By the numbers

Framatome recognises that the need to load follow intermittent renewables like wind and solar presents a growing technical challenge for the current and future nuclear fleet. As renewables are ‘must take’ in many markets, nuclear plants must respond to renewable production by starting up or shutting down, ramping up or down. If they are operating inflexibly, at baseload, they must pay to produce power at times when prices are negative.

In France, where electricity is mainly supplied by nuclear plants, and in other European countries that have been facing intermittency more recently, Framatome has learned how to manage the need for flexibility for various reactor and fuel types. The company now has significant experience designing the right fuel and employing the associated codes and methods to maximise flexibility of operation, taking into account the configuration of the reactor systems and the nature of the core components.

Flexibility in operations is a clear requirement when developing fuel technologies. For example, Framatome’s ‘PROtect’ enhanced accident-tolerant fuel (EATF) uses M5Framatome advanced cladding and chromia-doped UO2 pellets to increase margins during power level variations and allow for more frequent and longer power level reductions. These technologies also contribute to high safety margins and better fuel management. In the short term, the EATF design will further increase safety margins and operational benefits.

Framatome has also developed a new suite of codes and methods using state-of-the-art physics that improve modelling accuracy. This allows for more flexible core designs, optimised fuel management and safe operation, which improves fuel cycle economics. In many cases, plants relied on codes and methods developed many decades ago that were constrained by the computing power of that time. New levels of computing capability have enabled them to be re-examined and replaced. New codes offer a more accurate prediction of the capabilities and behaviour of reactors, and new methods allow for removal of overly conservative assumptions to gain margin.

The complete suite of enhanced codes, less penalising methods and higher computing power alongside advanced fuel designs have been critical in allowing plants to operate effectively in changing environments. For example, the ability to make accurate predictions of margins during reactor startup, the onset of reactor coastdown at the end of the cycle or the margins available in abnormal configurations can significantly improve operators’ ability to fulfil power generation targets.

Codes and methods in core design

New codes and methods needed to be developed around the problem of rod ejection accidents in PWRs — a historically limiting safety analysis for many plant types required to start flexible operations. This type of accident is a result of mechanical failure of the control rod drive mechanism housing, which can result in ejection of the attached control rod drive assembly and a power surge in the core.

Framatome initially began work on an advanced methodology known as AREA, utilising the Arcadia code suite to address this problem as a part of the mixed-oxide (MOX) fuel project in 2009. Work began in earnest in 2013. This development was motivated by completion of the Arcadia topical report and expected changes in the US Nuclear Regulatory Commission’s (NRC’s) rod ejection evaluation criteria.

In order to develop the AREA methodology, the Arcadia code suite was benchmarked against industry standards for rod ejection analysis. As part of this development, the code suite was coupled to the S-Relap5 system simulation code to give it the capability to accurately and dynamically model reactor core and system transient behaviours. Legacy PWR rod ejection accident codes were developed in the 1970s; the AREA method benefits from four decades of neutronic modelling and rod ejection analysis experience, so it can predict transient behaviour more accurately.

The topical report on AREA was submitted to the NRC for review in October 2015. NRC approval was received in December 2017 for all operating PWR plant types in the USA.

The AREA method was successful in an extensive set of sample problems for the topical report as well as during its recent first application to a US reactor. It showed the conservative nature of past rod ejection analysis methods. In addition to addressing updated analysis criteria, the advanced modelling and analysis capabilities demonstrate a significant reduction in the effects of such an event and present large margins to criteria. In most cases this advanced method, in conjunction with fuel hardware technologies, removes the rod ejection accident as a limiting factor in core design optimisation.

While the AREA method is an excellent example of applying the Arcadia framework, this advanced code suite can be used to cover a range of engineering activities from fuel assembly design to safety analysis. Framatome’s experts continue to use the suite to maximise plant margins, provide core design flexibility and allow for fuel cycle cost reductions.

When developing the AREA methodology, Framatome’s team pushed the methodology and codes to the limit of the computing power at the time, generating tens of gigabytes of data for a single rod ejection run and requiring a large number of processors in a high-performance computing system. Even with all this power, it still takes several days to perform just the computer runs necessary for an AREA analysis. However, the time required is decreasing as the computational infrastructure develops. As computers improve, the nuclear industry will be able to continue to take on the computationally intensive development of new codes and methods.

Enhancing flexibility with core monitoring

Another codes and methods option within the Arcadia framework, Framatome’s Argos core monitoring system, is designed to help operators run plants in various flexible operation modes.

An international team of nuclear physicists, nuclear engineers and software experts from five countries designed Argos to offer highly accurate power distribution monitoring and technical specification surveillance for all types of commercial light water reactors. Built as a modular system, Argos provides a general analysis of steady-state reactor periods and transient events, as well as a prediction module to plan and assist in the execution of projected load-following (see Figure 1&2).

When developing Argos, the team sought to use modern, open software standards. The graphical user interface is programmed in C++. It interacts with a Qt library, an open-source and cross-platform widget toolkit used to design graphical user interfaces, and data processing engine modules written in Python. The HDF5 format was chosen for binary files, and internal and external interface files are all in XML. The data backbone of the system is a PostgreSQL database.

The Framatome team worked closely with the team at a Siemens KWU three-loop PWR in Switzerland in the first implementation of the Argos system worldwide (see Figures 3&4). In this plant, Argos replaces Framatome’s Powertrax/S, a choice that was driven by Argos’s improved flexibility and configurability.

As of June 2019, after more than one year of successful parallel running of Argos together with the previous core monitoring system, Argos is now fully operational at the plant. During the implementation, Framatome’s experts had to make adaptations to stay on schedule while addressing the challenges inherent in any large-scale integration project.

The successful first implementation of Argos will serve as a blueprint for other Siemens KWU plants and will facilitate rollout for other plant types. The Framatome team will work on deploying it in other PWR designs, mainly with 17x17 geometry. The team will rely temporarily on the legacy 3D core simulator Prism as the nuclear engine of Argos, later replacing it with Framatome’s 3D neutronics and thermal-hydraulics reference code Artemis as a part of the company’s Arcadia suite.

Looking forward

While many uncertainties persist about the future of our energy mix and the composition of the grid, innovation only bolsters nuclear energy’s capability to meet the clean air and energy needs of the world into the future. In the numerical simulation domain, Framatome, together with EDF, has already launched the codevelopment of the next generation of code, Odysee, which will bring solutions to the next set of challenges faced by sustainable, safe and competitive nuclear energy generation.