Hiromitsu Ino *
Outline of Neutron Irradiation Embrittlement in Aging Nuclear Power Plants
Destruction of a reactor pressure vessel due to neutron irradiation embrittlement should be called an extreme severe accident. If the pressure vessel breaks, there is almost no way of preventing a runaway chain reaction. Such extreme damage must be avoided at all costs.
The benchmark for irradiation embrittlement is the ductile-brittle transition temperature (DBTT). If an extreme situation arises, such as pipe rupture due to an earthquake, it is necessary to cool the core using the emergency core cooling system (ECCS). However, if the DBTT is high, this becomes a dangerous operation. When cooled suddenly, a temperature difference arises between the inner and outer walls of the pressure vessel and strong tensile stress is brought to bear on the inner wall. If such tensile stress is applied when the temperature is below the DBTT, there is a danger that cracks could occur in the pressure vessel wall, leading to failure of the pressure vessel and a severe accident.
Table 1 shows Japanese nuclear power reactors in descending order of the DBTT of their pressure vessels. The table shows seven reactors in which DBTT exceeds 50°C. They are all old reactors that began operating in the 1970s.
|Table 1: Reactor Pressure Vessel Ductile-Brittle Transition Temperature (DBTT) – Worst 7|
|Source: Prepared by the author from “Results of Monitoring Tests on Steel in Nuclear Reactor Pressure Vessels,” CNIC
As of July 2011. A DBTT of 95 °C was later observed in Takahama-1.
Genkai-1 is the worst. The DBTT for this reactor was announced in October 2010. The figure comes from the most recent test of monitoring specimens in April 2009. The DBTT rose 42°C since the previous test result of 56°C in February 1993. This is a new record for Japan. This reactor will be discussed in detail in NIT 149.
All the reactors listed from second to fifth place in the table are located in Fukui Prefecture and are owned by Kansai Electric Power Company (KEPCO). In particular, we have been concerned about the continued operation of Mihama-1&2, where high DBTTs have been observed since the beginning of the 1990s. KEPCO asserts that results of pressurized thermal shock (PTS) analysis show that even if the ECCS was used in the event of a pipe rupture the pressure vessel would not fail. However, the evaluation methodology for the stress arising, KI, has not been released, so it is impossible to know whether this analysis is reliable.
PTS analysis assesses the pressurized thermal shock to the core of PWR pressure vessels in the case of accidents such as loss of coolant accidents and main steam pipe ruptures. It is necessary to confirm that the critical stress intensity factor KI does not exceed the fracture toughness KIC.
The reactors listed in sixth and seventh places in Table 1 are BWRs. The inner diameter of BWR pressure vessels is large compared to PWRs and the amount (flux) of neutron irradiation received in a given time is one or two orders of magnitude less than in PWRs. From the table it can be seen that the total amount (fluence) of irradiation received by Tsuruga-1 is about one thirtieth of that of Mihama-1, even though they began operating at much the same time. (There is a slight difference in operating time and also in the date the specimens were taken.) Consequently, it was thought that neutron radiation embrittlement was not such a big problem in BWRs as it was in PWRs. (Even now many researchers and engineers are still in the grips of that “common sense.”) However, after many years of operation, as we came to know the reality of irradiation embrittlement in BWRs, this “common sense” has been overturned. The total amount (fluence) of irradiation is not the only determining factor for irradiation embrittlement. It has become clear that the rate (flux) at which irradiation occurs is also a determining factor. As will be discussed in part two, this led to an amendment to the monitoring specimen method JEAC-4201 and to the situation where two BWRs are now listed among the worst seven and other BWRs are also known to have high levels of irradiation.
Why Does Irradiation Embrittlement Occur? – Basic Concept
Metal materials become degraded for all sorts of reasons. One reason is “radiation damage.” This phenomenon is investigated at the atomic level though the study of lattice defects. The Physical Society of Japan has had a section on lattice defects for over 50 years. As a personal note, I have devoted myself to this field of research since becoming interested in it as a university student. I became a tutor at Osaka University and experienced the student uprisings of the 1960s. In hindsight I can see that this field of research, which originated in the United States, developed in tandem with nuclear energy. Nevertheless, that fact did not lead me to abandon the field. I carried out materials research using radiation as a guest researcher at the Kyoto University Research Reactor Institute. However, it was difficult to see a connection between this research and the social problems associated with nuclear energy.
The reason why irradiation defects became an important research theme was because when neutrons generated by nuclear fission hit reactor vessels and pipes they damage the metal materials. This is called “neutron radiation damage.” If this causes materials to become brittle, it is called “neutron irradiation embrittlement.” Of particular importance is neutron irradiation embrittlement of the steel of the reactor pressure vessel, which is the heart of a nuclear power plant. If this is damaged it can lead directly to a severe and uncontrollable accident.
What type of lattice defects arise from neutron radiation? In crystals, atoms are precisely aligned in lattices, but if they are struck by a neutron they are displaced, leaving a hole. This is called a “vacancy.” Displaced atoms are called “interstitial atoms.” This phenomenon is called a “lattice defect.” In addition, secondary defects result when vacancies and interstitial atoms move about and accumulate, creating “vacancy clusters” and “interstitial atom clusters,” respectively. Impurities within the metal (copper atoms etc.) move to form “impurity clusters.” These “secondary lattice defects” cause metals to lose their characteristic ductility (plasticity) and become brittle. To compare it to the human body, it is like the hardening of the arteries which makes blood vessels vulnerable to rupture.
Usually, when a force is applied to steel it simply deforms without breaking, but below a given temperature, if the slightest force is applied, rather than deforming plastically it shatters like pottery. This critical temperature is called the ductile-brittle transition temperature (DBTT). This brittleness of steel used to be the bane of shipbuilders. Many ships sank due to this phenomenon. The Titanic, which sank exactly 100 years ago in 1912 when it struck an iceberg while crossing the North Atlantic Ocean, is a famous example. Subsequent studies showed that poor quality steel plate was used and that the DBTT was 27°C.
When reactor pressure vessels are bombarded by neutrons the DBTT rises. When designing nuclear reactors it is necessary to predict how high the DBTT will rise and whether they can survive for the period of their design lives. However, assuming a design life for nuclear reactors of 40 years, it is impossible to know what condition they will be in after 40 years until the 40 years has actually elapsed. That presents a problem, so accelerated experiments are conducted. Accelerated experiments are tests that are commonly used to assess endurance by, for example, applying forces beyond the normal load, or operating plants at greater than normal speed.
Likewise, when conducting tests for neutron irradiation embrittlement, the amount (flux) of neutron exposure in a given period of time is increased far above normal amounts. Materials test reactors can radiate materials at a rate of 1012n/cm2s (neutrons/square centimeter). This rate (flux) of exposure is between 100 and 10,000 times the rate of exposure in normal reactors, given that the rate of exposure for PWRs is 1010n/cm2s , while the rate for BWRs is 108n/cm2s . That means the amount of irradiation a BWR would sustain in 40 years can be applied in one or two days. Using such data a formula predicting embrittlement was produced. Furthermore, besides the normal monitoring specimens, accelerated monitoring specimens are also placed in BWR reactor vessels. They are placed not on walls of the vessel itself, but closer to the core, where the rate (flux) of radiation is an order of magnitude higher. The idea is to predict the future state of the reactor. Likewise, monitoring specimens are placed deeper inside PWRs than the walls of the reactor vessel. For example, in the case of Genkai-1, discussed in part two, the rate of radiation is about double the normal rate. This is an attempt to read the future.
However, there is an assumption underlying the notion that the future can be predicted. That is, regardless of the rate (flux) of irradiation, or, to put it another way, regardless of the period of exposure, if the total amount (fluence) of radiation is the same, the result will be the same. The formula for this assumption is as follows:
Rise in DBTT = material factor x F(f)
The material factor is determined by the type and the concentration of impurities in the steel. For example, if the steel contains a large amount of copper, the material factor will rise. F(f) is the irradiation factor. It is postulated to be a function of the fluence of neutron irradiation “f” alone.
With accumulated experience of operating nuclear power plants, it became possible to obtain long-term monitoring test data in real life conditions, and it became clear that this formula was suspect. In particular, with regard to BWRs, for which the rate of irradiation is slower, it became clear that the results for the normal monitoring specimens and the accelerated monitoring specimens placed in reactors did not agree. This trend is particularly pronounced in reactors like Tsuruga-1 and Fukushima Daiichi-1 where the steel of the reactor pressure vessels contains large amounts of copper impurity. It can be seen from this that the irradiation factor F(f) is dependent not only on the fluence (total amount) of neutron irradiation “f“, but also on the flux (amount in a given time) of irradiation.
We noticed this over ten years ago and alerted researchers to the issue. However, at the time, the results of American research refuting dependence on the flux of irradiation held sway, so Japanese researchers refused to take the matter seriously and did not alter the embrittlement prediction formula. Faced with data from Tsuruga-1 showing unpredicted high levels of DBTT, METI’s Aging Response Review Committee dismissed the results saying they were due to data scatter.
Thereafter, analysis of the micro-formation of copper progressed, and it became clear that when the rate of radiation is slow mainly clusters of copper atoms (obstructions) form, whereas in accelerated irradiation tests mainly clusters of vacancies form, so the cause of the hardening (embrittlement) is different. The results of this micro-analysis backed up our computer simulations. The outdated thinking described above was forced to give way and now the dependence of radiation embrittlement on the flux of irradiation is the shared academic understanding. The irradiation embrittlement prediction formula used in monitoring test methodology was changed and a new methodology (JEAC 4201-2007) was produced＊. Assessment of pressure vessels shifted to the 2007 formula from mid-2011, but when the increase of DBTT using this formula is smaller than that using the previous 2004 formula, the 2004 formula is included as a reference.
However, even the 2007 formula cannot explain the high DBTT for metal welds in Tsuruga-1 that we have drawn attention to. The metal welds in Tsuruga-1 have low levels of copper impurities, unlike the parent metal, and thus should not show a high DBTT. The amended JEAC-2007 was not adequately able to explain the complex nature of the reality of the metal materials.
Unpredicted Embrittlement in Genkai-1 Reactor Pressure Vessel
Further “unpredicted” monitoring specimen data were observed; these were the results from Genkai-1. At the October 25, 2010 meeting of Karatsu City Municipal Assembly’s Pluthermal Special Committee, Kyushu Electric Power Company announced that the DBTT observed in Genkai-1’s fourth monitoring test specimen, taken during a periodic inspection in April 2009, had reached 98°C. Previously, the highest DBTT for a reactor pressure vessel had been 81°C for metal taken from a weld at Mihama-1 (see Table 1). The Genkai-1 specimen exceeded this, so it would be fair to conclude that Genkai-1 is the most dangerous reactor pressure vessel in Japan.
|Figure 1: Genkai-1 Monitoring Test Sample Data and JEAC 4201-2004 Prediction Curve|
It is also very significant that this embrittlement was unpredicted. The DBTT observed in the previous (third) monitoring test (February 1993) was 56°C. That had increased by 42°C, which was contrary to the predicted result. Figure 1 is a diagram submitted by Kyushu Electric in its December 2003 Aging Technical Assessment, with a “×” added to the top right corner to show the result of the fourth monitoring test. Up until the third monitoring test the data points could be more or less plotted onto the predicted curve, but the latest data point is way above that curve. If you look closely at the diagram you will see that the broken line is the predicted curve and that a line is added above that showing the upper limit of the margin for error. However, actual embrittlement is way above that upper limit.
Kyushu Electric says that 98°C is the value predicted for 2060 (85 years after the start of operations), while the predicted DBTT for 2035 (60 years after the start of operations) is 91°C and for August 2010 (35 years after the start of operations) is 80°C. In part two, let us consider whether this is correct or not.
(To be continued in the next issue of Nuke Info Tokyo)