The Limitations of PEGASOS

Why were the results of PEGASOS not directly implemented?

The seismic hazard analysis of thePEGASOS project was carried out in the years 2000 to 2004 based on the methods and data available at the time and thus represents the level of knowledge at that time. Two key findings resulted from the project: 

  1. The seismic hazard determined in the project was higher than the assumptions that were made during the construction of the present power plants several decades ago.
  2. The methods used reached their limits given the data available in Switzerland. The results contain a large element of uncertainty and there are numerous open questions that are still to be answered.

PEGASOS revealed that the data available in Switzerland at the time were insufficient for achieving the specified goals. Unlike the situation in regions with a high earthquake risk, in Switzerland there are practically no empirical data for the major earthquakes that were being investigated. Models for other areas of the world that were verified using real data could not, at the time, be applied accurately to the tectonic and geological situation in Switzerland.

This applied in particular to modelling the attenuation of the vibrations during their migration from the earthquake hypocentre into the surrounding rock and the site effects at the individual power plant locations (subprojects 2 and 3 of the PEGASOS analysis). Every region really requires its own model, but at the time of the PEGASOS project there was no verified model available that was tailored to the Swiss situation.

In the meantime, scientific knowledge continued to develop (for example with a new generation of attenuation models in the USA and Europe).

International experts generally work in regions with high seismic activity and data available from current and historic events. In Switzerland, with its low seismic activity, it became apparent that the experts were assessing the seismic hazard very conservatively given the lack of available data. This was partly a result of the fact that uncertainties in the individual subprojects flowed more than once into the overall assessment (double counting) and thus distorted the results.

A comparison of the seismic hazard for Switzerland according to PEGASOS with other countries confirmed this finding. 

Intensive discussions lead to a refinement project

The problems associated with the pioneering character of the PEGASOS study led, in 2004 to 2008, to intensive discussions in the technical community. Against this background, the regulatory authority ENSI confirmed in its concluding report of 2004 that PEGASOS had met the highest scientific expectations, but that the range of uncertainty was still large. Added to this was the fact that the authorities then tightened the seismic hazard assumptions for the Swiss nuclear power plants in 2005.

Until a refined study becomes available, earthquake curves based on the hazard curves determined in PEGASOS have to be used for probabilistic safety analyses. Because of the known very large uncertainties and partly conservative assumptions made in PEGASOS, the hazard results can, however, be reduced by 20%.

ENSI drew certain conclusions from the demonstrated uncertainties and limitations of the PEGASOS study. It adopted a cautious approach to reducing the hazard – implementing this on a smaller scale than would have been justified by the most recent analyses from Switzerland and abroad.

Although the PEGASOS results were reduced by 20%, they are still significantly higher than estimates made up to now for Switzerland by earthquake experts. The result was that upgrades and improvements were carried out in the power plants.

In view of the still unsatisfactory situation, the power plant operators decided in 2007 to carry out a study to refine PEGASOS. The PEGASOS Refinement Project (PRP) began in September 2008 with the aim of quantifying the uncertainties more accurately and arriving at a realistic assessment.

Interim analysis following the Fukushima accident

Following the accident in the Fukushima-Daiichi power plant, ENSI called for an immediate review of the safety of the Swiss nuclear power plants in the event of earthquakes and floods. The power plants have since provided the required demonstration (Intermediate Seismic Hazard) and were able to make use of the available quality-assured data and models from the PRP.