The “Learning Curve” of Renewable Technology

by Patrick Quarberg

A 2007 study done by Patrik Soderholm and Thomas Sundqvist attempted to factor in “learning curve” expenses to renewable technology. They describe the “learning curve” expenses to be the increased cost of producing and installing a piece of equipment or technology while it is still a new product. As more of the product is implemented, implementation costs decrease. The study focuses on estimation methods for learning curve costs, and the importance of estimates in deploying technology like wind turbines and solar panels. Specifically, the study investigates issues regarding time as an important variable in learning rates, the interconnectedness of innovation and diffusion, and omission of other important variables in learning rate estimation. The reason for investigating time as an important factor, according to the researchers, is to find out whether costs are decreasing due to actual learning and innovation. The cost decreases should be explained by cumulative capacity—the implementation of additional units—not just time. This is indeed what was found in other studies, and was so included in Soderholm and Sundqvist’s estimation equations.

They then examined the falling prices of wind turbines in four European countries: Denmark, Germany, Spain, and the United Kingdom. The observed trends were as expected. As more wind turbines were built, the individual investment cost decreased, though the actual cost differed from country to country due to several variables depending on the characteristics of the country, as well as the average size of windmill being built. Soderholm and Sundqvist then used these data to further refine their estimation equations. They came up with several variations of their elementary learning curve equation, each taking different variables into account. Interestingly, some models, which accounted for omitted-variable bias, found that the time factor did not affect the model’s accuracy at all, so the learning-by-doing effect was completely isolated. After analyzing how each estimation equation performed, Soderhold and Sundqvist concluded that omitted-variable bias seriously affects the data, and must be accounted for in the equations. Additionally, they suggested that the relationship between innovation and diffusion—which they assumed to be dependent on each other—be tested in a later study, so as to confirm the validity of their estimates. They also noted that no matter what model is used, time will play some time of role in the trend observed, so steps must be taken to separate time and actual learning in order for estimates to be relevant.

Soderholm, P. Sundqvist T. 2007. Empirical Challenges in the Use of Assessing the Economic Prospects of Renewable Energy Technologies. Renewable Energy, Volume 32, pages 2559-2578.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s