In this work, in order to achieve the above mentioned objectives, we use pressure signals obtained over a period of ten years at Mont Terri (1996/2005), from which records as long as one year or more will be analysed in this text (see Section 4). But, to analyse the effects of a moving excavation front, we also focus on a period of 5 months of "syn-excavation" period for gallery "Ga98", and on a much shorter period of fast pressure changes lasting only a couple of weeks (see Section 5). In

more »

... , we study both long and short time scale records, from years to weeks. Some results of pressure signal pre-processing and long time scale analyses were presented in Fatmi et al. (2007 Fatmi et al. ( , 2008 at an early stage of this study. The present text provides a significant update of the results of long time scale analyses, and presents also new types of results on the analysis of excavation effects on pore pressures at very short time scales (weeks). Accordingly, the outline of this chapter is as follows. Section 2 describes the hydro-geologic site and the pore pressure measurements at the Mont Terri site (in the framework of experiment LP14). Section 3 is devoted to the mathematical and statistical methods of signal pre-processing (e.g. reconstruction of missing data) and of signal analysis (time-lag correlation functions, frequency spectra and cross-spectra, wavelets, envelopes). For hydro-geologic interpretation of pressure signals, the methods involve, not only statistical treatments, but also simplified elastic hydro-mechanical models: the concept of a specific storativity coefficient "Ss" (in relation to earth tides); and the barometric efficiency (involving both "Ss" and porosity "Φ") in relation to air pressure fluctuations. These models are briefly described at the end of the section (Box N°2) and their application is developed in the remaining sections of the text (see below). Section 4 develops the implementation of "long time scale" statistical analyses and hydrogeologic interpretations of signals, in the case where the objective is to obtain a global estimate of the two hydro-geologic coefficients (Ss,Φ) in two steps, first "Ss" from semidiurnal earth tide effects using multi-resolution wavelet analysis, and secondly "Φ" using the barometric spectral gain at diurnal frequency. Section 5 develops, in contrast, a somewhat more sophisticated implementation over "short time scales" with a strong non stationary trend in the pore pressure signal. This occurs in the case of a moving excavation front passing nearby the pressure sensor. The objective is to identify the effect of the evolving "EdZ" at the location of the pressure sensor, and in particular, its effect on the evolution of Ss(t). The time scale of analysis is typically a couple of weeks, and a new tool has been developed (based on statistical envelopes) in order to identify the modulation in time of the amplitude of earth tide fluctuations. Section 6 provides a summary of the methods and results, and suggests possible modifications and extensions of this work concerning statistical methods (multi-cross analyses) and hydro-geologic interpretations (enhanced hydro-mechanical models and inverse problems): see the "outlook" in Sub-Section 6.2. The text is completed by an appendix (for abbreviations and symbols), and by a list of references. Hydrogeologic setting and pressure signal measurements Hydrogeologic setting at Mont Terri (Opalinus clay rock) The Mont Terri Rock Laboratory was excavated within the "Opalinus Clay" of the Mont-Terri anticline, which has a very low permeability. The geological cross section shown in