By Finlay Colvile

The rush to add capacity through the valu chain during 2010 and 2011 has left many observers of the industry somewhat confused about what this all means. And more importantly, what implications does it have moving forward, and particularly in 2012.

The basic problem is one of definition, and the means through which people have historically looked at capacity in the PV industry. Up until the end of 2010, how capacity was defined was not that important. Therefore, the methods being used until then really didn’t matter too much. A limited number of players added capacity to meet growing demand and hopefully gain market share. Thin-film companies added capacity when investments were secured, often in the form of government loans or regional development funds. Simply adding up the announced capacities was largely an academic exercise, with large numbers often used as some kind of ‘health-check’ by industry advocates promoting long-term growth potential.

The expansion glut of 2010 and 2011 put a stop to this. And as a result, assessing capacity levels more systematically (and bottom-up) has now become essential. Historic methodology now looks very much out of place. This is characterized by the various announcements of figures up to─and in excess of─50 GW of capacity being available to satisfy global demand in the low-20-GW range for 2011. Based upon this type of analysis, there is barely an argument for any new capacity to be added for the next 3-4 years. This is not likely to happen.

There are several factors that must be included to move from an arithmetic summation of announcements to a figure that is indicative of commercial reality. This includes a number of steps that should be performed at the company-specific, production-line level:

• Use ramped capacity, not nameplate figures

• Use annualized capacity data, not yearend

values

• Include production line yield levels

• Redefine capacity as effective, not installed

Moving towards a ramped, effective and annualized figure is essential to frame the shipped manufacturing equipment with the demands of the marketplace. Indeed, equipment that does not provide (and simply cannot provide) output that is sellable at any given time represents capacity that should not be factored into any analysis that combines capacity with downstream supply/demand. Finally, it is essential to remove legacy capacity from the system that is typically used more for marketing purposes than for any meaningful production. Other factors have entered the equation also during 2H’11 that were largely absent in the past. Efficiency distribution has become a key metric in determining how much c-Si capacity is commercial. This is evidenced by the acceptance window for c-Si cells being tightened─indicative of the oversupply market that favors the buyers. Also, there is a strong argument for simply removing much of the tier 3 capacity that is currently offline owing to bankability or lack of access to any downstream project pipeline activity. Elimination of competition is likely to retain priority over any consolidation of companies stocked with equipment that offers limited value in acquiring. Therefore, keeping this capacity in the system as being useful in the industry moving forward is simply misleading.

The analysis discussed above is actually required for each of c-Si poly, c-Si ingot/ wafer, c-Si cell/module and thin-film stages of the value chain. Each has its own criteria of what should and should not be factored into the analyses. For now, let’s focus on the cell/module segment (including thin-film modules). Historically, this has been the gating figure used in assessing under/over-capacity.

Figure 1 pulls this all together, and helps to visualize the necessary methodology above. Here, it is clear where the large numbers come from─these are the ones more often cited in the press. As we apply more reality filtering, we move to a metric that is of far more value─the effective tier 1 ramped annualized capacity figure.

Moving into 2012, therefore, there is a nominal threshold of 30 GW of annual downstream demand that should be seen as perhaps the point at which incremental capacity is added by the industry leaders (tier 1 companies). So long as demand falls short of this level, increasing levels of capacity is simply going to be taken offline, retired or taken out the system owing to bankruptcy proceedings.

Finlay Colville is Senior Analyst at Solarbuzz focusing on upstream PV manufacturing, technologies and market supply. Colville previously served as Director of Marketing for Coherent, Inc.’s solar business unit. He holds a BSc in Physics from the University of Glasgow and a Ph.D. from the University of St. Andrews, Scotland.