Seismically safe – protecting power plants against earthquakes

5 November 2015 (Last Updated July 28th, 2020 13:36)

Nuclear power plants in the Nordic region could soon be better protected against earthquakes, if a new research project goes to plan. Elly Earls meets one of the project’s key players, Ludovic Fülöp, to find out more.

Seismically safe  – protecting power plants against earthquakes


A new Nordic research project, which will build on four years of previous study on the seismic safety of nuclear power plants, could reduce the risk of earthquakes affecting future nuclear installations in Scandinavia.

Since 2011, nuclear power plants in the Nordic area have given greater priority than ever before to understanding the safety implications of seismic events and how these affect plant design, despite the region's stability and not just because of the Fukushima accident in March of that year.

Yet, with so little empirical information available about seismicity in the region, one of the world's most stable, it's not going to be an easy task for researchers to gain actionable insights from the project.

Key to the programme was the 2011-2014 Finnish National Nuclear Power Plant Safety Research Programme's (SAFIR's) seismic safety research project (SESA).

"Decision makers understand that starting to build new nuclear plants calls for the re-evaluation of the knowledge on all potential hazards, including earthquakes. SAFIR SESA started before the events in Japan, and Fukushima only confirmed its relevance," says Ludovic Fülöp, principal scientist at the VTT Technical Research Centre of Finland, which was a collaborator on the SESA project and is also one of the partners working on a new four-year project on the same topic, funded partly by NKS, a platform for Nordic cooperation and competence in nuclear safety, and partly by SAFIR 2018, which follows on from SAFIR 2014.

The new project will build on the progress that's already been made through SESA, as well as other projects related to the country's new planned power plants. Planned installations include Olkiluoto 3 and Fennovoima Hanhikivi 1, both due to launch in 2018.

Is the government right to pursue Britain's first nuclear plant in 20 years

"A large database for all relevant earthquake recordings from Finland and Sweden has already been compiled and used to develop understanding of how vibration decays, or attenuates, with distance in Fennoscandian conditions," Fülöp notes.

"Moreover, using de-aggregation, a kind of decomposition of the sources of seismic hazards, we have also already been able to better localise the sources of potential design safety issues for power plants. And finally, we have tried to get some answers concerning the dangers of high-frequency shaking propagating to plant floors through very hard rock. This may have adverse effects on equipment located on those floors."

Modelling techniques: essential but early days

Yet, a lot more research still needs to be done to really understand how nuclear power plant design could be improved to take potential seismic events into account - largely because data in this area is pretty much non-existent.

"There is no locally measured empirical data on earthquakes with direct significance to design or plant safety," Fülöp remarks. "The measured data comes from earthquakes of too low magnitude, or the measurements are at distances too great from the source."

Modelling techniques are therefore crucial to even begin to understand the potential impact larger earthquakes could have on power plants in the region, something VTT started working on back in May when it ran a workshop aiming to identify and share the source modelling-based measurement data available on earthquakes in stable continental areas.

According to Fülöp, it's still very early days, though. "The workshop was a launching platform for the modelling activity," he explains, adding that modelling in relation to seismic events is always a 'sensitive topic'. "Uncertainties are significant. There is little known about many of the parameters playing key roles in the models. For instance an input is the size and type of earthquake faults, a topic with very little understanding in 'diffuse seismicity' regions [regions where there are no recognisable seismogenic features to which earthquakes can be associated]."

New research project, new modelling techniques

So what's the next step for research in this area? In a nutshell, the new SAFIR 2018 and NKS-funded four-year project, which will update existing earthquake source modelling techniques and develop new ground motion simulation models for stable continental regions.

"We hope to develop numerical models, which will result in credible vibration scenarios for larger magnitudes and lower source distances. These would be relevant for power plant design," Fülöp explains, noting that the other key partners working on the project are Aalto University, the University of Helsinki, Uppsala University, the Geological Survey of Denmark and Nordic technical consultancy ÅF-Consult Oy.

The team will start by looking at modelling work carried out at Uppsala University for nuclear waste depositories, after which they will collect calibration data from real events, and then create very simple models, which can be compared with the calibration data.

"The models can become complicated very fast, so we hope to capture some main features just through simple models," Fülöp says.

"The models can become complicated very fast, so we hope to capture some main features just through simple models."

"Then we will start to modify the input parameters within credible boundaries, and see how much the models are sensitive to inputs."
It's not going to be an easy road, however. "There are major challenges, starting with ensuring enough finances to be able to fully focus on this topic," Fülöp explains, adding that the aim is to have at least one person dedicated to the modelling effort, with some resources allocated to guidance and supervising.

"The existing knowledge barriers and how to overcome them is also a significant worry. We need to be efficient in concentrating on the most significant areas, going through the steps that are well understood as fast as possible. But, the main research worry is that we may find the models too sensitive to input parameters, making it impossible to use the results."

Cooperation will be absolutely crucial to success. "There is a lot we have to agree on as a group. As explained earlier, there is often none or very little data, and in these cases finding credible numbers relies on expert judgements, debating competing opinions and settling on some kind of consensus," Fülöp notes.

Widespread interest from all angles

The project has already garnered plenty of interest from various industry stakeholders, with both regulators and commercial companies getting on board. "Both the regulators from Finland and Sweden are supporting the work, while operators and equipment suppliers are also interested since they can estimate the safety margins of their engineering solutions," says Fülöp.

"As all findings are public, they can also anticipate the major research directions and explore them in the frame of their own products or designs, to see how they compare to what may become tomorrow's regulatory demand."

Ultimately, the researchers hope to expand the project to cover all Baltic countries. "Currently, we work under the SAFIR 2018 (a four-year extension of SAFIR 2014) and NKS frameworks, which are Finnish and Nordic projects. The plan for the Baltic countries is a longer-term goal," he explains. "The soil conditions in these countries are different, having to deal with sediments and soil-structure interaction.

But we do plan to propose a European cooperation mechanism in the future for integrating these countries into the broader 'diffuse seismicity' topic."