0.1 1 Source: Breyer and Gerlach, 2010.
Originally (Arrow, 1962), learning-by-doing was attributed to "increased workers' productivity" due to experience, other factors remaining constant. Nemet suggests this was a relatively minor factor underlining the drivers of cost reduction. The success of some new entrants was instead due to their capacity to raise capital and take on the risk of large investments. Ten out of the 16 major advances in module efficiency can be traced back to government and university R&D programmes, while the other six were accomplished in companies manufacturing PV cells. Finally, reductions in the cost of purified silicon were a spill-over benefit from manufacturing improvements in the microprocessor industry.
Increased deployment has thus been, through various channels, the most important driver of cost reductions. Nemet's analysis supports the policy prescription based on deployment-led cost reductions which is followed in IEA ETP projections to 2050. The projections are also based on an analysis of the cost reduction potentials to be mobilised in future learning phases, and milestones towards competitiveness in progressively broader electricity markets (see IEA Technology Roadmaps, in particular IEA, 2010c).
Discussion: locking-in, locking-out
The somewhat provocative expression "green serves the dirtiest" designates policies supporting renewable energy deployment as the only culprit of this paradoxical - but relative - advantage given to the most CO2-emitting fossil fuels. However, the problem, if there is one, arises from the interaction of the two policies. ETS alone would certainly give an advantage to the cleanest fossil fuels. RE policies alone would essentially disadvantage all types of fossil fuel.
The addition of an RE policy to an existing ETS does not lead to additional CO2 emission reductions from the entities covered by the ETS, as the overall cap remains unchanged. This is not a failure of the RE policy. It is a simple and logical consequence of the very design of the ETS, which is a fixed quantity policy. Things would be different if the policy directly addressing CO2 emissions were a price policy or a hybrid policy, or if the emission cap were to be set in conjunction with the expected contribution of RE support policies. In the case of a price policy - say, a carbon tax - CO2 reductions driven by the RE policy could add to the CO2 reductions driven by the carbon tax, depending on the strength of each. In the case of a hybrid policy, such as an ETS with a price floor, a reduction of the carbon price resulting from the RE policy could lead to additional CO2 emission reductions, in the event that the carbon price were to fall below the level of the price floor.
The remaining question is whether the short-term, relative advantage given to more CO2-intensive generation technology could lock-in of such technology, at the expense of efforts to cut emissions.
The feedback process from markets to technical improvements providing increasing returns tends to create "lock-in" and "lock-out" phenomena: it is not (always) because a particular technology is efficient that it is adopted, but (sometimes) because it is adopted that it will become efficient (Arthur, 1989). Technological paths may very much depend on initial conditions. As such, technologies having small short-term advantages may "lock-in" a society into technological choices that may have lesser long-term advantages than technologies that are "locked-out".
How does this apply to the issue of fuel shifting vs. renewables? Fossil-fuel technologies have had a very large global market for more than a century. They can still improve, but marginally, while the introduction and deployment of new renewable energy technologies from a very narrow basis holds the possibility of more considerable progress.
It is therefore unlikely that the rather minor advantage given to more CO2-intensive generation described by Bohringer and Rosendahl (2009) would enhance the lock-in of these fossil fuel technologies. By contrast, RE policy instruments will unlock the potential of renewables.
While this fuel shift in electricity generation does reduce GHG emissions, in climate-friendly scenarios like the 450 Scenario of the WEO 2010 (IEA, 2010a) the global consumption of natural gas decreases after 2025 while renewable energy production continues to expand. In this scenario, the contribution of renewables to electricity production reaches 14 500 TWh by 2035 (from 3 800 TWh in 2008), the contribution of modern renewables to the production of heat increases from 10% in 2008 to 16%, and the production of biofuels for transportation multiplies sevenfold.
A short-term fuel switch towards natural gas in existing capacities with low capacity factors entails immediate emission reductions; however, providing an incentive to build new gas-fired generating capacities - presumably efficient combined-cycle gas turbines - could make achieving deeper emissions reductions post-2020 more difficult and costly. CCGT plants established this decade would face competitive difficulties post-2020 in the face of tighter emissions targets/caps and higher carbon prices which, ultimately, would favour renewables. Although gas plants can be re-sold and moved, their owners would presumably argue for compensation and/or deferral or softening of post-2020 targets. From this viewpoint, less efficient but cheaper open gas turbines could paradoxically be preferred in that transition for their capability to serve later on, with quite low capacity factors, as back-up capacities for variable renewables.
Was this article helpful?