The PV industry experienced tremendous supply demand imbalance through the value chain in 2011 which was exacerbated by incentive adjustments in major solar markets and austerity measures in Europe.
2013 was tough for solar manufacturers. They were forced to reduce prices, decrease margins, close some manufacturing facilities, or even declare bankruptcy. In the 2013 solar market we saw increased manufacturing capacity but we also saw higher silicon supply, lower demand, oversupply and limited credit availability. Perhaps most significantly solar module prices experienced a sharp decline in part due to low cost Chinese PV cell manufacturers.
However, lower module prices have helped reduce the price of solar energy, making solar more competitive with other forms of electricity generation. Today’s manufacturing cost per watt can be as low as US$0.50. These prices can be expected to keep declining particularly as US homeowners are buying solar panels.
A still deteriorating global economic situation, government belt-tightening and a number of corporate incidents left the industry in an unfamiliar, unhealthy state at the end of 2011 but it was also the year of grid parity across many markets. With the end of the year fast approaching, it is time for a change.
These awkward market factors will have to be answered and many industry players are, thankfully, already taking steps to do so. Overcapacities, in particular, will present many companies with significant problems, as will the SolarWorld trade case and continuing falling prices. Increasing diversification and the proliferation of market penetration, consolidation and fast-approaching grid parity are the major upsides that will allow photovoltaics the opportunity to eventually emerge from the very real challenges it is currently facing.
Crystalline silicon technologies
Crystalline silicon is currently used in over 80% of PV cells. Despite the lower potential cost of emerging thin-film technologies (less material, higher theoretical efficiencies), crystalline silicon is expected to remain the dominant PV technology for at least the next 10 years due to its higher current average efficiency and its greater stability.
Monocrystalline (mono c-Si): Cells made from single silicon crystals offer a higher level of efficiency than polycrystalline cells, due to the lack of grain boundaries that impede the flow of electrons. Offsetting this benefit is the material’s greater cost, caused by the more advanced ingot-casting process required in production. To create a monocrystalline cell, a crystal ‘seed’ is placed into molten high purity silicon. The seed is then slowly pulled from the liquid, allowing a solid ingot to ‘grow’ from the seed as the material cools.
Polycrystalline (poly c-Si): A polycrystalline ingot is created through a simple casting process. While typically less efficient than a monocrystalline cell, polycrystalline cells benefit from their lower relative cost and their ease of production. As a result of this cost/efficiency trade-off, the per-watt cost of electricity tends to be similar for both cell types.
Ribbon technology: The basic ribbon process involves pulling a thin layer of silicon from a crucible of molten silicon in a continuous process. Of the different forms of ribbon technology currently under development, Evergreen Solar’s string ribbon and RWE Schott’s edge-defined-film-fed growth (EFG) appear to be the most advanced. The main advantage of ribbon technology is the reduction of silicon usage and silicon waste (caused by the sawing of ingots into solar wafers). Material loses during the sawing process can reach as high as 50%. This waste level can be reduced to 10% under ribbon technology methods, according to RWE Schott. Raising the efficiency level and obtaining economies of scale in production remains key to the technology’s longer-term success.
A technique borrowed from the semiconductor industry, the term ‘thin-film’ refers not the thinness of the material but to the method of building up the cell using numerous thin layers. In contrast to a typical crystalline silicon cell, a standard thin-film cell does not have a metal grid for its top electrical contact. Instead, a layer of a transparent conducting oxide (commonly tin oxide) is used.
The potential benefits of thin-film technology include the use of less material, the ability to be manufactured in complete modules (not individual cells) and the option of depositing the cells on to flexible substrate materials (to be used in building integrated products). The drawback (at least to date) is the technology’s lower level of efficiency and stability compared to that achieved with traditional crystalline technology. After many years of development, thin-film technologies are beginning to move from the laboratory to the manufacturing line.
Amorphous silicon (a-Si): An amorphous solid is one in which the atoms are not arranged in any particular order. While this disorganized structure limits the ability of electricity to flow through the substance and thus the cell’s efficiency, it has the advantage of absorbing solar radiation 40 times more effectively than a monocrystalline cell. Hydrogen is often added to the silicon film to increase efficiency. Such cells benefit from being able to be produced at lower temperatures and deposited on cost-efficient substrates (plastic, metal or glass for instance). This technology is commonly used in solar-powered consumer devices that have low power requirement characteristics. Combining amorphous silicon with microcrystalline silicon is yielding improved results.
Cadmium Telluride (CdTe): Currently the simplest thin-film technology to manufacture, Cadmium Telluride can be produced using an array of common industrial processes that do not require expensive capital equipment. Although the amount of Cadmium in the product is small (c. 0.033% of the module by weight according to First Solar), fear of its toxicity factor remains a major drawback, despite studies to refute it. This is particularly the case given the increasing importance of recycling initiatives in many countries (e.g. the Restriction on Hazardous Substances (RoHS) legislation in the EU). BP Solar, one of the early backers of the technology, closed its CdTe facility at the end of 2002.
Copper Indium (Gallium) Diselenide (CIS/CIGS): The biggest asset of this compound is its facility to absorb light – 99% of solar energy is absorbed in the first micrometer of the material. The compound also exhibits remarkably stable long-term conversion efficiencies. CIS typically contains copper, indium and selenium, with higher efficiencies gained through the addition of gallium. The downside of this particular thin-film technique is the use of indium, which is relatively limited in supply. Shell Solar has been a notable proponent of CIS technology, though others are also in the early stages of development. While potentially interesting for use in certain applications requiring high efficiencies (e.g. space applications, concentrator systems), supply restrictions are likely to prevent CIS becoming a significant proportion of long-term PV industry supply.
Gallium Arsenide (GaAs): The technology has mainly been used for applications requiring high-efficiency solar cells. While benefiting from a very high absorption rate and low sensitivity to heat and radiation, cost has remained a significant headwind. Gallium Arsenide produces some of the most efficient solar cells with efficiencies of up to 30%. However, with Gallium being rarer than gold and Arsenic being poisonous, such performance comes at a cost.
At the module level, cadmium telluride is proving to be a disruptive technology. First Solar Inc. has shown that in volume production the module cost/watt of CdTe technology can be roughly half that of average crystalline silicon. This is a long term sustainable advantage as both technologies progress down cost reduction curves. However, this is only part of the story. When the analysis is extended to the system level, the 50% cost advantage that CdTe has at the module level shrinks to roughly 10%, because balance of system costs (i.e., other than the modules and inverter) are proportional to the difference in conversion efficiency. So, a higher-efficiency, higher module cost, crystalline silicon-based system will have a significantly lower balance of system cost, and hence a competitive system cost/watt.
Material properties for different technologies result in modest variations in power output for a given amount of sunlight. Some thin-film technologies offer less pronounced temperature effects (i.e., % power degradation/°C that the solar cell operates above standard test conditions), and a wider spectral response (i.e., the absorption of a broader wavelength range of light). And, under certain conditions, these thin-film approaches can generate roughly 5% more power than an identically configured c-Si system.
With more than one viable technology, several variables including space, sunlight, and market efficiencies will determine which technologies are best suited to specific situations. Crystalline silicon will likely win in area-constrained applications in which power density is critical, and CdTe will likely lead a competitive run-off in a non-area constrained, lower-sunlight application. Based upon real data, and reasonable assumptions for technologies soon to enter production, cost/kWh for several solar PV technologies are clustered within 10% of each other. Despite claims of clear-cut winners, the economics of solar PV-generated electricity points toward more than one technology solution.
Looking into the future, we cannot ignore new technologies that could conceivably marry the conversion efficiency of crystalline silicon with the cost profile of thin films, offering game changing products. These technologies are developing the use of new materials, or novel approaches with existing materials like silicon. Suffice it to say that the solar PV industry will marry a host of complementary technologies to address an enormous market opportunity. Over time some will likely become obsolete, but for the near-term it seems clear that thin-film solar PV technologies will not vanquish crystalline silicon, but rather coexist competitively.
With the technology landscape pretty well defined over the next few years, economics becomes the focus. It is a simple fact that incentives drive the industry today, that standalone economic viability is still several years in the future. What is often underestimated, though, is the true value of solar PV generated electricity. Solar PV competes with grid supplied electricity when grid electricity is most expensive -- at peak power times. And an assessment of the true cost of grid supplied electricity by location and time of usage indicates that the elusive goal of grid price parity for solar PV may be closer than is commonly believed.
Solar PV electricity will become economically attractive without subsidies. It is simply a matter of time, and no technical breakthroughs will be needed to achieve this. Ultimately the cost of solar PV-generated electricity will determine the industry's success, and companies throughout the industry must realize that the goal is to sell or enable the sale of cheap energy, not simply a cheaper module.
When comparing the industry to its closest cousin, it is important to recognize a crucial difference between the semiconductor industry and the solar PV industry. The semiconductor industry competes with itself to constantly reduce cost/bit or cost/transistor, whereas the declining cost profile of the solar PV industry competes with the rising cost of grid supplied electricity. Although the solar PV industry has yet to establish its own Moore's Law, the hurdle rate becomes easier each year. When stand-alone economic viability is the industry's defining issue, who could ask for more