Last June, Canadian Solar Inc. (CSIQ) introduced a new ELPS solar cell technology that will provide monocrystalline solar cells efficiency of up to 19.5% and polycrystalline cell efficiency of up to 18%. According to CSIQ, ELPS stands for “efficient, long-term photovoltaic solution”. Yet almost a year later, at May 2011’s conference call of its first quarter of 2012, CSIQ said it shipped 5-6 MW ELPS high efficiency modules out of a 343 MW total shipments, a far cry from its previous annual target of 600 MW announced at November 2011. Given that CSIQ’s CEO had spoken about ELPS publicly several times in the past, the expectations on ELPS is quite high. Yet the production at Q1-2012 is a big disappointment.
Here is a short time line for ELPS:
June 2011 – CSIQ introduced its ELPS line of high efficiency modules, said to build a new 600 MW cell plant at Suzhou, Jiangsu Province of China.
July 2011 – CSIQ showcased its ELPS at North America Intersolar exhibition
July 2011 – CEO Shawn Qu said “we finally chose this [MWT] technology because we can have an immediate efficiency boost while at the same time we can almost maintain the same cost per watt”
August 2011 – At 2011-Q2 CC, CSIQ said it would start to commercially ship ELPS at 4th quarter
September 2011 - CSIQ showcased its ELPS at the 2011 EU PVSEC in Hamburg
November 2011 – Shawn Qu said the commercial production of ELPS started in October at an annual capacity of 50-60 MW.
March 2012 – At 2011-Q4 CC, CSIQ said ELPS sold at 10% price premium at 5% extra costs. No shipment number is disclosed.
April 2012 – CSIQ filed 2011 20-F, saying the ESE capacity is 280 MW by the end of 2011. No MWT capacity is mentioned. It said it planned to have 280 MW ELPS capacity by the end of 2012.
According to CSIQ, ELPS uses two cell technologies: Selective Emitter (SE) and Metal Wrap-Through (MWT), to improve conversion efficiency. Out of a number of technologies that can increase module efficiency, SE and MWT are two unrelated ones that can work independently. SPVI’s Robert Dydo has more details on the two here (https://solarpvinvestor.com/spvi-news/88-cell-manufacturing-process-and-future-cell-technologies). SE is a rather matured technology which can be implemented in different ways. It is relatively inexpensive and effective so CSIQ’s choice of SE appears to be natural. MWT is also an old concept which has stayed in the lab for a long time. For its ELPS, CSIQ partnered with Holland’s ECN to develop its own MWT production lines. In this case, CSIQ paid a solution fee to ECN and ECN provided the cell designs to CSIQ. Since ECN is only a research institute, it does not build the actual equipment for implementing MWT. It is up to CSIQ to come up with actual implementation, which greatly adds to its risk.
While there are no official news, there are information leaked here and there indicating CSIQ indeed has trouble ramping up its ELPS production – despite Shawn Qu has touted CSIQ’s capability in commercializing lab technologies. ELPS is still running on semi-automated production lines and the cost is much more than the “5% extra” claim. Since the premium is about 10% or 8 cents or so, CSIQ has no incentive to produce more ELPS modules until it can get the cost down.
ELPS might be a good example to highlight the difficulties to the high efficiency game. Despite there are quite a few lab technologies out there that can increase efficiency as described in Robert’s article, commercialization with effective cost control is hard. There are Taiwan cell makers commenting on MWT: there are 5-6 extra steps to make MWT cells and it is simply not worth the trouble for the efficiency gained. The more practical route might be the equipment makers to develop “turn-key solutions” for the module manufacturers. As long as the equipment makers can prove their products can help its customers, there will be a better chance for the module manufacturers to adopt. Of course, developing cost-effective “turn-key solutions” is no easy task for equipment makers either. They have to wrap the R&D costs into their products and most their offerings are too expensive for their customers to justify the investment.
