thickness of oxidation layer on fuel elements.

vanesch

I have a question: it is generally stated that 100 micrometers is the acceptable limit on the thickness of the oxidation layer on Zircaloy4 fuel rods. But I'm at loss as what is the reason for this limit ? Is it structural integrity ? Thermal resistance ? Mechanical tolerance ?

Homer Simpson

I've never really heard of a limit on ZrO2 layer, so my input is just for conversation.

A little oxide layer is a good thing: prevents the zirconium steam reaction at elevated sheath temps.

but like you say, if too much Zr is oxididized, it would be brittle with poor heat transfer.

Astronuc

It's mainly about structural integrity, but it's also about thermal resistance as it effects the cladding temperature, which affects structural integrity if rod internal pressure exceeds critical limit (lift-off).

That limit is somewhat arbitrary, and has received a lot of study in the last decade as a result of two concerns - 1) the response of the cladding to RIA transients and 2) response of cladding to LOCA and subsequent quenching when cooler water is introduced to the core.

Way back in the good old days, Westinghouse set limit of 100 microns which was calculated to yield an average value of 600 ppm H in the cladding wall. That value is based upon a hydrogen pickup rate (~0.15 - 0.17) during the course of operation. The actual H-pickup ratio is highly variable and has dependence on water/oxide chemistry. Hydrogen embrittlement is an issue at low temperatures.

In the past 15 years, there has also been increased attention to the spallation of cladding oxide, which can result in local cool areas on the cladding with the result the hydrogen migrates from the surronding area to the cooler area. Significant hydrogen (1000-3000 ppm), and this could present a problem for structural integrity under certain off-normal conditions (e.g. RIA) and potentially at high burnup during a power increase (e.g. so-called 'out-to-in' failures).

Westinghouse has developed ZIRLO (Zr-Nb-Sn-Fe) alloy and AREVA has developed M5 (Zr-Nb-O) alloy in order to reduce both the oxide thickness and hydrogen pickup as a function of burnup (exposure).

vanesch

Ah, so the oxide thickness is taken as a kind of rough measure of hydrogen absorption (which embrittles zirconium, I know). It is not the oxide layer itself which is the main difficulty, but rather the (more difficult to measure) hydrogen intake, but as both are correlated, we put a limit on the easiest observable one, is that it ?

Astronuc

Correct. One can measure the oxide thickness non-destructively, usually with an eddy-current liftoff probe, which is correlated to oxide thickness, as opposed to destructive testing where hydrogen is measured by metallography or hot-extraction (e.g. LECO test).

Eddy-current liftoff measurements must be corrected for the magnetic susceptibility Ipermeability) of the crud on the oxide layer. This has been an issue in plants which use Zn-injection in the primary cooling system. It was first noticed in BWRs, but now that PWRs are using Zn-injection, it is also an issue there.

Helmut Fischer GmbH makes thickness probes -
http://www.helmut-fischer.com/indexC...9&LanguageID=6

There is also the loss of the metal (structure). ZrO₂/Zr has a Pilling-Bedworth ratio of about 1.56, so there is that issue to consider in conjunction with the hydrogen.

In high burnup fuel, one may find hydrogen accumulation in the outer surface of the cladding wall, particularly in the region of the cladding adjacent to the pellet-pellet interfaces. It is the localization of hydrogen that is of concern, since that potentially may be the weakest point.

This might be of use (it's dated but informative) -

Waterside Corrosion of Zirconium Alloys in Nuclear Power Plants
IAEA TECDOC Series No. 996
http://www-pub.iaea.org/MTCD/publica...te_996_web.pdf (22.3 MB - use 'save target as')
1998.

vanesch

Yes, I remember that somewhere: it is due to the slightly lower temperature and a preferred migration of the hydrogen towards the cooler regions, no ?

Wow! The "everything you wanted to know about zirconium but didn't dare to ask" volume

Thanks a lot!

Astronuc

Yes, hydrogen migrates from the hotter regions to the cooler regions where it precipitates as ZrH₂, Zr hydride, which while it acts like a matrix stiffener (like a whisker with a high modulus), it also gives rise to local stresses. Furthermore, the migration of H through the metal leaves a high concentration of dislocations, which IMO can promote crack nucleation under the right conditions.

TECDOC 996 is a great overview of the state of the art in 1998. I know most of the people who co-authored that document and who are cited within.

We actually need an update since there has been a lot of work in the last decade on the subject.

Homer Simpson

as an aside, that sounds a bit like the method thru which 'delayed hydride cracking' can occur in Zr pressure tubes. As the unit comes down below 200 degrees C or so, the hydrogen which has migrated to cool/high stress areas comes out of solution leaving Zr hydride platelets, which are brittle, and will crack from the process pressure in the tube. This crack acts as a stress riser, attracting more hydride on the next cool down cycle.

Astronuc

It's the cycling of temperature, which drives the cycle of solution and precipitation (or resolution/re-precipitation) of the hydrides that drives DHC (delayed hydride cracking). Even though the Zr-Nb (or Zr-2) pressure tubes have low oxide thickness, they have a relatively large surface area, and IIRC it was the cold (inlet) end that had the DHC.

There is a similar mechanism in failed fuel, which is more of a HACP (hydrogen-assisted crack propagation).