In rock engineering, pilot holes are used to make a prediction of the expected rock conditions for excavation and rock reinforcement design. In some instances, the results coming from conventional upscaling can't be verified by observations. The differences between the data obtained from the pilot-hole analysis and the tunnel mappings have been collected in order to develop a method that will improve the upscaling results. The data that was taken into account came from the ONKALO rock characterisation facility (Olkiluoto, Finland). The upscaling bias is somewhat systematic, which may indicate that the results could be improved by implementing a method which takes into account the observed differences. It is also important to discover the cause of this bias. In order to use the prediction generated from the pilot-hole data in rock engineering design, the prediction has to be reliable. In some cases, the analysis of the pilot-hole data gives a different idea of the rock quality from the one ob-served during the tunnel mapping.
The Olkiluoto nuclear power plant consists of two operating reactors, a third one under construction and a fourth at a planning stage. The entire facility includes, as well, a LILW (low and intermediate level waste) repository and an underground HLW (high level waste) repository under construction on the Olkiluoto Island as well as the ONKALO Site, which includes the access tunnel and the final repository facility to be located at the depth of 437 m below the sea level. The site characterisation for the construction of the aforementioned repository has been carried out for about 20 years, including several kinds of investigations to in-crease the level of confidence in the Site description and to characterise the bedrock properties and groundwater characteristics (Posiva 2009–01).