Analysis of three sets of SWIW tracer test data using a two-population complex fracture model for matrix diffusion and sorption

This study has been undertaken to obtain a better understanding of the processes underlying retention of radionuclides in fractured rock by using different model conceptualisations when interpreting SWIW tests. In particular the aim is to infer the diffusion and sorption parameters from the SWIW test data by matching tracer breakthrough curves (BTC) with a complex fracture model. The model employs two populations for diffusion and sorption. One population represents the semi-infinite rock matrix and the other represents finite blocks that can become saturated, thereafter accepting no further diffusion or sorption. For the non-sorbing tracer uranine, both the finite and the semi-infinite populations play a distinct role in controlling BTC. For the sorbing tracers Cs and Rb the finite population does not saturate, but acts essentially semi-infinite, thus the BTC behaviour is comparable to that obtained for a model containing only a semi-infinite rock matrix. The ability to match BTC for both sorbing and non-sorbing tracers for these three different SWIW data sets demonstrates that the two-population complex fracture model may be useful to analyze SWIW tracer test data in general. One of the two populations should be the semi-infinite rock matrix and the other finite blocks that can saturate. The latter can represent either rock blocks within the fracture, a fracture skin zone or stagnation zones. Three representative SWIW tracer tests recently conducted by SKB have been analyzed with a complex fracture model employing two populations for diffusion and sorption, one population being the semi-infinite rock matrix and the other, finite blocks. The results show that by adjusting diffusion and sorption parameters of the model, a good match with field data is obtained for BTC of both conservative and non-conservative tracers simultaneously. For non-sorbing tracer uranine, both the finite and the semi-infinite populations play a distinct role in controlling BTC. At early times (the tracer peak) the finite-block population is most important, but at later times (the tracer tail), the finite-block population becomes saturated and the semi-infinite population controls the BTC. In contrast, for sorbing tracers Rb and Cs, the finite population does not saturate so a single-population model can be used to match these BTC. Hence, to match the behaviour of both non-sorbing and sorbing tracers, two populations, one finite and the other semi-infinite, are required to capture all the features of the BTC. The conclusion of this study using the three representative sets of SWIW data sets shows that the two-population complex fracture model may be useful to analyze SWIW tracer test data in general. One of the two populations should be the semi-infinite rock matrix and the other finite block that can saturate. The latter can be representing either rock blocks within the fracture, a fracture skin zone or stagnation zones.