URBAN & REGIONAL INFORMATION SYSTEMS PRACTICUM SOLAR FEASIBILITY ANALYSIS USING ARCMAP & ESRI CITYENGINE FLORIDA STATE UNIVERSITY SPRING 2014
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
URBAN & REGIONAL INFORMATION SYSTEMS PRACTICUM
SOLAR FEASIBILITY ANALYSIS USING ARCMAP & ESRI CITYENGINE
FLORIDA STATE UNIVERSITY SPRING 2014
1
ACKNOWLEDGEMENTS
This report was prepared in partial fulfillment of the requirements for the
completion of URP 5279: Urban and Regional Information Systems Practicum. This
project was commissioned to identify suitable building roofs on the Florida State
University (FSU) Campus to install photovoltaic solar panels, as well as develop
3D visualizations of the campus and the suitable buildings. We would like to
acknowledge and thank the following people for their help with this project:
COLLEGE OF ENGINEERING
Justin Vandenbroeck
ESRI
FSU FACILITIES AND UTILITIES
Mark Bertolami,
David Thayer,
Jim Stephens, and
Elizabeth Swiman
FSU OFFICE OF RESEARCH
Mary Jo Specter
TALLAHASSEE-LEON COUNTY GIS
Scott Weisman
Greg Maudlin
PROJECT MANAGER: SHAWN LEWERS
LIDAR Team Jeremy Crute
James Gaboardi
Doug Kossert
Andrew Uhlir
Wes Shaffer
City Engine Team Nicholas Alexandrou
Kathryn Angleton
Daniel Crotty
Susan Segura
Ana Thomas
2
ABSTRACT
The work detailed in this paper was done as a demonstration of GIS and 3D
GIS tools employed to identify potential photovoltaic (PV) solar panel installation at
Florida State University (FSU). There are two parts to the research. The first is a solar
suitability analysis utilizing LiDAR data to model solar radiation exposure on FSU's
buildings accounting for seasonal variations. Tools intrinsic to ESRI's ArcGIS version
10.2 were utilized to produce a series of raster datasets modeling solar radiation
exposure on rooftops. Several buildings were identified in this analysis that was
previously not strongly considered as candidates for PV installation. The models and
analysis outputs are part of this document.
The second effort created a 3D model/visualization of the campus using
ESRI's City Engine software. Here the project team leveraged existing GIS data sets
and procedural rule sets to create the model. The software offers numerous tools to
then customize individual building facades and add features to buildings such as
PV panels. This modeling effort provides end users with the ability to view realistic
renderings of the main campus and make changes to the models on the fly. The
software also offers analysis tool making the more than a mere visualization. These
models will be useful in the planning and political process as decisions are made
going forward with creating a renewable energy environment at FSU.
We conclude that many buildings on the campus are not suitable to be fitted
with PV panels due to issues such as structural, construction warranty and
architectural constraints. However, at least eighteen buildings including parking
structures, the Civic Center, the Indoor Practice Facility, and the Chemical Science
Buildings are strong candidates for photovoltaic panels and that under the right
circumstances can be cost effectively applied in cases where public/private
financing options are employed.
An extensive literature review of what other universities are doing to deploy
PV panels was also conducted, indicating at the time of this work several universities
have made significant strides in developing renewable energy sources. Among the
best was Arizona State University.
3
TABLE OF CONTENTS
Acknowledgements ..................................................................................................................... 1
Abstract ......................................................................................................................................... 2
Table of Contents ......................................................................................................................... 3
Table of Figures & Tables ............................................................................................................. 5
1.0 Introduction ............................................................................................................................. 6
2.0 Profiles of Photovoltaic Energy on University Campuses .................................................. 8
Arizona State University ........................................................................................................ 9
Rutgers, The State University of New Jersey ..................................................................... 13
2.1 Profiles of Photovoltaic Energy on University Campuses in Florida ............................ 14
University of Central Florida ............................................................................................... 15
University of Florida .............................................................................................................. 16
Florida Gulf Coast University .............................................................................................. 16
2.2 Lessons Learned ................................................................................................................ 17
3.0 Literature Review .................................................................................................................. 19
3.1 Estimating the Suitability of Rooftops for PV .................................................................. 19
3.2 GIS-Based Solar Suitability Studies .................................................................................. 21
4.0 Solar Suitability Methodology ............................................................................................. 23
4.1 Data .................................................................................................................................... 23
4.2 Methodology ..................................................................................................................... 25
5.0 Results of Solar Suitability Analysis ...................................................................................... 32
Parking Garages .................................................................................................................. 33
Civic Center and the Indoor Practice Facility ................................................................ 33
Additional Suitable Buildings.............................................................................................. 36
Leach Student Recreation Center and the Tully Gymnasium ...................................... 36
Earth Ocean and Atmospheric Science Building .......................................................... 37
Summary ............................................................................................................................... 38
6.0 CityEngine ............................................................................................................................. 39
CityEngine Software Application and Capabilities ........................................................ 39
Procedural Rules vs. Object-Oriented Modeling ............................................................ 40
Comparing CityEngine to SketchUp and Revit............................................................... 41
Using CityEngine to Complement Solar Suitability Analysis ........................................... 42
4
Using City Engine as a Campus Planning Tool ................................................................ 43
Methodology ....................................................................................................................... 43
Results.................................................................................................................................... 46
Visualization as a Planning Tool ......................................................................................... 47
Recommendations ............................................................................................................. 48
7.0 Conclusion ............................................................................................................................. 49
References .................................................................................................................................. 50
5
TABLE OF FIGURES & TABLES
Figure 1: ASU’s Total Annual Solar Energy Production .............................................................................. 10
Figure 2: Arizona State University .................................................................................................................. 10
Figure 3: ASU’s solar installation on the Tempe Campus ......................................................................... 11
Figure 4: ASU Solar Parking Facility examples ............................................................................................ 12
Figure 5: Rutgers Solar Parking Facility examples ...................................................................................... 13
Figure 6: PV Solar Resource of the United States ...................................................................................... 14
Figure 7: UCF Solar Umbrella ......................................................................................................................... 15
Figure 8: University of Central Florida solar panels .................................................................................... 15
Figure 9: Eastside campus installation at UF ............................................................................................... 16
Figure 10: Solar Field at FGCU ....................................................................................................................... 17
Figure 11: Radiation, Ground, Aspect and Slope Masks ......................................................................... 20
Figure 12: : Model 1 (a) and Model 2(b) results ......................................................................................... 21
Figure 13: Chemical Sciences Laboratory ................................................................................................. 23
Figure 14: Woodward Parking Garage ....................................................................................................... 24
Figure 15: Doak Campbell Stadium & Campus ........................................................................................ 24
Figure 16: Chemical Sciences Laboratory ................................................................................................. 27
Figure 17: Representative procedure of the model employed in this study ....................................... 28
Figure 18: Elevation, Slope, and Aspect Rasters derived from the LAS Dataset LiDAR data ........... 29
Figure 19: Suitable Slope and Aspect ......................................................................................................... 30
Figure 20: Solar Radiation Map ..................................................................................................................... 31
Figure 21: 2015 Solar Radiation clipped by building footprints .............................................................. 34
Figure 22: Suitable Parking Garages for PV Installations .......................................................................... 35
Figure 23: Solar Radiation for the Chemical Science Laboratory, 2015 ............................................... 36
Figure 24: Leach Student Recreation Center and the Tully Gymnasium ............................................. 36
Figure 25: Suitable sites for PV installations ................................................................................................. 37
Figure 26: Suitable sites for PV installations ................................................................................................. 38
Figure 27: St. Augustine St. Garage ............................................................................................................. 46
Figure 28: Indoor Practice Facility ................................................................................................................ 47
Table 1: Top Ten University Campuses by Total Installed Solar Capacity ........................... 10
Table 2: ASU’s Solar Profile per year (as of November, 2013)............................................... 11
Table 3: Estimations of Total Energy Production ..................................................................... 34
6
1.0 INTRODUCTION
A push for alternative energy
sources has become more
necessary due to the increasing
costs of oil and other finite resource
consumption, along with the political
and environmental ramifications
involved in their dependency. Over
the past few years, solar energy
system costs have dropped
drastically with innovations in
technology, increasing its feasibility
as an efficient generator of clean,
sustainable energy (U.S. Department
of Energy, n.d.). The most commonly
known solar energy source that is
becoming even more popular for
local domestic use is solar
photovoltaic cells, due to its ability to
be deployed at almost any scale.
Solar photovoltaic (PV)
systems use photovoltaic cell
technology to capture radiant
energy. The ability of these cells to
be incorporated into building and
roofing materials has led to their
successful implementation on
homes, businesses, manufacturing
facilities, and even utility projects
comparable in scale to large fossil-
fuel generation facilities(Morley,
2014) .
Based on a study done by
Denholm and Margolis, 60% of
commercial building rooftops in
warm climates and 65% of building
rooftops in cool climates display
optimal conditions for modular
efficiency at 13.5% from between
110 watts per meter squared to 135
watts meter squared (Lopez, Roberts,
Heimiller, Blair, & Porro, 2012).
However, the implementation of
solar energy systems is complex and
requires significant startup capital,
which has kept many cities and
universities from pursuing solar
energy on a large scale.
This study is not intended to
fully determine whether solar energy
is feasibility at FSU in terms of energy
production, economic cost, and
numerous other factors or make a
final decision on whether or not solar
power systems should be installed at
Florida State University. Instead this
study takes the first step in examining
the feasibility of photovoltaic energy
at FSU by evaluating whether FSU is
capable of producing sufficient solar
energy and to identify which
building roof tops are optimal
locations for photovoltaic panels
based on their location, elevation,
slope, aspect, and, most importantly,
the amount of solar radiation
received.
The following report uses
Geographic Information Systems
(GIS) to analyze the overall amount
of solar radiation potential for the
7
site-specific location of Florida State
University. By drawing from other GIS
modules, summarized in the
literature review, and using local
data, specific buildings on campus
were determined to have the
greatest potential for exposure to
solar radiation. These buildings offer
an initial understanding of solar
energy potential. The aspiration of
this report is to begin the discussion
of implementing photovoltaic arrays
on FSU’s campus, to identify the
optimal locations for pilot projects,
and to provide a base of information
to guide further research into the
financial feasibility of solar power.
In addition, this study utilizes
newly developed software called
City Engine to develop 3D
visualizations of FSU’s campus and of
the optimal buildings in particular.
This provides a visual reference what
buildings on campus would look like
with solar panels on them as well as
a useful tool for planning and
visualizing futures changes to
campus such as new building
construction.
The first section of this report
highlights the efforts of several other
universities across the country and in
Florida to utilize photovoltaic energy.
In addition, to providing an idea of
how feasible solar energy on college
campuses, the college profiles
uncover lessons learned and best
practices that FSU can learn from
mistakes and emulate innovative
ideas. The profiles are followed by a
discussion of existing research on
determining the optimal locations of
photovoltaic panels. This is used to
inform the next section which walks
through the methodology used in
this study to evaluate the suitability
of FSU’s campus for solar
photovoltaic energy production.
The results section highlights
eighteen building that were
identified as being suitable locations
for solar arrays and provides
estimates concerning the amount of
energy these rooftops could
generate. Finally, a brief overview of
the process and the benefits of using
CityEngine to generate 3D
visualizations of the optimal buildings
is laid out.
8
2.0 PROFILES OF PHOTOVOLTAIC ENERGY ON UNIVERSITY
CAMPUSES
The beginning of the literature
review involved finding current
university programs that are leading
the way in the implementation of
solar installations. Table 1 displays
ten universities with the largest
installed solar capacity in the U.S.
This offers two important insights.
First, many universities have
successfully installed large solar
arrays. If FSU were to pursue
installing solar arrays, it would not be
embarking on uncharted territory.
The examples of other universities
can be examined to gain insight into
suitable locations, appropriate
financing mechanisms, and
potential obstacles. Second, while
many of the top universities are
located in the western United States
and receive significantly more solar
radiation than the rest of the
country, some of the largest arrays
are located all across the country. In
fact, Rutgers has the second largest
array in the nation despite being
located in New Jersey, a state with
relatively low amounts of solar
radiation.
TABLE 1: TOP TEN UNIVERSITY CAMPUSES BY TOTAL INSTALLED SOLAR
CAPACITY
Source: http://www.aashe.org/resources/campus-solar-photovoltaic-installations/top10/#top-capacity
(AASHE, 2012)
Rank Inst itut ion Name State
Total Capacity
(kW)
1 Arizona State University Arizona 23,567
2 Rutgers, the State University of New Jersey New Jersey 17,417
3 Mount St. Mary's University Maryland 17,400
4 West Hills Community College District California 6,000
5 United States Air Force Academy Colorado 6,000
6 Colorado State University Colorado 5,539
7 California State University, Fullerton California 5,400
8 Arizona Western College Arizona 5,105
9 Butte College Montana 4,616
10 The University of Arizona Arizona 4,428
9
To provide an in-depth look at several universities efforts to install solar
arrays, the remainder of this section provides profiles of a few campuses that are
leading the nation in solar production. Several Florida universities will also be
highlighted to provide examples that are more relevant to FSU’s context. These
examples demonstrate that photovoltaic energy production can be successfully
implemented on research university campuses in Florida, and that universities
represent a unique opportunity for the implementation of these systems.
ARIZONA STATE UNIVERSITY
In October 2004, Arizona State University (ASU) began a comprehensive
sustainability effort to become a carbon neutral, zero-waste campus. A major
component of this endeavor involved implementing a large-scale solar energy
program. With its first solar panels installed in 2004, ASU was on track to become
the nation’s leader in solar energy. By November 2013, ASU had installed solar
arrays across all four of its campuses, and had the capacity to generate 23.5
megawatts per year of solar energy, as seen in Table 2. This represents 43% of
ASU’s peak daily load, and reduces ASU’s carbon footprint by 7.1% (Arizona
State University, 2014a).
TABLE 2: ASU’S SOLAR PROFILE PER YEAR (AS OF NOVEMBER, 2013)
Source: https://cfo.asu.edu/solar, 2014 (Arizona State University, 2014a)
ASU’s major installations included placing solar panels on academic
buildings, athletic facilities, parking structures, and the basketball arena. As
seen in Figure 2, the large majority of ASU’s solar installations are building
mounts. Like, FSU, much of ASU’s campuses are located in relatively urbanized
areas and, as such, they do not have enough available land for large ground
mounted solar arrays. In this way, ASU demonstrates that universities can
successfully retrofit buildings with solar panels at a large scale.
Date of First Solar Installat ion 2004
Total Solar Generat ion Capacity (MW) 23.5 MW
Total PV Generat ion Capacity (MW) 21.2 MW
Total Solar Systems 86
Total Number of PV Panels Installed 78,100
Total Number of Shaded Parking Spaces 5,447
10
23
1,642
3,665
10,846
24,204
29,531
-
5,000
10,000
15,000
20,000
25,000
30,000
35,000
2008 2009 2010 2011 2012 2013
En
erg
y P
rod
uc
tio
n (
mW
h)
FIGURE 2: ARIZONA STATE UNIVERSITY SOURCE: HTTPS://ASUNEWS.ASU.EDU/20120213_ASU_SOLAR, 2012
(ARIZONA STATE UNIVERSITY, 2012)
FIGURE 1: ASU’S TOTAL ANNUAL SOLAR ENERGY PRODUCTION SOURCE: ARIZONA STATE UNIVERSITY SOLAR, 2014 (ARIZONA STATE UNIVERSITY, 2014A)
11
Figure 3: ASU’s solar installation on the Tempe Campus
Source: https://www.asu.edu/fm/images/solarization/solar-map-tempe.pdf
(Arizona State University, 2014d)
12
One of the most unique and
exciting ways that ASU has
maximized their potential for solar
energy production is found in ASU’s
mounting of extensive solar arrays on
top of parking garages and surface
lots. Not only do parking garages
present large flat rooftops that are
perfect for generating solar energy,
but installing solar panels on parking
garages can simultaneously address
multiple other issues. In particular,
the question of whether solar panels
will harm the campus aesthetic
character that so often creates an
obstacle to solar installations is less of
an issue because solar panels may
even improve the aesthetics of most
parking garages. At the same time,
solar panels can provide much
needed shade to
vehicles on the top
floor. To capitalize
on this opportunity,
ASU has installed
solar arrays on nine
parking garages
and ten surface
parking lots such as
those seen in Figure
3 (Arizona State
University, 2014a).
The potential of
using parking
garages as an
opportunity of solar
array installations
has since be
recognized elsewhere as universities
including Rutgers, California State
University, UCF, and many others
have installed solar arrays on their
parking garages and surface
lots.Like most other universities, ASU’s
solar systems were financed through
a Purchase Power Agreement (PPA)
with a third-party solar installation
company. Under a PPA, a third-
party company installs, owns,
operates, and maintains the
installation for a predetermined time
period (United States Environmental
Protection Agency, 2014). During this
time, the university simply purchases
the energy produced by the solar
arrays. At the end of the time period,
the university has the option to
purchase the solar array for a fair
FIGURE 4: ASU SOLAR PARKING FACILITY EXAMPLES
SOURCE: HTTPS://ASUNEWS.ASU.EDU/20120213_ASU_SOLAR
(ARIZONA STATE UNIVERSITY, 2014A)
13
market price. In ASU’s case, one of
several companies including NRG
Solar, Encap Renewables, and
Renewable Energy I will own the
installations for fifteen years, at which
time, ASU can the purchase the
system (Arizona State University,
2014b). Thus, as of April 2014, ASU
only owned ten of its solar systems.
RUTGERS, THE STATE UNIVERSITY OF
NEW JERSEY
Rutgers is another nation-wide
leader among universities in solar
energy production. All of Rutgers’
17,417 kW solar capacity comes
from two very large solar arrays: a 7-
acre solar farm and a 28-acre solar
canopy over surface parking lots.
Together these two arrays generate
enough electricity to cover 63% of
the electricity demand of one of
Rutgers’ campuses and are
projected to yield $28 million in net
savings over the next 28 years
(Rutgers Climate Institute, 2014).
Rutgers is of particular interest
for its integration of photovoltaic
research into its efforts. Like FSU,
Rutgers is a Tier 1 research university,
and it has leveraged this status to
pursue a mutually reinforcing effort
that uses its solar arrays to research
the most efficient practices. With
the development of the Rutgers
Energy Institute, Rutgers became a
FIGURE 5: RUTGERS SOLAR PARKING FACILITY EXAMPLES SOURCE: HTTP://SOLAIREGENERATION.COM/PROJECT/RUTGERS-UNIVERSITY/ (SOLAIR GENERATION, 2013)
14
leader in photovoltaic research. This research has encouraged the
installation of Rutgers solar arrays,
but even more than that, it has
turned Rutgers campus into a living
laboratory for photovoltaic research.
2.1 PROFILES OF PHOTOVOLTAIC
ENERGY ON UNIVERSITY
CAMPUSES IN FLORIDA
It is apparent that universities
across the country have successfully
developed photovoltaic energy
systems that reduce their carbon
footprint and energy costs. Even so,
since the success of photovoltaic
energy is directly related to their
geographic location and the solar
radiation it receives (as seen in
Figure 6), questions could still remain
over whether photovoltaic energy is
feasible in Florida universities.
However, several universities in
Florida have already put these
questions to rest by successfully
installing large photovoltaic arrays.
FIGURE 6: PV SOLAR RESOURCE OF THE UNITED STATES SOURCE: HTTP://WWW.NREL.GOV/GIS/IMAGES/MAP_PV_NATIONAL_LO-RES.JPG (NREL, 2008)
15
UNIVERSITY OF CENTRAL FLORIDA
In 2011, the University of Central
Florida developed and installed 440 solar
panels on a 107kW ground-mounted
array capable of offsetting more than
half of the energy consumption of the
parking garage where it is housed, which
is one of the largest and busiest on
campus located near the UCF
Recreation and Wellness Center (Solar
World, 2012). Fully funded by a grant
from Progress Energy Florida, the array will
is projected to produce 157,000 kWh per
year which will save UCF $15,000 in
annual energy costs (University of Central
Florida, 2012). This project is a part of
UCF’s larger effort to produce 15% of its
energy with renewable sources within the
next decade.
Like Rutgers, UCF has made great strides to capitalize on the unique
opportunity research universities have to develop, test, install, and promote solar
power at the same time. Together with the Florida Solar Energy Center (FSEC),
UCF established the Solar Energy Center on it Cocoa campus. With the
capacity to house 2-megawatt solar arrays this 70,000 sq. ft. research facility is
the largest renewable energy research and testing facility in the U.S. (Florida
Solar Energy Center, 2007).
In addition to this dedicated
facility, UCF has also begun turning its
classroom buildings into a living solar
laboratory. In 2009, UCF installed a 21 kW
array on top of the Harris Engineering
Building. The array was purposefully built
with two different type of solar inverters to
test their efficiency. In this way, on top of
providing 11.2 kW of power every year,
the wall-mounted array serves as a
FIGURE 8: UNIVERSITY OF CENTRAL
FLORIDA SOLAR PANELS
Source: http://www.solarworld-
usa.com/blog/2012/may/university-central-
florida.aspx (Solar World, 2012)
FIGURE 7: UCF SOLAR UMBRELLA
Source: http://www.ideasforus.org/tag/ucf/
(IDEAS, 2011)
16
teaching aid for UCF’s Renewable Energy Program (University of Central Florida,
n.d.).
Much like the plan being implemented by UCF, one vision with this project
is to make photovoltaic panels an integral part of the Earth, Ocean, and
Atmospheric Sciences building that is scheduled for Phase 1 completion in July
2016 on Florida State University's campus. The idea is a fusion of design,
pedagogy, and self-sufficiency where the structure will not only be aesthetically
pleasing, but can sustain itself (to some degree) and also be an instrument for
research and instruction (Florida State
University Facility Program, 2013).
UNIVERSITY OF FLORIDA
The University of Florida began
installing photovoltaic panels in 2010, with a
75kW array at Powell Hall, and have since
added no less than 5 arrays at various
locations around campus which are
capable of producing nearly 300kW of
energy. Locations include the Department
of Microbiology and Cell Science, UF Hillel,
and Energy Research and Education Park
among others (Sustainable UF, 2014b). More
can be viewed at (Sustainable UF, 2014a),
which also displays real-time energy
production.
FLORIDA GULF COAST UNIVERSITY
Instead of mounting photovoltaic panels on structures, Florida Gulf Coast
University completed an array known as Solar Field in late 2010. The array is
200,000 square feet and consists of more than 10,000 panels on 15 acres of land
to the east of campus. The Solar Field array is capable of producing 2
megawatts of energy and supplies 18% of the university’s electricity, including for
three academic buildings (Florida Gulf Coast University, n.d., 2010a, 2010b).
FIGURE 9: EASTSIDE CAMPUS
INSTALLATION AT UF
Source:
http://sustainable.ufl.edu/topics/energy-
climate-change-at-uf/renewable-energy/
(Sustainable UF, 2014b)
17
2.2 LESSONS LEARNED
As the project team examined
the efforts of universities across the
country to become more sustainable
by developing solar energy systems,
several themes emerged. These
themes are compiled below as a list
of lessons learned and best practices
for FSU when considering whether to
develop solar energy systems.
Solar energy is a viable source
of renewable energy on university
campuses. Universities have
successfully installed solar arrays at
multiple different scales and in many
different geographic locations.
Some campuses have as much as
63% of their energy usage being
produced by solar arrays (Rutgers
Climate Institute, 2014). Other
universities, especially in Florida, are
only beginning to tap into their
potential for solar energy. However,
regardless of scale, solar energy is
being used to reduce the carbon
footprint of universities across the
country. Thus, any effort by FSU to
follow in these examples would be
following a well-trodden path.
Similarly, while geographic location
and the subsequent level solar
radiation will always play a role, it
does not appear to be the
determining factor of whether solar
Figure 10: Solar Field at FGCU Source: http://www.fgcu.edu/Facilities/SolarField.html (Florida Gulf Coast University, 2014)
18
energy is a viable option. Florida
universities have already
demonstrated that Florida’s
geography is conducive for solar
energy. Thus, the fact that FSU is not
located in Southwest does not
necessarily mean that photovoltaic
installations are not feasible.
Collegiate solar initiatives can
be successful regardless of whether
installations are ground-mounted
solar farms or retrofitted on existing
buildings with solar panels.
Universities have successfully installed
large ground-mounted solar farms. In
fact, all of Mount St. Mary's
University’s 17.4 mW capacity comes
from a single 100-acre solar farm
(Mount St. Mary’s University, 2011).
At the same time, universities with
land constraints such as ASU have
installed solar arrays on new and
existing buildings and parking
facilities. Thus, even though
mounting solar panels on buildings is
more expensive, urban universities
like FSU do not need large tracts of
land to successfully install large solar
arrays.
Parking facilities represent a
unique opportunity to solar arrays on
university campuses. Especially for
universities with land constraints,
parking structures are often able to
house the largest solar arrays at the
lowest cost. They avoid common
barriers such as the loss of aesthetic
value and the potential of causing
roofs to leak. Finally, parking arrays
provide the extra benefit of shading
cars.
Supportive state policies
including subsidies and tax credits
are important in successfully
implementing a solar installation
program. All of the profiled
universities made use of state
subsidies, tax-credits, or grants to
finance the installation of solar
panels. In this way, the viability of
solar energy will depend in part on
the availability of financial
incentives. While Florida’s incentive
structure of solar energy may not be
as extensive as California’s, it has
provided enough incentives for UF,
UCF, and FGCU to construct solar
arrays.
Power Purchase Agreements
are the standard financing
mechanism for large solar arrays.
Since universities pay no upfront
capital costs and are guaranteed
predictable energy prices, PPAs
minimize much of the financial risk
otherwise faced by the university
(United States Environmental
Protection Agency, 2014). Other
financial mechanisms are available,
but without a PPA universities often
rely on winning renewable energy
grants.
19
3.0 LITERATURE REVIEW
3.1 ESTIMATING THE SUITABILITY
OF ROOFTOPS FOR PV
The previous section demonstrated
that there is a growing movement
among universities to pursue the
installation of photovoltaic panels as
a cleaner and potentially more cost-
effective means of energy
production. This study seeks to
discern whether it is feasible for FSU
to join this movement, and if so, to
identify locations on campus that
are suitable for solar arrays.
Fortunately, the growing need for
sustainable energy has motivated a
large body of research evaluating
the suitability of rooftops for
photovoltaic energy production that
served as a guide for this study’s
approach and methodology.
Research on the efficiency of solar
panels has led to the development
of commonly accepted factors that
can maximize the energy production
of roof mounted photovoltaic arrays.
For example, in the northern
hemisphere south-facing rooftops
receive more annual solar radiation.
Similarly, rooftops that are flat or
have less than a 35-degree pitch
also receive more annual solar
radiation than sharply slanted roofs.
The challenge then becomes how to
systematically identify buildings that
meet these criteria and are, thus,
optimal locations for photovoltaic
arrays.
Three primary types of methods have
been utilized to evaluate the
suitability of rooftops for photovoltaic
energy production: constant-value
methods, manual selection methods,
and GIS-based methods. However,
GIS-based methods have
consistently demonstrated their
superiority over other methods in
their ability to provide objective,
accurate, and replicable results
(Melius, Margolis, & Ong, 2013).
Constant-value methods are based
upon far-reaching assumptions
concerning the amount of roof
space in the study area that meet
the ideal criteria. These are then
applied to the total amount of roof
area (typically estimated from
building footprints) in the study area
to estimate the amount of suitable
roof space. Manual-selection
methods, on the other hand, closely
examine aerial photography to
identify suitable rooftops. By relying
on assumptions and visual
approximations, constant value and
manual selection methods fail to
achieve the desired levels of
objectivity and precision.
GIS’ ability to use computer models
to systematically identify locations
that meet the ideal criteria sets it
apart as the most effective tool for
identifying the optimal locations for
solar panels and estimating the
photovoltaic energy potential of
identified rooftops (Melius et al.,
2013). However, even within GIS-
20
based methods, there remains
significant variation in the different
methodologies and models that
have been employed. Nevertheless,
as GIS software has advanced to
enable greater functionality and
more powerful radiation tools, the
literature has converged toward
common methodological
approaches. The remainder of this
section will follow this
methodological progression to gain
insight into how to maximize GIS’
potential as a solar assessment tool
within FSU’s context. To date very
few GIS-based solar suitability studies
have been conducted on college
campuses. As such, this study will
also draw upon several citywide
studies.
Figure 11: Radiation, Ground, Aspect and Slope Masks Source: Applied Energy Article, Kucuksari, 2013 (Kucuksari et al., 2014)
21
3.2 GIS-BASED SOLAR SUITABILITY
STUDIES
This study proposes a
framework to use GIS, mathematical
optimization, and simulation modules
to predict the optimal placement
and size of PV units. The study uses
LiDAR data to generate a digital
elevation model (DEM). Spatial
analysis tools in GIS then provide a
filtering based on specific masks:
solar radiation, elevation, slope,
aspect, and a human mask in order
to find the most suitable rooftops for
PV installation. The outputs of these
masks can be seen in Figure 11.
Once the suitable rooftops are
found, the optimization module
determines the set up of the PV
systems, finding the optimal location
and number of PV units to maximize
profit over the next two decades.
These results can be seen in Figure
12, below. However, the authors
determined that the optimization
module provides crude
approximations of output. The results
of each PV system were run through
a power flow study simulator in order
to find more realistic final outputs.
These simulators better capture
voltage magnitudes and each
systems limitations. This optimization
process is out of the scope of the
intended solar analysis of our report,
but provides a glimpse of other
factors to consider when moving
further into the implementation
process.
A similar analysis was
conducted by Santos et al. (2011)
demonstrating the ability to locate
photovoltaic systems in Lisbon,
Portugal using LiDAR utilizing nearly
identical methods to study done at
ASU (Santos et al., 2011). Their study,
however, took into consideration
only the LiDAR data and not the
socioeconomic factors of
installations. Another study by
Kodysh et al. (2013) uses very similar
Figure 12: : Model 1 (a) and Model 2(b) results Source: Applied Energy Article, Kucuksari, 2013 (Kucuksari et al., 2014)
22
methodology to site the optimal
locations of solar installation in Knox
County, Tennessee (Kodysh,
Omitaomu, Bhaduri, & Neish, 2013).
According to “Advantages of
using lidar in GIS” by Esri, there are
many advantages of utilizing LiDAR
data in GIS. It is accurate and quick
to collect. Also, there are no
geometric distortions and it can be
easily combined with other forms of
data to enhance geographic
representations and data
visualizations. Elevation can be
determined from the returns, which
cannot be done with aerial imagery.
This has it’s benefits in any
application in which elevation needs
to be derived (Esri, 2013).
LiDAR allows for the extremely
precise and detailed extraction of
data and a variety of derived
products; solar mapping is only one
aspect focused on in this study.
Most government entities have
invested in this type of aerial
mapping, and many citizens have
access to it, but most don’t
understand the concepts or have
the skills to utilize it. In this study, the
project team focuses on using LiDAR
LAS return to form the best estimate
for the placement and installation of
photovoltaic panels on rooftops at
Florida State University.
23
4.0 SOLAR SUITABILITY METHODOLOGY
4.1 DATA
The only data used to complete the suitability study was Light Detection and
Ranging (LiDAR) data, which is portmanteau of Light and Radar, data provided
by Tallahassee-Leon County GIS. LiDAR is a remote sensing tool that is similar to
SONAR except that it uses an airborne, (typically carried by an airplane) pulsing
laser instead of sound waves. The laser is reflected by the surface to a sensor on
the airplane, which records the latitude, longitude, and elevation of each laser
pulse. Viewing all of these points together creates an extremely detailed and
three-dimensional model of the surface. It is detailed enough to record cars,
people, and even birds. Figures 13-15 provide examples of LiDAR returns for
various locations around FSU’s campus.
FIGURE 13: CHEMICAL SCIENCES LABORATORY
24
FIGURE 14: WOODWARD PARKING GARAGE
FIGURE 15: DOAK CAMPBELL STADIUM & CAMPUS
25
Since LiDAR is extremely
expensive, the Tallahassee-Leon
County GIS cannot collect LiDAR
continuously and the most recent
LiDAR data for Leon County was
collected between 2009 and 2012.
As such any recently constructed
buildings, such as the Indoor
Practice Facility, were not
represented in the data. These
buildings were included in the
suitability analysis by digitizing the
building footprints prior to performing
the solar analysis described below.
4.2 METHODOLOGY
The methodology used to
evaluate solar suitability was derived
from two main sources: An
Integrated GIS, optimization and
simulation framework for optimal PV
size and location in campus area
environments (2014) and ArcUser
Online: Locating Sites for
Photovoltaic Solar (Chaves & Bahil,
2010; Kucuksari et al., 2014). Suitable
locations are considered based on
elevation, slope, aspect, and the
amount of solar radiation.
Individual raster layers were
created for these characteristics
(aspect, slope, solar radiation)
based off of the elevation raster
created through LiDAR data
provided by Tallahassee-Leon
County GIS department. Esri
ArcMap 10.2 was used for the data
processing and analysis. The process
is displayed in Figure 16 and outlined
in more detailed below.
Make a LAS dataset in ArcCatalog
and adding to Map
The LAS dataset obtained from
the Tallahassee-Leon County
department was created by
combining all the LAS files from
around campus. Surface
constraints, Breaklines and AOI
boundaries were then added.
Statistics were then calculated and
the output was added to the map.
Create Elevation raster from LAS
Dataset (Figure 18)
The elevation raster provides
the height values from which all of
the remaining rasters are
interpolated. For this reason, it is
important to have a high spatial
resolution (cell size) in order to
provide the effective analysis of the
data. A cell size of 5 was chosen,
which provides quality detail and
allows efficient model performance.
Clipping raster to the FSU campus
The raster was then clipped to
the study area to allow for quicker
performance when calculating
other rasters. This was done by doing
a selection by location of the streets
within 750 feet of the buildings layer.
This allowed us to use the clip raster
tool to output a raster of just the FSU
campus and the immediate area.
Create Slope raster (Figure 18)
The slope raster interpolates
the gradient by finding the rate of
elevation change between cells.
26
The preferred slope is between 0
degrees and 35 degrees (Chaves
and Bahill, 2010). This angle is ideal
for receiving optimal solar radiation
year-round and minimizes the cost
for installing PV solar panels.
Create Aspect raster (Figure 18)
An aspect raster provides a
look at the direction of a slope. By
interpolating the heights provided in
the elevation raster, the output
values of the raster will be the
compass direction of the aspect.
The aspect is in a range from 0 to
359.9, measured clockwise from
north. Since Tallahassee is located in
the northern hemisphere, solar
panels provide the best
performance on south-facing slopes.
As such, the aspect should be south
facing or horizon. The value range
was between 112 degrees and 248
degrees.
Aspect/Slope raster
The aspect raster and slope
raster were combined to identify the
optimal locations for PV panels.
Conditional Aspect/Slope raster
(Figure 19)
The conditional tool with an
SQL expression of “FSU_Aspect >=112
AND FSU_Aspect <=248 AND
FSU_Slope <=35” was then used on
the Aspect/Slope combined Raster.
This provides an output raster with
cells that have both a slope
between 0-35 degrees and an
aspect between 112-248 degrees.
Create Solar Radiation raster (Figure
20)
The Area Solar Radiation tool
derives incoming solar radiation from
a raster surface. The project team
ran the tool to get the Solar Output
for 2015. The output is in Wh/m2 and
this raster was once again clipped
with the buildings polygon.
Calculating Potential Energy
Production
To calculate how much solar
radiation each suitable roof was
receiving the Zonal Statistics tool was
applied to the solar radiation raster. This
calculated the average solar radiation
received per square meter (kWh/m2).
This was then applied to the total roof
area that was suitable for solar panels,
as determined by the conditional slope
and aspect rater. This provided an
estimation of the total radiation
received by each building. To
convert this to the amount of energy
that could be generated by each
building, an assumed efficiency
rating of 15% was used. Newer solar
panels are now capable of reaching
20% efficiency, but 15% was used as
a conservative estimate.
27
Finally the architecture, aesthetics, and roof types were taken into
consideration because FSU has a rich architectural history that needs to be
preserved. Thus, any evaluation of suitable locations for solar panels on FSU’s
campus must factor in the potential impact solar panels would have on
architectural integrity and aesthetic value. These factors were taken into
account to complement the technical analysis ensuring that only buildings also
meet these criteria are identified as optimal locations for solar panels.
To verify the accuracy of our results a solar radiation estimation tool
developed by the National Renewable Energy Laboratory (NREL) called the
PVWatts Calculator was used as a baseline comparison (NREL, 2014). As seen in
Figure 16, the PV Calculator provides a rough estimate of the potential array size
and electric generation potential for selected buildings. The PV Calculator
estimates closely matched the project team’s results providing greater validity
to this study’s results.
Figure 16: Chemical Sciences Laboratory Source: http://pvwatts.nrel.gov/pvwatts.php (NREL, 2014)
28
FIGURE 17: REPRESENTATIVE PROCEDURE OF THE MODEL EMPLOYED IN THIS
STUDY
29
FIGURE 18: ELEVATION, SLOPE, AND ASPECT RASTERS DERIVED FROM THE LAS
DATASET LIDAR DATA
30
FIGURE 19: SUITABLE SLOPE AND ASPECT
31
FIGURE 20: SOLAR RADIATION MAP
32
5.0 RESULTS OF SOLAR SUITABILITY ANALYSIS
The results of the GIS-based solar rooftop suitability analysis strongly indicate that,
from an energy production perspective, FSU is capable of utilizing solar arrays for
energy production. As seen in Figure 21, eighteen buildings were identified as optimal
locations for solar arrays. These buildings were found to have roofs with the appropriate
size, slope, aspect, and elevation to maximize solar radiation. As such, they are found
to receive a significant amount of solar radiation and are capable of generating a
substantial amount of solar energy. These buildings do not represent the only buildings
on FSU’s campus that are suitable for solar arrays. Other areas on campus are able to
utilize solar arrays at smaller scales. However, the specified locations provided the
greatest potential for solar energy output.
Table 3 displays the estimations of the total energy that could be produced by
each of the eighteen identified buildings. The eighteen suitable buildings were
estimated to be capable of generating 15,792,957 kWh of energy per year. To put this
into perspective, in 2013, ASU total solar energy production was 29,530,506 kWh (Arizona
State University, 2014c).
Bld.
Num. Building Name
Suitable
Roof Area
(m2)
Radiation
per m2
(kWh/m2)
Total Solar
Radiation
(kWh)
Total
Energy
Produced
(kWh *15%)
4006 Parking Garage #1 - Woodward 19,591 1,389 19,427,811 2,914,172
4090 Indoor Practice Facility* 31,775 1,461 14,225,451 2,133,818
4025 Parking Garage #6 - St. Augustine 19,340 1,355 12,399,937 1,859,991
378 Parking Garage #4 - Call 21,519 1,336 9,797,809 1,469,671
4014 Parking Garage #5 - Copeland 29,924 1,374 8,978,921 1,346,838
4028 Parking Garage #3 - Spirit Way 45,621 1,469 8,760,152 1,314,023
70 Parking Garage # 2 - Traditions 23,896 1,331 7,952,885 1,192,933
4546 Tucker (Civic) Center 14,349 1,488 6,509,458 976,419
132 Tully Gym 7,338 1,484 3,316,877 497,532
54 Housewright Building 5,837 1,478 2,629,793 394,469
26 Leach Student Rec Center 5,753 1,478 2,590,000 388,500
135 Sandels Building 3,658 1,480 1,648,667 247,300
40 Duxbury Hall - Nursing 3,505 1,471 1,572,022 235,803
78 Mendenhall Building B 3,223 1,500 1,473,812 221,072
4008 Chemical Science Labs 2,835 1,527 1,325,660 198,849
2 Diffenbaugh Building 2,614 1,481 1,179,532 176,930
469 Postal & Receiving Services 1,974 1,500 902,279 135,342
57 Pepper Building 1,273 1,536 595,314 89,297
4006 Total 244,023 105,286,380 15,792,957
TABLE 3: ESTIMATIONS OF TOTAL ENERGY PRODUCTION
33
Indeed, if FSU were to install solar
arrays on these eighteen buildings, the
output could rival Rutgers and other
leading universities for net solar energy
production. Granted these are
estimates and should not be treated as
exact, they certainly provide a general
indication that photovoltaic energy is a
viable option on FSU’s campus. The
remainder of this section will highlight
the eighteen buildings identified as
optimal for photovoltaic arrays and the
opportunities they provide.
PARKING GARAGES
Similar to the results of the profiles of
other university’s solar efforts, FSU’s
parking garages were consistently
shown to be optimal locations for solar
arrays. Their large, flat roofs give them
an ideal slope and aspect for solar
installations. Granted, as seen in Figure
22, the slope and aspect rasters do not
indicate that parking garages were
within the acceptable ranges, however,
this is due to the slope of the ramps and
the presence of vehicles on rooftops
when the LiDAR data was collected.
Placing solar arrays over top of the
parking structures would smooth out the
unevenness providing six of the largest,
flat rooftops on campus. As seen in
Table 4, parking garages made up six of
FSU’s top seven highest building in terms
of total potential solar energy
production. Estimates indicate that FSU’s
six parking garages alone could
generate approximately 10 million kWh
of energy every year. In addition,
placing solar arrays on parking garages
would have minimal impacts on
campus aesthetics, and would provide
the added benefit of shade and rain
cover to the top floor of the garage.
CIVIC CENTER AND THE INDOOR
PRACTICE FACILITY
The only two other buildings on campus
that can compete with the parking
garages in terms of their size, solar
suitability, and potential energy
production are the Civic Center and the
Indoor Practice Facility. Even though
only the south end of Civic Center’s roof
is technically south facing (see Figure
21), the large, wide-open, and almost
flat nature of much of its roof space is
conducive for solar radiation, making it
a prime candidate for a large solar
array. In this way, FSU could follow in
ASU’s footsteps by installing a signature
solar array on its basketball arena.
Unfortunately, since the only available
LiDAR data for Leon County was
generated prior to the construction of
the Indoor Practice Facility, it was not
initially included in the analysis.
However, simple observation indicates
its large southward facing rooftop with a
suitable slope and no obstructions as
ideal for photovoltaics. Assuming the
entire roof to be flat, the Area Solar
Radiation tool exhibits it to be capable
of producing the second most energy of
any building on campus.
34
FIGURE 21: 2015 SOLAR RADIATION CLIPPED BY BUILDING FOOTPRINTS
35
FIGURE 22: SUITABLE PARKING GARAGES FOR PV INSTALLATIONS
36
ADDITIONAL SUITABLE BUILDINGS
Even though the Civic Center, IPF,
and parking garages represent
the primary candidates for large
500+ kW arrays, numerous other
building were also identified as
optimal locations for photovoltaic
arrays. As seen in Figure 25, these
buildings include the Chemical
Science Building, Mendenhall B,
Duxbury Hall, the Claude Pepper
Building, the Sandels Building, the
Housewright Building, the
Diffenbaugh building, and the
Postal Receiving Building. Although
smaller in size, they are still more than
capable of producing a significant
amount of solar energy. For
example, the Chemical Science
Building’s 668 square meters of
suitable roof space could generate
198,849 kWh of solar energy based
off a 15% efficiency. With a south
facing roof, an ideal slope, and no
obstructions, the Chemical Science
Laboratory Building displays
remarkable potential for solar energy
production. In fact, the Chemical
Science Building was found to
receive the second most radiation
per square meter of any building on
campus.
LEACH STUDENT RECREATION CENTER
AND THE TULLY GYMNASIUM
FSU’s largest existing solar energy
system is located on the Leach
Student Recreation Center. The
Leach Center’s solar thermal array is
the largest in Florida and was
projected to provide 90% of the
heat Leach’s large indoor
swimming pool (Florida State
University, n.d.; fsunews.com, 2011).
The solar suitability analysis provides
further confirmation that the Leach
Center is an excellent location for
solar panels. In addition to verifying
the validity of the Suitability Analysis,
the success of the Leach Center’s
existing solar array provides greater
insight in the potential of solar
energy of FSU’s campus. Even
FIGURE 23: SOLAR RADIATION FOR THE CHEMICAL
SCIENCE LABORATORY, 2015
FIGURE 24: LEACH STUDENT RECREATION CENTER
AND THE TULLY GYMNASIUM
37
though the Suitability Analysis
indicated the Leach Center was a
prime location for solar arrays, other
buildings such as parking garages,
the Indoor Practice Facility, and the
Chemical Science Laboratory
Building were found to have even
more solar potential. So, the Leach’s
relatively small array is capable of
producing enough energy to heat
the indoor pool, how much more
energy could be produced by large
solar arrays on FSU’s six parking
garages.
EARTH OCEAN AND ATMOSPHERIC
SCIENCE BUILDING
Since the proposed Earth Ocean
and Atmospheric Science Building
(EOAS) was in the design and
approval process at the time of this
study, it was not included in the
suitability analysis. However, the
building’s function as an
environmental research facility
provides a unique opportunity to
simultaneously provide a state of the
art renewable energy research
facility while also providing the
university with clean, renewable
energy. Installing a solar array on
the EOAS Building would turn the
building itself into a teaching aid
and a living laboratory. It could
even be used as a test facility for
new types of photovoltaic
technology.
FIGURE 25: SUITABLE SITES FOR PV INSTALLATIONS
38
SUMMARY
The solar suitability analysis indicates that FSU has the potential to generate
significant amounts of solar energy. In fact, FSU’s concentration of large, multi-
story buildings along with its position as a Tier 1 research university likely make it
one of the best locations for solar panels in all of Tallahassee. In this way, FSU
has a unique opportunity and a responsibility to lead Tallahassee and the State
of Florida in the pursuit of clean and renewable energy. This study provides the
first steps in that pursuit by identifying the buildings that is best suited for
accommodating solar arrays and generating significant amounts of solar
energy. To that end, eighteen existing buildings were identified as optimal
locations for solar arrays (Figure 26). Of these FSU’s six parking garages, the
Indoor Practice Facility, and the Civic Center were identified as optimal
locations for large 500+ kW arrays. The Leach Center, Tully Gymnasium, and the
Chemical Sciences Laboratory Building, along with several others, represent
excellent locations for smaller arrays. Installing solar arrays on just a few of these
buildings could quickly make FSU a leader in solar production.
FIGURE 26: SUITABLE SITES FOR PV INSTALLATIONS
39
6.0 CITYENGINE
To complement the LiDAR-
based solar suitability analysis, a
modeling software called CityEngine
was utilized to develop a 3D
visualization of FSU’s campus. These
models paid special attention to the
eighteen buildings deemed suitable
for solar arrays. CityEngine’s ability
to leverage GIS data enabled the
project team to create realistic
renderings of what identified
buildings would look like with solar
panels on them. In addition to
ensuring that solar panels would not
compromise the aesthetic integrity
of FSU’s campus, these visualizations
are useful tools for planning and
political process as FSU moves
forward in its effort to pursue
renewable energy.
CITYENGINE SOFTWARE APPLICATION
AND CAPABILITIES
ESRI’s CityEngine software
program transforms 2D data into
realistic 3D city models. Creating
smart 3D city models with CityEngine
is an effective medium for
communicating data into a visual
setting. The software application has
many features that contain a long-
range list of user-friendly capabilities.
The main features that are
conducive towards modeling
efficient 3D cities include procedural
modeling, modeling pipeline and
general ideas behind the procedural
modeling of models and cities and
buildings (Esri R&D Center Zurich,
2013). Procedural modeling
technology is a unique tool that
contains pre-coded commands,
and utilizes cumbersome tasks with a
mass automated generic task. The
modeling pipeline consists of several
procedural modeling tools for
generating large-scale urban
layouts, as well as applying CGA
rules for the creation of detailed
building models (Esri R&D Center
Zurich, 2013). This modeling pipeline,
or methodology process, is used as a
path for transforming 2D data into
realistic 3D city models.
CityEngine also contains a
long-range list of user-friendly
capabilities. Its ability to be highly
versatile, realistic, and thematic
opens new vistas on how to
collaborate and communicate
complex ‘geospecific’ environments
using a simulated virtual environment
(GeoSpatial Intelligence Forum,
2014). CityEngine is equipped with
Web Scene, which is an interactive
browser that provides the ability to
share 3D city scenes with the general
public. The application is
compatible with all browsers,
computers, and mobile devices.
Other user-friendly CityEngine
capabilities for generating city
models include integrating existing
data sets, such as Census Tiger files,
40
GIS data, Google Earth, OpenStreet
Map, SketchUp, and AutoCAD.
Overall ESRI’s City Engine
software application emulates smart
3D cities using main features that
capture its user-friendly capabilities.
Its procedural modeling and the
modeling pipeline are seen as the
core components that effectively
communicate, and collaborate,
simulated 3D city models.
CityEngine software application and
its capabilities yield numerous
possibilities for all users.
PROCEDURAL RULES VS. OBJECT-
ORIENTED MODELING
CityEngine was chosen as our
3D Visualization program for a
number of reasons. The primary
choice for its utilization was due to its
procedural rule based modeling
ability as opposed to the object
oriented modeling of 3D visualization
programs that are commonly used
today. The team was also able to
draw upon ESRI for technical support
during the creation of the models.
The differences between object
oriented modeling, which programs
such as AutoCAD or Revit do, and
procedural rule modeling, which
CityEngine does, are rather large
and quite important when aspects
of a project such as the relative size
of the project, the level of accuracy
required, and the timeframe in
which a project needs to be
completed are considered.
The best way to understand
object oriented based modeling is to
picture the program as a set of
Lego’s. Blocks of code are written
and combined to form a larger
component; the larger component
being a single building, for instance,
and the small blocks of code
corresponding to individual pieces of
the building that define the physical
geometry and attributes of that
piece. For example, a string of code
written for the front door of a high
rise building in Boston, a piece of
code defining the windows on the
north side of the building, a piece of
code defining the type of material of
the building. Every piece of code is
assigned to each individual object
associated with a single building.
These codes can also house a
plethora of data regarding that
specific object, similar to attributes in
ArcMap attribute tables. For
instance, a code written for the front
door of a building could also contain
information regarding how much the
door costs to be replaced, its
physical dimensions, its color, how
often the hinges need to be
greased, etc. Anything
characteristic pertaining to that
specific door can be contained
within the attributes of that doors
specific code; all of these codes are
joined together to create a 3D
visualization of that building. This
process is important for architects,
engineers, and other users designing
a single building, but is rather costly
and time consuming to create on a
41
large 3D city or college campus
scale.
This is where the concept and
methodology of Procedural Rule
Modeling takes over. Procedural
rules can be thought of as defining
the physics and geometry of not
only a single building, but multiple
buildings at the same time. For
instance, a procedural rule can be
applied to all of the buildings on
FSU's campus based on the height
attribute defined to the specified
polygon. An additional rule can be
applied to all buildings to assign
them a roof, and once rendered, all
buildings will be display the assigned
roof. Once a rule is applied to an
entire set of buildings, a user can
select a single specific building;
adjust its settings, appearance, roof
type, etc. without altering the
attributes of other buildings. This
ability allows for the general mock-
up of a 3D campus to be created
very quickly, with the ability to fine
tune the appearance of individual
buildings as well.
The advantages of procedural
rule based modeling are not limited
to just the geometry of a city but
also the physics as well. Within these
rules that are applied globally to the
entire extent of campus, laws of
physics are also attached. For
example, Legacy Walk, which runs
West to East right through campus,
parallel to Tennessee Street, is
covered with the canopy of Live
Oaks. In certain object-oriented
models, a user would have to
specifically define that the branches
of Live Oaks do not grow into the
sides of buildings. With procedural
rule modeling, however, this rule is
inherent and when a building is
placed in an area that has a tree or
other 3D object, the portions of that
object that occupy the space that
the building now will are erased. This
is an example of how the procedural
rule modeling is constantly adaptive
to change by evaluating and
changing the dimensions
accordingly, making it a ‘living
model.’
COMPARING CITYENGINE TO
SKETCHUP AND REVIT
CityEngine is based off the
platform of ArcGIS and has many of
the analytical capabilities of GIS
software while also combining
visualization capabilities held by
such programs as SketchUp and
Revit. This section will compare the
capabilities and differences of each
program.
Through the use of LiDAR data
and campus landscape maps, the
project team was able to gather
exact building heights and overhead
imagery which was then used to
construct scenes of the FSU campus
and edit the resulting structure
façades to accurately resemble the
buildings. CityEngine also allows, to
a less degree, interior views of
buildings through cutouts. To
improve street scenes, the rule files
42
and geodatabase layers were
added to place people, vehicles,
trees and other elements in the
scenes and edit the appearances of
these elements in groups. By
applying rules to the extruded
buildings from the LiDAR data, this
process was much faster than would
otherwise have been possible. The
scenes can then be used to visualize
the placement of solar panels based
on the results of the solar radiation
analysis.
SketchUp also has the
capability of allowing the import of
LIDAR data and maps as the project
team did with CityEngine. However,
quick 3D renderings in SketchUp are
limited since the program only allows
for object-oriented modeling.
Structures and their components
must be created individually. To
improve scenes, textures can be
applied to façades or components,
such as windows, can be imported.
This same process of placing
individual components must be used
to import other elements such as
people, trees, and streets.
Unfortunately, without attributes,
SketchUp is not able to automate
the scaling of these renderings.
However, SketchUp has a greater
user-friendly interface for object-
oriented modeling. In fact, SketchUp
renderings can be imported into
CityEngine to provide site-specific
visualizations.
Finally, Revit, which is based
on an AutoCAD platform, is very
useful in the construction of 3D
structures with higher measurement
accuracy and interior visualization.
This software also allows higher
architectural detail and can be used
with LiDAR data for exact building
heights. In this way, Revit’s greatest
strength lies in its ability to construct
architectural building perspectives.
Revit’s ability to import people,
furniture and other features enables
the production of accurate and
realistic representations of individual
buildings. However, Revit is limited in
its ability to construct street or city
scenes.
For the above reasons, 3D
rendering on a large scale is most
easily accomplished with the given
resources by using CityEngine. All
three programs allow for 3D
rotational viewing and construction
of individual buildings using different
methods. The ability to integrate the
programs for specific purposes
provides a higher level of
visualization and analysis.
USING CITYENGINE TO COMPLEMENT
SOLAR SUITABILITY ANALYSIS
Much of the work for this
project performed in City Engine was
visualization based in terms of
creating a 3D model that was an
accurate representation of FSU's
campus in present shape, and in an
alternate form, one of which
consisted of photovoltaic solar
panels being placed on optimum
site locations. The majority of the site
43
suitability was determined through
LiDAR analysis; however, certain
buildings were identified
beforehand, based on the known
structural characteristics of a
building.
Namely, the Indoor Practice
Facility (IPF) and parking garages
were identified with visual
interpretations in City Engine as
optimum locations for solar
photovoltaic panels. The IPF was
chosen primarily due to its large
surface area at a slight angle facing
south, while the parking garages
were chosen based off of previous
case studies that found large
radiation returns for parking garages,
such as the ones identified at
Arizona State University. ASU has put
into action large scale photovoltaic
panels as an overhead shade
structure on the tops of parking
garage to absorb solar radiation for
energy production while providing
shade for cars. These areas were
identified as being suitable
candidates for photovoltaic panels,
and the process of creating a 3D
model visualizing how the installation
might look began.
USING CITY ENGINE AS A CAMPUS
PLANNING TOOL
CityEngine is not only useful for
creating a current 3D model of FSU's
campus, but can also be utilized as
a tool for planning potential built
environment changes. CityEngine
enables its users to create relatively
accurate models of the real world
quickly and easily. With the
advancement in technology and
the growing demands of visualizing
the built environment through
graphic programs, CityEngine
provides a bridge between GIS
analysis and 3D visuals.
What has been created for this
project can now be handed to the
planning, facilities, and potential
funding partners for further analysis
of future scenarios. This model gives
planners an accurate, 3D
presentation of FSU's campus but
more importantly, what particular
buildings could potentially look like if
the addition of photovoltaic solar
panels were implemented. A 2D
map highlighting buildings that are
suitable for PV panels is informative,
but a 3D representation of what the
entire campus could look like in the
future has the ability to instill a
common vision and stir excitement
over what could be.
METHODOLOGY
The first step in the process of
creating a three-dimensional model
of FSU's campus was data
acquisition. Streets, building
polygons, trees, a digital aerial
image of campus, and a digital
elevation model were needed.
Tallahassee-Leon County GIS was
more than helpful in providing us
with all of the data that we required
to create this model. After all the
data was acquired, the first step in
44
the process was to create a Digital
Elevation Model from the LiDAR data
set. Given the powerful tools in
ArcGIS, a DEM can be created from
a Digital Terrain Model quite easily. A
tool is ran that subtracts the top of
building elevation from the
surrounding ground elevation, and a
new interpolated DEM is created
that is roughly 5 ft. resolution.
The next step in the process
was to ensure that the aerial image
being used was sufficient. Trouble
was initially run into when the MrSID
files that were given to the project
team from Tallahassee-Leon County
GIS were examined. There was not a
single image that contained FSU's
campus in its entirety. Instead,
because of the planes flight path
during image capturing, the west
half of campus was on one image,
and the east half was on another.
These images were pieced together
into a mosaic to create one large
image, and then a subset was
made, or clipped, for the area of
focus, campus. This new image was
resampled to 3 ft. pixel resolution
and then draped on top of the DEM
to create an aerial image of FSU that
had a fairly accurate representation
of FSU's elevation. The process of
creating the 3D model was now
underway.
Since CityEngine operates
under procedural rules as mentioned
earlier, and since this is a very new
program and these rules appear to
be in somewhat of their own
language, the rules used in this
project were general rules acquired
through ESRI. They contained
modeling options for building
facades, roof types, window options,
tree types, buses, people, and most
importantly, solar panels.
A general rule was applied to
all the buildings on campus and
applied, then a 3D representation of
campus was quickly generated.
Granted, all of the buildings were
grey with no windows and flat roofs,
but they were 3 dimensional. From
here the process of selecting brick
facades, or whichever facade
appeared to match certain
buildings, and changing roof types
from flat to gabled was individually
applied to buildings. Most of the
focus of customizing was spent on
the buildings that were deemed
most feasible for photovoltaic
panels. These buildings actually did
not receive a generic facade.
Instead, images were captured
through Google Earth, and the
actual facades of those buildings
were pasted to them. This drastically
improved the aesthetic of the
model, and although time
consuming, if this project was
extended, the entire campus could
easily have been made with the
actual building facades.
A few buildings in particular
were not constructed with the
general rule set. For instance, Doak
Campbell Stadium was actually a
SketchUp file found on line that was
45
imported into the model as a KMZ
file. The file was georeferenced so
when it was added to our model it
snapped directly into place. Other
buildings were actually constructed
by hand, building them up from an
array of polygons, an example being
the Civic Center. Since the LiDAR
analysis found the south end of the
building to provide strong returns, this
building could not be left generic.
Unfortunately, this building is very
detailed in both the construction of
the walls as well as the multiple
sloped roofs. This building was
created by hand and each tiny
portion of the roofs and walls were
covered with images of the actual
facades.
Once the model was coming
to fruition, the task of solar panels
was next. It was quite simple to
attach solar panels to roofs; the
difficult part was moving them
slightly left, right, forward, or back to
make them give the appearance
that they are sitting properly on the
rooftops. Buildings with flat roofs
were quite easy to fit with PV panels,
whereas buildings with sloped roofs
such as the Chemical Sciences
Laboratory building, which was one
of the strong candidates for PV
panels, proved to be difficult given
its roof’s slope and orientation.
After all buildings were
generated and manipulated to
have the appearance of FSU's
campus, be it general brick with clay
tiled roof, or very detailed such as
the St. Augustine parking garage,
which was hand crafted and
created to have a solar panel
canopy constructed above the cars
parking on the top level, the task of
creating the trees on campus was
one of the final steps. A rule set
acquired again through ESRI
containing hundreds of different tree
types was used. Unfortunately, this
rule set was used for modeling a city
in California so there were no live
oaks, which are the prominent
vegetation on campus. Other trees
were found in the rule set that
resembled live oaks, with their
signature canopy, and they were
applied as well as cabbage palms
to all the trees across campus. This
really gave FSU's 3D campus a life-
like feel, especially in recognizable
areas such as the Legacy Walk in
front of Bellamy, which is lined with
large-scale live oaks. Without that
canopy there, the campus is not
accurately represented.
Other adjustments were made
to the model in terms of creating
future buildings that are going to be
constructed on campus such as the
proposed Earth, Oceanic, and
Atmospheric Sciences building that
will be constructed within this
decade. A rough sketch of what the
building will look like was found, the
polygon was created and then the
proposed building was constructed
exactly where it will be in the future.
This ability proves well for planning
and visual aid. It was also used to
create new bus stops on campus in
46
attempt at redesigning FSU and
Tallahassee's bus routes.
RESULTS
At the end of the project, the
project team yielded a complete
model of campus in 3D. All of the
buildings on campus were extruded
to scale based on how many floors
each building had recorded in their
attribute tables as part of the
geodatabase the building footprints
originated from. In addition to a
complete model with buildings to
scale, some, buildings on campus
were given realistic facades. For
example, the distinctive windows of
the Bellamy Building and the well-
known brick facades that
characterize most of the buildings on
campus were successfully rendered
in the model. On top of a
completed model of buildings to
scale with realistic facades, trees
and streets were also captured in
the model.
Finally, the completed 3D
model of campus also included PV
panels on the rooftops of many
structures, including parking
garages, which are good
candidates for harnessing solar
energy. Most of the visualization
process focused primarily on the
buildings highlighted in the LiDAR
analysis. For instance, the parking
structure on St. Augustine, which
serves as a large area capable of
capturing ample amounts of solar
radiation thru PV panels that are
suspended above the cars that are
parking on the top level. Another
building that was found in the LiDAR
analysis that could provide a lot of
solar power is the new Indoor
Practice Facility. Located
perpendicular to the intersection of
Pensacola St. and Stadium Drive, this
FIGURE 27: ST. AUGUSTINE ST. GARAGE
47
newly constructed building has a
very large, south facing roof with a
gentle slope.
VISUALIZATION AS A PLANNING TOOL
Visualization is a great tool for
planning because it provides the
ability to visualize how a plan will
appear upon completion and the
design impact it may have in an
area. By visualizing a plan, urban
design standards can be reviewed
to ensure compliance. Additionally,
people are naturally visual and by
producing visualizations for a
planning proposal, it can improve
the ability of planners to engage
with stakeholders who may not
otherwise understand it.
Visualizations were used in this
study to examine the impact solar
panels will have on buildings located
on the FSU campus. This includes a
review of the impact solar panels will
have on historic structures with clay
and asphalt tile roofs. New
construction, such as that proposed
for the Earth Oceanic and Science
Building, were also visualized through
the rough construction of
representative buildings in the
location. By doing this, the project
team was able to study the solar
exposure buildings would have as
well.
Additionally, the project team
created solar exposure visualization
maps from our LiDAR data to show
the levels of exposure areas of the
FUS campus experience. By color-
coding these maps, the project
team hope to allow the general
public to understand solar exposure
of each building on campus with
elevations incorporated for building
representation. Our ultimate findings
are then presented through overlays
of solar data onto CityEngine models
the project team created to
determine our proposed solar panel
installations.
FIGURE 28: INDOOR PRACTICE FACILITY
48
RECOMMENDATIONS
How Can CityEngine be Utilized
Moving Forward?
While the 3D model of campus
appears to be complete, there is still
plenty of work left for future courses
and cohorts within GIS and Urban
and Regional Planning. Future
courses should consider creating
more realistic buildings on campus
as well as using CityEngine to model
future development scenarios. While
certain polygons, like Bellamy
Building and Doak Campbell
Stadium, were broken down into
several polygons to create more
accurate representations of those
buildings, not every building on
campus that required this treatment
received it. For example, the
Westcott Building, which is famous
for its gothic-style turrets in the front,
was not custom-rendered. Future
courses should also consider
capturing art on campus in 3D.
Large statues, such as The
Integration Statues, are large
enough that they could be easily
created in CityEngine. In addition to
improvements and additions to our
course’s model of campus, future
courses should consider utilizing
CityEngine to portray how different
planning decisions, such as installing
a metro rail system, would affect
campus.
49
7.0 CONCLUSION
Utilizing GIS for the solar radiation analysis provided an understanding of
the potential for the implementation of solar energy technology at Florida State.
Based on the site characteristics of height, slope, and aspect, a number of
buildings exhibited high radiation returns, indicating a strong potential for
utilizing photovoltaic cells for solar energy production. The use of the CityEngine
software then allowed relatively quick three-dimensional modeling of how these
buildings might look with solar arrays. The combination of these tools provides
quantitative and qualitative results for the next step in the feasibility process. The
information is meant to provide future researchers, policy makers, and the
general public an understanding of where solar energy production is a viable
option while allowing those individuals to visualize its implementation.
This research was intended to provide the beginning steps in the
discussion. Other steps are needed in order to come to a final conclusion on
whether the construction of photovoltaic cells on the chosen buildings is
feasible, including analyses of both the fiscal and engineering needs. Lastly, it is
important to mention that this paper’s research has only breached the surface
of the LiDAR analysis and CityEngine’s potential. Future work has the potential to
create an even more seamless bridge between the two tools, yielding stronger
solar energy analysis and higher quality 3D renderings.
50
REFERENCES
AASHE. (2012). Top Ten Solar Photovoltaic
Installations Lists: Top Ten Campuses by
Total Installed Capacity*. Retrieved from
http://www.aashe.org/resources/campus-
solar-photovoltaic-installations/top10/#top-
capacity
Arizona State University. (2012). ASU solar
program shines, tops 14.5 megawatts. ASU
News [university]. Retrieved from
https://asunews.asu.edu/20120213_ASU_s
olar
Arizona State University. (2014a). ASU Solar.
Retrieved from https://cfo.asu.edu/solar
Arizona State University. (2014b). ASU Solar:
Financing & Incentives. Retrieved from
https://cfo.asu.edu/solar-financing
Arizona State University. (2014c). ASU Solar:
Solar FAQs. Retrieved from
https://cfo.asu.edu/solar-faq
Arizona State University. (2014d). Solar System
Installations: Tempe Campus. Retrieved
from
https://www.asu.edu/fm/images/solarizatio
n/solar-map-tempe.pdf
Chaves, A., & Bahil, A. T. (2010). Locating
Sites of Photovoltaic Solar Panels.
ArcUser, 24–27. Retrieved from
http://www.esri.com/news/arcuser/1010/fal
l2010.html
Esri. (2013). Advantages of using lidar in GIS.
ArcGIS Help 10.1. Retrieved from
http://resources.arcgis.com/en/help/main/1
0.1/index.html#//015w0000004w000000
Esri R&D Center Zurich. (2013). CityEngine:
Basics of Rule-based Modeling. Retrieved
from
http://cehelp.esri.com/help/index.jsp?topic
=/com.procedural.cityengine.help/html/ma
nual/cga/basics/toc.html
Florida Gulf Coast University. (n.d.). FGCU
Solar Field Goes Online. Retrieved from
http://www.floridaenergy.ufl.edu/wp-
content/uploads/FGCU-Solar-Field-Goes-
Online.pdf
Florida Gulf Coast University. (2010a). FGCU
goes solar! Pinnacle Magazine. Retrieved
from
http://www.fgcupinnacle.com/pinnacle-
archives/pinnacle-winter-2010
Florida Gulf Coast University. (2010b). Going
green. Pinnacle Magazine. Retrieved from
http://www.fgcupinnacle.com/pinnacle-
archives/pinnacle-spring-2010
Florida Gulf Coast University. (2014). Facilities
Planning: Solar Field. Retrieved from
http://www.fgcu.edu/Facilities/SolarField.h
tml
Florida Solar Energy Center. (2007). Florida
Solar Energy Center. Retrieved from
http://www.fsec.ucf.edu/en/about/index.ht
m
Florida State University. (n.d.). Alternative
Energy Sources. Retrieved from
https://www.facilities.fsu.edu/cup/altenerg
y.html
Florida State University Facility Program.
(2013). The Florida State University
Facility Program for Earth Ocean and
Atmospheric Sciences ( EOAS ). Retrieved
from
https://www.facilities.fsu.edu/FDC/Adverti
sements/EOAS-program.pdf
fsunews.com. (2011). FSU is among the most
energy efficient universities in Florida.
Retrieved from
http://www.fsunews.com/article/20110421/
FSVIEW/110420012/FSU-among-most-
energy-efficient-universities-Florida
51
GeoSpatial Intelligence Forum. (2014). Game
On For GeoInt. Retrieved from
http://www.kmimediagroup.com/geospatial
-intelligence-forum/424-articles-gif/game-
on-for-geoint/5633-game-on-for-geoint
IDEAS. (2011). UCF. Retrieved from
http://www.ideasforus.org/tag/ucf/
Kodysh, J. B., Omitaomu, O. a., Bhaduri, B. L.,
& Neish, B. S. (2013). Methodology for
estimating solar potential on multiple
building rooftops for photovoltaic systems.
Sustainable Cities and Society, 8, 31–41.
doi:10.1016/j.scs.2013.01.002
Kucuksari, S., Khaleghi, A. M., Hamidi, M.,
Zhang, Y., Szidarovszky, F., Bayraksan,
G., & Son, Y.-J. (2014). An Integrated
GIS, optimization and simulation
framework for optimal PV size and
location in campus area environments.
Applied Energy, 113, 1601–1613.
doi:10.1016/j.apenergy.2013.09.002
Lopez, A., Roberts, B., Heimiller, D., Blair, N.,
& Porro, G. (2012). U . S . Renewable
Energy Technical Potentials : A GIS-Based
Analysis, (July). Retrieved from
http://www.nrel.gov/docs/fy14osti/60593.p
df
Melius, J., Margolis, R., & Ong, S. (2013).
Estimating Rooftop Suitability for PV : A
Review of Methods , Patents , and
Validation Techniques Estimating Rooftop
Suitability for PV : A Review of Methods ,
Patents , and Validation Techniques,
(December). Retrieved from
http://www.nrel.gov/docs/fy14osti/60593.p
df
Morley, D. (2014). Planning for Solar Energy.
Retrieved from
https://www.planning.org/pas/reports/subsc
riber/archive/pdf/PAS_575.pdf
Mount St. Mary’s University. (2011). Solar
Energy at the Mount. Retrieved from
https://www.msmary.edu/about-the-
mount/sustainability/Solar-energy-at-
Mount.html
NREL. (2008). Photovoltaic Solar Resource of
the United States. Retrieved from
http://www.nrel.gov/gis/images/map_pv_n
ational_lo-res.jpg
NREL. (2014). PVWatts Calculator. Retrieved
from http://pvwatts.nrel.gov/pvwatts.php
Rutgers Climate Institute. (2014). Adressing
Climate Change in the Rutgers
Community. Retrieved from
http://climatechange.rutgers.edu/rutgers-
climate-stewardship
Santos, T., Gomes, N., Brito, M., Freire, S.,
Fonseca, A., & António, J. (2011). Solar
Potential Analysis in Lisbon Using LiDAR
Data, 13–19. Retrieved from
http://www.earsel.org/symposia/2011-
symposium-
Prague/Proceedings/PDF/Coastal Zones/3
ok25-a2393-Santos_et_al.pdf
Solair Generation. (2013). Rutgers University.
Retrieved from
http://solairegeneration.com/project/rutgers
-university/
Solar World. (2012). University of Central
Florida Knights get SunSense. Retrieved
from http://www.solarworld-
usa.com/blog/2012/may/university-central-
florida.aspx
Sustainable UF. (2014a). Please Select a
Renewable Energy Production Site.
Retrieved from
http://building.ufl.edu/kiosk/site.php
Sustainable UF. (2014b). Renewable Energy.
Retrieved from
http://sustainable.ufl.edu/topics/energy-
climate-change-at-uf/renewable-energy/
U.S. Department of Energy. (n.d.). Solar.
Retrieved from http://energy.gov/science-
52
innovation/energy-sources/renewable-
energy/solar
United States Environmental Protection Agency.
(2014). Solar Purchase Power Agreements.
Retrieved from
http://www.epa.gov/greenpower/buygp/sol
arpower.htm
University of Central Florida. (n.d.).
Sustainability & Energy Management:
Engineering I Building. Retrieved from
http://sustainable.ucf.edu/engineering-i-
building
University of Central Florida. (2012). UCF,
Progress Energy Florida Dedicate Energy-
Saving Solar Panel System. Retrieved from
http://today.ucf.edu/ucf-progress-energy-
florida-dedicate-energy-saving-solar-panel-
system/
DEPARTMENT OF URBAN & REGIONAL PLANNING
DEPARTMENT OF GEOGRAPHY
AT FLORIDA STATE UNIVERSITY
BELLAMY BUILDING, FLOOR 3
127 HONORS WAY
TALLAHASSEE, FL 32301
Related Documents
