LAUSR

Abstract Magnetite (Fe 3 O 4 ) deposits in the secondary circuit of nuclear pressurized water reactors (PWRs), lead to the fouling of the steam generators (SG) which decreases their thermal performances and increases the risk of corrosion of the SG tubes. As a counteraction, preventive and curative chemical cleanings (CCC) are industrially implemented to remove oxides sludges and deposits in SGs. The use of chelating agents in chemical cleaning processes could affect the passive layer of SG tubes, and modify their surface reactivity. In order to better understand these phenomena, two experimental loops have been designed and operated: the FORTRAND loop to perform and investigate SG tube oxidation, fouling and refouling and the ECCLIPS loop used to investigate the effects of the chemical cleanings. A three steps strategy has been implemented as follows: (i) reproducing magnetite deposits on oxidized SG tubes, (ii) applying three different industrial chemical cleaning procedures (a curative and two preventive ones) and (iii) studying the redeposition (refouling) of magnetite. The fluid physico-chemical conditions upon these steps have been thoroughly followed and controlled. Magnetite deposits formed on the SG tubes upon the first fouling have been characterized by scanning and transmission electron microscopies (SEM and TEM) and X-ray diffraction (XRD). They are shown to be composed of a dense layer of small magnetite crystals. Secondly, three different SG industrial cleaning processes were reproduced. Their timing and thermo-chemical conditions were strictly respected and they were found to dissolve most of the fouling deposit. Disperse magnetite crystallites were present on the tube surface. Moreover, TEM cross-section images showed that no general attack of the tube passive layer occurred. Finally, the cleaned tubes were fouled again in the FORTRAND loop using the same experimental conditions as for the first fouling step. It could be concluded that chemical cleanings have no effect on the fouling kinetics of the SG tubes for a short one month period and that the amount of deposit formed before and after the cleanings was identical. The small crystallite dense layer observed before cleaning was not present on refouled tubes and the size of the crystallites was bigger after the cleanings. For a short time period, this morphology could result in the formation of a fouling deposit with more porosity. As the increase of deposit porosity can impact the thermal transfer at the SG tube surface, morphological changes, hardly predictable, could be important for the SG thermal performance after chemical cleaning. For a longer period, frequent SG cleaning applications should prevent the densification of the deposit and thus delay performance loss over time. To the best of our knowledge, this experimental program is the first study of chemical cleaning impacts on SG tubes reactivity.