Mirror based spectral splitting CPV system

MIRROR BASED SPECTRAL SPLITTING CPV MIRROR BASED SPECTRAL SPLITTING CPV SYSTEMSYSTEM

Antonini A.1, Butturi M.A.1, Di Benedetto P. 1, Milan E. 1, Uderzo D. 1, Zurru P. 1,Sartore D.1, Parretta A. 2,3

1 CPOWER SRL, via Traversagno 33/3, 44122 – Ferrara (Italy)2 ENEA C. R. “E. Clementel”, Via Martiri di Monte Sole 4, 40129 Bologna (Italy)3 Università di Ferrara, Dipartimento di Fisica, Via Saragat 1, 44122 Ferrara (Italy)

Martedì 21 Settembre 2010 XCVI Congresso Nazionale SIF, Bologna, 20 - 24 Settembre, 2010

OUTLINE:

- CONCEPT

- DEVELOPMENTS

- ACHIEVEMENTS

- CONCLUSIONS


GoalGoal

< 3.5 < 3.5 €€//WpWp

< 2.2 < 2.2 €€//WpWp

-- Industrializing Industrializing designdesign

-- Cost effective Cost effective solutionssolutions

-- ReliabilityReliability

-- EfficiencyEfficiency

MultiMultiMultiMulti----APprOachAPprOachAPprOachAPprOach for high efficiency for high efficiency for high efficiency for high efficiency integrated and integrated and integrated and integrated and inteLLigent inteLLigent inteLLigent inteLLigent

cONcentratingcONcentratingcONcentratingcONcentrating PV modules (Systems)PV modules (Systems)PV modules (Systems)PV modules (Systems)

Work developed in the framework of The European APOLLON Project

THE GOAL OF THE PROJECT FOR THE MIRROR BASED SPECTRAL SPLITTING SYSTEM MBS3 APPROACH IS:

Next YearNext Year

In 3 In 3 YearsYears


The first approach:The first approach:SJ sc with different Band Gaps SJ sc with different Band Gaps

- Concentrate the light using geometrical optical solutions based on reflection

- Split of the spectrum using interference of the light (dichroism)

Advantages:

- No necessity of MJ SC- Current Matching avoided-“Distributed” cooling- different concentration levels

for different kind of cells

Problems:

- Many optical interfaces- multiparts system � complexity

Spatial & Spectral Spatial & Spectral SplittingSplitting

T % @ 20°of incidence angle

Wavelength (nm)


Starting point

11°° ProofProof of of ConceptConcept

-Open, modular structure- “Large” dishes

- high profile of the moduleArms give shadowing

“Large” dishes suffer of form errors producing optical losses at the second

receiver and current mismatches

Secondary optics

Dichroic mirror

CPower has focused its HCPV activity on the design o f concentrators where it’s possible to use a spectral splitting of the su nlight


Silicon Receiver

Dichroic Mirror

InGaP Receiver

Si Dense Array:The geometrical concentrationinside the yellow square is100x

InGaP Dense Array:The light concentration on the cells is 300x

Starting point

Non-uniform light on the cells in the modules, mainly due to different dishes deformations. It gives a stair shaped I-V

curve for Si receivers and high optical lossesat secondary optics


Light non-uniformity on the cells in the modules, mainly due to different dishes deformations, gives a stair shaped I-V curve

Starting point

Problem:


Design evolution

“Large” dishes (420mm x 420mm) Small dishes (110mm x 110mm)

Dense Si and InGaP arrays Subreceivers composed of 1 Si Cell and 1 InGaP cell

Open StructureClosed Box Structure

Dedicated heat sinks requiredDedicated heat sinks NOT required

With structural arms (shadowing) Without structural arms

-100x

-400x-40x

-600x


Design Evolution - 1

- Large symmetric dish with heavy heatsinks

- Flat dichroic mirror reflecting on a ligthpipe

- Small symmetric dish with mechanical structures working as heatsinks

- A lens improves the angular tollerance before the dichroic mirror

- The lightpipe is connected to the lens


Design evolution - 2

Concentrating parabolic dish (ReflectionReflection )

Lens to increase optical tolerance (RefractionRefraction )

Flat dichroic mirror (IntereferenceIntereference )

Glass Rod (Total Total internal Reflectionsinternal Reflections )

Patent Pending design


Silicon Cell

InGaP Cell

-- AllAll the the componentscomponents are are connected withconnected with the the rearrearside of the side of the modulemodule ((thermallythermally and and mechanicallymechanically))--No No shadowingshadowing

Design evolution

Patent Pending design

Point focus at the inlet of a quartz rod

With Ideal Surfaces,

optical eff. of 98%98%

- Concentrating parabolic dish (ReflectionReflection )

- Lens to increase optical tolerance (RefractionRefraction )

- Flat dichroic mirror(IntereferenceIntereference )

- Glass Rod(Total Total internal Reflectionsinternal Reflections )


Module Evolution

Step 1

Step 2

Step 3

Single Concentration Unit

Series of 5 Units

Tests on Optics and Circuits

Tests on a closedbox:- Thermal - Optical- Mismatching

Small Module of 15 Units (30W)

Final Tests


External conditions:- Tamb =35°C- DNI: 850W/m2- No wind- Free convection

TmaxTmax in the module: in the module: 5555 °°C C

Thermal Modelling

Thermal simulated performances (Finite Element Analysis)

Analysis performed by TECNALIA

Representative simulated sample

Receiver


LGBG (Laser Grooved Buried Grid)cells from Narec (19% of eff.)

InGaP solar cells from ENE (15% of eff.)

Receiver

Receiver under the Sun

Receiver without heatsinkReceiver without heatsinkFor opticalFor optical//electrical testselectrical tests ((FlashedFlashed))

Tests on Optics and Circuits

Single Concentration Unit


Receiver

With this filter and these solar cells:

Si sc power production:66% respect to total spectrum

InGaP sc power production:91% respect to total spectrum So, theoretically , with this filter and

with optical efficiency of 84 % on Si and 80% on InGaP:

Absolute eff. of Si: 0.84 x 0.66 x 0.19 = 10.5% Absolute eff. of InGaP: 0.8 x 0.91 x 0.15 = 11%Total efficiency = 21.5%

The The filter should havefilter should havea bit a bit longer cutofflonger cutoffwavelengthwavelength


Note*: The receiver substrate is not connected, in this testing prototype, to the larger heatsink (module rearside) as considered for the complete module. The cell temperature here is around 55°C. With 25°C, the“a bsolute efficiency contribution” for the module from Si is of about 10.4%

Si cell:

Isc: 2 AVoc: 630 mVFF: 71%

Eff.: 9.2% (eff. Contribution from this part of the concentrator, not cell efficiency…)

Note*

Si Receiver

DNI: 807 W/m2

Optical eff. on Si: 84%


Si Receiver – Angular Acceptance

The measured angular tolerance for the Si fits well the forecasted


InGaP Receiver

Problems arise with high concentration…

DNI: 812Isc: 0.49AVoc: 1.38VFF: 84%

Optical eff. on InGaP: 44%...Efficiency contribution

from InGaP: 5.8% (instead of 11%)


InGaP Receiver –Angular Acceptance

The measured angular tolerance for the InGaP Cells is largely higher than the forecasted

Larger focus area in the inlet of the quartz rod (Lower Efficiency –

Higher angular tolerance)


InGaP Receiver

Prototype optics (primary and secondary) directly machined (not moulded) � not perfect surfaces � focus enlargement

Larger focus area in the inlet of the quartz rod (Lower Efficiency

– Higher angular tolerance)

Slightly deviated ray from idealSlightly deviated ray from idealdirection are direction are lost withlost with High High ConcConc


In the range 900 – 1200nm, the Silicon produces about the 27% of its current under the indicate filter. So, it gives about 5.4% points of efficiency (20% under full spectrum). With 83% of Optical eff., it gives about 4.5% of eff. on full AM1.5D spectrum.A DJ solar of � = 31%, with 79% of Opt. Eff. contributes with about 24.5% .Globally, the module could achieve an � = 29%, with currently available technology

The current generated by the Ge with a proper filter doesn’t limit the TJ series; using a 39% TJ and a 20% Si SC could

be possible to achieve 34% for the module efficiency

MJ @ High Conc + Si @ Low Conc

Possible Evolution

Possible Evolution


Tracker

Mechanically designed by CPower

Electronic control by Tecnalia

Tracking error < 0.2° (Measured by Tecnalia) Tracker for 1.5kWp of dichroic modules

Cost for 1MW < 1 Euro/W

WithWith high high ConcConc ., the ., the pointing precision ispointing precision is FUNDAMENTALFUNDAMENTAL


ConclusionsConclusions

� Dichroic mirror based modules evolution to produce a competitive product (not just a demonstrator)

� Design with geometrical optical efficiency of 98% and angular acceptance of 3° on Silicon (geometrical 40x) and of 1.5° on InGaP (geometrical 600x)

� Design for a maximal temperature of 55°C on the receivers (simulated) connected to the rear side of the module

� Electrical efficiency on DNI of 16.2% on the first testing prototype

� Tracker for 1.5 kWp with pointing precision < 0.2°

� Possible module evolution to achieve 29% - (34%) of efficiency with currently commercially available technology


AcknowledgmentsAcknowledgments & & ContactsContacts

A particular acknowledgment to the partner of the Apollon project who have directly participated to this work:

- Eduardo Roman Medina (Tecnalia)- Ekaitz Olaguenaga Etxebarria (Tecnalia)- Michael Noak (ENE)- Keith Heasman (Narec)- Gianluca Timò (RSE – Apollon Project Coordinator)

Contacts:[email protected]

[email protected]

The work has been partially supported by the Europe an The work has been partially supported by the Europe an Commission under the Grant Agreement N.213514 (APOL LON Commission under the Grant Agreement N.213514 (APOL LON Project) in the Seventh Framework Program.Project) in the Seventh Framework Program.


Date post:	13-Nov-2023
Category:	Documents
Upload:	independent
View:	1 times
Download:	0 times

Mirror based spectral splitting CPV system

Documents