Fuel rod leak (failure) frequency

[mheslep]

After reviewing some old threads (e.g. here), anecdotally I see that fuel rods do suffer failures on occasion, but it appears not very often, as is intended. My question is, why not more frequently? Given a power reactor neutron flux of 10¹⁴ n/cm2-s, it seems like every atom in the cladding wall must suffer dislocation dozens of times over the couple years of fuel life in core, and the internal pressure from fission product gasses IIRC builds to many tons over the entire rod. Similarly, the structural supports directly in contact with the fuel assembly have no coolant-moderator separation to thermalize fast neutrons, so that those structural supports would also appear to be subject to rapid disintegration.

 

[rpp]

The short answer is that the industry has put a lot of effort into reducing the number of fuel failures.
See for example, the EPRI initiative "zero failures by 2010" http://www.ans.org/pubs/magazines/download/a_730

Some of the practices for reducing fuel failures include:
* monitoring and setting limits for power changes to reduce stress in the cladding (operating guidelines)
* aggressive programs to reduce foreign materials in the coolant (to reduce particles that create debris fretting)
* updated fuel assembly designs to include debris filters at the inlet
* improved spacer designs to reduce grid-to-rod fretting

You are correct that cladding materials suffer a lot of radiation damage. There are specific exposure limits on fuel assemblies
to reduce the amount of exposure that the fuel rods can take before being discharged. One challenge of core design is to
design efficient cores that do not exceed these exposure limits.

 

[Astronuc]

My question is, why not more frequently?

Since the 1960s, the industry has worked to identify and mitigate failures. In the 60s and 70s, corrosion, fuel rod collapse and primary hydriding (hydriding from sources within the fuel rod) were main failure modes. Improvements in alloys and manufacturing methods reduced and eliminated these modes of failure. Debris fretting and occasional grid-to-rod fretting became persistent failure modes in the 1980s and 1990s, respectively.

Some early experience with commercial nuclear fuel:
http://www.iaea.org/inis/collection/NCLCollectionStore/_Public/07/272/7272606.pdf

Debris fretting was resolved by fuel design features such as debris filter bottom nozzles (on PWR fuel) and tie-plates (on BWR fuel), and some PWR fuel designs employed a lower grid to capture the debris and hold it. The lower end plugs were lengthened such that the fuel cladding was above the debris-trap. Utilities developed programs to reduce the generation of debris. The debris is metal wires, turnings or shavings from maintenance operations.

Grid-to-rod fretting (GTRF) was mitigated through improved PWR grid spacer designs. GTRF has not been a problem in BWRs where the flow is about half that of PWR fuel.

There have been a few events of failures due crud intrusions leading to severe corrosion failures. These are rare.

Fuel rod failure rates used to be in the range of 1 E-4, but these days they are down in the 1 E-6 range. Some plants have never had a fuel failure, whereas most plants have had some failures in the past but not now or in a long time, while some plants have had persistent failures or periodic failure events.

it seems like every atom in the cladding wall must suffer dislocation dozens of times over the couple years of fuel life in core, and the internal pressure from fission product gasses IIRC builds to many tons over the entire rod.

The peak regions of cladding may experience about 19-22 displacements per atom (dpa) over 4.5 to 6 years. The atoms are displaced, but most return to proper lattice positions. Irradiation leads to increase strength and reduced ductility, but does not lead to failure. Failure may occur if corrosion reaches certain levels in conjunction with hydriding, which is why the burnup and residence times are limited. Rod internal pressure is also limited based on the fuel design, service limits and core design practices.

Similarly, the structural supports directly in contact with the fuel assembly have no coolant-moderator separation to thermalize fast neutrons, so that those structural supports would also appear to be subject to rapid disintegration.

In PWRs, the guide tubes and in-core spacer grids are made of the same or similar alloy as the cladding tubes. The top most and bottom most grids are often Inconel (usually 718). In BWRs, spacer grids are Inconel (X750 or 718) and Zircaloy-2 with Inconel springs. They experience about the same level of dpa as the cladding, but this does not lead to failure, and as a matter of fact, since the spacer-grids reduce moderation, the local fission rate is suppressed and the neutron flux is slightly less by a few percent as compared to the fuel region between the spacer grids.

It is a bit misleading to characterize atomic displacements as damage, although strictly speaking the displacements do disrupt or damage the order in the lattice. However, the damage does not cause failure or loss of function. Rather, the atomic displacement contributes to irradiation creep, and growth in the case of Zr-alloys, or voids and some swelling in the case of austenitic stainless steels.

 

[mheslep]

Thank you for the excellent response.

In regards to this last note:

It is a bit misleading to characterize atomic displacements as damage, although strictly speaking the displacements do disrupt or damage the order in the lattice. However, the damage does not cause failure or loss of function. Rather, the atomic displacement contributes to irradiation creep, and growth in the case of Zr-alloys, or voids and some swelling in the case of austenitic stainless steels

Is this then to say that there is no such thing as neutron related failure rating of material, regardless of the level neutron energy, density and time of exposure? If so, then it appears nuclear fuel failures are only due to the like of debris-crud-corrosion, and are little different from the problems of, say, heat exchangers in all thermal plants (coal, gas).

Neutron flux has long been a concern for the so called 'first wall problem' of D-T fusion concepts, i.e. the entire first wall of a reactor supposedly would require frequent replacement ( see the infamous essay from MIT's Lidsky ). If the record of fission fuel cladding is any guide, the fusion first wall problem is apparently overblown.

 

[mfb]

Fusion power plants would operate with something like 1000 displacements per atom, nearly a factor 100 higher than fission power plants. In addition, it is problematic if atoms from there get into the plasma, an issue fission power plants don't have.

 

[Astronuc]

Is this then to say that there is no such thing as neutron related failure rating of material, regardless of the level neutron energy, density and time of exposure?

It's more complicated. The fuel in a fission reactor spends limited time in the core. Most fuel would experience 2 or 3 cycles and achieve burnup limits. Some fuel may operate at lower power and therefore achieve additional cycles before reaching the same burnup limit. For now, BWR fuel is limited to 8 years of residence time. So fluence is limited by virtue of the limit on burnup.

As mfb indicated, components in fusion reactors, e.g., first wall and adjacent structures, are expected to experience much greater fluences/dpa, and the response to radiation is more severe. As fluence or dpa accumulates, many metal will experience swelling (I'll elaborate later). In addition, absorption of neutrons leads to activation and transmutation, i.e., nuclides become radionuclides and then decay into a different element. In addition, fast neutrons from fission fusion (14.1 MeV) can knock protons or alpha particles out of nuclei in (n,p) and (n,α) reactions, and the hydrogen and He will accumulate in the alloys causing potential failure. Irradiation creep is also a concern in addition to the swelling. There are ongoing programs to identify materials that will resist swelling at high fluences. Such work has been going on for the last 60 years.

Swelling occurs when vacancies (absence of an atom in the regular lattice) coalesce into voids. Groups of vacancies may also coalesce into dislocation loops, whereas interstitial atoms may form interstitial-type dislocation loops. Swelling is common for FCC austenitic steels and nickel alloys where vacancies tend to form volume clusters.
See - https://www.iaea.org/About/Policy/GC/GC51/GC51InfDocuments/English/gc51inf-3-att7_en.pdf - especially figures 1 and 2.

Radiation-induced defects can be visualized as regions where there is either a deficiency of lattice atoms (voids, vacancies, edge dislocations, vacancy type dislocation loops) or an excess of lattice atoms(self interstitial atoms, interstitial type dislocation loops). These deficiencies or excesses produce geometrical distortions in the lattice structure. Similarly, impurity atoms (interstitial impurity atoms, precipitates of impurity atoms, substitutional impurity atoms) also distort the lattice structure. All these distortions lead to changes in the mechanical and geometrical properties of the material. For example, vacancies (voids) lead to macroscopic swelling and distortion of the lattice structure, which alters the material’s strength and ductility.

Alloys often contain low levels of impurities, and radiation influences the migration of impurities. We sometimes refer to radiation-induced segregation in which elements preferentially migrate to grain boundaries or accumulate as different phases, which may present vulnerabilities in the material.

https://ocw.mit.edu/courses/nuclear...g-2015/lecture-notes/MIT22_14S15_Lecture6.pdf

Dimensional stability and mechanical integrity/capability are two critical areas for designers and there is much research.

it appears nuclear fuel failures are only due to the like of debris-crud-corrosion, and are little different from the problems of, say, heat exchangers in all thermal plants (coal, gas).

More or less that is true. As I mentioned above, the LWR fuel does not normally experience fluences where growth or creep would be especially limiting under normal operation. In addition to normal operation, we must also be concerned with anticipated and unexpected operational anomalies and severe accidents, where the fuel may be subject to thermal or mechanical transients. Some of the limits we impose on the fuel address the expected response of the fuel under such transients, so we have to design for 'just-in-case'.

 

[mheslep]

addition, fast neutrons from fission (14.1 MeV) can knock protons or alpha particles

You meant from *fusion* I assume at 14 MeV, though fission also produces 2 MeV fast neutrons which are still fast, I assume, on contact with the fuel rod walls, causing the same hydrogen or He accumulation problems. It is the rod walls that I seek to compare to would be fusion reactor walls.

 

[mheslep]

From Busby et al, "Structural materials for fission & fusion energy", 2009

http://www.sciencedirect.com/science/article/pii/S1369702109702949

 

[Astronuc]

You meant from *fusion* I assume at 14 MeV, though fission also produces 2 MeV fast neutrons which are still fast, I assume, on contact with the fuel rod walls, causing the same hydrogen or He accumulation problems. It is the rod walls that I seek to compare to would be fusion reactor walls.

Yes, I meant fusion. The neutron from dt-fusion is about 14.1 MeV. Fission also produces fast neutron, and actually the energies can be up to about 10 MeV, but the prompt neutron energy spectrum falls off by about 3 to 4 orders of magnitude over the range from 1 MeV to 10 MeV.

The hydrogen and helium accumulation is related to fluence. Most incore materials would be removed from service before the helium accumulation leads to failure. Interestingly, zirconium does not experience the (n,p) and (n,α) reactions, which is primarily a problem for Ni and other transition metals. The cross sections for these reactions have an energy threshold and are nuclide-dependent.

Transmutation is another issue, e.g., designers of Mn steels must be concerned about the transmutation of Mn to Fe. LWR fuel does not stay in core long enough for transmutation or (n,p) and (n,α) reactions to be a problem. The fuel on the core periphery operates at power much lower than core average power, and in some cases, special shield assemblies may be employed to reduce neutron fluence on the core barrel and other structures.

 

[mheslep]

Interestingly, zirconium does not experience the (n,p) and (n,α) reactions, which is primarily a problem for Ni and other transition metals

Very interesting. For this same reason, Zirc appears to be a good candidate for a fusion reactor first wall.

 

[Astronuc]

Very interesting. For this same reason, Zirc appears to be a good candidate for a fusion reactor first wall.

Zirconium undergoes a phase transition between 830 and 1000 C. The strength of Zr-alloys drops dramatically beyond 600 C, and irradiation creep is much greater than steels. Another potential disadvantage is the higher Z = 40 as compared to lower Z elements like V, Cr, Fe, etc.

Complex alloys often have phases, some of which embrittle the alloy, or may have lower melting temperature, reduced strength, or allow for localized corrosion, any of which can lead to failure.

 

[mheslep]

...

Complex alloys often have phases, some of which embrittle the alloy, or may have lower melting temperature, reduced strength, or allow for localized corrosion, any of which can lead to failure.

And yet Zirc fission fuel rods withstand high fission gas pressures.

 

[Astronuc]

And yet Zirc fission fuel rods withstand high fission gas pressures.

At normal operating temperatures (~300 - 350°C). Under LOCA or RIA conditions, or where temperatures are >> 600°C, especially > 800°C, Zr-alloys don't hold up so well. But then, failure is expected and the requirements are that the degree of failure and consequences not be underestimated, and the hydrogen generation from oxidation must be limited.

LOCA and RIA are hypothetical accidents, except in the case of Fukushima, it wasn't hypothetical, but rather 'Beyond Design Basis'.

Edit/update: Modern fuel designs are licensed and approved to operate above system pressure, some by up to 800 psid (116 MPa).

 

[Pengtaofu]

Thank you for the excellent response.

In regards to this last note:

Is this then to say that there is no such thing as neutron related failure rating of material, regardless of the level neutron energy, density and time of exposure? If so, then it appears nuclear fuel failures are only due to the like of debris-crud-corrosion, and are little different from the problems of, say, heat exchangers in all thermal plants (coal, gas).

Neutron flux has long been a concern for the so called 'first wall problem' of D-T fusion concepts, i.e. the entire first wall of a reactor supposedly would require frequent replacement ( see the infamous essay from MIT's Lidsky ). If the record of fission fuel cladding is any guide, the fusion first wall problem is apparently overblown.

I didn't log in the web and have researched some experience with fuel rod failure in PWR. The fuel rod failure rate has decreased significantly, just as Astronuc said. And IAEA publications (http://www-pub.iaea.org/MTCD/Publications/PDF/te_1345_web/t1345_part1.pdf) and (http://www-pub.iaea.org/books/IAEABooks/8259/Review-of-Fuel-Failures-in-Water-Cooled-Reactors) provide operation experience in detail. In recent year, only one or two fuel rod often defect and it usually happen in the first cycle and second cycle owing to debris which is difficult to avoid. On the other hand, there is no mechanism with fuel failure due to direct neutron radiation in PWR and BWR.

 

[anorlunda]

Wow, thanks all for the flash education on nuclear metallurgy. I had no idea that so much is known.

 

[Astronuc]

Wow, thanks all for the flash education on nuclear metallurgy. I had no idea that so much is known.

https://www.researchgate.net/profile/Frank_Garner/publication/255101170_Neutron-induced_swelling_in_cold-worked_titanium-modified_AISI_316_alloys/links/5685d77c08aebccc4e118b14/Neutron-induced-swelling-in-cold-worked-titanium-modified-AISI-316-alloys.pdfhttp://www-pub.iaea.org/MTCD/Publications/PDF/te_709_prn.pdf
http://www-pub.iaea.org/MTCD/Publications/PDF/te_1345_web/t1345_part2.pdf
http://www-pub.iaea.org/MTCD/Publications/PDF/Pub1445_web.pdfhttp://www-pub.iaea.org/MTCD/Publications/PDF/TE_1692_CD/PDF/IAEA-TECDOC-1692.pdf

Historically, there have been numerous conferences and topical meetings devoted to radiation effects on materials and nuclear fuel performance.

ASTM sponsors an International Symposium on Zirconium in the Nuclear Industry, which covers the technology of zirconium alloy development, manufacturing, testing and performance in nuclear fuel.

ASTM has another symposium on Effects of Radiation on Materials.
https://www.astm.org/BOOKSTORE/STP_SERIES/RadEffects.htm

ANS and counterparts in Europe and Asia sponsor a nuclear fuel performance meeting every year. It rotates among North America, Europe and Asia. It goes by various names such as TopFuel and Water Reactor Fuel Performance Meeting.

Finally, there is a more general meeting sponsored jointly by TMS/ANS/NACE Environmental Degradation of Materials in Nuclear Power Systems, which covers materials issues in the entire plant including the core.