Current developments: inspection techniques for VVER steam generators28 February 1998
In the past, advanced eddy current inspection technology was not as widely applied on Soviet-design VVERs as on Western-design PWRs. Recently, however, eddy current systems have been developed for the VVERs and a solid body of inspection experience has now been built up.
In comparison with other methods, such as ultrasonic inspection or leak testing, eddy current inspection of steam generator tubes has a number of advantages: fast data collection; absence of contact media; remote control; high reliability; permanent record of results; and low occupational doses.
The most common eddy current technique used for steam generator examination is the bobbin coil. This consists of two coils, with opposite electrical connections, wound on the probe body. The volume and depth of detected indications is determined by comparison of measured values with those for artificial defects on calibration standards. The bobbin coil is particularly suited to examinations where volumetric and predominantly axial defects are expected and where the whole tube length is to be examined, because it is fast and reliable.
However, the bobbin probe provides limited information on tube damage (depth). Also, circumferential damage is very hard to detect with a bobbin probe because it is parallel to eddy current flow and causes negligible conductivity changes. Advanced techniques are needed to give additional information about the nature of degradation.
Among these advanced eddy current techniques are rotating probes, which use single or multiple coils mounted on the probe body. Each of the coils has a limited field of view so tube inspection is achieved by translating the probe body axially as the probe head is rotated. The coil is spring loaded, riding the tube inner circumference in order to reduce lift-off effects.
Rotating probes are useful for diagnostics, to infer flaw morphology and/or orientation, for checking ambiguous bobbin coil indications, for detecting circumferential cracking and for examining tube regions where bobbin coil fill factor effects are pronounced (inner row U-bends, dented support plates, tubesheet intersections and expansion transition zones).
In the single-coil rotating design, a single pancake coil is mounted on the probe body. Circular eddy current flow in the tube wall provides equal sensitivity to axial and circumferential cracking, as well as sensitivity to volumetric tube wall degradation.
In the two-coil design, two pancake coils are mounted on the probe head 180 degrees apart and data acquisition rates can be doubled.
In the three-coil design a single pancake coil and two linearly wound orthogonally oriented coils are mounted on a common probe head at 120 degree intervals. One of the linear wound coils is more sensitive to circumferentially oriented discontinuities in the tube wall, while the other is more sensitive to axially oriented discontinuities. The pancake coil responds to both axially and circumferentially oriented discontinuities. All three coils respond to volumetric degradation. Other advanced eddy current techniques include array probes, and the plus point probe.
Techniques for the VVER
All these techniques have been successfully applied to Western PWR steam generator tubes. However, for a variety of reasons (economic as well as technical) eddy current examination of VVER units has not been developed to the same level as Western-design PWRs. INETEC of Croatia took up this challenge and has now applied eddy current techniques extensively to VVER steam generators, carrying out 100% tube inspection at many units. Today, inspection of VVER steam generators accounts for a significant portion of INETEC’s activities.
Maintaining the integrity of VVER steam generators requires periodic inspection of two essential components:
• Tubes (in VVER-440 and VVER-1000 steam generators). And
• Collectors (tubesheets) in VVER-1000 steam generators.
The standard technique for eddy current inspection of VVER steam generator tubes is the bobbin coil. It entails specially designed probes because of the VVER tube geometry, which consists of several sharp elbows along the tube length. For a variety of reasons (smaller resistance to probe movement, higher speed of probe pushing, better probe durability, etc) examination is usually accomplished in stages, from cold leg to antivibration (AV) bar 3 and from hot leg to AV bar 3. As well as standard bobbin probes, INETEC has also developed and applied specialist probes for VVER steam generator examination:
• Metal beaded. This has a special flexible hose, allowing examination of the whole tube length at once, and better penetrability in peripheral tubes with very sharp elbows.
• Metal covered. The coil is covered to give increased durability.
INETEC has also developed and used a rotating probe specially designed for VVER steam generator tubes (see Figure 3). Up until now the inconvenient geometry of VVER tubes has precluded the use of conventional rotating probe technology because the probe-mounted motor used for rotation could not be accommodated in the VVER tube configuration.
Some VVER-1000s have suffered from cracked tubesheets and this has required development of improved examination methods for detection of such cracks. For this purpose INETEC has developed a special rotating probe. Low frequencies are used to enable the magnetic field to penetrate inside the tubesheet material. This probe, as well as the examination procedure, has been qualified by INETEC and OKB Gidropress on pulled cracked tubesheet samples from the Zaporozhe PWR plant. Eddy current signals obtained on the calibration block and on a real crack, from Zaporozhe, are shown in Figure 4.
Analysing tube results
VVER-440 steam generators show significant variations in the number and distribution of indications, even those on the same plant. Figure 5 shows indication distributions in the tubesheets of two steam generators in one plant. Comparing the tubesheets, each steam generator seems to have its own pattern. In one, indications appear randomly while in the other they tend to be located to one side. The distribution depends on many factors (such as water chemistry, operating conditions, tube material properties, SG manufacturing process etc) and since the indications are often distributed over all the tubes, inspection needs to cover the whole tubesheet.
As well as variations between different steam generators in the same plant, there are also marked differences in the results obtained from successive in-service inspection campaigns. This is illustrated in Figure 6 which shows numbers of indications for all six steam generators in one unit over three consecutive inspections.
One possible conclusion is that there are no general rules governing when indications arise and their distribution over the tubesheets.
Looking at the typical distribution of indications along the tube length, we get a more consistent patters, as shown in Figure 7. The conclusion here is that indications are mainly on the hot leg side and in the U-bend on the longer side, while the cold leg side usually has significantly fewer indications. This can be explained in terms of temperature and flow distribution.
The depth distribution of indications (Figure 8) shows a normal distribution, with the greatest number of indications being between 30% and 40% wall thickness. Since the most damaged tubes are plugged after inspection (depending on plugging criteria), this distribution remains similar in subsequent examinations.
Figure 9 shows the proportion of indications where the damage depth was 60% or more of wall thickness (the plugging criterion for this particular unit). The proportion is usually about 20%, but can vary from around 10% to 30% of the total number of indications.
Generally, VVER-1000 steam generators have a small number of indications. But since they are randomly distributed over the tube bundle, examination of the whole bundle is needed. Due to the low incidence of indications, whole bundle examination can be accomplished through a series of partial in-service inspections. Inspection plans usually target different regions of the tube bundle with the aim of getting an overall picture of steam generator condition, providing timely detection of degradation processes and allowing corrective actions to be taken.
Based on inspections done to date on cracked tubesheet samples and the extensive experience INETEC analysts have obtained performing tubesheet inspections at Kozloduy, the eddy current technique that has been developed provides a sound basis for dealing with the issue of VVER-1000 tubesheet integrity. Each tubesheet should be inspected (especially cold leg) and as a minimum the scope should include the critical peripheral and wedge region, as recommended by OKB Gidropress (Figure 10).
In view of the big differences between the patterns of indications for the various steam generators in a single plant and at different plants, each steam generator needs to be analysed to determine its characteristics and establish a maintenance strategy. If any degradation mechanism is observed, the examination scope must be extended and, to monitor progress of indications, intervals between inspections need to be reduced.