Solar Site Evaluation

Tools & Techniques to Quantify

& Optimize Production
By Mark Galli and Peter Hoberg

T
Improve system performance here are multiple factors to consider when
evaluating a site for a photovoltaic or solar
thermal installation, and each may impact

and energy harvest optimal energy production. In addition to

latitude and longitude, which determine the
sun path characteristics, panel or collector

projections with a thorough orientation—tilt and azimuth—defines the

field of view that an array has of the sun. Shading from
trees, hillsides, buildings or other obstructions can cause

site evaluation that includes significant degradation in energy production. Additionally,

local and regional weather patterns result in site-specific
seasonal and daily fluctuations in solar insolation.

shading analysis and These factors combine and interact to determine the
solar energy incident on an array and therefore impact both
financial returns and customer satisfaction. Tools and tech-

insolation quantification. niques used in site evaluation emphasize shade analysis and
optimizing solar access.

54 S o l a r Pr o | December/January 2009
The Importance of Site Evaluations installation complexity and implementation cost,
An early and thorough site evaluation can lead to better such as conductor and trench routing; proximity
system designs that will result in the following benefits: of array to inverter; roofing material integrity; rafter
increased energy production by selecting the best loca- and beam spacing for engineering calculations; and
tion for the solar array; improved accuracy in energy pro- safety concerns and access issues.
duction estimates due to better quantification of shading
and other site-specific issues; optimized financial incen- • Direct contact with the client to discuss additional
tives, such as state-specific rebates that adjust for panel issues, including possible aesthetic concerns and
orientation and shading; improved system installation and financing plan options.
materials cost estimates; and increased customer satisfac-
tion and confidence, which in turn can lead to repeat or Sun Paths
referred business. Solar access will depend on the sun’s location, defined
Leading solar system designers and installers invest by elevation angle and azimuth direction, as it varies
significant efforts into on-site data collection and evalu- through each day and throughout the year. This path
ation, especially during customer qualification, initial can be plotted for a given latitude and longitude. An
design and proposal preparation. The site information example sun path chart is shown in both rectilinear and
gathered includes: polar formats in Graphs 1a and 1b. Typically, sun charts
are centered around south (180° azimuth) for sun path dia-
• Measurement of location parameters, including grams in the Northern Hemisphere, and around north for
available area for the array, roof pitch or site grade, sun path diagrams in the Southern Hemisphere. Examples
and azimuth. shown in this article are for the Northern Hemisphere with
references to summer and winter from a Northern Hemi-
• Measurement of solar access and impact of shade- sphere view.
causing obstructions, as well as evaluation of shade- The sun path is a function of latitude and longitude, and
reduction strategies, such as tree trimming or it shifts with changes in location. This effect is illustrated for
removal. two different locations in Graphs 2a and 2b (p. 56). Moving
north toward higher latitudes, the annual sun path chart
• Identification of issues that could jeopardize the shifts, indicating that the sun is at lower elevations. Mov-
viability of a project or result in increased design and ing south, the chart shifts, indicating higher sun elevations.

70º 12pm South

1pm
11am Jun 2 1 0º 21
15 0º
M a y 21
2pm
60º
10am
Apr 20
Dec 21
3pm
0º

24

50º
12

0º

Jan 21
Solar Elevation

9am
Mar 20
4pm Feb 20
40º
11h 12h 13h
10h 14h
D at a co u r t e s y s o l ard a t . uo reg o n . ed u ( 2)

8am
9h 15h
West
East

Feb 21 5pm 8h 16h Mar 20

30º
7h 17h
7am 80º
Jan 21 6h 70º 18h
20º 6pm Apr 20
D ec 2 1 60º
6am 5h 19h
50º
30
º
60

0º

10º 7pm 40º May 21

5am Jun 21
30º
20º
0º º 33
30 0º
60º 90º 120º 150º 180º 210º 240º 270º 300º 10º
East Solar Azimuth West
North

Graphs 1a & 1b A sun path chart for Portland, OR, in (a) rectilinear and (b) polar formats.

solarprofessional.com | S o l a r P r o 55
Solar Site Evaluation

Graphs 2a & 2b The polar

Dec Dec
sun path chart in 2a shows Nov Nov
Oct
Sacramento, CA (38.6N 121.5W, Oct

black lines), and a 2° shift in Sep Sep

latitude to the north (green lines). Aug Aug

Jul
Graph 2b shows a 2° shift in Jul
Jun Jun
12:00
longitude to the west (green 10:00
12:00
14:00 10:00 14:00

Da ta c o u r te sy so l me tr i c . c o m ( 2 )
lines) of Sacramento. 16:00 16:00
8:00 8:00

18:00 18:00
6:00 6:00

Definitions:
Terms Used in
Solar Site Evaluation Graph 2a Graph 2b

Insolation. The incident solar radiation on the Moving west toward greater longitude, the sun’s path remains
earth’s surface in a given time window, typically expressed the same, but the time for each sun location is shifted toward
in kWh/m2/day. later in the day.
Note that a latitude shift of 2° to the north (138 miles)
Solar access/shading. Solar access is the shifts the noon elevation angle by less than 2°. A longitude
ratio of the insolation in a given location, including shade, shift of 2° (108 miles at 38.6° N) causes a time shift of about
to the insolation available at that location without shade. 10 minutes at noon. When making measurements at a site,
Solar access is typically expressed in percent for a given using exact coordinates is ideal; but it is normally sufficient
time period, such as a month, season, or year. to use the sun paths for locations within 50 miles of the site
being surveyed and in the same time zone.
Optimum tilt and orientation.
For any location, the optimum tilt and orientation is the Shading Analysis
specific fixed tilt and orientation for solar arrays that absorbs Shade can have a dramatic impact on solar production.
maximum solar energy over the course of one year. Evaluating it is critical before getting too far into the system
design process. Various on-site analysis c o n t i n u e d o n p a g e 5 8
Tilt and Orientation Factor (TOF).
TOF is the solar insolation at the actual tilt and orientation
divided by the insolation at the optimum tilt and orienta-
tion, expressed in percent.

Total Solar Resource Fraction

(TSRF). TSRF is the ratio of insolation available account-
ing for both shading and TOF, compared to the total inso-
lation available at a given location at the optimum tilt and
orientation and with no shading. TSRF is also expressed
as a percentage. TSRF = solar access x TOF
Co u r t es y s o lm et r i c .c o m

Magnetic declination. Magnetic declina-

tion is the azimuth offset between magnetic north and
true north, expressed in degrees east or west.

Typical Meteorological Year (TMY).

TMY is a collection of weather information that includes
data about insolation for every hour of a typical year Image 1 Using a fish-eye lens, solar access measurements
computed using historical weather data. and computed results can be charted.

56 S o l a r Pr o | December/January 2009
Solar Site Evaluation

Extending Point
Measurements
This article focuses on ways of characterizing the solar access
from point measurements, for example with the fish-eye lens.
Specific modules in an array will experience shading at differ-
ent times of day and year. Typically, to get a good estimate of
a system’s performance, multiple points should be measured
C ou r te sy so l me tr i c . c o m

in and around the array, such as at each corner of the array.

Some techniques and trends for combining multiple point
readings are listed below. The different techniques vary in
accuracy and complexity.

PAverage multiple point measurements. Average the monthly

Image 2 A screen capture of a Google Earth image with an solar access values from each reading to generate 12
incorporated skyline plot map showing monthly shading data numbers that reflect the average monthly solar access for the
for Skyline Number 2 (Sky02). entire array. The California Solar Initiative program requires
that the measurements be taken at the four corners of the
array and averaged in this manner. More points can improve
70º accuracy but can be time consuming.
shading profile
60º P Interpolation techniques. Use linear interpolation to
estimate the solar access at locations in between measure-
50º ment points. Precise measurements of the relative locations
Solar Elevation

are necessary to enable accurate interpolation.

40º
P 3-D modeling from on-site data. In this case, the height
30º and elevation of each obstruction must be known. This can
D at a co u r t esy solm e tr ic .c om

provide some measurement challenges. With few obstruc-

20º tions, this approach is practical, but the complexity increases
with many obstructions.
10º
P 3-D modeling from aerial/satellite imaging. GIS and map-
0º ping technologies are advancing rapidly. Tools like Google
60º 90º 120º 150º 180º 210º 240º 270º 300º Earth, Microsoft Virtual Earth, and ArcGIS Explorer are extend-
East Solar Azimuth West
ing our ability to view buildings and obstructions online. In the
Graph 3 This graph represents a morning shade profile with future, these technologies may provide the 3-D details neces-
sun paths superimposed. sary for initial estimates and may provide a useful complement
to on-site evaluations.

tools and techniques can be used, including viewing reflec- Even for small residential arrays, shading analysis typi-
tions from a mirror dome (SolarPathfinder), multiple digi- cally requires taking multiple readings at various positions.
tal pictures (Wiley Electronics ASSET) or using a fish-eye These readings can then be averaged or processed in simu-
lens and digital camera to capture the whole sky in a single lation programs to modify the energy production estimates
image (Solmetric SunEye). The result is information about for the entire array. For larger commercial and utility scale
the shading obstacles’ elevation versus azimuth. The sun projects, readings from various locations can be tagged
path information described earlier can be overlaid directly with their GPS coordinates and then compared on a map,
onto these views, so that the impact of shading can be like the Google Earth plot map shown in Image 2. Shad-
determined either graphically or numerically. An exam- ing data can also be shown as elevation versus azimuth as
ple of a polar chart taken with a fish-eye lens is shown in depicted in Graph 3, using the same data as Image 1. Sun
Image 1 (p. 56). elevation and azimuth are also shown. c o n t i n u e d o n p a g e 6 0

58 S o l a r Pr o | December/January 2009
Solar Site Evaluation

the measurement plane (H) is shown in Illustration 1.

D a ta c o ur te sy so l me tr i c . c o m ( 3 )

The California Solar Initiative requires that the shad-

ing ratio (D:H) must be at least 2:1. This is equivalent to an
H
elevation angle of less than 26.6°. If a site meets these
θr
requirements, it is deemed to have good solar access, and a
D detailed shade analysis is not required.
An alternative way to specify shading is to determine
H/D = TAN (θr) a site’s shade-free hours, such as 10am–2pm or 9am–3pm.
θr = TAN-1 (H/D) In this case, obstructions are allowed to cast shadows only
before or after the specified time period.
Illustration 1 To quantify shading, determine the elevation The minimum D:H ratio can be specified for the shade-
angle (θ) as a function of the distance to the obstruction (D) free time periods for a given location. Some example calcu-
and its height above the measurement plane (H). lations are shown in Table 1. Note that this requirement is
worst-case and applies only at the lowest sun elevation of
the year within those time windows. This may be too con-
servative and restrictive for typical pitched roof applica-
θr tions, but it may be useful when considering row spacing in
X flat-roof or ground-mount system installations.
H When collecting shading data, it is possible to take the
θg data in one location and extrapolate it for another. This can
D
allow an analysis using ground-level data by extrapolating
up a distance H and over a distance X. This approach is
θr = TAN-1( (D x TAN(RADIANS(θg) ) - H) / (D + X) ) useful and often necessary when taking measurements at a
location where the building is not yet constructed or when it
Illustration 2 The necessary calculations for extrapolating is not practical to get shading data from the true height of the
ground level measurements to the roof, simplified for a single proposed array. The calculations can be complex for the
plane. full 3-D analysis. For reference, a 2-D equation is shown in
Illustration 2. For this equation to apply, the lines must all
Elevation Angles and Extrapolating Measurements be coplanar.
Elevation angle is a very useful way to describe obstruc-
tions. However, the angle alone may be insufficient to Array Orientation
describe the shape and direction of the shade on an array. The tilt and azimuth of an array establish the field of view,
For a more complete analysis, the distance to the obstruc- and the sun paths as seen within that c o n t i n u e d o n p a g e 6 2
tion and its height can be measured. Shading is sometimes
quantified in this way. The relationship between elevation ridge
angle (θ), distance to the obstruction (D) and height above

tree
Minimum Distance to Height Ratio west
string
east
string
chimney

Latitude Longitude 10am–2pm 9am–3pm

  ºN ºW EST EST

New York 40.7º 74.0º 2.8:1 4.4:1

repositioned
Atlanta 33.7º 84.4º 2.5:1 4.2:1 modules
South

Miami 25.8º 80.2º 1.7:1 2.6:1

Table 1 The minimum D:H at several example locations Illustration 3 Two modules in this array were repositioned to
guarantees shade-free status during the specified times. avoid the impact of shading from the tree and the chimney
east of the array.

60 S o l a r Pr o | December/January 2009
Solar Site Evaluation

field of view will determine the solar insolation. Note Annual Insolation as a Function of Panel Orientation
Location: Sacramento Metropolitan, CA. Optimal Tilt=30º, Azimuth=176º, Insolation=2050 kWh/m2.
that the field of view will also determine what shad- Station ID: 724839, Latitude: N38.70, Longitude: W121.58. 2
kWh/m
ing obstacles will impact solar access. For example, if 90º 100% (2050)

D a ta c o ur te sy so l me tr i c . c o m ( 3 )
the field of view is toward the west, it will decrease the 99% (2029)
roof pitch
impact of shading obstacles on the eastern horizon 24/12
95% (1947)
90% (1845)
60º
and increase the impact of shading obstacles on the 20/12
16/12 85% (1742)
western horizon. 12/12
10/12
80% (1640)
8/12 75% (1537)
For rooftop solar installations, the roof param- 30º
6/12 70% (1435)
eters—pitch and azimuth—typically determine the 4/12 65% (1332)
2/12 60% (1230)
array orientation and layout. Therefore the field of 0º 55% (1127)
view can be adjusted only by picking the best section 90º 120º 150º 180º 210º 240º 270º 50% (1025)

of the roof and the ideal location within that area.

Rooftop obstacles, such as chimneys and vent pipes, Annual Insolation as a Function of Panel Orientation
can cause shading in locations that are otherwise Location: Sacramento Metropolitan, CA. Optimal Tilt=36º, Azimuth=213º, Insolation=1745 kWh/m2.
Station ID: 724839, Latitude: N38.70, Longitude: W121.58.
desirable. In some cases, relocating modules that kWh/m 2
90º 100% (1745)
would be shaded can help significantly. Not only does 99% (1727)
this preserve the production of the affected modules, roof pitch 95% (1657)
24/12 90% (1570)
but, more importantly, it also can preserve the pro- 60º 20/12
85% (1483)
16/12
duction of one or more series strings by minimizing 12/12 80% (1396)
10/12
shading and keeping string voltage within the invert- 30º
8/12 75% (1308)
6/12 70% (1221)
er’s maximum power point tracking window. Multiple- 4/12 65% (1134)
point shading analysis makes this kind of performance 2/12 60% (1047)
0º 55% (959)
optimization possible. An example is shown in Illus- 90º 120º 150º 180º 210º 240º 270º 50% (872)
tration 3 (p. 60).
Graphs 4a & 4b Graph 4a plots the annual insolation versus tilt
Site Specific Weather Pattern and azimuth for Sacramento, CA, with no shading. Graph 4b adds
Quantification significant morning (eastern) shade and shows the resulting impact
The insolation values for specific sites are heavily on insolation availability.
influenced by the weather patterns for that partic-
ular location. Various weather and insolation data Annual Insolation as a Function of Panel Orientation
is available for locations worldwide. The National Location: Honolulu Intl. Arpt., HI. Optimal Tilt=22º, Azimuth=143º, Insolation=2060 kWh/m2.
Station ID: 911820, Latitude: N21.32, Longitude: W157.93.
kWh/m2
Renewable Energy Laboratory recently released 90º 100% (2060)
updated weather data for 1,020 locations in the US. 99% (2039)
The database, known as TMY3, records insolation roof pitch 95% (1957)
24/12 90% (1854)
during all hours of the year and can be very help- 60º 20/12
16/12 85% (1751)
ful in estimating weather-corrected insolation for a 12/12
10/12
80% (1648)
75% (1545)
given site. 30º
8/12
6/12 70% (1442)
The effects of shading and weather can be 4/12 65% (1339)
60% (1236)
observed using an annual insolation chart. Graph 0º
2/12
55% (1133)
4a shows a plot of annual insolation vs. tilt and azi- 90º 120º 150º 180º 210º 240º 270º 50% (1030)

muth values for Sacramento, California, without

shading. Graph 4b incorporates significant shade Graph 5 Annual insolation versus tilt and azimuth for Honolulu, HI,
to the east, as in the Skyline shown in Image 1. shows a significant eastward shift in the optimum annual values due
Notice that the optimum value shifts in azimuth to cloudy afternoon weather patterns.
and tilt, and the optimum available insolation is
reduced. Sacramento’s insolation chart is close to
symmetrical around the southern direction, which indi- The impact of climate change on insolation levels
cates that morning and afternoon insolation is similar. could be significant over an array’s 30-year operational life.
Honolulu, Hawaii, shown in Graph 5, reveals a significant Although general trends may be clear, precise modeling is
eastward shift in the optimum values, indicating significantly difficult, and results may be controversial. General trends,
less insolation in the afternoon than in the morning, most such as drier or wetter, could provide a qualitative indicator
likely due to patterns of afternoon clouds and rain. of either higher or lower solar production. c o n t i n u e d o n p a g e 6 4

62 S o l a r Pr o | December/January 2009
Solar Site Evaluation




 The Bottom Line

The financial plan for a photovoltaic or solar thermal proj-
ect involves costs and benefits, and a detailed site evaluation
can make or break a project’s success. Contractors, investors
and clients will all benefit from an open, honest review of
the solar access data upfront to avoid unwanted surprises
during or after project implementation.
  The site assessment itself represents a project cost.
Depending on the size of the array and complexity of the
site, a thorough site assessment can add a few hundred to
a few thousand dollars to the project. In the precontract
phase, contractors are typically not getting paid for this
work, so quick results with sufficient accuracy are critical.
Often a preliminary site analysis is performed as part of the
sales process with a more detailed analysis after a contract
is signed.
Solar energy project success requires a good site assess-
ment. Location, panel orientation, weather and shading
  all interact to influence solar access and therefore energy
production potential. With the right tools and techniques,
                   solar installers can dramatically improve the chances of
 their project’s success.
     
g C O N TAC T
Mark Galli / Solmetric / Bolinas, CA / mark@solmetric.com / solmetric.com
Peter Hoberg / Solmetric / Bolinas, CA / peter@solmetric.com /
solmetric.com

Resources:
California Solar Initiative /
gosolarcalifornia.org/csi/index.html
National Renewable Energy Laboratory /
nrel.gov/rredc/pvwatts
(photovoltaic performance calculator)
National Renewable Energy Laboratory, TMY3 /
rredc.nrel.gov/solar/old_data/nsrdb/1991-2005/tmy3
(weather data)
University of Oregon Solar Radiation Monitoring Laboratory /
solardat.uoregon.edu/SunChartProgram.html
(sun path chart program)
University of Oregon Solar Radiation Monitoring Laboratory /
solardat.uoregon.edu/PolarSunChartProgram.html
(polar sun path chart program)

Shading analysis tool:

SolarPathfinder / solarpathfinder.com
Solmetric / solmetric.com
Wiley Electronics / we-llc.com

64 S o l a r Pr o | December/January 2009