New Design Trends in Photovoltaic Converters
Alex Van den Bossche Ghent University,
EELAB Electrical Energy Laboratory
Sint –Pietersnieuwstraat 41 Gent Belgium
Outline
Introduction: Solar Energy: Concurrent Technologies?
PV Inverter overview of improvements
cost improvements
active components
approximate efficiencies
improve efficiency
topology
topology for large power
inductive components
system efficiency
suggestions for efficiency improvements
efficiency benchmark
improve reliability
direct DC use in vehicles
Solar Trackers
Conclusion
Solar Energy:concurrent
technologies?
Concentrated solar (CSP)
PV
Grid connected
inverters
PV
Stand alone use
AC or DC use?
converters
Solar thermal
-Water
-Air
-Absorption cooling
Stored
- Thermal
- Biomass
- Methanol
(With C02+ PV)
Introduction
3/30
CSP Concentrated solar power
+ Classical thermal
technology
+ possiblestorage in
molten salts
- Sunlight:
70% direct,30% scattered
- CSP cost: 4.8$/ peak
watt?PV 1.5 $?
Fig. 1. A project in California, “catching the sun“
177 000 mirrors, 377 MW
Introduction
4/30
European Supergrid
Introduction
“The Medgrid together with Desertec would serve as the backbone of the European Supergrid”
DESERTEC
en.wikipedia.org/wiki/Medgrid 5/30
Surface use in PV desert area (Tamanrasset), horizontal.
200W/m2 [*]= 487kWh/m2/year
For 50% used in PV panels at 14% efficiency (PV + converter)
200*8760*10^4*0.14/3600/1000*0.5=
=341 MWh/ha/year
Surface use of energy from Biomass in fertile areas1ha, 19 ton max. dry material
(miscanthus, sugar beet)70 MWh thermal/year/ha
25 MWh/year/ha at 35% efficiency to electricityStill a high price for the possibility to store energy!
Introduction
6/30
* http://www.umc.edu.dz/vf/proceeding/sigcle-2010/pages/themes/energies_renouvelables/Session%20III/35%20Fekih%20Sess%203.pdf
PV Inverter: overview of improvements
PV inverters,
converters,
Improve component efficiency?
Improve converter topology?
Improve reliability?
MTBF
MTTR
DC use?
Cost?
now 250-500 €/kW peak
Lifetime 10 years compared to 30 for PV panels
7/30
PV Inverter: cost improvements
Cost of inverter = major part?
now 250-500 €/kW peak
Lifetime 10 years compared to 30 for PV panels
PV panel
Cost of ownership
1000
[Euro/kWp]
2000
Inverter
0 10 20 30time [Years]
8/30
PV Inverters, active componentsTable 1. Three “well chosen” components
SiC MOSFETSCH2080KEC+ internal SBD
Si MOSFETIXFK64N50P
Si IGBTSGL160N60UFD
Ron [ ]25°C150°C
0.0800.150
<0.0850.216
0.0100.010 (125°C)
Vdrop [V] 0 0 1.1 (25°C-125°C)Vmax [V] 1200 500 600Qrr 25°C
150°C@Idcref
60nC 60nC*@ 10A
600nC2400nC**@25A
112nC420nC@25A
Etot [mJ]@ 25°C@125°C
I1.1
0.036 ***0.036 ***@300V, 10A
I1
0.360 ***1.44 *** @300V, 25A
I1.4
0.57 @25A0.725 @125°C@300V, 40A
Diode dropdiff R@100°C
0.9 V 0.06
0.6V0.004
0.8V @100°C0.02
Price/unit@100
25.75$ 9.19$ 7.23$
* the temperature influence of Qrr in SBD is almost negligible [10]** calculated on 4x more at 150°C*** calculated using (1) with K=2 and Vdc=300V
dcref
dc
dcref
dctotreftot
V
I
V
VEE
9/30
PV Inverters, active componentsTable 1. Three well chosen components
0 10 20 300
20
40
60
80
100
W
Psc I f( )
Ps I f( )
Pig I f( )
I
0 10 20 300
20
40
60
80
100
W
Psc I f( )
Ps I f( )
Pig I f( )
I
Fig. 3 Component losses at 25 kHz Fig. 4 Component losses at 125 kHz
Fast IGBT and ultrafast diode.
The SiC would like components in parallel, but at an even higher cost.
MOSFETS are good in soft switch topologies
Slow IGBTs (not shown) good slow switching multilevel topologies.
Vertical: Power in Watt, horizontal: current in A, 50% duty ratio, 300Vdc
sc = SiC MOSFET s= Si MOSFET ig=Si IGBT
10/3
0
PV Inverters, approximate efficiencies
Fig. 5. Approximate efficiencies of parts in the converter for a total of 6% loss
An example of total losses = 6%, often 3 converters in cascade. from grid to PV:
1.5 % DC-DC conversion: adapting voltage
2 % galvanic separation (soft switch, but transistors with high Vpeak and Ipeak
2.5% DC-AC conversion (PWM switching loss and EMC filters)
11/30
PV Inverters, improve efficiency?
Still improve active component efficiency?
Fig. 6. RCD (residual current device) to protect against leakage currents for safety.
- Less losses than galvanic separation
- Also ”Galvanic separation” still has some capacitor coupling
RCD
Inverter
12/30
PV Inverters, improve efficiency?
Still improve active component efficiency?
Fig. 7. Using a combination of MOSFETs and IGBTs to reduce
switching losses,
- Fast MOSFETS with slow diodes
- Fast diodes in IGBT
13/30
Without galvanic separation?
“Advanced ferrite material for photovoltaic systems”
http://www.powersystemsdesign.com/inverterefficiency1?a=1&c=4257
Problem in case of fire, risk of radio disturbance and residual current detectors
PV Inverters, improve efficiency?
Fig. 8. Converter to single phase without galvanic separation
14/30
PV Inverters: topology
Fig. 9. Filling in converter parts in a scheme (for simplicity drawn in single phase)
11
2
3
1
http://solar.smps.us/grid-tie-inverter-schematic.html 15/30
PV Inverters: topology for large power
Fig. 10. 1.5MW Multilevel DC/AC with in large PV
Fig. 13. Line to line voltage before filtering
16/30
PV Inverters: topology for large power
Fig. 11. 1.5MW Multilevel converter with boost converter in wind energy
17/30
PV Inverters: inductive components
Better design of inductive components?
http://www.powersystemsdesign.com/inverterefficiency1?a=1&c=4257
Ferrite core components are good at high
switching frequencies sized at 3kW
109 × 55 × 115 mm³. (689cc) http://www.powersystemsdesign.com/inv
erterefficiency1?a=1&c=4257
18/30
PV Inverters: inductive components
Better design of inductive components?
Similar purpose, but amorphous cut core for input and for output
inductor sized at 3kW, the saturation inductance is much higher.
71 x 45.5 x 101.3 mm³ (327cc) half of the volume
19/30
PV Inverters: system efficiency
The losses of wires, cabling, monitoring?
Use lower current/mm2 than in normal cabling. Typical 4mm2
up to 10A. One can compare investment in copper or more PV.
Avoid normal computers for measurements or monitoring
50W = 1.2kWh/day, = about 500W of solar panel installed for
nothing….
20/30
PV Inverters: suggestions for efficiency improvements
Overview of improving efficiency?
Better design of inductive components
Better materials Nanocrystalline, ferrites, better understanding of excess losses (eddy current loss,
DC bias, gaps, mechanical strain etc…)
Lower switching losses?
MOSFET IGBT combination, SiC or GaAs diodes
Without galvanic separation?
= one conversion less, but less safe, EMC problems
Without DC-AC stage?
AC is not necessary for storage in batteries of
electrical vehicles, PV at parking places
Without DC-AC stage?
Information technology,
Communication, computers are moving
from 48Vdc to 300Vdc.
More availability? Better/more HR needed in research and application.
Not just ‘hunting’: make studies research and job attractive
COST
EFFICI ENCY
COMPLEXITY
Fig. 15. The control compromise
21/30
PV Inverters: efficiency benchmark
Reported efficiencies
Remarks:
- With or without transformer = With or without insulation
- Parallel tandem working for higher efficiency at low power
http://labs.ti.bfh.ch/fileadmin/user_upload/lab1/pv/publikationen/wrt_dresden_06.pdf 22/30Source: Berner fachhochshule Switzerland, site not active any more
PV Inverters: improve reliability
Overview of improving reliability, causes of failure?
Not perfect design of undervoltage lockout of internal supplies.
Overcurrent protection, fast, also active if the processor hangs?
Surges or overvoltages on the grid due to switch operation or short circuit
surges in the grid.
Current/voltage surges by lightning,
Short circuits from the load to the ground.
ESD in manufacturing and handling
Not enough back up protection if a processor hangs.
Vandalism, sabotage, stealing
23/30
10 kW peak installed can generate 1750 kWh/year in a lot
of sunny regions in Africa.
A new generation economic electric car could use 12.5
kWh/100km which corresponds to 14 000 km year.
Ultra light vehicles need rather 2.5 kWh/ 100km and
would need only 2kW peak installed to reach a similar
distance, see
With bad roads, airco needs and storage losses and
retrofit vehicles one would need rather 25 kWh/100km,
which would correspond to 7 000 km per year for 10kW
peak installed, but it is still possible.
24/30
PV Inverters: direct DC use in vehicles?
Ultra light electrical vehicle
“ELBEV” concept
Ecologic low budget electric vehicle
Solar Trackers
Solar tracker? Project UGent electromechanical engineering by
3rd year students 2012 “interdisciplinary project
Diffuse light yield increases low inclination.
Morning and evening yield increases by rotating towards the sun.
The yield increase by trackers is sensitive to the utilized model:
Absorption by atmosphere, pollution, diffuse light model, weathers statistics.
10% increase in Belgium up to 20% subtropical regions?
Single axis Trackers could already get 90% of dual axis trackers
Vertical axis two axes polar axis
+ correction
25/30
Solar Trackers
Solar tracker? Project by 3rd year students 2012,
1D control reduced panel of 15W, simulating forces of 1m2 panel.
Control: 6….30V with PWM controlled DC motor, good for “12V” panel
Stand by current of the control:10mA , so very low own consumption.
Bill of materials <50 Euro for control, motor and gear is possible for 1D
May be economic for about 2m2
Control PCB 40mm*80mm, single side,
26/30
Solar Trackers
Solar tracker? Project by 3rd year students 2012, projects under
construction
Fig. 16. Solar tracker as a third year student project at EELAB UGENT
27/30
Solar Trackers
Solar tracker:
Could be discussed for lifetime and cost effectiveness
But generates more in the morning and evening
Or some panels eastwards and some westwards oriented?
28/30
Conclusion Cost effectiveness of solar inverters?
PV is more cost effective than solar thermodynamic and biomass if storage is not needed
Converters may cost more than the PV panels in total cost of ownership
Good active components exist, also in Si-IGBT types
Magnetic materials : amorphous and nano-crystalline materials can reduce cost and losses
Human are resources needed for improvement of MTBF and MTTR?
Solar tracker systems?
The cost has to be compared with the increased yield and own consumption
Student projects are used to generate modeling and mechanical arrangement ideas
Topology
Low power without galvanic separation has a higher efficiency but needs RCD for protection
MOSFETS for soft switch, fast IGBT for hard switch
Slow IGBTs for large power plants with multilevel
29/30
REFERENCES
[1] Mark Crawford, “Catching the sun” , Mechanical engineering March
2013.p33-37, https://www.asme.org/getmedia/44edaee0-d607-4ec4-b241-
1b7877bdbd01/Catching-the-Sun.aspx
[2] Medgrid project http://en.wikipedia.org/wiki/Medgrid
[3] Daniel Lynn, “Solar Powered Air Conditioning” http://www.machine-
history.com/Solar%20Powered%20Air%20Conditioning
[4] Van den Bossche Alex, Bart Meersman, 3Combined Heat Power System”
Patent WO2009007408
[5] Rik De Doncker, “Power Converter For PV-Systems”, ECPE Seminar:
Renewable Energies, PV farm, Ferbruary 9-10, IEST, Kassel Germany.
[6] Vencislav C. Valchev, Georgi T. Nikolov, Alex Van den Bossche, Dimitre
D. Yudov, Power losses and Applications of Nanocrystalline Magnetic
Materials” ICEST 2007, Ohrid 24 - 27 June 2007 4pp
[7] Vencislav C. Valchev, Alex P. Van den Bossche, David M. Van de Sype
“Ferrite losses of cores with square wave voltage and DC bias” IEEE 31th
Annual conference of the Industrial Electronics Society, IECON, November 6-
10 2005, Raleigh, NC, USA, pp. 837-841
[8] Alex Van den Bossche, Vencislav Chechov Valchev, Inductors and
Transformers for Power Electronics, February 23, 2005, CRC-press, Boca
Raton USA, ISBN 1574446797, hardcover, 480 pages.
[9] Alex Van den Bossche, “Inductive Components in Power Electronics”,
keynote paper, 33th International Telecommunications Energy Conference,
IEEE-Intelec 2011, 9-13 October 2011, USB-stick, 11pp.
[10] Anant Agarwal, Ranbir Singh, Sei-Hyung Ryu, James Richmond, Craig
Capell, Scott Schwab, Brice Moore and John Palmour, “600 V, 1- 40 A,
Schottky Diodes in SiC and Their Applications”
http://creepower.com/products/pdf/PWRTechnicalPaper1.pdf
[11] Miles C. Russell, “The Promise of Reliable Inverters for PV Systems: The
Microinverter Solution”, GreenRay Solar: June 18, 2010,
http://archive.is/9wrdE
[12] Chaz Andrews, "Solar Inverters incorporating RCM units - Why do you
need an RCD as well?" http://www.eponthenet.net/article/53565/Solar-
Inverters-incorporating-RCM-units-Why-do-you-need-an-RCD-as-well-.aspx
[13] Van den Bossche, Alex; Valchev, Vencislav; Marinov, Angel, “Reducing
switching losses through MOSFET-IGBT combination.” Proceedings XLIII
International Conference on Information communication and energy systems
and technologies (2008) 4pp.
[14] Jaime Alonso-Martı nez, Joaquı n Eloy-García, Santiago Arnaltes,
“Direct power control of grid connected PV systems with three level NPC
inverter”, Solar Energy, Volume 84, Issue 7, July 2010, Pages 1175–1186.
[15] Georgios A. Adamidis, Thomas G. Nathenas, Athanasios D. Karlis,
“Comparative investigation and improvement of wind farms based on wind
energy conversion and grid connection methods”, EPE’13 ECCE Europe Lille,
3-5sept 2013, 8pp.
[16] Alex Van den Bossche UGent, Peter Sergeant UGent and Isabelle
Hofman, “Towards low energy mobility using light and ultralight electric
vehicles”, First International Conference On Electromechanical Engineering,
Proceedings, keynote 2, Skikda Algeria, Nov20-21,2012, 9pp.
[17] Koen De Gussemé, DM VAN DE SYPE, Alex Van den Bossche UGent
and Jan Melkebeek, “Input-current distortion of CCM boost PFC converters
operated in DCM”, IEEE transactions on industrial electronics, 2007, vol 54,
issue 2, p.858-865.
[18] Alex Van den Bossche UGent, Peter Sergeant UGent and Isabelle
Hofman UGent (2012) First International Conference On Electromechanical
Engineering, Proceedings. Nov 20-21,2012
[19] Chris Turner, “The solar industry's Apple-sized ambitions”
http://www.mnn.com/earth-matters/energy/blogs/the-solar-industrys-apple-
sized-ambitions Apr07, 2011
[20] DESERT STAR – Solar Photovoltaic panels for very hot areas
http://www.solar-trackers.com/specific-pv-applications/high-temperature-pv-
panels
31/30