Australian research reactor fuel problem

[curie]

Any ideas about the causes of the incident at the new Australian research reactor (OPAL) where some fuel plates in some of the assemblies were found to be elevated? I.e., translated vertically upwards. This has been reported as the plates being "partially" dislodged, but the actual degree of dislodgement is fairly significant if given as % of the total element height.

 

[Astronuc]

Does one have a reference or link to a story that describes the problem.

It could be loss of mechanical integrity combined with hydraulic forces, or perhaps growth of the plates, which could be due to fuel swelling/growth or growth of cladding materials. Mechanical failure could be the result of a manufacturing/design flaw and/or stress/fatigue/corrosion problem.

It would help to have a description of the problem.

Edit (found this): http://www.ansto.gov.au/__data/assets/pdf_file/0003/17481/OPAL_media_release.pdf

http://www.ansto.gov.au/__data/assets/pdf_file/0011/16202/reactor_shutdown_background.pdf

A failure or anomaly very early in the life of the fuel usually indicates a manfuacturing/design problem. That's why one would have a rigorous QC/QA program - well supposedly one does.

 

[curie]

There is an INES report I believe but I couldn't find it online. They vary in detail anyway. In this case the operator & regulator disagreed on the INES rating. It has been given a provisional rating of 2.

I was surprised the plate elevation was not spotted earlier, only during the routine fuel change. And also that this issue was not predicted through modelling & testing. It does not give much confidence in that design process, given the possible consequences if actual uncontrolled fuel ejection occurred. I guess the desgin process is not as rigorous for research reactor fuel as for power reactor, but it should be suitably robust for designs of such reactor without containment.

 

[Astronuc]

There is an INES report I believe but I couldn't find it online. They vary in detail anyway. In this case the operator & regulator disagreed on the INES rating. It has been given a provisional rating of 2.

It may not have been too serious, technically, but the operator (ANSTO) should not be arguing with the regulator. The ANSTO operators screwed up!

I was surprised the plate elevation was not spotted earlier, only during the routine fuel change. And also that this issue was not predicted through modelling & testing. It does not give much confidence in that design process, given the possible consequences if actual uncontrolled fuel ejection occurred. I guess the desgin process is not as rigorous for research reactor fuel as for power reactor, but it should be suitably robust for designs of such reactor without containment.

It may be the first such shutdown. As I understand it, the reactor has only operated one year.

I doubt that the fuel plates would be ejected. OPAL is an open pool (i.e. not pressurized) reactor, so I expect the flow (and thermal duty) to be low. U-Si fuel is quite dense, so I doubt the plates would lift. Why they came out - is the question.

The manufacturer of the plate fuel should be following strict QC/QA procedures consistent with rules and regulations in Argentina and Australia. If not, then the regulators, manufacturer and ANSTO are at fault.

One of my colleagues has audited the fabrication of test/experimental reactor fuel and it was certainly covered by rigorous government requirements.

It could be a design fault and/or manufacturing failure. I also question the oversight by ANSTO and the Australian regulators.

 

[curie]

The fuel in OPAL is shuffled monthly (for optimum burn up profile/powershaping) so presumably this defect was not seen in previous fuel movements. The pool is monitored by camera but from directly overhead so apparently the vertical translation was not noticed. I still would have expected the maintenance schedule to include at least monthly in-tank visual inspections, preferably with an underwater camera (or at least binoculars!) but perhaps this inspection was combined with the fuel movements. The possibility exists that upward fuel movements could have occurred then dropped down again I guess.

By ejection I did not mean to imply forceful ejection, rather a vertical translation of a sufficient height to cause toppling of the element/s onto the fuel assembly.

 

[Morbius]

By ejection I did not mean to imply forceful ejection, rather a vertical translation of a sufficient height to cause toppling of the element/s onto the fuel assembly.

curie,

Again here is a case of being familiar with the terminology used in the field.

One of the things nuclear engineers consider when designing nuclear reactors is called a
rod "ejection". This means the fast, forcible ejection of a control rod from the core. One
has to show that this would not lead to dire consequences.

The pressure to forcibly eject a rod would be something found in a power reactor, and not a
research reactor like OPAL.

However, when you say the "magic word" - "ejection" - you naturally trigger nuclear engineers
into thinking of the rod ejection scenario.

Dr. Gregory Greenman
Physicist

 

[Astronuc]

The fuel in OPAL is shuffled monthly (for optimum burn up profile/powershaping) so presumably this defect was not seen in previous fuel movements. The pool is monitored by camera but from directly overhead so apparently the vertical translation was not noticed. I still would have expected the maintenance schedule to include at least monthly in-tank visual inspections, preferably with an underwater camera (or at least binoculars!) but perhaps this inspection was combined with the fuel movements. The possibility exists that upward fuel movements could have occurred then dropped down again I guess.

I would imagine that the depth of water is about 10 m, and it would be difficult to see a small vertical displacement. And if I remember correctly the heat is removed by natural convection rather than forced, so there is not a strong shear force on the fuel. Possibly there is some mechanical interaction during the fuel shuffling. Again, I'm not familiar with details, so I am only guessing. Nevertheless, any vertical displacement of the fuel is not appropriate.

By ejection I did not mean to imply forceful ejection, rather a vertical translation of a sufficient height to cause toppling of the element/s onto the fuel assembly.

Ejection is a strong word to use as Morbius indicated.

In the case of OPAL fuel, any relocation of fuel is undesirable due to local heating and power distribution effects. If it's not in the procedures or part of normal operation, it is to be avoided/prevented.

 

[Morbius]

And if I remember correctly the heat is removed by natural convection rather than forced, so there is not a strong shear force on the fuel..

Astronuc,

OPAL is a 20 Megawatt (thermal) reactor. I doubt that the heat is removed by natural convection.

The OPAL reactor website states that it is forced flow:

http://ftp.ansto.gov.au/opal/about1.html

" The core is cooled by demineralised water in a forced upwards flow."

This surprises me - due to the claimed direction of the flow.

Usually, for open pool reactors with powers greater than about 1 Megawatt - the cooling flow is
DOWNWARD! That's because if the power is greater than about 1 Megawatt - there would be
too high a dose due to activated oxygen or nitrogen.

An open pool reactor will have a duct beneath the core and water from the pool will be be forced
DOWNWARD through the core and out to where the activation products have some time to die
away.

A core with a forced flow upward through the core large enough to cool a reactor of > 1 Mw(t)
would have a flow that would force activation products to the surface of even a 20 foot deep pool;
where these activation products could disperse into the air of the reactor room.

For example, the forced flow through the now shutdown 2 MW(t) Ford Reactor at the University
of Michigan was DOWNWARD.

Dr. Gregory Greenman
Physicist

 

[Astronuc]

Greg, thanks for the clarification. I was reflecting on my experience with TRIGA reactors, but I should have checked the specs on OPAL.

I think the flow is probably fairly low anyway - certainly not like that of a 3-4 GWt power reactor.

I'd be interested in how they do the forced upwards flow. Presumably the reflector structure serves as a channel (?). Those webpages aren't very technical.

 

[Morbius]

I'd be interested in how they do the forced upwards flow. Presumably the reflector structure serves as a channel (?). Those webpages aren't very technical.

Astronuc,

For example, the Ford Nuclear Reactor at the University of Michigan had ducted fuel elements.

That is each fuel element was a square duct with plates that were cladded fuel "sandwhiches" -
so called "MTR" fuel. Each square duct had a cylindrical nozzle attached to the lower end.

The lower core plate of the core was a plate with a square array of circular holes. The lower
nozzle of a fuel assembly was placed in each of the holes - so the core consisted of a square
array of fuel assemblies. Below the core was a "funnel" that led to the cooling system.

http://www.umich.edu/~mmpp/

Coolant pumps drew water from the funnel. Therefore, the flow was from the pool into the top
of the fuel assembly, down the fuel assembly, which is a duct; through the nozzle and into
the funnel and out of the pool to the heat exchangers of the cooling system.

For low power research reactors the flow can be forced upwards through the core and out into
the pool. At some point there is a loop that cools the pool - a loop analogous to the filter loop
on a backyard in-ground swimming pool.

Dr. Gregory Greenman
Physicist