Fault-Tree Based Prevention Analysis of Cyber-Attack Scenarios for PRA Applications
Vaibhav Yadav, Robert Youngblood, Katya Le Blanc, Jeremy Perschon, Ryan Pitcher
Abstract
Vaibhav Yadav, Robert Youngblood, Katya Le Blanc, Jeremy Perschon, Ryan Pitcher
Abstract
Probabilistic risk assessment (PRA) based on event- and fault-tree analyses has long been a popular and powerful technique for formulating system- and plant-level risk scenarios in high-hazard facilities [1]. Event- and fault-tree-based PRA is commonly performed in the nuclear industry using tools like Systems Analysis Programs for Hands-on Integrity Reliability Evaluation (SAPHIRE) [2] or the Computer Aided Fault Tree Analysis System (CAFTA) [3]. Often, the goal of a PRA effort is to assess the risk of events having high consequences to the public or the environment. In such a case, the fault-tree and event-tree analyses in a PRA mostly focus on initiating events and system and component failures that would result in such consequences. However, a much broader spectrum of consequences is possible in principle from cyber-attack: a computer system could be exposed to an attack that could lead to disruption, financial loss or other damage to the system and its organization. It is not only a major threat for businesses, but has recently impacted infrastructure utilities. In a nuclear power plant, the potential consequences of cyber-attack may range from an inconvenience to unplanned reactor shutdowns or to plant damage, or (in principle) worse; but the low-consequence end of this accident spectrum is not typically addressed using PRA models. Hence it is important to understand better, and minimize the risk of, cyber-attacks in nuclear power plants. The proven fault-tree analysis methodology holds a strong promise of a comprehensive, robust, scalable, and efficient assessment of cyber-attack scenarios in NPPs. This paper presents a fault-tree based formulation for a cyber-attack scenario in a water flow-loop comprised of flow controllers and pumps, controlled via manual controls, wired signals and wireless signals that is susceptible to a cyber-attack. The fault-tree analysis technique is applied to a variety of cyber-attacks that may result in system failure. The analysis provides a comprehensive picture of the attack scenarios and an exhaustive list of attack pathways that are critical for causing system failure, paving the way for formulating strategies of performing cyber-attack prevention analysis.
OpenAlex reports 5 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Probabilistic risk assessment (PRA) based on event- and fault-tree analyses has long been a popular and powerful technique for formulating system- and plant-level risk scenarios in high-hazard facilities [1]. Event- and fault-tree-based PRA is commonly performed in the nuclear industry using tools like Systems Analysis Programs for Hands-on Integrity Reliability Evaluation (SAPHIRE) [2] or the Computer Aided Fault Tree Analysis System (CAFTA) [3]. Often, the goal of a PRA effort is to assess the risk of events having high consequences to the public or the environment. In such a case, the fault-tree and event-tree analyses in a PRA mostly focus on initiating events and system and component failures that would result in such consequences. However, a much broader spectrum of consequences is possible in principle from cyber-attack: a computer system could be exposed to an attack that could lead to disruption, financial loss or other damage to the system and its organization. It is not only a major threat for businesses, but has recently impacted infrastructure utilities. In a nuclear power plant, the potential consequences of cyber-attack may range from an inconvenience to unplanned reactor shutdowns or to plant damage, or (in principle) worse; but the low-consequence end of this accident spectrum is not typically addressed using PRA models. Hence it is important to understand better, and minimize the risk of, cyber-attacks in nuclear power plants. The proven fault-tree analysis methodology holds a strong promise of a comprehensive, robust, scalable, and efficient assessment of cyber-attack scenarios in NPPs. This paper presents a fault-tree based formulation for a cyber-attack scenario in a water flow-loop comprised of flow controllers and pumps, controlled via manual controls, wired signals and wireless signals that is susceptible to a cyber-attack. The fault-tree analysis technique is applied to a variety of cyber-attacks that may result in system failure. The analysis provides a comprehensive picture of the attack scenarios and an exhaustive list of attack pathways that are critical for causing system failure, paving the way for formulating strategies of performing cyber-attack prevention analysis.
Key concepts: Fault tree analysis, Event tree, Probabilistic risk assessment, Event tree analysis, Computer science, Reliability engineering, Nuclear power plant, Risk analysis (engineering)