Mohaghegh team receives NSF grant to use big data analytics in new-generation probabilistic risk assessment


“NPRE Assistant Prof. Zahra Mohaghegh is the Principal Investigator on a five-year National Science Foundation (NSF) project that integrates big data analytics into Probabilistic Risk Assessment (PRA).

Entitled “A Big Data-Theoretic Approach to Quantify Organizational Failure Mechanisms in Probabilistic Risk Assessment,” the $899,000 project is funded by two NSF programs: Science of Organizations (SoO) and Big Data Science and Engineering (BIGDATA).

“Catastrophic events such as Fukushima have made it clear that the capability of integrating physical and social causes of failure into a socio-technical modeling framework is the future of risk analysis,” said Mohaghegh. The NSF research aims to quantify this framework for the risk analysis of nuclear power plants. The methodology is also applicable for other high-risk industries, including aviation, healthcare, oil and gas.

Mohaghegh, an expert in PRA, is teaming up with specialists in Organizational Behavior (Co-PI Prof. Cheri Ostroff; University of South Australia) and Information Science (Co-PI Associate Prof. Catherine Blake; University of Illinois at Urbana-Champaign) to advance predictive causal modeling and big-data theoretic technologies for PRA. Current organizational risk contributors lack reliable data analytics, according to Justin Pence, an informatics PhD student and member of Mohaghegh’s group in the Socio-Technical Risk Analysis (SoTeRiA) Laboratory. These researchers are expanding the classical approach of data management for risk analysis by using big data analytic techniques and simulation to uncover organizational contributors to system risk.

This research will advance the Big Data-Theoretic Algorithm; a methodology for extracting and interpreting socio-technical information from unstructured textual communications. The PI’s research group has developed the Big Data-Theoretic Algorithm and initiated its application in the nuclear power industry to perform text mining on Corrective Action Program documents. In addition to unstructured data, this research deals with large volume of data to perform uncertainty analysis on large-scale risk frameworks. Mohaghegh is an investor in the Illinois Campus Cluster Program (ICCP), allowing her to use ICCP resources in risk research projects.

A member of the NPRE faculty for the past two years, Mohaghegh has pioneered research in the systematic causal modeling of physical and social failure mechanisms by incorporating Big Data Analytics and spatio-temporal dimensions into PRA. PRA is one of the key pillars of the Risk-Informed regulatory framework for the Nuclear Regulatory Commission (NRC). A growing number of other U.S. government agencies – the Department of Energy (DOE), the Federal Aviation Administration (FAA), the National Aeronautics and Space Administration (NASA), the Department of Defense (DOD), the Environmental Protection Agency (EPA), and the Food and Drug Administration (FDA) – also have begun to use PRA for policy setting and decision-making.

Dean’s Award for Excellence in Research
Mohaghegh was honored in Spring 2015 with the College of Engineering at Illinois Dean’s Award for Excellence in Research for her innovative and scientific contributions to risk analysis. She views the award as a result of state-of-the-art research emanating from the SoTeRiA team.

Since coming to Illinois, Mohaghegh and her group have been involved in a large-scale academia-industry project, sponsored by the South Texas Project Nuclear Operating Company. The work has contributed to developing an integrated risk methodology for the resolution of the nuclear industry’s 20-year-long and very challenging Generic Safety Issue 191 (GSI-191).

Examples of the lab’s current research projects include:

  • Fire PRA in nuclear power plants;
  • Location-specific Loss of Coolant Accidents (LOCA) leading to Emergency Core Cooling System failure;
  • Modeling the effects of human and organizational factors on nuclear power plant technical system failure;
  • Socio-technical risk-informed emergency preparedness, planning and response modeling for severe accidents;
  • Evaluating PRA’s monetary value.

International Impact
The SoTeRiA Laboratory aims to establish the University of Illinois as a global leader in Risk Analysis education and research. In March 2015, Mohaghegh’s group developed and presented the U.S.-China Probabilistic Safety Assessment Workshop on “Risk-Informed Regulation and Safety Culture,” in Shenzhen, China, as part of the DOE’s Peaceful Uses of Nuclear Technology (PUNT) program.

The SoTeRiA team coordinated and taught a one-week training workshop on risk-informed regulation and safety culture for the Chinese nuclear power industry. Seventy-two representatives from 28 Chinese nuclear energy organizations and entities attended, including plant managers, operators, engineers, and researchers. NPRE research affiliate Ernie Kee and Ph.D. candidate Tatsuya Sakurahara joined Mohaghegh and Pence in leading the workshop, in collaboration with U.S. regulatory experts Glenn Kelly (retired from the Nuclear Regulatory Commission) and Weidong He from AdSTM, LLC.

With 10 years experience in PRA research, Mohaghegh believes that “while the U.S. leads in risk analysis methods for nuclear safety, PRA researchers applying analysis techniques in countries such as China and Japan must take into account varying geographies, cultures and operating experiences to avoid potentially misleading results.” Mohaghegh and her team have initiated collaborations with national and international research institutions in order to achieve tailor-made solutions for high-risk operations around the world.

Impact at the University of Illinois
At the University of Illinois, Mohaghegh and her SoTeRiA team are presenting foundational undergraduate and graduate risk analysis courses to train the next generation for risk analysis in high-consequence industries. In addition to students from NPRE, the courses have attracted students from multiple departments including Civil and Environmental Engineering, Mechanical Science and Engineering, Aerospace Engineering, Industrial Systems and Enterprise Engineering, and Informatics. Risk Analysis courses address fundamental theories of risk-scenario modeling and accident phenomenology, uncertainty analysis, Bayesian and data analysis, probabilistic physics of failure, human error modeling, and next generation PRA methods and tools. Students are provided with hands-on opportunities using PRA software to address real-world risk analysis problems.

The momentum of Mohaghegh and her SoTeRiA research team is influenced by a shared vision to usher in a new era, void of catastrophic technological accidents, where industry-academia projects are leveraged to raise social responsibility for the protection of workers, the public, and the environment.”

CookieSettings CookieSettings