Source: TMS Conference Papers, February 14-18, 2010 – Seattle, Washington.
When considering the wire slicing process for cutting PV silicon wafers, recycling cutting slurry is an extremely important option to consider. There are, however, meaningful variations in recycling choices. By exploring these choices, we can help companies make effective choices. Understanding effective choices of options requires understanding the objectives which drive corporate recycling initiatives. In general the options are: no recycling of the slurry components, use of self-managed on-site systems, or using an outside vendor to manage the reprocessing system on- or off-site. Within this range of alternatives, there is a hierarchy designed to satisfy company objectives. This hierarchy includes satisfying the desire for risk management, cost improvement, stable process control, and direct process management. Customer data and industry experience tells us that these are all important drivers. Further consideration allows these objectives to be ranked, and to be further used in selecting the best recycling method for the company.
This article is about recycling slurry used to cut Silicon for photovoltaic (PV) applications. The slurry is comprised of two components: an abrasive grit (typically SiC or Silicon Carbide) and a carrier fluid (Oil or Polyethylene Glycol — aka PEG). Recycling these materials — grit and carrier fluid — is increasingly essential for reasons both internal and external to the company. The correct recycling decisions are dependent upon the objectives of the company, and the available recycling alternatives. Although it is an important topic, this paper does not deal with the recovery of elemental silicon from slurries.
Furthermore, this paper examines the economic and technical details of slurry recycling, and applies value-based thinking (Keeney 1992) to help structure the choices involved with recycling. This section describes why SiC slurry recycling is a key tactic for PV manufacturers and addresses the reasons that silicon recovery is not currently commercially viable. Section 2 discusses systems engineering methodologies which support the analysis of recycling tactics and objectives. Section 3 provides technical details, and section 4 discusses system alternatives for slurry recycling. Section 5 considers a range of company objectives in pursuing recycling. Section 6 uses a systems analysis technique of means-ends objective networks to evaluate the best recycling alternatives given organizational objectives. The paper concludes with recommendations for PV companies which are considering recycling alternatives.
The photovoltaic industry needs to guarantee itself affordable access to solar-grade feedstock, particularly over the shorter term (Woditsch and Koch, 2002). As a partial and short-term solution to maintaining security of supply, Sarti and Einhaus (2002) discuss recycling wasted silicon. Several firms have investigated the separation of the silicon from the waste by-product of slurry; however, there are significant technical and economic issues that currently limit reclamation efforts.
First and foremost is the price of the silicon itself. The price of silicon has risen and fallen dramatically in the past 3-4 years. In the 2004-2005 timeframe, spot prices for silicon ranged from $40 to $60 per kilogram (kg). From 2005 until 2008 the price rose dramatically to a high water mark of $350-450 per kg. The challenging environment coupled with a change in supply/demand dynamics have led to a collapse in the spot price back to 2004 prices ranging from $40-60 per kg (Silicon Price Index, 2009). In this context, recovery processes which might have been on the cusp of viability a year ago are economically unfeasible now. Add to this the technical difficulty of isolating the silicon from other silicon compounds in the waste stream (silicon carbide or silicon dioxide) and the problem has largely been set aside for wire saw slurries. Diamond wire slicing waste continues to enjoy some work in this area, since the process of silicon extraction is simpler (there is no SiC to worry about), however that too is still in a commercially nascent phase.
Nevertheless, the photovoltaic industry continues to work to reduce its environmental impact, which in part entails reuse of its material components. Previous researchers have noted the importance of recovering the SiC in slurry (Tsuo, 1998) as well as minimizing chronic health risks caused by environmental exposure to oil-based slurry (Ftenakis and Moskowitz, 2000).
Adding to this macroeconomic pressure is the larger global recession, which has radically increased the amount of volatility in the market. As a result, many governments have slashed subsidies for solar industry expansion. The Spanish market, a country which accounted for over 40% of the world’s solar demand in 2008, has collapsed in 2009 (Gonzalez, 2009). Bloomberg reported in June that Germany’s feed-in tariffs, which had built the solar industry up to $8.8 billion in sales, are also being reduced (van Loon, 2009).
As a result the world has seen a significant decrease in PV module demand along with oversupply in the entire PV supply chain. The effect of this oversupply has already been seen in the price of silicon wafers, which has dropped by more than 50% in 2009. Margins are getting tighter industry-wide; furthermore they are not expected to increase dramatically anytime soon. Figure 1, adapted from Hand (2009), shows the decreasing margins in the PV industry. Manufacturers expecting to endure this crisis must make meaningful and significant cost reductions.
Figure 1 – PV Margin Trends
2. SYSTEMS ENGINEERING METHODOLOGIES
The key components of a technical decision involve the system itself, the range of technical alternatives available for acting upon the system, and the objectives of the decision-makers tasked with making technical or design choices. Our purpose in reviewing the systems engineering literature is to select an appropriate technique for structuring the slurry recycling decision. Most particularly, this section introduces and motivates the means-ends network (Keeney 1992), which will be used to explore choices for slurry recycling. The structure of each of the following section entails a brief introduction to key concepts, and then a review of supporting methodologies utilizing the introduced concept.
2.1 Systems Specification and Analysis
The principal objective of systems engineering methodology is to provide structured guidance for the analysis and design of social and technical systems. Buede (2009) provides extensive details of the design of engineering systems, paying particular attention to how engineering alternatives are framed and communicated. Arthur (2009) offers a reflective account of the challenges of designing complex, interlocking technologies. A system is a structure of interacting or interdependent parts. Systems may be further composed of additional nested subsystems. System interactions may be material, but the system definition is broad enough to consider interactions based on the exchange of information as well as material exchanges. Systems may be both human as well as technical in character.
The principal tool for systems engineers when designing a system is the system diagram. This is basically an “input-output system” which considers the system itself, the material and informational inputs, and the outputs to the system. Some formulations also usefully consider the external environment, which represents a series of shocks or external events which can drive the system away from its desired purpose and functionality (Walker, 2000). As noted by Arthur (2009), complex technological systems are often hierarchical in character. The design and modification of these systems often requires an elaborate specification process; the management of such processes is naturally of major concern to system engineering (Sage, 1992; Sage and Rouse, 1999).
2.2 Design Alternatives
Any design is a means towards an end. Alternatives then, are possible solutions in light of specific social or technological problems. Unfortunately, given the complex interdependencies in technological systems, sometimes introducing new solutions creates problems elsewhere in the system (Arthur, 2009). McNerney (et al. 2009) argues that exploration of design alternatives, and the ramifying character of change in technical systems, is the major driver behind industrial learning curves. Moore’s law is one of the most famous of these learning curves. The original law entailed component density, but was rapidly expanded to consider a range of other technical specifications.
One of the most promising techniques for investigating alternatives, and their impact and intersection with technical systems is AISA — or the analysis of interconnected decision areas. This technique represents engineering alternatives as an undirected graph; traversing the graph shows the range of possible alternatives as they span all affected engineering systems and subsystems. Although originally invented in the 1960’s, the technique is seeing somewhat of a renaissance in the engineering design community for its clear and visual representation of technical alternatives (Harary et al. 1965). Axiomatic design is also a noteworthy technique for structuring and rationalizing design choices within a complex system (Suh, 2001).
2.3 Defining and Structuring Objectives
Objectives are therefore the expressions of the preferences of decision-makers; these preferences are variously called goals, objectives, ends, criteria or values. Design for new technologies, and decision-making for new technology, are interrelated problems. Systems analysis builds upon an older tradition of decision theory; the goal of decision theory is to assist designers and decision-makers in the consistent pursuit of their objectives. Decision theory prescribes the appropriate choices given existing approaches. This prescriptive approach, originally used in decision-making, is therefore prominent in the systems engineering literature. Hammond et al. (2002) for instance, describes a process for effective decision-making, based on systems engineering methodologies. Among other steps, the process entails specifying objectives and also — as discussed in previous sections — the clarification of possible alternatives.
The decision-maker may be simultaneously pursuing multiple objectives in any given technical situation. The principle technique for structuring objectives, particularly when there are multiple objectives at stake, is the objectives hierarchy. In such hierarchies values are seen as nested, with the most general and encompassing objectives at the top of the hierarchy, and the more specific and operational objectives at the foot of the hierarchy. Keeney (1992), in his work on value-based thinking, extended the objective hierarchy still further. Ends are the desired outcomes for decision-makers, while means are the specific tactics or alternatives the decision-maker finds available for pursuit of their objectives. The resultant structure is known as a means-ends objective network. Since there may be a many-to-one relationship between means and ends, the resultant structure of values and actions is no longer a “hierarchy” but a network. In the sections to follow, we use the means-ends objective network, to better clarify the key decisions involved in recycling. However, before fully characterizing the recycling decision we first highlight significant features of the slurry reprocessing system.
3. TECHNICAL DETAILS CONCERNING SLURRY
3.1 Slurry Composition
Silicon is formed into large blocks, or ingots, which must be cut into individual slices. Various means for cutting silicon slices exist, however the dominant technology in the marketplace today is the slurry wire saw. In slurry wire saws a thin wire, approximately 120 to 160 microns in diameter is fed multiple times around a series of roller guides. A thin layer of cutting slurry is deposited onto the wire. The wire serves only as the transport mechanism — the slurry does the actual cutting of the silicon ingot into wafers. Figure 2 shows a generic schematic of the slurry wire slicing process.
Figure 2 — Wire Saw Diagram
IThe slurry used to cut silicon is typically a mixture of an abrasive grain suspended in a carrier fluid. Silicon carbide (SiC — density 3.2 kg/liter) is used as the abrasive grain while the fluid can be one of a variety of liquids. Oil based liquids as well as glycol based liquids are used. The densities of the carrier fluid can vary from a low of approximately 0.85 kg/liter for oil-based fluids to over 1.12 kg/liter for polyethylene glycol (PEG). In order to create an effective cutting fluid various ratios of the abrasive powder to the suspending liquid are used. One common mixture is 48 percentage weight (wt%) of silicon carbide combined with 52 wt% polyethylene glycol. The resulting slurry has a combined density of 1.634 kg/liter.
3.2 Slurry Supply Options
When one considers the full range of slurry alternatives, a hierarchy emerges. This alternatives hierarchy is presented in figure 3, and then discussed more fully below. A firm can choose not to recycle which is expensive and generates a large volume of slurry that must be disposed of. Clearly recycling slurry is necessary part of running a silicon wafer manufacturing operation. Given that recycling is considered, firms can choose to the do it yourself model, wherein they purchase and run equipment themselves or they can opt to outsource this non-core activity to a firm that has the expertise to process the slurry for them. If the reprocessing is outsourced another decision to make is whether to perform this activity on-site or off-site. There are pros and cons to both methods. However on-site, outsourced reprocessing of wire saw slurry provides a compelling solution for the slurry cost problem. In general, there are four tactics when it comes to reprocessing slurry: do-it-yourself; off-site reprocessing; on-site reprocessing; or the null options, not to reprocess at all.
In a do-it-yourself scenario, the user purchases the equipment, owns it, and is responsible for operating it without outside assistance. Off-site reprocessing is an outsourcing option where the used slurry is transported to a remote location for rework and recovery of the usable slurry. The recovered material is then shipped back to the customer to be reused in the wire saw. On-site reprocessing is an outsourced model wherein the slurry never leaves the customer’s location, providing the benefits of outsourcing a non-core activity while limiting logistics and regulatory exposures. The final option, do no reprocessing, is difficult in this economic climate to justify from a cost perspective. In this model, the slurry is prepared and used in the wire saw, then sent directly to a waste stream at great cost
Figure 3 — Hierarchy of Slurry Supply Alternatives
4. TACTICS FOR SLURRY REPROCESSING
In this section we provide technical details of each of the principal alternatives used in slurry recycling.
4.1 The Choice to Recycle
Certainly manufacturers that use slurry wire saws can choose not to reprocess their slurry. In this situation, the slurry is used in the saw and then simply discarded. Virgin, or unused, slurry is introduced as the used slurry is exhausted from the saw. Based on a cost of $4.66 per liter — not including the cost to deal with the waste stream — this model is extremely expensive and becomes practically unsustainable from a cost and environmental perspective. For example, a manufacturing operation running 10 wire saws could easily generate a volume of 385 metric tons of used slurry per month. This used slurry would have to be replaced with virgin slurry at high cost and be sent to a waste handler at additional cost. The decision to reject the null option could result in a savings of $10-12 million per year, net of recycling costs, utilities, and infrastructure investment. A fuller discussion of this cost rationale is provided in the following sections.
4.2 Do-It-Yourself vs. Outsourcing
Some companies have elected to pursue the do-it-yourself method of slurry recycling. In this scenario, the firm must cover the capital cost of equipment, as well as the installation, operation, and maintenance of these tools. By taking the process in-house, they assume the responsibility for this all aspects of the reprocessing cycle.
Several vendors exist that can provide centrifuges, carrier recovery systems, and other equipment for reprocessing. The trade-off, of course, is that additional labor will be required to run the systems as well as technical expertise provided to ensure that the slurry is properly formulated and optimized.
Another downside to this method of recycling is that recoveries for do-it-yourself equipment rarely provide the maximum recovery of the usable material from the exhausted slurry stream. This impacts dramatically the overall cost of the system to the customer, rendering it expensive and inefficient. Additionally the firm is responsible to ensure that the equipment is running all the time, which means that extensive maintenance and repair time must be dedicated to a tool or system that is not at the core of their operations. Outsourcing the reprocessing of the slurry eliminates these downsides and allows for higher recovery of the usable material which lowers cost of ownership.
4.3 Off-Site vs. On-Site
If a firm has elected to outsource their slurry reprocessing they have several choices in the marketplace. These choices boil down to two large camps: off-site and on-site reprocessing. In off-site reprocessing, the material is removed from the customer’s location and sent to another site where the slurry is processed to recover usable grit and carrier fluid. In on-site programs, the recovery is done within a close proximity of the manufacturer.
Off-site reprocessing is one method selected by many firms. In this model, the used slurry is stored in totes or tanks. Once a sufficient quantity to merit transport has been collected, the material is moved off-site via truck, rail, or ship to the off-site vendor. Off-site reprocessing firms typically have several locations scattered through the world to minimize shipping costs to the customer. Nevertheless, the material must be picked up and transported to the off-site vendor for reprocessing.
Once at the off-site reprocessing firm’s location, the material is processed and the recovered slurry components are separated from the exhausted or unusable portion of the slurry. These recovered components are either returned via common carrier to the customer or if the customer has no need for the recovered slurry, sold on the open market to other customers. This model allows for demand to be aggregated at off-site locations but it requires significant logistics effort to manage the entire value chain. Lead times from when the slurry leaves the customers site until when it returns to their location can be anywhere from several weeks to several months. This means that firms that elect to use the off-site reprocessing route carry a significantly higher carrying cost of slurry “inventory”, as they must maintain material onsite, in transit, and at the reprocessing company’s location. A last matter is one of regulation: because the material is transported from site to site, most government jurisdictions require additional tracking and permitting in the off-site scenario.
For many manufacturers, on-site reprocessing is a more viable solution. In this model, the reprocessing is outsourced to a provider who designs, builds, and operates the reprocessing facility on the same property as the manufacturer. Because on-site reprocessing takes place within close proximity of the manufacturing, the significant cost of transportation is eliminated. Equally significant is that the amount of slurry that must be on hand is reduced. In a typical on-site scenario, a batch of slurry may need only three to five days to reprocess. This allows manufacturers to operate in a much leaner way and with significantly less day-to-day carrying cost for slurry inventory. Finally the regulatory hurdles required for on-site reprocessing are much less, since material is not being shipped from site to site.
5. SLURRY PROCESSING
In the hierarchy we have discussed the relative merits of choosing to recycle slurry vs. doing nothing, the decision to do-it-yourself vs. outsourcing the reprocessing, and if reprocessing is elected the differences in on vs. off-site reprocessing offerings. Now we will talk further about the costs and benefits of the various models in an attempt to further refine the framework and the issues surrounding each decision point. A preliminary review of the recycling alternatives shows that cost effectiveness, safety, process stability, and process control are all important considerations in choosing a slurry recycling method.
5.1 Cost Considerations for Recycling
The cost of the wafer within the hierarchy of the solar panel assembly cost can be highly significant. Consider the cost of one silicon wafer — one of 75 to 100 such wafers in a solar panel. Most wafers on the market today are made in the following size – 156 mm x 156 mm square. In the market place today, such a wafer is selling for approximately $4.00, down from $10.00 earlier this year. The cost of the silicon in such a wafer is about $1.22 assuming a wafer thickness of 180 microns, kerf loss of 180 microns, and a feed cost of silicon of $60 per kilogram. This represents 30% of the selling price of the wafer. Typically, the silicon is the most expensive component in a wafer, beating all other consumable items and labor cost.
The cost of SiC grit is typically $4.00 per kg and the cost of carrier fluid is typically $2.00 per liter. When mixed at a density of 1.634 kg / liter, slurry of this composition costs $4.66 per liter of slurry. That is $17.64 per gallon of slurry or more than 7 times the cost of a gallon of gasoline. This does not include the cost of disposing of the liquid when it is exhausted from the equipment as waste. On a per wafer basis, this equates to roughly $0.47 per wafer or 13% of the selling price of a $4.00 wafer. A company with 10 wire saws could spend more than $21 million a year for the silicon carbide slurry alone, not including disposal fees or labor. This cost places slurry in the top five operating expenses for the vast majority of wafer manufacturers. Clearly companies must work to control this expensive consumable cost.
5.2 Added Costs of Do-It-Yourself
The do-it-yourself option does have benefit for a customer that has concerns about opening their process to a long term partner — clearly any outsourced vendor charged with handling this waste stream will learn quite a bit about the slurry wire slicing process. For some companies, relinquishing this information may leave them with a sense of unease about the risk of having a competitor learn something about their process that is viewed as a key and crucial leverage point in the wafering value chain.
The downside for do-it-yourself options however tend to be much lower recovery rates for the slurry. Most do-it-yourself systems typically recover a smaller portion of the grit and carrier. In the case of the grit typical recoveries are 60%, leaving 20% or more on the table while carrier recoveries can be as low as 20-30%. This leaves 45-55% on the table in carrier recovery. In addition, a lack of attention to the recovery process in a do-it-yourself scenario may impact quality control and productivity, resulting poorer yield and increased downtime.
To put it in perspective, an optimal reprocessing solution that increased the recovery of both grit and carrier could easily decrease the overall costs for a 10-wire-saw operation by $275,000 to $375,000 a month. These factors, along with the need to fund resources that are not central to the business, reduce the potential savings that are expected by taking the process in-house.
5.3 Opening and Exposing Processes
From a quality control standpoint, on-site reprocessing brings full transparency to the manufacturer who has the advantage of immediate and verifiable slurry. Since the slurry does not leave the facility, it is reprocessed in a closed loop that eliminates the risk of outside contaminants entering the production stream. On-site testing, conducted by those who can easily and quickly adjust levels to achieve consistent and optimized slurry, helps ensure wafer yields are high with minimal waste. These types of attributes get to some of the less tangible, but no less important, benefits of on-site reprocessing. By having highly experienced staff on hand, supply is immediate and quality is at its peak. Better quality and higher yields, combined with best operating practices, translate to significant savings and a more stable market position in the long-term.
Contrast this with the off-site solution. In this case the slurry is reprocessed “out of sight and out of mind.” For customers with concerns about the losing a competitive edge by having an on-site provide always present, the off-site model can prove attractive since the recycling vendor does not have intimate access to the firm’s manufacturing processes. This can be a key issue for companies that fear loss of competitive advantage — firms that prefer to maintain an arm’s length from their suppliers in terms of proximity and intimacy. In this case, the selection of an off-site provider may best satisfy their perceived needs for the benefits of outsourced reprocessing without the perceived risks that transparency might bring.
5.4 Risk Management
Risk management is a large part of the decision factor for any reprocessing option. Firms that go the do-it-yourself route may think they have minimized risk, by maximizing the control over the process. In this situation, the manpower, materials, methods, machinery, and management are under the control of the firm. Since reprocessing is not core to the firm’s business however, it is questionable how much control will really be exercised over this important area of the firm’s supply chain. Perhaps a better method would be to consider the outsourcing of this activity to a firm that specializes in the reprocessing business.
Both off-site and on-site reprocessing solutions offer a different risk profile than doing nothing or doing-it-yourself. In this case, the reprocessing has been outsourced to a qualified expert that can manage the manpower, materials, machinery and methods in the best possible manner. Since this is a core competency of the reprocessing firm, these tools and practices will be well known to them and therefore employed with less cost or headache. There is one large difference between off-site and on-site reprocessing — with on-site the difficulty of managing logistics has been eliminated. On-site reprocessing radically reduces excessive working capital issues that spring from extended lead times to return the reprocessed material to the customer and mitigates the risk of a transportation mishap that might occur as the materials are transported from the customer to the vendor and back again.
Finally, there are the governmental regulations that apply to transporting this kind of material over long distances. These vary from country to country, but oil-based carriers in particular are considered hazardous materials in many countries including the U.S. and most of Western Europe. All materials are considered at some level to be waste product until they have been reprocessed and the usable material has been separated from the true waste stream. Any time a product of this nature is transported; the potential exists for accidents and spills to occur. When this occurs, there are numerous direct and indirect costs associated with an incident that negatively impacts the environment, the firm’s perception in the community, and its ability to maintain good regulatory relations with state and local governing bodies.
5.5 Hierarchy of Objectives
As discussed previously, slurry is at the heart of wafer manufacturing, comprising a large component of the cost of a finished wafer. Given this importance, manufacturers would be well served to consider a hierarchy of needs and objectives when evaluating the various options to manage slurry cost and performance (Figure 4). Firms need to clearly evaluate the options that provide the best cost of ownership for them.
Figure 4 — Hierarchy of Slurry Reprocessing Objectives
A clear primary objective that must be met is managing the slurry cost as a component of the overall cost of the wafer. Additionally, companies should consider how reprocessing impacts the operation of a stable process while maintaining a desirable level of internal control over that process. Without a doubt, process stability should be a desire of any firm in this business. The question is what option provides the perceived level of process stability to satisfy the firm’s needs. Additionally, the level of internal control desired is very important, but tends to be function of the company’s culture. Firms that have a strong desire to control every aspect may shy away from outsourcing the reprocessing of a key part of the manufacturing process. Finally, risk management is another key consideration when deciding which reprocessing option is the best fit. Some companies have a higher tolerance for risk coupled with a desire to minimize on-site infrastructure costs. Other firms may feel the risk to move slurry off-site is unacceptable and tend towards some sort of on-site solution, whether it is a do-it-yourself or an outsourced one. Figure 4 shows how these objectives can be structured in a hierarchical manner.
6. THE MEANS-ENDS OBJECTIVE NETWORK
As mentioned earlier, there are four options when it comes to reprocessing slurry: do-it-yourself; off-site reprocessing; on-site reprocessing; or the null option, no reprocessing at all. Manufacturers have a hierarchy of needs to address when considering these options. They include managing slurry cost, maintaining a stable process, deciding how much control they desire to have over each process step, and finally how much risk they are willing to incur in the value chain. Matching these needs to the various options available provides an interesting intersection of the means vs. the ends that can help firms choose the reprocessing solution that best fits their situation. Figure 5 shows a linking of the previously defined ends as they map to the various means. By assessing this network, firms can navigate the various options and end up with a solution that best fits their needs.
In the hierarchy, there certainly are companies with a high desire to maintain internal control over every aspect of manufacturing. These firms typically shy away from outsourcing any key process activity, preferring as much integration into their existing manufacturing framework as possible. For those firms that absolutely need to maintain internal control, the do-it-yourself option may be the best fit as it allows the firm ultimate control over the entire process, from equipment selection to process optimization, to how, when and where to run the recycling process. Ultimately however, these do-it-yourself tools lack the needed reliability and firms find it difficult to maintain focus on a non-core process in a way that allows for good cost and a stable process performance. When focus is lost on these attributes, the costs to operate the do-it-yourself system can increase dramatically, making it an unattractive proposition. In this situation, the need for stability may ultimately require firms to consider outsourcing.
Companies that really want to manage their risk do it on-site despite exposing their processes. This is because on-site outsourced service providers provide the managed slurry cost optimization needed to stay competitive, while allowing for stable process management by providing dedicated expertise and equipment that optimizes the slurry for the customer’s needs, and minimizing the risks associated with off-site processing.
Figure 5 — Means-Ends Objective Network
7. CONCLUSIONS: BEST STRATEGY FOR RECYCLING
Given the range of slurry reprocessing offerings in the market today, the on-site solution represents the best ends to satisfy demanding needs in a changing marketplace. On-site reprocessing brings full transparency to the manufacturer who has the advantage of immediate and verifiable slurry quality. By having highly experienced staff and equipment on hand, slurry is reprocessed in a closed loop that eliminates the risk of outside contaminants entering the production stream. Also, since the slurry does not leave the facility the logistical risks of transporting slurry to and from an outsourced partner are eliminated. All of this translates to better wafer quality and higher yields, netting significant savings and a more solid long-term future. For many companies, an on-site outsourced reprocessing solution provides optimal performance, and contributes to operating and production efficiencies. In this environment manufacturers can continue to focus on their core business.
- Arthur, W. B. (2009), The Nature of Technology: What it Is and How it Evolves, Free Press.
- Buede, D. M. (2009), The Engineering Design of Systems: Models and Methods, New York: Wiley.
- Elrod, R. H. and C. L. Hubbard (1979), “Applying Means-End Decision Trees,” Business, vol. 29, iss. 1, pp. 17-25.
- Fthenakis V. M. and P. D. Moskowitz (2000), “Photovoltaics: Environmental, health and safety issues and perspectives,” Progress in Photovoltaics, vol. 8, iss. 1, pp. 27-38.
- Gonzalez, A. and K. Johnson, “Spain’s Solar-Power Collapse Dims Subsidy Model”, Wall Street Journal On-line Edition, Business, 9 Sept. 2009, 15 Sept 2009, <https://online.wsj.com/article/SB125193815050081615.html>.
- Hammond, J. S., Keeney, R. L. and H. Raiffa (2002), Smart Choices: A Practical Guide to Making Better Decisions, Broadway.
- Hand, A., “PV Revenues Rebound but Margins Fall”, PVSociety.com, 24 Sept. 2009, 5 October 2009, <https://www.pvsociety.com/article/355189-PV_Revenues_Rebound_but_Margins_Fall.php?rssid=20347&q=pv+revenues>
- Harary, F., Jessop, N., Luckman, J., and J. Stringer (1965), “Analysis of interconnected decision areas: An algorithm for project development,” Nature, vol. 5, p. 118.
- Keeney, R.L. (1992), Value-focused thinking: A path to creative decision-making. Cambridge, MA: Harvard University Press.
- McNerney, J., Redner, S., Farmer, J. D. and J. E. Trancik (2009), “The role of design complexity in technology improvement,” Proceedings of the Atlanta Conference on Science and Technology, IEEE.
- Sage, A. P. (1992), Systems Engineering, New York: Wiley.
- Sage, A. P. and W. B. Rouse (1999), Handbook of Systems Engineering and Management, New York: Wiley.
- Suh, N. P. (2001), Axiomatic Design: Advances and Applications, Oxford: Oxford University Press.
- Sarti, D. and R. Einhaus (2002), “Silicon feedstock for the multi-crystalline photovoltaic industry,” Solar Energy Materials and Solar Cells, vol. 72, iis. 1-4, pp. 27-40.
- “Silicon Price Index: November 2009”, Photon Consulting, November 2009, <https://www.siliconpriceindex.com/>.
- Tsuo S. (1998), “Silicon solar cell and module manufacturing: current and alternative, more environmentally benign processes, Photovoltaics and the Environment, Keystone, Colorado, 23±24 July, BNL-52557, Brookhaven National Laboratory, Upton, New York, 1998.
- Van Loon, J., “Germany Slashes Solar Subsidies, Threatening Industry (Update2)”, Bloomberg.com, 2 June 2009, 15 Sept 2009, <https://www.bloomberg.com/apps/news?pid=20601085&sid=aHTxRgadpTpg>.
- Walker, W. (2000), “Policy analysis: A systematic approach to supporting policymaking in the public sector,” Journal of Multi-Criteria Decision Analysis, vol. 9, iss. 1-3, p. 11-27.
- Woditsch, P. and W. Koch (2002), “Solar grade silicon feedstock supply for PV industry,” Solar Energy Materials and Solar Cells, vol. 72, iss. 1-4, pp. 11-26.