CLEANING THE PRIMARY CIRCUIT AFTER THE FIRST HIGH TEMPERATURE OXIDATION OF STEAM GENERATOR TUBING: WHY AND HOW TO DO IT

摘要

When they are first exposed to the primary coolant at high temperatures, nickel-based alloys get covered by an oxide layer that strongly slows down their ulterior generalized corrosion. For the Thermal Treated alloy 690 recently built steam generators are made of, this corrosion can become extremely slow, which makes the nickel source term in the primary circuit really small, and by consequence little ~(58)Co is generated by activation of natural nickel. Many studies aimed at determining the best chemistry to get a very protective passive layer and it has long been recognized that an important mass of nickel is oxidized during its formation. Oxidized nickel is not immediately released to water because at high temperatures, nickel oxide, hydroxide and ferrite have very low solubility and precipitate, giving the well known chromium-poor external oxide layer. However, the release of nickel to water is not inhibited, but only delayed. It is therefore desirable to clean nickel out of the primary circuit, but despite a few trials, especially during Sizewell B commissioning, this cleaning has probably never been really efficient. The new idea developed here is that the cleaning of the primary circuit can be achieved by an acid reducing chemistry at a temperature near 170°C, which would enable the dissolution of nickel containing oxides, while preserving the chromium rich passive layer on primary circuit surfaces. Before this cleaning step, a pre-oxidation of the tubes is necessary, and should lead to a stable and as protective as possible passive layer, while minimizing the mass of nickel made available for later release. The main benefit associated with this cleaning would be to avoid elevated ~(58)Co oxygenation peaks following new steam generators operation, and to obtain quickly a plant with both low release rates and low nickel inventory. Other effects will probably be a reduction in dose rates, no more need for prolonged purification steps during shutdown and a generally cleaner reactor. With the current state of knowledge, the best compromise for the pre-oxidation step would be an alkaline reducing chemistry. Several lab tests have shown that a deaerated alkaline conditioning is likely to lead to an increased nickel inventory. However, doing the pre-oxidation step with the primary coolant borated and deaerated, but with no lithium addition could be an interesting option, especially if hydrogen injection is not possible, but more testing is required to assess the consequences of this kind of conditioning. The optimum conditions for dissolving nickel after the pre-oxidation step should focus on the different possible nickel-bearing phases, which are predicted to be slightly different just after commissioning new steam generators to what is usually encountered during PWR shutdowns. Based on the general knowledge on the behavior of corrosion products, a compromise should be made between solubility, which is higher at lower temperatures and dissolution kinetics which show the opposite trend. Acidic pH is favorable for the dissolution of all phases, and hydrogen is required for the dissolution of nickel ferrites.