Doel 3 & Tihange 2: flaw indications in the reactor pressure vessel steel
Case chronology and scientific background
Summer 2012: discovery of flaw indications
In the summer of 2012, the reactor pressure vessel of Doel 3 was inspected using ultrasonic probes as part of a scheduled periodic 10-year safety review. In additional to usual ultrasonic inspection, specific ultrasonic inspection was performed in order to test the weldings for the presence of ‘under-clad defects' (gaps between the steel and inox inner surface of the reactor pressure vessel). No under-clad defects were found, but the ultrasonic test did reveal other unexplained flaws inside the steel wall of the reactor pressure vessel. It was subsequently decided to maintain Doel 3 shut-down until the nature and source of the flaw indications was identified.
In September 2012, the reactor pressure vessel of Tihange 2, that was manufactured by the same company (Rotterdamsche Droogdokmaatschappij), was subjected to the same ultrasonic testing. Tihange 2 proved to have similar flaw indications to Doel 3.
The flaws were soon revealed to have been caused by hydrogen flaking. When steel components are manufactured, there is a risk of excess hydrogen building up in the cast steel during the process of cooling and casting. When the hydrogen evaporates it leaves behind tiny bubbles in the steel. In this case, the bubbles were rolled flat into tiny hairline cracks during the forging process.
These tiny bubbles have an average length of 1.2 to 1.6 cm and the thickness of a cigarette paper. These are known as hydrogen flakes. The hairline cracks are laminar, which means they run parallel to the curve of the wall of the reactor pressure vessel.
The occurrence of hydrogen flakes in steel structures is a well-known and widely studied phenomenon in metallurgy. However, this was the first time the phenomenon had been observed in the walls of the reactor pressure vessel of a nuclear power plant in operation.
Following this discovery, the FANC ordered that the Doel 3 and Tihange 2 nuclear reactors were not to be restarted until Electrabel had conclusively proved that the presence of the hydrogen flakes did not form a risk for the integrity and hence the safety of the reactor pressure vessels.
>> Link to the FANC Technical Information Note issued on 03/09/2012
>> Link to the FANC Technical Information Note issued on 15/10/2012
May 2013: The FANC authorised restart
Following a series of extensive studies, Electrabel submitted two Safety Case reports in December 2012, in which it argued that the safety of Doel 3 and Tihange 2 was not at risk despite the presence of hydrogen flakes in the walls of the pressure vessels. In January 2013, the FANC concluded that there was no cause for permanent shutdown of these plants, however they did require that Electrabel provide them with additional information. The licensee subsequently submitted two addenda to the Safety Case reports.
In May 2013, after carefully analysing the case documents and consulting with various national and international experts, the FANC concluded that the evidence provided by Electrabel answered satisfactorily to short-term safety concerns and was sufficient to guarantee the safe operation of the Doel 3 and Tihange 2 reactors for one cycle, until answering to long-term safety concerns. On 17 May 2013, the FANC authorised the restart of the reactors for one cycle of operation.
FANC did require Electrabel to asnwer before the end of the cycle a number of long-term safety concerns, requiring the reactors to be operated to be answered on this authorisation.
>> Links to the Electrabel Safety Case Reports:
- Safety case report: Doel 3 - Reactor Pressure Vessel Assessment
- Safety case report: Tihange 2 - Reactor Pressure Vessel Assessment
>> Links to the Electrabel Health Physics Department's independent analyses on the Doel 3 and Tihange 2 Safety Case Reports:
- Report on independent analysis and advice regarding the safety case - Doel 3 Reactor Pressure Vessel Assessment
- Report on independent analysis and advice regarding the safety case - Tihange 2 Reactor Pressure Vessel Assessment
>> Link to the FANC Provisional Evaluation Report issued on 30/01/2013
>> Link to the Report of the National Scientific Expert issued on 11/01/2013
>> Link to the Final Report of the International Expert Review Board issued on 15/01/2013
>> Link to the FANC Technical Information Note issued on 01/02/2013
>> Link to the Tihange 2 Safety Case Report Addendum
>> Link to the Doel 3 Safety Case Report Addendum
>> Links to the Electrabel Health Physics Department's independent analyses on the Doel 3 and Tihange 2 Safety Case Report Addenda:
- Report on independent analysis and advice regarding the safety case addendum - Tihange 2 Reactor Pressure Vessel Assessment
- Report on independent analysis and advice regarding the safety case addendum - Doel 3 Reactor Pressure Vessel Assessment
>> Link to the FANC Final Evaluation Report
>> Links to the reports supporting the decision of the FANC:
- Tihange 2 – ISI 2012 – Justification of the Reactor Pressure Vessel (RPV) shell (AIB-Vinçotte)
- Doel 3 – ISI 2012 – Justification of the Reactor Pressure Vessel (RPV) shell (AIB-Vinçotte)
- Safety Evaluation Report - Flaw indications in the RPV's of Doel 3 and Tihange 2 (Bel V)
- Report of the National Scientific Expert Group on the RPV Doel 3
- Report of the National Scientific Expert Group on the RPV Tihange 2
March 2014: fracture toughness test produced unexpected results
One of the measures that Electrabel had to complete before the end of the first reactor cycle involved mechanical testing of irradiated steel components affected by hydrogen flakes. The aim of these tests was to find out to what extent exposure to intensive radiation influences the mechanical properties of steel that contains hydrogen flakes. Electrabel was required to conduct these tests in order to empirically confirm some hypotheses and calculations in its Safety Case reports.
The tests used samples provided by a French steam generator (“VB395”) that was constructed for use in a nuclear power plant but was rejected because the steel contained hydrogen flakes. The samples were first placed in the research reactor of the Belgian Nuclear Research Centre (SCK•CEN) in Mol, where they were intensively irradiated for four weeks in order to simulate the cumulative radiation exposure of 40 years of operation. The samples were then subjected to a series of tests to evaluate their mechanical properties. All the tests produced the expected results, with the exception of one test: the fracture toughness test. The results of this test suggested that, under irradiation, the steel from the VB395 steam generator samples became brittle much faster than expected.
Because the researchers did not have an immediate explanation for these unexpected test results, it was decided on 25 March 2014 to shut down the Doel 3 and Tihange 2 reactors again as a precautionary measure. The FANC ordered that the reactors were not to be restarted until Electrabel had conclusively proved that the presence of the hydrogen flakes in the reactor pressure vessels did not have a greater influence on the steel than initially assumed, despite these unexpected results. During the following months, Electrabel commissioned the SCK•CEN to conduct a number of new tests.
Because the study of the mechanical properties of irradiated materials is a highly specialised research domain, the FANC called in the help of a number of leading international experts in this area to review this particular aspect of the Safety Case. This team of experts (the International Review Board) was asked to make a statement on the methodology that Electrabel should use for those parts of the Safety Cases that were related to the mechanical properties of steel under the influence of radiation.
February 2014: number of detected hydrogen flakes re-evaluated
Alongside the mechanical tests, Electrabel was also required to complete a number of other actions. A number of these actions were related to the qualification of the ultrasonic testing method with which the hydrogen flakes were discovered in the summer of 2012. Electrabel was required to provide more evidence that the testing method used was capable of detecting and accurately sizing and locating all the hydrogen flakes.
The new tests revealed that the ultrasonic inspection method was able to identified flaw indications if any, but that the detection thresholds needed to be adjusted in order to be able to guarantee that literally every hydrogen flake was detected and accurately sized. It was also concluded that the method used to interpret the test results left room for improvement.
In 2014, new ultrasonic tests based on this new qualified UT procedure were conducted on the reactor pressure vessels. The use of lower detection thresholds led to the detection of some 60% more flaw indications than during the previous inspection. The researchers also noted an increase in the average and maximum length of the cracks. This was because the new interpretation method prescribed that hairline cracks that were close together were to be treated as single elongated cracks.
However, an analysis of the results from the 2012 and 2014 ultrasonic tests – both conducted using the same interpreting method and detection thresholds – revealed that there had not been any evolution in the number and size of the hydrogen flakes during the 1-year operation of the two reactors between 2012 and 2014.
The FANC called in the expertise of the certified inspection company AIB-Vinçotte to evaluate the results of the qualification of the ultrasonic testing method.
July 2015: Electrabel submitted its Safety Case Reports to the FANC
On 17 July 2015, Electrabel submitted its Safety Case Reports for Doel 3 and Tihange 2 to the FANC.
As required by the FANC, these Safety Case Reports focussed on three major themes, whereby the research results of themes 1 and 2 were to support the evidence for theme 3:
1. Detecting, sizing and locating the flaw indications.
2. Determining the mechanical properties of the material containing hydrogen flakes and the influence of irradiation on these mechanical properties.
3. Assessing the structural integrity of a reactor pressure vessel that contains hydrogen flakes.
The FANC forwarded copies of the Safety Case Reports to the external experts involved in the definitive safety review.
The certified inspection company AIB-Vinçotte wrote an evaluation report on the first theme.
The Oak Ridge National Laboratory (USA) was commissioned to conduct a thorough evaluation of the hypotheses, methodology, calculations and interpretation of the Electrabel's results. The US laboratory subsequently carried out structural integrity calculations using its own hypotheses, methodology and alternative computer codes.
Bel V, the FANC's technical branch, conducted a separate evaluation of the entire Safety Case.
Finally, a working group of four Belgian professors considered the question of whether the hairline cracks might grow due to the migration of hydrogen through the walls of the reactor pressure vessel.
The reports of the various external experts were sent to the FANC over the following months and compared with the conclusions of the FANC's own experts.
November 2015: the FANC authorised the restart of Doel 3 and Tihange 2
Based on its own review and the reports of the various external experts, the FANC concluded that there was no cause to prevent the Doel 3 and Tihange 2 reactors from restarting.
Electrabel was able to demonstrate that the unexpected test results of March 2014 were probably due to the specific material properties of the specimen that was used. Tests on another material and on the material of the reactor pressure vessels themselves demonstrated that long-lasting irradiation will not have an abnormal impact on the mechanical properties of the reactor pressure vessels of Doel 3 and Tihange 2.
The structural integrity of the reactor pressure vessels of Doel 3 and Tihange 2 complies with the safety requirements and the presence of hydrogen flakes does not pose a risk to the safety of the nuclear power plants.
>> Links to the Electrabel's Safety Case Reports:
- Safety Case 2015 - Doel 3 - Reactor Pressure Vessel Assessment
- Safety Case 2015 - Tihange 2 - Reactor Pressure Vessel Assessment
>> Links to the Electrabel Health Physics Department's independent analyses on the Doel 3 and Tihange 2 Safety Case Reports:
- Report on independant analysis and advice regarding the Safety Case 2015 - Doel 3
- Report on independant analysis and advice regarding the Safety Case 2015 - Tihange 2
>> Links to the reports supporting the decision of the FANC:
January 2016 : The FANC invited safety authorities to international information meeting
On 11 and 12 January 2016 the FANC hosted an international workshop in Brussels for the safety authorities. About fifty experts originating from fifteen countries (Austria, Belgium, Czech Republic, England, France, Germany, Hungary, Japan, Luxemburg, Netherlands, Poland, Spain, Sweden, Switzerland and United States) - along with EU, OECD and IAEA experts - took part in this meeting.
June 2017: No evolution of hydrogen flakes: Full Doel 3 and Tihange 2 inspection report
On 17 November 2015, FANC gave the go-ahead for the Doel 3 and Tihange 2 reactors to restart after three and a half years of investigations following the discovery of microscopic hydrogen flakes in both reactor vessels. At the end of this assessment period, and based on the opinions of a number of independent expert groups, FANC reached the conclusion that all safety concerns raised by this issue had been satisfactorily addressed. Nonetheless, FANC stipulated that ENGIE Electrabel must carry out re-inspections of these reactor vessels at regular intervals in line with recommended practices for pressure equipment. As a result, FANC required ENGIE Electrabel to reinspect the two reactor vessels during the first unit outage for refuelling, and at least every three years thereafter, using the same qualified method for detecting microscopic hydrogen flakes.
Given the specific nature of the Doel 3 and Tihange 2 reactor fuel cycles (12 and 18-month periods respectively), and their effective restart dates, these re-inspections were carried out by Electrabel in November 2016 on the Doel 3 reactor vessel and in April 2017 in the case of the Tihange 2 reactor vessel.
These inspections were conducted using an ultrasonic testing technique with a view to monitoring the population of hydrogen flakes. The hydrogen flake population regarded as the reference basis for comparison purposes is the figure obtained during the inspections carried out in 2014, since the accuracy of the measuring equipment and the procedure for interpreting results were improved significantly at that time. The ultrasound inspection procedure and the method used to interpret the results are the same as the procedures used in 2014 and formally qualified by the safety regulator at the end of 2015 for detecting, locating and measuring the size of hydrogen flakes.
Following the re-inspections carried out at Doel 3 in November 2016 and at Tihange 2 in April 2017, FANC published the main conclusions of its assessment on its website. Based on this information, it had no objection to the reactors restarting and resuming their next fuel cycles.
The purpose of this document is to provide the full results of these re-inspections and to set out the safety regulator's detailed conclusions on this subject. It also underlines FANC's transparent communication policy.
In the first instance, FANC stresses that the characterisation of hydrogen flaw indications using a non-destructive ultrasonic testing method is an experimental technique with results that vary from measurement to measurement. This technique is based on the principle of emitting ultrasound signals through the metal of the reactor vessels and measuring the extent to which the reflected signals are attenuated to provide an image of any hydrogen flakes contained in this metal. These ultrasonic inspections are performed by placing an ultrasound emitting and measuring machine inside the reactor vessel. This machine can be used to inspect all the vessel shells in approximately 2 cm stages, thus resulting in the characterization of flaw indications with a relatively high level of accuracy. However, this depends on a wide range of parameters such as the focal length of the machine, the thickness of the reactor vessel, the position of the flaw indication, or properties such as the water temperature inside the reactor, etc.
The re-inspection process is made even more complex by the fact that repositioning the machine at exactly the same starting position in each re-inspection is a major challenge.
As a result, the properties of all hydrogen flakes detected in the reactor vessels are subject to slight variations between re-inspections due to the technique and measurement positioning. This was evident when comparing the results of the inspections conducted in 2013 and 2014 using an identical inspection procedure (although this was less accurate than the qualified procedure used in 2015). This phenomenon is well known to experts when detecting flaws by ultrasound methods and, as a result, strict criteria have been defined to assess whether or not a variation in the size of a flaw indication is representative of a genuine change in the size of this flaw. In other words, for the very same flaw, the figures used to define its size and inclination as obtained during an inspection may differ very slightly from the results recorded during previous inspections.
The documents attached to this report provide full and detailed results of the re-inspection procedures carried out at Doel 3 in 2016 and at Tihange 2 in 2017, compared with the reference population recorded during the inspection performed in 2014. As predicted by theory and anticipated in the inspection procedures, these documents reveal size variations in practically all hydrogen flakes detected in the vessel shells. These slight variations are entirely in line with the criteria used to confirm that these hydrogen flakes have not changed in size for a number of reasons: statistical analyses show that the distribution of the size variations of the hydrogen flakes is around zero across the entire hydrogen flake population. In other words, the number of hydrogen flakes with larger dimensions is cancelled out by the number of hydrogen flakes with smaller dimensions. On the other hand, the size variations recorded (both decreases and increases) are fully in line with the no-change criteria. This suggests that size variations are less than the variations regarded as acceptable when repeating the ultrasonic inspections. In other words, these variations are not representative of an actual phenomenon, but are due to the process used to measure the hydrogen flakes and in no way reflect any deterioration in the strength of the vessel or, as a logical consequence, the safety of the reactor as a whole. The general public, workers and the environment are therefore still protected from the negative effects of ionising radiation.
Just over 300 additional flaw indications at Doel 3 and 70 additional flaw indications at Tihange 2 also exceeded the recording threshold for the first time during the re-inspections carried out in 2016 and 2017 respectively. During the 2014 inspections, these flaw indications were actually present, but the ultrasound signals that they reflected were too attenuated to exceed the threshold defined in the acceptance criteria for regarding a signal as a flaw indication. Due to a number of changes in the inspection conditions, these flaw indications exceeded this recording threshold during the recent re-inspections. Along similar lines, other flaw indications have moved in the opposite direction, falling below the recording threshold to the extent that they no longer appear in the flaw population reported in 2016 and 2017.
Since we have been able to find scientific explanations for all these newly reported hydrogen flakes, or they have been accounted for by signals recorded in previous inspections, the analysis of these results allows us to conclude that no new hydrogen flakes have appeared and that there has been no change in the size of the hydrogen flakes already detected.
Full results of the ultrasonic re-inspection of the Doel 3 reactor vessel and partial results from the Tihange 2 reactor vessel re-inspection are provided in the appendix to this document. Given the time required to analyse and interpret the results in full, FANC has agreed that Electrabel is entitled to request that the reactors be started up again after a partial analysis of the results. However it has stipulated that the full analysis should be completed within three months of restarting. In order to guarantee safety, the safety regulator and Electrabel were meticulous in defining the hydrogen flake sample to be analysed during the provisional assessment process: this sample contains all hydrogen flakes with the largest dimensions and the greatest inclination, but also those that are most detrimental when calculating the structural integrity of the reactor vessels, with a minimum of 500 hydrogen flakes. The licensee is also required to check that there are no radial connections between two hydrogen flakes across the entire vessel, and that no new flaws have come to light over a preselected volume. In the case of Doel 3, the results of the full analysis confirmed the results of the provisional analysis across the board. Electrabel must provide the full analysis results for all hydrogen flaw indications for Tihange 2 by September 2017.
The parties involved
Due to the unique and complex nature of this case, the FANC set up a specific review process. It called in the help of numerous experts and took its final decision taking into account the opinions of each expert group as well as its own assessment. The role of each party involved in this process is described as follows:
- Electrabel: demonstrate to the FANC that the structural integrity of the Doel 3 and Tihange 2 RPV is not jeopardized by the presence of thousands of hydrogen flakes.
- FANC: review the Safety Case Reports submitted by Electrabel and the opinions issued by the experts acting on its behalf and, as a result, take a final decision on the future of each reactor.
- Bel V (technical branch of the FANC): along with the FANC – review thoroughly the Safety Case Reports submitted by Electrabel.
- AIB-Vinçotte (certified inspection company) : specialist in ultrasonic testing techniques and the analysis of data related to the detection, location and sizing of hydrogen flakes. Along with the FANC – review thoroughly the Safety Case Reports submitted by Electrabel.
- Belgian Nuclear Safety Authority: FANC, Bel V and AIB-Vinçotte
- International Expert Review Board (2012-2013) (panel of experts who are specialized in the reactor pressure vessels): review the studies conducted by Electrabel to assess the Doel 3 and Tihange 2 reactor pressure integrity.
- International Review Board (2014-2015) (panel of experts who are specialized in the effects of radiation on the mechanical properties of materials): review the tests conducted by Electrabel to assess the impact of radiation on the Doel 3 and Tihange 2 reactor pressure vessel steel embrittlement.
- National Scientific Expert Group (2012-2013) (group of four Belgian university professors, acting on behalf of the Scientific Council for Ionising Radiation): assess and challenge the safety demonstration provided by Electrabel.
- National Scientific Expert Group (2014-2015) (group of four Belgian university professors, acting on behalf of the Scientific Council for Ionising Radiation): assess the consistency of the hypothesis of a potential flaw growth due to a diffusion of molecular hydrogen from the RPV to the walls, leading to the growth of the detected flakes.
- Oak Ridge National Laboratory (US laboratory): conduct a thorough and independent evaluation of the Electrabel Safety Case Reports on behalf of the FANC.
What about the other Belgian reactors?
- An ultrasonic test was conducted on the Tihange 1 reactor pressure vessel during its planned outage for refuelling in May 2013. No hydrogen flaking was detected.
- The Tihange 3 reactor pressure vessel was also subjected to similar tests during its outage in September 2013. No hydrogen flaking was detected.
- The reactor pressure vessels of the Doel 4 and Doel 1 units were inspected respectively in September and October 2015. No hydrogen flaking was detected.
- The inspection of the Doel 2 RPV was conducted in November 2015. No hydrogen flaking was detected.
|08/06/2017 - 16:11|