In the hydrogeological characterization of sedimentary rocks, hydraulic properties of fault and cap rock structure are important factors. Although, the certainty of hydrogeological model depends on the amount of geological or hydrogeological data mainly obtained from boreholes, it is indispensable to make an outlined model from small amount of data obtained in the early stage of field investigation, and improve the details of the model as the data are obtained. In sedimentary rocks, each aquifer often has the different hydraulic head caused by various reasons such as surface topography, long-term change of sea level, or cap rock structures that were not explicitly observed in geological investigation in boreholes.
In this paper, we use the vertical hydraulic head distribution obtained in the pressure monitoring or the water level during the drilling of boreholes as the observed data, and applied numerical inversion code iTOUGH 2 to construct alternative hydrogeological model in Horonobe underground research laboratory site of Japan Nuclear Cycle Development Institute. We applied random sampling for the coverage of assumed cap rock, and two models for the hydrogeological structure of faults in the study were performed. The followings are the main results of this study.
1) It is necessary to assume the low permeable cap rock structure that was not clearly observed in boreholes, to reproduce the deep high-pressure zone as a characteristic hydraulic feature in this site.
2) The numerical inversion with random sampling of cap rock model shows that if the cap rock coverage is greater than 75%, the observed hydraulic head profile can be reproduced.
3) Hydraulic head profile of the deeper zone is controlled by the hydrogeological structure of steep fault in this area. Thus, the increase of hydraulic head profile in the deeper zone, and the numerical inversion can determine the hydrogeological structure of fault.
4) Hydrogeological structure of fault dominates the vertical groundwater flow in the vicinity of faults.
From these results, the numerical inversion using the head distribution is proven to be a useful method to construct preliminary but quantitative hydrogeological model from small amount of data obtained in the early stage.
View full abstract