Electromagnetic Non-Destructive Evaluation II, Proceedings of the 4th International Workshop on E'NDE, Chatou France, September 1998. ISBN: 90 5199 375 7 - published by IOS Press

TABLE OF CONTENTS

Introduction E'NDE Applications and NPP Maintenance The SG Tubes The Rod Cluster Control Assemblies Safety/Competitiveness Strategy Prospects

INTRODUCTION

In opening this workshop, allow me to describe the position of Electromagnetic Non-Destructive Evaluation in the maintenance of Nuclear Power Plants (NPP) and its share in the stakes of EDF. For the first time in its history, nuclear energy in Europe is on the verge of a decisive turning point. Several parameters have recently come into play to modify the scenario, such as the deregulation of the electricity market, the appearance of new means of production (fluidised bed coal boilers, small unit co-generation, etc.), but the essential parameter is the drop in the price of fossil fuels, particularly of natural gas.

Figure 1 shows the reference cost curves used to forecast the price of 1 kWh produced by an up-to-date power plant, burning different types of fuels.

You can clearly see that the benefit drawn between 1975 and 1990 from the low price of nuclear energy is going to disappear. Therefore, the main objectives of nuclear operators for remaining competitive are :

1. to reduce the maintenance costs, which represent 53% of direct operating expenses,
2. to obtain maximum availability in the European energy network,
3. to extend the life of financially amortised plants.

Figure 1 : Forecast reference cost per kWh

E'NDE APPLICATIONS AND NPP MAINTENANCE

Non-Destructive inspection of NPP components forms part of the maintenance work. By the timely diagnosis of degradation capable of implicating the safety or the availability of the plant, it enables the replacement of the necessary components to be programmed and anticipated. It represents 10% of maintenance expenditure and has been estimated for 1996 at more than 900 million Francs. E'NDE techniques, essentially those which employ Eddy Currents (ECT) represent more than half this expenditure, which is in the main devoted to two applications:
• The examination of Steam Generator (SG) tubes.
• The examination of Rod Cluster Control Assemblies (RCCA).
The strategy employed for these two examples illustrates the procedure applied also to other components.

THE SG TUBES

Figure 2 : Location of main degradations in Steam Generators tubes

These tubes are the place of transfer of all the energy from the reactor core to the turbine through the boiling of the water in the secondary circuit.

Substantial degradation in these tubes is due to the choice of the Inconel 600 alloy as the tube material. This choice was made for the great majority of PWR plants throughout the world. The stress corrosion cracks (IGSCC), appear from the primary side in the deformed zones like the U bends or the transition at the top of the tube sheet. But also it was initiated from the secondary side in the areas of concentration of pollutants, which are the deposits above the tube sheet or in the interstices between the tube and the tube support sheet. Generalised micro-cracking (Intergranular Attack IGA) due to corrosion has also been found under the deposits accompanying the IGSCC. Metal losses through wear have been observed at the anti-vibration bars (AVB) level in the top of the bundle and on the periphery in contact with loosing objects fallen into the SG. Figure 2 shows the zone where each type of defects appears.

Inspection of the tubes is done with an axial probe with two coils over the whole length of the tube. 320,000 tubes were examined in this way in 1996. In addition, to measure the length of cracks and be able to characterise them more readily, several rotating probes with two differential coils with ferrite cores are employed locally.
More than 300,000 tubes were examined in the same year.

Identification of the defects from the ECT probe signals is difficult. As can be seen above, the signals are always mixed :

• either two simultaneous defects
• or the presence of an interference signal due to the environment (such as the magnetite or copper deposit, the presence of steel sheet or bars, local deformation such as rolled expansion, bending or denting).

The policy followed by EDF has been not to plug the tubes on a defect signal associated with a theoretical depth without knowing the nature of that defect and assessing its behaviour and its harmfulness. Hence, besides the very numerous laboratory simulation studies conducted to learn how to recognise the type of defect, some tubes of the SG in operation have been pulled out for checking and validating the diagnosis made by END. Up to the end of 1996, 375 tubes were pulled out which represented approximately 2 milliard Francs for investigations, without counting the loss of production due to the duration of the removal operation.

This whole procedure has enabled EDF to define a plugging policy suited to each degradation and closely linked to the ECT signal identified.

THE ROD CLUSTER CONTROL ASSEMBLIES

Figure 3 : Location of main degradations in RCCA rodlets

These are the fine rodlets, 10 mm in diameter, the 0.5 mm cladding of which encloses a neutron absorber which enables the fission reaction to be controlled and stopped. They are grouped in clusters of 24 rods and inserted into different zones of the core.

Wear and tear, due to vibrations induced by the circulation of the fluid in the core, appears in the areas of friction by the rod against its guide. Locally it can lead to the puncturing or indeed the fatigue rupturing of the rod. For certain clusters, the end of which is inserted into a dense neutron flux, the irradiation causes swelling of the absorbent which expands the cladding until it cracks. The swelling also leads to a slowing down in the descent of the cluster or its jamming. Figure 3 shows the location of the defects.

The examination of each rodlet by an encircling ECT probe allows the presence of defects to be discerned. 24 coils enable the simultaneous inspection of a complete cluster. In 1996, the number of clusters examined was 32,000.

The hot laboratory investigation of several rods has enabled a relationship to be established between the mean section of wear and the ECT signal. However, some kinds of wear in the form of a V were poorly estimated. Similarly, in the swollen area, the cracking may be masked by the swelling signal. The solution to these two problems could not be found by the eddy current method. Therefore, profile measurement by a rotating ultrasonic probe is used each time the ECT signal in any area of the rod exceeds a predetermined threshold.

The establishment of criteria linked to the ECT signal has enabled a policy to be defined for scrapping a cluster or for using it for one or two cycles after its neutral position has been modified.

SAFETY/COMPETITIVENESS STRATEGY

The efforts expended in finding out the true mechanisms of degradation and the soundness of the END diagnoses, demonstrated in numerous cases of verification by analysis of the parts removed, have enabled the safety of these components to be established. This policy necessitates perfect reproducibility of the measurements and their interpretation over the years in order to guarantee rigorous monitoring of the kinetics of evolution of degradation. For example, the accuracy of the length of the cracks in the tubes is of the same order as their lengthening in one cycle; this enables them to be monitored over several cycles before reaching the size required for plugging which, that too, leaves a margin in relation to the critical size defined by the safety studies.

The strategy therefore has consisted in :

• stabilising the type of probe (with stringent requirements on manufacturing acceptance),
• EDF developing algorithms for processing the signal and for diagnosing the defects,
• enforcing service companies utilising these measures in their own industrial tooling.

Thanks to this strategy, productivity gains have been obtained by placing performances in competition on the basis of quality and rapidity in data acquisition. Similarly the reliability of the diagnosis over a long period has enabled savings to be made in expenses, for example :

• by establishing a programme for replacing the SG in the medium term with anticipation of orders
• or of eliminating a succession of cluster inspection campaigns to arrive directly at their replacement.

PROSPECTS

The development of the policy in matters of safety is pushing the regulators to demand that processes be qualified by an independent organisation. This is going to lead to formalising requirements and to conducting demonstrations of performances. So as to limit the tests, there will increasingly be recourse to mathematical modelling in order to explore the parameters of the methods. Similarly, the common use of recognised mock-ups (via the ENIQ for example) will allow the costs of creation to be shared.

As has been noted above, to reach a final diagnosis, two or more ECT probes are employed, or indeed another method, that of ultrasonics, on the same component. To reduce the examination times, and therefore costs, we are going to try using multi-technique "probes" simultaneously. The data processing systems should be available commercially and should enable a diagnosis to be produced in real time. Despite the use of different techniques and different service providers, it must be possible to compare easily the results from one campaign to another, hence the obligation to work with a common data format. These three notions (shown in bold) constitute the main lines of progress along which EDF is counting on moving in the future with its partners.