Projects

Selected engineering projects in probabilistic risk assessment, nuclear safety, risk-informed design, and engineering software development.

PRA Model Modernization and Level 1–Level 2 Integration

National Commission on Nuclear Safety and Safeguards (CNSNS), Mexico City | BWR Mark II | 2024-2025

This two-year project focused on the modernization, technical review, and consolidation of a nuclear power plant Probabilistic Risk Assessment (PRA) model. The work strengthened the Level 1 model, developed the interface with Level 2 PRA (Plant Damage States), incorporated human reliability and accident-management strategies, and established the initial framework for future dynamic PRA applications. The project supported the development of a technically consistent and traceable model suitable for regulatory analysis and risk-informed decision-making, including future applications related to the aging management of structures, systems, and components.

Selected Technical Contributions

  • PRA model restructuring and documentation. Reorganized large fault trees using external transfers and modular system representations to improve model readability, maintainability, and technical review. Developed descriptions for basic events, logic gates, systems, and accident sequences to provide greater traceability throughout the model.
  • Python-based model automation and quality assurance. Developed Python tools to interpret structured basic-event identifiers, automatically generate technical descriptions, and create files that could be imported into SAPHIRE. Additional scripts were implemented to identify unclassified basic events, verify partition-rule coverage, and support the systematic review of large quantities of model information.
  • Event-tree and sequence-logic enhancement. Revised event-tree linking rules, sequence numbering, transfer logic, flag sets, and unused model elements. New event trees were developed for steam-line LOCA sequences and for transfers originating from loss-of-offsite-power and station-blackout scenarios, improving the logical relationship between initiating events, system responses, and final accident-sequence outcomes.
  • ATWS model improvement. Updated the Anticipated Transient Without Scram event-tree logic. Additional top-event probabilities and reactor shutdown functions were incorporated to improve the treatment of reactor protection, manual shutdown actions, recirculation-pump trip, and turbine-bypass availability.
  • Common-cause failure analysis. Conducted a comparative assessment of common-cause failure modeling using the Beta Factor, Multiple Greek Letter, and Alpha Factor methods. 
  • Human Reliability Analysis using SPAR-H. Applied the Standardized Plant Analysis Risk–Human Reliability Analysis methodology to quantify human error probabilities for diagnostic and action-related tasks. Performance-shaping factors—including available time, stress, task complexity, procedures, training, ergonomics, operator condition, and communication—were evaluated using engineering judgment and accident-sequence context. Dependencies among related operator actions were also considered.
  • Level 1–Level 2 PRA interface development. Developed and refined partition rules to transform Level 1 minimal cut sets into Plant Damage State vectors for Level 2 analysis. The classification considered reactor shutdown status, vessel pressure, availability and recoverability of high- and low-pressure injection systems, containment spray capability, containment condition, and timing of core damage.
  • FLEX and accident-management strategies. Incorporated post-Fukushima accident-management, portable mitigation, equipment recovery, and containment-protection strategies into the probabilistic model.
  • RCIC and station-blackout model refinement. Used MELCOR simulation results to assess the temporal relationship among DC power depletion, RCIC operation, wetwell pressurization, and containment response during loss-of-offsite-power sequences. These deterministic insights were used to revise PRA assumptions, system failure logic, and severe-environment survivability modeling.
  • Integrated Level 1–Level 2 execution framework. Developed an event-tree interface in SAPHIRE in which Level 1 sequence results and partition-rule outputs activate the corresponding Level 2 accident-progression models. This structure improved the traceability between core-damage sequences, Plant Damage States, containment response, and potential release pathways.
  • Initial framework for dynamic PRA. Developed a Python-based MELCOR simulation controller as the foundation for future dynamic PRA applications. The controller included automated steady-state execution, initiating-event management, time-dependent controls, a graphical time interface, and routines for modifying MELCOR restart files. The architecture was designed to support future communication between deterministic MELCOR simulations and SAPHIRE probabilistic models. 

Project Impact

The project substantially improved the technical quality, internal consistency, transparency, and maintainability of the PRA model. The combined review of fault trees, event trees, linking and partition rules, common-cause failures, human actions, and accident-management strategies established a more coherent representation of plant risk.
The use of MELCOR severe-accident simulation results provided a stronger physical basis for evaluating accident-sequence timing and system performance. At the same time, the development of the Level 1–Level 2 interface strengthened the connection between core-damage contributors, Plant Damage States, containment response, and accident progression.
The resulting model provides a more reliable technical foundation for regulatory evaluations, risk-informed applications, aging-management assessments, and future dynamic PRA development.

Tools and Methods

SAPHIRE | Python | MELCOR | Level 1 and Level 2 PRA | Fault Tree Analysis | Event Tree Analysis | Minimal Cut Sets | Plant Damage States | Common-Cause Failure Analysis | SPAR-H Human Reliability Analysis | FLEX Strategies | ATWS | Station Blackout | Risk-Informed Decision-Making | Dynamic PRA

Development and Integration of a Full-Scope Level 2 PRA

National Commission on Nuclear Safety and Safeguards (CNSNS), Mexico City | BWR Mark II | 2026

This project involved the development and completion of a full-scope Level 2 Probabilistic Risk Assessment for a BWR Mark II. Working as part of a small technical team, I assumed primary responsibility for completing and integrating the model, including the containment event trees, supporting fault trees, severe-accident top events, recovery and mitigation functions, SAPHIRE processing logic, and model quality assurance.
The resulting model connects the Plant Damage States obtained from the Level 1 PRA with the potential progression of severe accidents inside the primary containment. It provides a structured representation of containment challenges, mitigation opportunities, vessel and containment response, and the principal phenomena that can influence accident progression.

Selected Technical Contributions

  • Level 2 PRA model architecture. Defined and implemented the overall architecture of the Level 2 model, establishing the logical connections among Plant Damage States, containment event trees, supporting fault trees, system availability and recovery models, accident-progression sequences, and Level 2 end states and associated release categories.
  • Containment event-tree development. Developed the containment event trees required to represent accident sequences associated with the principal Plant Damage States. The event-tree structure was progressively optimized to provide sufficient technical detail while avoiding unnecessary model complexity.
  • Definition of severe-accident top events. Reviewed and optimized the top events used in the containment event trees based on the physical phenomenology of severe accidents. This ensured that the probabilistic structure reflected the relevant progression of the accident rather than relying exclusively on generic or purely logical branching.
  • Pre-vessel-failure recovery modeling. Developed fault-tree models for the recovery of electrical power, restoration of reactor-vessel injection, and recovery of primary-containment spray before vessel failure. These models represented the principal opportunities for preventing or delaying further accident progression.
  • Post-vessel-failure mitigation modeling. Developed additional models for the recovery of core-injection and containment-spray systems after reactor-vessel failure. This allowed the Level 2 PRA to consider mitigation actions that may remain relevant during later stages of a severe accident.
  • Reactor-vessel failure and ex-vessel progression. Developed the supporting fault-tree logic associated with reactor-vessel failure and molten-core concrete interaction. These elements were incorporated into the containment event trees to represent important transitions between in-vessel and ex-vessel accident progression.
  • Steam and hydrogen explosion modeling. Developed probabilistic logic for severe-accident challenges associated with steam explosions in the drywell and wetwell, as well as hydrogen combustion and explosion phenomena within the primary containment.
  • Containment venting model. Developed the fault-tree model for primary-containment venting and incorporated the principal accident phenomena and containment conditions occurring after vent activation. The model considered venting as both an accident-management action and a transition affecting subsequent containment progression.
  • SAPHIRE logic implementation. Developed the process flags, linking rules, and post-processing rules required to connect the containment event trees with their supporting fault trees and Plant Damage State conditions. This logic ensured that each accident sequence activated the appropriate systems, recovery actions, containment conditions, and severe-accident phenomena.
  • Minimal cut set review and model debugging. Performed iterative reviews of the minimal cut sets generated for each containment event tree to verify and refine dependency treatment, event combinations, and processing-rule behavior through repeated model quantification and verification.

Project Impact

The project produced an integrated Level 2 PRA model capable of tracing dominant Level 1 core-damage sequences through their subsequent progression within the primary containment. The model provides a structured representation of recovery actions, vessel failure, containment mitigation systems, severe-accident phenomena, venting, and potential containment challenges. The systematic review of minimal cut sets and SAPHIRE processing logic significantly improved the internal consistency and traceability of the model. Each containment accident sequence could be connected to its initiating Plant Damage State, applicable mitigation functions, physical accident-progression conditions, and corresponding containment outcome. This work strengthened the regulator’s independent analytical capability for reviewing severe-accident risk, evaluating accident-management strategies, and supporting risk-informed regulatory decision-making.

Tools and Methods

SAPHIRE | Level 2 PRA | Containment Event Trees | Fault Tree Analysis | Event Tree Analysis | Plant Damage States | Minimal Cut Sets | Process Flags | Linking Rules | Post-Processing Rules | Severe-Accident Phenomenology | Reactor-Vessel Failure | Molten-Core Concrete Interaction | Steam and Hydrogen Explosions | Containment Venting | Accident-Management Strategies | Risk-Informed Decision-Making

Independent Risk-Informed Regulatory Evaluation

National Commission on Nuclear Safety and Safeguards (CNSNS), Mexico City | BWR Mark II | 2026

Performed an independent regulatory review of a risk-informed application involving surveillance requirements for selected nuclear plant equipment. The assessment evaluated potential effects on component unavailability, initiating-event contributions, Core Damage Frequency (CDF), and Large Early Release Frequency (LERF), including independent verification of probabilistic calculations and consistency with the plant PRA model.

Selected Technical Contributions

  • Independently evaluated changes in Core Damage Frequency (ΔCDF) and Large Early Release Frequency (ΔLERF) associated with modified surveillance intervals.
  • Reviewed and reproduced component unavailability calculations for periodically tested equipment, including latent failure and repair-time contributions.
  • Assessed the treatment of Common Cause Failures (CCF) and the use of alpha-factor models for redundant component groups.
  • Evaluated the representativeness and uncertainty of generic and plant-specific failure-rate data used in the risk model.
  • Reviewed Bayesian updating of generic reliability data using plant-specific operating experience.
  • Analyzed the relationship between surveillance frequency, component failure probability, initiating-event frequency, and conditional core damage probability.
  • Reviewed minimal cut sets and PRA model modifications to identify the risk contributions directly affected by the risk-informed application.
  • Evaluated sensitivity studies and identified key modeling assumptions controlling the calculated change in risk.
  • Developed technical information requests to improve model traceability, parameter justification, uncertainty treatment, and reproducibility of PRA results.

Project Impact

Strengthened the technical basis for regulatory decision-making by identifying the parameters and assumptions controlling the calculated risk change and independently verifying the submitted PRA results. The work demonstrated the ability to move beyond reported risk results and independently assess the underlying probabilistic model, including component reliability, common-cause failure modeling, conditional probabilities, initiating events, uncertainty, sensitivity, and CDF and LERF calculations.

Tools and Methods

CAFTA | Level 1 PRA | Core Damage Frequency (CDF) | Large Early Release Frequency (LERF) | ΔCDF and ΔLERF Evaluation | Fault Tree Analysis | Event Tree Analysis | Minimal Cut Sets | Component Unavailability Modeling | Surveillance Test Interval Analysis | Common Cause Failure Modeling | Alpha-Factor Method | Bayesian Updating | Plant-Specific Reliability Data | Generic Reliability Data | Initiating-Event Analysis | Conditional Core Damage Probability | Sensitivity Analysis | Uncertainty Analysis | Risk-Significant Assumptions | Risk-Informed Decision-Making

ASME ST-LLC Project 200. Risk-Informed Plant Systems Design

ASME Standards Technology, LLC (ASME ST-LLC) | Plant Systems Design Standard | 2023-2025

This project was conducted as part of ASME STLLC Project 200: Target Reliability for Mechanical Fluid Systems, an international initiative supporting the development and technical validation of risk-informed guidance for the ASME Plant Systems Design Standard (PSD-1).The project developed a probabilistic and performance-based framework for incorporating safety, production, reliability, availability, and economic objectives directly into the design of engineered facilities. Unlike conventional PRA applications that evaluate an established design, this approach uses risk information during the conceptual, preliminary, and final design stages. The methodology progressively translates high-level owner objectives into quantitative targets for accident or production-loss sequences, functions, systems, components, and individual failure modes. PRA results, engineering judgment, reliability information, and multi-objective optimization are integrated to compare design alternatives and identify configurations capable of satisfying safety and operational objectives.

Selected Technical Contributions

  • Contributed to the development and refinement of a structured methodology for allocating plant-level risk and performance objectives to sequences, functions, systems, and components during different design stages.
  • Developed and tested Python-based calculation modules and graphical interfaces for sequence-frequency allocation, system unavailability targets, production availability analysis, and component-level reliability and cost optimization.
  • Implemented a proportional risk-allocation methodology to distribute an overall plant risk objective among accident or production-loss sequences while preserving the relative importance of their consequences and preliminary risk estimates.
  • Integrated PRA information, including event trees, fault trees, minimal cut sets, and basic-event probabilities, and SAPHIRE results into the Project 200 optimization framework.
  • Applied the NSGA-III multi-objective optimization algorithm to identify design alternatives that minimize cost and deviation from established unavailability targets. Evaluated and interpreted Pareto-based solutions located above and below the target to support engineering decision-making.
  • Developed the VVER-type pressurized water reactor case study used in the associated research papers. This work included defining safety functions and accident sequences, developing event-tree and fault-tree logic, establishing risk and unavailability targets, identifying influential systems and basic events, and evaluating alternative design configurations.
  • Participated in the conceptual-design safety study for the Mizton multipurpose nuclear microreactor, a fast-spectrum design based on liquid-sodium heat pipes. The work included applying the ASME availability-target approach, reviewing event-tree and preventive-maintenance assumptions, and supporting the initial-design risk evaluation corresponding to Section H.3 of PSD-1.
  • Contributed extensively to the preparation and technical revision of the final project report, software documentation, calculation examples, and proposed guidance for the probabilistic risk evaluation appendix of PSD-1.

Project Impact

ASME PSD-1 provides a structured risk-informed and performance-based framework for using safety, reliability, availability, production, and economic objectives as design inputs rather than evaluating risk only after a design has been established. The standard can be used to allocate plant-level objectives to accident sequences, functions, systems, components, and individual failure modes; establish quantitative reliability and unavailability targets; and compare alternative configurations throughout the conceptual, preliminary, and final design stages.
My contributions helped translate this framework into an implementable engineering workflow through the development of risk-allocation methods, computational modules, multi-objective optimization capabilities, and representative reactor case studies. These contributions make tradeoffs among risk, reliability, availability, performance, and cost more explicit and traceable, supporting the selection of technically feasible design alternatives for both water-cooled and advanced reactors. The work also provided technical evidence and practical application experience to support the continued development and use of the probabilistic risk evaluation guidance in ASME PSD-1.

Tools and Methods

Python | SAPHIRE | Probabilistic Risk Assessment | Event Trees | Fault Trees | Minimal Cut Sets | Reliability and Unavailability Allocation | Risk Target Allocation | Expert Judgment | NSGA-III | Pareto Optimization | Risk-Informed Design | VVER/PWR Safety Analysis | Advanced Microreactor Safety

Scroll to Top