Circumferential austenitic welds of main circulation piping, made of 08Kh18N10(12)T forged stainless steel, have been examined by semi-mechanised UT system at Dukovany NPP since 1998. The examinations have been performed by ŠKODA Nuclear Engineering. Qualified inspection procedures have been used for mechanised UT inspections by Nuclear Research Institute REZ (NRI) especially for the inspection areas selected and examined within Phare project PH 1.02/94 since 1999. At the beginning three inspection areas have been selected to be examined preliminarily outside the ISI programme for the first time in the plant history. Within 3 outages in May, July and September 1999 the staff of NRI examined together 10 welds. Four 245/500 mm nozzle to MCP welds (at Unit 1 and Unit 2), two 108/500 mm nozzle to MCP welds (at Unit 1 and Unit 2) and four steam generator collector dissimilar welds (at Unit 2 and Unit 4 on cold and hot legs).

Within examinations of 245/500 mm and 108/500 mm nozzles to MCP welds have been used, according to the experience received within Phare project 1.02/94, five probes for detection (0°longitudinal wave mode 5 MHz probe, 45°shear wave 2 MHz probe and three focused longitudinal wave 2MHz probes (55°, 60°and 70°) and a similar set of probes as well for sentencing. In majority of cases the best sizing has been reached in compliance with the results valid for practical trials performed on open test assemblies with 60°probe focused on 35 mm sound path with a suitable 2 MHz frequency for austenitic forged steel weld.

In the frame of UT mechanised inspections of steam generator collector dissimilar weld two probes (55°shear wave 1, 5 MHz and 60°longitudinal wave 1, 5MHz) were selected for practical trials and also used within in-service inspections in correspondence with Phare project PH 1.02/94 results. Similarly like in some other cases 60°probe provide usually good results. In spite of the fact that the qualification of inspection procedures revealed some weak points, which will be solved in the near future by additional open tests, it can be concluded that mechanised UT in-service inspections of all the three above inspection areas would be very difficult to be conducted without previous experience gained within inspection procedures qualification. Substantial deficiencies in case of missing qualification trials on open test assemblies could be expected especially in the following aspects:

Incorrect understanding of signals coming from geometrical indications especially for saddle type complex weld geometry,

A small practice with virtual angles application for indication sentencing,

A very low justification for sizing of various types of specific and postulated defects as lacks of fusion, fatigue cracks, etc.

As a conclusion can be told that the above qualified inspection procedures meet the requirements of operator Dukovany NPP. Remaining deficiencies will be removed by additional testing and practical trials.