Digitalising maintenance and planning of unit outages to optimise the performance of the nuclear fleet in operation

Christelle DELEUIL

Digital Performance and Innovation Manager

After graduating from INSA Val de Loire as an engineer specialising in industrial and environmental risk management, Christelle began her career working in nuclear power plants, where she managed large-scale projects in the fields of logistics, waste management and dismantling of Graphite Gaz facilities. In 2015, Christelle joined Assystem and moved into the fields of owner engineering. She supervised the teams in the Centre and Aquitaine regions while working on business development and project delivery for the French existing fleet. Since 2020, Christelle has worked as Digital Performance and Innovation Manager and takes charge of the digitalisation of practices, the improvement of processes and the valorisation of data, making the bridge between Assystem’s digital and operational teams for the Installed Base.

The fight against climate change requires a global transition to make safe and sustainable energy accessible to all populations. The use of nuclear power is an appropriate response to these challenges, enabling the production of affordable, controllable, low-carbon electricity, which also promotes energy independence. In order to achieve a sufficient, safe and sustainable level of nuclear power generation in the coming decades, EDF has been undertaking major programmes in France for several years to maintain and improve the reliability of its facilities, known as the ‘Grand Carénage’ programme. To guarantee its success, EDF must ensure that the interventions are technically mastered, and that deadlines and financial budgets are met. In this context, the digitalisation of processes to optimise the performance of industrial maintenance has quickly become one of the major areas of focus for the operator.

As EDF's trusted partner for existing fleet projects, Assystem is also convinced that the operating conditions of power plants can be improved by taking advantage of innovation and the latest digital technologies. These technologies help reduce operating charges in a context where safety requirements are becoming increasingly stringent. This is particularly the case for the planning of maintenance shutdowns, which are increasingly complex, and where planning software coupled with algorithms for processing big data helps optimise outage times and thus guarantee the availability of the electricity production tool as soon as possible.

To go further, other aspects of the maintenance operations ecosystem, such as project management, resources and methods, are also integrated into this digitalisation approach. The whole therefore contributes to the operator's challenges in terms of plant maintainability and thus to extending the lifespan of the facilities under optimal conditions of performance, safety and improvement of business processes.

Optimisation of multi-annual maintenance planning: the Solvertools use case

Ensuring the maintenance of complex industrial infrastructures requires precise planning of every aspect of work and activities, while coordinating the intervention of multiple stakeholders and skills; it is also essential to take into account the constraints associated with the equipment, particularly in terms of quantity, availability and budget. These forecasts must be anticipated over long time scales in order to comply with the strategy of production and operational safety.

The multi-annual planning of maintenance work on the French nuclear power plant fleet is a perfect illustration of this: coactivity, safety/security constraints and availability of resources, are among the many factors to be taken into account. The complexity is further increased by the number of units to be planned in parallel and the need to consolidate a national approach for all the work. The use of digital tools based on the latest Artificial Intelligence and simulation technologies is therefore an essential solution for the effective management of these thousands of tasks and the numerous constraints and recurring updates to the associated schedules:

A power plant shutdown project can involve up to 15,000 different operations in a confined space and a limited time, while some areas such as the Reactor Building are only accessible once the plant has been shut down for maintenance

says Christelle Deleuil, New Services and Innovation Manager at Assystem.

In order to respond effectively to these issues, Assystem has supported EDF in the deployment of the SolverTools solution. SolverTools is a decision support application solution that can be interfaced with EDF's EAM (Enterprise Asset Management) and multi-year planning software. As a planning simulation engine, SolverTools enables:

Optimal planning of shutdowns: scheduling of major maintenance work and modifications to installations during shutdowns, taking into account the operational constraints and the packing of the worksites;
Optimal dimensioning of outdoor worksites for operating units with a global vision of the site.

Thanks to dynamic schedule simulations, it very quickly generates ‘feasible’ solutions, while resolving conflicts by staggering activities, prioritising high-priority tasks.

Developed since 2019 by Assystem in co-investment with the Tricastin Nuclear Power Plant, the use of the tool has enabled the teams of the Multi-Annual Plan to improve their operational performance, facilitating trade-offs or iterations within short deadlines; it also represents a robust means of better detecting conflicts, but also of capitalising on practices and knowledge about the constraints of unit shutdowns between sites or with engineering centres

explains Christelle Deleuil, who has been working in the deployment of the solution since 2019.

Modelling for on-site collision detection

The Tricastin Nuclear Power Plant was not only the place where Solvertools was created. The plant, which has been a partner in numerous innovative projects to improve the performance of power facilities, has also worked with Assystem to set up a 4D planning visualisation tool (addition of the time dimension to the 3D model to determine ‘who does what, when and where?’ at any point in the project).

For outdoor work on buildings, the initial approaches made it possible to define scripts and methodologies adapted to dialogue with EDF's planning software. The aim was to first prove the value of the approach, then to create a dynamic of exchange and real synergy between the different business and project teams. This digital approach improves data sharing and facilitates decision-making thanks to better synchronisation between stakeholders. It is therefore perfectly suited to optimising the planning of complex projects by enabling the modelling and improvement of the phasing of operations.

As part of the multi-annual maintenance, the 4D visualisation of the schedules is intended to ensure secure planning in terms of activity scheduling and workflow.

In an environment where so many stakeholders communicate and collaborate together, this 4D approach makes it possible to visualise current and future activities in concrete and visual way, to better identify interfaces in the delivery process, or to plan the flow of vehicles and people.

completes Christelle Deleuil.

To date, 4D visualisation has been developed by Assystem in collaboration with EDF at the Tricastin nuclear power plant and then deployed at other 900MW nuclear power plants.

The benefits that our customers achieve from 4D visualisation clearly demonstrate that digitalisation makes it possible to decompartmentalise data, to link it to modelled processes and to optimise the synergies between digital and business lines, with the aim of mastering the complexity of projects, improving their performance and promoting collaboration between stakeholders.

Management, analysis and quality of data: the use case of spare parts process improvement

In the industrial sector, efficient data management and in-depth data analysis are powerful levers for accelerating project delivery and promoting innovation. Faced with a growing volume and diversity of data, industrial data science makes it possible to:

Increase productivity through the automation of tasks and business processes,
Improve the quality of planning, operation and maintenance,
Facilitate project management through the precise monitoring of performance indicators, thus contributing to informed decision-making,
Reduce project risks and make the most of feedback (deadlines, earned value, risk management).

The volume of data generated by the multi-annual programme of EDF DPN (Nuclear Production Division) obviously makes rigorous data management crucial: proper control of information not only makes it possible to anticipate problems, but also to optimise the organisation of maintenance operations, thus reducing downtime and repair costs.

Reaffirmed in the objectives of its 2025 strategy, one of EDF's priorities is to guarantee that the level of production of the fleet is maintained under optimal safety conditions. This involves industrialisation and therefore making maintenance operations more reliable by drawing on feedback. Industrial performance then means having and capitalising on comprehensive, reliable and regularly updated maintenance databases. This data must also be secured, so that it is only accessible to authorised stakeholders, guaranteeing its confidentiality and integrity.

explains Christelle Deleuil.

Assystem supported EDF on one of their strategic projects for optimising the performance of maintenance operations : the project was dedicated to improving the spare parts circuit, which aims to ensure the proper connection of data between industrial models (a set of articles or spare parts constituting a piece of equipment) and the equipment installed on site.

During maintenance operations, in order to obtain the right spare part, in the right place, at the right time, the databases listing the equipment installed on site were required to be exhaustive and reliable. Assystem has made this task easier by implementing an ‘intelligent extraction of targeted data from different information system databases’, enhanced by algorithms that can predict the best possible connection between a piece of equipment and its industrial model. By exploiting the information contained in EDF's tools, Assystem's business and digital experts have been able to develop a decision-making tool that:

Proposes relevant industrial models based on the data sources available for each piece of equipment,
Argues these proposals with details of their relevance scores to inform the choice of technical preparers,
Identifies anomalies in the data already available, to create a reliable database that shows which equipment model is actually installed on a tranche.

For the two EDF nuclear power plants where we have implemented this methodology, the teams have been able to increase their efficiency thanks to a tool-based approach, based on well-argued proposals rather than laborious manual research. This predictive tool, based on advanced data analysis, therefore contributes to better forecasting and improving the quality of spare parts orders, as well as optimising the associated costs.

Several potential areas of work have been identified so far to industrialise the use of data modelling for the spare parts supply chain. This work is also completed by modelling work on the multi-annual process and by reflections on the feasibility of a workload/resource adequacy assessment system (anticipating the volume, type and placement of maintenance activities by suppliers over several years).

In conclusion, the digitalisation of processes associated with industrial maintenance operations improves the performance, durability and safety of industrial infrastructures, particularly in critical sectors such as nuclear power. By using simulation, modelling tools and advanced technologies, it makes it possible to better anticipate and manage the complex challenges of infrastructure maintenance while optimising technical interventions and the ‘Quality-Cost-Deadline’ triptych. It therefore helps to maintain the high level of performance of the facilities while supporting EDF's energy transition ambitions, and to extend the lifespan of these critical assets under optimal conditions.

A question, a project?