Energy for Sustainability

Randolph & Masters, 2008

Chapter 11:
Photovoltaic Systems
PV Applications
170 kW Rooftop PV
Integrated PV windows
PV Roofing
PV Array roof tile
PV array tile installed
Germany and Japan accounted for over ¾ of
total photovoltaic power installed in 2005
 Global shipments of photovoltaics are mostly
destined for grid-connected, residential, rooftop
systems
2500
Industry Shipments (MW/yr)

2000
Grid-connected
Residential

1500

1000

Off grid residential

500
Commercial
Industry
0
1 2 3 4 5
2000 2002 2004 2006 2008
Crystalline silicon forms a three-dimensional
tetrahedral structure (a); but it is easier to draw
it as a two-dimensional flat array (b).

silicon 
nucleus
+4 +4 +4
shared 
valence 
electrons
+4 +4 +4

(a)  tetrahedral
(b)  2­D version
(a) Hole-electron pair forms (b) Hole moves
An n-type material consists of immobile positive charges with
mobile electrons while p-type materials have fixed negative
charges and mobile, positively-charged holes.
When n-type and n-type semiconductors are brought together,
electrons diffuse from the n-region into the p-region filling holes and
creating immobile charges on each side of the junction.
The electric field created by those fixed charges opposes further
diffusion keeping holes on the p-side and electrons on the n-side.
When photons create hole-electron pairs near the junction, the electric field in
the depletion region sweeps holes into the p-side and electrons into the n-side
of the cell. Electron flow is clockwise through the load; conventional current
is in the other direction.
 The clear-sky solar spectrum at AM 1.5. For silicon, over half of
the incoming solar energy is wasted because photons either don’t
have enough energy or they have more than is needed to create
hole-electron pairs
1800
Photons with not
1600 Photons with too enough energy
much energy: Losses = 20.2%
Radiant power (W/m2-µm)

1400 Losses = 30.2%

1200
AM 1.5
1000

800 Silicon Bandgap

1.11 µm (1.12 eV)
600

400
NEAR IR
200 UV VIS 44%
2% 54%
0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6
Wavelength (µm)
PV performance measures usually assume sunlight passes through 1.5 times
as much air before it reaches the earth’s surface (designated as AM1.5) as it
would if the sun were directly overhead (AM1). AM1.5 is equivalent to the
sun being about 42 degrees above the horizon
 Maximum Wavelength (µm)
1.24 1.04 0.89 0.78 0.69 0.62
1.5

Lower voltage Higher voltage

Higher current Lower current
Theoretical Efficiency

Cu(In,Ga)Se 2 (CIGS)

OPTIMUM

GaAs
CuInSe2 (CIS)

a-Si
CuGaSe2
Si

CdTe
0
1 1.2 1.4 1.6 1.8 2
Band Gap Energy (eV)

Increasing the band gap increases cell voltage but reduces current, and vice
versa. The product of voltage and current is power, which means there is an
optimum band gap at which maximum power will be produced
The best laboratory, single-crystal, single-junction silicon
cells are approaching the theoretical efficiency limit of 31%.
35
Shockley Theoretical Limit 31%
30

25
Cell Efficiency (%)

20

15

10

0
1950 1960 1970 1980 1990 2000 2010
 Efficiencies of various photovoltaic technologies.
The rectangles show the range of large-module efficiencies
while the arrows indicate laboratory cells

Single crystal Si

Polycrystal Si Module
efficiency
range
Si Ribbon

CIS

CdTe Best laboratory

efficiencies
Amorphous Si

Graetzel

Polymer

0 2 4 6 8 10 12 14 16 18 20 22 24 26
PV Module Efficiency (%)
A high-
efficiency 95-
watt module on
the left
compared to a
lower-efficiency
75-W
conventional
module on the
right.
Higher
efficiencies
lower
area-related
costs.
 Types of PV cells:
Quartz crucible
Czochralski method for growing
single-crystal silicon
(a) Growing the crystal

Seed crystal Saw
Seed crystal
ystalline  Si  Ingot
Si  Ingot
Crystalline 
ingotSi ingot Molten Si
Molten Si Wafer
Wafer

 crucible
Quartz crucible

kerf
kerf
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.

(b) Slicing into wafers

(a) Growing the crystal
(a)  Growing the crystal (b) Slicing into wafers
Saw
Si  Ingot
Organic and nano photovoltaics can be printed onto
flexible plastic rolls. Cross section is for a dye-sensitized
or Graetzel solar cell

DSCC Substrate
Conducting SnO2 layer
TiO2-semiconductor
Dye
Electrolyte

Catalyst
Conducting SnO2 layer
Substrate
Putting the electrical contacts on the underside of the cell
and texturing the surface to bounce reflected sunlight into
the cell, helps boost efficiency of this device to over 20%

li n e sil ico n wafer

o nocrysta l
N-ty p e m
N P
P
N
P
N
N
P
- + - +

- +
A triple-junction (tandem) cell. Higher-energy photons (shorter
wavelengths) are captured in the upper junction. Longer
wavelength photons may be captured in subsequent junctions.

High-energy photons

Medium energy

Low energy

High band gap

Medium band gap

Low band gap

PV Cost, $/WP
 Photovoltaic cells, modules, and an array. Modules can be
connected in series to increase voltage and in parallel to
increase current, the product of which is power.

Modules in parallel
Increase current

Modules in series
increase voltage

Cell                        Module                  Array
Grid-connected systems allow you to spin your electric meter
backwards when your PVs generate more power than you need,
in essence using the grid for energy storage
 Two de-rating factors are likely to be encountered. For
rebate programs, the de-rating from (DC,STC) to (AC,PTC)
is often used. For system performance estimates, an overall
de-rating from (DC,STC) to AC is needed. Reasonable
estimates for these de-ratings are shown

PAC = PDC,STC x (De-Rating Factor) (11.3)

De-Rating (DC,STC) to (AC,PTC) - 0.84

kW DC, STC kW AC,PTC kW AC

Dirt, Mismatch
Inverter Other losses

De-Rating (DC,STC) to (AC) - 0.75

Four different power outputs for a PV system with representative
values based on 1 kW dc at standard test conditions. Also shown
are corresponding ways to express the $/Watt cost of system
having a pre-rebate cost of $7/Watt (DC,STC).

1 kW (DC STC)
­ 0.89 kW (DC, PTC) ­ 0.84 kW (AC, PTC) ­ 0.75 kW(AC)
Cell 25oC, 1-sun

Dirt, Mismatch
Inverter Other factors
­ 0.94 ­ 0.9

$7.00/W(DC,STC) $7.85/W(DC,PTC) $8.33/W (AC,PTC) $9.30/W(AC)

Monthly and annual average insolation for Denver for various
south-facing collector tilt angles. On an annual basis, insolation
varies only a few percent for a wide range of tilt angles
8.0

Horizontal
7.0
L-15
6.0
Insolation (kWh/m2-day)

Lat

5.0
Tilt = L+15

4.0
Average Annual kWh/m2
3.0 Tilt Year
Horizontal 4.6
2.0 Latitude - 15 5.4
Latitude 5.5
1.0 Latitude + 15 5.3

0.0
Jan Feb
1 2
Mar
3
Apr
4
May
5
Jun
6
Jly
7
Aug
8
Sep
9
Oct
10
Nov
11
Dec
12
 Annual average insolation in kWh/m2-day for
south-facing surfaces and for various tracking
collectors: same as average annual hours per day
of 1-sun (1 kW/m2)

San Francisco, CA

Boulder, CO NM

Minneapolis, MN
Sacramento, MN
CA
MA CA

Miami, FL OH
Albuquerque,

Roanoke, VA
Los Angeles,

TXHI
Honolulu, HI

OR
Minneapolis,
Boulder, CO

Seattle, WA
Raleigh, NC

Atlanta, GA
Atlanta, GA

Boston, MA
Chicago, IL
IL
Tucson, AZ

Cleveland,
Miami, FL

Honolulu,
Medford,
Chicago,

Boston,

Austin,
Tilt Angle
.6Horizontal
4.8 4.6 5.4 5.7
3.9 4.9
3.9 4.9
3.9 4.7
4.4 4.6
5.6 4.8
3.8 4.6
4.4 5.4
4.9 3.9
4.2 3.9
3.3 3.9
.4L - 15
5.1 5.0 5.5 6.3
4.4 5.5
4.5 5.5
4.6 5.3
4.9 5.4
6.3 5.1
4.2 5.0
4.9 5.5
5.2 4.4
4.8 4.5
3.8 4.6
.5Lat 5.2 5.1 5.7 6.5
4.4 5.6
4.6 5.5
4.6 5.4
5.0 5.5
6.4 5.2
4.1 5.1
4.9 5.7
5.3 4.4
4.8 4.6
3.7 4.6
.3L + 5.1
15 4.9 5.5 6.3
4.1 5.4
4.4 5.2
4.4 5.1
4.8 5.3
6.2 5.1
3.9 4.9
4.5 5.5
5.1 4.1
4.6 4.4
3.5 4.4
.890 3.0 3.1 2.9 3.9
3.0 3.5
3.2 3.4
3.3 3.4
3.2 3.8
4.1 3.0
2.7 3.1 2.9
3.1 3.0
3.2 3.2
2.6 3.3
.41-Axis
6.2(Horiz)
6.0 N,S
7.2 8.1
5.1 6.4
5.2 6.8
5.3 6.3
5.7 6.4
7.8 6.2
4.8 6.0 7.2
6.4 5.1
5.6 5.2
4.3 5.3
.21-Axis
6.5(Latitude)
6.4 7.4 8.7
5.5 7.0
5.7 7.4
6.0 6.9
6.2 7.2
8.5 6.5
5.1 6.4
6.5 7.4
6.7 5.5
6.1 5.7
4.7 6.0
.42-Axis
6.7 6.6 7.7 9.0
5.7 7.2
5.9 7.6
6.2 7.1
6.4 7.4
8.8 6.7
5.3 6.6
6.7 7.7
7.0 5.7
6.3 5.9
4.9 6.2
 The “Peak-Hours” Approach to Sizing a Grid-
Connected PV System

 Energy (kWh/yr) = PAC(kW) x (hours/day of 1-sun) x 365 day/yr

(11.4)

Where PAC is the ac power produced by the array when exposed to 1-sun of
insolation. Since modules are rated according to their dc output under standard test
conditions (DC, STC) we can insert (11.3) into (11.4) and get the following simple
sizing equation:

 Energy (kWh/yr) = PDC,STC(kW) x (De-Rating) x (hr/da of 1-sun) x 365 day/yr

(11.5)

We would also like to know how big an array must be to deliver the energy found in
(11.5). To do that, we need to know the PV efficiency under standard test conditions,
η, which is easy to obtain from manufacturer specifications.

 PDC,STC(kW) = 1 kW/m2 insolation x A (m2) x η (11.6)

The area, A, found in (11.6) is in square meters. The conversion to square feet is 1
m2 = 10.76 ft2).
 Simple sizing estimator from annual insolation to energy
generated (kWh/yr) per kW of (DC,STC) rated power.
Assumes a de-rating factor of 0.75
2400

2200
(kWh/yr) per kWDC,STC

2000

1800

1600

1400

1200

1000
3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9
Annual Insolation (kWh/m2day)
 30
PV efficiency 18% 16%
kWh/yr per square foot of collector

14%
25
12%
20
10%

15
8%

10

Annual energy production from a PV array

5
per square foot of collector using efficiency as a parameter

0
3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0

Annual insolation (kWh/m 2-day)

 The Total Potential for Roof-Top PV in the U.S.
 There are 3.5 billion square meters of residential roof-top area and
2.9 billion square meters of commercial roof area available for PV,
accounting for roof orientation, shading and structural issues. If we
assume the best 17-percent-efficient collectors available today, the
total area of 6.4 billion square meters would allow
P (DC,STC) = 6.4x109 m2 x 1 kW/m2 x 0.17 = 1,088 x 106 kW

of installed capacity. If exposed to an average of 5 kWh/m2-day of

insolation (equivalent to 5 hrs of 1 kW/m2 sun), and assuming a de-
rating factor of 0.75, that would be enough to generate
Annual energy = 1088 x 106 kW x 0.75 x 5 hr/day x 365 day/yr
= 1490 billion kWh/yr

which is just over one-third of the entire net output of all U.S.
power plants in 2005 (4339 billion kWh). In fact, if we include
transmission losses from traditional power plants to end-users, which
is avoided by on-site generation, this full build-out of PVs would be
sufficient to supply half of the total electricity demand of all U.S.
buildings.
 Average installed cost for 625 residential PV
systems installed between 1994 and 2000

INSTALLATION + BOS
$1.40 / Wdc
MODULES
$4.20 / Wdc

INVERTER
$1.20 / Wdc
TOTAL $6.80 / Wdc
 Cost of Electricity from PVs
 Suppose you install a 2 kW array in Los Angeles that costs
$7/WDC,STC. If you borrow the money at 6% interest on a 30-yr
loan, Let’s find the cost of electricity generated if it is installed on a
south-facing roof with tilt = L-15 and we assume a de-rating factor of
0.75.

 From Table 11.1, the average annual insolation is 5.5 kWh/m2-day

(5.5 hrs/day @ 1-sun), so the system generates
2 kW x 0.75 x 5.5 hr/day x 365 day/yr = 3011 kWh/yr

 The system costs $14,000. CRF (6%,30yr) = 0.07265/yr, so the

annual cost of the loan is $14,000 x 0.07265/yr = $1,017/yr.
Combining the $/yr and kWh/yr, we find the cost of electricity to be

$1017 / yr
$ / kWh = =$0.338 / kWh
3011 kWh / yr
 Showing the impact of net system cost $/Wdc,stc (after tax credits and
rebates) on annualized electricity cost. Assumptions:
De-Rating = 0.75, 6%, 30-yr loan, no deduction for mortgage interest
40
a y)
2 -d
35 Poor site (4 kWh/m2-day) m
h/
kW
Electricity cost (¢/kWh)

30 (5
te
si
g e
25 er
a
Av
Good site
20 (6 kWh/m2-day)

15

10

0
0 1 2 3 4 5 6 7 8 9 10
Photovoltaic Net System Cost ($/W)dc,stc
 PV economics with a rebate, tax credit and tax
deduction
 Let’s continue the example that resulted in PV electricity
costing $0.34/kWh.
 Assume we are eligible for the 30% (max $2000) Federal
Tax Credit and the California “Emerging Renewables”
rebate of $2.80/WAC,PTC. And, let’s assume your income
puts you in the 28% marginal tax bracket.
 The rebate of $2.80/W applies to the AC/PTC watts, not
the 2000 DC, STC watts. It takes a de-rating factor of 0.84
to convert PDC,STC into PAC,PTC, so our rebate is therefore
based on a system rated at
PAC,PTC = 2 kWDC,STC x 0.84 = 1.68 kW
 So, the system cost after applying the Emerging Renewables rebate is
System cost = $14,000 - 1680 W x $2.80/W = $9,296

 The Federal Tax credit is 30% of the price you pay after other rebates, up to
$2000, so it is
Tax Credit = Min. (0.30 x $9,296 = $2,789, or $2,000) = $2,000

 As a credit, this actually reduces your taxes by $2,000 making the system
cost now
After-tax credit system cost = $9,296 - $2,000 = $7,296

 Borrowing that on a loan with CRF (6%,30) = 0.07265 gives you annual loan
payments of
A = $7,296 x 0.07265 = $530.05/yr

 The system still generates 2 kW x 0.75 x 5.5 hr/day x 365 d/yr = 3011
kWh/yr. So our cost per kWh is now

$530.05 / yr
$ / kWh = =$0.176 / kWh
3011 kWh / yr
Including Tax-Deductible Interest
 We have a system that costs $7,296 after tax credits and
rebates. Now let’s add the impact of that 6% loan interest.
Assume you are doing well and are in the 28% marginal tax
bracket (MTB). In the first year your loan balance is the full
amount you borrowed; that is, $7296.
Tax reduction for interest (year 1) = $7296 x 0.06 x 0.28 =
$122.57
 The cost of your loan in the first year is now the loan payment
of $530.05 minus the tax savings on interest of $122.57,
which is $407.48. The cost of PV electricity to generate those
3011 kWh/yr is therefore

$407.48 / yr
$ / kWh = =$0.135 / kWh
3011 kWh / yr
 How does this compare to California
Electricity Rates?

PG&E Regular Residential (E-1)

Baseline Usage $0.11430/kWh
101% - 130% of Baseline $0.12989
131% - 200% of Baseline $0.21314
201% - 300% of Baseline $0.29007 Big users make out better with PV
Over 300% of Baseline $0.33039
Solar electricity can provide valued peak-demand reduction. This
graph shows the potential peak reduction impact of 2500 MW of
PVs on the system load for a major California utility during the 10
peak days of the year
 PVs For a Modest Cabin
 Let’s estimate the monthly energy demand for a cabin with a 22-cu. ft refrigerator,
five 16-W compact fluorescents used 6 h/day, a 20-in LCD TV used 4 h/day, a 1000-
W microwave used 6 min/day, and a 300-ft well that supplies 150 gallons of water per
day.
22 cu. ft Refrigerator: 1,250 Wh/d
5 16-W CFLs used 6 h/day: 5 x 16 W x 6 h/d = 480 Wh/d
20-in LCD TV used 4 h/day: 3.6 W/in x 20 in x 4 h/d = 288 Wh/d
20-in LCD TV standby 20 h/day: 10 W x 24 h/d = 240 Wh/d
1000-W Microwave, 6 min/day: 1000 W x 6 min/60min/h = 100 Wh/d
150 g/d @ 1.5 gpm, 180 W: 150 g/d/1.5g/min/60 min/h x 180 W = 300 Wh/d
Total: 2,658 Wh/d
 Notice the energy required by this TV when it isn’t turned on is almost as much
as when it is. This user might consider a power strip to really turn that thing off
when not in use.

 Suppose the average sun available is 5 kWh/m2-day (5 hrs of full sun) and suppose
the De-Rating Factor is assumed to be 0.75. From (11.5), a first-cut at the rated
power needed for this PV array would be
2658 Wh / d
PDC,STC = =709W ≈ 0.7 kW
(5h / d @1−sun) x 0.75
 the area of a 14% PV array to meet this load would be about 55 ft2.