In the article “Geriatric design assessment” (1) the failure of around 200 pressure and boiling water reactor major components incorporating an Inconel 600 alloy steel was examined. The cracking of a tubular control rod access “penetration” or branch in a reactor vessel head due to stress corrosion cracking led to a near disaster at Davis-Besse nuclear power plant in Ohio in 2002. It allowed the moderating boric acid in the cooling water to attack the outer thick ferritic shell of the reactor head, eating away a 200 mm wide hole reaching the inner 9 mm thin stainless steel liner, which although it ballooned out, fortunately held the internal pressure.
The Inconel 600 metal, which has a chromium content of
around 15%, has been largely replaced with Inconel 690, which contains more
chromium, comprising around 30% of it. As tubing it is stress relieved by
thermal treatment, but it is not clear whether this is just the tubing before
the fabrication of the major component, or if the finished object is stress
relieved as a whole. The latter relieves stresses set up by bending of the
tubes, the rolling of the tubes or welding of them into tube plates or other
welding during fabrication. This requires the complete insertion of the
component into a large oven.
See (2) which describes the construction of a modern PWR steam generator.
The steam generators at Sizewell B from the outset employed Inconel 690 tubing, but already some tubes have been “plugged” to prevent leaks of cooling water into the turbine housing, due to the “fretting” failure of some of the tubes. This is caused by vibration fatigue where the tubes are in contact with support plates.
Tubes are also subject to wastage, i.e. thinning, to which Inconel 690 is liable (3), but it is considered to offer better resistance to intergranular attack, pitting and general corrosion than Inconel 600.
These problems have been well known for more than a decade.
(4) and (5). The subject of suitable materials for nuclear reactors is a
continuing topic (6), while the irradiation of all the materials employed leads
to intergranular penetration. The ageing problem with Inconel 600 arose after 15
to 20 years of operation, so a similar time must elapse before Inconel 690 can
be assured as a suitable material. As was referenced in (1), Inconel 690 is not
entirely free of stress corrosion cracking. (7)
The 5,980 tubes employed in an Areva EPR steam generator
have an outer diameter of 19.05 mm and have only 1.09 mm wall thickness, so only
a moderate thinning (or emaciation) can be tolerated.
John Busby 30 July 2009