A recent shortfall in polysilicon stock has compounded the cost and efficiency challenges of manufacturing silicon solar cells. Thin-film solar technology could be the way forward.
Dr. John R. Tuttle, Chairman and CEO, DayStar Technologies.
The worldwide demand for photovoltaic electrical generating systems is increasing at a staggering rate. A dramatic increase in manufacturing capacity is required to satisfy the increasing global need for this renewable energy technology.
Many factors suggest that energy demand in general, and electricity demand in particular, will grow substantially in the coming decades. The near- to mid-term availability of hydrocarbon resources to fuel this demand remains debatable, as does the ability of the earth’s atmosphere to maintain its present state of quasi-equilibrium in the face of the resulting CO2 emissions.
Achieving a one per cent – or better, ten per cent – share of the world electricity supply in the next 20 years (based upon two different demand growth scenarios), would enable the Compound Annual Growth Rate (CAGR) for the photovoltaic industry to be a staggering 31-48 per cent.
Expanding photovoltaic market penetration will significantly depend on these factors:
- Cost reductions at the system level would need to be reduced from the present benchmark of $6-8/Peak Watt (WP) to approximately $1/WP (Keshner & Arya).
- CO2 emissions would need to be reduced and then remain at below 1990 levels, requiring a substantial carbon-free power generation infrastructure of up to 10 000 GW by 2050 (Hoffert).
A growing industry
With the projected demand for renewable energy approaching terawatt levels, today’s photovoltaic industry is presently able to manufacture only about 1.5 GW of generating capacity per year – equivalent to one nuclear power plant. Crystalline silicon solar cells comprise 95 per cent of annual photovoltaic output; the balance is in thin-film solar cells and more exotic products used in specialized applications. The photovoltaic industry has reached current levels of production primarily through incremental scaling of manufacturing equipment derived from the silicon semiconductor industry. This approach has been moderately effective at growing the industry, but the inherent complex and expensive nature of manufacturing crystalline solar cells also puts into question whether they can be produced at the quantities and $1/W costs required to meet terawatt-scale demand.
A recent shortfall in polysilicon feedstock, the base material used for silicon crystal ingots, has compounded the inherent cost and efficiency challenges in manufacturing silicon solar cells. Delivery dates for new polysilicon now extend three to five years, and contract spot market prices are rising. The supply shortfall is likely to continue to hamper the ability of silicon solar cell manufacturers to increase their production volumes, and will cause increases in final module costs.
Figure 1. 2002 world electricity production
Thin-film cell technologies have long been considered a promising technology to break through the cost, scaling and feedstock barriers that plague the crystalline silicon solar cell industry. These cell technologies have such unique advantages as lower manufacturing costs, greater production scalability, higher material yields and lower requirements in manufacturing energy. The three leading commercially available thin-film solar cell materials are: amorphous silicon, cadmium-telluride and copper-indium-gallium-selenide (CIGS). In contrast to silicon solar cells, which are 100-300 microns thick, typical thin-film PV technologies use only a few microns of their semiconductor materials. These solar cell materials are so thin that they require a structural component known as a substrate for ease of handling. Typically made of low-cost glass, metal or plastic, these substrates can be manufactured much less expensively than the semiconductor materials themselves, thus enabling a lower cost final solar cell product.
Of the three mature commercially available thin-film technologies, CIGS-based solar electric products and modules have been successfully demonstrated in the field for more than 15 years and offer the highest performance and best reliability. As a benchmark for performance, the best laboratory-scale solar cells for the CIGS, CdTe and amorphous silicon technologies are 19.3 per cent, 16.7 per cent and 13.1 per cent, respectively.
Despite its superior performance, CIGS technology has not yet attracted sufficient financial investment for substantial market penetration. Nonetheless, CIGS solar cells fabricated on flexible metal substrates by scalable, high volume manufacturing methodologies appear to be the most promising implementation model for industry-leading, multi-gigawatt scale manufacturing. DayStar Technologies, Inc., has chosen the CIGS cell technology for its commercialization efforts, offering its Photovoltaic Foils product as discrete solar cells on multiple substrate materials, depending on the intended market application.
Compared to monolithic designs, manufacturing discrete thin-film solar cells offers several advantages that allow production up-scaling to achieve the volumes necessary to overcome the thin-film deposition challenges. This technique also allows for wide web widths, excellent economies of scale and volume, as well as high yields. It also utilizes ‘off the shelf’ equipment already designed for the magnetic hard drive industry. Once the semiconductor layers are applied, the cells are cut into either traditional square formats used to make today’s solar modules (100 mm, 125 mm, or 150 mm) or into innovative shapes and sizes depending upon the customer’s requirements for the production of new and unique products. Some products leverage the physical flexibility of the cells to achieve a wider range of applications.
In general, unlike batch-processed crystalline silicon cell wafers, thin-film solar foils lend themselves to high volume production techniques well proven in other industries. Thin-film foils can be manufactured in a continuous process rather than in batches, with an ability to scale the process in both width and processing speed (length). As the film deposition quality is perfected based on much larger manufacturing volumes (as with the flat panel display industry), monolithic integration techniques can be revisited to reduce final product costs further. However, until that time, and likely thereafter, discrete cells will enable much faster market acceptance and further accelerate market penetration.
The incremental scale-up of CIGS solar cell and module manufacturing to gigawatt-scale involves three primary components:
- Expansion of primary semiconductor operations
- Vertical integration of secondary product material production operations to achieve cost minimization
- Replication of optimized production line sizes to achieve the targeted factory gigawatt capacity.
In the thin-film sector of the photovoltaic industry, the first component above has traditionally been the major stumbling block to achieving cost-competitiveness and profitability. This is a consequence of the relatively small capital investments made to date in manufacturing development and in the choices made regarding product design. Seeking a monolithic solution has at times obscured the attraction of the incremental near-term introduction of discrete commodity solar cells sizes.
By focusing on producing discrete cells, manufacturing scale-up can be rapidly achieved by employing continuous processing methodologies and by trading off economies of volume with economies of scale. Based on our research at DayStar, we believe that a 100 MW/yr continuous line will offer the best combination of economy of scale (relative to a 20 MW line) and economy of volume (10 lines required to achieve 1 GW/yr capacity).
Figure 2. Thin-film module manufacturing processes
Achieving the lowest manufacturing costs – not just for the solar cell, but also for the complete solar panel – requires evaluation of vertical integration within each of the solar cell material supply chains. Figure 2 illustrates vertical integration within a CIGS solar cell factory. Vertical integration along each supply chain is most recommended where it is economically advantageous. Previous product development has determined that the manufacture of glass and aluminum for the solar panel packaging should be vertically integrated [Keshner & Arya]. Adding the refinement of indium and/or gallium, as well as the synthesis and formation of CIGS and other source materials to the production chain, shows similar promise. Additionally, when manufacturing CIGS solar cells on a stainless steel substrate, the steel supply chain should also be vertically integrated.
While thin-film solar cells hold great promise to reduce cell costs and scale to increasingly larger production volumes, feedstock supplies must be secured, product designs must be optimized and production equipment must be scaled to meet the required product demand.
The availability of such rare minerals as indium, gallium, tellurium and selenium constitute a perceived impediment to large scale manufacturing of thin-film solar cells. Gigawatt-scale production of thin-film solar cells may require a significant percentage of the current extraction and refinement capacity of these resources. For example, the use of indium in gigawatt-scale production of CIGS solar cells could require as much as ten per cent of today’s indium production, assuming current cell thickness and material usage efficiencies of 40 per cent. Expected improvements in cell efficiency, reductions in cell thickness and material yield improvements can lower the requirement to the one- to two-per cent level. This mineral is already in short supply due to demand for clear, conductive oxides in the flat panel display industry. To achieve large scale thin-film solar cell production, rare minerals must be used as efficiently as possible, requiring exploration of methods for increased material extraction and the substitution of viable substitutes for these materials in other applications.
Naturally, significant financial commitments must be made into scaling the thin-film photovoltaic industry as the likely best solar cell technology to fulfill future photovoltaic demand. Once established, a standard, optimal thin-film solar cell fabrication platform need only be replicated to achieve gigawatt-scale production facilities. Development and construction of a fully integrated plant will require commitments on the order of $0.5 billion in the near-term. Interest is growing in the international investment community, along with awareness of these possibilities.
Simply put, we as a society must commit to the investment required to meet the ‘terawatt challenge’ over the next half-century. The incremental build out of CIGS gigawatt factories is the best technology platform to meet that challenge.