Search

KR-20260067011-A - Risk Evaluation Method based on Hydrogen Charging Station Leakage Data

KR20260067011AKR 20260067011 AKR20260067011 AKR 20260067011AKR-20260067011-A

Abstract

The present invention relates to a technique for evaluating leakage risk based on hydrogen leakage data of a hydrogen refueling station, comprising: a first step of collecting hydrogen concentration data leaked through a hydrogen leakage sensor for each facility of the hydrogen refueling station for a predetermined period; a second step of calculating a probability distribution by separating only the data where leakage occurred among the hydrogen concentration data collected in the first step, or calculating a probability distribution by separating only the maximum value and the upper value close thereto among the daily hydrogen concentration data for a predetermined number of cases; a third step of dividing the probability distribution calculated in the second step into predetermined intervals and setting a risk grade according to the hydrogen concentration value of each interval; a fourth step of evaluating the risk for each facility by comparing the hydrogen concentration value of the hydrogen concentration data collected in real time through the hydrogen leakage sensor with the risk grade set in the third step; and a fifth step of determining signs of hydrogen leakage for each facility based on the risk grade evaluated in the fourth step and the frequency of occurrence, mean, or standard deviation during a predetermined period.

Inventors

  • 김정훈
  • 오정석

Assignees

  • 한국가스안전공사

Dates

Publication Date
20260512
Application Date
20241105

Claims (7)

  1. This relates to a technique for evaluating leakage risk based on hydrogen leakage data from hydrogen refueling stations, A first step of collecting hydrogen concentration data leaking through hydrogen leak sensors for each piece of equipment at a hydrogen refueling station for a preset period; A second step of calculating a probability distribution by separating only the data where leakage occurred from the hydrogen concentration data collected in the first step, or by separating only the maximum value and the upper values close to it from the daily hydrogen concentration data for a predetermined number of cases and calculating a probability distribution; A third step of dividing the probability distribution calculated in the second step into pre-set intervals and setting a risk level based on the hydrogen concentration values of each interval; A fourth step of evaluating the risk level for each facility by comparing the hydrogen concentration values of hydrogen concentration data collected in real time through a hydrogen leak sensor with the risk level set in the third step; and, Step 5, determining signs of hydrogen leakage for each facility based on the risk level evaluated in Step 4 and the frequency, mean, or standard deviation of occurrence over a preset period; A hydrogen refueling station leakage data-based risk assessment technique characterized by including
  2. In paragraph 1, Stage 3 is, Risk grades are divided into Grade A, which signifies safety, and the remaining grades, which signify danger, and Grade A is, In the probability distribution, it is set to an interval where the hydrogen leakage concentration is 0 ppm or higher and less than the mean + 1 * standard deviation, or is set to an interval where the hydrogen leakage concentration is 0 ppm or higher and less than the mean + 2 * standard deviation, and A risk assessment technique based on hydrogen refueling station leakage data, characterized in that the remaining grades representing risk are divided into equal intervals according to a preset number of grades, excluding the section set as Grade A in the probability distribution, and grades are set in alphabetical order starting with Grade B, in order from the section with the lowest hydrogen leakage concentration to the section with the highest.
  3. In paragraph 2, A risk assessment technique based on hydrogen refueling station leakage data, characterized by dividing the entire remaining section, excluding the section set as Grade A in the probability distribution, into four equal intervals, and setting grades from Grade B in alphabetical order to Grade E in order from the section with the lowest hydrogen leakage concentration to the section with the highest.
  4. In paragraph 2, Stage 4 is, A hydrogen refueling station leak data-based risk assessment technique characterized by evaluating the risk and assigning a grade based on a preset risk grade for hydrogen concentration data collected in real time through hydrogen leak sensors for each piece of equipment at the hydrogen refueling station.
  5. In Paragraph 3, Stage 4 is, For hydrogen concentration data collected in real-time through hydrogen leak sensors for each piece of equipment at a hydrogen refueling station, the risk is evaluated according to a preset risk grade, and a grade from A to E is selected and assigned. Stage 5 is, A hydrogen refueling station leak data-based risk assessment technique characterized by determining that there is a sign of hydrogen leakage if the risk grade assigned in the fourth stage corresponds to Grade C, Grade D, or Grade E.
  6. In paragraph 5, Stage 5 is, Even if the risk level assigned in Stage 4 corresponds to Grade A or Grade B, (1) If there are 3 or more cases where a Grade B was assigned during the past 5 consecutive days as of the current point in time, or (2) Calculate the daily average of only the data where leakage occurred among the hydrogen concentration data collected over the past 5 consecutive days based on the current point in time, or calculate the daily average of the maximum value and the top 100 data points closest to it among the daily hydrogen concentration data, and perform linear regression analysis on 5 daily average data points where the slope is positive and the coefficient of determination is 0.5 or higher, or (3) Calculate the daily standard deviation of only the data where leakage occurred among the hydrogen concentration data for the past 5 consecutive days based on the current point in time, or calculate the daily standard deviation of the maximum value and the top 100 data closest to it among the daily hydrogen concentration data, and perform linear regression analysis with 5 daily standard deviation data and if the slope is positive and the coefficient of determination is 0.5 or higher, A risk assessment technique based on hydrogen refueling station leakage data, characterized by assigning a risk grade of C and determining it as a sign of hydrogen leakage if any one of the following cases applies.
  7. In any one of paragraphs 1 through 6, Step 6, which assigns an overall risk grade to the hydrogen refueling station based on the highest risk grade among the risk grades assigned to each facility of the hydrogen refueling station; A hydrogen refueling station leakage data-based risk assessment technique characterized by further including

Description

Risk Evaluation Method based on Hydrogen Charging Station Leakage Data The present invention is a technology for preventing accident risks in advance by evaluating the risk level of a hydrogen refueling station based on leakage data generated from each major facility constituting the hydrogen refueling station. Hydrogen refueling stations are required to install hydrogen leak sensors on the upper parts of compressors, dispensers, etc., for safety management against leaks, and through this, the presence and concentration of hydrogen leaks are detected 24 hours a day. The safety of hydrogen refueling stations is managed by a method of transmitting an alarm when the concentration of leaked hydrogen detected by these hydrogen leak sensors reaches 10,000 ppm, which is about one-fourth of the explosion limit concentration (for example, alarms are set at 20% of the explosion limit concentration (20% LEL = 8,000 ppm) and 25% of the explosion limit concentration (25% LEL = 10,000 ppm)), but there is a problem in that no separate alarm or measures are taken for minute leaks that occur continuously during normal times in the main equipment of the hydrogen refueling station, such as compressors, dispensers, and panels, so the risk of the hydrogen refueling station, which is a hazardous facility, cannot be predicted in the near future or the risk level of the current state at all. Furthermore, if major equipment fails or deteriorates, leakage usually occurs; there is a high risk that such leaks will rapidly escalate at some point, leading to a major accident involving a sudden explosion. There are numerous prior technologies for evaluating risks related to hydrogen refueling stations or hydrogen facilities, which can be briefly reviewed as follows. 1) Registered Patent No. 10-2374469 relates to an apparatus for calculating the risk of a hydrogen refueling station accident, comprising: a feature map generation unit that, upon input data including the location of the hydrogen refueling station and the scale of the hydrogen refueling station, generates a feature map including an accident feature map, a topographic feature map, and a population distribution feature map based on the input data; a risk prediction unit that inputs the feature map into a prediction model and causes the prediction model to perform multiple operations in which weights learned between multiple layers are applied to the feature map to generate a personal risk map representing the personal risk level due to the accident and a social risk map representing the social risk level due to the accident; an evaluation unit that evaluates the risk level of the hydrogen refueling station based on the personal risk map and the social risk map; and a storage unit that stores the location of a hydrogen refueling station where an accident occurred in the past, the scale of the hydrogen refueling station where the accident occurred, and the original personal risk map and original social risk map corresponding to the hydrogen refueling station where the accident occurred. The method is characterized by including a model generation unit that extracts the location of the hydrogen refueling station where an accident occurred in the past, the scale of the hydrogen refueling station where the accident occurred, and the original personal risk map and original social risk map corresponding to the hydrogen refueling station where the accident occurred as training data, and trains a prediction model to generate a simulated personal risk map and a simulated social risk map using the extracted training data. 2) Registered Patent No. 10-2574441 relates to a real-time risk prediction system for hydrogen facilities, comprising: a current status measurement unit that monitors the current status of a hydrogen-based facility in real time using a plurality of sensors installed in the hydrogen-based facility; a risk assessment unit that evaluates the risk of the hydrogen-based facility using a hydrogen-based facility map, hydrogen-based facility design specifications, and the amount of hydrogen supplied to the hydrogen-based facility; and a system that generates and specifies an accident scenario using the risk of the hydrogen-based facility, and in the generated accident scenario The system includes an accident response unit that generates a corresponding accident response scenario; wherein the risk assessment unit includes a real-time risk determination module that determines the real-time risk of the hydrogen infrastructure using real-time changes in the current state obtained through multiple sensors installed in the hydrogen infrastructure in real time; wherein the real-time risk determination module tracks whether changes in values measured by multiple sensors installed in the hydrogen infrastructure change to abnormal values, calculates cases where values exceeding a preset safety range are measured or values change with a slope exceeding the safety range, and determines the r