Friday, October 10, 2014

Analyzing Trends in Solar Cell Efficiencies for Fun and Profit

In Future Solar Cost Reductions Hinge on Raising Solar Cell Efficiencies (Solar Industry Magazine, October 2014), Michael Puttre suggests that understanding trends in the efficiency of PV cells is instrumental to understanding the future economics of solar energy.  This is unequivocally true, but it is often accompanied by the implication that one can do little more than track such developments after the fact.   Fortunately, a small cadre of specialists in the fields of strategic technology planning and technology intelligence have invested careers studying the dynamics of technological advance and have honed analytical approaches to characterize the state of the art (SOA) and predict the future direction and rate of advance in technologies of interest to their businesses.  This field of endeavor, often known as technological forecasting, consists of numerous extrapolative and normative techniques for predicting the direction and rate of technological advance in a given field and for assessing emerging technologies.  Importantly, for our purposes, these approaches have been shown to be particularly amenable to analyzing advances in renewable and alternative energy technologies, and the growth in PV cell efficiencies presents us with an ideal application.

Mr. Puttre’s article makes clear the need to understand advances in PV cell efficiencies and their importance to reducing the cost of solar energy.  But, with the plethora of PV technologies available, understanding the landscape can be difficult.  Particular varieties of silicon and thin film dominate the landscape now, but for how long?  The history of technological advance is replete with competing technologies vying for dominance in the marketplace and the current energy arena is no different.  For example, perovskites have recently garnered much attention in the press and appear to be on a steep upward trajectory.  How do we assess their potential in comparison, not to where the incumbents are today, but to where they will be in the future?

Most people have heard of Moore’s Law which describes the exponential growth in computing power over time but few know that Moore’s Law is merely a singular application of a broad class of growth models that apply to technological advance in general.  Fewer still know how to apply such models to the vast quantity of technology trend data that has become available in photovoltaics, wind energy, energy storage, and a host of technologies of interest to energy planners, utilities, developers, manufacturers, researchers, R&D managers, strategic planners, venture capitalists, and investors in emerging technologies.

Is there a Moore’s law for energy in general and PV in particular?  Of course there is.  The problem is there are many, and the first order of business is to adequately define the technology of interest – think silicon vs CdTe vs CIGS vs organic PV vs etc., etc., etc. – and the metrics that will allow you to assess their performance over time.

But, there is more to it than simply turning such technology assessments into a data dredging exercise.  It is common to consult with individual “experts” but unfortunately their biases often limit, albeit unintentionally, their ability to provide an objective picture of technological advance in a given field.  More worrisome, perhaps, would be relying on overpriced reports from the big market research houses whose generally optimistic growth projections are typically unsubstantiated.  Fortunately, more rigorous approaches for aggregating the diverse opinions and experiences of multiple experts are available to help arrive at a more complete picture of future advances in technologies of interest.  And, combining these qualitative approaches with the aforementioned trend analyses can equip technology planners with a fairly robust suite of methods to analyze emerging technologies and plan technology investments.

It is incumbent on those who are making and managing investments in new technology, as well as those who are at the heart of contributing to such innovations, to have at least some understanding of the dynamics of technological advance to help guide their decision making.  The tools to achieve this are available.  And, it is not so much that a given forecast must be proved accurate, but it is the learning gained from engaging in the exercise that enables better decision making.

So where do you start?
  1. Determine the technological performance parameters that can help you describe technological advance (hint: sales growth is NOT a technological performance parameter).  For PV, cell efficiencies are a good place to start but there are others.  For wind energy, rotor diameter, hub height, and turbine nameplate capacities are also a good starting place.
  2. Obtain trend data on the metrics of interest but be careful not to mix different technologies.  For example, if looking at trends in the speed of aircraft, do not mix propeller driven aircraft with jet aircraft as they are fundamentally different technologies and this will only make a mess of your analysis.
  3. Understand where you are on the technology s-curve of the technology you are researching.  There is always some physical limitation to the performance of every technology.  Are you close to it? (See the figure below).  Know how and when to apply the proper growth models (Gompertz, Pearl-Reed, Fisher-Pry, etc.) to complete your analysis.
  4. If looking at the substitution of an emerging technology for an incumbent one, consider that such substitutions often take longer than expected – in spite of what you may read in the press about “game changers” – and it is not uncommon for such an attack to motivate improvements in the incumbent technology.
  5. Last, the use of such analyses in planning your technology strategy is only a starting point.  Work to understand what the data is telling you.  Most importantly, understand that trend is not destiny.

For years, as principal in Technology/Engineering Management Int’l (TEMI) I have taught the principles of technology forecasting and technology intelligence to researchers, planners, analysts, R&D managers and others charged with strategic technology planning, competitive analysis, technology roadmapping, technology scouting, technology assessment, and technology transfer.  On November 13-14, I will be teaching a 2-day workshop in Colorado to introduce other professionals in renewable and alternative energy (and other technologies) to the methods for characterizing and forecasting technological advance.  Course information is available at or by contacting me personally at