1. Introduction
Since the Three Mile Island accident in 1979, Chernobyl accident in 1986 and Fukushima accident in 2011, a great emphasis has been put on assessing emergency response capabilities for different types of stakeholders including electric utilities, regulatory bodies and local communities. Estimation of radioactive materials released to the atmosphere, which is usually termed as source term estimation, constitutes an important part of nuclear emergency response. This source term can provide a necessary input for the consecutive simulation of atmospheric dispersion and radiological dose assessment to the public. Based on the dose assessment, corresponding emergency response actions could be recommended such as whether people should be guided by evacuation or shelter‐in‐place protective actions (Zhao et al., 2015). Radiological source term estimation, atmospheric dispersion analysis and radiation dose assessment have been considered as the most important safety concerns in the area of emergency response after a nuclear accident. Therefore, public concerns after nuclear accidents emphasized an absolute necessity to identify and address potential health effects which were closely related with the accurate dose assessment to the public around nuclear power plants (NPPs) (Ahangari et al., 2017).
The siting of NPP is one of important parameters from the point of view of emergency preparedness and response. There are two kinds of site, i.e., on the site (on‐site) and off the site (off‐site). The on‐site area is the area under the control of operators or first responders, and the off‐site area is the area beyond that area. In addition, for the off‐site, there are two kinds of emergency planning zone (EPZ), precautionary action zone (PAZ) and the urgent protective action planning zone (UPZ). Protective actions within PAZ are to be taken before or shortly after a release of radioactive materials on the basis of conditions at NPPs, and protective actions within UPZ are to be mainly taken on the basis of environmental monitoring. The EPZ is divided into PAZ and UPZ which have a radius of 5 km and a radius of 20~30 km from the nuclear facility, respectively (Lee et al., 2016).
Radiation exposure following the release of radioactive materials in nuclear accident can occur through various pathways of exposure. Total effective dose equivalent (TEDE) can be generally calculated from three pathways including external cloud shine, external ground shine and internal inhalation which contributes committed dose equivalent (CDE). In this study, we estimated off‐site radiological doses for severe nuclear accidents of Shin Kori unit 3 which is named an advanced power reactor 1400 (APR1400), a two‐loop pressurized water reactor (PWR) with rated thermal output of 4,000 MWt or electrical power of 1,400 MWe. It is located at the latitude of 35.3236o and the longitude of 129.2915o in the administrative district of Ulsan city. The elevation of the site is 9.5 m from the sea level. The geological location and reactor design parameters of Shin Kori unit 3 are shown in Fig. 1 and Table 1, respectively, according to design control documents of APR1400 (KEPCO and KHNP, 2014). The residential population around Shin Kori unit 3 and the relevant meteorological data are shown in Tables 2 and 3, respectively (KMA, 2019). Using the information above, the effectiveness of emergency management for Shin Kori unit 3 was addressed in terms of radiological source term, TEDE and thyroid CDE by a regulatory accident consequence analysis code. Multiple simulations were performed for two accidents of long‐term station blackout (LTSBO) and loss of coolant accident (LOCA) considering the atmospheric conditions in the time of morning and evening of four seasons. Through the preliminary analysis, the evening time of winter was found to have the most significant effect on the accident (Uddin, 2019). The worst scenario in the evening time of winter was chosen and analyzed for two hypothetical accidents of LTSBO and LOCA in this study.
2. Materials and Methods
Radiological assessment system for consequence analysis (RASCAL) is the software developed by United States Nuclear Regulatory Commission (NRC) emergency operations center to estimate projected doses from radiological emergencies (U.S. NRC, 2017a). The code can provide the rapid dose assessment for an incident or accident and help any decision‐making process such as whether the public should be considered shelter‐in‐place or immediate evacuation. RASCAL can compute nuclear reactor’s source terms, airborne transport of activities through both Gaussian plume and puff models, and the resulting off‐site doses (Athey et al., 2015).
Atmospheric dispersion is generally governed by a differential equation called the diffusion equation. With a set of assumptions that can be reasonably applied to the atmospheric processes, this diffusion equation has a specific and closed‐form algebraic solution of Gaussian function. In a Cartesian coordinate system with x and y axes in the horizontal plane and z in the vertical, the normalized concentration in the vicinity is as follows.
where χ is the concentration (Ci/m3 or g/m3) and Q is the amount of material released (Ci or g). σ is the dispersion parameter (m) which is a function of distance from the release point. When combined with a transport mechanism to move the center of the puff (xo, yo, zo), Eq. (1) is a simplified version of the puff model in RASCAL (Ahangari et al., 2017).
The basic scenario of LTSBO was assumed to have been initiated by an external event which caused the loss of offsite and onsite AC power at Shin Kori unit 3. It was assumed that the reactor was shut down at 5:30, 12 December 2017. Reactor cooling was maintained for a period of several hours but was ultimately lost, at which point the coolant boiled away. Core damage and releases from the core began after core uncovery. Emergency diesel generators (EDGs) were providing the electric power and emergency core cooling systems (ECCS) were available and operating for 5 hrs. These inputs were combined with the shutdown time to determine when the reactor coolant would begin to boil off. Then, RASCAL added a fixed boil‐off duration of 8 hrs according to NUREG‐1935 to determine the time of core melt and release (U.S. NRC, 2012). EDGs became inoperable after 5 hrs and the core began to heat to the point of meltdown. The containment sprays then became inoperable which increased containment pressure and leakage to a release rate of 5% per day. Power was restored at 21:00 and the core was recovered. Containment pressure was reduced by 04:30 of the next day and then containment leakage was stopped (U.S. NRC, 2017b).
In case of LOCA, it was also assumed that there was a major rupture in the primary coolant system after the extreme natural disaster. As a result, the reactor was tripped at 18:00, 12 December 2017. Due to significant loss of coolant passed into the containment building, the core was uncovered at 19:00. Release from the containment to the atmosphere was assumed to follow the leak rate of design control documents of APR1400 (KEPCO and KHNP, 2014). Release rate was assumed to be 3% per day. At 01:00 of the next day, operators were able to recover the core. Containment remained at high pressure which kept the release ongoing and pressure was not reduced until 05:00 of the next day. At that time, the release rate was set to be 0.1% per day (U.S. NRC, 2017c).
3. Results and Discussion
The accident scenario and radiological source terms for LTSBO were shown in Table 4. Noble gases, iodine and particulate nuclides were remarkably released to the atmosphere. Among the released activities to the atmosphere, the proportional portion of noble gases was higher by 51.4% and that of iodine was 27.7%. This was because the reactor core was melted in the case of LTSBO. It was observed that activities were increasing until the time of 21:00 as shown in Figs. 2 and 3. After that, the activities of 135Xe, 132I and 132Te were decreasing slowly because of their relatively short half‐lives which were 9.1 hrs, 2.3 hrs and 3.2 days, respectively. But, 133Xe remained steady after reaching the maximum point compared to others. 133I and 131I were showing the similar characteristics and they decreased rapidly with half‐lives of 20.8 hrs and 8 days, respectively. 135I had a half‐life less than 7 hrs and it shortly decayed to 135Xe. It was also observed that the radioactive plume was moving towards the south‐east direction from Shin Kori unit 3 and accumulated TEDEs during 48 hrs since 18:30 of 12 December were shown in Fig. 4. TEDE and thyroid CDE for LTSBO were shown in Tables 5 and 6, respectively. TEDE and thyroid CDE were 14 mSv and 160 mSv, respectively, at about 5 km of PAZ in the evening time of winter season. Calculated values of TEDE and thyroid CDE in PAZ were higher than the Korean regulations on the indoor sheltering of 10 mSv in 2 days and thyroid protection level of 100 mSv, and therefore it was found that indoor sheltering and distribution of iodine prophylaxis should be immediately implemented for short‐term protective measures. In this situation, an immediate evacuation should be decided by the off‐site emergency management center (OEMC) which was responsible for emergency decision‐making process under the nuclear safety and security commission (NSSC) in Korea.
The accident scenario of LOCA and top eight radionuclides released to the atmosphere for both accidents of LTSBO and LOCA were shown in Tables 7 and 8, respectively. Among the released activities to the atmosphere, the proportional portion of noble gases was higher by 99% but that of iodine was smaller by 0.3% unlike LTSBO. This was because reactor core was not melted in case of LOCA. It was observed that released activities were increasing very fast to their maximum values within 2 hrs and then, the activities were decreasing slowly until 05:00 of the next day as shown in Figs. 5 and 6. 133Xe and 135Xe were showing almost similar characteristics and decreased slowly with 5.2 days and 9.2 hrs of half‐lives, respectively. The behaviors of 88Kr and 85mKr were different because both particles’ half‐lives were relatively short, i.e., 2.8 hrs and 4.5 hrs, respectively. 87Kr decreased very quickly with a short half‐life of 76.3 minutes and there was little contribution to the population to radiation exposure during the early phase of a nuclear accident. 135mXe also decreased very fast to an unstable isotope of xenon with a half‐life of 9.2 hrs. As shown in Fig. 7, the radioactive plume whose TEDE was above 10 mSv did not reach to the distance of 8 km (5 miles) from the facility. TEDE and thyroid CDE were 0.08 mSv and 0.46 mSv, respectively, at about 5 km of PAZ in the evening time of winter season. Both values were sufficiently low when compared with the criteria of 10 mSv for 2 days because most radioactive isotopes released to the atmosphere belonged to noble gases. The accident scenario of LTSBO proved to be more dangerous than that of LOCA in this study, As shown in Figs. 8 and 9, this result could be convinced by the fact that estimated TEDEs of Shin Kori unit 3 for LTSBO and LOCA showed similar trends with those of Arkansas unit 2 and Braidwood unit 1 in the United States, whose reactor powers were 3,026 MWt and 3,645 MWt, respectively, and other reactor design parameters were similar with Shin Kori unit 3 (U.S. NRC, 2017b, 2017c).
4. Conclusions
The objective of this study was to analyze the off‐site radiological consequences from hypothetical severe accidents of Shin Kori unit 3 along with the emergency protective levels in Korea. In order to investigate the effect of released radioactive materials to the public, TEDE and thyroid CDE were evaluated using a regulatory software tool developed by U.S. NRC. Multiple simulations were performed considering reactor conditions and seasonal effects. In the event of LTSBO, TEDE and thyroid CDE were estimated to be 14 mSv and 160 mSv, respectively, at about 5 km of PAZ in the evening time of winter season. Estimated values of TEDE and thyroid CDE in PAZ were higher than the Korean regulations on the indoor sheltering of 10 mSv in 2 days and thyroid protection level of 100 mSv, and therefore it was found that more careful public protective measures of emergency management should be taken in the accident of LTSBO.