SCK CEN is shaping the impact models of the future through MULTIPLEX and CIThARA
SCK CEN has recently been awarded two new European research projects in the fields of site remediation and radiological protection. As a partner in the European research projects MULTIPLEX and CIThARA, we develop forward-looking models designed to better support both the management of complex environmental sites and nuclear emergency planning. In this manner, SCK CEN not only consolidates its expertise but also tackles socially significant questions.
Research and development with European relevance
SCK CEN’s involvement in two new European research projects strengthens its role in the Pianoforte consortium and advances both its expertise and societal impact. SCK CEN is developing an integrated model to assess environmental impacts on complex sites through MULTIPLEX. Within CIThARA, the group is exploring how AI can improve both the speed and reliability of emergency planning. Both projects combine qualitative data enrichment with smart technologies to enhance knowledge and transparency regarding nuclear activities, for both policymakers and public debate.
As a knowledge and research centre, SCK CEN works daily on improving impact analyses for historically contaminated sites, new nuclear activities, and regions requiring heightened vigilance, such as active nuclear power plants in conflict zones. The knowledge and modelling tools to manage these sites are steadily gaining significance. European legislation now explicitly calls for consideration of environmental impacts and protective measures in environmental permitting, while radiological vigilance has returned to the forefront in several parts of Europe.
MULTIPLEX is developing a model that calculates the interaction between chemical and radioactive substances in the environment
In the MULTIPLEX project, SCK CEN collaborates with a regulator, a knowledge centre and two universities from France and Norway. Each partner investigates the development of a model to predict how freshwater ecosystems on complex local sites respond to radioactive and chemical contamination. This includes sites with previous nuclear activity and sites containing elevated natural radioactive substances.
“The impact of these stressors is often calculated separately, which can lead to an under- or overestimation of environmental risks when remediating a site,” explains researcher Nele Horemans. “During site remediation, more soil is sometimes removed than needed as a precaution.” This also has an impact on local ecosystems. Overlooking the interaction between nuclear and chemical stressors can also result in certain risks being missed. By better assessing the combined environmental effects, we aim to help prevent such unnecessary interventions."
Within MULTIPLEX, the Grote Nete (Belgium), the Rhône (France), and the Nitelva (Norway) serve as case studies. Each of the project partners works on their own case study to enrich existing data with new environmental datasets, and then use it to develop models that take into account various elements and their interactions. In addition, MULTIPLEX partners will actively engage with local residents and stakeholders at complex sites to monitor public opinion and increase transparency around the impact of both nuclear and non-nuclear contamination.
Alongside their individual work, the project partners regularly integrate their models to adjust and refine them. As one of the most experienced partners in predictive model development, SCK CEN takes the lead in this work.
"We are developing the impact models of the future," says Jordi Vives i Batlle. "We translate theoretical and experimental knowledge into a toolbox for policymakers and decision-makers."
CIThARA explores ways in which AI can improve both the speed and quality of emergency responses.
In the CIThARA project, SCK CEN works together with regulators and research and knowledge institutions from France, Germany, Italy, Norway, Poland and Greece. They are working together to examine how AI can make models for nuclear emergency response faster and more reliable.
"In emergency planning, it is crucial to predict how a radioactive cloud will move through the air in the first hours and days," explains researcher Pieter De Meutter. "Existing models must constantly strike a balance between computational speed and accuracy. “Using AI, we could operate up to ten thousand times faster. That could create opportunities to simulate other scenarios and calculate uncertainties. In short: more reliable forecasts, with no wasted time."
Within the partnership, SCK CEN focuses exclusively on developing AI-driven surrogate models, using validated physical models. These can predict the movement of a radioactive cloud based on parameters such as weather conditions, and in this project they are used to train the AI models and test their quality. No laboratory experiments are carried out here, although the project partners do make use of citizen-generated data. “Although citizen-generated data can enrich predictions, the quality of such measurements is difficult to guarantee at present,” says Pieter De Meutter. “With the help of AI, we also hope to improve the accuracy and usability of this data.”
Moreover, surrogate models can not only save time during emergencies but also, in the long term, help detect clandestine nuclear activities, improve emergency planning, and enhance impact analyses of historically contaminated and NORM sites.
Pieter De Meutter: “The aim is not to replace existing emergency plans or impact analyses, but to explore the benefits that surrogate models can offer and how they might be applied.”
