Thank you, Mr. Chairman, ladies and gentlemen of the committee.
I provided a brief biography to the committee. I would like to expand on that just a little, and I'd like to touch on two points: aspects of the production of moly-99 and the positive power coefficient of reactivity.
The positive power coefficient of reactivity is not a mystery. It is not an unsolvable engineering problem. It is a small thermal mechanical effect in a prototype design that requires a simple engineering fix. The power coefficient can be restored to a value of close to zero, and a safety case can be made for these conditions.
I'll come back to my biography. I started working at AECL in 1975. In 1981 I became section head of physics at the Whiteshell nuclear research establishment. In 1982, at the request of my branch head, I started to design the core for a new research reactor concept for the purpose of developing a new product. At that time, it had no name. The design concept was that it would be a multi-purpose research reactor, be based on low-enriched uranium, have competitively high thermal neutron flux levels, and not require development of any new technologies.
“Multi-purpose” meant at that time that it should be able to provide neutrons for fundamental nuclear materials research as well as produce a wide range of both medical and industrial isotopes.
During the next six years I led the development of the reactor concept, assembled and developed the computer codes, developed the various initial models required to simulate the reactor, and built an analysis group to support the effort.
By 1984 we felt that we had the basis for this new product. The group was asked to come up with a name and to make it Canadian in some way. In the next week or two, I coined the acronym “MAPLE” reactor. It stands for “Multipurpose Applied Physics Lattice Experiment” reactor.
In 1985 we hosted a team of ten Korean physicists and thermal hydraulics specialists to start the design of a MAPLE for Korea, which became known as the HANARO reactor. The HANARO reactor went critical in the early 1990s and functioned successfully at up to 30 megawatts, carrying out all of the described functions. It is mainly devoted to research, and production of moly-99 is not a priority. You cannot just throw out research programs that take years of implementation.
I left AECL at the end of 1988 to advance my career and experience by working for the CEA at Cadarache in France. I returned two years later to work in the nuclear safety group of Ontario Hydro. Three or four years later, I was offered the opportunity to work in Moscow on contract to AECL, consulting as an expert on western safety analysis methodology.
In 1997 I returned from Moscow to work as the head of the physics group on the newly revived MAPLE project. I led the physics effort in the preliminary and final safety reports, became a commissioning supervisor for MAPLE, and then became a nuclear commissioning manager.
I and my team took both MAPLE 1 and MAPLE 2 to criticality. We measured the positive PCR and we participated in the subsequent efforts on the positive power coefficient of reactivity.
I understand from the newspapers that there has been a team of experts who claim that MAPLE would never be functional. I now ask the rhetorical question: who are these people? If anybody qualifies as an expert on MAPLE, I think I'm it. Nobody has asked me or anybody else involved in the project what we think.
Let me talk briefly about the production of moly-99.
From the project's inception, we had focused attention on how to make sufficient quantities of fission product moly-99. If it were easy to do, everybody would be doing it. Working from the known demand at the time in the mid-1980s and using the estimated demand growth for a 30-year reactor lifetime, we built in the capacity to deliver double the world requirement at any time. This is achieved by high thermal flux levels, flexible target removal schedules, and the capability of the reactor to be shut down and started up every 24 hours. This is not an easy task.
I emphasize the word “deliver” since the reactor must produce at least twice the amount that has to be delivered, because you're going to essentially lose half of what you've produced by the time you've extracted it, purified it, and delivered it to where it's going. As was pointed out by Dr. Koclas, the half-life of moly is only 2.7 days. You have to work very quickly. Processing is almost a military-style operation. That is why you cannot store it.
But to make sufficient quantities to meet these demands, the reactor needs to have high flux levels. Without delivering a nuclear theory course, please accept that it is the nature of the production and decay processes.
You cannot make more in a low power reactor by operating for longer periods, because what you have made will be destroyed by neutron absorption and decay. You cannot arbitrarily raise the maximum power level of an existing reactor to increase flux levels to produce more moly-99. You could produce more in this manner, but reactors are designed for a certain maximum power level. Raising that maximum value involves redesign to provide additional cooling and compensate for whatever safety margins have been eroded, and possibly fuel redesign.
These changes would be increments of 5%, 10%, and 15% on their current capability, as you've already heard in the press. When people talk about how they're going to work on their reactors, they're talking about 5%, 10%, and 15%.
It's the compactness of the MAPLE core that permits the required flux levels at a relatively low power of 10 megawatts. Having said that, I note that MAPLE 1 operated at 80% full power and is capable of making the world requirement for moly-99 at that power level. MAPLE 1 was producing moly-99. We did not extract it because we're still commissioning the reactor and we did not want to destruct the continuity of the process.
Let me just speak briefly, then, about the power coefficient of reactivity. When the positive PCR was measured during commissioning, further tests were put on hold until it could be investigated. We reviewed our calculations. We've re-performed the calculations using the latest tools and data libraries. The design tools that we had used were in the original tool kit developed in the eighties. We contracted external expert groups to review our test analyses and calculations, and another group, as was pointed out, to recalculate the PCR.
Nobody came up with a result that was significantly different from the original results. From this, we concluded that there must be an unmodelled effect taking place. The regulator required that we understand it.
We executed a PIRT study for a phenomenon, importance, and ranking table, in which every component of the reactor is examined by a group of specialists on each system against a list of physical phenomena to decide if a particular phenomenon can contribute to the observed effect. This systematic approach led to the identification of 20 possible candidates, but only three stood out: bowing at the targets, bowing of the fuel elements, and possible heat-up of water between the reflector wall and the flow tubes, because there was some recirculation.
A test program was planned for the execution of these tests that focused on each candidate to the maximum extent possible. However, in each case it was not possible to completely isolate the factors in each test; i.e., there was some interdependence of each of the three different effects. So the end result needed the answers from all three tests to determine each individual contribution uniquely.
By taking out the targets and replacing them with fuel bundles, test one showed a reduction of a positive PCR by about one-third. Test two straightened out the recirculating water, but did not change the value of the PCR. That's been called failure. That was a measurement of what we intended to measure.
That left the third test, on the fuel bundles, to be executed, which was to use the simple engineering fix, which is to restrain the bowing of the elements--and it's a very tiny amount--when the project was suddenly terminated. The contribution from the targets and the contribution from the fuel both depend upon the same physical effect.
This same effect will happen to any material that expands when heated. If one side is hot and the other side is cool, there's a temperature asymmetry from side to side.
The fuel assemblies in MAPLE and HANARO reactors are very similar. HANARO is a larger MAPLE. I personally worked on the transfer of the technology. The fuel assemblies are made by the same people. HANARO has a negative PCR of a value we calculate for MAPLE. It is mostly a property of the fuel constituents.
The temperature asymmetry results from a high-flux gradient across the outer elements of a fuel bundle making one side hotter and the other side cooler. The fuel element will bow as one side tries to expand more than the other due to the temperature difference. This bowing movement moves the fissile material in the fuel element up the flux gradient in such a manner. This makes it more important to the core. This is the source of the positive reactivity coefficient.
This last statement could be labelled as speculation, since we did not have the opportunity to perform that test. In fact, the project was terminated in May, and the test was scheduled for October.
While the PCR would be negative for a core fuelled without moly-99 targets, we see that they do make a positive contribution, so putting the same number of targets back in the core would bring the net effect to approximately zero PCR--perhaps slightly positive, perhaps slightly negative--unless the targets were also modified to resist bowing. So you cut off the effect in both types of assemblies.
In conclusion, I'd like to repeat that if making moly-99 was easy to do everybody would be doing it. Other reactors may be upgraded, but will only be able to contribute small fractions of the demand. Other proposed methodologies are still in the experimental stage, and there are two MAPLE reactors, each with the capacity to deliver more than the current world requirement. Positive PCR requires a relatively simple engineering fix to restrain the bowing of the elements and reduce the PCR to approximately zero.
I thank you for your attention, and I hope this doesn't turn into another Avro Arrow.