To reduce characterisation costs, statistical techniques may be employed where a limited number of samples are taken and inferences are made for the whole area. There are two techniques:
| Technique | When to Use | Description |
|---|---|---|
| Unbiased sampling | Activity expected to be uniformly distributed | Samples taken at random or on a grid pattern across the area |
| Biased sampling | Activity distribution expected to be non-uniform (e.g. hot spots present) | Samples targeted at areas of suspected elevated activity |
In both cases, the level of confidence in sampling and measurements should be specified.
Note: Characterisation activity is much less intense if the facility was shut down in a planned way at the end of its operational life. If the facility was shut down as a result of an accident, characterisation activity is immense and the cost is unpredictable.