A new approach to determine abnormality of radioactive discharges from pressurized water reactors and to derive abnormality indicators correlated with a specific causal event

3.1.1 Calculation and distribution characteristics of RCF for US PWRs

In Section 2.2, RCF values were derived for US PWRs with 1,300 RYs mentioned. Using Equations (1) and (2), the values of s⁻¹ and f were calculated for radionuclides (groups) for every RY, and the results were plotted in ascending order in Figure 2. In Figure 2, the y-axis represents RCF values, where (a) indicates the range where the value of s⁻¹ is less than 1, and (b) indicates the range where the value of f is greater than 1, which is the significant range.

Figure 2:

Distribution curves of RCF by radionuclides (groups) for 1,300 RYs in US PWRs. (A) and (B) represent the values of s⁻¹ and f, respectively. The legend of the graph: G, I, P, T, and C represent gaseous radionuclides (groups) denoting fission products, radioiodine, particulates, gaseous tritium, and gaseous ¹⁴C, respectively. F, D, and H represent liquid radionuclides (groups) denoting fission and activation products, dissolved gases, and liquid tritium, respectively.

In Figure 2, the reason for the relatively low number of data points for ¹⁴C compared to other radionuclides (groups) is that Regulatory Guide 1.21 was revised in 2009, adding the record item of ¹⁴C to the ARERR in US, leading to the recording of ¹⁴C discharge data from US NPPs starting in 2010 (USNRC 2009).

Fundamentally, Figure 2 represents a curve of the ‘upward and downward parabola’ type, where the narrower the gap between the exponential growth and logarithmic growth parts and the steeper the slope, the greater the variability of increases or decreases in annual activities during the respective periods. From Figure 2A, it can be inferred that radionuclides (groups) with relatively large variations in annual activities are radioiodine (I), particulates (P), and dissolved gases (D) in the range where s⁻¹ values are less than 1, and from Figure 2B, it can be inferred that radionuclides (groups) with f values greater than 1 are I, P, D, and fission and activation products (F).

To assess the relative distribution of annual activities by radionuclides (groups), the coefficient of variation (CV), which is the standard deviation divided by the mean, was calculated (Nairy and Rao 2003). Table 2 shows the statistical values (mean and CV) of annual activities by radionuclides (groups) and the ratio of RCF values distributed within the range of 0.1–10 in Figure 2.

In this study, the ratio of RCF values distributed between 0.1 and 10 was used to differentiate between relatively large or small differences in annual activities during normal and abnormal releases due to events. As indicated in Table 2, ¹⁴C (C) and liquid and gaseous tritium (H and T) have the highest proportion, with both s⁻¹ and f values falling within the range of 0.1–10, exceeding 95 %. Therefore, radionuclides (groups) with a high proportion of RCF values distributed between 0.1 and 10 are considered that their abnormal releases contribute less to the annual activities. The analysis results indicating that I, P, and D show relatively large variations in annual activities in Figure 2 align with Table 2, where the proportion of RCF values for these radionuclides (groups) falling between 0.1 and 10 is relatively low.

The CV values for the annual activities of C, T, and H are 0.74, 1.18, and 2.08, respectively, indicating very small variations in annual activities. This aligns with the fact that both s⁻¹ and f values fall within the range of 0.1–10, which is the highest proportion among the three radionuclides (groups). These results suggest that the relative magnitude of variations in annual activities between normal and abnormal releases can be reasonably represented by the distribution of s⁻¹ and f values.

3.1.2 Estimation ratio of abnormal releases according to RCF

As stated in Equation (5), s⁻¹ less than 1 or f greater than 1 implies that the discharges for the current year compared to the previous year or average from the reference year (i.e., 2000) up to the previous year has increased. In this study, it is assumed that as the values of s⁻¹ decrease or f increase, the estimation ratio of abnormal releases, as presented in Equation (3), increases. In this case, it is expected that there would be a high correlation with abnormal releases.

Figure 3 depicts the estimation ratio of abnormal releases, ER_{RCF,i,00–20}, by radionuclides (groups) according to Equation (3) for the dataset of discharges from 1,300 RYs in the US from 2000 to 2020. Figures 3A and B represent the estimation ratio of abnormal releases where the values of s⁻¹ is less than 1 and f is greater than 1 according to Equation (3), respectively. The horizontal dashed line in Figure 3 represents the Freq_ab,00–20 mentioned in Equation (4) for the entire dataset of discharges from 1,300 RYs in the US from 2000 to 2020.

Figure 3:

Estimation ratio of abnormal releases by radionuclides (groups), ER_{RCF,i,00–20}, based on RCF values for 1,300 RYs in US PWRs. (A) and (B) represent the estimation ratio of abnormal releases according to the values of s⁻¹ and f, respectively.

In Figure 3B, there are a total of 29 RYs across all radionuclides (groups) except for the I and D where the y-axis ER_{RCF,i,00–20} is 0 %. This is expected when the average activities up to the year before abnormal releases are significantly low compared to the years when abnormal releases occur. As an example, at the Beaver Valley NPP, during maintenance activity from 2000 to 2002, discharges of H were observed to be approximately 0.1 times lower than in other years (FENOC 2002). To ensure meaningful result analysis in Sections 3.1.2 and 3.1.3, f values where ER_{RCF,i,00–20} is 0 % for each radionuclide (group) were excluded.

In Figure 3, for fission products (G), the estimation ratio ER_{RCF,G,00–20} is consistently lower than the Freq_ab,00–20 for all RCF values. Similarly, in Figure 3A, for I, the estimation ratio ER_{RCF,I,00–20} is consistently lower than the Freq_ab,00–20 for all RCF values as well. In Figure 3A, for T, C, H, and F, as the value of s⁻¹ decreases, the estimation ratios ER s − 1 , T , 00 − 20 , ER s − 1 , C , 00 − 20 , ER s − 1 , H , 00 − 20 , and ER s − 1 , F , 00 − 20 respectively increase, with most s⁻¹ values having higher values than the Freq_ab,00–20. In Figure 3B, for P and D, as f increases, the estimation ratios ER_f,P,00–20 and ER_f,D,00–20 respectively increase, with most f values having higher values than the Freq_ab,00–20. Based on the above results, G and I were excluded from the analysis, and in the s⁻¹ value, T, C, and H were analyzed, while in the f value, P and D were analyzed.

To assess the correlation between the increases of the activities of radionuclides (groups) discharged due to abnormal releases by radionuclides (groups) and the causal events, linear regression analysis based on Equation (5) was conducted on the dataset of discharges from 1,300 RYs in the US as depicted in Figure 3. Subsequently, the satisfaction status and values of the criteria for identifying sensitive radionuclides (groups) with high correlation between the increases of the activities of radionuclides (groups) discharged and causal events, as criteria presented in Table 1, were summarized in Table 3.

In Table 3, T, C, and H satisfy all three criteria, indicating the highest correlation between the increases of the activities of radionuclides (groups) discharged and causal events in terms of s⁻¹ values, while D satisfies all three criteria in terms of f values. Although not all criteria are satisfied, radionuclides (groups) that meet C-① and either C-② or C-③ are F and D in terms of s⁻¹ values and P in terms of f values. Radionuclides (groups) that meet only C-① are I in terms of f values.

As mentioned in Section 2.3, T and C are confirmed to be generated whenever major events (primary system leakage, secondary system leakage, leakage of RWMS, leakage from defected SF during handling). They can be classified as radionuclides (groups) with high estimation ratios of abnormal releases according to the results of Figure 3A. Moreover, based on the results of Table 3, it can be anticipated that T and C may experience an increase in activities due to abnormal releases when events occur, indicating a high correlation with events.

In Section 3.1.1, it was confirmed that the distribution of RCF in T, C, and H is predominantly between 0.1 and 10, with a proportion exceeding 95 %, and the CV values are low, ranging from 0 to 3. Based on these results, it was expected that the estimation ratio of abnormal releases would be low due to the small magnitude of the increases of activities of radionuclides (groups) discharged during abnormal releases. However, the results from Figure 3 and Table 3 indicate a very high correlation between the increases of activities of radionuclides (groups) discharged and abnormal releases, as well as a high correlation with causal events, for the same radionuclides (groups).

3.1.3 Derivation and verification of reference values of RCF

The RCF derived from a consistent set of discharge data can be utilized to estimate the correlation with abnormal releases for succeeding discharge datasets. In this study, to verify the applicability of the proposed RCF, discharge data from US PWRs from 2000 to 2020 were divided into two groups: Group 1 discharge data (i.e., preceding discharge data) from 2000 to 2010 and Group 2 discharge data (i.e., succeeding discharge data) from 2011 to 2020.

Group 1 discharge data consists of a total of 681 RYs, with 163 RYs occurring abnormal releases. Group 2 discharge data comprises 619 RYs, with 156 RYs occurring abnormal releases. Therefore, the Freq_ab,11–20 calculated using Equation (4) is 25.2 % (=156 RYs/619 RYs). Figure 4 represents the estimation ratio of abnormal releases ER_{RCF,i,00–10} by radionuclides (groups) based on the RCF values from the Group 1 discharge dataset of 681 RYs from 2000 to 2010 in the US, following the same methodology as Figure 3. The horizontal dashed line in Figure 4 represents the Freq_ab,11–20 calculated for the entire Group 2 discharge dataset of 619 RYs from 2011 to 2020.

Figure 4:

Estimation ratio of abnormal release by radionuclides (groups) based on RCF values for Group 1 discharge dataset of 681 RYs from 2000 to 2010 in US PWRs. (A) and (B) represent the estimation ratio of abnormal releases based on s⁻¹ value and f value, respectively.

In Figure 4A, it is observed that for G, I, and P, the estimation ratios ER s − 1   , G , 00 − 10 , ER s − 1   , I , 00 − 10 , and ER s − 1 , P , 00 − 10 are lower than Freq_ab,11–20 for all s⁻¹ values. Similarly, in Figure 4B, for G and T, the estimation ratios ER_f,G,00–10 and ER_f,T,00–10 are lower than Freq_ab,11–20 for all f values.

In Figure 4A, it is observed that for the T and H, as the value of s⁻¹ decreases below 1, the estimation ratios ER s − 1   , T , 00 − 10 and ER s − 1 , H , 00 − 10 increase. For most s⁻¹ values, these ratios are higher than Freq_ab,11–20. Similarly, in Figure 4B, it is observed that for the P and D, as f increase above 1, the estimation ratios ER_f,P,00–10 and ER_f,D,00–10 increase. For most f values, these ratios are higher than Freq_ab,11–20.

The C in Figure 4A, as mentioned in Section 3.1.1, has discharge records starting from 2010, thus limiting its usability for data analysis (USNRC 2009). In Figure 4A, it was confirmed that for s⁻¹ values less than or equal to 0.1, the estimation ratio of abnormal releases ER s − 1 , C , 00 − 10 is 0 %, indicating that there are no NPPs corresponding to RY ( s − 1   <   RCF ) , a b , i , k in Equation (3).

To ascertain the correlation between the increase of activities of radionuclides (groups) discharged due to abnormal releases in Figure 4 and the causal events, a linear regression analysis was conducted on the RCF values of Group 1 discharge data according to Equation (5). Subsequently, the satisfaction status and values of the criteria for identifying sensitive radionuclides (groups) with high correlation between the increases of activities of radionuclides (groups) discharged and events were summarized in Table 4.

In Table 4, the radionuclides (groups) that satisfy all criteria are T for the s⁻¹ value and D for the f value. The radionuclides (groups) that satisfy C-① while also satisfying either C-② or C-③ are H for the s⁻¹ value and P for the f value. The radionuclides (groups) that only satisfy C-① are D and F for the s⁻¹ value. For all radionuclides (groups) that satisfy C-①, the reference values of RCF were determined. The radionuclides (groups) that meet these criteria are T, F, and H for the s⁻¹ value, and P and D for the f value.

Based on the results in Table 4, T, H, F for the s⁻¹ values, and P, D for the f values were prioritized as sensitive radionuclides (groups). The reference values for RCF selected from the Group 1 discharge data were determined based on the condition that ER_{RCF,i,00–10} is greater than Freq_ab,11–20 (25.2 %) but not exceeding 100 %, and were chosen as the mean or median values of s⁻¹ and f values. The reference values for RCF can be chosen by the user, and apart from the mean or median values, RCF values corresponding to estimation ratios of abnormal release ER_RCF,i,k can also be selected according to the user’s needs. The values of ER_{RCF,i,11–20} when applying the reference values to Group 2 discharge data from 2011 to 2020 have been calculated and summarized in Table 5.

In Table 5, excluding D, it was observed that when the reference values were determined based on the mean and median values, they were respectively 1.1 to 1.2 times and 1.2 to 1.4 times higher than Freq_ab,11–20, which is 25.2 %. For D, the estimation ratio of abnormal release was lower at 11.76 % when based on the median value and 25 % when based on the mean value. This discrepancy is likely due to the intermittent discharge of D, accounting for 121 RYs out of 691 RYs (19.55 %) during the period, suggesting insufficient discharge data for reliable analysis.

In Section 2.1, among the 705 events of multiple abnormal releases reported in US PWRs within a year, detailed information on the specific annual activities of radionuclides (groups) discharged attributed to actual abnormal releases was provided for 259 events. When examining the ratio of the activities of radionuclides (groups) discharged during abnormal releases to the total annual activities of radionuclides (groups) discharged, it was found that the majority (accounting for 85 % of the total) had ratios of less than 20 %. The average ratio was 11.82 %, with a standard deviation of 28.06 %. This indicates that abnormal releases have a relatively minor impact on the total annual activities of radionuclides (groups) discharged, and there are many cases where a specific event of abnormal release does not dominantly influence the total annual activities of radionuclides (groups) discharged. Therefore, it may be challenging to establish a clear correlation between specific events and abnormal releases. However, by introducing the concept of RCF in this study, it was confirmed that abnormal release could be estimated at values exceeding approximately 1.1–1.5 times (26.79–35.71 %) the Freq_ab,00–20 value of 24.54 % for 1,300 RYs in US PWRs.

3.2.1 Leakage from the gas decay tank

Out of the 1,300 RYs in the US, abnormal release due to leakage from GDT amounted to 97 RYs, resulting in an Freq_{ab,GDT,00–20} of 7.46 % according to Equation (4). As mentioned in Section 2.3, abnormal release from leakage from GDT can result in the release of fission gases, including ⁸⁵Kr, ^85mKr, ⁸⁷Kr, ⁸⁸Kr, ¹²⁷Xe,^131mXe, ¹³³Xe, ^133mXe, ¹³⁵Xe, ^135mXe, ¹³⁷Xe and ¹³⁸Xe. Figure 5 illustrates the RCF values when abnormal release occurs due to leakage from GDT. As mentioned in Section 2.2, if the s⁻¹ value is less than 1 and the f value is greater than 1, it can be expected that the activities of radionuclides (groups) discharged will increase during abnormal release. While the average s⁻¹ value in Figure 5A is not consistently less than 1, radionuclides with s⁻¹ values less than 1 in over 50 % of cases include ¹³⁸Xe (85 %), ^133mXe (59 %), ⁸⁸Kr (52 %), ¹³⁵Xe (51 %), ¹²⁷Xe (50 %), and ^135mXe (50 %). The average value of f in Figure 5B is greater than 1 for all radionuclides, and the radionuclide with a ratio of over 50 % of cases where the actual f value exceeds 1 is ¹⁴C (73 %).

Figure 5:

The RCF values of fission gases discharged abnormally due to leakage from the GDT in US PWRs. (A) and (B) represent the values of s⁻¹ and f by radionuclides, respectively.

When applying the criteria for determining the correlation between the increase of activities of radionuclides (groups) discharged and causal events as described in Equation (5) and Section 2.3, the radionuclides that satisfy all criteria (C-①, C-② and C-③) are ⁸⁸Kr, ¹³⁷Xe and ¹³⁸Xe in the f values. The radionuclides satisfying C-① while also meeting either C-② or C-③ are ⁸⁵Kr and ⁸⁸Kr in the s⁻¹ values, and ⁸⁵Kr, ⁸⁷Kr, and ^133mXe in the f values. Therefore, when considering the RCF values and the satisfaction of judgment criteria, radionuclides with a high correlation with the increase of the activities of radionuclides discharged and events are likely to be ⁸⁸Kr, ¹³⁷Xe and ¹³⁸Xe in the f values.

3.2.2 Leakage from the steam generator’s power operated relief valve

Out of the 1,300 RYs in the US, abnormal release due to leakage from SG’s PORV amounted to 39 RYs, resulting in an Freq_{ab,SG,00–20} of 3 % as defined in Equation (4). This event could result in the leakage of gaseous tritium. While the average value of s⁻¹ is not consistently less than 1, the proportion of s⁻¹ values less than 1 is 48.6 % when considering all s⁻¹ values. Although the average value of f is greater than 1, the proportion of f values greater than 1 is 44.4 %, indicating that not all activities of radionuclides discharged increase during abnormal release causing events. While T does not satisfy all three criteria for determining the correlation between the increase of activities of radionuclides (groups) discharged and causal events as described in Equation (5) and Section 2.3, it does satisfy C-① and C-② for f values. Since C-①, which is essential in s⁻¹ values, is not satisfied, it can be expected that the correlation between the increase of the activities of radionuclide discharged and causal events is higher in f values.

3.2.3 Leakage from defected spent fuel during handling

Out of the 1,300 RYs in the US, leakage from defected SF during handling amounted to 4 RYs, resulting in an Freq_{ab,Fuel,00–20} of 0.31 % as defined in Equation (4). As mentioned in Section 2.3, leakage from defected SF during handling results in the generation of T and C (ORNL 1977, 1980; USNRC 2020). It has been confirmed that fission gases emitted due to leakage from defected SF during handling include radionuclides such as ⁸⁵Kr, ^85mKr, ⁸⁷Kr, ⁸⁸Kr, ^131mXe, ¹³³Xe, ^133mXe, and ¹³⁵Xe.

Figure 6 depicts the RCF values when there is abnormal release due to defected SF during handling. In Figure 6A, the average value of s⁻¹ is less than 1 for the ¹⁴C. The radionuclides ¹²⁷Xe and ³H have proportions where 50 % are less than 1 of the s⁻¹ value. The T and C, identified as radionuclides with a high estimation ratio of abnormal release in s⁻¹ values in Section 3.1.2, correspond to radionuclides with high correlation between the increase of the activities of radionuclides and events. In Figure 6B, the f values for ⁸⁵Kr, ⁸⁸Kr, ^131mXe, ¹³³Xe, ^133mXe and ¹⁴C have an average value greater than 1, and the ⁸⁸Kr corresponds to cases where all f values are greater than 1, with ¹⁴C having a proportion of 75 % greater than 1. The radionuclide commonly associated with an increase in the activities when abnormal release occurs due to defected SF during handling is consistently ¹⁴C.

Figure 6:

The RCF values of fission gases, ³H, and ¹⁴C discharged abnormally due to leakage from defected spent fuel during handling in US PWRs. (A) and (B) represent the values of s⁻¹ and f by radionuclides, respectively.

Applying the criteria for determining correlation between the increase of the activities of radionuclides discharged and event in Equation (4) and Section 2.3, radionuclides that satisfy all three criteria are ^133mXe for s⁻¹ values, and ⁸⁸Kr and ¹⁴C for f values. Isotopes that satisfy criterion ① while also meeting criterion ② are ⁸⁵Kr, ^131mXe, ¹³³Xe, ^133mXe and ³H for f values. Among these radionuclides, ⁸⁵Kr, ¹³³Xe,^133mXe and ³H have coefficient of determination, R², all greater than 0.9, indicating a very high correlation between the increase of the activities or radionuclides discharge and leakage from defected SF during handling. Indeed, abnormal releases of ⁸⁵Kr and ¹³³Xe were confirmed in cases of leakage from defected SF during handling at Calvert Cliffs Units 1 and 2, Catawba Unit 1, and Millstone Unit 2 (Dominion Energy 2015; Duke Energy 2015; Exelon 2016).

3.3.1 Derivation and analysis of RCF based on discharge data

RCF values for 310 RYs from 2000 to 2020 were derived for Korean PWRs, and their distribution characteristics were analyzed. Using Equations (1) and (2), s⁻¹ and f values were calculated for radionuclides (groups) of each PWR for every RY, and the results were plotted in ascending order in Figure 7. This is comparable to the RCFs of US PWRs shown in Figure 2. On the y-axis of Figure 7, A represents the range where the value of s⁻¹ is less than 1, and B indicates the range where f values are greater than 1, which is considered a significant range.

Figure 7:

Distribution curves of RCF by radionuclides (groups) for 310 RYs in Korean PWRs. (A) and (B) represent the values of s⁻¹ and f, repectively.

In Figure 7A and B, the radionuclides (groups) with steep slopes and narrow gaps between the exponential growth and logarithmic growth parts in the ‘upward and downward parabola’ type curve are ranked in descending order as D, I, P, F, G, C, T, and H. Especially noteworthy are the top three ranked radionuclide groups, D, I, and P, which demonstrate results similar to those observed in Figure 2 for US PWRs. A steep slope indicates significant fluctuations in the annual activities of the corresponding radionuclides (groups). Reviewing the discharge data for Korean PWRs, it can be inferred that D, I, and P show relatively low frequencies of exceeding the Minimum Detectable Activity (MDA), with discharge frequencies of approximately 18 RYs (5.8 %), 92 RYs (30 %), and 111 RYs (36 %) out of 310 RYs, respectively. Consequently, it is anticipated that these radionuclide groups may exhibit larger fluctuations in the activities of radionuclides (groups) discharged due to their infrequent occurrences of exceeding the MDA.

Table 6 represents the distribution ratios of RCF values and statistical values of annual radioactive effluent release for 310 RYs from 2000 to 2020 for Korean PWRs. Comparing the statistical values of annual activities of radionuclides (groups) between Table 2 for US PWRs and Table 6 for Korean PWRs, it can be observed that Korean PWRs exhibit lower annual average activities, ranging from 6.14 × 10⁻⁵ to 0.82 times that of US PWRs. Furthermore, for radionuclides (groups) excluding C and D, the CV ranges from 0.23 to 0.67 times, indicating lower variability in annual activities compared to US PWRs. It is worth noting that discharge data for C has been recorded since 2012 in Korea, resulting in limited available data. Additionally, D accounts for only 5.8 % of total discharge data among the 310 RYs, indicating a limited frequency of occurrence and suggesting limitations in comparison with statistical values from US PWRs.

Meanwhile, the proportion of radionuclides (groups) distributing within the range of 0.1–10 for s⁻¹ values is highest in the order of H, T, C, G, F, P, I, and D, and this order remains the same for f values as well. In terms of CV, the order from lowest to highest is H, C, T, G, F, I, D, P, indicating a similarity to the order of proportions distributing within the range of 0.1–10. Similarly, in Table 2 for US PWRs, the top three radionuclides (groups) with the highest proportion of distribution within the range of 0.1–10 are H, C, and T, parallel those of Korean PWRs, and the order of lowest CV is also similar to that of US PWRs.

In Figure 7, it can be observed that the distribution within the range of 0.1–10 is denser in Korean PWRs compared to that in US PWRs shown in Figure 2. Table 7 compares the statistical values of RCF corresponding to US PWRs in Figure 2 and Korean PWRs in Figure 7.

In Table 7, The σ value of RCF is smaller in all radionuclides (groups) for Korean PWRs compared to those for US PWRs in terms of s⁻¹ values, and it is also smaller for all radionuclides (groups) except for T and C in terms of f values. When comparing the σ value with that of US PWRs, it can be observed that the annual activities of liquid and gaseous radionuclides (groups) in Korean PWRs exhibit less variability. Consequently, it is expected that Korean PWRs may exhibit less pronounced increases in activities of radionuclides (groups) during abnormal release causing events compared to US PWRs.

3.3.2 Estimation of correlation for specific event

In Korea, unlike the US, abnormal releases are not categorized and reported separately in annual radioactive effluent reports. Instead, the Korea Institute of Nuclear Safety (KINS) records accidents and failures in NPPs in OPIS (KINS 2024). However, identifying accidents or failures that may lead to abnormal releases individually is necessary. Nonetheless, cases that could be classified into the four groups of abnormal releases causes in Section 2.3 were not identifiable in OPIS. Therefore, a search for cases of abnormal release was conducted by referring to publicly available Safety Regulation Status for NPPs from 2000 to 2020 (NSSC 2024). From these data, events leading to abnormal release from 2000 to 2020 were expected to include the secondary system leakages and leakage from fuel defects. For these two groups of events, an analysis was conducted on the correlation between the increase of the activities of radionuclides (groups) discharged and causal events.

Additionally, using data recording the number of fuel defects from 1978 to 2003 for each NPP (Kori Units 1 to 4, Hanbit Units 1 to 5, and Hanul Units 1 to 4), events were defined as fuel defects, and occurrences were counted based on records from 2000 to 2003. For Korean PWRs from 2000 to 2020, abnormal release occurred in 1 RY due to leakage from SG’s valve and in 30 RYs due to leakage from fuel defects, resulting in an Freq_ab,00–20 of 10 % (=31 RYs/310 RYs).

In Korean PWRs, the fission gases discharged during normal operation include ⁷⁹Kr, ⁸⁵Kr, ^85mKr, ⁸⁷Kr, ⁸⁸Kr, ^129mXe,^131mXe, ¹³³Xe, ^133mXe, ¹³⁵Xe, and ^135mXe. The fission gases discharged due to fuel defects in Korean PWRs are ⁸⁵Kr, ^85mKr, ⁸⁷Kr, ⁸⁸Kr, ^131mXe, ¹³³Xe, ^133mXe, ¹³⁵Xe, and ^135mXe, as well as ³H and ¹⁴C.

Figure 8 represents the RCF values when abnormal release due to fuel defects in Korean PWRs. As mentioned in Section 3.2, it can be expected that the activities of radionuclides discharged will increase when the s⁻¹ value is less than 1 and the f value is greater than 1. In Figure 8A, radionuclides with average values less than 1 include ^85mKr, ⁸⁷Kr, ^133mXe, and ^135mXe. Radionuclides with a ratio of s⁻¹ values less than 1 exceeding 50 % include ^85mKr (100 %), ⁸⁷Kr (100 %), ^133mXe (80 %), ^131mXe (73 %), ⁸⁵Kr (50 %), ⁸⁸Kr (50 %), and ^135mXe (50 %). In Figure 8B, radionuclides with average f values greater than 1 include ⁸⁵Kr, ^85mKr, ⁸⁷Kr, ⁸⁸Kr,¹³³Xe, ^133mXe, ¹³⁵Xe, and ³H. Radionuclides with a ratio of f values greater than 1 exceeding 50 % include ^85mKr (100 %), ¹³³Xe (89 %), ³H (67 %), ⁸⁵Kr (63 %), and ¹³⁵Xe (57 %). The radionuclides commonly associated with an increase in the activities when abnormal release occurs due to fuel defects are consistently ⁸⁵Kr and ^85mKr.

Figure 8:

The RCF values for the abnormal release of fission gases, ³H, and ¹⁴C of leakage from fuel defects in Korean PWRs. (A) and (B) represent the values of s⁻¹ and f, respectively.

When applying the three criteria to derive sensitive radionuclides highly correlated with an increase in activities in Korean PWRs using Equation (5), Freq_{ab,Fuel,00–20} is 9.68 % (=30 RYs/310 RYs). Radionuclides satisfying C-① for the s⁻¹ value include ⁸⁵Kr and ¹³³Xe, with ¹³³Xe also meeting C-③. For the f value, radionuclides satisfying all criteria, C-①, C-②, and C-③, are ⁸⁵Kr and ¹³⁵Xe, while radionuclides satisfying C-① and C-② are ^85mKr, ⁸⁷Kr, and ⁸⁸Kr. Among these, ⁸⁵Kr is deemed to be the radionuclide with the highest correlation with an increase in the activities due to fuel defects.

When comparing the abnormal release due to leakage from defected SF during handling between US PWRs in Figure 6 and the Korean PWRs in Figure 8, it was observed that common radionuclides discharged abnormally include ⁸⁵Kr, ^85mKr, ⁸⁷Kr, ⁸⁸Kr, ^131mXe, ¹³³Xe, ^133mXe, and ¹³⁵Xe, as well as ³H, and ¹⁴C. In both countries, the radionuclides most strongly correlated with abnormal releases and an increase in the activities due to leakage from defected SF during handling are ⁸⁵Kr and ⁸⁸Kr.