(268d) Digital Network Twin of the UMass Lowell Research Reactor Facility

The on-campus UMass Lowell research reactor (UMLRR) is a light-water moderated and cooled, graphite-reflected, heterogeneous, open-pool type research reactor. There is great potential for the UMLRR facility to be used as a test bed for wireless control systems. The facility is fully functional (with availability of current and historical data), well preserved, and consists of key components that can be used for data generation of prospective wireless control components. This work reports on first steps towards developing a digital twin of key components of the facility as described next.

We adopt a divide-and-conquer approach for modeling and simulation of the UMLRR facility wherein significant sub-systems are modeled individually while coupled via an external framework. The computational modules (nodes) in the network are built from stand-alone, reduced order models of the sub-system of interest. Some sub-systems appear repeatedly, either physically or logically, for instance, holdup and sump tanks consist essentially of the same unit operation, therefore a single class module can be programmed for modeling the necessary physicochemical phenomena; similarly the same approach holds for pumps and controllers. In both cases of physical and logical sub-systems, a network can be built for all the components, and a dynamic model assembled to represent a digital network twin of the physical system. A key capability requirement of the framework is to allow fo scalability and re-usability of computational modules. Given the status of modern computing power, extended networks can now be created to various levels of detail; for instance, valves and pipes can be included explicitly in the network. An extended network can be mapped on large scale parallel computing platforms to allow for increasingly modeling fidelity and faster simulation of the evolution of the dynamics of the network.

1) V. F. de Almeida et al., “Cortix: Network Dynamics Simulation,” http://cortix.org/ (2018), download at http://cortix.org.
2) T. Azzaoui, K. L. Lyon, and V. F. de Almeida, “Design and Implementation of a Flowsheet Extraction Simulator,” Trans. Am. Nuc. Soc., 120, 106–9 (2019).
3) V. F. de Almeida, T. Azzaoui, and K. L. Lyon, “Solvent Extraction Flowsheet-Level Modeling and Simulation with Radiolysis,” Proceedings of the Global 2019 International Nuclear Fuel Cycle Conference, ANS, 386–9 (2019).
4) A. Rotker and V. F. de Almeida, “Network Dynamics Simulation of a BWR Plant,” Trans. Am. Nuc. Soc., 124, 211–4 (2021).
5) N Kemper, J. Kennedy, and V. F. de Almeida, “Dynamically Coupled Extended Balance of Plant for an SMR,” Trans. Am. Nuc. Soc., 124, 215–7 (2021).
6) V. F. de Almeida, “Plant-Level Components for Nuclear Fuel Recycling: Commonality,” Letter Report ORNL/LTR-2011/176, Oak Ridge National Laboratory, Tennessee, TN 37831-6181, U.S.A. (Jul. 2011)