Development of a
High Efficiency Solar Cell
Using Adaptive Self-Cooling
Tiasha Joardar
© 2016
Rate of decrease in price of solar panel
beginning to stagnate
PHOTOVOLTAIC EXPERIENCE CURVE 1976 - 2014 ($/W)
Solar energy: promising source of clean
energy but not cost competitive with
fossil fuels
• Solar panel prices have dropped
100X in the last 40 years
• Approaching commercial viability
Cost per Watt
($/W)
The Economics
Solar energy is
30% more
expensive
Motivating Factors
Installed Capacity (MW)
Cost Analysis and Modeling
Cost / Watt (CPW) =
System cost = Cost of Solar Cells +
Peripheral Costs (cost of land, labor,
electronics, etc.)
Cost of solar cell system
Output power of system
• 25% of total system cost (@ current
market prices)
• Increases exponentially with its
power output capacity
Peripheral costs
• 75% of total system cost (@ current
market prices)
• Decreases inversely as power output
capacity of solar cell increases
Spectrolab
Solarworld
Entech
Power Ouput / P_solar
Peripheral Cost per 1000W ($)
CC0 = 0.15 PP0 = 0.09
Cost per sq. cm ($)
The Mathematics
Cost of solar cell
Power Ouput / P_rated
The Mathematics
Cost Analysis and Modeling (contd..)
30% decrease in CPW
needed to compete
with fossil fuels
25% increase in power output
needed to make solar energy
competitive with fossil fuels
= Ultimate solar cell efficiency target
The Self-Cooled High Efficiency Solar Cell
• Solar cells lose about 10% of power output capacity due to heating in sunlight - can this be
reduced / eliminated?
• Open-circuit voltage decreases with increasing temperature (approximately = - 2.2mV / °C )
• Fill-factor (FF) and short-circuit current (Isc) are not very sensitive to cell temperature
P
ΔT
Useful Part of Solar
Radiation
QIR
Infrared
(IR) Part of
Solar
Radiation
+
Σ
_
+
_
RTH
PV Solar
Cell
ΔT
POUT
÷
QPump
Heat Pump
PHeatPump
Qrej
By diverting a fraction of the solar cell output power to a high
efficiency heat pump in a feedback loop, its temperature can be
maintained at ambient air temperature
Thermoelectric Heat Pump
• Thermoelectric (TE) heat pump: an attractive option for solar cell temperature control
• Based on Peltier effect
• Current flowing through a junction composed of two materials generates or absorbs
heat depending on the direction of current
COLD SIDE
Conductor
Peltier Effect
Semiconductor
pellet
p
n
HOT SIDE
Heat
sink
HEAT
ADVANTAGES OF THERMOELECTRIC
• Very energy efficient at low heat loads
• Very energy efficient at low DT
• Max theoretical efficiency = Tc /(Th - Tc)
•
•
•
•
•
Easy to integrate into solar panels
Continuously adaptable to heat load
No refrigerants – eco-friendly
No moving parts – highly reliable
Cost of modules dropping over time
Objectives and Variables
VARIABLES
OBJECTIVES
•Develop a thermoelectric cooled hybrid solar
cell that
• produces at least 5% more power
when compared to a similarly sized
conventional solar cell operating under
identical environmental conditions
• Optimize design of hybrid cooled solar cell
such that the
• system cost /watt is reduced 5%
compared to a similarly sized conventional
solar cell
•Independent variables:
• Type of solar cell
• Thermoelectric cooled
hybrid
• Conventional
• Switching duty cycle of selfcooled cell
•Dependent variables:
• Power generated by each type of
solar cell
• Cost of solar cell
• System cost per watt (CPW)
•Constants:
• Size and rating of solar cell
• Ambient temperature
Hypothesis
• Based on theoretical calculations the hypothesis was that the project objectives could be met :
• by using model 12706 TE modules, and
• by operating the TE modules such that the cell temperature is equal to the ambient
temperature (ΔT = 0)
• A mathematical model was set up using Microsoft Excel to calculate the cooling power and
energy usage of various TE modules
• Thermal and electrical specifications from datasheets provided by TE vendors were used
• Thermal resistance of 500 K.cm2/W used
• Voc change of - 2 mV/C was assumed
• Standard solar radiation data used to compute heat load and cell temperature rise
Implementation of Concept
Sunlight
D
Switch S1
Solar cell
Thermoelectric
module
(1 - D)
Switch S2
Load
• A small fraction of the solar power output is diverted to the
TE heat pump
• Synchronous switching + duty cycle control
• power the load for a fraction D of the switching
frequency and
• power the TE module for the remaining fraction (1 - D)
TEC
SOLAR CELL
Electrical Implementation
Implementation of Concept
R1
OUTPUT
DISCHG
THRSHLD
TRIGGER
5
5
5
C2
OUTPUT
DISCHG
R2
THRSHLD
TRIGGER
5
5
5
3m
10K
C1
0.1μ
SWITCHES
SWITCHING CONTROL CIRCUIT
• Synchronous gate drive signals generated using 555 timers
• Duty cycle controlled by varying resistor R1
• Ripple cap C2 used to smooth out voltage across load during switching
L
O
A
D
0.5Ω
Implementation of Concept
CELL
R1a OR R1b
OR
R1c OR R1d
DIGITAL
OUTPUTS
ADC
INPUTS
TEMP SENSOR
THERMOCOUPLES
Software Implementation
AIR
R1a
ATMEGA168
μCONTROLLER
R1b
105K
92%
• Atmel Atmega168 mcontroller adjusts
duty cycle digitally
• Temp sensors connected to ADC detect
cell temp.
• Software determines appropriate
output port bit to set low
• MOSFET connected to high bit turns on
R1c
146K
94%
R1d
230K
96%
480K
98%
dT = cellTemp - airTemp;
if (dT > 8) {
PORTB = 0b00010000;
dutyCycle = 0.92;
}
else {
if (dT > 6) {
PORTB = 0b00001000;
dutyCycle = 0.94;
}
else {
if (dT > 4) {
PORTB = 0b00000100;
dutyCycle = 0.96;
}
else {
if (dT > 2) {
PORTB = 0b00000010;
dutyCycle = 0.98;
}
else {.....
Implementation of Concept
Solar Cell
Acrylic shield
- ve
Upper assembly
plate (Copper)
+ ve
Mechanical Implementation
Hybrid solar cell assembly
TEC
Lower assembly
plate (Aluminum)
Foam
insulation
TECs
Thermal
paste
Preliminary Checks
Experiment 1
• Gate drive signals
switching in sync as
designed
• Peak - peak swing is
12V
Experiment 1 (contd..)
50 mV drop in Voc
10% drop in Pmax
• Optimum load for max
power out = 0.5Ω
• Open cct. voltage drops 50
mV as cell heats up
• Max power output drops
about 10%
Duty Cycle = 80%, Vout = 0.406V
TEC
GATE
DRIVER
CIRCUIT
SOLAR CELL
Pmax versus Duty Cycle (Self-cooled)
Experiment 2
C2
3m
L
O
A
D
0.5Ω
Duty Cycle = 85%, Vout = 0.429V
SWITCHES
Purpose: Study effect of duty
cycle on cell power output with
self-cooling
Experiment 2 (contd..)
Duty Cycle = 91%, Vout = 0.471V
~ 10% boost
no selfcooling
Duty Cycle = 96%, Vout = 0.483V
RESULTS: Output power boosted
10.25% with self-cooling at 96%
duty cycle
• Exceeds the project target of 5% increase in power output
Thermocouple
(AIR)
GATE
DRIVER
CIRCUIT
TEC
DUTY
CYCLE
ADJUST
μC
ATMEL
ATMega168
ADC In
Thermocouple
(SOLAR CELL)
SOLAR CELL
Automatic Temperature Control
Experiment 3
C2
3m
L
O
A
D
0.5Ω
SWITCHES
Purpose: Study effectiveness of automatic control of
duty cycle based on atmospheric conditions
Experiment 3 (contd..)
RESULTS:
• Output power
consistently higher
with auto adjust selfcooling
• Total energy output is
about 6% higher with
self-cooling
Cost and CPW Estimation
Type of
cell
Cell Cost ($)
Peri.
Cost
TE
PV
Total
($)
Module
Conventional 1.25
TE Cooled
Hybrid
1.25
CPW
Pout
(W)
($/W)
n/a
1.25
3.75
0.423
11.8
0.06*
1.31
3.75
0.467
10.8
• Cost per watt of TE self-cooled cell is 8% lower than
conventional solar cell
• This exceeds the project target of 5%
• (*) TE module cost stated on per watt basis
Results Summary
Independent
Variable
Dependent
Variable
Effect on Dependent
Variable
(TE self-cooled vs
conventional)
TE cooled solar cell has
Power output 10.25% higher power
output than conventional
TE cooled solar cell is 4.8%
Cost of cell
more expensive
Self-cooled cell
duty cycle
TE cooled solar cell has 8%
lower cost per watt
Pout has a max value at
about 96% duty cycle
Type of solar cell
Cost per watt
Power output
• Original project targets met / exceeded
• Statistical data indicated 96.2% probability of exceeding target power
(hypothesis satisfied)