Rack with samples exposed in the reactor vessel of the GKN boiling-water reactor in Dodewaard

Figure 1

Flow scheme of the Dodewaard BWR in 1981

Figure 2

Position of the sample rack in the reactor vessel

Figure 4.

The sample rack, ready for exposure in the reactor vessel.

Figure 7

SEM pictures of the CRUD, deposited on C-steel (C5), AISI 304 (E-5), 254SMO (E-4) and Ur50 (B-2)

Figure 8

Optical microscopic pictures of the corrosion pits on C-steel (A) and 10CrMo9.10 (B)

Deposition of CRUD in BWR water on various steels exposed in the Dodewaard nuclear power plant

W.M.M. Huijbregts & P.J .C. Letschert

Kema Scientific & Technical Reports 4 (2): 15-25. ISSN 0167-8590; ISBN 90-353-0037-8. (paper 33)

JAF conference Tokio 1987

pdf of full paper available

Abstract

A rack composed of different materials and surface treatments was exposed in the reactor vessel of the GKN boiling-water reactor in Dodewaard for one reactor cycle in order to study the CRUD (Corrosion Residual Unidentified Deposit) deposition. Ferritic, austenitic and ferritic-austenitic steels and lnconel 600 were mounted in the rack. The Co-60 and Mn-54 activity of the CRUD and of the adherent oxide layer were measured.

The more CRUD was deposited on the samples, the higher the Co-60 and Mn-54 activi- ties of the adherent oxide layer appeared to be. The Co-60 activity of the adherent oxide layer increased greatly with higher CRUD deposition for the high-chromium austenitic steels. The effect of CRUD deposition on Co-60 activity of the adherent oxide layer was slight for the ferritic low-chromium steels and for lnconel 600. The influence of CRUD deposition on the Mn-54 activity of the adherent oxide layer was greater for the ferritic steels than it was for the austenitic steels. The surface pretreatments of AISI 304, annealing in air at 623 K and autoclaving at 583 K, resulted in higher activity levels.

Figure 3 Samples mounted in the rack, before exposure. Flow direction from left to right.

Results

The activity levels and the morphology of the samples after CRUD deposition were the object of special attention during the work reported here.

Activity

The activity of the samples with the CRUD was measured with a 10% high-purity Ge-detector and a 4000-channel multichannel analyzer ( detector Canberra model 7229P, analyzer Canberra S40). The distance of the samples from the detector was 75 cm. Co-60 and Mn-54 activity was calculated from the gamma spectra per unit area of sample surface. The statistical counting error (3a) was < 1% .After measurement of the activity, the samples were cleaned ultrasonically in water with a wetting agent added for 45 minutes (US cleaner Sonogen model L TH 80/4R, average power 150 W, peak power 300 Wand 24.4 kHz). Subsequently the samples were flushed and measured again.

The results of the activity measurements are given in Tables 4 and 5 (not included in this site). The values (for AISI 304 and 316 the average values) were plotted as bar diagrams (Figs. 5 and 6).

Fig.5

The activities measured on the inside position samples, all glass blasted.

Fig.6

The activities measured on the inside position samples AISI 304 and AISI 316.

There is a clear difference in the Co-60/Mn-54 ratios of the CRUD and of the adherent oxide. The Co-60 is preferably incorporated in the adherent oxide layer. The ferritic steels show lower activities per unit of surface of the adherent oxide than do the austenitic and ferritic-austenitic steels. There is some indication that the autoclaved and air-annealed sam- ples of AISI 304 have higher activities per unit of surface than the steels subjected to the other surface pretreatments. This effect was not found for AISI 316.

Morphology

The diameter of the CRUD particles ranges from 0.5 to 2.0 micron. The loose oxide removed from the unalloyed and low-alloyed steels contained larger pieces of oxide, i.e. corrosion products of the samples. The texture of the CRUD on the various samples did not differ. Several types of crystals could be distinguished in the CRUD (Fig. 7):

1. small round particles (0.5 to 2 !1m);
2. large round spheres (up to 30 !1m), covered with small oxide particles;
3. needle- or plate-like oxide occasionally.

Conclusions

Incorporation of Co-60 and Mn-54 in the adherent oxide layers appears to depend very much on the amount of CRUD deposited. This effect of the CRUD layers cannot yet be explained in detail. The local chemical conditions of the water under CRUD presumably change more a5 the layer becomes thicker.

Incorporation of Co-60 occurs more in the adhe- rent oxide of chromium-containing austenitics than in the unalloyed and low-alloyed steels. Much Mn-54 is incorporated in the adherent oxide layers of both of the latter steel types, which is in contrast to the .austenitic steels.

Pre-oxidizing of the austenitic steel AISI 304 gave a higher activity than the pickled and passivated sam- pIes. This effect is in contrast with other results and could be explained by differences in test conditions.

Figure 9 Co-6Q activity of the adherent oxide, plotted against the Co-6Q activity of theCRUD plus that of the adherent oxide.	Figure 10 Mn-54 activity of the adherent oxide, plotted against the Mn activity of the CRUD plus that of the adherent oxide.
Figure 11 Co-60 activity of the adherent oxide, plotted against the calculated amount of CRUD.	Figure 12 Mn-54 activity of the adherent oxide, plotted against the calculated amount of CRUD.