compressing CO₂. Limitations of existing boilers and turbines could mean that reductions of emissions to something like those of natural gas power plants without CCS, about half that of a typical coal plant, would be more likely to be implemented than the 90 percent reduction that is technically possible. Achieving more substantial reductions in emissions will require more extensive retrofitting of existing coal plants; their replacement with new coal plants (which have higher greenhouse gas-capture efficiencies) or with some combination of renewable-energy and nuclear-energy sources; or reductions in energy use.

Consequently, achieving substantial reductions in CO₂ emissions from the electricity sector is likely to require a portfolio approach involving the accelerated deployment of multiple technologies: energy efficiency; renewables; coal and natural gas with CCS; and nuclear. However, the following two kinds of demonstrations must be carried out during the next decade if we are to more fully understand the range of available options:

• Assess the viability of CCS for sequestering CO₂ from coal- and natural-gas-fired electricity generation. This will require the construction of a suite (~15–20) of retrofitted and new demonstration plants with CCS, featuring a variety of feedstocks (diverse coal types and natural gas); generation technologies (ultrasupercritical pulverized coal, oxyfuel, integrated gasification combined cycle, natural gas combined cycle); carbon capture strategies (pre- and post-combustion); and geologic storage locations (enhanced oil recovery sites, coal seams, deep saline formations). A few retrofits of existing natural gas plants and new gas plants with CCS should be included among the demonstrations to prepare for the possibility that optimistic forecasts of domestic natural gas availability and price prove correct. The commercial-scale demonstration of CCS would also enable the integration of this technology into plants that produce liquid fuels from coal and biomass.

• Demonstrate the commercial viability of evolutionary nuclear plants in the United States by constructing a suite of about five plants in this country during the next decade. Evolutionary plants are already in operation and are being built in some other countries, so there are no technological impediments to their construction in the United States. However, plant construction requires multi-billion-dollar investments—very large for the size of nuclear plant owner-operators in the United States. The long lead times (6–10 years) required for planning, licens-