Uploaded by Yellow 37

Хорошая преза по СБ 1

advertisement
MECH/ENGR 372
Space Systems Design and Engineering II
Power Systems
Dr. Christopher Kitts
Associate Professor, Santa Clara University
Satellite Power Systems
Required Functions
– Generate power to support average demands of components
– Store power to provide energy when arrays are not illuminated and during peaks
– Condition, regulate, suppress transients, and convert power as required by
components
– Distribute power throughout spacecraft to components
– Include some level of robustness to power-related faults
– Include special functions such as safety inhibits, ordnance firing support, etc.
Considerations
– Degradation over mission life and variation (radiation, thermal cycling, etc.)
– Orbit parameters (radiation, solar parameters, etc.)
– Spacecraft configuration (spinner vs 3-axis control, mechanisms, etc.)
Space Systems - Power: C. Kitts, Santa Clara University
2
Satellite Power Systems
Power
Generation
Source
Control
Regulation
& Control
Storage
Control
Energy
Storage
Space Systems - Power: C. Kitts, Santa Clara University
3
Power
Distribution
Satellite
Loads
Satellite Power Systems
Generation Options
Photovoltaics
RTGs
Primary batteries
Tethers
Solar dynamic
Nuclear reactor
Power
Generation
Source
Control
Regulation
& Control
Storage
Control
Energy
Storage
Space Systems - Power: C. Kitts, Santa Clara University
4
Power
Distribution
Satellite
Loads
Satellite Power Systems
Generation Options
Photovoltaics
RTGs
Primary batteries
Tethers
Solar dynamic
Nuclear reactor
Power
Generation
Source
Control
Regulation
& Control
Storage
Control
Storage Options
Secondary batteries
Fuel Cells
Flywheels
Energy
Storage
Space Systems - Power: C. Kitts, Santa Clara University
5
Power
Distribution
Satellite
Loads
Satellite Power Systems
Generation Options
Photovoltaics
RTGs
Primary batteries
Tethers
Solar dynamic
Nuclear reactor
Power
Generation
Source
Control
Source Control
Shunt Dissipation
Series Dissipation
Storage Options
Secondary batteries
Fuel Cells
Flywheels
Regulation & Control
Peak power tracking
Direct energy transfer
Regulation
& Control
Power
Distribution
Storage
Control
Energy
Storage
Space Systems - Power: C. Kitts, Santa Clara University
6
Storage Control
Charge Control
Reconditioning
Satellite
Loads
Satellite Power Systems
Generation Options
Photovoltaics
RTGs
Primary batteries
Tethers
Solar dynamic
Nuclear reactor
Power
Generation
Source
Control
Source Control
Shunt Dissipation
Series Dissipation
Storage Options
Secondary batteries
Fuel Cells
Flywheels
Regulation & Control
Peak power tracking
Direct energy transfer
Regulation
& Control
Distribution
Cabling
Switching
Fault Protection
Power
Distribution
Storage
Control
Energy
Storage
Space Systems - Power: C. Kitts, Santa Clara University
7
Storage Control
Charge Control
Reconditioning
Satellite
Loads
Satellite Power Systems
Generation Options
Photovoltaics
RTGs
Primary batteries
Tethers
Solar dynamic
Nuclear reactor
Power
Generation
Source
Control
Source Control
Shunt Dissipation
Series Dissipation
Storage Options
Secondary batteries
Fuel Cells
Flywheels
Distribution
Cabling
Switching
Fault Protection
Regulation & Control
Peak power tracking
Direct energy transfer
Regulation
& Control
Storage
Control
Energy
Storage
Space Systems - Power: C. Kitts, Santa Clara University
8
Power
Distribution
Satellite
Loads
Demand
Average & Peak
Duty cycling
Storage Control
Charge Control
Reconditioning
Lecture Topics
Generation (emphasis on photovoltaics)
Storage (emphasis on batteries)
Regulation & Control
Distribution
Systems Issues
– Power budgeting
– System interactions
– Commands and telemetry
– Fault Tolerance
Space Systems - Power: C. Kitts, Santa Clara University
9
POWER GENERATION
Space Systems - Power: C. Kitts, Santa Clara University
10
Power Generation
Primary batteries
– Chemical reaction produces energy (primary = not-rechargeable)
– Launch vehicles, very short duration spacecraft (~hours to days)
Photovoltaic cells
– Photovoltaic effect in p-n junction semiconductors generates power
Static thermal-to-electric
– Heat source drives a thermal-to-electric energy conversion process
– Sources: solar, radioisotope decay, nuclear fission reaction
– Thermoelectric, thermionic
Dynamic thermal-to-electric
– Heat source and heat exchanger to drive an engine
– Sources: solar, radioisotope decay, nuclear fission reaction
– Cycles: Stirling, Rankine, or Brayton
Space Systems - Power: C. Kitts, Santa Clara University
11
Power Generation Comparison
Wertz, SMAD
Space Systems - Power: C. Kitts, Santa Clara University
12
Power Generation Comparison
Fortescue, Fig 10.1
Space Systems - Power: C. Kitts, Santa Clara University
13
POWER GENERATION
Photovoltaics
Space Systems - Power: C. Kitts, Santa Clara University
14
Solar Power - Introduction
Illumination
Photovoltaic effect
Solar cell
– Types
– Performance
Solar Arrays
– Layouts
– Performance
– Practical Issues
Solar Array Design Process
EUVE, Courtesy UC Berkeley
Space Systems - Power: C. Kitts, Santa Clara University
15
Solar Power – Solar Illumination
Illumination a function of distance
– 1358 W/m2 = Average solar power density at Earth-Sun distance outside of
Earth’s atmosphere
– Varies: ~ 1310 W/m2 at aphelion to 1400 W/m2 at perihelion
– Solar intensity decreases w/square of distance to sun – so, can be much less for
outer planetary probes (Venus~2600 W/m2, Mars~585 W/m2, Jupiter~50 W/m2)
Time in eclipse (worst case approximation, circular polar orbit)
RE 
−1 


ρ = sin 
 h + RE 
Max t eclipse =
Where: ρ=Earth’s angular radius (deg)
P=orbital period
h=orbit altitude (km)
RE=6378 km
2ρ
P
o
360
Sellers Fig 13-20
Space Systems - Power: C. Kitts, Santa Clara University
16
Solar Cell Theory: Semiconductors
Semiconductor lattice (Silicon example):
– Each atom shares 4 electrons within lattice to fill valance band with 8 electrons
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Kitts
Space Systems - Power: C. Kitts, Santa Clara University
17
Solar Cell Theory: Semiconductors
N-type doped semiconductor – creates extra electrons:
– Pentavalent atoms (atoms w/5 electrons in valance band, like Antimony) added as impurities
– Note – the lattice is still neutrally charged… it simply has mobile electrons in this form
Si
Si
Si
Si
Si
Sb
Si
Si
Sb
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Sb
Si
Si
Si
Kitts
Space Systems - Power: C. Kitts, Santa Clara University
18
Solar Cell Theory: Semiconductors
P-type doped semiconductor – creates extra holes:
– Trivalent atoms (atoms w/3 electrons in valance band, like Boron) added as impurities
– Note – the lattice is still neutrally charged… it simply has mobile holes in this form
Si
Si
Si
Si
Si
B
Si
Si
B
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
Si
B
Si
Si
Si
Kitts
Space Systems - Power: C. Kitts, Santa Clara University
19
Solar Cell Theory: Semiconductors
Create a P-N Junction
– Electron diffusion across
junction fills holes
– Creates a “depletion
region” where charge
carriers don’t exist
– Field in the depletion
region inhibits additional
flow of charge
– Coulomb forces limit size
of region
P-type doping
N-type doping
p-type positive carriers
n-type negative carriers
Note: free hole carriers,but the material does
NOT have a net positive charge
Note: free electron carriers,but the material
does NOT have a net negative charge
P-N Junction
- -
For solar cells, the
potential difference
drives the photovoltaic
power generation
+
+
+
+
+
+
+
Kitts
Depletion region
Extra electron
- Negative ion
Space Systems - Power: C. Kitts, Santa Clara University
20
Extra hole
+ Positive ion
Solar Cell Theory: Photovoltaic Effect
Simple model of the photovoltaic effect
–
–
–
–
–
Photon strikes cell and its energy moves an electron to the conduction band
Electron-hole pair created and separated / driven by p-n potential difference – E field
Electrons flow out n-type, through load, and recombine with hole at p-type
Note: some electrons reflected by cell, some pass through cell
Note: for absorbed photons, difference btw photon energy and gap energy becomes heat
Energy (eV)
Front contact
N-type semiconductor
E
Field
+
Depletion region
Electron
travel
-
LOAD
Conduction band
Band Gap
Energy
Valence band
Thermal energy or energy from a
photon can create an electron-hole
pair by moving the electron from the
valence band to the conduction band
P-type semiconductor
Rear contact
Kitts
Space Systems - Power: C. Kitts, Santa Clara University
21
To help protect y our priv acy , PowerPoint prev ented this external picture from being automatically downloaded. To download and display this picture, click Options in the Message Bar, and then click Enable external content.
Solar Cells
Typical layered architecture for a solar cell
– N and P layers (usually thin N layer on top of thicker P layer)
Electrons (minority carriers in P layers) have much longer diffusion length than holes in N layers; since
radiation damages cells by reducing effective diffusion paths, a P base layer is less susceptible to damage
– BSR (back surface layer) used on back to reflect photons back through
– Electrical contact layers (top one is windowed to allow light to pass through)
Top layer contact trade-off: Cover it all and light can’t pass through. Cover only at the edges and electrons
have to travel far to get to contact. But there is “series resistance” within the semiconductor, so want to limit
this since it is a source of losses.
– Antireflective coating on top used to prevent reflections
– Cover glass – hermetic seal used for structural, thermal, radiation protection
Often smooth for pointed panels but textured if not actively pointed in order to increase photon capture
Clear Coverglass
Antireflective Coating
Front Electrical Contact
N-doped Si
P-doped Si
Kitts
Back Surface Reflector
Back Electrical Contact
Space Systems - Power: C. Kitts, Santa Clara University
Example: Plastecs
22
Additional types of solar cells
Si semiconductor lattice
– Si atoms share 4 electrons to fill
valence band w/8 electrons
– Ge (Germanium) can do the same
Apapted from Honsberg & Bowden, Photovoltaics
Other possibilities – element
combinations with 3+5 or 2+6
valence electrons
– III-V: Ga+As (Gallium Arsenide),
Ga+P (Gallium Phosphide), etc.
– II-VI: Cd+S (Cadmium Sulfide),
Cd+Te (Cadmium Teluride), etc.
Different combinations of
elements lead to differences in:
–
–
–
–
Energy conversion efficiency
Ease of manufacturing
Radiation tolerance
Structural integrity
Space Systems - Power: C. Kitts, Santa Clara University
23
Cell Efficiency
photons is converted to electricity
An incident photon must transfer an amount
of energy equal to the band gap energy in
order to generate an electron-hole pair
– Photons with energy below this level are lost
– Photons with energy above this level are
converted, but excess energy is lost
– Note that photon energy is a function of frequency
(wavelength)
Different semiconductor lattice structures
have different band gap energies, thereby
leading to different efficiencies
What is the “most efficient” band gap energy
– Too high and lower energy photons lost
– Too low and excess photon energy lost
– Given photon energies in the solar spectrum,
empirical estimate ~1.4 eV (single layer cells)
Energy (eV)
Only a portion of the energy of incident
Conduction band
Band Gap
Energy
Valence band
If EPho<EBG, no
electron-hole generation
If EPho=EBG, an electronhole pair is generated
If EPho>EBG, an electronhole pair is generated and
the photon continues with
EPho_new=Epho_orig-EBG
Thermal energy or energy from a photon can create
an electron-hole pair by moving the electron from
the valence band to the conduction band
Cell
Material
Band
Gap (eV)
Max Wavelength
(µ
µm)
Silicon
1.12
1.12
CdS
1.2
1.03
GaAs
1.35
0.92
GaP
2.24
0.56
CdTe
2.1
0.59
CuGaInSe2
1.0
1.24
Extended from Fortescue Table 10.1
Space Systems - Power: C. Kitts, Santa Clara University
24
Comparison of Cell Types (Single Layer)
Note: quoted numbers are often for lab quality cells tested under ideal conditions
Silicon: production cells ~ 15%, lab cells ~24%
– Crystalline silicon (example used in previous discussions), inexpensive, rugged
– Polycrystalline silicon & amorphous silicon: cheaper but less efficient (5-10%)
Gallium-Arsenide: production cell ~19%, lab cells ~22%
– Larger band gap energy (high efficiency) and less susceptible to radiation (2-3 x) than Si
– More expensive material & doping process; also brittle (difficult to manufacture)
– Ultimately, manufacturing solution is to put on layer of germanium
Indium Phosphide: production cells ~18%, lab cells ~20%
– Outstanding radiation tolerance (more than an order of magnitude better than Si)
Cadmium Telluride: lab thin films ~17%
– Thin films use less material, allow flexibility, and have potential to be less expensive
Copper Indium Diselenide: lab thin films ~20%
Space Systems - Power: C. Kitts, Santa Clara University
25
Solar Cell Efficiency
Space Systems - Power: C. Kitts, Santa Clara University
26
Multi-layer (multi-junction) Cells
Improved efficiency with integrated “stacks”
of cells – “multi-junction”
– Highest band gap junction on top so highest
energy photons absorbed first
– Each junction is sensitive to different parts of the
solar spectrum
– Junctions connected in series – higher voltage
with less relative current density
Dual Junction Example - (see figure)
– ~21-24% production cells
– GaAs (Gallium Arsenide) with GaInP2 (Gallium
Indium Phosphide) on top
Sandia
Triple Junction Example –
– ~28%+ production cells
– Dual junction GaAs and GaInP
– 3rd junction: activating the interface with the
germanium substrate on the bottom
Pisacane Fig 6.10
Space Systems - Power: C. Kitts, Santa Clara University
27
Solar Cell Power Generation
Electrical model of a solar cell extends the concept of a diode
Diode review
Forward bias
Once external voltage is high enough,
it overcomes internal field and pulls
n-region electrons through the
depletion region
I
(conventional current)
Reverse bias
External voltage pulls n-region
electrons away from the
depletion region, inhibiting
current flow
V-I Curve
Characterization of performance:
small forward bias enables current
flow; no negative current flow until
large negative breakdown is reached
no conventional current
I
+
p
p
-
-
+
V
-------+++++
-
free
electrons
pulled
through
V
-------+++++
n
n
electron flow
-10’sV
+
no electron flow
Space Systems - Power: C. Kitts, Santa Clara University
28
free
electrons
pulled
away
Not to scale
~0.5V
V
Solar Cell Power Generation
Solar cell model: superposition of a diode and the photovoltaic effect
qV
 nkT

I = I O  e − 1


I
V-I curve shifts down with increased
illumination. Interpretation is that
light creates a forward voltage bias
combined with a negative current
flow, meaning that the device supplies
current rather than consumes current
V
qV
 nkT


I = I O  e − 1 − I L


+
-
Simplified model neglecting
internal resistance
I
qV
 nkT

I = I L − I O  e − 1


Where: I=device current
Io=device leakage current in dark
q=||electron charge||
V=device voltage
k=Boltzmann’s constant
T=absolute temperature
IL=Light generated current
n = measure of how ideal device is
(n=1 is ideal; 1<n<2)
Space Systems - Power: C. Kitts, Santa Clara University
29
V
As a convention, the V-I curve is
typically inverted with the
understanding that power is now being
generated when cell is illuminated
Solar Cell Power Generation
Solar cell operation
ISC
VL, IL
I
ISC (short circuit current): Max current
when load is a short; no voltage drop
VOC (open circuit voltage): Max voltage
when load is open circuit; no current flow
VOC
V
+
LOAD
VL and IL (load voltage and current):
These are the operating conditions of the
cell given the characteristics of the load.
Note that if the load characteristics change,
the cell operating point slides along the V-I
curve. If the load has a fixed voltage, the
V-I curve shows the resulting current that
the cell will produce for that voltage.
-
Space Systems - Power: C. Kitts, Santa Clara University
30
Solar Cell Power Generation
Solar cell maximum power
Power = Voltage x Current
V3, I3
I
P3
V2, I2
V1, I1
P2
P1
Power Generated: For the V-I curve, power
generated at a given operating point (generally
dictated by the characteristics of the load) is the
area in the box defined by VL and IL for that
operating point
Maximum Power: At the “knee” of the curve,
where the prescribed box has maximum area
V
Remember: The load might vary, and the V-I
curve may vary based on intensity, temperature,
and other factors…
Space Systems - Power: C. Kitts, Santa Clara University
31
Solar Cell Power Generation - Variation
Effect of intensity and temperature variation on power production
Piscane Fig 6.17 and 6.16
Pisacane
Pisacane
Intensity Variation
Temperature Variation
An intensity increase causes an increase in the
number of photons striking the semiconductor,
thereby increasing the number of free charge
carriers. The intensity curve shifts up: ISC scales
proportionally while VOC increases only slightly.
Temperature increases electron energy in the semiconductor,
effectively reducing the band gap and increasing current since
less energetic photons can now cause current generation.
More dramatically, VOC decreases due to increase in Io with
temperature . Note power increase at low temperatures.
Space Systems - Power: C. Kitts, Santa Clara University
32
Solar Cell Power Generation - Degradation
Primary cause of degradation is due to radiation
– Radiation damage in the semiconductor creates “trapping centers” that cause
electron-hole pairs to recombine. These diminish the number of free electrons
available to produce current.
Space cells are generally n-type deposited on a p-type base
– Radiation damages base
region by reducing diffusion
length of minority carriers
– So, use p-type for base
material given its inherently
longer diffusion length
Pisacane
Space Systems - Power: C. Kitts, Santa Clara University
33
Solar Arrays
Common panel architectures
Body mounted arrays
Deployed planar arrays
Space Systems - Power: C. Kitts, Santa Clara University
34
Solar Array Cell Layout
Interconnects
– Major array failure hazard due to thermal cycling
Cell and substrate materials differ, allowing differential
expansion with temperature swings of 100°C in a few
minutes – stress relief required
– Atomic oxygen interacts with interconnect material (often
silver foil)
Causes silver oxides which flake, thin out the interconnect,
and raise resistance thereby resulting in power loss (Hubble
launch delay led to replacement of interconnect material
from Silver to Molybdenum with Silver coating)
Fortescue Fig. 10.3
Substrate Material
– Typical: Kapton (possibly reinforced) mounted on
honeycomb panel for rigidity
– Want high thermal conductance through panel to move
heat away from cells
– Thermal cycling on aluminum face sheets historically has
led to faults; carbon fiber composite facesheets often used
given their thermal stability, but then must address thermal
and electrical conductivity
Space Systems - Power: C. Kitts, Santa Clara University
35
Pisacane Fig. 6.24
Solar Array – Multi-cell Arrangements
Groups of cells used together in arrays
Arranging cells in series – a “string”
– Voltages add, and the currents through each cell are equal
Arranging cells in parallel – a “block”
– Currents add, and the voltages across each cell are equal
VT=V1+V2
V1
V2
IT=I1=I2
Cell 1
Cell 2
I1
I2
Series Configuration:
Voltages add, Currents
are equal
Parallel Configuration:
Voltages are equal,
currents add
Space Systems - Power: C. Kitts, Santa Clara University
36
VT=V1=V2
IT=I1+I2
Cell 2
Cell 1
I1 V1
I2 V2
Solar Array – Multi-cell V-I Curves
Cells in series:
– Series: Voltages add, currents same
– So, at a given current, the voltages
add across each series cell
– Construct curve by adding cell
voltages for each current value
I
Cell1
Multi-cell
string
Cell2
V
I
Cells in parallel:
– Parallel: Currents add, voltage same
– So, at a given voltage, the currents
add from each parallel cell
– Construct curve by adding cell
currents for each voltage value
Multi-cell
block
Cell2
Cell1
V
Space Systems - Power: C. Kitts, Santa Clara University
37
Solar Array – Simple Layout (Block of Strings)
PSA = N x Pcell
Use this to determine N, the total number of cells
VSA = Vstr = Nseries x Vcell
Arrange cells in series to achieve panel voltage
ISA = Nstr x Istr = Nstr x Icell
Arrange strings in parallel to achieve panel current
Note: N = Nseries x Nstr
A “block of strings”
Vcell, Icell
VSA
Simple
Given: PSA=2kW with a VSA= 28V
Example Given: Vcell=0.5V and Icell=150mA
N = PSA/(Vcell x Icell) = 26,667 cells
Nseries = VSA / Vcell = 56 cells
…
… …
Istr
Istr
…
Istr
Nstr = N / Nseries = 476 strings
Note: Often VSA is set higher than Vbus given
regulation losses, the desire to be able to directly
charge batteries w/out boost converters, etc.
Space Systems - Power: C. Kitts, Santa Clara University
38
Solar Array – Simple Layout V-I Curve
Ideal case shown below:
– Start with V-I curve of individual cell
– Form string V-I curve: sum voltages for a given current through all cells
– Form array V-I curve (block of strings): sum currents for a given voltage through all strings
I
Cell V-I Curve
I
ISC-cell
String V-I Curve
I
ISC-cell
Array V-I Curve
ISC-Array= Nstr x ISC
VOC-cell
V
V
V
VOC-string = Nseries x VOC
Space Systems - Power: C. Kitts, Santa Clara University
39
VOC-string = Nseries x VOC
Solar Array – Blocking Diodes
Practical Issues
Protection Diodes
– Put in series w/each string to prevent reverse charging
– Prevents a fully charged battery from reverse biasing cells when they aren’t illuminated
VBUS = VSA - Vdiode
-Vdiode
Vcell, Icell
VBUS
VSA
Battery
…
… …
Istr
Istr
…
Istr
Space Systems - Power: C. Kitts, Santa Clara University
40
Loads
Solar Array – Cell Mismatch Issues
Practical Issues
Cell Mismatch – What if the cells aren’t exactly the same? Mismatch caused by:
– Cells with different physical properties (leading to different VOC, ISC, etc.)
– Uneven operating conditions (illumination, temperature, etc.) across the array
– Examples: cell fault, manufacturing differences, shading, etc.
Cell mismatch for series-connected cells – Case 1: VOC mismatch
VT=V1+V2
VOC mismatch: Benign effect
IT=I1=I2
At a typical
operating point,
current through
cells is equal,
voltages across
cells are different
I
V1
Cell 1
I1
V2
Cell 2
I2
Cell1
Cell2
Array
V
Space Systems - Power: C. Kitts, Santa Clara University
41
Solar Array – Cell Mismatch Issues
Practical Issues
Cell mismatch for series-connected cells – Case 2: ISC mismatch
– REMEMBER: ISC is dramatically affected by illumination (shadowing)
VT=V1+V2
IT=I1=I2
ISC mismatch: Potential for a dramatic effect. Total
current limited by ISC of bad cell – meaning that IArray is
effectively ISC of the bad cell
I
V1
Cell 1
I1
Cell2
V2
Cell 2
I2
Cell1
Array
V
This might be OK if we planned on this. But when this
happens, it is usually unplanned – and it can lead to
disastrous effects
Space Systems - Power: C. Kitts, Santa Clara University
42
Solar Array – Cell Mismatch Issues
Practical Issues
Cell mismatch for series-connected cells – Case 2: ISC mismatch
– REMEMBER: ISC is dramatically affected by illumination (shadowing)
How this often goes wrong
– Cells are nearly matched
– Array operates near max power by design
I
Cell2
Cell1
Array
V
Space Systems - Power: C. Kitts, Santa Clara University
43
Solar Array – Cell Mismatch Issues
Practical Issues
Cell mismatch for series-connected cells – Case 2: ISC mismatch
– REMEMBER: ISC is dramatically affected by illumination (shadowing)
The bad cell can become
reverse biased if there is a
low voltage demand
How this often goes wrong
Cells are nearly matched
Array operates near max power by design
I
But now, let one cell become shadowed
Array produces max current near ISC-bad cell
Near this point, bad cell is reverse biased
Cell1
Cell2
Array
V
Space Systems - Power: C. Kitts, Santa Clara University
44
Solar Array – Cell Mismatch Issues
Practical Issues
Cell mismatch for series-connected cells – Case 2: ISC mismatch
– REMEMBER: ISC is dramatically affected by illumination (shadowing)
The bad cell can become
reverse biased in attempt to
maintain current flow
How this often goes wrong
Cells are nearly matched
Array operates near max power by design
I
But now, let one cell become shadowed
Array can produce max current of ~ISC-bad cell
Near this point, bad cell is reverse biased
Result is power dissipation in bad cell
This can lead to severe overheating
Gets even worse for larger strings
Cell1
Cell2
Array
V
Power dissipated in bad cell – leads to
massive overheating that can damage cell
or even parts of the local array
Space Systems - Power: C. Kitts, Santa Clara University
45
Solar Array – Cell Mismatch Issues
Practical Issues
Cell mismatch for series-connected cells – Case 2: ISC mismatch
– REMEMBER: ISC is dramatically affected by illumination (shadowing)
How this often goes wrong
Cells are nearly matched
Array operates near max power by design
But now, let one cell become shadowed
Array can produce max current of ~ISC-bad cell
Near this point, bad cell is reverse biased
Result is power dissipation in bad cell
This can lead to severe overheating
Gets even worse for larger strings
Honsberg & Bowden, Photovoltaics
Space Systems - Power: C. Kitts, Santa Clara University
46
Solar Array – Cell Mismatch Issues
Practical Issues
Cell mismatch for series-connected cells –Case 2: ISC mismatch
CURE = Bypass diodes
ISC mismatch: Using bypass diodes for each cell alters the
V-I curve for reverse biasing.
VT=V1+V2
IT=I1=I2
I
V1
Cell 1
I1
Cell2
V2
Cell 2
Array
I2
NOTE - Bypass diodes are often
manufactured into the solar cell construct
rather than being externally wired
Cell1
Power dissipated in bad cell is now limited since
the operating voltage is much lower due to diode
Note minor dissipation in bypass diode as well
Space Systems - Power: C. Kitts, Santa Clara University
47
V
Solar Array – Cell Mismatch Issues
Practical Issues
Cell mismatch for parallel-connected cells – Case 1: ISC mismatch
ISC mismatch: Benign effect
I
VT=V1=V2
IT=I1+I2
Cell2
Cell 2
Cell 1
I1 V1
Array
I2 V2
Cell1
V
Space Systems - Power: C. Kitts, Santa Clara University
48
Solar Array – Cell Mismatch Issues
Practical Issues
Cell mismatch for parallel-connected cells – Case 2: VOC mismatch
VOC mismatch: For voltages to the right of the
arrow, Cell 2 begins to dissipate heat due to
current reversal through cell
I
VT=V1=V2
IT=I1+I2
Cell 2
Cell 1
I1 V1
Array
I2 V2
Cell2
Cell1
V
Space Systems - Power: C. Kitts, Santa Clara University
49
Solar Array – Alternate Layouts
Parallel Strings
Series of Parallel Blocks
…
…
…
… …
Parallel Strings w/Bypass
…
If one cell fails or is shadowed,
its string is limited. Because
string’s voltage must match
voltage of other strings, bad cell
could be reversed biased and
permanently damaged.
… …
…
If one cell fails or is shadowed, overall
current is drops by 1/Nparallel; additional
single failures (in independent blocks)
have no effect. But – the affected block is
now in series with higher current blocks –
so now, the entire block could be reverse
biased – damaging all cells in the block!
Gets worse with increasing Nseries.
Space Systems - Power: C. Kitts, Santa Clara University
50
…
…
…
For high voltage arrays where
Nseries is high, the “series of
parallel blocks” layout is not
preferred. Rather, bypass diodes
are used for each cell or for small
groups of cells, and the preferred
layout may revert to “parallel
strings”
Solar Array Design Process (Wertz & Larson)
Step 1 – Determine key system requirements
Step 2 – Determine required array power
Step 3 – Determine idealized cell performance
Step 4 – Determine necessary BOL power performance
Step 5 – Determine array area
Space Systems - Power: C. Kitts, Santa Clara University
51
Solar Array Design Process – Step 1
Step 1 - Requirements & constraints:
Mission life:
– note that cells degrade over time due to radiation
Average power required:
– the more power, the bigger the array
Orbital specifications
– Governs illumination intensity, angle, solar flux, eclipse durations, etc.
Approach:
– Array sized to meet EOL need
– Implies extra power at BOL which must be addressed (dissipation issues)
Example:
– 10 year mission life
– 500W continuous
– Circular with 800 km alt, 35.1 min eclipse duration
Space Systems - Power: C. Kitts, Santa Clara University
52
Solar Array Design Process – Step 2
Step 2 – Estimate Array Power
Determine average power required during illuminated portion of orbit
 PeTe Pd Td 


+
Xe
Xd 

PSA =
Td
Where Pe=required eclipse power,
Pd=required daylight power,
Te=time in eclipse per orbit,
Td=time in daylight per orbit,
Xe=system efficiency in eclipse,
(typ: Xe= 0.65 direct; 0.6 ppt)
Xd=system efficiency in daylight
(typ: Xd= 0.85 direct; 0.80 ppt)
Example:
–
–
–
–
Pe=Pd=500W
Te=35.1 min, Td=65.9 min
Assume ppt: Xe=0.6, Xd=0.8
Psa=1069 W
PSA: The amount of Power you
need to generate given the
demand, system efficiency, and
orbit characteristics
Space Systems - Power: C. Kitts, Santa Clara University
53
Solar Array Design Process – Step 3
Step 3 – Trade Cell Type to Estimate Panel Power Density
Cell selection criteria
– Efficiency, radiation-degradation sensitivity, cost, mass, risk, etc.
– Note that quoted numbers are often:
for ideal lab conditions
for cells, with array fabrication leading to additional losses
Cell performance:
'
O
P = η cell *1358W / m
2
Notation:
P power
P’ power density
Rough estimation by cell type
– For Silicon cells = 14%, Gallium Arsenide cells = 18%, Solar flux ~ 1358 W/m2
– Idealized performance: Si ~ 190 W/m2, GaAs ~ 244 W/m2
Example
P’O: The ideal power density you get
given your choice of solar
cell type
– Assume Si cells, P’o= 190 W/m2
Space Systems - Power: C. Kitts, Santa Clara University
54
Solar Array Design Process – Step 4
Step 4 – Determine necessary P’BOL performance
Realistic degradation in arrays
– Packing factor ~ 0.85: spatial loss due to spacing and interconnects of cells
– Shadowing: often this can be avoided, but take into account if present
– Temperature: warmer arrays less efficient, Si looses ~0.5%/ ° above 28°
I d = I packing ∗ I shading ∗ I temp
Account for pointing losses (worst case):
BOL Power Density (exposed area):
cos θ
Assumes constant pointing
loss throughout mission
'
PBOL
= Po' ∗ I d ∗ cosθ
Example
– Packing factor = 0.85, no shadowing, Temperature loss = 0.9, Id = 0.77
– θ=23.5°
P’BOL: The realistic power density you get
2
– P’BOL=134 W/m
given your choice of solar cell type as
Space Systems - Power: C. Kitts, Santa Clara University
55
well as how you install it and the
operating conditions for your satellite
Solar Array Design Process – Step 5
Step 5 – Determine array area
Account for degradation over operating life
– Radiation causes significant damage: ~ 1.5 (Si) - 2.5 (GaAs) %/yr in LEO
– Additional degradation causes an additional ~ 1.25 %/yr
thermal cycling, micrometeoroid strikes, plume impingement, outgassing, etc.
Ld = (1 − (degradation/yr)) mission life
BOL Power Density:
Solar Array Area:
'
'
PEOL
= PBOL
∗ Ld
'
ASA = PSA / PEOL
Example
Assumes all cells are
nominally illuminated (e.g.,
not body mounted, etc.)
P’EOL: The realistic power density you get
AT END OF LIFE GIVEN
DEGRADATION OVER TIME given the
previously determined beginning of life
realistic power density
– 3.75% degradation/yr, Ld=0.68
– P’EOL=91.4 W/m2
– ASA=11.7 m2
Space Systems - Power: C. Kitts, Santa Clara University
56
Solar Array Design Process
Mass Estimate (for planar arrays – Wertz & Larson, SMAD):
M SA = (0.04 kg/W ) ∗ PSA
Cell Layout (from earlier in this lecture)
– Select layout: parallel strings, series of parallel blocks, parallel strings w/bypass
– Select VSA – often a little higher than Vbus
– Select cells – get Vcell and Icell
– Nseries = VSA / Vcell ; N = PSA/(Vcell x Icell) ; Nstr = N / Nseries
– Lay out strings so all cells in string are nominally illuminated at the same time
(for example, for body mounted arrays, don’t lay strings out around vehicle –
keep strings on a single side).
Example:
– MSA=1069/25 = 42.76 kg
– See previous example for string computations
Space Systems - Power: C. Kitts, Santa Clara University
57
Power Generation – Summary of Losses
Cell Efficiency
– 15-30% efficiency (70-85% loss) typical
Path efficiency through batteries to load
– 60-85% efficiency (15-40% loss) typical
“Inherent” degradation
– Packing factor, temperature losses, shadowing
– 12-50% loss typical
Pointing loss
– Cos θ given angle between the panel normal and the direction of the sun
Lifetime degradation
– Radiation, thermal cycling, micrometeoroids, plume impingement, outgassing
– 2-4% loss typical per year
– Ld=(1 – annual degradation)lifetime
Space Systems - Power: C. Kitts, Santa Clara University
58
Solar Panels Fabrication Example
Sapphire microsatellite
Cells: GaAs on Ge substrate w/cover
glass; 15% eff, VOC=0.8V
Strings: 20 cells for 16V, each diode
protected; within strings, cells spacing is
0.016” on side and 0.30” on ends per
Lockheed spec; interconnects are silver
plated molybdenum
8 Body Mounted Panels: 4 strings on
top/bottom,2 strings on sides. Aluminum
panels w/insulating dielectric coating;
cells bonded to panels using special
RTV; panels are grounded to frame
Space Systems - Power: C. Kitts, Santa Clara University
59
POWER GENERATION
RTGs: Radioisotope Thermoelectric
Generators
Space Systems - Power: C. Kitts, Santa Clara University
60
RTGs – Theory of Operation
Thermoelectric effect (Seebeck effect)
– A voltage can be generated between two materials (usually p-type and n-type
semiconductors) if a temperature difference is maintained
Thermal Source
Heat Flow
Electrical Insulation
Conductive strap
Electron Flow
P
N
Doped semiconductors
+
Conductive strap
Electrical Insulation
Thermal Sink
Space Systems - Power: C. Kitts, Santa Clara University
61
RTGs – Basic Design
RTGs
– Heat source created by a material that radioactively decays
Emitted high energy particles typically absorbed by a secondary material that heats up
– Semiconductor “legs” arranged in series modules to create desired voltage
High temperature semiconductors: Lead telluride (300-500°C), silicon germanium (> 600 °C)
– Modules can then be arranged in parallel to create desired current
Radioactive heat source
P
N
P
N
P
N
P
N
P
N
P
N
+
-
+
-
+
-
+
-
+
-
+
-
Cold Space
Space Systems - Power: C. Kitts, Santa Clara University
62
RTG – Example
General Purpose Heat Source RTG (GPHS-RTG)
– Developed by GE (now LM) for JPL in 1990’s
– Designed with modular heat sources that could be sized to future missions
– Modular design approach allowed qualification to be done once
– ~300 W BOL with ~260 W after 10 yrs; designed to survive Earth re-entry
Cassini RTG
18 GPHS modules
SiGe thermoelectric couples
~1000°C inner core
~300°C cold junction
Cassini used 3 of these RTGs
Launched in 1998
Courtesy JPL
Space Systems - Power: C. Kitts, Santa Clara University
63
RTG – Design Considerations
Radioisotope selection
– Lifetime should be compatible with mission life
– Pu-238 (Plutonium) generally used due to half-life
(~87 yr half-life), radiation profile, engineering
properties, etc.
P(t ) = Po * e −.693t /τ where τ = halflife
– Power output degrades over time
– Size BOL output with excess power
– Degradation: fuel decay, material degradation, etc.
Sizing
Control - Power can’t be turned off, so shunt excess
Use in actual missions
– RTGs are only real option for deep-space missions
– All long-duration US missions have used either
photovoltaic/battery or RTG power systems
– Cassini, Galileo, Ulysses, Voyager I & II, TRANSIT
Space Systems - Power: C. Kitts, Santa Clara University
64
New Horizons RTG, Courtesy NASA
RTGs – Pros and Cons
Advantages:
– Power generation independent of distance from the sun, satellite orientation and selfshadowing effects, eclipse shadowing effects (e.g., landing on the dark side of a moon,
planet), environmental dust and debris
– Can provide low power levels for long periods of time (decades)
– Not susceptible to radiation damage
– Highly reliable with proven track record – no moving parts, no maintenance
– Very mass efficient compared to photovoltaic systems
Disadvantages:
–
–
–
–
–
–
–
Creates a radiation environment for satellite (shielding, RTGs on booms, etc.)
Hazardous material handling issues prior to launch
High temperature operation leads to thermal design issues
Relatively low overall efficiencies (conversion of heat energy to electrical energy)
Can’t be turned off (so active power management necessary when power draw is low)
Can interfere with some science instruments relating to plasma measurements
Politics of radioactive material on satellites given launch failures, de-orbits, etc.
Capsules design to survive re-entry, dramatically driving up cost
Space Systems - Power: C. Kitts, Santa Clara University
65
Download