We study the adaptation of new technologies by renewable energy-producing firms in a dynamic general equilibrium model where energy is an input in the production of goods. Energy can come from fossil or renewable sources. Both require the use of capital, which is also needed in the production of nal goods. Renewable energy firms can invest in improving the productivity of their capital stock. The actual improvement is random and subject to spillovers. Productivity improvements by renewable firms require "scrapping" some of their existing capital. Together with spill-overs, this leads to under-investment in improving the productivity of renewable energy capital. In the presence of environmental externalities, the optimal allocation can be implemented through a Pigouvian tax on fossil fuel, together with a policy which promotes adaptation of new renewable technologies by taxing firms proportional to their under-scrapping. An implication of our analysis is that it is not optimal to make large investments in new technologies where progress is fast and where current capital becomes obsolete before long. We calibrate the model using world-economy data in order to study the implications of various proposed tax/subsidy scenarios for economic growth.