Anthony Scopatz

Kathryn D. Huff, Matthew J. Gidden, Robert W. Carlsen et al.

As nuclear power expands, technical, economic, political, and environmental analyses of nuclear fuel cycles by simulators increase in importance. To date, however, current tools are often fleet-based rather than discrete and restrictively licensed rather than open source. Each of these choices presents a challenge to modeling fidelity, generality, efficiency, robustness, and scientific transparency. The Cyclus nuclear fuel cycle simulator framework and its modeling ecosystem incorporate modern insights from simulation science and software architecture to solve these problems so that challenges in nuclear fuel cycle analysis can be better addressed. A summary of the Cyclus fuel cycle simulator framework and its modeling ecosystem are presented. Additionally, the implementation of each is discussed in the context of motivating challenges in nuclear fuel cycle simulation. Finally, the current capabilities of Cyclus are demonstrated for both open and closed fuel cycles.

Anthony Scopatz

A method for quickly determining deployment schedules that meet a given fuel cycle demand is presented here. This algorithm is fast enough to perform in situ within low-fidelity fuel cycle simulators. It uses Gaussian process regression models to predict the production curve as a function of time and the number of deployed facilities. Each of these predictions is measured against the demand curve using the dynamic time warping distance. The minimum distance deployment schedule is evaluated in a full fuel cycle simulation, whose generated production curve then informs the model on the next optimization iteration. The method converges within five to ten iterations to a distance that is less than one percent of the total deployable production. A representative once-through fuel cycle is used to demonstrate the methodology for reactor deployment.

Anthony M. Scopatz, Matthew J. Gidden, Robert W. Carlsen et al.

The current state of nuclear fuel cycle simulation exists in highly customized form. Satisfying a wide range of users requires model modularity within such a tool. Cyclus is a fuel cycle simulator specifically designed to combat the lack of adaptability of previous generations of simulators. This is accomplished through an agent-based infrastructure and treating time discretely. The Cyclus kernel was developed to allow for models, called archetypes, of differing fidelity and function depending on need of the users. To take advantage of this flexibility, a user must write an archetype for their desired simulation if it does not yet exist within the Cyclus ecosystem. At this stage, a user graduates to the title of archetype developer. Without automation, archetype development is difficult for the uninitiated. This paper presents the framework developed for simplifying the writing of archetypes: the Cyclus preprocessor, or cycpp. cycpp addresses the computer science and software development aspects of archetype development that can be addressed algorithmically, allowing the developer to focus on modeling the physics, social policies, and economics. cycpp passes through the code three times to perform the following tasks: normalizing the code via the C preprocessor, accumulation of notations, and code generation. Not only does this reduce the amount of code a developer must write by approximately an order of magnitude, but the archetypes are automatically validated.