Perrochet, Pierre

2013, Preisig, Giona, Perrochet, Pierre

Field observations and laboratory experiments have clearly demonstrated that heavily perturbed / exploited aquifers are subject to 3D deformations, which may cause significant socio-economic impacts at regional scale. Most common examples include: (1) excessive pumping of groundwater from deep aquifers leading to land subsidence; (2) deep excavation of tunnels in permeable geological units resulting in dangerous differential consolidation, especially for dams; and (3) fluid injection into deep reservoirs causing ground uplift and microsismicity. These manifestations are due to a substantial modification of water pressures within the aquifer, leading to effective stress variations, and deformations. Moreover, such deformations modify hydrodynamic parameters, i.e., hydraulic conductivity, porosity and storage coefficient. In confined or deep aquifer systems, hydrodynamic parameters have to be considered as effective stress-dependent variables. In such environments the assumption of constant parameters can lead to significant quantitative errors. The afore-mentioned fluid-to-solid hydromechanical processes seem to be essentially governed by hydrogeological, geomechanical and structural properties of the aquifer system. In order to take into account the major processes, as well as their principal properties, regional coupled hydromechanical simulation necessarily requires simplification of the governing equations to be operational in real, large scale, hydrogeological systems.
In the present thesis, model functions relating effective stress to hydrodynamic parameters are developed from fundamental hydrogeological and physical concepts, and implemented in the groundwater flow equation. Proposed stress-dependent equations are verified by a comparison with laboratory and field data. This is carried out for (1) fractured aquifers, i.e., consolidated rocks whose porosity results principally from the presence of fractures, cracks, joints and faults, and (2) granular porous aquifers, i.e., unconsolidated rocks whose porosity results from voids between solid grains. The relation between porosity and stress is also used to elaborate a deformation model for solving aquifer vertical volume change, i.e., ground settlement / uplift. A modelling approach is proposed in order to solve fluid-to-solid hydromechanical processes at regional scale, considering detailed geological structures. In this numerical method, hydrodynamic parameters are considered as stress-dependent variables.
Exact analytical solutions solving flow in a media under stress are developed in order to verify the numerical method. The proposed approach is then applied to the analysis of real case studies. In particular, to (1) the abnormal deformation of the Zeuzier arch dam (Wallis, Switzerland) due to the drainage of an unexpected confined aquifer by the Rawyl exploratory adit; (2) the problematic of water inflow into tunnels based on the geological investigations undertaken by the Lyon-Turin railway project for the 57 km basis tunnel; as well as (3) the anthropogenic land subsidence affecting the Mexico City basin.
Quantitative studies of deep aquifer systems considering constant hydrodynamic parameters result in non-accurate volumetric discharge rates and pressure head fields. On one hand, increasing effective stress leads to decreasing hydrodynamic parameters. This results in a diminution of volumetric flow rates through a deep reservoir or in a deep excavation. Moreover, the decrease of water pressure is slowed down due to the decrease of hydraulic conductivity. This has repercussion on consolidation time. On the other hand, if - and only if - the rock is elastic, decreasing effective stress can lead to increasing hydrodynamic parameters and volumetric discharge rates.
For analytical solutions of volumetric discharge rate in deep wells or into a tunnel, the dependency of hydrodynamic parameters on effective stress can be taken into account by using a factor allowing stress consideration; whereas, in numerical analysis, such a process can be considered by implementing stress-dependent parameters in the groundwater flow and aquifer deformation models.
The proposed numerical approach for fluid-to-solid coupled hydromechanical processes, is computationally simple, based on few unknowns, and efficiently reproduces regional consolidations in geologically oriented 3D meshes. This method is critical for hydrogeological and geomechanical quantitative studies investigating the sensitivity of deep aquifers on decreasing / increasing effective stresses. In particular for regional scale projects where water pressure may be subject to substantial modifications, such as dam construction or tunnel excavation, geologic radioactive waste repositories, deep reservoir exploitation for CO₂ sequestration, geothermal energy production, as well as extraction of groundwater and / or hydrocarbons.

1997, Tacher, Laurent, Perrochet, Pierre, Parriaux, Aurèle

A discrete reduced distance method to generate 2-D and 3-D granular porous media is presented. The main property of the method is to produce heterogeneous and/or anisotropic packed beds of joined grains with arbitrary shapes and optimum fitting (i.e., minimum porosity). The iterative generation process starts with the coarsest grain and adjusts the size and location of the next ones depending on the updated available space. Hence, grain size distribution cannot be specified directly but is merely the consequence of user defined input parameters. The latter consists of a set of randomly distributed initial points, a few typical predefined grain shapes as well as the minimum and maximum grain diameters. The simulated granular media can readily be processed by an appropriate mesh generator to allow for subsequent numerical solutions of differential equations.

2003, Marechal, Jean-Christophe, Perrochet, Pierre

The present paper addresses two major problems encountered during tunnel drilling and related to the hydraulic interaction with surrounding groundwater bodies. The first one is the prediction of water discharge into the tunnel, as a function of the geometric and hydrogeological data. The second problem is related to the assessment of the draining effects on surface waters (springs, lakes, wetlands). Surface monitoring campaigns are costly and evaluating their duration is a sensitive question. Both problems are tightly related and depend on aquifer dynamics. It is shown that in a geological context with steeply dipping structures, nearly vertical, inducing series of aquifers and aquicludes such as in the Alps, the drainage of the aquifer by the tunnel can be modelled by the analytical solution of Jacob and Lohman [1952] for artesian wells. First developed for horizontal, confined unsteady flow towards a vertical well with constant drawdown, it is adapted here to a horizontal tunnel by a rotation of pi/2. The main difference between this solution and more classical Theis' solutions is that a constant drawdown condition replaces the constant discharge rate condition. Hence, a relation is obtained for the time-dependent discharge rate Q(t) detected at the tunnel after drilling, as a function of aquifer transmissivity (T), storage coefficient (S), initial drawdown (s(o)) and tunnel radius (r(o)). This analytical solution is compared to a finite-elements model simulating a draining tunnel in a simplified 2D vertical cross-section. The comparisons show that the decay of the tunnel discharge can be divided into two periods. During the first period, radial drawdown develops around the tunnel and there is excellent match between analytical and numerical results. Tunnel discharge results from the decompression of rock and water (storage effects) as a response to the sudden initial drawdown at the tunnel location. During the second period, the drawdown cone reaches the aquifer limits (lateral and upper) and numerical discharge rates decrease faster than analytical rates because of hydraulic heads decline at the aquifer limits. In the Alps, such trends were observed for the discharge rates into the Simplon and Mont-Blanc tunnels, and the analytical solution of Jacob and Lohman [1952] was applied to the first discharge period to evaluate aquifer transmissivity and storage coefficients. As indicated by the simulations, and corroborated by field observations, the analytical solution is only valid during a first period after tunnel opening, the duration of which scaling with the inverse of the aquifer diffusivity (T/S). In the second part of the paper, dimensionless type-curves are presented to enable rapid evaluation of the time where a given drawdown is observed at a given distance from the tunnel. Accounting for tunnel geometry (radius and depth) and aquifer parametres (T and S), these curves could for instance help in practice to determine when surface waters would start to be affected by a draining tunnel underneath. Although neglecting the boundary effects discussed in the first part of the paper, these type-curves demonstrate the great inertia of mountain aquifers, and could be used to adjust the duration of surface monitoring campaigns according to the specific tunnel/aquifer settings..

2001, Guyonnet, Dominique, Perrochet, Pierre, Come, Bernard, Seguin, J J, Parriaux, Aurèle

In the context of municipal solid waste and hazardous waste disposal, the notion of "equivalence" between different barrier designs appears in regulatory documents from several industrialized countries. While in the past, equivalence has been thought of mainly in terms of contaminant travel times, in recent years it has been defined more in terms of the magnitude of a disposal site's potential impact on groundwater resources. This paper presents some original analytical solutions to the problem of contaminant migration through a multi-layered mineral barrier. The solutions account for the two major mechanisms of subsurface contaminant migration, namely, advection and diffusion-dispersion. An example application using the proposed solutions and a numerical model illustrates how one multi-layered mineral barrier can be considered superior to another from a strictly hydro-dispersive viewpoint. The influence of partial saturation of the mineral barrier is investigated using a numerical solution to the Richards equation for unsaturated flow. It is emphasized that conclusions relative to the superiority of one multi-layered barrier, with respect to another, should not only consider hydro-dispersive aspects, but also other processes such as the mechanical and chemical evolutions of the different barrier components. Although such phenomena are poorly addressed by existing models, failure to take them into account, at least in a qualitative fashion, may lead to unconservative conclusions with respect to barrier equivalence. (C) 2001 Elsevier Science B.V. All rights reserved.