[en] Rapid atomic power deployment may be possible without using fast breeder reactors or making undue demands on uranium resource. Using by-product D2O and thorium-U233 in CANDU and RBMK piles may circumvent need for either fast breeder reactors or seawater uranium. Atmospheric CO2 is presently increasing 2.25%/year in proportion to 2.25%/year exponential fossil fuel consumption increase. Roughly 1/3 anthropologic CO2 is removed by various CO2 sinks. CO2 removal is modelled as being proportional to 45-year-earlier CO2 amount above 280 ppm-C Water electrolysis produces roughly 0.1 kg-D20/kWe-y. Material balance assumes each electrolysis stage increases D2O bottoms concentration times 3. Except for first two electrolysis stages, all water from hydrogen consumption is returned to electrolysis. The unique characteristic of this process is the ability to economically burn all deuterium-enriched H2 in vehicles. Condensate from vehicles returns to appropriate electrolysis stage. Fuel cell condensate originally from reformed natural gas may augment second-sage feed. Atomic power expansion is 5%/year, giving 55000 GWe by 2100. World primary energy increases 2.25%/y, exceeding 4000 EJ/y by 2100. CO2 maximum is roughly 600 ppm-C around year 2085. CO2 declines back below 300 ppm-C by 2145 if the 45-year-delay seawater sink remains effective