Interview with Dr. Lewis M. Fraas president JX Crystals

CPVToday.com Special

 It was almost five years ago when Dr. Lewis M. Fraas president of JX Crystals had termed the perception about solar energy being intrinsically too expensive to compete with the traditional forms of electric power generation as "wrong".

 Reflecting on the current status Dr. Fraas whose company is doing R&D on 40 percent efficiency solar cells and systems including a new Cassegrainian concentrator concept with distinct advantages for 4 junction PV stacks says demonstrating 44 percent "will be very straight forward using the dual focus Cassegrain geometry."

 However speaking with CPVToday.com he said "Commercialisation will still take a long time and lots of money "

Dr. Fraas' association with CPV

 It was Dr. Fraas's group at the Boeing HiTech Center that demonstrated the 35 percent MJ cell at Boeing in 1989.

 "We left Boeing with a licence on this to JX Crystals in 1992. I also wrote the first paper on the InGaP/GaInAs/Ge triple junction cell in 1978 predicting an efficiency of 40 percent at 300 suns " reminisces Dr. Fraas who is scheduled to speak during CPV summit, USA 2009 (scheduled to take place on February 3-4 in San Diego).

 "JX Crystals is actually an early pioneer in LCPV having started work in this area in 2003 well before OPEL or SV Solar started their activities. Our LCPV is patented based on a thin aluminum sheet heat spreader at the back of a traditionally laminated module with cut-up 1 sun cells with linear optics. Our LCPV modules have been deployed in systems in China and in the US and operating successfully for several years now. Our patents cover the use of either linear mirrors or linear lens geometries."

"If OPEL or SV Solar were to use metal back sheet heat spreaders with cut-up cells and linear optics they will violate our patents. If they do not use a metal back sheet heat spreader their modules will operate at higher temperatures decreasing their module performances relative to ours " says Dr. Fraas.

Concentrator cells have been reaching increasingly impressive efficiencies inspiring new interest in the high-efficiency high-concentration approach. Currently the record efficiency is 40.7 percent for a three-junction GaInP/GaInAs/Ge cell.

 Commenting on such developments Dr. Fraas said JX Crystals sees the LCPV approach as fast to market with minimal risk.

 "It is a simpler approach to understand in terms of reliability and O&M because it is evolutionary from the traditional planar silicon module. Because of its simplicity LCPV may be more suited to commercial building flat rooftops. For the longer term JX Crystals is also involved in HCPV. We have a dual focus Cassegrain approach where the combined cell efficiencies can easily hit 44 percent. Our view is that while the 35 percent monolithic cell is attractive because of the penalty of DNI vs global (78 percent) and the optical losses (85 percent) a 23 percent silicon planar cell competes (0.78x0.85x35 percent=23 percent). However in the longer term 44 percent provides a real significant advantage."

Three domains for CPV

 It is said that CPV follows a complementary approach and uses concentrating optics to focus the light onto small cells. The optics may be designed for low or high concentration. Low-concentration concepts use silicon or other low-cost cells; high-concentration optics may use more expensive higher-efficiency cells. The higher-efficiency cells can reduce the cost per watt if the cost of the small cells is minimal.

 But Dr. Fraas says such categorisation of CPV options is over simplified.

 "I think there are really three domains for CPV. They are LCPV which I define as less than 5 suns MCPV (mid) which I define as 5 to 100 and HCPV which is >100.

Why do I make this partition?

I think it is my experience over 30 years with cells packaging for thermal management and environmental durability and CPV indoor and outdooor testing " said Dr. Fraas.

 "Every new comer in CPV starts with MidCPV. My view is that this is the worst place to start.

Why?

You can't use screen printed cells and cells are not available. Packaging is difficult for durability " he said.

 Explaining further he added: "We worked with another CPV company on this for 1.5 years in 2004-05.

We are working with a second CPV company on this. We found a special cell supplier for them in China where screen printed cells with 50 percent grid covers are used with silicone prismatic covers. And then there is Euclidides and there is the original Martin Marietta system where the 2" diameter cells on alumina substrates eventually failed in the field in the 80's." (He pointed out that the SAI Incubator program just awarded a contract to work in this domain and the DOE SBIR solicitation called out this domain."

 For HCPV according to Dr. Fraas one can use small cellls with surface mount technology.

 He further elaborated on LCPV's definition in a following manner:

€¢

One can use screen printed single crystal or multi-crystal cells in high volume production.

€¢

One can use EVA for encapsulant polypropylene for voltage isolation and glass for weather proofing.

€¢

One can use standard 1-sun module assembly semi-automated stringing and lamination equipment or OEM with low cost labour as we did for our 4 kW and 100 kW systems.

The only addition is an aluminum sheet heat spreader in the lamination process for thermal management. There are JXC patents on all of this.

 And on LCPV vs HCPV he said HCPV has the obvious advantage of 40 percent cells.

However LCPV has several other advantages.

 

€¢

Higher optical throughput ie. 94 percent vs 85 percent.

€¢

Less sensitivity to dust accumulation.

€¢

Can use less precise lighter and simpler single axis tracker on commercila building rooftops to generate retail priced electricity.

€¢

Sees direct + 1/3 of sky around sun.

Based on our data from Las Vegas over 1 year this is 87 percent of global instead of 78 percent of global.

€¢

Fast to market because 1-sun manufacturing infrastucture is already in place.

CPV summit, 2009 San Diego

 CPV summit, 2009 is scheduled to take place in San Diego (February 3-4 2009). For more information click here:http://www.cpvtoday.com/usa/programme.shtml

 or Contact: Joshua Bull by email josh@cpvtoday.comSource: CPV Today