An Analysis of Photovoltaic Performance and its Relationship to Mounting Height in Buildings

MISSING IMAGE

Material Information

Title:
An Analysis of Photovoltaic Performance and its Relationship to Mounting Height in Buildings
Physical Description:
1 online resource (70 p.)
Language:
english
Creator:
Wilhoit,Randall R,Jr
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Master's ( M.S.B.C.)
Degree Grantor:
University of Florida
Degree Disciplines:
Building Construction
Committee Chair:
Sullivan, James
Committee Members:
Srinivasan, Ravi
Lucas, Elmer

Subjects

Subjects / Keywords:
mounting -- output -- performance -- photovoltaic -- pv -- renewable -- solar -- temperature
Building Construction -- Dissertations, Academic -- UF
Genre:
Building Construction thesis, M.S.B.C.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
Stationary, roof-mounted residential photovoltaic arrays are typically installed at a height of three to four inches from the roof surface. Anecdotal evidence has found this to be problematic from a maintenance standpoint, as leaves and other debris tend to accumulate under the modules over time and are difficult to remove when the panels are so close to the roof surface. Solar Impact, a solar contractor in the Gainesville, Florida area has found that by mounting the panels at a height of six to eight inches above the roof surface, they are much easier to maintain. The purpose of the study was to determine if there were any measureable benefits, specifically power or temperature associated with mounting the modules at the increased height. The study was conducted between 11 and 30 May, 2011 on an existing photovoltaic system in Gainesville, Florida. The system was divided into two identical arrays. The first array remained at four inches and the second array elevated to approximately eight inches from the roof surface. Power output was recorded from each module within the arrays, as well at one temperature reading per array. The samples were taken at one minute intervals from 12:00PM to 3:30PM every day resulting in 211 samples per day, or 4220 samples over the length of the study. The array mounted at four inches had an average power output per panel of 110.82 Watts with a standard deviation of 38.84 Watts, and an average temperature of 62.70 degrees Celsius with a standard deviation of 10.78 degrees Celsius. The array mounted at eight inches had an average power output per panel of 113.16 Watts with a standard deviation of 39.74 Watts, and an average temperature of 54.07 degrees Celsius with a standard deviation of 9.14 degrees Celsius. The results were that the eight inch array had an average of 2.43% more power output per panel and a 15.77% lower average temperature than the four inch array.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Randall R Wilhoit.
Thesis:
Thesis (M.S.B.C.)--University of Florida, 2011.
Local:
Adviser: Sullivan, James.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2011
System ID:
UFE0043425:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 AN ANALYSIS OF PHOTOVOLTAIC PERFORMANCE AND ITS RELATION TO MOUNTING HEIGHT IN BUILDINGS By RANDALL RICHMOND WILHOIT, JR. A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE IN BUILDING CONSTRUCTION UNIVERSITY OF FLORIDA 2011

PAGE 2

2 2011 Randall Richmond Wilhoit, Jr.

PAGE 3

3 To everyone who believes that we can still fix this thing

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank my wonderful wife, Amber, for tolerating me all these years. I would like to thank Barry and Elaine Jacobson, for giving me a future, and without whom this study wou ld not have been possible. Lastly, I would like to thank God for making the sun, which is pretty cool.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ................... 10 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 12 Statement of Purpose ................................ ................................ ............................. 12 Objective of the Study ................................ ................................ ............................. 12 2 LITE RATURE REVIEW ................................ ................................ .......................... 14 Photovoltaic Panels ................................ ................................ ................................ 14 History ................................ ................................ ................................ .............. 14 Panel Composition ................................ ................................ ........................... 14 Installati on Techniques ................................ ................................ ..................... 15 PV Performance ................................ ................................ ................................ ..... 16 Theoretical Solar Limitations ................................ ................................ ............ 16 Limiting Factors ................................ ................................ ................................ 17 Available light ................................ ................................ ............................. 17 Weather ................................ ................................ ................................ ..... 17 Temperature ................................ ................................ .............................. 17 Orientation ................................ ................................ ................................ 18 Shading ................................ ................................ ................................ ...... 18 Spacing and mounting height ................................ ................................ ..... 19 3 RESEARCH METHODOLOGY ................................ ................................ ............... 26 Location ................................ ................................ ................................ .................. 26 Site Description ................................ ................................ ................................ ....... 26 Array Descri ption ................................ ................................ ................................ .... 26 Study Period ................................ ................................ ................................ ........... 27 Sampling ................................ ................................ ................................ ................. 27 4 RESULTS ................................ ................................ ................................ ............... 38 5 CONCLUSIONS ................................ ................................ ................................ ..... 66

PAGE 6

6 6 RECOMMENDATIONS ................................ ................................ ........................... 67 LIST OF REFERENCES ................................ ................................ ............................... 68 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 70

PAGE 7

7 LIST OF TABLES Table page 4 1 Results of Analysis ................................ ................................ ............................. 45

PAGE 8

8 LIST OF FIGURES Figure page 2 1 Thin Film Photovoltaic Array ................................ ................................ ............... 21 2 2 Crystaline Photovoltaic Modules ................................ ................................ ......... 22 2 3 Single Axis Pole Mounted Array ................................ ................................ ......... 23 2 4 Stationary Photovoltaic Array ................................ ................................ ............. 23 2 5 Photovoltaic Module Exposure at Different Angles of Incidence ......................... 24 2 6 ................................ ................................ .............................. 25 3 1 Google Earth Imagery of Site. ................................ ................................ ............ 29 3 2 Solar Pathfinder Imagery. ................................ ................................ ................... 29 3 3 Sample Monitoring During April 2011. ................................ ................................ 30 3 4 Site Photos ................................ ................................ ................................ ........ 31 3 5 Unir ac Mounting System ................................ ................................ .................... 32 3 6 Photovoltaic Array Layout. ................................ ................................ .................. 32 3 7 ................................ .......................... 33 3 8 Extensions ................................ ................................ ................................ .......... 34 3 9 Site S hading ................................ ................................ ................................ ....... 35 3 10 Tigo Equipmen t ................................ ................................ ................................ .. 36 3 11 Tigo Energy Site Interface ................................ ................................ .................. 36 3 12 Output Demonstrating Partial Shading. ................................ .............................. 37 4 1 May 11, 2011 Power and Temperature ................................ ........................... 46 4 2 May 12, 2011 Power and Temperature ................................ ........................... 47 4 3 May 13, 2011 Power and Temperature ................................ ........................... 48 4 4 May 14, 2011 Power and Temperature ................................ ........................... 49 4 5 May 15, 2011 Power and Temperature ................................ ........................... 50

PAGE 9

9 4 6 May 16, 2011 Power and Temperature ................................ ........................... 51 4 7 May 17, 2011 Power and Temperature ................................ ........................... 52 4 8 May 18, 2011 Power and Temperature ................................ ........................... 53 4 9 May 19, 2011 Power and Temperature ................................ ........................... 54 4 10 May 20, 2011 Power and Temperature ................................ ........................... 55 4 11 May 21, 2011 Power and Temperature ................................ ........................... 56 4 12 May 22, 2011 Power and Temperature ................................ ........................... 57 4 13 May 23, 2011 Power and Temperature ................................ ........................... 58 4 14 May 24, 2011 Power and Temperature ................................ ........................... 59 4 15 May 25, 2011 Power and Temp erature ................................ ........................... 60 4 16 May 26, 2011 Power and Temperature ................................ ........................... 61 4 17 May 27, 2011 Power and Temperature ................................ ........................... 62 4 18 May 28, 2011 Power and Temperature ................................ ........................... 63 4 19 May 29, 2011 Power and Temperature ................................ ........................... 64 4 20 May 30, 2011 Power and Temperature ................................ ........................... 65

PAGE 10

10 Abstract of Thesis Presented to the Graduate School of the University o f Florida in Partial Fulfillment of the Requirements for the Degree o f Master of Science in Building Construction AN ANALYSIS OF PHOTOVOLTAIC PERFORMANCE AND ITS RELATION TO MOUNTING HEIGHT IN BUILDINGS By Randall R. Wilhoit, Jr. August 2011 Chair: James Sullivan Major: Building Construction Stationary, roof mounted residential photovoltaic arrays are typically installed at a height of three to four inches from the roof surface. Anecdotal evidence has found this to be problematic from a maintenance standpoint, as leaves and other debris tend to accumulate under the modules over time and are difficult to remove when the panels are so close to the roof surface. Solar Impact, a solar contractor in the Gainesville, Florida area has found that by mounting the panels at a height of six to eight inches above the roof surface, they are much easier to maintain. The purpose of the study was to determine if there were any measureable benefits, specifically power or temperature associated with mounting the modules at the increased height. The study was conducted between 11 and 30 May, 2011 on an exist ing photovoltaic system in Gainesville, Florida. The system was divided into two identical arrays. The first array remained at four inches and the second array elevated to approximately eight inches from the roof surface. Power output was recorded from e ach module within the arrays, as well at one temperature reading per array. The

PAGE 11

11 samples were taken at one minute intervals from 12:00PM to 3:30PM every day resulting in 211 samples per day, or 4220 samples over the length of the study. The array mounted a t four inches had an average power output per panel of 110.82 Watts with a standard deviation of 38.84 Watts, and an average temperature of 62.70 degrees Celsius with a standard deviation of 10.78 degrees Celsius. The array mounted at eight inches had an a verage power output per panel of 113.16 Watts with a standard deviation of 39.74 Watts, and an average temperature of 54.07 degrees Celsius with a standard deviation of 9.14 degrees Celsius. The results were that the eight inch array had an average of 2.43 % more power output per panel and a 15.77% lower average temperature than the four inch array.

PAGE 12

12 CHAPTER 1 INTRODUCTION Statement of Purpose The guidelines for photovoltaic (PV) module installations recommend a mou nting height of three to four inches from the roof surface when mechanically attaching PV arrays to a sloped roof. However, anecdotal evidence has found this to be problematic from a periodic maintenance perspective. Leaves from surrounding trees and othe r debris tend to accumulate under the modules, much like they do in roof gutter systems. When the modules are mounted at four inches from the roof surface, there is approximately two inches between the bottom of the mounting rails and the roof surface. B ecause of this small space, the debris can be difficult to remove. It is important to keep the area under the modules clear because the debris restricts airflow, which in turn increases the module temperature. This increase in module temperature correspo nds to a decrease in module power output. Solar Impact, Inc., a solar contractor in the Gainesville area, has found that by mounting the modules above the recommended height at six to eight inches, it is much easier to access the area under the modules an d kee p it clear of debris. What wa s not known is if the increase in mounting height alone actually decreases module temperature and/or increases the module performance. Th is stud y sought to determine if there was any performance advantage in mounting PV mo dules at six to eight inches from the roof surface as opposed to three to four inches. Objective of the Study The objective of this research was to determine if there was a performance advantage to mounting the PV modules at eight inches instead of four inches.

PAGE 13

13 The specific task s consisted of comparing the average power output of two PV arrays over 20 days, comparing the temperature of the arrays over 20 days, and determin ing i f a performance advantage existed at the increased mounting height

PAGE 14

14 CHAPTER 2 LITERATURE REVIEW Photovoltaic Panels Histor y The birth of photovoltaics dates back to 1839 when French physicist Edmund Becquerel discovered that a pair of electrod es increased their electron emissions when they were placed in a conductive solution and exposed to sunlight (Dunlop, 2007). However, it was not until 1954 that the first useable PV cells were developed in the U.S. by Bell Laboratories (Dunlop, 2007). Wi th the oil crisis of the 1970 s, the U.S. government became increasingly interested in the use of PV, and began to implement tax credits and incentives for its increased use. These incentives, along with refinements in PV efficiency and manufacturing have r esulted in the PV systems that are seen today. Panel Composition PV panels (or modules) come in a variety of sizes and can be broken down into two main types: thin film and crystalline. The different types of modules have different efficiency, or ability t o capture available light and convert it to electricity. Regardless of type, both will produce an electric current (DC) when exposed to a light source. Additionally, both types must be connected to a power inverter in order to produce AC power that can be used directly or connected to the local electrical grid. Thin film modules are produced by combining a silicone film or another material with semiconducting properties to a highly flexible base. Because of its flexibility, these thin film modules can be used in a variety of construction applications, from PV roof tiles to a kind of rolled veneer that can be applied directly to the roof surface, and example of

PAGE 15

15 which can be seen in Figure 2 1 (Green is Global, 2009). Of the two types of PV, thin film is th e least efficient (Holladay, 2009). Crystalline PV modules, which can further be broken down into mono crystalline and poly crystalline, are produced by taking a thin layer of crystalline silicon and pressing it between two layers of glass and polymer resu lting in a kind of crystalline silicon sandwich, examples of which can be seen in Figure 2 2. Crystalline PV modules are more common and more efficient than thin film modules, with the best Crystalline PV modules capable of capturing approximately 20% of available light, meaning that of the direct light hitting the module surface, the module is only able to capture 20% of it, with the remaining 80% being reflected. However, emerging technologies may push module efficiency as high as 30 40% (Deb, 2000). Th is study focused on mono crystalline PV modules (easily recognized by their mottled composition) in panel configuration. Installation Techniques The most common form of PV panels are flat plate collectors which are essentially flat surfaces covered with PV cells that can capture sunlight directly from the sun or diffused by the atmosphere or clouds and reflected off of other surfaces. Newer technologies are emerging that use lenses to concentrate greater areas of available light onto smaller PV surfaces, b ut they are beyond the scope of this study. PV panels can be installed in a solar tracking or stationary configuration. Solar tracking can further be broken down into dual and single axis systems. Dual axis systems, most commonly pole mounted, are the mo st efficient in terms of capturing available sunlight and can increase energy production by up to 40% (Dunlop, 2007). Dual axis systems track the sun as it moves through the sky over the course of a day as

PAGE 16

16 well as adjust the angle of incidence over the co urse of the year. Single axis systems only follow the sun throughout the day and must be adjusted throughout the year to achieve optimal efficiency An example of a single axis pole mounted system can be seen in Figure 2 3. Stationary systems are either roof or other surface. Stationary systems are not as efficient as solar tracking systems, but have the advantage of no moving parts resulting in decreased maintenance. An added benefit of roof mounted sy stems is the reduction in cooling load due to the partial shading of the roof surface (Malkawi et al 2005.) An example of a stationary system is shown in Figure 2 4. PV Performance Theoretical Solar Limitations phere, only a portion of that light can be passes through it without interference. Direct radiation accounts for roughly 80 90% of available light, even under clear s kies. Diffuse radiation is the remaining 10 20% of can absorb both direct and diffuse radiation, but only a portion of diffuse radiation actually reaches the ground. It is also important to note that the proportion of diffuse radiation can increase to 100% on cloudy or rainy days. Of the light that does actually to 1100 nanometers (Dunlop, 2007). Of this light range, even the best panels only

PAGE 17

17 Limiting Factors Limiting factors affecting PV performance can be broken down into uncontrolled and controlled factors. Availab le light One uncontrolled factor is the amount of available light striking the PV surface. day. Solar insolation varies by geographical region, with latitudes n earer to the equator typically having higher solar insolation than those at higher latitudes. However, this is not uniform, with nearly all of Florida receiving an average of four to four and a half peak hours of sunlight per day while parts of Nevada and Southern California, while at a higher latitude, receive seven to seven and a half peak hours of sunlight per day (Dunlop, 2007). Weather Weather is another uncontrolled factor. Weather patterns affect the amount of available light. Areas with a higher level of precipitation (and associated cloud cover) will have a corresponding decrease in available light. Florida, while commonly referred 53.49 inches per year resulting in decreased available light compared to a more arid region like Nevada, which has an average annual rainfall of 9.5 inches (U.S. Department of the Interior, 2005). Temperature Another uncontrolled factor is the ambient tem perature in which PV panels operate affects their efficiency. PV panels operating in high temperature environments typically have dramatically reduced output (Kim et a l 2011). Additionally, these high operating

PAGE 18

18 temperatures cause the individual PV cells to degrade at an accelerated rate, further reducing panel efficiency (Dunlop, 2007). Orientation One factor that can be controlled is the angle and orientation at which the PV panel array is mounted. A good rule of thumb for installations in the Nor thern Hemisphere is to orient the PV panel array due south with an angle of incidence equal to the sites latitude (Dunlop, 2007). An example of how module orientation can affect the amount of module face exposure can be seen in Figure 2 5. However, this is not the most important factor as a panel optimized at 30 degrees will only lose 4.3% of its maximum output when oriented 45 degrees east of south (Hussein et al 2003). Shading Shading is another factor that can be controlled. Just 10% shading of PV p anels can result in a 90% reduction in output (Ubisse, 2009). This means that should 10% of Shading can be soft or hard. Hard shading is the result of debris or staining on the PV panels and can generally be avoided through maintenance. Soft shading is the result of a structure or object casting a shadow on the PV panel. Both can seriously degrade PV panel performance (Lamont et al 2010). It is extremely important that thorough shading analysis be conducted prior to installing any PV system. While no shading is ideal, it is often unavoidable due the long shadows cast by object in the early morning and late evening hours. It is recommended that PV panels be arranged to avoid shading of any kind between 9:00 AM and 6:00 PM to access the maximum amount of sunlight when the sun is highest and at its most intense. In certain situations where some shading is unavoidable, these systems can be optimized for the time of year as

PAGE 19

19 well as the time of day. Sometimes it is more advantageous to optimize a system for total production, rather than average production throughout the year. For example, in ity pays the customer a premium for the electricity produced from customer owned roof mounted PV systems. Therefore, because the electricity produces has a monetary value associated with it, systems should be optimized for the summer months when the days are longer and the sun is higher in the sky resulting in the greatest output, with less focus on the winter months when the sun is lower in the sky and the days are shorter, resulting in lower output regardless of optimization (Jacobson, 2011). Shading an alysis can be performed using the profile angle method, or more commonly using a sun path calculator Figure 2 6). Spacing and mounting height The spacing and mounting height of the PV panels also affect their performance. To det ermine the rise in PV cell temperature above the surrounding air temperature, the temperature rise coefficient is used. To determine PV cell temperature using the temperature ( C) = Ambient Temperature ( C) + {Temperature Rise Coefficient ( C/kW/m^2) x Solar Irradiance (kw/m^2)} (Dunlop, 2007). For example, if a module has a temperature rise coefficient of .5% at given temperature and solar insolation, then should the temperature in crease 20 degrees Celsius and the solar insolation remain the same, power production will decrease 10%. PV p anels that are installed with their back surfaces facing the wind have a temperature rise coefficient of 5 15 C lower than PV panels mounted on or close to another surface (Dunlop, 2007). With increased spacing and mounting height comes increased ventilation, which reduces the temperature rise

PAGE 20

20 coefficient, which increases PV panel performance. A general guideline for panel installation is three to f our inches off of the roof surface (Holladay, 2009).

PAGE 21

21 Figure 2 1. Thin Film Photovoltaic Array Photo courtesy of greenisglobal.net.

PAGE 22

22 A B Figure 2 2. Crystaline Photovoltaic Modules A) Mono Crystaline B) Poly Crystaline Photos by a uthor.

PAGE 23

23 Figure 2 3. Single Axis Pole Mounted Array Photo by a uthor. Figure 2 4. Stationary Photovoltaic Array Photo by author.

PAGE 24

24 Figure 2 5. Photovoltaic Module Exposure at Different Angles of Incidence A) 90 Degrees B) 75 Degrees C) 60 Degrees D) 45 Degrees Figure by author created with Google Sketchup

PAGE 25

25 A B Figure 2 B) View of Face Photos by author.

PAGE 26

26 CHAPTER 3 RESEARCH METHODOLOGY Location The study was conducted at a home in northwest Gainesville Florida (Figure 3 1). The location of the study was the residence of Barry and Elaine Jacobson, the president and vice president (respectively) of Solar Impact, Inc., a solar contracting firm in the north central Florida region. The site was selected because the PV system, to include all of the monitoring equipment was in place and functioning before the study took place. Site Desc ription The PV arrays were mounted on the southern facing roof surface of the structure. The roof surface was conventional asphalt shingle and was dark brown in color. It had a 3:12 roof pitch. There were existing trees to east and west of the site, some o f which provided partial shading throughout the day. After conducting onsite analysis with a Solar 2), it was determined that the roof surface was partially shaded before 11:00 AM and after 4:00 PM. From previous monitoring, the roof s urface temperature had an average temperature of 65 degrees Celsius in the month preceding the study (Figure 3 3). There was a solar hot water system located directly above the 2 PV arrays. It did not shade either array, and did not affect their operation in any way. Array Description The modules used in the study were Sharp ND 167U1 which were rated at 167 watts. They had been installed and functioning for approximately 4 years. There were 30 total modules ( Figure 3 4)

PAGE 27

27 The arrays were mechanically attach mounting system (Figure 3 5). There was approximately one inch of space between modules in the same row and approximately eight inches of space between rows. The space between the arrays was roughly three feet. The m odules were divided into two arrays, Array 1 and Array 2. To better understand the physical layout and the shading effects of the surrounding trees, a 3D model of the site was developed using Google Sketchup (Figure 3 6). Each array consisted of 15 module s and was arranged into 3 rows of 5. Array 1 was furthest west. Both arrays were mounted at four inches from the roof surface prior to the study (Figure 3 7). In order to raise Array 2 to the desired height, a custom aluminum extension was fabricated and installed, allowing Array 2 to be raised to a mounted height of seven and seven eighths inches (Figure 3 8). Study Period The study will take place during the early summer of 2011, specifically between May 11 th and May 30 th Because of shading issues from the trees to the east and west (Figure 3 9), the samples were taken between 12:00 PM and 3:30 PM daily to avoid temperature differences due to sun exposure. Sampling ipment (Figure 3 10). A a ximizer ES was attached to the back of every module, which tran ateway CG Two modules were additionally equipped with emperature Sensor TS ES that transmitted temperature data to the module gatew ay. The gateway wirelessly transmitted the power

PAGE 28

28 Tigo Energy website. Sampling was set at one per minute, which translated to 211 samples per day. The data was stored onli ne at www.tigoenergy.com (Tigo, 2011). The site utilized a graphical interface (Figure 3 11) which allowed the researcher to collect data remotely at any time. By using the Tigo site (Tigo, 2011), it was determined that partial shading existed on Array 2 before 11:00 AM and on Array 1 after 4:00 PM (Figure 3 12), hence the sampling period of 12:00PM to 3:30PM. The data was downloaded daily to a Microsoft Excel spreadsheet. The data contained the power output for each panel in Watts, as well as the tempera ture recorded by the Temperature Sensors in degrees Celsius. Occasionally individual modules would fail to record data at a given moment. To account for these gaps in the data, the arrays were analyzed by finding the average of the recorded outputs for e ach array. These averages were then plotted on a graph. The differences between the two averages were calculated as well as the percent difference. Because only one temperature sample was recorded per array per minute, no averaging was necessary. Howeve r, because the occasional gaps in data occurred for the temperature samples as well, these gaps in the daily temperature data are displayed as actual gaps, as shown in the data plots in the following chapter.

PAGE 29

29 Figure 3 1. Google Earth Imagery of S ite Photo court esy of Google Earth. Figure 3 2. Solar Pathfinder Imagery Photo by author.

PAGE 30

30 Figure 3 3. Sample Monitoring During April 2011

PAGE 31

31 A B C Figure 3 4. Site Photos A) Southeast View B) Northwest View C) Southwest View Photos by author.

PAGE 32

32 Figure 3 5. Unirac Mounting System Photo by author. Figure 3 6. Photovoltaic Array Layout Image by author created with Google Sketchup.

PAGE 33

33 Figure 3 Photo by author.

PAGE 34

34 A B Figure 3 8. Extensions A) Aluminum Extension B) Extension Installed Photos by author.

PAGE 35

35 A B Figure 3 9. Site Shading A) East Trees B) West Trees Photos by author.

PAGE 36

36 A B C Figure 3 10 Tigo Equipment A) M odule Gateway B) Module Maximizer C) Temperature Sensor D) Tigo Information Flow Photos by author. Figure 3 11. Tigo Energy Site Interface Photo courtesy of tigoenergy.com

PAGE 37

37 A B Figure 3 12. Output Demonstrating Partial Shading A) 11:00 AM B) 4:00 PM Photos courtesy of tigoenergy.com.

PAGE 38

38 CHAPTER 4 RESULTS 4220 samples were taken in total. Array 1 had an average power output per panel over the 20 day period of 110.82 Watts with a standard deviation of 38.84 Watts. The average temperature of Array 1 was 62.70 degrees Celsius with a standard deviation of 10.78 degrees Celsius. Array 2 had an average power output per panel over the 20 day period of 113.16 Watts with a standard deviation of 39.74 Watts. The average temperature of Array 1 was 54.07 degrees Celsius with a standard deviation of 9.14 degrees Celsius. A comparison of performance before and after Array 2 was elevated can be seen in Table 4 1. A brief discussion of the individual data plots for each day follows. May 11, 2011 ( Figure 4 1): Scattered clouds throughout the afternoon resulted in sporadic partial shading on both arrays throughout the sampling period. Array 1 had average power outputs ranging from a low of 79.82 Watts to a high of 135.62 Watts. Array 2 had average power outputs ranging f rom a low of 78.18 Watts to a high of 136.70 Watts. Array 1 operated in panel temperatures between a low of 57.35 degrees Celsius to a high of 71.94 degrees Celsius. Array 2 operated in panel temperatures between a low of 49.13 degrees Celsius to a high of 63.27 degrees Celsius. May 12, 2011 ( Figure 4 2): S cattered clouds throughout the afternoon resulted in sporadic partial shading on both arrays throughout the sampling period. Array 1 had a verage power outputs ranging from a low of 28.01 Watts to a high of 138.20 Watts. Array 2 had average power outputs ranging from a low of 28.07 Watts to a high of 140.60 Watts. Array 1 operated in panel temperatures between a low of 58.70 degrees

PAGE 39

39 Celsius to a high of 72.39 degrees Celsius. Array 2 operated in panel temperatures between a low of 50.14 degrees Celsius to a high of 63.05 degrees Celsius. May 13, 2011 ( Figure 4 3): Scattered clouds th roughout the afternoon resulted in sporadic partial shading on both arrays, with light showers in the late afternoon. Arra y 1 had average power outputs ranging from a low of 30.66 Watts to a high of 123.96 Watts. Array 2 had average power outputs ranging from a low of 33.37 Watts to a high of 125.94 Watts. Array 1 operated in panel temperatures between a low of 55.58 degree s Celsius to a high of 70.95 degrees Celsius. Array 2 operated in panel temperatures between a low of 48.51 degrees Celsius to a high of 61.68 degrees Celsius. May 14, 2011 ( Figure 4 4): Significant showers throughout the late morning and afternoon, with t he skies beginning to clear after 2:40PM. Array 1 had average power outputs ranging from a low of 0 Watts during periods of heavy rain to a high of 167.41 Watts. Array 2 had average power outputs ranging from a low of 0 Watts during periods of heavy rain to a high of 172.10 Watts. Array 1 operated in panel temperatures between a low of 19.61 degrees Celsius to a high of 69.51 degrees Celsius. Array 2 operated in panel temperatures between a low of 18.68 degrees Celsius to a high of 61.65 degrees Celsius May 15, 2011 ( Figure 4 5): Heavy clouds and light showers throughout the afternoon. Array 1 had average power outputs ranging from a low of 16.18 Watts to a high of 179.01 Watts. Array 2 had average power outputs ranging from a low of 17.59 Watts to a h igh of 182.84 Watts. Array 1 operated in panel temperatures between a low of 51.24 degrees Celsius to a high of 65.95 degrees Celsius. Array 2 operated in panel

PAGE 40

40 temperatures between a low of 42.72 degrees Celsius to a high of 55.03 degrees Celsius. May 1 6, 2011 ( Figure 4 6): Sporadic clouds early with heavier clouds and light rain in the late afternoon. Array 1 had average power outputs ranging from a low of 14.90 Watts to a high of 164.04 Watts. Array 2 had average power outputs ranging from a low of 13 .75 Watts to a high of 166.21 Watts. Array 1 operated in panel temperatures between a low of 50.94 degrees Celsius to a high of 66.14 degrees Celsius. Array 2 operated in panel temperatures between a low of 41.71 degrees Celsius to a high of 55.75 degree s Celsius. May 17, 2011 ( Figure 4 7): Heavy clouds throughout the afternoon. Array 1 had average power outputs ranging from a low of 22.97 Watts to a high of 182.42 Watts. Array 2 had average power outputs ranging from a low of 20.30 Watts to a high of 18 0.29 Watts. Array 1 operated in panel temperatures between a low of 43.31 degrees Celsius to a high of 64.81 degrees Celsius. Array 2 operated in panel temperatures between a low of 35.37 degrees Celsius to a high of 53.87 degrees Celsius. May 18, 2011 ( F igure 4 8): Mostly clear skies throughout the afternoon. Array 1 had average power outputs ranging from a low of 121.17 Watts to a high of 141.48 Watts. Array 2 had average power outputs ranging from a low of 123.41 Watts to a high of 144.21 Watts. Arra y 1 operated in panel temperatures between a low of 59.06 degrees Celsius to a high of 66.27 degrees Celsius. Array 2 operated in panel temperatures between a low of 48.20 degrees Celsius to a high of 57.96 degrees Celsius.

PAGE 41

41 May 19, 2011 ( Figure 4 9 ): Most ly clear skies throughout the afternoon. Array 1 had average power outputs ranging from a low of 120.47 Watts to a high of 134.17 Watts. Array 2 had average power outputs ranging from a low of 84.44 Watts to a high of 136.91 Watts. Array 1 operated in p anel temperatures between a low of 62.50 degrees Celsius to a high of 69.71 degrees Celsius. Array 2 operated in panel temperatures between a low of 52.98 degrees Celsius to a high of 60.14 degrees Celsius. May 20, 2011 ( Figure 4 10 ): Clear skies throughou t the afternoon. Array 1 had average power outputs ranging from a low of 112.86 Watts to a high of 132.25 Watts. Array 2 had average power outputs ranging from a low of 115.22 Watts to a high of 134.84 Watts. Array 1 operated in panel temperatures betwe en a low of 66.43 degrees Celsius to a high of 72.41 degrees Celsius. Array 2 operated in panel temperatures between a low of 55.97 degrees Celsius to a high of 63.51 degrees Celsius. May 21, 2011( Figure 4 11 ): Clear skies throughout the afternoon. Array 1 had average power outputs ranging from a low of 108.99 Watts to a high of 127.33 Watts. Array 2 had average power outputs ranging from a low of 111.78 Watts to a high of 129.40 Watts. Array 1 operated in panel temperatures between a low of 65.07 degre es Celsius to a high of 71.83 degrees Celsius. Array 2 operated in panel temperatures between a low of 55.50 degrees Celsius to a high of .62.54 degrees Celsius. May 22, 2011 ( Figure 4 12): Mostly clear skies throughout the afternoon. Array 1 had average power outputs ranging from a low of 105.15 Watts to a high of 125.24 Watts. Array 2 had average power outputs ranging from a low of 96.99 Watts to a high of 127.16 Watts. Array 1 operated in panel temperatures between a low of 63.84

PAGE 42

42 degrees Celsius to a high of 70.61 degrees Celsius. Array 2 operated in panel temperatures between a low of 54.14 degrees Celsius to a high of 62.73 degrees Celsius. May 23, 2011 ( Figure 4 13): Clear skies early on with partly cloudy skies in the late afternoon. Array 1 had a verage power outputs ranging from a low of 19.41 Watts to a high of 131.08 Watts. Array 2 had average power outputs ranging from a low of 20.56 Watts to a high of 134.51 Watts. Array 1 operated in panel temperatures between a low of 50.99 degrees Celsius to a high of 71.62 degrees Celsius. Array 2 operated in panel temperatures between a low of 45.76 degrees Celsius to a high of 62.81 degrees Celsius. May 24, 2011 ( Figure 4 14): Clear skies early on with partly cloudy skies in the mid to late afternoon. Array 1 had average power outputs ranging from a low of 24.17 Watts to a high of 136.03 Watts. Array 2 had average power outputs ranging from a low of 25.24 Watts to a high of 139.26 Watts. Array 1 operated in panel temperatures between a low of 61.51 de grees Celsius to a high of 69.84 degrees Celsius. Array 2 operated in panel temperatures between a low of 53.34 degrees Celsius to a high of .61.92 degrees Celsius. May 25, 2011 ( Figure 4 15): Partly cloudy early on with mostly clear skies in the late aft ernoon. Array 1 had average power outputs ranging from a low of 91.20 Watts to a high of 136.15 Watts. Array 2 had average power outputs ranging from a low of 92.02 Watts to a high of 137.84 Watts. Array 1 operated in panel temperatures between a low of 61.73 degrees Celsius to a high of 71.03 degrees Celsius. Array 2 operated in panel

PAGE 43

43 temperatures between a low of 52.15 degrees Celsius to a high of 62.16 degrees Celsius. May 26, 2011 ( Figure 4 16): Partly cloudy throughout the afternoon. Array 1 had a verage power outputs ranging from a low of 9.82 Watts to a high of 144.32 Watts. Array 2 had average power outputs ranging from a low of 13.42 Watts to a high of 145.85 Watts. Array 1 operated in panel temperatures between a low of 50.86 degrees Celsius to a high of 71.80 degrees Celsius. Array 2 operated in panel temperatures between a low of 44.44 degrees Celsius to a high of 63.38 degrees Celsius. May 27, 2011 ( Figure 4 17): Partly cloudy in the early afternoon with heavy clouds and sporadic showers in the late afternoon. Array 1 had average power outputs ranging from a low of 10.69 Watts to a high of 171.90 Watts. Array 2 had average power outputs ranging from a low of 10.39 Watts to a high of 176.11 Watts. Array 1 operated in panel temperatures bet ween a low of 27.21 degrees Celsius to a high of 66.70 degrees Celsius. Array 2 operated in panel temperatures between a low of 24.99 degrees Celsius to a high of 59.17 degrees Celsius. May 28, 2011 ( Figure 4 18): Partly cloudy throughout the afternoon. Array 1 had average power outputs ranging from a low of 25.77 Watts to a high of 152.80 Watts. Array 2 had average power outputs ranging from a low of 23.40 Watts to a high of 153.11 Watts. Array 1 operated in panel temperatures between a low of 57.93 de grees Celsius to a high of 72.52 degrees Celsius. Array 2 operated in panel temperatures between a low of 48.88 degrees Celsius to a high of 64.16 degrees Celsius. May 29, 2011 ( Figure 4 19): Mostly cloudy throughout the early and late afternoon Array 1 had average power outputs ranging from a low of 12.11 Watts to a high of

PAGE 44

44 153.47 Watts. Array 2 had average power outputs ranging from a low of 14.06 Watts to a high of 156.64 Watts. Array 1 operated in panel temperatures between a low of 52.57 degrees C elsius to a high of 71.41 degrees Celsius. Array 2 operated in panel temperatures between a low of 44.75 degrees Celsius to a high of 62.46 degrees Celsius. May 30, 2011 ( Figure 4 20): Heavy clouds throughout the afternoon with light showers in the late af ternoon. Array 1 had average power outputs ranging from a low of 12.86 Watts to a high of 157.79 Watts. Array 2 had average power outputs ranging from a low of 15.35 Watts to a high of 163.37 Watts. Array 1 operated in panel temperatures between a low o f 51.74 degrees Celsius to a high of 71.80 degrees Celsius. Array 2 operated in panel temperatures between a low of 43.60 degrees Celsius to a high of 62.97 degrees Celsius.

PAGE 45

45 Table 4 1 Results of Analysis Date Array 1 AVG Array 2 AVG Difference Percent Array 1 Temp Array 2 Temp Difference Percent Before Array 2 Elevation 22 Apr 128.98 129.39 0.45 0.31% 65.02 63.45 1.56 2.51% After Array 2 Elevation 11 May 114.79 116.18 1.99 1.14% 66.83 58.36 8.47 14.53% 12 May 113.66 115.93 2.83 1.84% 67.54 59.19 8.36 14.15% 13 May 110.05 112.26 2.22 2.07% 65.21 56.54 8.67 15.36% 14 May 36.68 37.88 1.57 12.97% 28.99 26.82 2.17 6.78% 15 May 119.36 121.37 4.46 1.23% 60.34 50.87 9.47 18.62% 16 May 123.46 124.83 2.16 1.34% 61.14 50.41 10.73 21.32% 17 May 111.05 112.25 3.81 0.02% 55.80 46.38 9.41 20.36% 18 May 136.55 138.95 2.62 1.72% 64.26 54.11 10.15 18.83% 19 May 130.08 132.16 2.53 1.47% 67.33 57.64 9.70 16.87% 20 May 126.84 129.20 2.37 1.84% 70.36 60.43 9.93 16.47% 21 May 122.41 124.47 2.06 1.67% 69.44 59.52 9.91 16.68% 22 May 118.77 120.64 2.09 1.53% 68.01 59.43 8.57 14.47% 23 May 107.00 109.28 2.72 2.37% 66.21 57.33 8.88 15.50% 24 May 114.21 115.91 2.21 1.53% 66.56 58.21 8.36 14.39% 25 May 123.43 125.57 2.25 1.71% 67.24 58.35 8.89 15.31% 26 May 107.01 112.65 6.58 5.05% 64.77 56.45 8.32 14.77% 27 May 56.11 56.95 2.29 0.50% 46.88 41.31 5.57 13.06% 28 May 118.14 122.65 5.78 3.12% 68.03 58.83 9.20 15.74% 29 May 108.24 111.32 3.62 2.54% 64.38 55.65 8.73 15.73% 30 May 109.48 113.52 4.95 3.95% 64.55 55.40 9.15 16.57% Average 110.82 113.16 3.06 2.43% 62.70 54.07 8.63 15.77%

PAGE 46

46 Figure 4 1. May 11, 2011 Power and Temperature

PAGE 47

47 Figure 4 2. May 12, 2011 Power and Temperature

PAGE 48

48 Figure 4 3. May 13, 2011 Power and Temperature

PAGE 49

49 Figure 4 4. May 14, 2011 Power and Temperature

PAGE 50

50 Figure 4 5. May 15, 2011 Power and Temperature

PAGE 51

51 Figure 4 6. May 16, 2011 Power and Temperature

PAGE 52

52 Figure 4 7. May 17, 2011 Power and Temperatur e

PAGE 53

53 Figure 4 8. May 18, 2011 Power and Temperature

PAGE 54

54 Figure 4 9. May 19, 2011 Power and Temperature

PAGE 55

55 Figure 4 10. May 20, 2011 Power and Temperature

PAGE 56

56 Figure 4 11. May 21, 2011 Power and Temperature

PAGE 57

57 Figure 4 12. May 22, 2011 Power and Tem perature

PAGE 58

58 Figure 4 13. May 23 2011 Power and Temperature

PAGE 59

59 Figure 4 14. May 24, 2011 Power and Temperature

PAGE 60

60 Figure 4 15. May 25, 2011 Power and Temperature

PAGE 61

61 Figure 4 16. May 26, 2011 Power and Temperature

PAGE 62

62 Figure 4 17. May 27, 2011 Power and Temperature

PAGE 63

63 Figure 4 18. May 28, 2011 Power and Temperature

PAGE 64

64 Figure 4 19. May 29, 2011 Power and Temperature

PAGE 65

65 Figure 4 20. May 30, 2011 Power and Temperature

PAGE 66

66 CHAPTER 5 CONCLUSIONS Due to the dramatic effect of even partial shading from cloud cover, days that experienced even partially cloudy skies had a significantly greater range of power output than days with mostly clear skies. For this reason, it is the opinion of this researcher that da ys 18, 19, and 20 are probably the best examples of the differences in power output and module temperature due to elevating the second array to approximately eight inches. On those days, Array 2 outperformed Array 1 by an average of 2.51%. Array 2 also h ad an average temperature of 9.93 degrees Celsius less than Array 1 on those days. While initially the average differences in power production hovered between 1% and 2%, toward the end of the study these differences ranged between 3% and 5%. The data appe ars to indicate that Array 2 seems to perform even better than Array 1 under less than ideal conditions. The average increase in power over the course of the study was 2.43%. To better understand the value of such an increase, a PV array producing 20,000 kWh of electricity per year at $0.16 per kWh would see a $77.76 increase in revenue per year. Of particular interest is the temperature differential between the two arrays. As previously mentioned in the literature review, higher module temperatures cause the modules to degrade more rapidly.

PAGE 67

67 CHAPTER 6 RECOMMENDATIONS Because of the tendency for sporadic clouds in the afternoon in Florida during the summer months, it is the recommendation of the research that the study period be extended to cover at least one year. This will accomplish several things I t will provide data over a wider range of ambient temperatures, particularly during the winter months. I t will also provide data at a wider variety of angle of incidence, particularly when the su n is lower in the sky and for shorter periods of time. Additionally, i t will further support whether the 8 to 10 degree temperature differential is maintained throughout the year, which could be of interest to photovoltaic module manufacturers as it would affect their warranties. Beyond the purposes of this study, increasing the panel height from the roof surface makes sense if only from a maintenance standpoint, specifically in areas with many trees or airborne debris.

PAGE 68

68 LIST OF REFERENCES Clark, Ant hony, 2011. Feed in tariff means feast or famine for solar contractors. [Online] (Updated Saturday, January 15, 2011). Available at http://www.gainesville.com/article/ 20110114/ARTICLES/110119642?p=2&tc=pg (Accessed on 02 March 2011). Deb, Sayten B., 2000. Recent Developments in High Efficiency PV Cells, In: World Renewable Energy Congress VI, July 1 7, 2000, Brighton, U.K. Dunlop, Jim., 2007. Photovoltaic Systems Ameri can Technical Publishers, Inc. Green is Global, 2009. PV Laminate. [Online] Available at http://greenisglobal.net/building integrated photovoltaic systems for metal roofs (Accessed on 20 April 2011) Holladay, Martin (2009). Making Your Own Electricity: Onsite Photovoltaic Systems. Environmental Building News 19(11), pp. 2. Hussein, H.M.S.; Ahmad, G.E.; El Ghetany, H.H.,2003. Performance evaluation of photovol taic modules at different tilt angles and orientations, Energy Conversion & Management 45(15/16), pp.2441 2452. Jacobson, Barry. 2011. Optimizing photovoltaic systems. Personal Interview. 12 January, 2011. Kim, Jong Pil; Lim, Ho; Song, Ju Hun; Chang, Young June; Jeon, Chung Hwan., 2011. Numerical analysis on the thermal characteristics of photovoltaic module with ambient temperature variation Solar Energy Materials & Solar Cells 95(1), pp.404 407. Lamont, Lisa A.; El Chaar, Lana, 2010. Enhancement of a st and alone photovoltaic Renewable Energy: An International Journal 36(4), pp.1306 1310. Malkawi, Ali M.; Yun Kyu Yi; Lewis, Geoffrey, 2005. Integrated Evaluation of a Photovoltaic Installation. Jour nal of Architectural Engineering 11(4), pp.131 138. Sherwood, Larry, 2010. US Solar Market Trends 2009, [Online] Available at http://irecusa.org/wp content/uploads/2010/07/IREC Solar Market Trends Report 2010_7 27 10_web1.pdf (Accessed on 01 March 2011). Tigo Energy, 2011. Jacobson Home Installation. [Online] Available at https://datacenter.tigoenergy.com/main/installations.php?lang=0 (Accessed on 11 May 2011).

PAGE 69

69 Ubisse, A.; Sebitosi A., 2009. A new topology to mitigate the effect of shading for small photovoltaic installations in rural sub Saharan Afric a, Energy Conversion & Management 50(7), pp.1797 1801. U.S. Department of the Interior., 2005. Annual Precipitation for Florida. [Online] Available at http://www.n ationalatlas.gov/printable/images/pdf/precip/pageprecip_fl3.pdf (Accessed on 01 March, 2011).

PAGE 70

70 BIOGRAPHICAL SKETCH Randall received his Master of Science in Buildin g Construction with a focus in sustainable c onstruction from the University of Florid a in the summer of 2011. He re ceived his Bachelor of Arts in p sychology from the University of Central Florida in the Signal Officer s Basic Course, and Signal Captain