Carbon capture and storage (CCS) is one of the most promising abatement options to curb CO2 emissions of the energy sector. Usually, in models where the atmospheric carbon stock is constrained to not exceed a given ceiling and under constant average costs, it is never optimal to deploy CCS before the time at which this ceiling is reached. In this paper, we show that, when the CCS technology is submitted to decreasing returns to scale, abatement activities must begin earlier, i.e. before the climate constraint binds. It must also cease strictly before the end of the ceiling period.