skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Conceptual Software Reliability Prediction Models for Nuclear Power Plant Safety Systems

Abstract

The objective of this project is to develop a method to predict the potential reliability of software to be used in a digital system instrumentation and control system. The reliability prediction is to make use of existing measures of software reliability such as those described in IEEE Std 982 and 982.2. This prediction must be of sufficient accuracy to provide a value for uncertainty that could be used in a nuclear power plant probabilistic risk assessment (PRA). For the purposes of the project, reliability was defined to be the probability that the digital system will successfully perform its intended safety function (for the distribution of conditions under which it is expected to respond) upon demand with no unintended functions that might affect system safety. The ultimate objective is to use the identified measures to develop a method for predicting the potential quantitative reliability of a digital system. The reliability prediction models proposed in this report are conceptual in nature. That is, possible prediction techniques are proposed and trial models are built, but in order to become a useful tool for predicting reliability, the models must be tested, modified according to the results, and validated. Using methods outlined by this project,more » models could be constructed to develop reliability estimates for elements of software systems. This would require careful review and refinement of the models, development of model parameters from actual experience data or expert elicitation, and careful validation. By combining these reliability estimates (generated from the validated models for the constituent parts) in structural software models, the reliability of the software system could then be predicted. Modeling digital system reliability will also require that methods be developed for combining reliability estimates for hardware and software. System structural models must also be developed in order to predict system reliability based upon the reliability of the individual hardware/software components. Existing modeling techniques--such as fault tree analyses or reliability block diagrams--can probably be adapted to bridge the gaps between the reliability of the hardware components, the individual software elements, and the overall digital system. This project builds upon previous work to survey and rank potential measurement methods which could be used to measure software product reliability 3. This survey and ranking identified candidate measures for use in predicting the reliability of digital computer-based control and protection systems for nuclear power plants. Additionally, information gleaned from the study can be used to supplement existing review methods during an assessment of software-based digital systems.« less

Authors:
; ;
Publication Date:
Research Org.:
Lawrence Livermore National Lab., CA (US)
Sponsoring Org.:
USDOE Office of Defense Programs (DP) (US)
OSTI Identifier:
791856
Report Number(s):
UCRL-ID-138577
TRN: US0300442
DOE Contract Number:  
W-7405-Eng-48
Resource Type:
Technical Report
Resource Relation:
Other Information: PBD: 3 Apr 2000
Country of Publication:
United States
Language:
English
Subject:
21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS; 99 GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE; ACCURACY; CONTROL SYSTEMS; DIGITAL SYSTEMS; DISTRIBUTION; FORECASTING; NUCLEAR POWER PLANTS; PROBABILITY; RELIABILITY; RISK ASSESSMENT; SAFETY; STRUCTURAL MODELS; TREES; VALIDATION

Citation Formats

Johnson, G, Lawrence, D, and Yu, H. Conceptual Software Reliability Prediction Models for Nuclear Power Plant Safety Systems. United States: N. p., 2000. Web. doi:10.2172/791856.
Johnson, G, Lawrence, D, & Yu, H. Conceptual Software Reliability Prediction Models for Nuclear Power Plant Safety Systems. United States. doi:10.2172/791856.
Johnson, G, Lawrence, D, and Yu, H. Mon . "Conceptual Software Reliability Prediction Models for Nuclear Power Plant Safety Systems". United States. doi:10.2172/791856. https://www.osti.gov/servlets/purl/791856.
@article{osti_791856,
title = {Conceptual Software Reliability Prediction Models for Nuclear Power Plant Safety Systems},
author = {Johnson, G and Lawrence, D and Yu, H},
abstractNote = {The objective of this project is to develop a method to predict the potential reliability of software to be used in a digital system instrumentation and control system. The reliability prediction is to make use of existing measures of software reliability such as those described in IEEE Std 982 and 982.2. This prediction must be of sufficient accuracy to provide a value for uncertainty that could be used in a nuclear power plant probabilistic risk assessment (PRA). For the purposes of the project, reliability was defined to be the probability that the digital system will successfully perform its intended safety function (for the distribution of conditions under which it is expected to respond) upon demand with no unintended functions that might affect system safety. The ultimate objective is to use the identified measures to develop a method for predicting the potential quantitative reliability of a digital system. The reliability prediction models proposed in this report are conceptual in nature. That is, possible prediction techniques are proposed and trial models are built, but in order to become a useful tool for predicting reliability, the models must be tested, modified according to the results, and validated. Using methods outlined by this project, models could be constructed to develop reliability estimates for elements of software systems. This would require careful review and refinement of the models, development of model parameters from actual experience data or expert elicitation, and careful validation. By combining these reliability estimates (generated from the validated models for the constituent parts) in structural software models, the reliability of the software system could then be predicted. Modeling digital system reliability will also require that methods be developed for combining reliability estimates for hardware and software. System structural models must also be developed in order to predict system reliability based upon the reliability of the individual hardware/software components. Existing modeling techniques--such as fault tree analyses or reliability block diagrams--can probably be adapted to bridge the gaps between the reliability of the hardware components, the individual software elements, and the overall digital system. This project builds upon previous work to survey and rank potential measurement methods which could be used to measure software product reliability 3. This survey and ranking identified candidate measures for use in predicting the reliability of digital computer-based control and protection systems for nuclear power plants. Additionally, information gleaned from the study can be used to supplement existing review methods during an assessment of software-based digital systems.},
doi = {10.2172/791856},
journal = {},
number = ,
volume = ,
place = {United States},
year = {2000},
month = {4}
}

Technical Report:

Save / Share: