High-Density Gamma Radiation Spectrometry Surveys of Contaminated Land

The design of gamma radiation surveys of contaminated land has always had to compromise between the density of sampling and the accuracy of measurement. Large-area contamination is readily detected by a variety of measurement techniques, while the detection of small areas or ‘particles’ requires a high density of measurement, generally one measurement per square metre. High Resolution Gamma-radiation Spectrometry (HRGS) can provide accurate qualitative (radionuclide identification) and quantitative (Bq.g⁻¹ of the radionuclide for a stated scenario) assessment of the state of the land, but are not cost effective at high density. In contrast, simple walk-over or ‘scan’ surveys using standard Health Physics instruments can provide a high density of measurement, but cannot provide the qualitative and quantitative accuracy of HRGS. In 1996, Nuvia Limited adopted methods to address some of these issues, by allowing scan surveys to include a degree of qualitative analysis of the gamma radiation detected using Low Resolution Gamma-radiation Spectrometry systems with Sodium Iodide detectors, while maintaining a high density of measurement. While low-resolution systems (including medium-resolution Lanthanum Bromide) have become the de-facto standard for large area land surveys, the use of the technology has changed little. High density scan surveys can be conducted using, for example, simple gross-gamma techniques, doserate or region-of-interest logging. However, if spectra are required for qualitative purposes, they are normally collected at fixed locations and for inconveniently long counting periods. The ideal would be to collect spectra at every measurement location, preferably once per second, and then ‘aggregate’ the spectra using spatially-aware techniques. This would allow a scan survey to be performed rapidly, while losing no spectral information. It would thus be possible to analyse the data in a number of ways, for example, producing justifiable doserate measurements or using neural-network techniques to search the measurements for spectral anomalies. This paper describes the work done by Nuvia to develop a system to collect spectra efficiently and to make the spectra readily available for a number of analyses. The efficiency of spectrum acquisition and compression are discussed, along with methods of managing the potentially large volumes of data. The analysis of the data, using a customised Geographical Information System is described, including spatial aggregation of spectra and semi-automatic analysis of spectral structure for identifying common ‘background’ features. Examples of the use of these facilities in Nuvia’s ‘Groundhog’ site survey service are provided.