Skip Navigation Links.

Search for publications:

Limit search to the fields

ECN publication
Learning from the Sun; Analysis of the use of experience curves for energy policy purposes: The case of photovoltaic power. Final report of the Photex project
Schaeffer, G.J.; Seebregts, A.J.; Beurskens, L.W.M.; Moor, H.H.C. de; Alsema, E.A.; Sark, W.G.J.H.M. van; Durstewitz, M.; Perrin, M.; Boulanger, P.; Laukamp, H.; Zuccaro, C.
Published by: Publication date:
ECN Energy in the Built Environment 1-8-2004
ECN report number: Document type:
ECN-C--04-035 ECN publication
Number of pages: Full text:
99 Download PDF  


Since the 1990s energy policy scientists have started to explore the possibilities of using the experience curve approach for energy policy making. The concept of the experience curve is simple, at least in principle. It states that, for every doubling of cumulative produced capacity of a product or technology, the cost for making it declines with a fixed percentage (learning rate). Historical statistical analysis can be used to define this percentage. Extrapolating the trend thus found, into the future will then give relevant information about future cost developments and will also give information how much ?learning money? will be needed to get to the break-even point. The Photex project used the development of solar PV as a case to further explore this approach, and also to deduce lessons for PV policy. Other aims were to look at learning rates of the different components of PV-systems and to combine the experience curve analyses with bottom-up engineering studies. The main conclusion with regard to the use of experience curves for energy policy making is that this is an interesting approach, but that such an analysis should be done with much care. For the historical analysis the availability of reliable and firm data is essential. As cost data often are not available, price data could be used as a proxy, as long as sufficient long time ranges are used. Also, the analyst should take care he considers the right learning system boundaries. Furthermore the number of years to be included in the statistical analysis should at least be 10 years and this period sample should not over-represent stable price periods or periods of steep price decline. If possible, data uncertainties should be taken into account as well. An interesting finding was that, at least in the case of PV, the learning rate is not a constant, but can vary over time. In the case of PV it improved from 20% to 23%. Extrapolations into the future should take uncertainties into account and always be done scenario-wise. Doing this for grid-connected PV, the analysis shows that self-sustaining markets could be expected in Europe at specific locations before 2010. Reaching break-even for the wholesale electricity market will most probably not happen before 2030. To reach break-even within a reasonable time-frame, the market deployment growth rate should be at least 15%-20% over the next coming decades, assuming that the learning rate remains at the current 20%-25%.

Because the learning rate is not fixed, it might be possible to influence it by a right combination of market deployment policy (learning investment) and R&D-policy supporting the market development (investment in learning). An experience curve analysis can contribute interesting information to policy discussions on the balance between market support and R&D. In the case of PV it seems that a substantially increased support for learning processes, such as R&D, and interactive learning networks within and between users, producers and component suppliers, which supports current and future market deployment instruments, can yield a huge public benefit in the longer term. Furthermore the experience curve approach shows that from a European perspective the cost-effectiveness of PV- policy could be improved. This effectiveness could be reached if a European-wide market-based incentive scheme were implemented. As cost-effectiveness is just one policy goal among many others, such a European market-based deployment instrument should not exclude additional Member State PV deployment policies.

Back to List