Elsevier

Energy Policy

Volume 98, November 2016, Pages 505-512
Energy Policy

A retrospective analysis of compact fluorescent lamp experience curves and their correlations to deployment programs

https://doi.org/10.1016/j.enpol.2016.09.023Get rights and content
Under a Creative Commons license
open access

Highlights

  • We develop a segmented regression technique to estimate historical CFL learning curves.

  • CFL experience curves do not have a constant learning rate.

  • CFLs exhibited a learning rate of approximately 21% from 1990 to 1997.

  • The CFL learning rate significantly increased after 1998.

  • Increased CFL learning rate is correlated to technology deployment programs.

Abstract

Experience curves are useful for understanding technology development and can aid in the design and analysis of market transformation programs. Here, we employ a novel approach to create experience curves, to examine both global and North American compact fluorescent lamp (CFL) data for the years 1990–2007. We move away from the prevailing method of fitting a single, constant, exponential curve to data and instead search for break points where changes in the learning rate may have occurred. Our analysis suggests a learning rate of approximately 21% for the period of 1990–1997, and 51% and 79% in global and North American datasets, respectively, after 1998. We use price data for this analysis; therefore our learning rates encompass developments beyond typical “learning by doing”, including supply chain impacts such as market competition. We examine correlations between North American learning rates and the initiation of new programs, abrupt technological advances, and economic and political events, and find an increased learning rate associated with design advancements and federal standards programs. Our findings support the use of segmented experience curves for retrospective and prospective technology analysis, and may imply that investments in technology programs have contributed to an increase of the CFL learning rate.

Keywords

Experience curve
Learning curve
CFL

Cited by (0)