Methodology for Hydrogeochemical Sampling to Characterise Groundwaters in Crystalline Bedrock: Developments Made within the Swedish Radwaste Programme

The search by SKB (Swedish Nuclear Fuel and Waste Management Co.) for a site to locate the deep geological repository for spent nuclear fuel in Sweden has involved geoscientific investigations at several locations since the 1970s. The objectives were to characterise geologically a bedrock volume as well as its hydrogeology and hydrochemistry. To acquire high-quality hydrogeochemical data, a complete system for groundwater sampling and analysis, as well as for interpretation strategies, has been developed through a continuous process of modification and refinement. Since the largest part of the Swedish bedrock is composed of granitoids, the site investigations had to adapt to the special difficulties of fractured crystalline rocks. This paper discusses the problems with groundwater sampling that are specific to fractured crystalline rocks and describes the solutions adopted and methods developed by SKB since the early 2000s during the site investigations. The methodology described in this paper for the characterisation of deep groundwaters in crystalline rocks is not only applicable in the context of radioactive waste disposal but also useful when sampling groundwaters for any purpose in such rocks. Sampling of groundwaters in fractured rocks at depth, often down to approximately 1,000 m, involves special challenges since the natural conditions of the groundwater are easily disturbed, especially by the initial drilling, but also by every subsequent activity performed in the borehole, including the actual groundwater sampling. The sampling strategy presented in this paper shows that planning of the sampling preferably starts already when the drilling procedure is decided. Each following step is described in detail and includes tracing the drilling fluid, selecting the best borehole sections to sample, procedures for the actual sampling, and selection of analytical protocol; all this with the goal of taking representative samples. Although the evaluation of the sampling uncertainties is not a straightforward procedure, an adequate categorisation routine has been established to classify groundwater samples regarding sample quality, representativeness, and suitability for further interpretations and modelling.