CNS Annual Conference 2026

The Electrics Power and Research Institute (EPRI) is modernizing the MAAP computer program by converting the code from FORTRAN to C++ and by consolidating the PWR, BWR, CANDU and VVER versions into a unified, maintainable platform. For MAAP-CANDU this modernization also includes the development of a new dynamic linked library (DLL) named CHANNELS v1.0, which contains all physics related to the heatup, disassembly and eventual collapse of a series of horizontal fuel channels under severe accident conditions.

This portion of MAAP-CANDU has been maintained by Kinectrics on behalf of CONEXUS. This paper will outline the steps taken to qualify CHANNELS v1.0 in accordance with CSA N286.7-16. A comprehensive CHANNELS regression test suite was prepared with near-full code coverage to assist with the verification process and to support future code maintenance. Verification steps include verification of the initial FORTRAN to C++ conversion, verification of the added fuel quench functionality and verification of a number of code error fixes. Following verification the DLL will be released to the MAAP-CANDU community for beta testing and further feedback prior to full integration into the modernized MAAP platform.

Presenter:
John Kennedy, Principle Analyst, Nuclear Safety & Licensing, Kinectrics
June 15, 2026, 1:50PM to 2:10PM | Technical Room - Mayfair

A major challenge in new reactor design is assessing the behaviour of irradiated fuel bundles before the reactor is operational. This challenge arose during the development of the ACR-1000 reactor. To compensate for the absence of irradiated bundles, a campaign was launched to fabricate a mixed-oxide (MOX) fuel bundle with neutronic characteristics approximating those of an irradiated bundle at a specified burnup – under both full-power ACR-1000 conditions and ZED-2 experimental conditions. Historically, determining the MOX composition required extensive manual effort. This work proposes an optimization methodology to automate this process. The approach employs the Ant Colony Optimization (ACO) algorithm as the optimization engine, while perturbation ranges are derived from a companion study using Latin Hypercube Sampling (LHS) to define the parameter space. This methodology offers a systematic and efficient alternative to manual design, enabling accurate replication of irradiated bundle behaviour for testing and validation purposes.

Presenter:
Doddy Kastanya, Principal Engineer, Nuclear Safety & Licensing, Kinectrics
June 16, 2026, 11:30AM to 11:50AM | Technical Room - Nakiska

One of the key challenges in designing a new reactor is testing the behavior of irradiated fuel bundles before the reactor itself exists. This issue arose during the development of the ACR-1000 reactor several decades ago. To address the absence of irradiated bundles, a campaign was initiated to create a mixed-oxide (MOX) fuel bundle with neutronic characteristics closely matching those of an irradiated bundle at a specific burnup level—both under full-power ACR-1000 operating conditions and under ZED-2 conditions where various tests were planned. The MOX bundle composition was determined largely through manual effort. Drawing on that experience and anticipating similar requirements for next-generation CANDU reactors, the development of an optimization tool for this purpose is now being explored. The optimization process must account for multiple independent parameters. This paper presents a methodology, based on Latin Hypercube Sampling, to define the appropriate parameter ranges for use in the optimization.

Author:
Doddy Kastanya, Principal Engineer, Nuclear Safety & Licensing, Kinectrics
June 17, 2026, 11:10AM to 11:30AM | Technical Room - Bow Valley

Elementary nuclear engineering textbooks typically introduce extrapolation distance as an approximate function of the material through which neutrons exit the reactor core, making it proportional to the transport mean free path. However, in practical reactor analyses, the structure at the core boundary (such as the end of a fuel assembly in light water reactors or a fuel channel in pressurized heavy water reactors) rarely consists of a single material. Variations in material composition at these boundaries can significantly influence the extrapolation distance. This study employs the MCNP code to develop a methodology for determining axial extrapolation distance. The findings serve as a foundation for future work, which will incorporate variations in neighboring channels and establish correlations with key factors such as the number of fuel bundles, fuel burnup, and power distribution. Sensitivity analyses confirm the robustness of the derived axial extrapolation distance, which ranges from 4.576 cm to 4.734 cm across tested scenarios. These results represent an initial step toward creating a fuel-channel-dependent extrapolation distance function for integration into full-core CANDU reactor physics models.

Presenter:
Doddy Kastanya, Principal Engineer, Nuclear Safety & Licensing, Kinectrics
June 17, 2026, 10:50AM to 11:10AM | Technical Room - Mayfair

The CANDU (Canada Deuterium Uranium) reactor is equipped with multiple, robust safety systems designed to prevent accidents and mitigate their consequences. Among the three fundamental safety functions, controlling reactor reactivity is critical. Reactivity control involves managing nuclear chain reactions within the core, including increasing, decreasing, or completely terminating the reaction as required. During normal operation, reactor power is regulated by adjusting reactivity control devices located in the core. These adjustments are based on feedback from an array of sensitive detectors strategically distributed throughout the reactor. To ensure reliability, detector responses must be verified periodically, beginning at installation. Given the large number of detectors involved, reducing the time required for these checks is essential for operational efficiency. This paper presents a methodology to optimize the detector installation verification process through pre-simulation of planned activities. The approach utilizes a variant of the Ant Colony System (ACS) algorithm as an optimization engine to determine a quasi-optimal set of rods for commissioning. Compared to conventional methods, where the best rod is selected for each detector individually, the proposed technique achieves more than an 80% reduction in testing duration. This approach offers a significant improvement to reactor commissioning procedures.

Presenter:
Doddy Kastanya, Principal Engineer, Nuclear Safety & Licensing, Kinectrics
June 17, 2026, 10:30AM to 12:00PM | Technical Room - Mayfair

Canada currently operates 17 reactors with a producing capacity of 12,714 MWe, representing ~15% of the country’s energy consumption. The changing climate poses potential threats to existing and planned infrastructure. Canada does not currently have tailored industry standards for assessing climate change effects on NPPs. Current design parameters for existing and planned infrastructure might require revision to consider local climate change and maintain suitability. Canadian stakeholders are interested in understanding the implications of climate change on plant design and operation. This additional consideration is an important aspect of a broader framework to ensure public safety and prevent reductions in the Country’s capacity to meet energy demands. Kinectrics worked with Utilities across Canada to assess existing and planned infrastructure for vulnerabilities to climate change, guided by INPO and EPRI publications. Climate data, including both projections and bounding conditions, were considered in the evaluation of nuclear infrastructure to determine potential risks to continuity of operations and safety. As such, Kinectrics applied climate change vulnerability assessment methodologies which project climate impacts and make recommendations to mitigate vulnerabilities using existing plant procedures and programs. This paper discusses the lessons learned of applying these methodologies in Canada.

Presenter:
Melissa Diamantakos, Assistant Analyst, Kinectrics
June 16, 2026, 9:10AM to 9:30AM | Technical Room - Bow Valley

Presenter:
Larkin Mosscrop, Manager, Environment and Community, Kinectrics
June 16, 2026, 10:50AM to 11:10AM | Technical Room - Bow Valley