Waste Assay & Characterisation

A large part of the past decade has seen my involvement in specification and design of assay systems for radioactive waste. When in a cynical mood, which happens occasionally, I have been known to say that assaying the waste is the easy bit. Usually with gamma assay systems no more than two isotopes are measured and the results used to estimate the inventory of a much larger number of isotopes. (I accept that neutron systems are more technically difficult, but the following still broadly applies). The challenge is in understanding what the results mean, interpreting them in a form useful to the customer, and estimating the uncertainties. Most waste assay systems do not need to be accurate unless operating near a legislative or safety margin. There are usually many factors outside the control of the operator, such as lack of knowledge of the waste history, or the applicable nuclide vector, which dominate the overall uncertainty of the result.

The challenge is often not so much in estimating activities, as in making a credible and supportable estimate of the uncertainty in the activities. This comes from a detailed analysis of all of the sources of uncertainty, and combining them properly to give a global estimate. A tool which I have used a lot is Monte-Carlo modelling of uncertainties, in particular in analysis of configurations of active material in drums, boxes or trays of waste. An example of this was a recent set of measurements which I completed on the first waste to be retrieved from the Berkeley Active Waste Vaults. A Monte-Carlo analysis of possible configurations of waste in the drum being assayed gave both bounding values of activity as well as a conservative estimate of 1-sigma uncertainties.

Further Monte-Carlo modelling can help in assessing the uncertainty in waste fingerprints / nuclide vectors - by, for example, modelling distributions of decay times or activation conditions (neutron flux, irradiation duration, precursor impurity levels etc).

Assay techniques which I have used include gross gamma measurements (using dose probes) - useful when only a single nuclide dominates the gamma field; low-resolution gamma ray spectrometry, and high-resolution gamma ray spectrometry. The latter, although more complex and expensive, usually rapidly repays the investment in terms of its ability to detect low levels of activity against high levels of interfering nuclides, and in giving unequivocal identification of nuclides. All of these methods have been used for both installed assay systems on retrieval / processing plant,as well as for ad-hoc assays of drums or items using portable in-situ  measurements.