Both Sides Now: The Thin-Film/Crystalline Debate

A financial industry client recently asked whether “thin-film” PV product can ever compete with highly established crystalline PV technology based product which currently has 85% market share.  His question was prompted by a spate of recent press articles that talk about the 50% drop in module sale prices in the last year, with claims of margin pressure on the thin-film category.  A good summary of the situation, here.

Like all things related to solar energy generation, there are many factors to consider.  First, thin-film is a broad term; there are many technology types with different performance capabilities, cost points, and structures which factor heavily in a comparison to crystalline products.

First, amorphous silicon thin-film from various vendors is currently offered in both flexible and rigid glass type products. The product is generally 6 – 9% efficient (efficiency coming off production line, not to be confused with handmade laboratory efficiencies which are much higher) in converting sunlight and has a wide manufactured cost range depending on vendor and type. While a-Si is becoming less and less relevant in the marketplace due to a weak value proposition (low efficiency with high production cost), there is promising work being done with a-Si involving nanotechnology  and manufacturing process (Oerlikon for instance) to improve performance and bring costs down. Vendor examples include Sharp, Uni-solar, Sunfilm and Kaneka.

At the higher performing end of the thin-film market is product made of CIS/CIGS and cadmium telluride (CdTe). Production line CdTe is 9% – 12% efficient with highly competitive manufactured costs.  First Solar is a mature, dominant thin-film supplier with over a gigawatt of manufacturing capacity. The CIS/CIGS family of thin-film has been in development for over 20 years by a variety of companies, but only recently have companies begun to hit their large-scale commercialization stride.  With production line product in the 10% – 13% efficiency range, vendors such as Solar Frontier and others are scaling up to reach manufactured costs similar to First Solar in the very near future.

A discussion about CIGS would not be complete without mentioning start up efforts from Solyndra and Nanosolar which have different form factors and cell designs from the more conventional “flat plate” products mentioned above. Solydra utilizes a tube type form factor where the CIGS solar cells are deposited on the inside. Nanosolar’s cells are placed on individual “coupons” similar in form to a crystalline solar cell. While both of these products have intriguing stories, neither have reported competitive cost structures or performance.

Second, most thin-film PV is deployment location specific which affects internal rate of return for generation asset financiers.  While crystalline product can be 12% to 22% efficient, this solar cell type needs direct sunlight and only reacts to a certain light spectrum. Most thin-film reacts to many light spectrums and can produce useful energy amounts even when partly cloudy. Many thin-films also work well in high heat conditions where crystalline energy output suffer, such as in the desert southwest of the U.S., Northern Africa and Middle East.

In Germany (not the sunniest location), there are enormous amounts of data where thin-film produces more useful energy on an annual basis and provides financiers with stronger IRR than crystalline product in solar farms that compete with wholesale cost energy from utilities. Of course thin-film, with its lower efficiency, requires more land and more balance of systems (wiring, inverters, combiner boxes, racking etc.). Generally smaller thin-film module sizes (90 watts or smaller) require more installation cost compared to installing fewer larger 200 watt crystalline modules. The land issue can be especially problematic for utility scale solar farms using thin-film in some locations. Even with these extra costs, the thin-film levelized cost of energy (LCOE, where all project costs are taken into account) and resulting IRR is still better for project investors in locations where the above positive factors line up.

On rooftops, where PV energy is competing with retail cost energy from utilities, crystalline has had strong advantages due to the higher efficiency in a space-constrained setting. But with the price of high performance thin-film dropping and its efficiency increasing, this situation is changing as witnessed by the 500MW program from the utility Southern Csualifornia Edison. This program is aimed at bigger rooftops of150,000 ft.2 and larger; First Solar is one of the vendors for this program.

Third, there are inherent design and manufacturing differences to be considered. Crystalline is a one-by-one paradigm with many steps in the process to make one solar cell which then has to be interconnected, packaged and integrated with many other cells into a module with a glass front and metal framing.  In contrast, thin-film manufacturing uses significantly less material for the solar cells, and the deposition of the cells is usually done all at once on glass or other materials. The thin-film semiconductor is typically encased with glass on both sides, with less packaging (weatherization) time. Overall, thin-film production throughput, a key manufacturing metric, is usually considerably less, resulting in lower cost.

These manufacturing advantages have been illustrated recently with the Q4 2009 SEC filings from First Solar and a number of publically traded crystalline product providers. First Solar’s average gross margins were over 50%, while the crystalline providers were in the range of 12% – 20%.  Current oversupply issues in the market will drive gross margins down for First Solar and all other supplier types.

Clearly there are many factors to consider when answering the question about thin-film photovoltaic competitiveness in the future. Claims which contend that crystalline product is so inexpensive now that thin-film can’t compete are full of holes. Suppliers throughout the crystalline supply chain are operating at very low margins and the technology type does not compete on a system basis in many locations where temperature and shading from clouds is present. As PV industry analyst Paula Mints from Navigant Consulting said recently regarding the minimal margins, “this is not sustainable.” Crystalline producers, whether Asia based or elsewhere, will eventually need gross margin improvement via higher average sales prices.

Crystalline product, especially high performing product from companies like Sunpower Suntech, Trina and others, will always enjoy large market share, as it has applications unique to its performance. But a few high performance thin-film companies will increase market share in the coming 5 years as manufacturing scale is achieved.  These thin-film cost and performance leaders will have the luxury of high gross margins to compete from a position of strength on a project by project basis.

Share this:

Comments are closed.