Policies for the development of RES in the European Union - The Sustainability of Renewable Energy in Europe - page 64

of learning curves was observed already in the 1930s by Wright ( 1936 ) for air-
plane production. The learning curve began to receive attention during the Second
World War because the governments were trying to predict the cost and time needed
to build ships and aircraft to use during the war (Yelle 1979 ). Some attempts to
use data generated from shipbuilding companies were published by Montgomery
( 1943 ).
Arrow ( 1962 ) was the first economist to officially note the existence of a learn-
ing curve. In “The economic implications of learning by doing” (p. 156), Arrows
wrote: “The role of experience in increasing productivity has not gone unobserved
though the relation has yet to be absorbed into the main corpus of economic theory”.
Then he proposed an endogenous theory of changes in knowledge which leads to
improvement in production functions over time. Subsequently, many authors have
used learning curves to study the diffusion of new technologies and recently a major
focus in this regard has been placed at the RES (Gruber the et al. 1999 ; McDonald
and Schrattenholzer 2001 ).
The learning curve is a tool that can be used both for assessments relating to stra-
tegic production competitiveness and to design or reorganize production systems
taking into account variations that occur over time as a result of the phenomenon of
learning. The basic concept is that the production efficiency of each activity increas-
es continuously to the recurrence of such activities. This basic concept, translated
into mathematical models, allows to predict with a reasonable precision, if wisely
applied, the variation in the time-dependent variables from learning and progress
in which the cost of a product unit, the time required to build it, maintenance hours
required per unit volume of production, etc. (Table 3.1 ).
The learning can be considered as the sum of the factors of discrete and continu-
ous improvements. Discrete factors are triggered by significant events, such as in-
ventions, findings, widespread applications in a very concentrated of fundamentally
new technologies that can cause a change virtually instantaneous and easily notice-
able in the sector observed. Continuous improvements are events not perceivable
individually to a superficial observation and which are attributable to the areas of
design, technological/technical and organizational/management.

Table 3.1 Estimated learning rates. (Source: Azevedo et al. 2013 )
Technology Range of learning rates (%) Time period

Coal 5.6-12 1902-2006

Natural gas 0.65-5.3 1980-1998

Nuclear 0-6 1975-1993

Wind (onshore) − 3 to 32 1980-2010

Solar PV 10-53 1959-2001

Biomass production 12-45 1971-2006

Biopower generation 0-24 1976-2005

Hydropower 0.5-11.4 1980-2001