Answer to Question #4227 Submitted to "Ask the Experts"

In case of an accident at a nuclear power plant it is necessary to estimate the consequences of released dose. This dose depends on the power at which the plant was operated and the time of operation. To estimate the dose to the population it is necessary to calculate the source term and dose conversion factor. Please tell me the current methodologies used to evaluate this source term, that is,  radionuclide concentration.

Estimating the source term for a release is, as you point out, an essential first step to assessing any exposures. There are several ways that this is currently approached. The "traditional" approach relies on a deterministic analysis of a particular accident path, such as a large pipe break loss-of-coolant accident (LOCA) and some sort of analysis or assumption of the fraction of the core inventory that escapes. These analyses are generally part of the licensing and technical documentation for the power plant, such that they exist already.

A more "modern" approach that is increasingly being used employs probabilistic safety analysis (PSA) to assess all possible accident scenarios and assign a probability to their occurrance. PSA Level-1 analysis assesses the pathways that could lead to core damage. PSA Level-2 assesses, in addition, pathways that could lead to releases to the environment. PSA Level-3 takes the last step and assesses pathways that could then lead to population exposures.

All these techniques now use large computer models, or decision-support systems. Many such systems now exist; for example, the European system called RODOS has been developed over many years by many countries to assess all phases of a nuclear emergency and to provide information supporting decisions regarding any countermeasures that governments may need to implement. Similarly, in the United States, the Department of Energy ARAC (Atmospheric Release Advisory Center) located at Lawrence Livermore National Laboratory also provides a host of analytical services, mostly dealing with atmospheric dispersion after the release, but also modelling source terms.  In fact, there have been incidents when national radiation detection monitoring systems in Europe have detected increased radionuclide content (very slightly increased, but still increased) and these have used atmospheric dispersion models to back-calculate where and how the source might be.

Although this is not a particularly specific answer to your question, if you have any more detailed aspects that you would like to explore, please feel free to submit another question to our Ask the Experts Feature.

Ted Lazo, CHP, PE, PhD
OECD Nuclear Energy Agency