• Nuclear Engineering and Technology
• /
• v.56 no.6
• /
• pp.2266-2273
• /
• 2024

A machine learning platform is proposed for the diagnosis of a severe accident progression in a nuclear power plant. To predict the key parameters for accident management including lost signals, a long short term memory (LSTM) network is proposed, where multiple accident scenarios are used for training. Training and test data were produced by MELCOR simulation of the Fukushima Daiichi Nuclear Power Plant (FDNPP) accident at unit 3. Feature variables were selected among plant parameters, where the importance ranking was determined by a recursive feature elimination technique using RandomForestRegressor. To answer the question of whether a reduced order ML model could predict the complex transient response, we performed a systematic sensitivity study for the choices of target variables, the combination of training and test data, the number of feature variables, and the number of neurons to evaluate the performance of the proposed ML platform. The number of sensitivity cases was chosen to guarantee a 95 % tolerance limit with a 95 % confidence level based on Wilks' formula to quantify the uncertainty of predictions. The results of investigations indicate that the proposed ML platform consistently predicts the target variable. The median and mean predictions were close to the true value.

• Journal of Radiation Protection and Research
• /
• v.41 no.2
• /
• pp.87-91
• /
• 2016

Background: A hand-held detector for an emergency response was developed for nuclide identification and to estimate the information of the ambient dose rate in the scene of an accident as well as the radioactivity of the contaminants. Materials and Methods: To achieve this, the most suitable sensor was first selected as a cadmium zinc telluride (CZT) semiconductor and the signal processing unit from a sensor and the signal discrimination and storage unit were successfully manufactured on a printed circuit board. Results and Discussion: The performance of the developed signal processing unit was then evaluated to have an energy resolution of about 14 keV at 662 keV. The system control unit was also designed to operate the CZT detector, monitor the detector, battery, and interface status, and check and transmit the measured results of the ambient dose rate and radioactivity. In addition, a collimator, which can control the inner radius, and the airborne dust sampler, which consists of an air filter and charcoal filter, were developed and mounted to the developed CZT detector for the quick and efficient response of a nuclear accident. Conclusion: The hand-held CZT detector was developed to make the in-situ gamma-ray spectrometry and its performance was checked to have a good energy resolution. In addition, the collimator and the airborne dust sampler were developed and mounted to the developed CZT detector for a quick and efficient response to a nuclear accident.

• Journal of the Korean Institute of Gas
• /
• v.19 no.5
• /
• pp.87-97
• /
• 2015

Accidents of HAZMAT transport vehicle are generally dealt by unskilled firefighters. As a result, firefighters became 2nd victims of accidents in some case because of this reason, and it is required to prepare an accident response guidelines against HAZMAT transport vehicle accidents. In this study, risk assessment methods and making methods of guidelines were investigated to make accident response guidelines of HAZMAT transport vehicle accidents. and It identifies hazards and combines with guideword for making guideline items. At last, we determine criteria or detailed methods by referring survey regulations and existed methods.

• Journal of Environmental Policy
• /
• v.11 no.2
• /
• pp.17-54
• /
• 2012

This paper researched market responses for listed companies in several industries affected by the major oil spill accident off the coast of Taean, in December 7, 2007. The Taean accident triggered considerable concerns in people over the possibility of potential future regulation in shipbuilding and petroleum industries. However, the accident also provided an unexpected business opportunity for environmental clean-up industry and shipbuilding industry. The oil spill triggered the acceleration of the enactment of policies that require all new oil tankers to be constructed with double hull, which is interpreted as a good news for shipbuilding industry. Increased public pressure coupled with the prospect of tightened regulation is expected to decrease the market values of firms in the affected business fields. The stock prices of shipbuilding companies dropped after the incident but dramatically surged after the enactment of the policy in January 31, 2008. Our study also found that companies with more extensive prior environmental disclosure had less negative market reactions during the first sixteen days following the accident.

• Journal of Navigation and Port Research
• /
• v.46 no.3
• /
• pp.227-235
• /
• 2022

The purpose of this study was to analyze the risk and vulnerability of marine accidents based on statistical data on marine accidents at Ulsan Port, which has the largest amount of liquid cargo in Korea. It was found to be quite vulnerable to the risk of marine accidents, environmental damage, and vulnerability to environmental pollution accidents. Based on analysis results, marine accident scenarios and accident response strategies were prepared. Additionally, as a response strategy to prepare for large-scale marine pollution accidents at Ulsan Port, it is necessary to establish control equipment and infrastructure, as well as establish a control center to integrate marine accident safety functions. In particular, in the case of liquid cargo specialized ports such as Ulsan Port, considering the size of the cargo volume and the frequency of marine pollution accidents, it is urgent to build professional safety management institutions, which should make the port safer.

• Journal of the Society of Disaster Information
• /
• v.9 no.4
• /
• pp.521-531
• /
• 2013

Temporary residence in disaster scenarios emergency response system designed for utilizing emergency incident types identified through the rapid emergency response procedures and proper application of the resulting accident is very important in order to minimize the damage. In this study, emergency disaster expected to analyze the potential for accidents, the possible accident scenarios emergency disaster preparedness These were derived on the basis of temporary residence in the emergency scenario, the scenario for the emergency response and recovery measures were derived. In addition, each event of an emergency response scenario to be taken during the emergency response requirements for emergency response to a step-by-step behavior was defined by the subject. In addition, subject-specific, step-by-step emergency disaster scenarios emergency response is quick and can be done in a comprehensive foundation was laid.

• Proceedings of the Safety Management and Science Conference
• /
• 2011.11a
• /
• pp.583-591
• /
• 2011

A number of Brain-Computer Interface (BCI) studies have been performed to assess the cognitive status through EEG signal. However, there are a few studies trying to prevent user from unexpected safety-accident in BCI study. The EEGs were collected from 19 subjects who participated in two experiments (rest & event-related potential measurement). There was significant difference in EEG changes of both spontaneous and event-related potential. Beta power and P300 latency may be useful as a biomarker for prevention of response to safety-accident.

• Journal of the Korea Safety Management & Science
• /
• v.14 no.1
• /
• pp.1-6
• /
• 2012

A number of Brain-Computer Interface (BCI) studies have been performed to assess the cognitive status through EEG signal. However, there are a few studies trying to prevent user from unexpected safety-accident in BCI study. The EEGs were collected from 19 subjects who participated in two experiments (rest & event-related potential measurement). There was significant difference in EEG changes of both spontaneous and event-related potential. Beta power and P300 latency may be useful as a biomarker for prevention of response to safety-accident.

• Nuclear Engineering and Technology
• /
• v.53 no.10
• /
• pp.3256-3263
• /
• 2021

The main steam line break accident is an essential initiating event of the pressurized water reactor. In present work, the fuzzy set theory and the signal-based fault detection method has been used to detect the occurrence and diagnosis of the location and break area for the small scale MSLB. The models are validated by the AP1000 accident simulator based on MAAP5. From the test results it can be seen that the proposed approach has a rapid and proper response on accident detection and location diagnosis. The method proposed to evaluate the break area shows good performances for small scale MSLB with the relative deviation within ±3%.

• Nuclear Engineering and Technology
• /
• v.50 no.8
• /
• pp.1412-1420
• /
• 2018

An experiment was performed using the large-scale test facility (LSTF), which simulated a 1% vessel upper head small-break loss-of-coolant accident with an accident management (AM) measure under an assumption of total-failure of high-pressure injection (HPI) system in a pressurized water reactor (PWR). In the LSTF test, liquid level in the upper head affected break flow rate. Coolant was manually injected from the HPI system into cold legs as the AM measure when the maximum core exit temperature reached 623 K. The cladding surface temperature largely increased due to late and slow response of the core exit thermocouples. The AM measure was confirmed to be effective for the core cooling. The RELAP5/MOD3.3 code indicated insufficient prediction of primary coolant distribution. The author conducted uncertainty analysis for the LSTF test employing created phenomena identification and ranking table for each component. The author clarified that peak cladding temperature was largely dependent on the combination of multiple uncertain parameters within the defined uncertain ranges.