildings) was close to the actual measured savings of 28% (2008). However, they
found significant variation among individual buildings, where over half of the buildings dif-
fered from the design predictions. Specifically, over 30% performed significantly better and
25% performed significantly worse.
There is ongoing research that aims to identify key issues that need to be addressed in
order to reduce this gap between simulated and measured energy performance (Korjenic and
Bednar 2012, Ham Golparvar-Fard 2013, Katunsky et al. 2013, de Wilde 2014). For example,
one of the investigated approaches is to use dynamic simulations and precise input data, where
the input data is coming from measurements of energy usage for each part of the HVAC
system and equipment (Korjenic and Bednar 2012, Katunsky et al. 2013). This approach has
been investigated for commercial office buildings and industrial buildings. Another method
is to use digital and thermal imagery to analyze existing buildings and develop spatio-thermal
models, which can be used to detect deviations from simulated data and improve simula-
tion accuracy through model calibrations (Ham Golparvar-Fard 2013). This approach has
been applied to residential and educational buildings. However, both of these methodologies
are only applicable to existing buildings and determination of retrofitting design strategies to
improve their energy performance. It is much more challenging to address these issues during
the design process of new buildings. The literature suggests that the performance gap for new

166 Volume 10, Number 3

buildings can only be bridged by a broad, coordinated approach that combines model vali-
dation and verification, improved data collection for predictions, improved forecasting and
change of industry practice (de Wilde 2014). It is necessary to expand collection of data,
provide deep insights into individual cases, apply validation and verification to energy predic-
tion methods, and investigate applicability of different tools and software applications and
their accuracy. Therefore, the following case study provides an in-depth analysis of simulated
vs. actual energy usage for a specific building, specifically investigating results of two software
programs that are aimed for early architectural design and their correlation to actual, mea-
sured data.

Downloaded from http://meridian.allenpress.com/jgb/article-pdf/10/3/161/1768560/jgb_10_3_161.pdf by Moldova, Republic of user on 23 December 2020

3. Case Study Building Description
The case study building that was used for this study is a research laboratory building located in
Tacoma, Washington. The facility is primarily used for studying and analyzing water samples,
but is also used for educational activities. Its area is 51,000 ft2 (4,740 m2). The program
includes laboratories, offices, conference rooms, an exhibit center, a cafeteria, and related
building services. The building is located on a long and narrow site along the industrial water-
front of the Thea Foss Waterway. The geometry of the site led to a narrow building design,
oriented roughly north and south (Figure 3).

Figure 3: Building site and typical

floor plan.

Journal of Green Building167

 The building design used passive sustainable design strategies, which were strongly influ-
enced by the site’s orientation. The major programmatic elements are grouped into two zones:
a laboratory zone facing inland and an office zone along the waterway. Because of the pro-
grammatic requirements of the research activities, the laboratories required mechanical venti-
lation. Locating them adjacent to the industrial neighborhood, with its reduced opportunities
for fresh air, was a practical response to the site.
On the other hand, natural ventilation for the office spaces was considered highly desir-
able. By facing the waterway, the offices benefit from natural ventilation. The office spaces
on the north end of the building, with the laboratories to the east, use single-sided natural

Downloaded from http://meridian.allenpress.com/jgb/article-pdf/10/3/161/1768560/jgb_10_3_161.pdf by Moldova, Republic of user on 23 December 2020

ventilation. At the south end of the building, where the offices have west and east exposures,
natural cross-ventilation is provided. Operable windows in the west and south facades allow
occupants to control the amount of natural ventilation. Landscaping was used to create a
buffer zone to the east of the offices, keeping out air and noise produced by the neighboring
industrial activities. Solar orientation was also a factor in the design of the west and south
facades. The glazed curtain wall on the south facade uses horizontal shading elements to block
midday sun, while providing unobstructed views to the water. Figure 4 shows natural ventila-
tion and shading strategies.

Figure 4: Passive
design strategies for
shading and natural
ventilation.

168 Volume 10, Number 3

 The western facade consists of an aluminum rainscreen system with punched high-per-
formance windows, and automated exterior blinds. Similar to venetian blinds typically used
for interiors, the closed blinds prevent solar heat gain within the building during afternoon
hours. The south facade consists of a curtain wall with fixed exterior horizontal sunshades,
and fritted glass. On the east and north facades, a rainscreen facade system with corrugated
metal panels was used. The overall window-to-wall ratio (WWR) for all four facades was low,
around 32%. Glass selection was based on the orientation of the windows and the functional
requirements of the interior spaces. The vision areas for all facades consist of double-glazed
air-insulated glazing units with low-e coating. The opaque areas of the facades were designed

Downloaded from http://meridian.allenpress.com/jgb/article-pdf/10/3/161/1768560/jgb_10_3_161.pdf by Moldova, Republic of user on 23 December 2020

for an average thermal resistance of R-19 hr-ft2-F/Btu (3.36 m2-K/W).

Figure 5: Building systems and sustainable design approaches.

Journal of Green Building169

 HVAC systems include a radiant heating and cooling system in the floors, vertical geo-
exchange wells, and a heat-recovery system in the laboratories and office spaces (Figure 5).
Other energy-efficiency and sustainability strategies include vegetated roofs, stormwater col-
lection, water reuse, use of recycled and reclaimed materials, as well as a measurement and ver-
ification system that tracks actual building performance and informs users of real-time energy
use. During the design process, the modeled energy consumption showed that the EUI for
this facility would be 81 kBtu/ft2 (256 kWh/m2). This indicated that energy savings would be
36% compared to an ASHRAE 90.1 baseline building’s EUI of 123 kBtu/ft2 (388 kWh/m2).
The building was completed in 2010, and achieved LEED Platinum certification by the U.S.

Downloaded from http://meridian.allenpress.com/jgb/article-pdf/10/3/161/1768560/jgb_10_3_161.pdf by Moldova, Republic of user on 23 December 2020

Green Building Council.

4. Energy Modeling
The building design incorporated several advanced design methods, as discussed in previous
section. Table 1 shows the modeling capability of such design features by the studied energy
simulation tools.
Table 1: Summary of modeling capabilities of investigated energy simulation tools.

4.1 Energy Modeling with Vasari/GBS

Vasari/GBS is one of Autodesk’s design software tools that integrates conceptual model-
ing with WBEA, allowing the designers to make important design decisions in earlier phases
of the project. It is still in beta version, but it is becoming increasingly used by design profes-
sionals due to its dynamic and integrated features, as well as automated modeling capabilities
that reduce time and effort needed during the conceptual design. The apparent benefit of
this tool is that BIM-based design information and geometry can be used for energy analysis
during the earliest stages of the design process. It supports performance-based design via inte-
grated energy modeling and analysis features. GBS is a web-based energy modeling software
that can be used for early design decision-making, and allows for data exchange between BIM
design programs and an energy modeling engine. GBS differs from Vasari slightly, where the
parameter settings can be altered post-simulation without creating a new project. Another
difference is that GBS offers a more detailed list of component and condition parameters
to change for the design alternatives, such as construction methods and building systems
(R-value of the building envelope, type of glazing, sizing of HVAC equipment, etc.). Vasari/
GBS use DOE-2.2 for its energy analysis engine.
The modeling of the case study building began by modeling its geometry in Vasari, as
seen in Figure 6. Then, inputs for the building’s occupancy patterns, systems, equipment,
lighting and plug loads were selected that describe the building in more detail. Cloud-based
simulations are performed by using Autodesk’s subscription to Autodesk 360, which prepares
the model and uploads to Autodesk’s servers for analysis. The data exchange is performed

170 Volume 10, Number 3

Figure 6: Energy model of the case study building in Vasari.

Downloaded from http://meridian.allenpress.com/jgb/article-pdf/10/3/161/1768560/jgb_10_3_161.pdf by Moldova, Republic of user on 23 December 2020

through the gbXML file schema. Simulations are conducted, and results are returned directly
to Vasari, allowing the user to visualize the basic results. Changes to the model can be made,
and different sets of simulations can be run fairly quickly. For the case study building this was
not conducted since the building is already built and occupied.

4.2 Energy Modeling with Sefaira

Sefaira is a web-based sustainability analysis platform specifically built for conceptual design.
Sefaira is targeted towards architects, engineers, consultants and building designers. It per-
forms whole-building analysis of energy use, carbon and renewable energy potential allowing
designers and architects to explore different design options. The software runs simulations on
a specified geometry from a SketchUp model, and produces results that the designer is able
review, compare, and manipulate in a web-based interface. Sefaira uses the radiant time series
(RTS) method as the core of their proprietary energy simulation engine.
As a general overview, the process in this study was to first build the conceptual model
in Vasari, as outlined in the previous section. In order to have an identical model to compare
WBEA results to Sefaira, the Vasari model had to be imported into SketchUp. Before Sefaira
can run a simulation, building components (i.e. walls, floors, roofs, and glazing) must be
assigned as “entities” using the Sefaira plug-in for SketchUp. This is how Sefaira is able to
assign values and understand the geometries in order to run simulations. Figure 7 shows the
model coming from SketchUp and the overall results.
Building components are automatically organized into categories known as “entities”.
The user can manually select and change the “entities” if certain components are designated
incorrectly. The SketchUp model file then must be uploaded on Sefaira’s website in order for
WBEA simulations to be calculated, which was conducted for the case study building. The
building parameters for space use, zones, HVAC systems, occupancy patterns and loads were

Journal of Green Building171

Figure 7: Energy model of the case study building in Sefaira.

Downloaded from http://meridian.allenpress.com/jgb/article-pdf/10/3/161/1768560/jgb_10_3_161.pdf by Moldova, Republic of user on 23 December 2020

Figure 8: Comparison of actual and all modeled energy usage data.

172 Volume 10, Number 3

then assigned, and simulations were run. Sefaira also offers an ability to create different design
alternatives to investigate different design strategies and their effects on energy consumption,
but that was not conducted for the case study building.

5. Comparison of Results to Actual Building Performance Data

The actual energy usage data for the case study building was collected over a period of one
year, from May 2012 to June 2013. The data was collected almost two years after the building
occupation in order to allow continuous operation of building systems and commissioning.
Since the building is a research laboratory building, the energy usage is relatively constant due

Downloaded from http://meridian.allenpress.com/jgb/article-pdf/10/3/161/1768560/jgb_10_3_161.pdf by Moldova, Republic of user on 23 December 2020

to equipment loads and cooling loads that are present during the entire year. The measured
EUI for the building is 94 kBtu/ft2 (321 kWh/m2), and the total annual energy consumption
is 4,774 MBtu (1,399 MWh).
Figure 8 shows a summary of results, and comparison between modeled energy usage
and actual energy usage data for the two different software programs. Modeled energy data
from Vasari/GB is significantly higher than the actual data, while modeled energy data
from Sefaira is comparable (but does not match the monthly energy loads for the build-
ing). Table 2 shows comparison between the monthly modeled data and measured values,
and Table 3 shows a comparison between the overall modeled and measured EUIs.

Table 2: Comparison of monthly modeled and measured energy consumption data.

Table 3: Comparison of modeled and measured EUIs.

Comparing the modeled energy usage results from Vasari/GBS to the actual per-
formance data, it is evident that the modeled energy usage in Vasari/GBS is significantly
higher than the actual. Simulation results from Vasari/GBS indicated that lighting loads
would be a significant part of the overall energy usage for the building and that it would

Journal of Green Building173

be constant throughout the year, which is not the case. The building design incorporated
several advanced design methods for providing daylight (narrow building plate, daylight
redirecting mechanisms such as light-shelves, occupancy sensors), which the modeling
capabilities in Vasari/GBS cannot fully support. Therefore, it was found that modeled
energy data from Vasari/GBS may not accurately portray the effects of advanced design
strategies on energy consumption. The modeled EUI from Vasari/GBS for the building
was 148 kBtu/ft2 (505 kWh/m2), and the overall annual consumption was 7,527 MBtu
(2,206 kWh).
Comparing the modeled energy usage results from Sefaira to the actual performance

Downloaded from http://meridian.allenpress.com/jgb/article-pdf/10/3/161/1768560/jgb_10_3_161.pdf by Moldova, Republic of user on 23 December 2020

data, it is evident that the monthly energy usage data is close to the actual energy usage
data, but there are some discrepancies for the monthly loads. Generally, the modeled
energy consumption is lower than the actual energy usage for the majority of months.
However, summer loads tend to be lower than the winter loads, which even for the tem-
perate climate of Tacoma is typically not the case. The modeled EUI from Sefaira for the
building was 95 kBtu/ft2 (324 kWh/m2), and the overall annual consumption was 4,821
MBtu (1,413 kWh).
When we analyze monthly results, results from Vasari/GBS are higher than actual
energy consumption for all months, and especially during summer season. Monthly results
from Sefaira show variation compared to the actual data. For example, modeled results are
higher than actual for January, March, May, June, November and December. During Febru-
ary, July, September and November are lower; and are comparable during April and October.
These results suggest that significant deviations may exist between modeled and
actual energy consumption for buildings, and more importantly between different simu-
lation software programs. The two investigated simulation programs are geared towards
early stages of the architectural design process, and designers should be cautious in select-
ing and choosing the appropriate tools to investigate energy performance during the early
stages of design process. Understanding the limitations and suitability of specific simula-
tion programs is crucial for successful integration of performance simulations with the
design process. Especially important are the early stages of the design process, since the
design decisions at these stages can have a significant impact on the building performance.
Therefore, it is necessary to select the tools and applications that have the ability to take
into account passive design strategies (shading, building envelope treatment, daylighting,
natural ventilation), as well as active design strategies (HVAC equipment) and occupancy/
building operation inputs, and infer how combined energy-efficiency design strategies will
influence building performance.

6. Conclusion and Future Work

Design of energy-efficient and high-performance buildings requires that building performance
and simulations tools are used and integrated with the design process. The purpose of this
research was to document a comparative analysis of different simulation tools that are appro-
priate for early conceptual stages of the design process. We analyzed and simulated energy
usage of a recently constructed and occupied building in two different programs aimed for
early energy analysis (Vasari/GBS and Sefaira), and compared it to the actual measured energy
usage. The results show that there may be large discrepancies between simulated results and the
actual energy data. Moreover, results have indicated that different simulation programs may

174 Volume 10, Number 3

provide different results, where some tools may predict significantly larger energy consump-
tion and some tools may predict lower energy usage in buildings, compared to actual mea-
sured data. This is a major concern, since the design and construction industry may require
performance-based design and delivery in the near future, where design professionals might
be legally liable for the energy performance of the buildings that they design. This changing
paradigm would require use of simulation and modeling tools from the earliest stages of the
design, through design development. Therefore, it is crucial to understand the limitations of
different tools in order to successfully integrate building performance analysis in early stages
of the design process.

Downloaded from http://meridian.allenpress.com/jgb/article-pdf/10/3/161/1768560/jgb_10_3_161.pdf by Moldova, Republic of user on 23 December 2020

Next step of this research is to extend the simulations and modeling, and complete
energy models for the case study building in two additional software programs: eQuest and
EnergyPlus. These two software programs are geared more towards the schematic design
and design development stages of the design process. Those results will give us an insight
into similarities and discrepancies between results coming from conceptual energy model-
ing tools, more robust modeling tools geared towards later stages of the design process, and
the actual energy usage.

References
Agami, R. 2006. “Literature Review on Calibration of Building Energy Simulation Programs: Uses, Problems,
Procedures, Uncertainties, and Tools.” ASHRAE Transactions 112 (1): 226-240.
Aksamija, A. 2009. “Integration in Architectural Design: Methods and Implementations.” Design Principles and
Practices: An International Journal 3 (6): 151-160.
Aksamija, A. 2010. “Analysis and Computation: Sustainable Design in Practice.” Design Principles and Practices:
An International Journal 4 (4): 291-314.
Aksamija, A. 2012. “BIM-Based Building Performance Analysis” In Proceedings of ACEEE 2012 Summer Study
on Energy Efficiency in Buildings, http://www.aceee.org/files/proceedings/2012/data/papers/0193-000367.pdf
Aksamija, A. and A. Abdullah. 2013. “Building Technology Research in Architectural Practice: Lessons Learned
from Implementations of Energy-Efficient Advanced Building Technologies” In Proceedings of ACEEE 2013
Summer Study on Energy Efficiency in Industry.
Aksamija, A. and Z. Mallasi. 2010. “Building Performance Predictions: How Simulations Can Improve Design
Decisions.” Perkins+Will Research Journal 2 (2): 7-32.
ASHRAE. 2001. ANSI/ASHRAE Standard 140-2011 Standard Method of Test for the Evaluation of Build-
ing Energy Analysis Computer Program. American Society for Heating, Refrigeration and Air-Conditioning
Engineers.
Attia S., Beltran L., de Herde A. and J. Hensen. 2009. “Architect Friendly: A Comparison of Ten Different
Building Performance Simulation Tools”, In Proceedings of IBPSA ‘09 Buildings Simulation Conference, 204-
211. Glasgow: The International Building Performance Simulation Association.
Augenbroe, P., de Wilde, H., Moon, J. and A. Malkawi. 2004. “An Interoperability Workbench for Design
Analysis Integration.” Energy and Buildings, 36 (8): 737-48.
Crawley, D., Hand, J., Kummert, M. and B. Griffith. 2008. “Contrasting the Capabilities of Building Energy
Performance Simulation Programs.” Building and Environment 43 (4): 661-673.
de Wilde, P. 2014. “The Gap Between Predicted and Measured Energy Performance of Building: A Framework
for Investigation.” Automation in Construction 41 (1): 40-49.
Diamond, R., Opitz, M., Hicks, T., Neida, B. and S. Herrera. 2006. Evaluating the Energy Performance of the
First Generation of LEED-Certified Commercial Buildings, Lawrence Berkeley National Laboratory.
Fowler, K. and E. Rauch. 2008. Assessing Green Building Performance: A Post Occupancy Evaluation of 12
GSA Buildings.
Fumo, N. 2014. “A Review on the Basics of Building Energy Estimation.” Renewable and Sustainable Energy
Reviews 31(1): 53–60.

Journal of Green Building175

Gratia, E. and A. de Herde. 2002. “A Simple Design Tool for the Thermal Study of an Office Building.” Energy
and Buildings, 34: 279-289.
Ham, Y. and M. Golparvar-Fard. 2013. “EPAR: Energy Performance Augmented Reality Model for Identifica-
tion of Building Energy Performance Deviations between Actual Measurements and Simulation Results.”
Energy and Buildings 63: 15-28.
Katunsky, D., Korjenic, A., Katunska, J., Lopusniak, M., Korjenic, S. and S. Doroudiani. 2013. “Analysis of
Thermal Energy Demand and Saving in Industrial Buildings: A Case Study in Slovakia.” Building and Envi-
ronment (67): 138-146.
Korjenic, A. and T. Bednar. 2012. “Validation and Evaluation of Total Energy Use in Office Buildings: A Case
Study.” Automation in Construction 23: 64-70.
Maile, T., Fischer, M. and V. Bazjanac. 2007. “Building Energy Performance Simulation Tools – a Life-Cycle

Downloaded from http://meridian.allenpress.com/jgb/article-pdf/10/3/161/1768560/jgb_10_3_161.pdf by Moldova, Republic of user on 23 December 2020

and Interoperable Perspective”, Center for Integrated Facility Engineering Working Paper #WP107, Stanford
University.
Newsham, G., Mancini, S. and B. Birt. 2009. “Do LEED-Certified Buildings Save Energy? Yes, but…” Energy
and Buildings 41(8): 897-905.
Neymark, J., Judkoff, R., Knabe, G., Le, H-T., Durig, M., Glass, A. and Zweifel, G. (2002). “Applying the
Building Energy Simulation Test (BESTEST) Diagnostic Method to Verification of Space Conditioning
Equipment Models Used in Whole-Building Energy Simulation Programs.” Energy and Buildings 34: 97-931.
NREL. 1995. International Energy Agency Building Energy Simulation Test (BESTEST) and Diagnostic
Method. National Renewable Energy Laboratory. NREL/TP-472-6132.
Punjabi, S. and V. Miranda. 2005. “Development of an Integrated Building Design Information Interface”, In
Proceedings of IBPSA ’05 Buildings Simulation Conference, Montreal, pp. 969-976.
Ryana, E. and T. Sanquist. 2012. “Validation of Building Energy Modeling Tools under Idealized and Realistic
Conditions.” Energy and Buildings 47: 375–382.
Scofield, J. 2009. “Do LEED-Certified Buildings Save Energy? Not Really…” Energy and Buildings 41(12):
1386–1390.
Stoppel, C. and F. Leite. 2013. “Evaluating Building Energy Performance of LEED Buildings: Identifying Poten-
tial Sources of Error through Aggregate Analysis.” Energy and Buildings 65: 185-196.
Turner, C. 2008. LEED Building Performance in the Cascadia Region: A Post Occupancy Evaluation Report,
Cascadia Region Green Building Council.
Turner, C. and M. Frankel. 2008. Energy Performance of LEED for New Construction Buildings, New Build-
ings Institute.
Wetter, M. 2011. “A View on Future Building System Modeling and Simulation”, In Building Performance Simu-
lation for Design and Operation, Hensen, J. and Lamberts, R., (eds.), Routledge, UK.
Widener, D. 2009. Regional Green Building Case Study Project: A Post-Occupancy Study of LEED Projects in
Illinois: Year 1 Final Report.

176 Volume 10, Number 3