Lining liners

There are basically only three causes of pool liner leaks: mechanical damage, weld failure or corrosion. Mechanical damage can be divided into two cases: mechanical impact (dropping tools etc) and mechanical stresses (thermal stresses, load changes). Damage due to mechanical impact, such as from dropping tools or similar, are immediately recognized and countermeasures can be initiated at once. Also in such cases the location of the defect area is obvious and intensive search for leak or damage location is not required.

Damage due to mechanical stresses is more difficult to detect and is not revealed immediately. Depending on the location, the applied stress, and the material condition, these failures will cause penetration only over the medium term. Areva NP has experience with cases where unequal heat removal in a reactor cavity lead to mechanical stresses that caused cracks after several years of operation.

After shutdown, if the reactor cavity is filled with cold water, rapid temperature changes cause thermal and mechanical stresses that lead to fatigue cracks. Such failures may also take years until they become apparent. Although the location of such leaks is usually in the vicinity of weld seams, locating the crack is still a challenge in a cavity with thousands of feet of weld.

In the experience of Areva NP, weld failures have not been a significant root cause contributor for failures found after commissioning. After the stainless steel liner has been installed in the pool, it is standard procedure to check all weld seams by non-destructive examination such as a PT test or a leak test. All surface penetrating failures are found and repaired during this phase.

For this type of welding, pores are the most commonly found indications. Such pores do not grow or propagate in the weld seam. Therefore if such failures are found and repaired, no further problems are to be expected. Failures that appear after a certain amount of time are likely to be caused by corrosion, because it does propagate unpredictably, depending on the environment.

However, weld failures can also be indirect causes for leaks in liners. Weld failures may promote the formation of leaks, e.g. by hidden weld pores in the middle of the weld. Such pores are not detectable by usual NDE testing methods, however the material thinning caused by such failures obviously reduces resistance to applied stresses.

Repair of obvious failures may leave undetected discontinuities that will reduce the initial resistance of the material against the applied stresses. Furthermore, removal of superficial failures by grinding without rebuilding will reduce the thickness and thus the resistance of the material against mechanical loads.

Corrosion dominates

The experience gained in the past two decades shows that corrosion as root cause for liner leaks outranges mechanical and weld failures by far. Corrosion failures are dominated by chloride-induced transcrystalline stress corrosion cracking (SCC) as root cause of the failures.

The essential preconditions for transcrystalline SCC are susceptible material, chloride and stresses. Such preconditions are not present on the water side of steel liners in nuclear power plants. Because the water in the reactor cavity is so pure, it does not contain nearly enough chloride to initiate SCC, and furthermore that water is only present during a limited amount of time.

However at the concrete side of the liner the environment is completely different, particularly in the areas where the liner sheet is welded on the embedded framework. There, conditions exist that strongly promote the initiation of SCC. In the area where the liner sheets are welded to the framework, all essential parameters for SCC are present. First, there is a susceptible material. The material commonly used for such liners is austenitic stainless steel TP 304 or TP 321. This material is susceptible to pitting corrosion and chloride induced stress corrosion cracking. The general susceptibility is additionally increased by sensitization due to the heat input during welding.

Second, chloride is present. In the normal construction process, the basic framework is embedded in the concrete and then the liner sheets are welded on this framework. In this process, the contamination of the liner and framework with chloride- or chlorine-containing material is rather likely. However the determining factor is the actual concentration of chlorides. Concentration is obviously a matter of amount of material and volume of the solvent. The construction of the liner creates crevices with obviously very low volume. Accordingly the absolute amount of chlorides required to reach critical concentrations is also extremely low.

Third, there are stresses in the material. As experience shows, stresses induced by heat input during welding are sufficient for the initiation of SCC. Particularly in reactor cavities there are additional stresses when the hot cavity is cooled down with cold water during refuelling. Furthermore the changing stress load with and without water has to be taken into account.

Thus, it can be concluded that at the concrete side of steel liners, and particularly in the area where the liner is welded to the framework, conditions prevail that allow initiation and also promote chloride induced SCC.

Leak Detection

A significant problem of liner corrosion is in detection. Owing to the fact that the corrosion starts from the non-visible side, it cannot be detected until a complete penetration of the steel sheeting has occurred. Therefore leak detection systems or telltales are installed which collect leaking water and indicate the presence of leaks. A weakness of this type of leak detection is that precipitating boric acid may temporarily block cracks and thus prevent a continuous leak rate.

Also, these methods will not detect corrosion that is in progress, but which has not yet penetrated. With these methods, efforts for search and repair work will have to be carried out repeatedly.

The least problem with such leaks is the water loss; lost water can easily be replaced. There are two more worrying possibilities. First, a leak could cause environmental contamination. Second, boric acid along the leak paths could reduce the alkalinity of the concrete, and so reduce the possibility to maintain passivity of carbon steel in this environment. This does not only affect reinforcing material but may also affect unprotected material of the containment vessel.

Areva NP’s solution is to coat the affected or susceptible areas with a material not sensitive to corrosion. The coating covers the susceptible area from the interior of the tank, including over the entire weld seam. This method removes the need to find individual defect areas. This method also allows in cases with many potential leak areas to omit the locating of the individual defect areas. This method avoids additional heat input which may lead to new sensitization or mobilization of detrimental materials. Furthermore this method allows the use of elastic material, which is important for pools where changing loads have to be expected. Either an epoxy or a silicone-based material can be applied dry or under water. The silicone AK system is applied as a sandwich of double-component silicone resin-based matrix with stainless steel sheet as cover. The epoxy system may be applied on its own or in a sandwich. In underwater applications users can apply the coatings with a manipulator.