Abstract
The presence of high levels of moisture in the annulus gas system of a CANDU reactor indicates that a leaking crack may be present in a pressure tube. This will initiate the shutdown of the reactor to prevent the possibility of fuel channel damage. It is also desirable, however, to keep the reactor partially pressurized at hot shutdown for as long as it is necessary to unambiguously identify the leaking pressure tube. A premature full depressurization may cause an extended shutdown while the leaking tube is being located. However, fast fracture could occur during an excessively long hot shutdown period. A probabilistic methodology, together with an associated computer code (called MARATHON), has been developed to calculate the time from first leakage to unstable fracture in a probabilistic format. The methodology explicitly uses distributions of material properties and allows the risk associated with leak-before-break to be estimated. A model of the leak detection system is integrated into the methodology to calculate the time from leak detection to unstable fracture. The sensitivity of the risk to changing reactor conditions allows the optimization of reactor management after leak detection. In this report we describe the probabilistic model and give details of the quality
More>>
Citation Formats
Walker, J R.
MARATHON - a computer code for the probabilistic estimation of leak-before-break time in CANDU reactors.
Canada: N. p.,
1990.
Web.
Walker, J R.
MARATHON - a computer code for the probabilistic estimation of leak-before-break time in CANDU reactors.
Canada.
Walker, J R.
1990.
"MARATHON - a computer code for the probabilistic estimation of leak-before-break time in CANDU reactors."
Canada.
@misc{etde_10128292,
title = {MARATHON - a computer code for the probabilistic estimation of leak-before-break time in CANDU reactors}
author = {Walker, J R}
abstractNote = {The presence of high levels of moisture in the annulus gas system of a CANDU reactor indicates that a leaking crack may be present in a pressure tube. This will initiate the shutdown of the reactor to prevent the possibility of fuel channel damage. It is also desirable, however, to keep the reactor partially pressurized at hot shutdown for as long as it is necessary to unambiguously identify the leaking pressure tube. A premature full depressurization may cause an extended shutdown while the leaking tube is being located. However, fast fracture could occur during an excessively long hot shutdown period. A probabilistic methodology, together with an associated computer code (called MARATHON), has been developed to calculate the time from first leakage to unstable fracture in a probabilistic format. The methodology explicitly uses distributions of material properties and allows the risk associated with leak-before-break to be estimated. A model of the leak detection system is integrated into the methodology to calculate the time from leak detection to unstable fracture. The sensitivity of the risk to changing reactor conditions allows the optimization of reactor management after leak detection. In this report we describe the probabilistic model and give details of the quality assurance and verification of the MARATHON code. Examples of the use of MARATHON are given using preliminary material property distributions. These preliminary material property distributions indicate that the probability of unstable fracture is very low, and that ample time is available to locate the leaking tube.}
place = {Canada}
year = {1990}
month = {Feb}
}
title = {MARATHON - a computer code for the probabilistic estimation of leak-before-break time in CANDU reactors}
author = {Walker, J R}
abstractNote = {The presence of high levels of moisture in the annulus gas system of a CANDU reactor indicates that a leaking crack may be present in a pressure tube. This will initiate the shutdown of the reactor to prevent the possibility of fuel channel damage. It is also desirable, however, to keep the reactor partially pressurized at hot shutdown for as long as it is necessary to unambiguously identify the leaking pressure tube. A premature full depressurization may cause an extended shutdown while the leaking tube is being located. However, fast fracture could occur during an excessively long hot shutdown period. A probabilistic methodology, together with an associated computer code (called MARATHON), has been developed to calculate the time from first leakage to unstable fracture in a probabilistic format. The methodology explicitly uses distributions of material properties and allows the risk associated with leak-before-break to be estimated. A model of the leak detection system is integrated into the methodology to calculate the time from leak detection to unstable fracture. The sensitivity of the risk to changing reactor conditions allows the optimization of reactor management after leak detection. In this report we describe the probabilistic model and give details of the quality assurance and verification of the MARATHON code. Examples of the use of MARATHON are given using preliminary material property distributions. These preliminary material property distributions indicate that the probability of unstable fracture is very low, and that ample time is available to locate the leaking tube.}
place = {Canada}
year = {1990}
month = {Feb}
}