Study of the radioactive liquid waste treatment by coprecipitation: from modelling to design of new processes

To decontaminate liquid nuclear wastes, the coprecipitation process is the most commonly used in nuclear field because it can be applied to any type of aqueous effluents whatever their composition may be. This process deals with the in situ precipitation of solid particles to selectively remove one or more radioelements. The aim of this PhD work is to investigate phenomena which take place during the coprecipitation of a trace component. To reach this objective, we have proposed a new modelling of the coprecipitation mechanism. The originality of this new approach lies in the possibility to simulate the phenomenon in non equilibrium conditions and at the reactor scale. This modelling combined with the resolution of the population balance, enable to identify the influence of process parameters (flowrates, stirring speed...) on crystal size and ultimately on decontamination. To test this new modelling, simulations of the coprecipitation of strontium ions with barium sulphate have been performed in continuous and semibatch reactors. Thanks to these simulations, laws of the treatment efficiency variation as a function of several process parameters (mean residence time, stirring speed, BaSO4 concentration) have been determined and experimentally verified. This study leads to the determination of optimal treatment conditions. Three apparatus (recycling apparatus, fluidized bed and reactor/settling tank) providing these optimal conditions have been successfully tested and offered significant outlooks for the reduction of the volume of sludge produced by the process. Two new processes are patent pending. (author)