A Sickputer Review of:
In-Vessel Degradation in LWR [Light Water Reactor]
A State of the Art Report to CSNI
http://www.oecd-nea.org/nsd/docs/1991/csni-r91-12.pdf Comments: 13.2 MB PDF (in picture format). No easy way to make the content text available for copying without printing out this document and scanning wth OCR software).
Page 10: “Degraded core quench is not well understood or modelled. This is important in view of the potential for subsequent hydrogen generation and core heatup during the addition of water to the degraded core.”
Page 11: SP: discussion of the lack of data on late phase degraded core behavior.
“The most important of these are fission product release and transport, molten-fuel coolant interactions (particularly steam explosions) and thermal hydraulics (including natural circulation and debris bed dryout and rewet).”
Page 19: SP: The authors of this report omitted coverage of fission product release and transport and fuel-coolant interactions as these were specialists topics about which there was little progress in research experiments.
Page 20: SP: Speaks of the major accident assessment reports in large nuclear plants (refers to the 1957 WASH-740 report also known as the Brookhaven Report). The assessments went through several revisions culminating in the January 2012
State-of-the-Art Reactor Consequence Analyses (SOARCA) which includes information on Fukushima Daiichi. Available here:
Also on Page 20: The development of BOIL as a a core meltdown model to calculate metal-water reaction netween the fuel cladding and steam. “The heat and hydrogen gas that are produced from this reaction are also calculated.
SP: Interesting paragraph about how much they can tell about the slumped fuel and the steam based in part on the dead weight of the core debris on the lower head and the ablation of the lower head by the hot debris. P. 21
Also on 21…the vast amount of information retrieved from Three Mile Island, Unit 2 (TMI-2 reactor) with regard to in-vessel degradation of a reactor core.
“There is also the possibility of a return to criticality during certain BWR severe accidents.”
Page 23: Blowdown and Boiloff: When there is a large pipe break in the primary system to the reactor vessel, the reactor vessel fluid will be forced out of the vessel in about a minute. For a smaller pipe break the flow rates will be much lower and the blowdown would be followed by water boiloff.
Page 23 cont…When fuel is uncovered the zircaloy cladding oxidizes and generates hydrogen. Clad ballooning may also occur. Significant oxidation makes the cladding brittle, which can result in fuel rod ruptures.
Page 24: The oxidation of the zircaloy cladding gets really going at 1300 K and makes the core heat-up. In a BWR the masses of stainless steel in the core region at 6% of core inventory, and 24% core weight for zircaloy. “Thus oxidation of stainless steel is another potential source of hydrogen generation.”
“The rate of steel oxidation is is small relative to the oxidation of Zircaloy at temperatures below 1400 K. At higher temperatures and near the steel melting point, the rate of steel oxidation exceeds that of Zircaloy.”
SP: Interesting 1991 terminology in describing fissile fuel rods as “poison rods.” Page 24.
Pages 25 and 26: Discussion of boron carbide pellets in BWR control rods and their stainless steel cladding. Detailed account of how zircaloy cladding melts and falls through the control rod tubes in a process called “candling”.
Page 27: Discussion on how after zircaloy and uranium pellets melt and relocate downward there is a significant amount of hydrogen that can be generated after this relocation, but it is difficult to predict accurately.
Page 28: Description of core sample boring of TMI-2. They discovered a molten pool surrounded by a crust in the lower region.
Also on page 28 (Very Interesting!) “Relocation of molten core debris into a water pool can involve a nonenergetic interaction (steam spike) or an energetic interaction (steam explosion). A steam spike can cause significant steam generation which can oxidize metals remaining in the core region and increase hydrogen generation. A steam explosion, however, has the potential to cause significant damage to the reactor vessel. It has also been suggested that steam explosions are capable of generating missiles which could threaten containment integrity.”
(SP: Goddard’s steam theory explosion for Unit 3 may have new life in my opinion)
Page 28 cont.: Boron control holes in bottom of BWR number 55 instrument guide tube penetrations through the lower head .
Page 29: “In a lower plenum with a large number of penetration tubes, the molten fuel may flow preferentially in a film along the the penetration tubes resulting in the failure of these tubes…”
Page 30: Three Mile Island came oh so close to a melt-through or perhaps it actually did have a melt-through:
“Data from the TMI-2 vessel indicate that wall failure occurred in several instrument penetration tubes and many tubes were plugged by debris. In Fact, some of the tubes were plugged in sections well outside the reactor vessel.”
Page 33: Discussion of fission inside the core meltdown. This disturbing statement:
“Fuel liquefaction (i.e., dissolution of fuel pellet with with molten Zircaloy) destroys the crystal structure of the UO2 pellet so that the release of fission products is governed by the migration of atoms and bubbles in a liquid. This is a much faster process than diffusion in the core, most of the fission products in gaseous form would be released within a few minutes.”
Pages 41-49: A number of core melt tests were conducted in several labs (Sandia, Idaho, and Chalk River, Canada), the CORA program in Germany, and severe fuel damage tests in Cadarache, France in the early 1980s. This section details some of the melt results.
Pages 50-70: In several core melt tests the researchers dropped control rods into the molten mass and one woth stainless steel blades to see what moderating effect it might create. A dismal failure as the boron barely dropped 50K off the temperature in one test and in the other a eutectic interaction between the Boron and stainless steel actually caused the stainless steel to liquefy at 200K below the melting point of stainless steel (1700K). This weird eutectic effect was also mentioned in earlier pages with the interaction of zircaloy cladding and stainless steel.
Page 80: The discussion of quenching core melts with water only may promote recriticality. The need for boron in the quench water is emphasized for debris beds.
One of the most important: “Significant hydrogen production continues at temperatures well above Zircaloy melting.”
Page 175: More interesting disussion on Zircaloy cladding and late phase core meltdown on hydrogen production. “…hydrogen generation continues after melting and relocation. There appears to be no ‘cut-off’ temperature for hydrogen generation. Uncertainties in hydrogen generation (which are large) are mainly due to uncertainties in estimating the amount of metallic Zircaloy surface available for oxidation.”
Pages 183-184: Lots of adequate models exists for the early phase of core degradation, but “uncertainties increase for the late phase phenomena. These include: chemical interactions between materials; transitions from highly oxidized or declad fuel pellet stacks to debris geometry; crust formation and integrity; quenching and long-term coolability; fuel-coolant interactions; mode of vessel rupture and debris ejection.”
Pages 189 to 276: Appendices of research findings