|
Title:
|
|
Learning curves for hydrogen production technology: An assessment of observed cost reductions
|
|
|
|
Author(s):
|
|
|
|
|
|
Published by:
|
Publication date:
|
|
ECN
Policy Studies
|
5-5-2008
|
|
|
|
ECN report number:
|
Document type:
|
|
ECN-W--08-029
|
Article (scientific)
|
|
|
|
Number of pages:
|
|
|
16
|
|
Published in: International Journal of Hydrogen Energy (Elsevier), , 2008, Vol.33, p.2630-2645.
Abstract:
At present three key energy carriers have the potential to allow a transition towards a sustainable energy system: electricity, biofuels and hydrogen. All three offer great opportunity, but equally true is that each is limited in different ways. In this article we focus on the latter and develop learning curves using cost data observed during the period
1940–2007 for two essential constituents of a possible ‘hydrogen economy’: the construction of hydrogen production facilities and the production process of hydrogen with these facilities. Three hydrogen production methods are examined, in decreasing order of importance with regards to their current market share: steam methane reforming, coal gasification and electrolysis of water. The fact that we have to include data in our analysis that go far back in time, as well as the uncertainties that especially the older data are characterized by, render the development of reliable learning curves challenging. We find only limited learning at best in a couple of cases, and no cost reductions can be detected for
the overall hydrogen production process. Of the six activities investigated, statistically meaningful learning curves can only be determined for the investment costs required for the construction of steam methane reforming facilities, with a learning rate of 11 ± 6%, and
water electrolysis equipment, with a learning rate of 18 ± 13%. For past coal gasification facility construction costs no learning rate can be discerned. The learning rates calculated for steam methane reforming and water electrolysis equipment construction costs have large error margins, but lie well in the range of the learning reported in the literature for
other technologies in the energy sector.
Back to List