The photovoltaic industry has been low margin for so many periods in its history that the concept of a margin healthy enough to profitably run an entire company is anathema to it. Focus on the benefits of multi gigawatt economies of scale on margins belies the lengths that PV manufacturers have undergone to salvage margin; lengths that include sourcing lower cost and thus lower quality backsheets, EVA, glass, junction boxes and polysilicon.

This acceptance of poor quality runs from one end of the photovoltaic value chain to the other as all participants from raw material manufacturers through to developers and end users seek to preserve margins by sourcing all components at the lowest available price.

Currently manufacturers from China are relocating manufacturing equipment to Vietnam, Thailand and other countries in South East Asia and are operating on the assumption that pilot scale production can be circumvented and commercial production more rapidly achieved.

Pilot scale production is an essential step in the photovoltaic timeline. This timeline begins with research and development in the laboratory and ends with commercial production. The timeline from R&D through pilot scale production is ~20 years or more. PERC research and development began in the late 1980s. Moving from pilot scale production to commercial manufacturing can take five years. The point of pilot scale production is to arrive at an average efficiency and average durability and reliability through an iterative process. The point of pilot scale production is to repeat the process again and again and again and again until a reliable result is achieved. Even when equipment is relocated the pilot scale function should not be truncated.

A discussion of cost in the photovoltaic industry has to include the effect of incentives (tax and other), subsidies, grants, loans with favorable repayment rates, labor costs including benefits such as health care and unemployment, variable costs such as electricity, working conditions and frankly government goals concerning support. To ignore the rapid growth of manufacturing in low cost countries or to explain China’s lower manufacturing costs simply by virtue of money earned through the U.S. stock market ignores the significant differences country by country. All manufacturers have benefited at one time or another from significant government support for manufacturing. First Solar enjoyed support from the German government when it set up manufacturing in Germany in 2006/07 and no doubt irked more than one government official in that country when it exited in 2013.

Governments support domestic manufacturing for many reasons with the top three being a) tax revenues b) local jobs and c) GDP growth. The U.S. has inarguably the lowest level of support for its domestic manufacturing and China has some of the highest support. These two countries have completely different economic, political, cultural and social ideologies and systems.

It cannot be ignored that in terms of PV manufacturing costs China, with >50 percent of cell manufacturing capability in 2016, has an effect on the global manufacturing cost average. China’s manufacturing costs are extremely difficult to untangle and this untangling cannot be accomplished through public statements.

Components and raw materials used to produce solar cells and modules in China are half the cost and half the price of components sourced from other countries. The difference is a result of a completely different perspective about margins. For example, a low cost backsheet in the U.S. can be produced for $0.22 per square foot, while in China a low cost backsheet is sold for $0.22 per square foot. A higher performing formula such as fluorinated and Tedlar-based is produced in the $0.54 per square foot to $0.60 per square foot range in the U.S. and is sold in China for half of the U.S. production cost.

A normal manufacturing operation for a similar product to a PV module requires a >35 percent to run a healthy, thriving business that includes all the operations required to run a business including R&D. When the margin is truncated quality eventually suffers because all manufacturers will eventually seek methods of controlling costs and retaining a semblance of margin.

Cost behavior is a better measure of industry learning than price behavior. This is because price is most directly a market function and as the photovoltaic industry has little to no control of its markets the price function is measure of market reactions to government subsidies and not a measure of learning.

By any reasonable measure and despite its erratic and low margin pricing behavior, manufacturing cost learning in the photovoltaic industry has been outstanding as the following figures, 4.5 through 4.8, will illustrate.

Figures 4.5 through 4.8 observe costs (2016 constant and current) from 1976 through 2016. As a reminder:

• Current dollars do not adjust for inflation that is, current dollars are in the perspective of the year the goods were purchased.
• Constant dollars adjust for inflation and allows for direct price comparison from one year to another. Constant dollars are in the perspective of the year analyzed and in this case, 2016 dollars are presented. For example, $20.00 in 1976 would be ~85.00 in 2016.