To investigate the rational evaluation method and criteria of additional holes of blanket grouting for embankment dam foundations, we make fundamental numerical simulations. In addition, we propose the effective grouted area evaluation method to simply estimate the spatial distribution of the permeability. After the permeability field in the blanket grouting area of an existing rockfill dam is estimated by this method, we carry out numerical simulations and evaluate a main flow path quantitatively using the percolation theory. It is confirmed that the streamwise distribution of high permeability elements is important for forming a main flow path, and it is found that the inequality between the sequential order of 15% exceedence probabilities and percolation thresholds is reversed according to the spatial distribution of permeability.
Blanket grouting is conducted beneath the impervious zone of embankment dams to control the seepage through the shallow part of the rock foundation. The rectangular pattern is generally used as the layout of blanket grouting holes in Japan, and the grouting is done according to the split-spacing method. In this layout, primary holes are first drilled at the same intervals, and secondary holes are drilled diagonally between them. This halving sequence continues until the permeability of the sequence decreases sufficiently. These sequences are divided into two: one group of sequences is called the sequence of design holes where all holes are drilled at the same interval in each sequence, and the other is called the sequence of additional holes where such holes are only drilled in locations that are below the required standard in imperviousness. To execute blanket grouting effectively, it is necessary to widely and precisely measure the permeability of the dam foundation before and during blanket grouting, because a main flow path is caused not only by the local high-permeability but also by the spatial distribution of the permeability. In Japan, the effect of blanket grouting is now evaluated based on the sequential order of 15% exceedence probabilities (hereafter, it is expressed as '15% exceedence probability value') in permeability in the designed area at each sequence. But this statistic cannot be considered as the representative value in the dam foundation when additional holes are constructed locally. In the split-spacing method, there are some advantages in that the locations of additional holes are chosen easily, but there are also some disadvantages: the evaluation of the improvement in the imperviousness by blanket grouting is retarded and the number of additional holes increases in a geometrical ratio with the progress of the sequences. We should tackle these problems to respond to the recent social demand for the reduction of the cost of dam construction, so it is necessary to examine more rational methods of evaluating grouting effects. We had already carried out two-dimensional seepage analyses using a simple model in order to clarify the relationship between the spatial distribution of unimproved elements and the location of a main flow path (Satoh et al 1999a, I999b) using the percolation theory (Odagaki 1997).
