2.1 Crud scraping

In this study, a single stroke of a scraper with an area of about 2 cm² on the fuel rods was used to define the sampling area. Crud was removed using a stream of water, which was then passed through a vacuum filter (0.45 μm pore size) for sample collection. Crud samples were scraped at 5 mid-span axial locations (spans 1-5) on two rods located on different sides (A5 and L5) of each assembly, while held underwater in the spent fuel storage pool. The spacer numbering and span elevations of the ATRIUM-10 fuel assembly for the sampling locations are presented in Fig. 1. The locations on the rods from which crud samples were obtained are presented in Fig. 2. All samples were loaded into casks for shipment to laboratories at the Institute of Nuclear Energy Research (INER) for further analysis.

Fig. 1.Full-screen View

The dimensions and span elevations of the fuel assembly.

Fig. 2.Full-screen View

Location of crud sampling on used rods.

2.2 Sample pretreatment

X-ray diffraction (XRD) and scanning electron microscopy (SEM) measurements were performed on small samples cut from the filter paper. Chemical and radiochemical analysis of the crud using inductively coupled plasma-atomic emission spectroscopy (ICP-AES) and gamma spectrometry required pretreatment using wet oxidation and the total dissolution of the filtered samples. The samples were completely dissolved in a solution of concentrated HNO₃ (69%) and HCl (37%) (1:1). Following the addition of several drops of H₂O₂ (30%) to destroy any remaining organic substances, the samples were evaporated to near-dryness in Teflon containers on a hot plate. Finally, the samples were diluted in DIW (100 ml) in a volumetric flask for quantitative analysis.

2.3 SEM/EDS

The topography of the samples was analyzed using a scanning electron microscope with an integrated energy dispersive X-ray analysis system (JEOL, JSM-6510) at high resolution (3.0 nm at 30 kV). After drying under an infrared lamp (250 W) for approximately 1 h, the voltage was set to 20 kV and images taken at an appropriate magnification ratio were selected.

2.4 XRD

The analysis of crud samples was performed using a Bruker D8 Advance diffractometer operated at 40 kV and 40 mA in a ceramic Cu X-ray (1.5418 Å) tube. Scanning was performed in the range of 20–60 (2θ) at a speed of 1 °/min. A LynxEYE compound silicon strip detector with lower background counts (< 0.1 cps) was employed. Diffraction spectra of the sample and references were analyzed using TOPAS Spectra software to calculate the crystal content of the samples.

2.5 ICP-AES

The following major elements were detected using ICP-AES analysis (Horiba JY Ultima 2): Fe, Co, Cu, Ni, Cr, Zn, and Mn.

2.6 Gamma spectrometer

Activity concentrations in the crud samples were obtained using a high purity Ge-detector (HPGe; Canberra GC 4020) with a 1024 multi-channel analyzer (Canberra DSP 9660). The activity concentrations of ⁵⁴Mn, ⁶⁰Co, ⁵⁸Co, and ⁵⁹Fe were calculated according to the gamma spectra using Canberra Genie 2000 software.

3.1 Elemental composition and activity concentration of deposits

During the operation of the plant, impurities accumulate in the reactor water due to the ingress of soluble and insoluble metallic species through the feed water system. Table 2 lists the average concentrations of metallic species (as determined by ICP-AES) in the reactor water and feed water. The most common impurity is iron and the minor impurities include copper, zinc, and nickel, which accumulate in the form of crud on the surfaces of fuel cladding and can be viewed as a product of the boiling that occurs in the core. Table 3 lists the average metallic compositions and activity concentrations of corrosion products accumulated on the surface of fuel rods. More than 97.0 wt.% of the content is made up of iron, with other elements accounting for less than 3.0 wt.%. The Fe can be traced to carbon steel, stainless steel, and other iron-based alloys. After Fe, the most common elements in the crud are Ni and Cu, which can be attributed to the use of a copper-nickel condenser tube at this plant. Ni and Cr can also be traced back to the stainless steel and Inconel, whereas the Cu mostly leeches from copper-based alloys. Mn is common in stainless steel and carbon steel. The Co content in the crud is very low; however, it is an important source of the activated product ⁶⁰Co. The average element compositions during the refueling outages of EOC 15 (fuel cycle 3) and EOC 16 (fuel cycle 4) under NWC conditions are consistent with the results obtained after the implementation of HWC [13].

The main nuclides included the following activated corrosion products: ⁵⁴Mn, ⁵⁹Fe, ⁶⁰Co, and ⁵⁸Co. The axial concentration distributions of Fe with ⁵⁴Mn and ⁵⁹Fe activities, Ni with ⁵⁸Co activities, and Co with ^6oCo activities in the crud deposits are presented in Figs. 3, 4 and 5. Similar trends can be observed in the two groups of analytical results for the EOC 21 (K1D015 and K1D059) and EOC 22 (K1D085 and K1D131) crud samples. The activity of radionuclides in the adherent oxide layer is proportional to that of the crud deposited on the fuel rods. The distribution patterns of ⁵⁴Mn and ⁵⁹Fe in Figs. 3(b) and 3(c) are similar to those in Fig. 3(a). This is not unexpected, due to the fact that ⁵⁴Fe and ⁵⁸Fe are the mother isotopes of ⁵⁴Mn and ⁵⁹Fe via the activation reactions of [⁵⁴Fe(n, p)⁵⁴Mn] and [⁵⁸Fe(n, γ)⁵⁹Fe], respectively. Furthermore, ⁵⁴Mn and ⁵⁹Fe exist for only a short period following the shutdown of the reactor due to their short half-lives. As shown in Figs. 4(a) and 4(b), the axial distribution of ⁵⁸Co was similar to that of Ni. ⁵⁸Co is the daughter isotope of ⁵⁸Ni (natural isotope; 68.08% abundance), formed through the activation [⁵⁸Ni(n, p)⁵⁸Co] of nickel from stainless steel and nickel-based alloys. Similarly, the distribution pattern of ⁶⁰Co, as obtained by neutron irradiation [⁵⁹Co(n, γ)⁶⁰Co], is similar to that observed in Co, as shown in Figs. 5(a) and 5(b). A greater amount of crud was deposited on spans 1–2 (at a height of 50∼100 cm from the bottom) than on spans 3–5 (at a height of 00∼ 250 cm from the bottom) of the rods. Spans 1–2 (where the boiling starts) were fully immersed in the water phase and thus had a higher probability of deposition, whereas spans 3–5 are in the upper part of the core close to the steam phase [13], where deposition is less likely to occur.

Fig. 3.Full-screen View

(Color online) Axial profiles of A5 crud samples collected in 2010 and 2012: (a) Fe; (b) ⁵⁴Mn; and (c) ⁵⁹Fe.

Fig. 4.Full-screen View

(Color online) Axial profiles of A5 crud samples collected in 2010 and 2012: (a) Ni; and (b) ⁵⁸Co.

Fig. 5.Full-screen View

(Color online) Axial profiles of A5 crud samples collected in 2010 and 2012: (a) Co; and (b) ⁶⁰Co.

The axial profiles of the massic and activity concentrations of EOC 15-KBE 301 (fuel cycle 3) and KBE 302-EOC 16 (fuel cycle 4) under NWC conditions are consistent with the results obtained following the implementation of HWC [13]. We also analyzed a large number of L5 fuel deposit samples. The results of the concentration and activity distributions along the length of the fuel rods present trends similar to those obtained from the A5 crud samples [12, 15]. Lin [14] claimed that the amount of fuel surface deposit and its chemical composition depend on the corrosion product input from the condensate and feed water systems.

3.2 Interpretation of ⁵⁸Co/⁶⁰Co ratios in the crud samples

Figure 6 presents the activity ratios of ⁵⁸Co/^6oCo in the crud samples. K1D015 and K1D059 were collected at EOC 21 (3^rd burn up) in 2010; K1D085 and 131 were obtained at EOC 22 (4^th burn up) in 2012. Jung et al. [16] suggested that similarities in the values of radioactivity ratios of ⁵⁸Co to ^6oCo were an indication of fuels with a similar burn up history in the reactor core. The average radioactivity ratios of ⁵⁸Co/⁶⁰Co in the fuel assemblies of K1D085 and K1D131 are considerably lower than those of K1D015 and K1D059. This can be partly explained by the fact that the former remained in the spent fuel pool for a longer duration than did the latter. It can also be attributed to the considerable difference in the half-lives of the two radionuclides (i.e., 70.8 days for ⁵⁸Co and 5.27 years for ⁶⁰Co). Thus, the radioactivity ratio could be used to determine how long the crud was present in the reactor core.

Fig. 6.Full-screen View

(Color online) Radioactivity ratios of ⁵⁸Co/⁶⁰Co in crud samples.

3.3 SEM and XRD analysis

Figures 7(a)-7(d) present SEM images showing the morphology of crud deposits on the fuel rods. Most of the corrosion products display clusters of irregularly-shaped smaller (0.1 to 0.3 μm) or larger (0.5 to 2.0 μm) particles in the submicron range. A small number of the samples contained visible ‘‘flakes’’ within the crud deposits. These flakes were defined as fragments of crud deposit with the largest dimension greater than or equal to 5 μm. Orlov et al. identified several types of cubic clusters as spinel crystal structures [17]. The constituent materials may include trace quantities of franklinite (ZnFe₂O₄), trevorite (NiFe₂O₄), and jacobsite (MnFe₂O₄). EDS analysis identified cubic spinel-type crystals of MnFe₂O₄ with a stoichiometric Mn/Fe ratio of 1: 2.

Fig. 7.Full-screen View

SEM photos of crud samples: (a) KID015 L5 span 5; (b) KID059 L5 span 2; (c) K1D085 A5 span 5; (d) K1D131 A5 span 4.

Crud samples were scanned using XRD and analyzed with TOPAS to characterize the crystal content. As shown in Fig. 8, reddish α-Fe₂O₃ (hematite) was the primary form of iron in the loosely-adhered outer layer of the BWR fuel rods. XRD analysis of the crud deposits gave no indication that they contain Fe₃O₄, which could theoretically change into Fe₃O₄ if hydrogen were used as a reductant. Nevertheless, this reaction would require very high temperatures (∼600 ℃) [18]. These results corroborate the supposition that the transition of α-Fe₂O₃ to Fe₃O₄was kinetically inhibited at temperatures of only 288 ℃. Under HWC conditions, the results obtained in the fourth fuel cycle were the same as those obtained in the third [15]. In the report by Liang and Chien, hematite was the major constituent deposited on the fuel rods in an NWC environment, accounting for more than 95 wt.% of the total [13]. The crud samples in the report by Judd et al., included almost 100% α-Fe₂O₃ with no detectable Fe₃O₄ type under HWC. They reported no adverse transitional effects associated with bundles that were exposed to NWC, as well as HWC [19]. Beverskog and Puigdomenech stated that the nature of the crud depended on the Ni/Fe ratios at 285 ℃; i.e., both NiFe₂O₄ and α-Fe₂O₃ can form when Ni/Fe is < 0.5 [20]. In this study, the Ni/Fe ratio observed in the four fuel assemblies was in the range of only 0.005. These negligible quantities of Ni resulted in the formation of α-Fe₂O₃ as the preferred phase.

Fig. 8.Full-screen View

X-ray diffraction patterns of crud on fuel rods of four fuel assemblies.