Energy and Buildings 38 (2006) 421428

www.elsevier.com/locate/enbuild

Applying computer-based simulation to

energy auditing: A case study
Yimin Zhu *
Department of Construction Management, College of Engineering and Computing, Engineering Centre,
Florida International University, Miami, EC 2956, 10555 W. Flagler Street, Miami, FL 33174, USA
Received 15 March 2005; received in revised form 20 July 2005; accepted 29 July 2005

Abstract
Through a case study, this research explores an approach, which uses computer simulation technology, to evaluate different energy conservation
alternatives and to assist facility managers to select reliable and feasible solutions. The subject facility is located in the Southeast region of the
United States. One of the major challenges and operation goals of the General Services Administration, who manages the facility, is for that facility
to achieve the Energy Star designation. However, due to the complexity of the facility, the requirements from building occupants, as well as other
difficulties, finding a path for optimizing the operation of the facility in order to achieve the Energy Star designation is not always easy.
This project uses eQuest, a simulation software tool, to create a virtual environment, in which the operations of the HVAC (heating
ventilation air-conditioning) system and the lighting of the facility are studied. Subsequently, recommendations initially made by experts through
traditional energy audit approaches are evaluated in the virtual environment in order to determine the best solution to achieve the goal of the
facility managers. This paper discusses major aspects of the project, including the challenges, the values and the limitations of applying computer
simulation techniques in such a facility with complicated structural, occupancy and operation features.
# 2005 Elsevier B.V. All rights reserved.
Keywords: Energy efficiency; Computer-based simulation; Energy audit

1. Background
Energy consumption of buildings is a research topic that
retains a lot of research attention [e.g., 15], since many studies
suggest that energy consumption is one of the major types of
cost throughout the lifecycle of a building. In addition, due to
the rising cost of energy, looking for efficient ways of building
operation such as load management [6] or renewable energy
[7] becomes more and more important for facility managers,
who often require a tool that can assist them for decisionmaking regarding various alternatives of a retrofitting project or
the alteration of an existing facility.
Facility managers, nowadays, face many technical challenges, including identifying problematic areas in a facility,
isolating different types of problems, prioritizing the impact of
those problems, developing solutions and implementing
selected solutions, as well as psychological challenges, i.e.,
to accept risks associated with the selected solution. Once an
* Tel.: +1 305 348 3517; fax: +1 305 348 6255.
E-mail address: zhuy@fiu.edu.
0378-7788/$ see front matter # 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.enbuild.2005.07.007

energy saving plan is decided and executed, the process is

difficult to reverse. In addition, many energy-auditing studies
are often constrained by budget and time, yet expectations from
the results of those studies are usually high. All those factors put
a facility manager in a very risky position at the time of making
decisions in order to ensure that, the selected energy
conservation plan will deliver whatever it promises.
In general, energy auditing has been an effective tool that can
assist facility managers to develop their energy saving plans and
to achieve their energy saving goals [e.g., 4,812]. However,
many existing energy auditing approaches may overlook the
intricate relationships between different factors that will affect
the energy consumption of a large facility [e.g., 10,13], and the
cross-effect that two or more different solutions may result in.
Therefore, a more effective validation tool is needed to fine-tune
the results from an energy auditing process.
Computer-based simulation is accepted by many studies as a
tool for evaluating building energy [e.g., 14,15]. There are
many different types of computer-based simulation tools that
are available for performing whole-building simulation, e.g.,
the application of DOE 2 [10,16] and an expert system for

422

Y. Zhu / Energy and Buildings 38 (2006) 421428

supporting energy auditing [17]. eQuest, a quick energy

simulation tool developed by Doe2.com, is a sophisticated,
yet easy to use, freeware building energy use analysis tool
which provides professional-level results with an affordable
level of effort (http://www.doe2.com). This is the reason that
this study chooses to use eQuest over many other simulation
tools such as DOE2 (http://www.doe2.com) and EnergyPlus
(http://www.eere.energy.gov/buildings/energyplus/).
Another issue related to an energy audit is benchmarking.
The Energy Star Program, developed and maintained by the
Environmental Protection Agency in the United States,
provides an excellent source for benchmarking building energy
consumption. The Energy Star Program is a voluntary labeling
program of the United States Environmental Protection Agency
(EPA) and the United States Department of Energy that
identifies energy efficient products, including buildings.
Qualified products exceed minimum federal standards for
energy consumption by a certain amount, or where no federal
standards exist, have certain energy saving features. Such
products may display the Energy Star label. The Energy Star
label is awarded to buildings that perform in the top 25% in the
country (a score of 75 or better out of 100).
In addition, the Environmental Protection Agency
provides an online tool, the portfolio manager, for energy
evaluation of an existing facility with respect to the Energy
Star designation (http://www.energystar.gov/). The software
tool allows a user to manage several facilities at the same
time. To evaluate each facility, a user needs to input facility
data (e.g., the name and the address of a facility and contact
information of a facility manager), facility space information
(e.g., the types of spaces, gross floor area, occupants, the
number of computers and operation hours) and energy meters
(e.g., the types of energy, the types of spaces that are
associated with the energy data, time period of the energy
data, the energy data and the associated cost). Once such
information is input to the software, the software tool will
calculate a score for the facility. Therefore, this study will
rely on the online tool provided by the Environmental
Protection Agency for evaluating energy performance of an
existing facility.
In the following, the paper will discuss a case using an
off-shelf software tool, eQuest, as a complementary tool to
validate recommendations from a conventional auditing
process. Details of the modeling process, as well as the results
of using the simulation technology, are also discussed.

Table 1
Building physical attributes and operation data
Floor space (m2)

Operating
hours/week

Space name

Space type

Start date

Data center

Computer
data center
Mercantile
and services
Office

7/26/1999

1040.79

168

7/26/1999

2917.34

40

7/26/1999

134241.18

84

Food service
Office
Total

138199.31

An initial study on the energy performance of the facility

indicates that the facility may not be very far from achieving the
designation of the Energy Star. The initial study, based on the
monthly electricity consumption data from August 2000 to July
2001, showed that the building had a score of 62 for Energy Star
rating, which was lower than the required score, 75, for Energy
Star designation (see Table 2). The analysis also implies that the
energy consumption rate of this building may be above the
national average rate, which is 50 according to the portfolio
manager. In addition, the energy consumption for a typical office
in the USA is US$ 1.5 per square foot (US$ 16.15 per square
meter) per year [18]. For this building, the energy cost per square
foot is US$ 0.98 (from July 1999 to June 2000) (or US$ 10.55 per
square meter) and US$ 1.15 (from July 2000 to June 2001) (or
US$ 12.38 per square meter), which further supports the
observation that the current status of energy consumption is
better than the national average. Other comparisons using US
office cost indexes [18] also reveal similar results.
Before conducting the computerized simulation of the
energy performance of the facility, an ALERT (assessment of
load and energy reduction techniques) was assembled to
perform an assessment of the facility, which resulted in several
recommendations. The experts in the ALERT (assessment of
load and energy reduction techniques) team performed various
studies by working through the facility, interviewing facility
managers, collecting and studying as-built information, studying the building control system and analyzing the data and the
information for developing recommendations.
Among the recommendations, three areas are targeted
for improvements, i.e., heating energy reduction, the HVAC
(heating ventilation air-conditioning) fan runtime reduction
and savings from lighting. Since the implementation of those
recommendations has different cost implications as well,
the facility managers would like to see cost effective solutions,
as well as to have a second opinion on the recommendation.

2. An introduction to the case

The facility consists of a high-rise tower (of 25 stories), a
bridge crossing the Forsyth Street that connects the high-rise to a
12-story mid-rise tower, the mid-rise tower and the 1924 building
(the historic Richs department store) directly connected to the
mid-rise. The total area for these structures is more than 1.4
million square feet (about 130,000 m2) (see Table 1). In addition,
there is a 10-story parking garage attached to the tower building;
four stories of which are located underground and the rest are
above grade and blended in with adjacent office space.

Table 2
Results of the initial energy efficiency assessment
This building
Year ending

National average
Your target

25 July 2001
Score
Energy use (k Wh/m2)
Energy cost ($/year)

62
248.98
1704037

75
212.42
1457501

50
291.84
1927760

Y. Zhu / Energy and Buildings 38 (2006) 421428

423

To meet the needs of facility managers, this project intended

to explore those recommendations by utilizing the computer
simulation tool, eQuest, and the Energy Star evaluation tool, the
portfolio manager. Major activities of the execution plan for the
project include:
1. Modeling by using eQuest.
2. Validating the model with actual electricity data.
3. Evaluating each recommendation, i.e., recommendation 1,
heating energy reduction; recommendation 2, HVAC fan
runtime reduction; recommendation 3, lighting energy
reduction, separately as well as in different combinations.
4. Using the electricity data generated from each recommendation, as well as their combinations, from step 3 to determine
the Energy Star score by using the portfolio manager.
5. Making the final recommendation.

3. Methodology
In the following, the paper will discuss the methodology
applied in this study, including data collection, modeling and
model assessment.
3.1. Data collection
Data acquisition is primarily concerned with collecting data
relevant to the energy performance of the facility to ultimately
simulate the actual energy behavior of the facility. The data
substantially covers fundamental building characteristics such
as geometrical configuration, internal loads, building shell,
energy systems including the water-side and the air-side of the
HVAC (heating ventilation air-conditioning) systems and
operations.
The bulk of collected data is gathered based on the as-built
blueprints (e.g., architectural, electrical, HVAC (heating
ventilation air-conditioning), lighting, etc.) and the specifications of the facility, while interviewing with facility management personnel augments, validates and ensures that data and
information are up-to-date and conforming with the current
building conditions. The building automation system, Metasys
from Johnson Controls, is another substantial data-source for
building operation data. A wealth of data such as operating
hours, and temperature and pressure set points are referenced
from the Metasys.
The following are the major types of the building and the
HVAC (heating ventilation air-conditioning) system data that
are collected for assessing building energy performance (for
details please refer the model itself):
 Geometrical Configuration and building footprint.
 Building shell and construction materials.
 Internal loads including occupancy/un-occupancy loads of
employees during day time and after-hours, office equipment,
lighting systems and heating and cooling loads.
 Operating schedules including occupancy and after-hour
schedules.

Fig. 1. Geometric representation of the facility (south-east view).

 HVAC (heating ventilation air-conditioning) system equipment and operation, including the water-side and the air-side
systems.
3.2. Modeling
The model is created from the building characteristics
acquired from our building survey and the as built documentations, specifications and drawings.
3.2.1. Geometric modeling
In order to create a building model for the eQuest simulation,
a geometric model of the building is created and then the
characteristics of each modeled space (see Figs. 1 and 2) are
specified accordingly. The layout of the geometric model or the
connectivity of the thermal zones is based on the architectural
and the HVAC (heating ventilation air-conditioning) drawings.
Afterwards, the HVAC (heating ventilation air-conditioning)
system of the building is modeled, followed by the AHUs
(air handling units), the controlling units in this project.
The geometry model of the facility is first created based on
the world coordinates of the facility and then the model is
rotated 388 clockwise according to the azimuth angle of the

Fig. 2. Geometric representation of the facility (south-west view).

424

Y. Zhu / Energy and Buildings 38 (2006) 421428

Fig. 3. Thermal zone design.

actual facility. The interior walls in the model are classified into
two types: solid wall and air wall. The interior partitioning in
the model, which forms the space boundaries of thermal zones
of the model, is modeled by two criteria: the thermal zones
according to the HVAC (heating ventilation air-conditioning
system) drawings and the thermal characteristics of the physical
partitions of the floors above or below, for the consistency in
specifying the adjacent space or the space next to or on the
other side of the wall. For example (see Fig. 3), the floors of
space B and C can be specified as interior walls that are next to
space A, instead of specifying the entire floor above A as an
interior wall next to space B or C, because the floors of space B
and C may have different thermal characteristics. Although this
will lead to more thermal zones in the model than those
specified in the drawings, it will improve the accuracy of the
simulation results. On the other hand, with space D (see Fig. 3)
acting as a plenum space above space A, the job can be done
more easily, because no matter how busy the partitions of the
spaces above or below are, the plenum space can be shared as
the next to type of space for the interior wall specification
and can act as a transition to the next floor.
The plenum spaces above the ceiling are modeled to be the
return air spaces for the AHUs (air handling units). Like other
thermal zones, the plenum layout has to be consistent with the
adjacent spaces on the other side of the ceiling or floor in terms
of geometry. To separate each plenum from one another, air
walls are used as partitions. In the high-rise building, the
plenum spaces are partitioned into mainly three parts, the
plenum in: the tower area, the connector area and the bridge
area. The mid-tower building and the Rich building also have
their own plenum spaces to separate the return air, since they do
not share the same HVAC (heating ventilation air-conditioning)
system.
To simplify the model, the model uses multipliers for typical
floors. However, the complexity of the building, especially the
types of walls that differ from one floor to another, limits the use
of the multiplier to only the typical floors of the tower despite
the fact that many other floors have the same, or similar, floor
layout in terms of geometry.
Thermal zones consist of perimeter, core and plenum spaces.
When modeling thermal zones, the perimeter spaces, strips of
space around the building, are modeled as 15 ft (4.6 m) wide;
except for the ones of the bridge, which are modeled as 9 ft
(2.7 m) wide. The HVAC (heating ventilation air-conditioning)
drawings are used when the thermal zones are developed.
The high-rise and the mid-rise have curtain wall systems
with mainly two types of glass. The remainder of the exterior
wall system is masonry concrete blocks masked and concealed

with decorative pre-cast concrete boards on the outside. The

Rich building, with more traditional window openings, is enclosed with pale-yellow brick veneer exterior walls.
The roofing systems are 8-in. (20 cm) concrete roofing slabs
insulated with felt-bitumen and rigid insulating stucco.
Waterproofing and drainage matting are used to enhance the
resistance against the severe weather conditions. The roofing
systems are finished with polished and honed terrazzo marble.
The interior floors are finished with tufted carpet.
The features of the facility discussed above are major
considerations when developing the geometric model.
3.2.2. Internal loads
The types of internal loads considered in the model include
human occupants, overhead lighting, task lighting, plug-loads
and data centers. The data is mostly collected via surveys. In
many cases, data collected is not directly associated with each
thermal zone; rather the data is the total for a specific floor, or
even the entire building. In such cases, this model uses an
averaging approach based on the percentage of a particular
thermal zone area to the total floor area.
When defining internal loads for each thermal zone, eQuest
allows a user to specify many types of information, e.g.,
geometric information about the thermal zones in the model,
energy consumption for equipment and lighting in the zone. In
addition, infiltration methods and day lighting can also be
specified in the model.
3.2.3. Water and air-side systems
The central chiller plant consists of five Trane chillers, four
1310 t units and one 500 t unit, and is located in the basement
of the high-rise, with chilled water pumped throughout the
facility (bridge, mid-rise and 1924 building). An Alfa-Laval
plate heat exchanger allows water-side economizer operation,
with 2000 t of cooling capacity when the outdoor wet-bulb
temperature is 38 8F (3.3 8C) and full cooling tower flow is
maintained. Cooling towers are on the roof. The economizer is
not utilized at the time of this study and thus is not modeled
either.
In the primary loop, the configuration of chillers and
pumps are parallel in that one chiller is served by one pump
for the condensation cycle and another pump for the
evaporation cycle. The configurations of the primary pumps
and secondary pumps are serial. The data required for
simulating the working conditions of chillers, pumps and
cooling towers is obtained from equipment specifications,
design documents and surveys.
The air-side units are floor-by-floor variable speed air
handlers with chilled water coils. Interior zones are controlled
by VAV (variable air volume) boxes to maintain interior space
conditions at reasonable comfort levels (with no heating
available) and exterior zone control is supplemented with
PIUs (powered induction units) with heating coils. Minimal
outdoor air is provided and no air-side economizers are used.
The data for simulating the operation of air-side units is
obtained from equipment specifications, design documents
and surveys.

Y. Zhu / Energy and Buildings 38 (2006) 421428

425

Table 3
Electricity data for model assessment (units: k Wh  1,000,000)
Month

10

11

12

Actual
Simulated

4.24
4.20

3.48
3.46

3.30
3.32

3.10
2.72

2.73
2.75

2.92
2.80

3.00
2.98

2.81
3.12

2.82
2.58

2.51
2.60

2.90
3.14

3.85
3.68

Table 4
Paired-samples t-test
Paired differences
Mean

Standard deviation

Standard error mean

0.0256

0.19125

0.05521

3.3. Model assessment

The model was assessed based on one-year (2001)
electricity data, because that year the operation was relatively
stable. Before or after that year, there were operation changes
made to the HVAC (heating ventilation air-conditioning)
system and eQuest does not support the co-existence of various
operations for simulation, which limits our use of actual
electricity data.
Paired-samples t-tests were applied to compare the data
from simulation and from the actual electricity bills to
determine if there is a significant difference. A statistics
software tool, SPSS version11, was used to facilitate the
analysis.
Table 3 shows 12-m electricity data collected from two
sources, i.e., the actual electricity bills (labeled as Actual)
and the simulation model (labeled as Simulated). Fig. 4
descriptively demonstrates the patterns of the two data series.
The two patterns are very similar except for data for April,
August and September, which show some irregularities. In
order to verify that there is no significant difference between the
two data series paired sample tests were performed. The results
of the tests from the SPSS software are shown in Tables 46.
The probability value, 0.652, labeled as Sig. (two-tailed)
in Table 4 indicates that there is no significant difference
between the two data sets at the significant level of 0.05.
Meanwhile, the correlation analysis shows that these two data
sets are significantly correlated (see Table 5). The statistics

Fig. 4. Actual vs. simulated electricity usage.

d.f.

Sig. (two-tailed)

0.463

11

0.652

95% confidence interval of the difference

Lower

Actual vs. simulated

0.1459

Upper
0.1971

Table 5
Paired-samples t-test correlations

Actual vs. simulated

Correlation

Level of significance

12

0.925

0.05

Table 6
Paired-samples t-test statistics

Actual
Simulated

Mean

Standard deviation

Standard error mean

3.1381
3.1125

12
12

0.50087
0.48769

0.14459
0.14078

(see Table 6) also show that the means for the two data sets are
very similar. The statistics analysis indicates that the model has
generated viable data.
4. Optimization
Before the recommended energy conservation plans were
input to the model for evaluation, observations were made
about the existing facility operation and maintenance. It has
been noticed that the fans of AHUs (air handling units) are
locked into manual mode, meaning that they run all the time
regardless of what the Metasys tells them to do. This has
negative ramifications as to how much fan energy is wasted
especially during unoccupied periods. Also, the Metasys
programming does not map the control of the PIUs (powered
induction units) to an unoccupied setting for controlling
heating, which results in the PIUs (powered induction units)
supplying heating energy even when the central AHU (air
handling unit) fans are off. This means that these terminal units
operate during after-hours to maintain space temperatures at
72 8F (22 8C) or above all the time. Implementation of heating
setback involves mapping the unoccupied period control of the
PIUs (powered induction units) to Metasys, so that the
space heating temperature setpoint is 72 8F (22 8C) during
occupied periods and 68 8F (20 8C) during unoccupied periods

426

Y. Zhu / Energy and Buildings 38 (2006) 421428

(all other times). These two approaches do not cost too much to
implement.
Another more expensive approach, which may result in more
energy savings, is to reduce the lighting energy consumption in
the facility. Current lighting practices and schedules do not
reflect optimized usage and, if monitored and adjusted, may
provide significant energy savings. Main overhead office lights
are supposed to have regular occupancy schedules, yet the
reality indicates that lighting schedules extend after-hours to
allow cleaning crews to perform their jobs but with a huge time
window. This is essentially due to lighting controllers, which
are centralized on each floor so that a whole floor has to be on if
after-hour lighting operation is desired. Compounded with
these practices, the high-rise cafeteria lighting set up
illuminates approximately 7 W/f2 (96 W/m2) and does not
make advantageous use of daylight harvesting of the perimeter
lighting coming through the northern and southern cafeteria
glass curtain walls. On the other hand, a threshold of 1.8 W/f2
(19 W/m2) has been documented based on as-built lighting
drawings and could provide sufficient illumination for the
cafeteria area.
Multiple recommendations are considered to optimize
lighting practices at the facility. Having lights on more than
12 h/day is not effective energy management, so stringent
occupancy schedules are to be enforced to control the facility
lighting systems. For after-hour office demands and/or
janitorial purposes, it is ideal to decentralize lighting
controllers (localization) or install occupancy-sensing light
controls similar to the Environmental Protection Agency
lighting retrofitting project to suit occupancy loads in the
workplace and trim on energy consumption. This can be
accomplished through means of detecting human presence and
signal to turn lights on by floor zones for a specified period.
Decent energy savings can be observed if normal occupancy
schedules have been adapted to control lighting on the bridge
area as well. Furthermore, daylight harvesting along with asdesigned lighting operations has to be considered when
improving cafeteria lighting. Although it is difficult for eQuest
to simulate the schedule, it is possible to find out the results of
reducing lighting in the cafeteria area by reducing the number
of watts per square footage.
Meanwhile, upon reviewing the Environmental Protection
Agencys overhead lighting metered bill for 30 October 2002, a
power density figure of 0.88 W/f2 (9.47 W/m2) has been
measured after the completion of the Environmental Protection
Agencys lighting retrofitting project. Office overhead lighting
load of 1.2 W/f2 (12.92 W/m2) has been substituted for

Fig. 5. Electricity consumption by different scenarios.

Environmental Protection Agencys calibrated figure in all

office zones.
In addition, normal occupancy schedule (6:00 a.m. to
6:00 p.m. with automatic shutdown for the night) takes place
at the bridge area to regulate lighting instead of the 24 h/day,
7 days/week current operational mode. These features are then
introduced to the eQuest model for evaluation. Fig. 5 and
Table 7 show the results of the evaluation.
Table 7 includes four data series by month. The first is the
series of actual data collected from electricity bills of the
facility. The second, third and fourth series include data
generated from the simulation model after optimization steps
were introduced to the model respectively, i.e., heating
setpoints, the fan operation and lighting energy reduction.
Resetting heating setpoints was first introduced to the
system, which resulted in a 4% energy reduction (see Table 7)
compared to the total energy consumption of the Actual data
series. Then optimizing fan operation was added to the model,
which resulted in 17% of energy reduction cumulatively compared to the actual total energy consumption. On top of these
two strategies, lighting energy reduction was introduced to the
model. There was a total energy saving of 22% observed from
the simulation.
5. Energy evaluation
Three data series in Table 7, i.e., the Heating, the
Heating + Fan and the Heating + Fan + Lighting, were then
used as three situations of energy consumption and were input in
the portfolio manager separately for the evaluation. Table 8
below shows the results. The model shows that when
implementing various energy-saving scenarios the facility can
achieve Energy Star label status of 75 or above. Upon inspection,

Table 7
Metered and simulated electricity usage (units: k Wh  1,000,000)
Month

10

11

12

Total

% Saved

Actual
Heating
Heating + fan
Heating + fan + lighting

4.24
3.97
3.30
3.20

3.48
3.22
2.68
2.59

3.30
3.12
2.72
2.58

3.10
2.70
2.41
2.27

2.73
2.75
2.42
2.26

2.92
2.80
2.45
2.30

3.00
2.98
2.60
2.43

2.81
3.12
2.77
2.58

2.82
2.58
2.22
2.07

2.51
2.60
2.32
2.16

2.90
2.94
2.52
2.39

3.85
3.40
2.71
2.62

37.66
36.18
31.12
29.45

4.00
17.00
22.00

Y. Zhu / Energy and Buildings 38 (2006) 421428

Table 8
Energy Star evaluation results
Solution

Floor space (m2)

Actual energy
intensity (k Wh/m2)

Score

Heating
Heating + fan
Heating + fan +
lighting

138199.31
138199.31
138199.31

262.25
245.23
213.39

66
77
80

using only the low-cost solutions, resetting heating setpoints or

fan operation, can reduce the overall energy consumption
significantly, but may not guarantee an Energy Star rating for the
facility considering variations in the simulated results. When
applying a combination of all three strategies, considerable
savings can be attained, which offers a high possibility of leading
to the Energy Star designation (see Table 8)
6. Conclusion
The study has indicated that computer-based simulation is a
valuable technique to assist facility managers in determining
energy conservation solutions. This project, combining the
power of the portfolio manager developed by the Environmental Protection Agency and eQuest by DOE2.com, has
analyzed the operation of an existing facility in order to
formulate methods for achieving the Energy Start designation.
Based on expert observations and qualitative analysis, three
strategies are formulated. This project is to quantify their
cumulative results to determine whether they are sufficient for
the energy saving goal. According to the study, resetting
heating setpoints alone will not lead to sufficient energy savings
to achieve the designation. The combination of resetting
heating setpoints and controlling the fan operation of air
handling units will result in sufficient energy savings, which
may be enough for getting the designation. However, a
combination of the three strategies will most likely be sufficient
enough for achieving the Energy Star designation.
In addition, through this study, some observations regarding
the use of the simulation tool have been made. Although
computer-based simulation offers a valuable tool to energy
studies like the one performed in this project, the process of
creating the simulation model is very time consuming and
resource demanding. Especially, for a complex facility such as
the facility, correctly defining thermal zones becomes a major
challenge, because thermal zones will significantly affect the
simulation results. On the other hand over defining thermal
zones will complicate the work. Another challenge is acquiring
data, which is the base for building a viable model. The success
of a simulation project is often at the mercy of the availability of
data, such as as-built building data, system specifications,
operation schedules and so on. This requirement may put
some facility off the limit for simulation, as there might be
tremendous difficulty in getting proper data. Otherwise, the
accuracy of the model will be compromised. In addition, there
are still limitations in the tools, which prohibit the model from
reflecting the reality. For example, there is only one type of

427

operation schedule that can be used in one simulation

calculation. However, in realty, during the time period that a
model covers, there might be several different operation
schedules of the same type. In existing simulation tools such as
eQuest, this cannot be modeled. All these limitations of the
existing tools will eventually contribute to the disparity
between the results from the actual electricity bills and
the models.
Nevertheless, the simulation offers a reusable and effective
tool for energy efficiency studies. For example, the cross-effect
of two energy saving plans such as lighting energy reduction
and HVAC fan operation schedule can be calculated by the
simulation model, while a manual-based audit may overlook
such effect.
The study has also observed that existing simulation tools
are in general lack of capabilities to integrate with other
software tools such as cost estimation. It would be more helpful
to facility managers, if a simulation tool can also generate a cost
estimate with respect to a specific energy conservation solution,
because eventually cost savings are one of the most convincing
parameters that facility managers will consider in order to make
a decision of selecting an energy conservation solution.
Acknowledgements
The author would like to express thanks to the General
Services Administration field office for providing as-built
information, as well as participation in various surveys. The
author also would like to thank General Services Administration for providing financial support for this project.
References
[1] L. Barelli, G. Bidini, Development of an energetic diagnosis method for
the buildings: example of the Perugia University, Energy and Buildings 36
(2004) 8187.
[2] P. Thollander, M. Karlsson, M. Soderstrom, Dan Creutz, Reducing
industrial energy costs through energy-efficiency measures in a liberalized
European electricity market: case study of a Swedish iron foundry,
Applied Energy 81 (2) (2005) 115126.
[3] M. Bojicb, F. Yika, Cooling energy evaluation for high-rise residential
buildings in Hong Kong, Energy and Buildings 37 (2005) 345351.
[4] S.M. Deng, J. Burnett, A study of energy performance of hotel buildings in
Hong Kong, Energy and Buildings 31 (2000) 712.
[5] C.S. Canbay, A. Hepbasli, G. Gokcenc, Evaluating performance indices of
a shopping centre and implementing HVAC control principles to minimize
energy usage, Energy and Buildings 36 (2004) 587598.
[6] D.V. Bush, J.L. Maestas, Effective load management planninga case
study, Journal of the Association of Energy Engineering 99 (3) (2002) 71
79.
[7] E. Vine, E. Mills, A. Chen, Energy-efficiency and renewable energy
options for risk management and insurance loss reduction, Energy
(Oxford) 25 (2) (2000) 131147.
[8] P.N. Botsaris, S. Prebezanos, A methodology for a thermal energy building
audit, Building and Environment 39 (2) (2004) 195199.
[9] S.T. Anderson, R.G. Newell, Information programs for technology adoption: the case of energy-efficiency audits, Resource and Energy Economics 26 (1) (2004) 2750.
[10] S. Chirarattananon, J. Taweekun, A technical review of energy conservation programs for commercial and government buildings in Thailand,
Energy Conversion and Management 44 (5) (2003) 745764.

428

Y. Zhu / Energy and Buildings 38 (2006) 421428

[11] A. Larsen, M. Jensen, Evaluations of energy audits and the regulator,

Energy Policy 27 (9) (1999) 557564.
[12] V. Butala, P. Novak, Energy consumption and potential energy savings
in old school buildings, Energy and Buildings 29 (3) (1999) 241
246.
[13] W.G. Haman, Total assessment audits (TAA) in Iowa, Resources, Conservation and Recycling 28 (2000) 185198.
[14] Waltz, P. James, Computerized Building Energy Simulation Handbook,
Marcel Dekker, 2000.

[15] M.S. Al-Homoud, Computer-aided building energy analysis techniques,

Building and Environment 36 (2001) 421433.
[16] Simulation Research Group, Lawrence Berkley National Lab, Overview
of DOE 2.2, http://ww.doe2.com/, 1998.
[17] B. Caudana, F. Conti, G. Helcke, R. Pagani, A prototype expert system for
large scale energy auditing in buildings, Pattern Recognition 28 (10)
(1995) 14671475.
[18] Johnson Controls, USA Office Costs Index, http://www.jci.com/ifm/
research/, 2001.