University of Southern California Dissertations and Theses

Energy simulation in existing buildings: calibrating the model for retrofit studies

(USC Thesis Other)

Energy simulation in existing buildings: calibrating the model for retrofit studies

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
ENERGY SIMULATION IN EXISTING BUILDINGS:
CALIBRATING THE MODEL FOR RETROFIT STUDIES

by

Guang Yang

A Thesis Presented to the
FACULTY OF THE USC SCHOOL OF ARCHITECTURE
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF BUILDING SCIENCE

August 2012

Copyright 2012 Guang Yang
ii

ACKNOWLEDGEMENTS

My committee chair Professor Karen Kensek had guided me through the whole
academic program. She had guided me every step and had given large amount valuable
suggestions in research as well as in writing. The thesis could not be accomplished
without her help. I would extend my heartfelt gratitude to her.
Professor Marc Schiler had pointed out many mistakes I made in the thesis and had
given useful advices in solving problems I faced. I would express my great appreciation
to him.
I would acknowledge Tianxin Xing of Glumac. He had helped me a lot in creating models
and calibrating process.
I wish to acknowledge Carol Fern of Energy Services at Facilities Management Services,
USC. She had provided necessary design documents used in cased study.
This research was supported in part by Southern California Edison. My special
acknowledgement goes to Doug Avery.
I would like to thank my friends Sukreet Singh and William Vicent who had helped me
during the research process.

iii

TABLE OF CONTENTS

ACKNOWLEDGEMENTS .................................................................................................................... ii
ABSTRACT ...................................................................................................................................... viii
CHAPTER 1: INTRODUCTION ............................................................................................................ 1
1.1. Why is building calibration important ............................................................................. 1
1.2. What is building energy simulation ................................................................................. 3
1.2.1. Definition of building energy simulation ................................................................. 3
1.2.2. History of building energy simulation ...................................................................... 3
1.3. Usage of building energy simulation ............................................................................... 4
1.3.1. Conceptual Design (Pre-Design)............................................................................... 4
1.3.2. Design Development ................................................................................................ 5
1.3.3. Commission .............................................................................................................. 6
1.3.4. Building Retrofits...................................................................................................... 6
1.4. Energy simulation tools.................................................................................................... 7
1.4.1. DOE-2 ....................................................................................................................... 8
1.4.2. eQUEST..................................................................................................................... 8
1.4.3. EnergyPro ................................................................................................................. 8
1.4.4. EnergyPlus ................................................................................................................ 8
1.4.5. IES-VE ....................................................................................................................... 8
1.4.6. ECOTECT ................................................................................................................... 9
1.4.7. Autodesk Green Building Studio .............................................................................. 9
1.4.8. HEED ......................................................................................................................... 9
1.4.9. DesignBuilder ........................................................................................................... 9
1.4.10. DeST ......................................................................................................................... 9
1.5. Energy Model Calibration .............................................................................................. 10
1.5.1. Advantages and disadvantages of energy simulation ........................................... 10
1.5.2. What is Energy Model Calibration ......................................................................... 12
1.5.3. General Steps of Calibrating an Energy Model ...................................................... 12
1.5.3.1 Create the virtual model. ....................................................................................... 12
1.5.3.2 Comparing calculation results with real energy consumption .............................. 13
iv

1.5.3.3 Analyzing Possible Reasons .................................................................................... 13
1.5.3.4 Improving models .................................................................................................. 15
1.5.4. Choose Proper Software ........................................................................................ 15
1.6. Summary ........................................................................................................................ 16
CHAPTER 2: BACKGROUND ............................................................................................................ 17
2.1. Calibrated Energy Simulation Approach ........................................................................ 17
2.1.1. Making decision of using calibrated simulation method ....................................... 17
2.1.2. Producing Simulation Plans.................................................................................... 19
2.1.3. Collecting Building Information ............................................................................. 19
2.1.4. Creating models, inputting data and starting simulation ...................................... 21
2.1.5. Comparing simulation results with measured data ............................................... 22
2.1.6. Refine models until it is calibrated ........................................................................ 22
2.1.7. Calculating in pre-retrofit (baseline) and post-retrofit (condition) ....................... 22
2.1.8. Estimate and report saving .................................................................................... 23
2.2. Case Study ...................................................................................................................... 23
2.2.1. Basic Information of KAP ....................................................................................... 23
2.2.2. General introduction of details of KAP .................................................................. 24
CHAPTER 3: RESEARCH METHODOLOGY ....................................................................................... 25
3.1. Introduction ................................................................................................................... 25
3.2. Deciding whether or not create a calibrated simulation .............................................. 25
3.3. Producing Simulation Plan ............................................................................................. 26
3.4. Collection Building Information ..................................................................................... 28
3.4.1. Original Drawings and Digital Record Drawings .................................................... 28
3.4.2. Site Survey and Operation Interview ..................................................................... 29
3.4.3. Energy Consumption Data ..................................................................................... 30
3.4.4. Special Equipment measurement .......................................................................... 30
3.4.5. Weather Data ......................................................................................................... 31
3.4.6. Codes and Standards ............................................................................................. 32
3.5. Creating Virtual Models ................................................................................................. 34
3.5.1. Geometric Models ................................................................................................. 34
3.5.2. Thermal Zones ........................................................................................................ 37
3.5.3. Building Panels ....................................................................................................... 38
v

3.5.4. Windows and Doors ............................................................................................... 40
3.5.5. Lighting system ...................................................................................................... 43
3.5.6. Internal Equipment ................................................................................................ 46
3.5.7. Occupancy .............................................................................................................. 47
3.5.8. HVAC systems ........................................................................................................ 48
3.5.8.1 Type of HAVC systems ........................................................................................... 49
3.5.8.2 Details of HAVC systems ........................................................................................ 49
3.5.9. Weather Data ......................................................................................................... 50
3.5.10. Conclusion .............................................................................................................. 50
3.6. Calibrating Models ......................................................................................................... 50
3.6.1. Collecting and displaying energy data ................................................................... 51
3.6.2. Error Analyzing ....................................................................................................... 51
3.7. Revising Models ............................................................................................................. 53
3.7.1. Mistakes made by modelers .................................................................................. 53
3.7.2. Improper simplification .......................................................................................... 54
3.7.3. Analyzing uncertain factors of collected building information .............................. 55
3.7.4. Operation Schedules .............................................................................................. 57
3.8. Conclusion ...................................................................................................................... 58
CHAPTER 4: CASE STUDY ................................................................................................................ 59
4.1. Introduction ................................................................................................................... 59
4.2. Deciding whether or not to create a calibrated simulation........................................... 59
4.3. Producing a Simulation Plan .......................................................................................... 60
4.4. Collecting Building Information ..................................................................................... 60
4.4.1. Original Drawings and Digital Record Drawings .................................................... 61
4.4.2. Site survey and operation interview ...................................................................... 62
4.4.3. Energy Consumption Data ..................................................................................... 62
4.4.4. Special Equipment Measurement .......................................................................... 65
4.4.5. Weather Data ......................................................................................................... 65
4.4.6. Codes and Standards ............................................................................................. 66
4.5. Creating Virtual Models ................................................................................................. 66
4.5.1. Geometric Models ................................................................................................. 66
4.5.2. Thermal Zones ........................................................................................................ 69
vi

4.5.3. Building Panels ....................................................................................................... 70
4.5.4. Windows and Doors ............................................................................................... 71
4.5.5. Lighting system ...................................................................................................... 73
4.5.6. Internal Equipment ................................................................................................ 75
4.5.7. Occupancy .............................................................................................................. 78
4.5.8. HVAC systems ........................................................................................................ 81
4.5.9. Weather Data ......................................................................................................... 84
4.6. Calibrating Models ......................................................................................................... 84
4.6.1. Collecting and displaying energy data ................................................................... 84
4.6.2. Error Analyzing ....................................................................................................... 88
4.7. Revising Models ............................................................................................................. 89
4.7.1. Mistakes made by modelers .................................................................................. 89
4.7.2. Improper simplification .......................................................................................... 89
4.7.3. Operation Schedules .............................................................................................. 89
4.7.4. Final Simulation Results ......................................................................................... 90
4.8. Conclusion ...................................................................................................................... 91
CHAPTER 5: PARAMETERS INFLUENCED SIMULATION RESULTS ................................................... 92
5.1. Introduction ................................................................................................................... 92
5.2. Area of Windows ............................................................................................................ 92
5.3. Panels ............................................................................................................................. 94
5.4. Weather Data ................................................................................................................. 96
5.5. Lighting Power Densities ................................................................................................ 97
5.6. Conclusion ...................................................................................................................... 98
CHAPTER 6: CONCLUSIONS AND FUTURE WORK .......................................................................... 99
6.1. Conclusions .................................................................................................................... 99
6.2. Future work .................................................................................................................. 101
6.3. Summary ...................................................................................................................... 102
BIBLIOGRAPHY ............................................................................................................................. 103
APPENDICES
APPENDIX A: PARAMETERS IN MODELS .................................................................................. 105
vii

APPENDIX B: INTERMEDIATE SIMULATION RESULTS FOR REVISING MODEL .......................... 106
APPENDIX C: INTERMEDIATE SIMULATION RESULTS FOR INFLUENCE ANALYZING ................ 112

viii

ABSTRACT

Buildings in United States are responsible for close to half of the total energy consumed
in the country, with existing building stock as the main responsible factor. While just a
small proportion of that energy is consumed during the extraction of raw materials and
construction phases, the majority of that energy is used for ongoing building operations
after construction is complete (EIA, 2011). New technologies and management
methods of building energy conservation have greatly decreased energy consumption in
new buildings.
Compared with new constructions, energy performance of a large amount of existing
buildings is poor. There are many methods of upgrading these buildings to achieve
lower energy consumption, including upgrading building mechanical systems, using high
quality windows and doors, adding extra insulate layers, etc., could be done in order to
avoid energy wasting and decrease operation price. However, not all these methods
may be very effective if one does not know how the building is currently performing at a
more detailed level than just looking at energy bills. It is very useful perform energy
simulations as part of a well-managed process of improving performance of an existing
building. As a result, calibration of an energy model to the existing building’s actual
performance is considered as one of the best important steps in whole retrofit process.
In order to get an accurate simulation result, proper building information collection and
simplification are needed for simulation. However, different software, even when used
ix

by a careful, knowledgeable person, may have different results even with similar input.
Compared to the real world situation, simplifications of energy models are unavoidable.
Therefore, there are always differences between simulation results and real energy
consumption. Yet, the first step in predicting future savings is to create a calibrated
model that can then be used for energy simulation before the retrofit happens.
KAP on the USC campus was used as a case study for developing the process of creating
a calibrated energy models that could be used for future upgrades to the building
General steps for calibration and energy simulation were also developed and discussed
in this thesis. A set of guidelines has been developed as a result of the calibration study.
1

CHAPTER 1: INTRODUCTION

1.1. Why is building calibration important
Buildings in United States are responsible for close to half of the total energy consumed
in the country, with existing building stock as the main responsible factor. While just a
small proportion of that energy is consumed during the extraction of raw materials and
construction phases, the majority of that energy is used for ongoing building operations
after construction is complete (EIA, 2011). New technologies and management
methods of building energy conservation have greatly decreased energy consumption in
new buildings. For example, in lighting systems, fluorescents and LEDs are widely used,
daylight is often considered in design, and sensors and lighting control system are used
in order to reduce electricity usage. In building mechanical systems, high efficiency
equipment like lithium bromide absorption units, ground-source heat pumps, waste
heat recovery units, etc., have been more and more popular. High-quality windows and
doors would significantly reduce heat transfer and infiltration. New materials like foam
glass have been used in insulation layers. Stricter building codes force designers to pay
more attention on building energy conservation in order to make their buildings meet
code requirements. Some rating systems, like LEED, also attract developer’s interests on
reducing energy consumptions.
2

Compared with new constructions, energy performance of a large amount of existing
buildings is poor (Iron, 2006). There are many methods of upgrading these buildings to
achieve lower energy consumption, including upgrading building mechanical systems,
using high quality windows and doors, adding extra layers of insulation, etc., which
could be done in order to avoid energy wasting and decrease operation expenses.
Moreover, in some conditions, improvement in operation process could also reduce
energy consumption. For example, in some existing HVAC systems, changing some
operating parameters would increase the efficiency of whole systems. In other
conditions, cleaning blocked air channels and pipes will significantly reduce energy
usage of fans and pumps. Under code requirements, one of the simplest ways to reduce
HVAC system consumption is to increase indoor air temperature setting points in
summer and to reduce them in winter.
However, not all these methods may be very effective if one does not know how the
building is currently performing at a more detailed level than just looking at energy bills.
In some projects, because the chief factors which influence energy consumption were
either not found or not dealt with properly, there is no significant energy conservation.
It is very useful to perform energy simulations as part of a well-managed process of
improving performance of an existing building. As a result, calibration of an energy
model to the existing building’s actual performance is considered as one of the best
important steps in whole retrofit process.
3

1.2. What is building energy simulation
1.2.1. Definition of building energy simulation
Building energy simulation means using computer-based tools to simulate energy usage
of a building through a period of time, such as a specified day, an entire year, or whole
building life (ASHRAE14-2002). Building information, including building structure and
orientation, characteristics of external surfaces, characteristics and operation schedules
of HVAC system, characteristics and operation schedules lighting system, location or
weather data, and other related information, would be input in simulation programs.
Many kinds of software include both annual energy consumption and energy
consumption for different usage including cooling, heating, lighting, hot water,
equipment, and other usage, in their simulation results. Usually annual, monthly, daily,
and hourly change curves of energy consumptions, as well as energy usage contribution
will be calculated. In some tools, related parameters like energy cost and CO
2
emission
would be given in final report; while in others, simulation results will be compared with
code requirements.
1.2.2. History of building energy simulation
In 1967, in some Post Office Programs, thermal loads were calculated by response factor
method, and it was considered to be the beginning of building performance simulation.
SHADOW developed at UCLA was the first energy design tools in the world. Blast,
developed in 1976, DOE-2, developed in 1979, and EnergyPlus, developed in 1990, have
4

played an important part in history, and latter two are still widely used till now. (Milne,
2008)

1.3. Usage of building energy simulation
1.3.1. Conceptual Design (Pre-Design)
Energy simulation can be used at the very beginning of the project. Conceptual design,
which is a rough plan of basic ideas about what a building would be like with the
requirements and limits from developers and real conditions, is at the beginning of the
design process and often considered to be the most important part because early
decisions like the orientation of the building (major axis east-west versus north-south,
for example) can have a large influence on energy consumption. Basic elements of a
building, like position, orientation, type, style, etc., would be determined during this
step. Usually the designer is most worried about whether the design is acceptable,
which means his design should either meet code requirements or fulfill the developer’s
requirements. During conceptual design, a simplified model could be built by the
designer with limited information such as site position, building type, basic building
information like stories, mass type, orientation, etc. Even if a large amount of
information is unknown, and simulation results are not exactly accurate, it is still useful
to help designers and developers to have an overall impression of the energy
5

consumption of their buildings. Different studies can be made during conceptual design
that are critical for understanding the potential energy consumption of the building.
1.3.2. Design Development
Detailed energy simulation is more often used during design development. Design
development is the process where details of the building are decided. During this
process, architects and designer of different area would work together, and final
building plans would be completed by the end of this step. The building usually has the
following details already worked out: structure, materials roofs, floors, exterior and
interior walls, windows and doors, lighting system, HVAC system, water system, and
other details of the building. Most of these energy related designed will be finished by
engineers; this type of engineer is called building mechanical engineer. It is important
that they help the architects and other designers reach the owner’s design goal and
fulfill the appropriate codes. After some details of the building have been determined,
the modeler could get simulation results on the influence of these designs, and these
results could help designers to evaluate these designs. For example, when walls and
windows have been determined, what is the best window-to-wall ratio of the building?
Whether high solar heat gain coefficient (SHGC) glass should be used? Should heat
pump be used instead of furnaces in winter? Will this design fulfill sustainability goals
for this project? A large amount of questions could be answered with help of energy
simulation results. Capacities of building mechanical systems could also be determined
6

according to simulation results. With the simulation results, architects and designers
could alter their designs, and modelers would re-calculate these designs.
1.3.3. Commission
After final design is submitted, it is a good idea for a developer to judge whether it is
energy efficient with energy simulation results before starting construction. Usually
engineers from a third party will finish the simulation and develop a report. Simulation
models could also be used for comparisons between proposed designs and code-based
buildings. Energy commissioning results are important support documents for green
building certifications such as LEED.
1.3.4. Building Retrofits
For historical reasons, some existing buildings may not meet current code requirements.
Or a renovation might require that the building be upgraded to new standards and
codes. Even if buildings currently meet code requirements, owners might want to
decrease their energy bills. Building retrofits would be done in order to increase building
energy performances. There are many methods of upgrading these buildings to achieve
lower energy consumption, including upgrading building mechanical systems, using high
quality windows and doors, adding extra insulate layers, etc., could be done in order to
avoid energy wasting and decrease operation price. However, not all these methods
may be very effective. As a result, before retrofits, in order to find out which solutions
7

would beneficially influence energy consumption, energy simulations should be
completed by some service company.
1.4. Energy simulation tools
An energy simulation tool is usually composed of three parts: geometric design module
to create the geometry of the model, parameter input module, and calculation module.
Sometimes the former two modules are combined together and considered to be
interface module.
For most simulation tools, it is sometimes difficult to create accurate geometric models
for a complicated project with their own geometric design modules. There are two ways
to solve this problem: one is to create a simplified geometric model; another is to use
other professional design tools, like AutoCAD or Revit to create the model and import it
into the simulation software program.
Building energy simulation tools have developed rapidly recent thirty years, hundreds of
different tools have been developed. US Department of Energy has listed 129 different
kinds of software.
(http://apps1.eere.energy.gov/buildings/tools_directory/subjects_sub.cfm)
Here are ten software programs that are widely used in the world. (Crawley, et al, 2005)
8

1.4.1. DOE-2
DOE-2 is one of the most widely used energy used analyzing program in the world. It can
calculation energy consumption and cost for almost all type of building. It has also been
used as core engine in many different kinds of simulation software.
1.4.2. eQUEST
eQUEST is a DOE-2 based free tool for building energy simulation. It has developed in
scope and features in recent years.
1.4.3. EnergyPro
EnergyPro is a quick-modeling energy simulation tool related to California Code (Title
24). It provides a shell user interface and uses the DOE-2 as its simulation engine.
1.4.4. EnergyPlus
EnergyPlus is a popular energy simulation program since 2001. It can calculate energy
consumption for complicated buildings with multi-zones. As a core engine, it has been
adopted for use in many other simulation programs.
1.4.5. IES-VE
IES-VE is a tool that can simulate many different things like energy use, CO2 emissions,
occupant comfort, light levels, airflow. Through plug-ins, it can also work as an
extension of Autodesk Revit and Google SketchUp.
9

1.4.6. ECOTECT
ECOTECT has a 3D design interface and module of solar, thermal, lighting, acoustic, and
cost analysis. Data in ECOTECT can easily be export to other software for analyzing.
1.4.7. Autodesk Green Building Studio
Green Building Studio can be used for energy, water, and carbon analysis. DOE-2 is used
as its engine for simulation. Models in GBS can be transferred to formats for DOE-2 and
EnergyPlus. Project Vasari and Revit Architecture from Autodesk are using Green
Building Studio for calculating conceptual whole building energy consumption.
1.4.8. HEED
HEED is a whole building simulation tool used during concept design. It is easy to use
and is applicable for all 16 California climate zones. It can be downloaded free from
Professor Murray Milne’s website (http://www.energy-design-tools.aud.ucla.edu/).
1.4.9. DesignBuilder
DesignBuilder can be used for many types of buildings, including naturally ventilated
buildings, buildings with daylighting controls, double facades, advanced solar shading
strategies etc. EnergyPlus is used as its engine.
1.4.10. DeST
DeST is a fast calculation tool for HVAC system design. It was developed by Department
of Building Science, Tsinghua University, and is widely used in China. With help of DeST,
10

designers could quickly know energy effects of their designs in short time. Only Chinese
versions have been developed for this tool.
1.5. Energy Model Calibration
1.5.1. Advantages and disadvantages of energy simulation
Building energy simulation is a convenient and efficient method to evaluate building
energy consumption. An experienced modeler could create models and receive
comprehensive results in a relatively short period of time. Moreover, effects of design
changes or improvements could easily be anticipated according to simulation results.
However, there is a serious flaw of simulation (not only building energy simulation, but
also all other kinds of simulations): are simulation results believable? There are three
reasons that cause simulation results to be incredible or inaccurate: simplifications,
inaccurate input, and errors made by modelers.
First, for all existing software, there are many simplifications during modeling and
calibration. For instance, natural ventilation is often ignored; however, in some
situations, it has significant influence on energy consumption. Another example is that
air conditioning systems are usually considered to be a simple system with a constant
COP number, while real air-conditioner-efficiency varies with environmental
temperature changes. Excessive or improper simplification, either made by modelers or
by software, may lead simulation results apart from real situation.
11

Inaccurate input might be another important reason that causes error results. Building
information used in simulation is a kind of ideal (which could also be defined as a kind of
simplification), for example, indoor air temperature and humidity might consider being
constant; thermal resistance of walls could be calculated with material types and layer
thicknesses. Nevertheless, characteristics of real products might vary from ideal
condition. These differences may also cause errors.
Besides, there might be input mistakes and improper simplifications during simulation
processes, especially by some inexperienced modelers.
Another cause for error or at least a difference in values between one simulation and
another is that software programs do not always have the same features, and defaults
can be different. For example, the default R-value for a standard 8-inch concrete wall
might different in different programs, while the same thing happens in SHGC of a double
glazed window. Unless modelers are familiar with parameters of real products and do
not use any default settings, these kinds of errors are unavoidable.
Sometimes, because of the way that the model is being used, it is not critical that the
values are exactly correctly. For example, in a conceptual model, the designer might
change the window to wall ratio to give a relative idea of how to understand the
relationship between energy consumption and overall amount of glass. However, for
more detailed studies, especially for retrofits and renovations, a calibrated energy
model is needed.
12

1.5.2. What is Energy Model Calibration
Most simulations are run before a building is built. They use some kind of standard
weather for a given location. This rarely matches the weather for a given specific year,
which may be colder or hotter or both. However, even when the weather for a specific
year and location is available, the simulation results often do not match the recorded
energy consumption for a building. It is important to calibrate energy models before
using simulation results. The most common method of calibrating energy models is to
compare simulation results with real utility data in retrofits. When substantial error
occurs, it should be analyzed if any improvement of the model could be made. It is
possible that the simulation is wrong, but it is most likely that the building was not
designed or commissioned as intended.
1.5.3. General Steps of Calibrating an Energy Model
Basic steps of calibrating an energy model are the following: creating a virtual model,
comparing calculation results with utility data or real energy consumption, analyzing
reasons, and revising the model. Only basic principles are talked here, and more details
about calibrating steps will be discussed in Chapter 2.
1.5.3.1 Create the virtual model.
In this step, a model of the building will be created. Information needed in the model,
including structures, envelope characteristics, HVAC systems, lighting systems, and
other related information are usually gathered from design data. Sometimes, detailed
13

information that cannot be found in design documents has to be tested, calculated, or
estimated. With the model, first draft of simulation results could be calculated.
1.5.3.2 Comparing calculation results with real energy consumption
Many existing buildings will have historical utility cost data. And some of them may have
installed sub-meters and have more detailed data, such as energy consumption every
day, every hour or even every minute, hot water and chilled water usage, electric usage
in lighting system, and mechanical system, and some other data. All these data could be
calculated in the simulation. Ideally these data should match each other; however, there
are usually some differences between them. These differences need to be reconciled.
This can be a difficult process.
According to collected utility data, several conditions, including annual energy usage,
monthly energy usage, daily energy usage in different seasons, hourly energy usage in
typical days, and other details, could be analyzed.
1.5.3.3 Analyzing Possible Reasons
When there are notable differences between simulation results and real data, the
reason for the discrepancies need to be carefully researched. Here are some possible
reasons: use of inappropriate software, inaccurate input, and user errors.
a. Use of Inappropriate Software
14

There might be too many simplifications in the calculation process of some software.
These simplifications may not be ignored in some cases. Besides, some complicated
parts of a building, like glass facades or ground source heat pumps, could not be
simulated in some software.
b. Inaccurate Input
Building information of a real project may not be measured, thus, input parameters
have to be estimated or forecasted. Sometimes characteristics of real building materials
are different from what is given in a material’s handbook, while other parameters of a
real project may vary from design condition. As an important part in model, weather
data, which is usually gathered by weather stations, usually is different from weather
conditions around the building. Besides, measurement errors also exist, which means
utility or real energy usage data might be inaccurate. These inaccurate inputs may cause
unbelievable simulation results.
c. User errors
There might be mistakes made by modelers, especially by inexperienced ones and in
complex cases. For example, there might be errors or impropriate simplifications of
structure while creating a model. Styles and parameters of HVAC systems in model may
be set mistakenly. Moreover, improper building materials might be selected by
modelers who are not familiar with the correct values. Sometimes this kind of errors
may become chief reason for inaccurate simulate results.
15

1.5.3.4 Improving models
Based on analyzed reasons, improvements of energy models could be created.
For software reasons, if simulation software could not meet project requirements, it
should be replaced. For example, some programs like HEED are good for conceptual
design, but may not be used during design development.
Input data needs to be checked carefully, especially for those which might not be very
accurate, for instance, infiltration and average R-value of the envelope, SHGC and U-
value of windows, COP of air conditioners, human and equipment loads, etc.
Whole modeling process need to be examined carefully in order to avoid errors made by
modeler.
1.5.4. Choose Proper Software
There are so many different tools for building energy simulation, and it is important for
engineers to choose a proper one. Here are some suggestions:
a. Choose widely used software. For modelers and researchers who are not that
familiar with these tools, it would be a good idea to start with one that has
already been used widely. Usually there will be less substantial defects in
them or at least more users available who can help get around problems.
b. Clarify what can and what cannot be simulated in software. Different software
has different specializations. Some are good at lighting simulation, while
16

others efficient in calculating ventilation. A proper tool should satisfy
requirements of the objective building.
c. Other factors, such as complexity of software, policy or code requirements,
price, etc., also need to be considered.
1.6. Summary
In order to study future improvements to a building, it is useful to employ simulation
programs. But the results are only as good as the information put in. In order to verify
the results, it is compared with the actual building energy use. From what is learned,
changes are made to the virtual model until it matches closely to the real data.
A calibrated energy model will be constructed for Kaprielian Hall on the USC campus.
Effort will be made to make it as accurate as possible while documenting the changes
that had to be made to the original model to make it match the real data. This thesis is
trying to figure out whether these four kinds of simulation tools are acceptable for
campus buildings. A set of guidelines are will be developed by the as a result of the
result of the calibration and documents.
17

CHAPTER 2: BACKGROUND

2.1. Calibrated Energy Simulation Approach
As an important method of the measurement and verification (M&V), the approach of
calibrated simulation has been described in different standards and guidelines, such as
ASHARE Guideline, FEMP, IPMVP, etc. Generally speaking, the whole approach can be
summarized into 8 steps: deciding if the project qualifies for simulation, producing
simulation plans, collecting building information, creating models and simulating,
comparing simulation results with utility data, refining models, analyzing renovation
effects with simulation results, and estimating and reporting savings. Details of these
steps are discussed below.
2.1.1. Making decision of using calibrated simulation method
At the very beginning of the process, it is necessary to judge that whether calibrated
simulation method are accessible. Two aspects must be considered.
1. Requirement of calibrated simulation method
In FEMP’s M&V Guideline, calibrated simulation can only be used in cases that meet
following requirements (FEMP, 2008):
 “For complex equipment replacement and controls projects
 When interactive effects between ECMs are too complex for retrofit isolation
approaches, but need to be quantified
18

 When the Option C utility data analysis approach is not viable due to the overall
level of savings being less than 20% of metered use
 When complex baseline adjustments are expected during the performance period
 When energy savings values per individual measure are desired
 When new construction projects are involved
 When savings levels are sufficient to warrant the cost of simulation
 When either baseline or performance period energy data, but not both, are
unavailable or unreliable.”

Sometimes a computer model cannot simulate special building types or building
elements. In these conditions, calibrated simulation is not a proper method. EVO-IPMVP
gives some special building types and building elements which are not easy to simulate,
including buildings with large atriums, a large space underground, complex or unusual
shapes and HVAC systems, etc. (EVO, 2010). With development of simulation
technology, many complex situation can or will be possible to simulate.

2. Benefits and Costs
As mentioned in ASHRAE Guideline, “every project must find its own balance between
the benefits and costs of measurement and resultant accuracy” (ASHRAE 14-2002).
Compared with other methods, calibrated simulation can be used to determine
19

individual ECM in complex systems. However, it also needs experienced modelers, and
calculation results are hard for people with little technical background to understand.
Sometimes perhaps retrofit isolation or whole building measurements are better choice
than calibrated simulation. For instance, if retrofit plan in a project is to replace light
equipment to energy efficient lights, retrofit isolation measurement method is a better
choice than calibrated simulation.
2.1.2. Producing Simulation Plans
In this step, the most important thing is to choose a proper simulation program. A
proper simulation should either satisfy project’s requirement or be widely accepted.
FEMP gives several characteristics of simulation programs:
 “The program is commercially available, supported, and documented.
 The program has the ability to adequately model the project site and ECMs.
 The model can be calibrated to an acceptable level of accuracy.
 The program allows the use of actual weather data in hourly format.”
(FEMP, 2008)

2.1.3. Collecting Building Information
Building information is needed either in creating models, or in adjusting accuracy of the
model during calibrating process. Without adequate building information data, accurate
20

simulation results could not be calculated. Even worth, proper simulation model could
not be created.
In FEMP’s Guideline, building information data are divided into eight types:
1. “Utility bill record”
Utility bills are real energy usage data, which usually contains monthly electricity usage
and gas consumption.
2. “Architectural, mechanical, and electrical drawings”
These drawings contain most of design data that are important for creating models.
Usually information of about the geometry, lighting systems, mechanical systems,
windows and doors, etc., are included in these drawings.
3. “Site survey data”
By viewing building operators in site survey, operation conditions and data of building
systems and equipment, including HVAC systems, lighting systems, plug loads, building
envelope and thermal mass, etc. will be gathered.
4. “Short-term monitoring”
Energy (electricity, gas, chilled water, hot water, etc.) consumptions over time are
collected. Usually there are hourly data available.
21

5. “Spot measurements of specific equipment”
For special equipment, like pumps and fans in HVAC system, lights, office equipment,
etc., energy usage capacity records will significantly improve the simulation model.
6. “Operation Interviews”
Buildings often do not operate on designed conditions. Proper operation interviews
should be used to gather this information.
7. “Weather data”
Representative local weather data are required for calibration. Usually they can be
downloaded from Internet, or could be found in database of simulation tools.
Sometimes on site weather data for specific period also exist.
8. “Minimum code performance standards”
Most buildings should meet code requirement based on location. For instance, ASHREA
90.1 is widely used all around U.S., while Title 24 is used in California. These codes or
standards should also be collected, read, and understood before creating models.
(FEMP, 2008)
2.1.4. Creating models, inputting data and starting simulation
With necessary building information, a virtual model could be created. In this step, all
required building information should be transferred to proper format for input. As
22

mentioned in FEMP, “the more site-specific data incorporated the more accurate the
savings calculations, but the greater the costs” (FEMP,2008). During this process,
accuracy of input date needs to be adjusted.
2.1.5. Comparing simulation results with measured data
This step is greatly on both resolution of exist energy usage data and accuracy
requirement of the project. Monthly utility bill or hourly electricity usage or energy flow
could be used during this process. Real energy consumption data should be examined
carefully to find any errors.
2.1.6. Refine models until it is calibrated
In many cases, there are considerable gaps between measured data and simulation
results. Modelers should study reasons for these gaps. Usually models will be revised to
make calculation results match measured data. When differences are acceptable, the
model could be considered as calibrated model. Error analysis is needed in this step, and
CVRMSE and MNBE are two indices that most often uses. Details about these indices
will be discussed in Chapter 3.
2.1.7. Calculating in pre-retrofit (baseline) and post-retrofit (condition)
With calibrated models, energy consumption in before and after the retrofit could be
calculated. During this process, the simulation model may need refinement based on
the energy conservation methods used in retrofits. Modelers should pay attention to
these changes in order to verify accuracy of simulation results.
23

2.1.8. Estimate and report saving
With calculated results from calibrated models, savings could easily be calculated.
IPMVP gives a formula for saving calculation. (EVO, 2010)
“Savings = Baseline energy from the calibrated model without ECMs – Reporting-period
energy from the calibrated model with ECMs”
If there are sufficient energy consumption data before retrofit, measured energy usage,
which is more accuracy, could be used to replace calculated result with ECMS.
Results and details during calibrated simulation should be contained in final reports.

2.2. Case Study
In this thesis, Kaprielian Hall (KAP) is chosen as a case study.
2.2.1. Basic Information of KAP
KAP (Kaprielian Hall) a four floor building built in 1989, located near Gate 6 on USC
campus. Most rooms in the building are offices and classrooms. Three different
departments of USC are in it. There are two main reasons to choose KAP as a case study:
it is a fairly common building type and has a good amount of historical energy data.
24

2.2.2. General introduction of details of KAP
There are four floors above ground and one floor basement underground of the building.
Brick walls and concrete floors are used in KAP. The flat roof is also made up of concrete.
Windows of the building are single glazed with aluminum frames.
As a part of USC campus chilled water loop, temperature of most parts of the building
are controlled by VAV systems with chilled water from campus. In central area of
basement, a small water system with chilled coil units is used.
Hourly chilled water consumption data and 15-minute-electricity usage data of the
building are collected. No separate gas or hot water meter was assembled for KAP, thus
there’s no real usage data of heating and domestic hot water.
More detailed building information will be given in Chapter 4 as it applies to the energy
simulation.

25

CHAPTER 3: RESEARCH METHODOLOGY

3.1. Introduction
In chapter 2, the basic steps of creating a calibrated simulation are introduced. In this
chapter, more details about the whole process are described. The intent of this chapter
is to be an introductory guideline of how to create calibrated simulation for engineers,
students, or researchers. There is no fundamental difference between the approach or
steps introduced in some standards or articles and intents in this chapter. The method
discussed here is summarized based these existing guidelines and some experience in
case studies, and it is more suitable to use. Chapter 4 will describe the KAP case study
using the method outlined here.
3.2. Deciding whether or not create a calibrated simulation
The goal of this step is to determine whether calibrated simulation is appropriate for a
specific project. As mentioned in 2.1, two questions should be asked by engineers: one
is whether simulation method is useful for the project; another is if simulation method
is worthy of the project.
For the first question, 2.1.1 gives some rules and requirements in standards. In fact, with
development of modeling techniques, more buildings and conditions that were
considered difficult to simulate several years before could be simulated today. However,
there are still some complicated cases that could not be simply simulated with building
26

energy models. Sometimes, CFD (computational fluid dynamics) is an appropriate
alternative choice.
In most whole building renovations, simulation is a viable choice. However, sometimes
creating a calibrated model is not sensible or cost-efficient. An extreme example is that
in a very simple case, when the only retrofit approach is using new energy-efficient
fluorescent lamps to replace old, high wattage incandescent bulbs. Isolation
measurement or utility data analysis methods are much better than whole building
simulation method.
Generally speaking, many aspects must be considered in this step. Three chief factors
that influence developers and engineers to decide include building conditions, retrofit
goals and methods, and budget requirements.
3.3. Producing Simulation Plan
The complexity of the simulation model needs to be decided; this choice is not
irrevocable, but is useful to consider when determining the desired use of the model.
Similarly with the previous step, building condition, simulation goals, and budget limit
should be considered in order to determine whether be detailed simulation or a quick-
and-dirty model is more appropriate.
Simulation software should to be chosen in this step. Appropriate software means it
either satisfies project requirement or could be accepted by modelers. Since there are
27

numerous energy simulation tools, it can be difficult to make a decision. Here are some
suggestions for choosing a proper tool:
a) Choose familiar software
Modelers could easily find out whether a tool reaches project’s requirements when they
are familiar with it. Besides, working with familiar simulation software will also greatly
increase work efficiency during the simulation process.
b) Choose popular software
There must be some reasons that caused the software to be popular. Usually popular
tools have more advantages while their simulation results are more acceptable. Often
more help is available, both formal (for example, the software manufacturers website)
or informal (for example, blogs about the software) as more people are using the
software).
c) Choose software that does not need extra payment
For modelers, there is no fundamental difference among most popular software like
eQuest, EnergyPlus, or IES-VE. Some of them have different interfaces but same core
engine, such like eQuest and EnergyPro. As a result, it is a good idea for entry-level
students or researcher to choose some free software (or software which has a free
education version). For engineers, they may use tools that have already been purchased
by their companies.
28

Generally speaking, in this step, modelers need to choose acceptable simulation tools
and decide detail levels in future modeling.
3.4. Collection Building Information
Building Information covers almost all aspects of a building or project. As one important
area of Building Information Modeling (BIM), only a part of building information are
needed for calibrated building energy simulation, thus modelers only need to gather
information from related areas. The following information should be available for the
simulation team to use: original drawings or digital record drawings (formerly known as
"as-builts"), site survey and operation interview, energy consumption data, special
equipment measurements, local weather data, and applicable codes and standards.
3.4.1. Original Drawings and Digital Record Drawings
There are different types of drawings for a building, such as architectural drawings, civil
drawings, mechanical drawings, electrical drawings, and landscape drawings. For energy
simulation, usually architectural drawings, mechanical drawings, and electrical drawings
are needed (FEMP, 2008).
From the original drawings, many parameters needed in creating models could found,
including geometric structure, characteristics of building envelope, thermal zones,
mechanical system, lighting system, etc. Details about these elements will be discussed
later in the modeling part. For some more recent constructions, 3D models are already
29

finished by architects and engineers and could conveniently be used in energy
simulation.
There might be changes and renovations on the building after the initial construction,
and these changes may or may not be found in drawings. These possible changes should
be carefully checked when using drawings. More recently digital BIM are available that
sometimes include the changes to the building made during construction and other
retrofits during the lifetime of the building.
3.4.2. Site Survey and Operation Interview
Through interviewing building operators, more building information, especially some
not contained in the design drawings, will be collected. FEMP had concluded that
equipment and operation policies of HVAC systems, lighting system, operation
schedules and occupant numbers, plug loads, and other possible energy consumers
need to be contained in site surveys (FEMP, 2008). However, in many situations it is
impossible to get all these information in site survey; but for a detailed study, the more
complete, accurate, and up-to-date information that can be gathered, the better.
In addition to interviewing building operators, another approach is visiting the target
building. It could either help modelers gathering necessary or optional building
information which may not found before, or double check building information already
collected.
30

3.4.3. Energy Consumption Data
In a simulation of retrofit project, utility data are considered to be one of the most
important parts, for they are used to calibrate calculation results. Usually electricity and
gas consumption will be included. Sometimes “peak electric demand” is required (FEMP,
2008). Other kind of energy resource like hot or chilled water might also be contained
based on different system styles.
Utility bills are usually monthly because of energy supplier’s policies. In some cases,
energy consumption data are monitored and recorded hourly or every 15 minutes.
These records might be more helpful in calibrating process, while they need to be
examined carefully, since sometimes monitors might not work in correct modes or do
not work at all intermittently.
3.4.4. Special Equipment measurement
Capacities and operation schedules of building equipment, including fans, pumps, and
chillers in HVAC system, lighting system, plug loads, and any other special equipment of
the building, can be collected through the design documents and site survey. However,
real products and operation conditions are different from ideal conditions and may
cause significant errors in simulation results. One way to solve this problem is to
monitor parts where the data may be suspect. This use of monitored data will make the
model more accurate.
31

However, measurements will significantly increase complexity and cost of calibrated
simulation. In many simulation cases, especially for entry level studies, this step might
be skipped.
3.4.5. Weather Data
For most regions and cities, weather files for a standard-year can be found. These data
are generated based on historical climate data of multiple years and are used to present
local weather condition. They are widely used in designing and modeling.
Standard-year climate data are usually published by the government and are freely
available for thousands of locations across the world. Some institutes, like energy
simulation software companies, will collect these data, and may generate weather files
in different format. There are different kinds of weather file format, such as DOE-2
weather file (.bin), EnergyPlus weather files (.epw), ASHRAE Design Conditions Design
Day Data file (.ddy), summary report on the data (.stat), California Climate Zones
Revision 2 (.cz2), etc. Sometimes modelers have to use some existing software programs
to convert files to get proper format.
EnergyPlus has collect weather files for more than 2000 locations. All these data file can
be downloaded from its website:
http://apps1.eere.energy.gov/buildings/energyplus/cfm/weather_data.cfm.
Even though standard-year weather files are widely used, in calibrated simulation
process, weather conditions for specific year are different from the standard condition,
32

and it would an important reason that caused the simulation results to be different from
the utility data.
One way to solve this problem is using historical weather data instead of general
weather data. Usually these historical data of calibrated period were site-monitored or
came from a weather station nearby. Often, it is difficult to find these specific weather
data. Some private companies sell this information for different locations.
3.4.6. Codes and Standards
Sometimes it might be a little confusing for entry-level engineers or architects to
understand the difference between codes and standards. A simple explanation made by
National Fire Protection Association is that “a code tells you what you need to do, and a
standard tells you how to do it”. (NFPA, http://www.nfpa.org/ )
For all building design or redesign projects, a set of codes and standards must be
followed through whole process. Since usually it is designers’ responsibility to decide
which codes and standards are used in a project, original documents or parameters used
in modeling might already follow these requirements. However, a set of codes and
standards still need to be familiar by modelers, either in guiding simulating and
calibrating process, or in judging the following retrofit methods. Below is a list of codes
and standards that often used in building energy simulation.
1) ANSI/ASHRAE 90.1, Energy Standard for Buildings Except Low-Rise Residential
Buildings. This is an energy related code, which is widely used in U.S. and all
33

around world. It is also used in LEED rating system. The latest version is ASHRAE
90.1-2010.
2) California Title 24, Building Energy Efficiency Standards for Residential and
Nonresidential Buildings. Building energy code used in California. The latest
version is 2008.
3) ASHRAE Guideline 14, Measurement of Energy and Demand Savings. This
guideline introduced standard procedures of energy saving calculation with
different methods, among which calibrated simulation is an important part. The
latest version is ASHRAE Guideline 14-2002.
4) FEMP M&V Guidelines: Measurement and Verification (M&V) for Federal Energy
Projects. A standard that can be used for guiding energy simulating and
calibrating procedures. Widely used in U.S. The latest version is V3.0 published
for 2008.
5) IPMVP (International Performance Measurement and Verification Protocol):
Concepts and Options for Determining Energy and Water Savings Volume 1.
Another standard, similar to FEMP M&V Guidelines, and more global accepted.
The latest version is IPMVP-2012.
Based on specific locations and rating requirements, other local codes or standards
might be needed for energy calculation.
34

3.5. Creating Virtual Models
After creating the simulation plan and collecting sufficient building information, a virtual
model can be created for energy calculation. Details for modeling in different software
are specifically dependent on that program; however, the basic principles are similar.
The procedure for this step is to transfer collected building information and to input
them in simulation tools. There are many parameters that need to be considered in the
modeling process: geometry, thermal zones, building panels, windows and doors, the
lighting system, internal equipment, occupancy and their schedules, HVAC systems, and
the weather data, which were mentioned earlier.
3.5.1. Geometric Models
All simulation software needs a 3d geometric model for energy calculation. Building
information, including the envelope’s shape, floor height, interior areas, and building
orientations, are included in the geometric model. In most cases, geometric models are
created based on original architectural drawing, including floor plans and elevations, or
3d building information models.
General speaking, there are two approaches in creating a geometric model: within the
interface of the selected energy simulation tool or creating the 3d model in another
software program specifically designed for 3d modeling, like Revit or SketchUp, and
transferring the model to the simulation software.
35

The chief advantage of creating a geometric model in the energy software is that it will
make the modeling procedure proceed more smoothly with fewer errors. However,
since many simulation tools have simplified 3d modeling module in their interface, while
target buildings might have complicate shapes and structures, it may impossible or not
worthwhile to create the geometry models within the simulation tools.
As a result, it has become a comprehensive adoptable solution to create geometry
models in design tools. While choosing this approach, an important question must be
asked that whether the geometric model generated in design tools can be transferred to
a format that can be accepted by energy simulation model. If the answer is no, then
some intermediate tool has to be used in format transfer. Table 3.1 shows format
transfer procedures between several popular design tools and energy simulation tools.

36

Table 3.1 Format Transfer Procedures of Geometric Models between Different Tools
Design Tool Energy Tool Format Transfer Procedure
Revit eQuest Revit (gbXML) – Green Building Studio(BIN) – eQuest
Revit EnergyPro Revit (gbXML) – EnergyPro
Revit EnergyPlus Revit(gbXML) – Ecotect (IDF) – EnergyPlus
Revit IES-VE Revit – IES-VE (Using IES-VE Plug-in)
SketchUp EnergyPro
SketchUp – DesignBuilder (gbXML) – EnergyPro
(using DesignBuider Plug-in called gModeller)
SketchUp EnergyPlus
SketchUp – EnergyPlus
(using EnergyPlus Plug-in called OpenStudio)
SketchUp IES-VE SketchUp – IES-VE (Using IES-VE Plug-in)
ArchiCAD EnergyPro ArchiCAD – DesignBuilder (gbXML) – EnergyPro
ArchiCAD EnergyPlus ArchiCAD – EnergyPlus (using EnergyPlus Plug-in)

In some cases, 3d BIM models have already existed as original design documents. These
3d models can be used as geometric models in energy simulation, but must be revised
first in order to avoid errors. For example, some structure components, like beams and
columns must be removed, and small gaps need to be fixed.
According to purpose, accuracy requirements, and other limits, geometric models can
be simplified and many accurate details, which may not influence energy consumption
results much, can be ignored.
37

3.5.2. Thermal Zones
Mechanical engineers divide the entire space into different thermal zones based on
space location, usage, loads, and other factors when doing HVAC system design. Usually
a thermal zone is treated by one HVAC system (while several thermals zones may share
one HVAC system) and has an individual thermostat.
The definition of thermal zones in modeling is similar to that used by designers.
However, done does not need to follow the original classifications made by designers.
On one hand, different thermal zones in design might be combined in the model for
simplification purpose; on the other hand, since a thermal zone in many modeling
software has to have a unique setting, including indoor environment, HVAC system,
occupancy schedule, lighting density and schedule, and equipment operation schedule,
a thermal zone in design might be divided into several smaller thermal zones in the
energy model in order to achieve more accurate results.
There are multiple ways for modelers to decide the appropriate division of thermal
zones in a building. In the most simplified case, the whole building is considered to be
one thermal zone. For multi-floor buildings, there could be one thermal zone for each
floor. In an extremely detailed model, there might be even one thermal zone per room.
In some software, thermal zones need to be defined while creating geometric models.
Though not necessary, it is a good idea to follow designer’s original design when
creating thermal zones in a model.
38

3.5.3. Building Panels
Building panels in an energy model include walls, roofs, floors, and ceilings. Since their
positions and areas have already been determined in geometric models, only thermal
related parameters need to be inputted in this step. Two parameters must be
considered in energy simulation:
1) Heat transfer coefficient, often called U-Value or U-factor. U-value presents heat
transfer rate per unit area. Unite of U-value is BTU/(h °F ft²). Sometimes, another
parameter called R-Value (ft² °F h/BTU), which means thermal resistance, is also
used in software. Relationship between these two parameters is:

2) Heat Capacity (C). This parameter means heat storage capability of components.
A large C number will greatly decrease influence of fluctuation of outdoor
temperature. This parameter can be ignored in light structure buildings, but
must be considered in heavy components like concrete walls.
Characteristics of walls in a building are based on materials and thickness, which can be
found in original drawing. Generally speaking, there are three ways to input the
parameters of walls:
1) Choose the default walls that are closest to the real situation from the software’s
database. Most tools have a set of standard components with different materials
39

and characteristics, which can be used as an appropriate alternative. Figure 3.1
shows some of standard walls stored in EnergyPro’s Library.

Figure 3.1 Standard Walls in EnergyPro’s Library
2) Create a layer of walls in the simulation software. Some software may already
have a large amount of common used building materials in its database.
Modelers only need to input materials and each layer of the wall, and
parameters will automatically calculated in the software.
3) Input U-value and C-value directly. These two parameters can be based either
from calculation results based on handbook or other tools, or measured from
on-site testing.
40

For roofs, floors, and ceilings, procedures of inputting parameters are similar to that of
walls. Sometimes, ceilings even are considered to be a part of roofs and floors.
3.5.4. Windows and Doors
Compared to building panels, more information is available about windows that
influence indoor environment and energy consumption, including window size,
positions, types of glass, frame materials, and shading devices.
1) Size and positions
In most detailed models, widow width, window height, still height, and central
position of each widow will be included. Usually these details will be created while
generating the geometric models.
2) Thermal properties of doors and windows
Two parameters are used for the thermal properties of window: U-Value and SHGC.
U-Value has same definition as mentioned in panels. Usually U-Value of windows is
lower than walls, which means heat loss through windows is more serious than walls.
SHGC, Solar Heat Gain Coefficient, is an index to indicate the percentage of solar
radiation energy admitted through a window. It is usually written as a number
between 0 and 1.
41

These two parameters are influenced by window structure and the materials of the
glass and frames. Like parameters in walls, they can either be set or calculated by
energy tools or be input directly by modelers.
Title 24 has a rough default suggesting U-Value and SHGC for different type of
windows. For more accuracy, modeler may check handbook or product manual, or
do test experiment to measure these two values.
Table 3.2 Default Fenestration U-Values in California Title 24 (CEC, 2008)

42

Table 3.3 Default Fenestration SHGC in California Title 24 (CEC, 2008)

3) Shading
Different shading methods are used to block direct sunlight going through windows
in order to reduce heat gain and indoor glazing. Generally speaking, there are four
types of shading strategies currently used:
a. Exterior horizontal shadings, including overhangs, awnings, and eaves. They are
often used on south façade of a building.
b. Exterior vertical shadings, like fins, are often used on east and west façades.
c. Interior shadings, like blinds, curtains, and other window coverings.
43

d. Sometimes trees near the building are considered shades. This is specifically
true in eQuest.
e. In some software, overhangs and fins are considered to be a part of windows
while there might be several default settings for interior shadings. Usually
shading information needs to be collected through a site interview.
For doors, the procedures are similar. However, since in most cases, total areas of
outdoors are so small comparing with total wall areas or window areas that its influence
of energy usage may often be ignored. As a result, some energy models do not have a
door, or simply chose an existing door in database. Sometimes, doors are treated as
windows in the model for convenience.
3.5.5. Lighting system
The electricity use by lighting is an important part of total energy consumption,
especially in commercial buildings. In the energy model, there are two parameters that
are used for lighting system in calculation: Lighting Power Density (LPD) and Lighting
Schedule.
(1) LPD (Watt/ft
2
). It can either be calculated by dividing the total lighting
instrument capacity (gathered from design document or from onsite
measurements) by area of each thermal zone, or simply follows requirements or
suggestions in related codes or standards. Table 3.4 shows LPD requirements in
Title 24. Most often, lighting designers also follow code requirements, as a result,
44

there might be little difference between two approaches. However, it needs to
be examined carefully in order to avoid significant errors in simulation results.
Table 3.4 Lighting Power Density Values (Watt/FT
2
) in Title 24 (CEC,2008)

Since thermal zones in model are often combined with multiple types of rooms,
average the LPD based on area can be used.

∑

∑

45

(2) Lighting Schedule. In most models, this is an hourly fraction schedule. For a zone,
there might be different operation schedule in weekdays, weekends, holidays, or
different period based on various operation and occupancy behaviors.
For design purposes or new building modeling, code or standard based lighting
schedules can be used. Figure 3.2 is an example of weekday lighting schedule in
hotel/motel suggested by ASHRAE 90.1. Both ASHRAE 90.1 and California Title 24
have given a set of recommended operation schedules (not only for lighting, but
also for occupants and HVAC operations).

Fig. 3.2 Weekday lighting schedule in hotel/motel suggested by ASHRAE 90.1
Energy consumption of lighting system is also a procedure of generating heat by lighting
equipment. It can either increase cooling loads in summer, or decrease heating loads in
winter. For pendant lamps or reading lamps that are completely located in air-
conditioned spaces, all generated heat need to be removed by mechanical systems. For
lamps embedded in ceilings, a part of heat does not need to be considered by HVAC
systems. Ratios need to considered carefully based on assemble styles.
46

3.5.6. Internal Equipment
Electric usage of equipment, such as computers, printers, copying machines, washing
machines, refrigerators, etc., is another essential part of total energy consumption of a
building. Internal equipment loads are often also called plug loads. In most simulation
software, the chief components of mechanical systems, like pumps, fans, are simulated
separately and are not considered to be part of internal equipment loads. However, for
some accessories, like monitors or automatic control machines, if they cannot be
simulated in mechanical system part, they can be contained in equipment loads.
Like lighting systems, equipment load information also contains two parameters:
equipment load density and equipment operation schedule. Operation condition of
different kinds of equipment may totally different from each other, for example,
computer servers are usually kept on 24 hours a day and seven days a week, while
personal computers may only be turned on when employees arrived and off when they
leave for the day. As a result, equipment needs to be examined, classified, and
scheduled carefully when creating models.
There are also suggested numbers for equipment loads density and schedules for
different types of areas in codes and design manuals. However, in real operation
conditions, this data can be quite different from design requirements. Also, occasional
or unexpected events usually have greater influence on equipment loads rather than on
47

lighting loads or occupant schedules. As a result, internal equipment loads might be an
important uncertainty factor that causes simulation result discrepancies.
3.5.7. Occupancy
Occupancy behavior greatly influence different aspects of whole building operations,
including lighting, internal equipment, and HVAC systems. Similar to lighting and
internal equipment, maximum occupant densities and occupancy schedules are two
chief parameters used to describe occupancy behavior.
Occupant densities are usually determined by mechanical system designers according to
code requirements. For example, ASHRAE 62.1 lists a series of default values of
occupant density in different categories like waiting rooms, daycare, classrooms, science
laboratories, computer labs, bars, offices, lobbies, etc. These numbers determine the
maximum capabilities of mechanical systems and can be directly used in energy
simulation models.
There are also suggested default occupant schedules given by codes and standards. For
example, figure 3.3 shows a default school occupant schedule given by ASHRAE 90.1. In
some cases, if information of real occupancy schedules is not collected, these default
settings can be used in models. However, in some cases, the real operation schedules
might completely different from these default settings. For example, some classrooms
in a campus building might have very few classes during weekdays, while be
extraordinarily busy on weekends for student originations. In these conditions, occupant
48

schedules used in models need to be created or amended by the real situation;
otherwise it may cause disparities between simulation results and energy consumption
data.

Fig. 3.3 Weekly School Occupant Schedule suggest by ASHRAE 90.1
3.5.8. HVAC systems
HVAC systems are used to control the indoor environment of a building, and usually
they are chief parts of building energy consumption.
Generally speaking, there are two steps for creating HVAC systems in a model: selecting
correct types of HVAC systems in the model, and setting proper parameters for
components and parameters. Types and details of HVAC systems can be determined
according to collected building information.
49

3.5.8.1 Type of HAVC systems
There are different types of building mechanical systems. A modeler must understand
basic principles of these systems and choose the correct type in simulation software.
In most cases, an HVAC system can be divided into three parts: cooling (air conditioning)
system, heating system, and ventilation (fresh air) system. These three systems usually
share some equipment, such as fans and wind channels, pumps and water piles, etc.
Sometimes, ventilation systems are combined into cooling and heating systems.
There are some common types of cooling systems, including constant air volume (CAV)
systems, variable air volume (VAV) systems, direct expansion (DX) unitary systems,
variable refrigerant volume (VRV) Systems, etc. Many detailed simulation software, like
eQuest or EnergyPro, already have these default types. But in some pre-design tools like
HEED, there are only some limited types that cannot cover exiting cooling systems. As a
result, modeler may have to select alternative systems to replace existing ones.
For space heating, there are three types of heating sources used in most projects: fuel
furnace, direct electrical heating, and heat pump. Generated heat may either be send to
space directly or throw indirect air pipes.
3.5.8.2 Details of HAVC systems
Details of existing HAVC system can be divided into two parts: component capacities,
and environment control settings.
50

After setting type of HVAC systems, parameters of components need to be set in
software, including pumps, fans, chillers, cooling towers, air tubes, pipes, etc. Different
types of systems may have different kinds of components that need to be set.
Parameters of environment control include design temperatures, thermostat set points,
airflow settings, etc. Sometimes operation schedules are also need to be set in software.
These parameters determine all-year-round control policies of HVAC systems.
3.5.9. Weather Data
Details about weather data had been discussed in 3.4.5. In this step, an appropriate
weather data files (a yearly summary or real data) would be used in the simulation
program.
3.5.10. Conclusion
In this step, it was discussed that how to isolate parameters in models from collected
building information. Based on different software, there will be specific method to input
these parameters into the program. Table A in Appendix A contains the inputting
parameters discussed in this step.
3.6. Calibrating Models
A calibrated model has simulation results within a specified tolerance range compare to
existing utility bill or energy consumption data. As a result, in this step, calculated and
51

monitored data will be compared, and the differences between these data will be
analyzed to decide whether simulation results are acceptable.
3.6.1. Collecting and displaying energy data
Before comparing with two series of data, an important thing is to displaying numbers in
order to keep them in same format. In most cases, original energy consumption data
and calculated data work in different units, for example, pervious electricity use might
be 15 minutes of data in kWh, while simulation results are monthly data with units of
kBTU. In these situations, final format used for calibrating must be considered. Both
calculated data and real energy consumption data should be at same time periods
(hourly, daily, monthly or annually) and with same unit (kWh, kBTU, etc).
3.6.2. Error Analyzing
Differences between calculated results and real usage use data need to be analyzed in
order to verify whether the model is calibrated or not. There are several different
indices used to describe differences between two series of data. In calibrating
procedure, two indices named CVRMSE and NMBE are often used.
(1) CVRMSE (Coefficient of Variation of the Root Mean Square Error)

(∑(

̂

)

( ))

̅

52

(2) NMBE (Normalized Mean Bias Error)

∑(

̂

)
( ) ̅

In the two formulas, y
i
are values of real energy usage data (measured data); are values
of simulated results; ̅ is arithmetic mean value of energy usage data; n is total number
of data; p is number of parameter (for calibrating energy models, p=1).
In most cases, monthly data are analyzed in this step; in some detailed cases, hourly
data might need to be checked; while in simplified models, maybe only annually results
are used. According to different codes and standards, a calibrated model may have
different error range requirements. Table 3.5 shows some details in several standards
After deciding which standard to use for verification and calculating error indices, a
conclusion of whether or not the current model is calibrated could be generated.
Table 3.5 Required CVRMS and NMBE in Different Standards
ASHRAE 14 FEMP IPMVP
CVRMSE
Month
±5% ±5% ±10%
NMBE
Month
±15% ±20% ±15%
CVRMSE
Hour
±10% - -
NMBE
Hour
±30% - -

53

3.7. Revising Models
In many cases, there are significant gaps between the simulation results and real data,
especially in the first draft of the model. In this situation, input parameters need to be
checked, and models need to be revised. In some respects, revising a model can be
considered as the most important part of whole procedure of creating a calibrated
model, since modelers may spend most of their time during this step.
To refine a model, all input parameters must be examined carefully. A good idea is to
follow steps of creating and inputting those parameters in 3.5 for double checking.
These parameters might be amended according to three reasons: mistakes, improper
simplification, and inaccurate building information.
3.7.1. Mistakes made by modelers
There is no doubt that human errors occur everywhere including creating a building
energy model. This kind of mistakes often is caused by lack of experience as well as
absentmindedness. Even though they can be considered to be simplest mistakes to
solve, modelers still need to patiently and accurately fix the mistakes.
Here are some common man-made errors.
a. Mistakes made in creating geometric models, such as incorrect input of length
and position of a wall, improper location of different floors, incorrect
orientations, incorrect floor heights, etc.
54

b. Errors in dividing thermal zones.
c. Mistakes in characteristics of walls, roofs, floors, and ceilings.
d. Mistakes in windows and doors, including total areas, positions, glasses and
frames.
e. Errors in lighting densities and schedules.
f. Errors in internal equipment densities and schedules.
g. Mistakes in determine occupancy schedules in different zones.
h. Mistakes in HVAC systems, such as choosing wrong heating and cooling styles,
improper equipment and operation settings, etc.
i. Use of wrong weather files.
3.7.2. Improper simplification
No matter how detailed a model is, there are always simplifications, and these
approximations may cause simulation results different from the real situation. A proper
simplification would greatly reduce the complexity of a model while have little influence
on calculated results. However, usually not all simplifications or approximations are
proper in the model.
The most significant differences between improper simplifications and man-made
mistakes is that man-made mistakes are caused by unconscious conducts of modelers,
hard to find but easy to revise, while simplifications are created by design, sometimes
easy to find but often difficult to determine whether to be improper or not.
55

However, there is no essential difference between these two kinds of mistakes. For
example, a simplification made by an entry level modeler might be considered to be a
mistake by a senior engineer based on his sufficient experience; a negligence in creating
geometric model may accidently become an exquisite approximation. For a modeler,
although focus might be different in finding improper simplification, procedure of
double-checking parameters is similar with in revising man-made mistakes in 3.7.1.
3.7.3. Analyzing uncertain factors of collected building information
In some situations, a modeler have tried his best to finding and revising man-made
mistakes and improper simplifications, however, the simulated results still could not
match real utility data. There might be three reasons: first, more mistakes have not
been found; second, select software does not meet project requirements; third, building
information used for creating parameters is not accurate.
The first reason might be caused by modelers’ capability and experience, and one way
to solve this problem is to find an experienced engineer for double-checking. The
second reason rarely occurs, since simulation tools should already be carefully chosen
before step of creating models, and it means that most of the previous work becomes
useless.
Inaccurate building information is probably most impossible reason for mismatch results.
There are two different conditions:
(1) Missing or inaccurate historical data
56

Historical data includes utility bill or energy consumption data, lighting system
operation conditions, internal equipment conditions, occupancy schedules, and
historical building mechanical system operations. These historical data are quite
important in creating and calibrating energy models. However, it is almost
impossible to gather all these information in a retrofit project. For modelers, lack of
historical data is always an annoying problem. To solve this problem, modelers
usually use ideal parameters in design, suggested data in standards and handbooks,
real data in other similar project, or invented data to instead. Admittedly, these
replacements will cause errors in inputting parameters and simulation results.
(2) Errors of products and components
Many parameters input in models are based on design documents and products
manuals, for instance, U value of walls and roofs, capability of lighting equipment,
and mechanical system components like chillers or pumps, etc. However, differences
occur between ideal conditions and real products. U-values of exterior walls might
be 10% smaller or 30% larger than design value; electricity consumption of lamps
may be 20% larger than that written in manuals after using for six years; efficient
curves of chillers or system curve of pumps may have changed a lot year by year.
These kinds of errors would also influence energy consumption results.
For entry-level modelers, it might be too complex to analyzing all these factors; however,
this could not be ignored on the process of becoming a senior energy modeler.
57

3.7.4. Operation Schedules
Operation schedules are a part of missing or inaccurate historical data. However, in
most cases, it can be considered most uncertain factors that influence final calculated
results.
There are different kinds of operation schedules, including lighting schedules, internal
equipment schedules, HVAC system operation schedules, and occupancy schedules.
Definitions of these schedules have been discussed in 3.5. Usually occupancy schedules
are core factors while all other schedules might have relationship with it.
It is almost impossible to gather historical operation schedules in a whole year directly.
As a result, modelers have to find some indirect way to determine input schedules in the
model. For example, operation and occupancy conditions of an office building during
weekdays and weekends might be investigated; historical course calendar of a campus
building might be checked; survey information of similar buildings might be used in an
energy model. As a result, operation schedules input in a model might be quite different
from real schedules.
Electricity usage of a building has a sensitive relationship with operation schedules. For
example, usually 30% to 50% of electricity usage of office buildings is for lighting, and
electricity usage of lighting system is approximately direct proportional to lighting
schedules. So if lighting schedule has a 30% increasing, total electrical consumption
58

might increase 10% to 15%, which is a significant difference. Sometimes differences
between operation schedules might even larger.
3.8. Conclusion
In summary, the whole process of calibrated simulation can be divided into six steps: 1.
deciding if the project qualifies for simulation, 2. producing simulation plans, 3.
collecting building information, 4. creating models, 5. comparing simulation results with
utility data, 6. refining models. Step 1 - 3 can be considered as pre-modeling part; step 4
can be considered as modeling part; and steps 5 and 6 can be considered as calibrating
part. Steps of how to using simulation results for analyzing renovation effects
mentioned in 2.1.7 and 2.1.8 were not discussed in this chapter, since they were not
considered to be a part of the whole procedure of how to create a calibrated model.

59

CHAPTER 4: CASE STUDY

4.1. Introduction
In this chapter, a university building, KAP on the USC campus, is chosen to be the
calibrated simulation case study. The simulation process follows the steps outlined in
Chapter 3.

Fig. 4.1 Kaprielian Hall at USC Campus
4.2. Deciding whether or not to create a calibrated simulation
KAP is a multi-use campus building with offices and classrooms. The building is not
overly complex and does not have any unusual requirements. As a result, a calibrated
simulation method can be used for this building. There is no specific retrofit plan (it has
been recently retrofitted) or any budget requirements in this case.
60

In this case, it is not a real renovation project and only building conditions are
considered; the choice made in this step is atypical and does not have much reference
value. In a real construction or retrofit project, retrofit goals and methods, and budget
requirements have to be considered carefully at the same time.
4.3. Producing a Simulation Plan
In this case, the purpose of the calibrated simulation is for research. As a result, detailed
analyses are required, and the calibrated simulation method is used.
eQuest is chosen to be the chief simulation software in this project for three reasons:
eQuest can be used to simulate hourly energy consumption of a building; it is popular in
United States, especially in California, which means it is easier to get help from local
experienced engineers when errors or complicate problem occurs; and it is free. As
there are no collaborators on the project (for example energy consultants), it is not
necessary to consider the software that they are using.
4.4. Collecting Building Information
The classification of collected building information of KAP follows the outline described
in 3.4. Most of the design and monitoring data, including original drawings and energy
data comes from the University of Southern California Facilities Management Services
(FMS).
61

4.4.1. Original Drawings and Digital Record Drawings
More than 100 original drawings were collected from FMS. Based on the original
classification on cover sheet, these drawings are divided into three types: architectural
drawings, electrical drawings, and structural drawing. Details of these drawings are
shown in table 4.1.
Table 4.1 List of Original Drawings of KAP
Type Drawings
Architectural
Floor Plans, Roof Plans, Ceiling Plans, Building Sections, Elevations,
Building Materials, Window Schedules, Door Schedules, Connection
details, Stairs, Toilet Rooms, etc.
Electrical
Lighting Plans, Power Plans, Mechanical Equipment Schedules, Fire
Protection Schedules, Fixture Schedules, etc.
Structural
Floor Plans, Roof Plans, Sections, Elevations, Shoring Plans and Details,
Structure Details, Connection Details, etc.

Architectural drawings and electrical drawings are the most valuable to energy
simulation.
Most of the drawings were hand drawn in late 1980s, thus they are harder to work with
compared to recent CAD plots. A lot of effort was needed to determine some specific
details like some geometric size data.
62

4.4.2. Site survey and operation interview
In this step, engineers from FMS who are familiar with historical operating conditions of
KAP were interviewed and important building information about the mechanical system
was collected. In 2007, the mechanical system of KAP had been rebuilt. As a result, the
collected mechanical drawings of KAP are invalid. Subsequent diagrams of HVAC
systems, operation schedules, thermal zone divide, and indoor thermal environment
were collected. And a Ph.D. student who had studied in KAP during 2010 was
interviewed.
Site visits of KAP have also been done in order to understand operating conditions of
the building, while some details of building envelope, lighting system, equipment, and
occupant conditions have been double-checked.
4.4.3. Energy Consumption Data
Historical energy consumption data in 2010 was collected from FMS. It contains
historical electricity usage data and chilled water consumption data.
Fig 4.2 shows total electric power of the whole building, which was measured every 15
minutes in kW. Electricity usage data contains energy consumption data of many
aspects of the building, including lighting systems, internal loads, pumps and fans, etc.
Space cooling, heating, and domestic hot water are not included.
63

Fig. 4.2 2010 15-min Electric Usage Data of KAP
KAP is a part of the USC campus chilled water loop, and cooling capacity of the building
comes from supplied chilled water. Hourly chilled water usage data has been recorded
by FMS. In the original data, there was inlet water temperature (t
in
, ℉), outlet water
temperature (t
out
, ℉), and chilled water flow rate (F, gpm). Cooling capacity (energy
consumption for cooling, BTU/h) can be calculated based on following formula:

(

)

64

Fig. 4.3 2010 Hourly Chilled Usage Data(original)

Fig. 4.4 2010 Hourly Chilled Usage Data(Revised)
Figure 4.3 shows hourly chilled water usage data. From the chart, it can be seen that
there is no chilled water data for some days in 2010, including all of November and most
65

days in October and December. Based on the operation interview in 4.4.2, it is known
that no special events happened these days, and the mechanical system of KAP worked
properly. The conclusion is that the monitoring system data is incorrect. Based on
existing data and weather conditions, these gap numbers have been filled in, and a
revised original hourly chilled water usage chart created (Figure 4.4). The revised data
are used for the calibrated simulation.
4.4.4. Special Equipment Measurement
Special equipment measurement can be considered as advanced step in calibrated
simulation, and it usually cost more money and takes time. In this case, no special
measurement was done.
4.4.5. Weather Data
The building is located not far from Los Angeles Downtown area, which is in California
Climate Zone 6. As a result, weather data of California Climate Zone 6 can be used as
general data in this case. The general weather file for Los Angeles can also be found in
weather file database of eQuest.
A special weather file of the Los Angeles Downtown area specifically for 2010 was also
used. These data were collected from a company called “Weather Analytics” and
recorded in a weather station (+34° 3' 5.19", -118° 14' 5.92") located in Downtown Los
Angeles area. (Singh, 2012)
66

4.4.6. Codes and Standards
A series of codes and standards are used for reference, including ANSI/ASHRAE 90.1 -
2010, California Title 24-2008, ASHRAE 14-2002, FEMP M&V Guidelines-2008, and
IPMVP-2012. These documents are used either for guiding general steps for simulating
and calibrating or for analyzing details of the model and obtaining default setting values.
There were no specific requirements for this building that would change the energy
calibration study.
4.5. Creating Virtual Models
An eQuest model of KAP is created according to procedure discussed in 3.5.
4.5.1. Geometric Models
There’s no existing 3d building models for KAP; as a result, a 3d geometric model had to
be created based on current 2d drawings. The floor plans of KAP were the primary
reference materials for creating the 3d-model (Figure 4.5-4.9, drawn by FMS in 2011)

Fig. 4.5 KAP Floor Plans (Basement) Fig. 4.6 KAP Floor Plans (1st Floor)
67

Fig. 4.7 KAP Floor Plans (2nd Floor) Fig. 4.8 KAP Floor Plans (3rd Floor)

Fig. 4.9 KAP Floor Plans (4th Floor)
In eQuest, there are two methods for creating geometric models: one is creating a
model in another 3d design software like Revit; another is doing it directly in the eQuest
Wizard. In this case, the second method for creating the geometric model was used,
since KAP is not too complicated or unique, and it is not too difficult to use the Wizard in
create 3d model in eQuest. Figure 4.10 shows the interface for creating geometric
models in eQuest Wizard, and Figure 4.11 is the final 3d geometric model of KAP.
68

Fig. 4.10 eQuest Wizard Interfaces for Creating Geometric Models

Fig. 4.11 3d Model of KAP in eQuest

69

4.5.2. Thermal Zones
The model is divided into seven thermal zones. Details are shown in Table 4.2.
Table 4.2 Thermal Zones in KAP
Name Location Usage
Area
(sqft)
HVAC
System
Zone 1
Central
Basement
Science lab (100%) 4,284 CT-0
Zone2
Basement
(exclude
Zone 1)
Science labs (21.8%), Offices
(33.3%), Mechanical space (17.0%),
Computer labs (0.8%), Storage
space (8.0%), Restrooms (1.4%),
Corridors (17.6%)
22,370 AHU-1
Zone 3 First Floor
Classrooms (51.3%), Offices (26.6%),
Computer Labs (3.3%), Restrooms
(1.7%), Corridors (17.1%)
26,650 AHU-1
Zone 4 Second Floor
Offices (69.1%), Classrooms (7.0%),
Computer Labs (5.2%), Restrooms
(1.8%), Corridors (16.4%)
26,260 AHU-2
Zone 5 Third Floor
Offices (74.8%), Classrooms (7.0%),
Restrooms (1.8%), Corridors (16.4%)
25,670 AHU-2
Zone 6
Northeast of
Forth Floor
Offices (70.5%), Conference rooms
(18.0%), Corridors (10.8%)
12,150 AHU-1
Zone 7
Southwest of
Forth Floor
Offices (76.2%), Conference rooms
(8.9%), Restrooms (5.4%), Corridors
(9.5%)
12,150 AHU-2

70

4.5.3. Building Panels
Details of the building components, including walls, roofs, ceilings, and floors, can be
found in the original drawings (see Figure 4.12). Since they are relatively common, the
characteristics can either be calculated by simulation software or by modelers. In this
case, layers of building materials are directly input in eQuest.

Fig. 4.12 Section Details of KAP
(Drawn by Abbott Marshall Architecture Planning in 1990)

71

Table 4.3 Panels of KAP
Panels Zones Layers
U-Values
(BTU/(h °F ft²))
Underground
Exterior Walls
Zone 2 12” CMU 0.363
Exterior Walls
Aboveground
Zone 3, Zone 4,
Zone 5, Zone 6,
Zone 7
4” Bricks, 3.5” R11
Insulation, 5/8” Gypsum
Board
0.080
Underground Slabs Zone 1, Zone 2
12” Concrete, 3/4"
Plywood, 3/4" Cement
0.010
Ground Floors
Zone 3, Zone 4,
Zone 5, Zone 6,
Zone 7
5.5” R19 Insulation,
3/4" Plywood, 3/4"
Cement
0.053
Ceilings
Zone 1, Zone 2,
Zone 3, Zone 4,
Zone 5, Zone 6,
Zone 7
1/2” Gypsum Board 0.885
Roofs Zone 6, Zone 7
3/8” Built-up Roof, 3/4"
Plywood, 5.5” R19
Insulation
0.053

4.5.4. Windows and Doors
Information of windows’ details is contained in the original drawings. Figure 4.13 shows
the window schedule that includes all types of windows in the building. On exterior
walls, most windows are 5”*6.6” single light tint windows with aluminum frames (Figure
72

4.14). For simplification, all exterior windows in the model are assumed to be this type,
and characteristics of windows in the model are shown in Figure 4.15.

Fig. 4.13 Window Schedules in Original Drawings

Fig. 4.14 Details of 5’–0” x 6’–6” Windows in Original Drawings
73

Fig. 4.15 Characteristics of Windows in the Model
There are two entrance doors in KAP. Since the frame and type of glass of both windows
and doors are the same, no special doors are created in the model. Total areas of
exterior windows and doors of KAP equate to those areas the model.
4.5.5. Lighting system
Information about the lighting system can be found in lighting plans and lighting
schedules in the original drawings. Since there are no system retrofits of lighting
systems in KAP after it was built, the drawings of the lighting systems are valid.
Table 4.4 shows suggested LPD and designed LPD of different areas exist in KAP.
Suggested values are from ASHARE 60.1-2010. Table 4.5 shows calculated LPD in
different thermal zones.
74

Table 4.4 Lighting Power Density (LPD) in Different Areas of KAP
Area Suggested LPD (W/ft
2
) Designed LPD (W/ft
2
)
Science lab 3.19 0.95
Offices 1.49 1.37
Mechanicals 0.81 0.68
Computer Rooms 1.45 1.42
Classrooms 1.45 1.44
Storages 1.19 0.88
Corridors 0.57 0.42
Conference rooms 0.92 1.13
Restrooms 0.77 0.68

In the first draft of the model, the school lighting schedules suggested in ASHRAE 90.1
are used (Figure 4.16).

Fig. 4.16 ASHRAE 90.1 School Lighting Schedules
75

Table 4.5 Lighting Power Density (LPD) in Different Thermal Zones of KAP
Name Usage
Designed LPD
(W/ft
2
)
Zone 1 Science lab (100%) 0.95
Zone2
Science labs (21.8%), Offices (33.3%), Mechanical (17.0%),
Computer labs (0.8%), Storages (8.0%), Restrooms (1.4%),
Corridors (17.6%)
0.94
Zone 3
Classrooms (51.3%), Offices (26.6%), Computer Labs (3.3%),
Restrooms (1.7%), Corridors (17.1%)
1.22
Zone 4
Offices (69.6%), Classrooms (7.0%), Computer Labs (5.2%),
Restrooms (1.8%), Corridors (16.4%)
1.20
Zone 5
Offices (74.8%), Classrooms (7.0%), Restrooms (1.8%),
Corridors (16.4%)
1.21
Zone 6 Offices (70.5%), Conference rooms (18.0%), Corridors (10.8%) 1.22
Zone 7
Offices (76.2%), Conference rooms (8.9%), Restrooms (5.4%),
Corridors (9.5%)
1.22

4.5.6. Internal Equipment
Compared with lighting, there are more uncertainties in internal loads. In the design
process, the electrical designer can only determine maximum capability for plug loads,
and it is usually much higher than internal peak loads, let alone continuous loads.
76

In the energy model of KAP, the recommended values of internal load density from
ASHRAE 90.1 are used. Table 4.6 shows these values in different areas in KAP, and Table
4.7 shows calculated internal loads in each thermal zone.
Table 4.6 Internal Load Density in Different Areas of KAP
Area Recommended Internal Load Density(W/ft
2
)
Science lab 1.00
Offices 0.75
Mechanicals 0.10
Computer Rooms 5.00
Classrooms 0.50
Storages 0.00
Corridors 0.00
Conference rooms 0.10
Restrooms 0.10

77

Fig. 4.17 ASHRAE 90.1 School Equipment Schedules
Similarly with lighting schedules, school equipment schedules recommended in ASHRAE
90.1 are used in first draft of the model.

78

Table 4.7 Lighting Power Density in Different Thermal Zones of KAP
Name Usage Internal Load Density (W/ft
2
)
Zone 1 Science lab (100%) 1.00
Zone2
Science labs (21.8%), Offices (33.3%),
Mechanical (17.0%), Computer labs (0.8%),
Storages (8.0%), Restrooms (1.4%), Corridors
(17.6%)
0.52
Zone 3
Classrooms (51.3%), Offices (26.6%), Computer
Labs (3.3%), Restrooms (1.7%), Corridors
(17.1%)
0.62
Zone 4
Offices (69.6%), Classrooms (7.0%), Computer
Labs (5.2%), Restrooms (1.8%), Corridors
(16.4%)
0.81
Zone 5
Offices (74.8%), Classrooms (7.0%), Restrooms
(1.8%), Corridors (16.4%)
0.60
Zone 6
Offices (70.5%), Conference rooms (18.0%),
Corridors (10.8%)
0.55
Zone 7
Offices (76.2%), Conference rooms (8.9%),,
Restrooms (5.4%), Corridors (9.5%)
0.58

4.5.7. Occupancy
The procedure to determine occupancy density and schedules are similar to that of
figuring out the internal loads: first, determine values for each type of area; then
calculate values for each zone based on area distribution.
79

Recommended values of occupancy density from ASHRAE 60.1 are given in Table 4.8,
and occupancy densities in each thermal zone are shown in Table 4.9. School occupancy
schedules recommended in ASHRAE 90.1 are used in first draft of the model.
Table 4.8 Occupancy Density in Different Areas of KAP
Area Occupancy Density (/1000 ft
2
) Occupancy Density (ft
2
)
Science lab 25 40
Offices 5 200
Mechanicals 0 -
Computer Rooms 25 40
Classrooms 65 15
Storages 2 500
Corridors 0 -
Conference rooms 50 20
Restrooms 0 -

80

Table 4.9 Occupancy Density in Different Thermal Zones of KAP
Name Occupancy Density (/1000 ft
2
) Occupancy Density (ft
2
)
Zone 1 25 40
Zone2 8 134
Zone 3 36 28
Zone 4 9 107
Zone 5 8 120
Zone 6 13 80
Zone 7 8 120

Fig. 4.18 ASHRAE 90.1 School Occupancy Schedules
81

4.5.8. HVAC systems
Although there are mechanical drawings in the original drawing set, these drawing are
invalid because the mechanical systems of KAP were rebuilt in 2007. Detailed
information of KAP was collected from FMS.
In KAP, VAV (variable air volume) systems are used for space cooling and heating. Since
KAP is a part of the USC chilled water loop, there is no chiller or cooling tower in the
building. Fig 4.19 and Fig 4.20 are diagrams of air side and water side of HAVC systems
in KAP.

Fig. 4.19 Diagram of VAV system in KAP (From FMS Monitoring Systems)
82

Fig. 4.20 Diagram of Water System in KAP (From FMS Monitoring Systems)

Fig. 4.21 Diagrams of HVAC systems in the Model (Water and Air system in eQuest)
83

Fig. 4. 22 Mechanical System settings in eQuest.
84

However, in eQuest (as well as in most simulation software), a mechanical system like
this is considered to be “incomplete” and cannot be simulated. There must be a chiller
and a cooling water system in the model. As a result, virtual components are created in
models. Fig 4.21 are diagrams of HVAC systems in the eQuest model, and Fig 4.22
shows the input parameters of mechanical systems in the model.
4.5.9. Weather Data
In the first draft of the model, general weather file of Los Angeles in eQuest’s database
was used.
4.6. Calibrating Models
After creating the virtual model of KAP in eQuest, simulated energy consumption results
were calculated. In this step, collected historical energy consumption data in 2010 are
used for calibrating simulation results.
4.6.1. Collecting and displaying energy data
Detailed simulation results of eQuest are hourly data. According to ASHRAE 14, either
hourly data or monthly data can be used to calibrating. Hourly data are more accurate
compared to monthly data, but they need more detailed building information and
experienced modelers. In this case, monthly data are used for calibrating. As a result,
both simulated results and historical data must be transferred to hourly format. Figure
4.23 is monthly electricity usage data in simulation results, while Figure 4.24 is monthly
gas usage date.
85

Fig. 4.23 Monthly Electricity Consumption Results (First Draft)

Fig. 4.24 Monthly Gas Consumption Results (First Draft)
86

It can be seen that an important part of calculated electricity usage is space cooling,
which is energy consumed by the virtual chiller defined in the model. However, in the
real building, energy inlet for space cooling was chilled water rather than electricity and
chilled water usage was monitored separately. As a result, energy for space cooling
must be treated independently and a transfer is needed from chilled water consumption
to end-use electricity.
In a real situation, there are efficiency curves for chillers, and it is a bit complicated to
transfer input and output energy for real products. In this case, a simplification is that
the efficiency of the chiller in the model is assumed to be constant and equal to the
nominal COP. As a result, following formula is used.

P
c
is electricity consumption (kWh), E
w
is energy consumption in chilled water(kBTU),
and C is cooling load efficient of chiller (kW/ton). In this case, C equals to 0.837. Figure
4.25 compares the simulation results for cooling and real chilled usage data.
87

Fig 4.25 Simulated results and real data for cooling (First Draft)
In figure 4.26 total electricity usage data between simulation and actual data are
compared. In this chart, real energy data are summations of electricity usage data and
converted chilled water data. Note that the simulation is generally higher than the
actual values, but the shape of the curves are similar.
Since there is no actual gas consumption data, simulated results for heating and
domestic hot water cannot be calibrated in this case.

88

Fig. 4.26 Electricity Usage Result(First Draft)
4.6.2. Error Analyzing
Looking at Figure 4.26 it is apparent that there are large gaps between the simulated
results and real energy usage data. To judge the severity of the differences, one refers
to CVRMSE and NMBE for what the allowable tolerances are. The draft model is not
calibrated by these standards (Table 4.10).
Table 4.10 Error Analysis of Simulated Results (First Draft)
Chiller Water Total Electric ASHRAE 14 FEMP IPMVP
CVRMSE
Month
23% 128% ±5% ±5% ±10%
NMBE
Month
9% -132% ±15% ±20% ±15%

89

4.7. Revising Models
Since simulation results are quite different from real energy usage data, the model
needs to be revised. In a real situation, there is a circle of revising – simulation – revising
until the model is calibrated. Revising details are summarized here, and intermediate
simulating and calibrating results are shown in Appendix B.
4.7.1. Mistakes made by modelers
After carefully researching the model, several mistakes were found and the following
changes are made:
(1) Change glass type from single clear 1/8” to single green 1/8”,
(2) Change Light to Space Ratio from 1 to 0.8,
(3) Affiliate Zone 7 to AHU-2 instead to AHU-1.
These mistakes do not influence simulation significantly.
4.7.2. Improper simplification
In this part, the special weather file of 2010 from the weather station in downtown Los
Angeles is used instead of a general weather file.
4.7.3. Operation Schedules
According to Figure 4.26, it can be seen that in simulation results, either the energy
consumption for lighting or for equipment is too large compared with the total
90

electricity usage. Based on the assumption that there is no significant error in power
density, the only possible conclusion is that the input schedules’ values are too large
compared with the real situation.
Based on site interviews, and simulation results, operation schedules, including lighting
schedules, equipment schedules, and occupancy schedules were revised. In the new
version, each zone has separated schedules since area distributions are different.
4.7.4. Final Simulation Results
Figure 4.27 shows final simulation results and table 4.11 concludes that the model can
be considered as “calibrated” according to IPMVP, or “almost calibrated” based on
ASHRAE 14.

Fig 4.27 Simulated results and real data for cooling (Final)
91

Fig. 4.28 Electricity Usage Result (Final)
Table 4.11 Error Analysis of Simulated Results (Final)
Chiller Water Total Electric ASHRAE 14 FEMP IPMVP
CVRMSE
Month
19.8% 11.3% ±15% ±20% ±15%
NMBE
Month
-3.2% -6.7% ±5% ±5% ±10%

4.8. Conclusion
In this chapter, an energy model of KAP was created and calibrated following the steps
outline in Chapter 3. Whereas Chapter 3 can be used as a methodology for building
calibration, Chapter 4 has shown how to apply the steps and principles toward a specific
example.
92

CHAPTER 5: PARAMETERS INFLUENCED
SIMULATION RESULTS

5.1. Introduction
While revising models, it can be seen that some parameters had greatly influenced
simulation results while others have less influence. Analyzing sensitive parameters can
help modelers creating and calibrating virtual models more easily and provide further
supports for engineers and designer to create efficient building retrofit plans.
In this chapter, the relationship between several parameters (window areas,
characteristics of panels, weather data, and lighting power density) and simulation
results were analyzed based on current calibrated model of KAP. Simulation results of
final draft model of KAP were used as the baselines.
5.2. Area of Windows
The area of windows is often considered to be an important element that influences
building cooling/heating loads. In many related standards, such as ASHRAE 90.1 and
California Title 24, Window to Wall Ratio (WWR) is discussed as an important design
parameter. In the process of creating simulation models, the total areas of windows
might be vary from the real situation because of inaccurate measurements, incorrect
calculations, or improper simplification, and this may influence final calculation results.
93

For this sensitivity analysis, the total area of windows was changed by adding and
reducing quantities of windows in the eQuest model, while other parameters including
the characteristics of the windows remained constant.
Figure 5.1 shows how electricity consumptions changed while changing total areas of
windows of KAP, and Table 5.1 shows details of these differences conditions. Details of
simulation results are in Appendix C.
According to the calculation results, the area of window does not influence total energy
consumption much in KAP. It is because that energy loss of windows in the original
building was not too large; as a result, reducing window area did not greatly influence
total energy consumption.

94

Table 5.1 Electricity consumption differences under multiple window areas
Window Area (sqft) 3482 4862 5583 6274
6964
(original)
7684
Window Area Compared To Original 50% 70% 80% 90% 100% 110%

Fig 5.1 Electricity consumption curves under multiple window areas
5.3. Panels
Building panels can be divided into two types according to their positions: exterior
panels (including exterior walls and roofs), and interior panels (including interior walls,
ceilings, and floors). When there is no significant difference between interior thermal
environments in different thermal zones, influences of interior panels is limited. In many
simulation models, interior walls are often ignored.
95

As a result, in this case only exterior panels were analyzed. Compared with roofs,
exterior walls of multi-story building like KAP have much larger areas, and that means
they have greater influence on simulation results. In this part, only U-values of exterior
walls were changed.
Figure 5.2 shows calculated electricity consumptions with different U-values of exterior
walls of KAP. Details of simulation results can be seen in Appendix C.
Based on calculated results, it can be seen that variety of exterior U-values rarely
influence simulating results, which means panels are not sensitive factors for building
energy consumption of KAP.

Fig 5.2 Electricity consumption curves under different U-values of exterior walls
96

5.4. Weather Data
In most cases, there are multiple choices weather data for simulation models. Details
had been discussed in 3.4.5. In this part, following three different weather files were
used for analyzing influence of calculated results in the eQuest model of KAP.
(1) 2010 special weather data used in revised model. Details of these data had been
discussed in 4.4.5.
(2) General weather data of Los Angeles. The file was downloaded from Energy Plus
weather file database.
(http://apps1.eere.energy.gov/buildings/energyplus/cfm/weather_data.cfm)
(3) General weather data of California Climate Zone 8. It was the default weather file of
eQuest.
Figure 5.3 showed the calculated results with different weather files. It can be seen that
more than 10% differences, which could not be ignored, occurred between final results
by using special and general weather files in this case.

97

Fig 5.3 Electricity consumption curves under different weather files
5.5. Lighting Power Densities
According to the simulation result calculated in 4.7.4, it can be seen that electricity
usage of lighting was an important proportion of total energy usage. Light power
densities (LPD) were analyzed in this part.
Different zones of KAP had their dependent LPD numbers. All these LPD Numbers were
changed at same radio every time. Figure 5.4 shows curves of the simulation results,
and details of simulation results are in Appendix C. It can be seen that changing LPD
numbers would greatly influenced simulation results.

98

Fig 5.4 Electricity consumption curves under different LPD numbers
5.6. Conclusion
In this chapter, four parameters, including area of windows, thermal characteristics of
panels, weather data and lighting power densities (LPD) were subjects to a sensitivity
analysis. The area of windows and thermal characteristics of panels were considered to
be insensitive parameters that did not influence simulated energy usage results much,
while weather data and LPD were considered to be sensitive that need to be carefully
determined during modeling and calibrating process.
All analyses were based on the existing eQuest model created in Chapter 4. As a result,
the previous sensitivity conclusions were only used for the case of KAP.
99

CHAPTER 6: CONCLUSIONS AND FUTURE WORK

6.1. Conclusions
This thesis created simple guidelines for simulating and calibrating energy models for
existing buildings and then used this guideline in simulating and calibrating building
energy consumption of a campus building. During the modeling and calibrating
procedures of the case study, details of the guidelines had been added and revised.
The whole process of calibrated simulation can be divided into six steps in three parts:
(1) pre-modeling, including steps of deciding if the project qualifies for simulation,
producing simulation plans, and collecting building information; (2) modeling, including
step of creating models; (3) calibrating, including steps of simulating comparing
simulation results with utility data, and refining models. Steps of how to using
simulation results for analyzing renovation effects were not discussed. Students and
engineers can create and calibrate simulation models by following these steps.
The step of collecting building information was one of the most important parts in
whole process. There were multiple sources of building information that needed to be
found by the modelers. Usually it was considered that with more details of building
information collected the simulation results would be more accurate compared with
real energy usage data.
100

However, it is neither possible nor necessary for modelers to find all details of building
information needed in modeling. As a result, some assumptions had to be made when
creating virtual models. On one hand, these assumptions were helpful and necessary for
simplifying the building energy models and significantly reduced the workloads for
modelers; on the other hand, they were often considered to be for errors in the
calculated results.
Two mathematical indices standards, CVRMS and NMBE, were used to compare
differences between the simulation results and real energy consumption data. In many
cases, including the case study of KAP in this thesis, monthly data were used for
analyzing. Hourly data would have been better if it were available. In a real project,
proper codes or standards need to be chosen by modelers to judge whether differences
between simulation results and real data meet and whether models can be considered
to be calibrated.
In most cases, the models had to be revised several times during the calibrating process.
Either errors or simplification of input parameters must be examined carefully.
Sometimes, analyzing the contributions of variables is helpful or performing sensitivity
analysis to check to see if effort is warranted in getting more exact values to input into
the simulation.
For the KAP case study, sensitivities of four parameters, including window areas,
characteristics of panels, weather data, and lighting power density (LPD), were analyzed
101

based on the eQuest model. Window area and panels had little influence on final
simulation results; that means models did not need to pay much attention on them
either in revising models or creating building retrofit plans. Other two parameter,
weather data and LPD, had more influence on simulation result, and could not be
ignored in calibrated simulating process. These summaries were only for KAP building in
this case, and could not be extend to be general conclusions for other calibrated
simulation projects.
6.2. Future work
Guidelines created in this thesis are straightforward, but inexperienced models or
students may still find it hard to create calibrated energy models. Part of this is due to
the fact that getting the information necessary is often difficult. Also, more details
could to be added to improve these guidelines. Other possible related research is also
discussed below.
1. Study more calibrated simulation cases
Detailed guidelines should be supported by sufficient cases. These cases might have
different types of building in different thermal areas, different styles of mechanical
systems, light systems, and operation schedules. Other case studies could provide
additional insight into issues to consider when calibrating a building model.

102

2. Study more about operation schedules
Operation schedules for existing building can be considered to be one of the most
difficult parts to simulate. It is worth and helpful to find better procedures to create
more accurate schedules in models.
3. Uncertainty factors analysis and error analysis
Sometimes it is hard to get “calibrated” models by the end. In these conditions,
uncertainty factors analysis and error analysis might be helpful for modelers to know
the reason and for future improvements.
6.3. Summary
Buildings use energy. It can be quantified. Renovations may decrease or even increase
the amount of energy that buildings use, and those results quantified. However, it
makes sense to simulate the building’s performance first to determine energy saving
strategies, quantify the predicted savings, and decide financial trade-offs. Having a
calibrated building model is the first step towards being able to predict future
performance. These guidelines will help in developing this.

103

BIBLIOGRAPHY

Avinash, Y., Geetha, T., & Sivasubramaniam, A. (2010). “Activity based energy model for
calibrated simulation approach in retrofitting a multi use facility.” Paper presented at
the 2010 9th International Power & Energy Conference (IPEC 2010), 951-6. Retrieved
from http://dx.doi.org/10.1109/IPECON.2010.5696986
EIA. (2011). “Annual Energy Outlook 2011 With Projections to 2035.” DOE/EIA-
0383(2011)
EVO. (2012). International Performance Measurement and Verification Protocol:
Concepts and Options for Determining Energy and Water Savings.
FEMP. (2008). M&V Guidelines: Measurement and Verification for Federal Energy
Project, Version 3.0
Fernando Simon Westphal, & Roberto Lamberts. (2005). “Building simulation calibration
using sensitivity analysis.” Paper presented at the Building Simulation 2005, Canada.
Hubler, D., et al. (2010). “Pulling the Levers on Existing Buildings: A Simple Method for
Calibrating Hourly Energy Models.” ASHRAE Transactions. Volume 116, Part 2.
Itorn, Inc. (2006). “California Commercial End-Use Survey.” CE-400-2006-005
Libowitz, G. G., Feldman Jr., K. T., & Stein, C. (1997). “Thermodynamic properties of
metal hydrides for a novel heat pump configuration.” Journal of Alloys and Compounds,
253-254(1-2), 673-676. Retrieved from http://dx.doi.org/10.1016/S0925-
8388(96)02948-9
Milne, M. (2008) “History of simulation software.”
Mui, K. W., & Chan, W. T. (2006). “Building calibration for IAQ management.” Building
and Environment, 41(7), 877-886. Retrieved from
http://dx.doi.org/10.1016/j.buildenv.2005.04.019
Ohashi, Y., Genchi, Y., Kondo, H., Kikegawa, Y., Yoshikado, H., & Hirano, Y. (2007).
“Influence of air-conditioning waste heat on air temperature in Tokyo during summer:
Numerical experiments using an urban canopy model coupled with a building energy
104

model.” Journal of Applied Meteorology and Climatology, 46(1), 66-81. Retrieved from
http://dx.doi.org/10.1175/JAM2441.1
Pan, Y., Huang, Z., & Wu, G. (2007). “Calibrated building energy simulation and its
application in a high-rise commercial building in Shanghai.” Energy and Buildings, 39(6),
651-7. Retrieved from http://dx.doi.org/10.1016/j.enbuild.2006.09.013
Pedrini, A., Westphal, F. S., & Lamberts, R. (2002). “A methodology for building energy
modelling and calibration in warm climates.” Building and Environment, 37(8-9), 903-
912. Retrieved from http://dx.doi.org/10.1016/S0360-1323(02)00051-3
Qing, L., Ming, L., Chaofeng, X., & Xu, J. (2010). “Design and heating performance of
solar-assisted heat pump water heater with truncated-cone-type evaporator.” Paper
presented at the 2010 Asia-Pacific Power and Energy Engineering Conference (APPEEC
2010), 4. Retrieved from http://dx.doi.org/10.1109/APPEEC.2010.5448577
Raftery, P., Keane, M., & O'Donnell, J. (2011). “Calibrating whole building energy models:
An evidence-based methodology.” Energy and Buildings, 43(9), 2356-2364. Retrieved
from http://dx.doi.org/10.1016/j.enbuild.2011.05.020
Yoon, J., Lee, E. J., & Claridge, D. E. (2003). “Calibration procedure for energy
performance simulation of a commercial building.” Transactions of the ASME.Journal of
Solar Energy Engineering, 125(3), 251-7. Retrieved from
http://dx.doi.org/10.1115/1.1564076

105

APPENDIX A: PARAMETERS IN MODELS

Table A Parameters in Models
Type Parameters Source
Geometric
Envelopes, Floor Heights, Interior
Areas, Building Orientation, etc.
Original drawings
Thermal
Zones
Distributions, Areas, Usage, Related
HVAC systems
Original drawings and documents,
operation interviews
Panels
Materials and thermal parameters(U-
value, heat capacity) of walls, roofs,
ceilings and floors
Original drawings, onsite interviews and
tests
Windows
and Doors
Sizes, positions, total numbers, types
and materials, thermal parameters (U-
value, SHGC)
Original drawings and documents,
onsite interviews and tests
Lighting
Systems
LPD, lighting schedules
Original drawings and documents,
operation interviews, onsite interviews
and tests, related codes and standards
Internal
Equipment
Equipment loads densities and
schedules
Operation interviews, onsite interviews
and tests, related codes and standards
Occupancy Occupancy schedules
Operation interviews, onsite interviews,
related codes and documents
HVAC
Systems
System types, components,
environment settings
Original drawings and documents,
operation interviews and tests

106

APPENDIX B: INTERMEDIATE SIMULATION
RESULTS FOR REVISING MODEL

B.1 Round 1
Changes made for revising:
(1) Changed glass type from single clear 1/8” to single green 1/8”,
(2) Affiliated Zone 7 to AHU-2 instead to AHU-1,
(3) Used the special weather file of 2010 from the weather station in downtown Los
Angeles instead of general weather file.
Simulation Results and Real Data

Fig. B.1 Chilled Water Usage Data (First Round)
107

Fig. B.2 Total Electricity Usage Data (First Round)

B.2 Round 2
Changes made for revising:
(1) Revised lighting schedules
Simulation Results and Real Data
108

Fig. B.3 Chilled Water Usage Data (Second Round)

Fig. B.4 Total Electricity Usage Data (Second Round)
B.3 Round 3
Changes made for revising:
(1) Revise internal equipment schedules
109

Simulation Results and Real Data

Fig. B.5 Chilled Water Usage Data (Third Round)

Fig. B.6 Total Electricity Usage Data (Third Round)

110

B.4 Round 4
Changes made for revising:
(1) Revise internal equipment schedules
Simulation Results and Real Data

Fig. B.7 Chilled Water Usage Data (Forth Round)

111

Fig. B.8 Total Electricity Usage Data (Third Round)

B.4 Error Analysis of Intermediate Results
Table B Error Analysis of Intermediate Results
Chiller Water Usage Total Electricity Usage
R1 R2 R3 R4 R1 R2 R3 R4
CVRMSE 64% 47% 40% 19.8% 159% 68% 26% 11.3%
NMBE -56% -40% -33% -3.2& -164% -69% -23% -7.0%

112

APPENDIX C: INTERMEDIATE SIMULATION
RESULTS FOR INFLUENCE ANALYZING

C.1 Changing Area of Windows
Table C.1 Electricity consumption differences under multiple window areas
Window Area (sqft)
3482 4862 5583 6274
6964
(original)
7684
Window Area Ratio
(compared to original)
50% 70% 80% 90% 100% 110%
Jan 1496 1496 1496 1496 1496 1499
Feb 1498 1499 1499 1499 1499 1500
Mar 1714 1719 1721 1721 1721 1727
Apr 1617 1619 1619 1619 1619 1621
May 1636 1642 1647 1647 1652 1671
Jun 1941 1951 1967 1967 1972 2027
Jul 1874 1903 1926 1935 1952 2007
Aug 1960 1990 2015 2022 2040 2090
Sep 1988 2012 2027 2037 2057 2090
Oct 1735 1749 1755 1758 1767 1780
Nov 1526 1527 1527 1527 1529 1531
Dec 1394 1394 1395 1395 1395 1396
CVRMSE

2.4% 1.6% 0.9% 0.6% 0% 1.7%
NMBE 1.7% 1.0% 0.6% 0.4% 0% -1.3%

113

C.2 Changing U-Values of Walls
Table C.2 Electricity consumption differences under different U-values of exterior walls
U-Value Change Radio 50% 80% 90%
100%
(Original)
120% 150% 200%
Jan 1497 1496 1496 1497 1497 1498 1498
Feb 1498 1499 1499 1499 1499 1499 1500
Mar 1715 1715 1715 1716 1716 1717 1721
Apr 1619 1619 1619 1619 1619 1620 1621
May 1637 1639 1640 1640 1641 1646 1652
Jun 1945 1949 1950 1952 1954 1958 1977
Jul 1885 1899 1903 1909 1918 1932 1957
Aug 1968 1979 1981 1984 1995 2014 2042
Sep 1997 2013 2016 2022 2024 2043 2058
Oct 1739 1739 1740 1740 1744 1748 1760
Nov 1528 1528 1529 1529 1529 1530 1531
Dec 1393 1496 1394 1394 1395 1396 1397
CVRMSE

0.7% 0.3% 0.2% 0% 0.3% 0.8% 1.6%
NMBE 0.4% 0.2% 0.1% 0% -0.2% -0.5% -1.1%

114

C.3 Changing Weather Files
Table C.3 Electricity consumption differences under different weather files
Weather File Type 2010 Special (Original) Los Angeles General CZ8 General
Jan 1497 1235 1238
Feb 1499 1275 1339
Mar 1716 1366 1410
Apr 1619 1397 1469
May 1640 1364 1429
Jun 1952 1597 1822
Jul 1909 1728 1841
Aug 1984 1787 1874
Sep 2022 1757 1772
Oct 1740 1486 1577
Nov 1529 1356 1436
Dec 1394 1321 1361
CVRMSE

0% 15.1% 10.9%
NMBE 0% 15.1% 10.3%

115

C.4 Changing Lighting Power Density
Table C.4 Electricity consumption differences under different LPD numbers
LPD No. Change Radio 50% 80% 90%
100%
(Original)
120% 150% 200%
Jan 1344 1436 1469 1497 1555 1647 1800
Feb 1338 1436 1473 1499 1566 1671 1825
Mar 1540 1651 1689 1716 1797 1906 2093
Apr 1440 1547 1587 1619 1686 1795 1982
May 1476 1580 1619 1640 1714 1818 2007
Jun 1765 1888 1931 1952 2064 2208 2417
Jul 1744 1866 1904 1909 2045 2166 2377
Aug 1823 1955 1999 1984 2127 2281 2492
Sep 1860 1975 2019 2022 2136 2258 2480
Oct 1589 1693 1733 1740 1835 1943 2126
Nov 1358 1460 1496 1529 1599 1705 1884
Dec 1226 1324 1362 1394 1462 1563 1729
CVRMSE

10.20% 3.60% 1.40% 0% 5.60% 12.80% 24.30%
NMBE 10.60% 3.70% 1.20% 0% -5.80% 13.10% 25.10%

Abstract (if available)

Abstract Buildings in United States are responsible for close to half of the total energy consumed in the country, with existing building stock as the main responsible factor. While just a small proportion of that energy is consumed during the extraction of raw materials and construction phases, the majority of that energy is used for ongoing building operations after construction is complete (EIA, 2011). New technologies and management methods of building energy conservation have greatly decreased energy consumption in new buildings. ❧ Compared with new constructions, energy performance of a large amount of existing buildings is poor. There are many methods of upgrading these buildings to achieve lower energy consumption, including upgrading building mechanical systems, using high quality windows and doors, adding extra insulate layers, etc., could be done in order to avoid energy wasting and decrease operation price. However, not all these methods may be very effective if one does not know how the building is currently performing at a more detailed level than just looking at energy bills. It is very useful perform energy simulations as part of a well-managed process of improving performance of an existing building. As a result, calibration of an energy model to the existing building's actual performance is considered as one of the best important steps in whole retrofit process. ❧ In order to get an accurate simulation result, proper building information collection and simplification are needed for simulation. However, different software, even when used by a careful, knowledgeable person, may have different results even with similar input. Compared to the real world situation, simplifications of energy models are unavoidable. Therefore, there are always differences between simulation results and real energy consumption. Yet, the first step in predicting future savings is to create a calibrated model that can then be used for energy simulation before the retrofit happens. ❧ KAP on the USC campus was used as a case study for developing the process of creating a calibrated energy models that could be used for future upgrades to the building General steps for calibration and energy simulation were also developed and discussed in this thesis. A set of guidelines has been developed as a result of the calibration study.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Facade retrofit: enhancing energy performance in existing buildings

PDF

Digital tree simulation for residential building energy savings: shading and evapotranspiration

PDF

Energy efficient buildings: a method of probabilistic risk assessment using building energy simulation

PDF

Post occupancy energy evaluation of Ronald Tutor Hall using eQUEST; computer based simulation of existing building and comparison of data

PDF

Energy use intensity estimation method based on building façade features by using regression models

PDF

A simplified building energy simulation tool: material and environmental properties effects on HVAC performance

PDF

Life cycle assessment: existing building retrofit versus replacement

PDF

Zero net energy institutional building

PDF

Net zero energy building: the integration of design strategies and PVs for zero-energy consumption

PDF

Visualizing thermal data in a building information model

PDF

Simulation-based electric lighting control algorithm: integrating daylight simulation with daylight-linked lighting control

PDF

Building retrofitting evaluation: Energy savings and cost effectiveness of building retrofits on Graduate Art Studios at the University of California, Los Angeles

PDF

Geothermal heat pump's energy effect and economical benefits

PDF

Building energy performance estimation approach: facade visual information-driven benchmark performance model

PDF

ctrl+z: exploring the effects of undoing retrofits to pre-war buildings in Los Angeles

PDF

A parametric study of the thermal performance of green roofs in different climates through energy modeling

PDF

District energy systems: Studying building types at an urban scale to understand building energy consumption and waste energy generation

PDF

Landscape and building solar loads: development of a computer-based tool to aid in the design of landscape to reduce solar gain and energy consumption in low-rise residential buildings

PDF

Bridging performance gaps by occupancy and weather data-driven energy prediction modeling using neural networks

PDF

Daylight prediction: an evaluation of daylighting simulation software for four cases

Asset Metadata

Creator Yang, Guang (author)

Core Title Energy simulation in existing buildings: calibrating the model for retrofit studies

School School of Architecture

Degree Master of Building Science

Degree Program Building Science

Publication Date 07/31/2012

Defense Date 05/09/2012

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag building energy,calibration,eQuest,OAI-PMH Harvest,simulation

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Kensek, Karen M. (committee chair), Schiler, Marc (committee member), Xing, Tianxin (committee member)

Creator Email yangguang00@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-78482

Unique identifier UC11289938

Identifier usctheses-c3-78482 (legacy record id)

Legacy Identifier etd-YangGuang-1087.pdf

Dmrecord 78482

Document Type Thesis

Rights Yang, Guang

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA