Technical Report NTB 91-16

At a meeting of Nagra and BGS representatives, held at Keyworth on May 24^th, 1989, a, number of processes were examined which might, conceivably, be responsible for the anomalously low hydraulic heads observed in borehole testing in the Opalinus Clay. An anomalous hydraulic head was defined as a head which cannot be interpreted in terms of normal gravitational advective flow under steady-state conditions.

In examining the problem it was apparent that three types of process might be invoked, the first associated with long-term transient flow, the second with coupled flow (in particular osmotically-driven flow), and the third with borehole effects such as swelling, the perturbation of in situ stresses and plastic deformation of the borehole walls.

This report represents a preliminary study of these identified processes. The approach taken has been to examine each type of process individually, first at a general level, and then with specific reference to the Opalinus Clay. Simple theoretical frameworks are established and a number of scoping calculations are performed in an attempt to distinguish first-and second-order effects. Baseline hydrogeological, geotechnical and geochemical parameters for the clay-shale are obtained from the available test data or are calculated from the known interrelationships between properties. Where no data exists, "generic shale" properties are taken from the literature.

What emerges from this study is that, although the mechanisms of groundwater movement in mudrocks are at present uncertain, certain processes could have such profound hydrogeological effects that their significance goes way beyond the interpretation of borehole test results, with repercussions in the development of a strategy for site characterisation, in modelling and in safety assessment.

Our calculations suggest that long-term transient flow may be occurring in the Opalinus Clay as a consequence of the stress changes associated with the process of overconsolidation and the removal of sedimentary cover by erosion. In addition, based on the strong evidence for neotectonic deformation of the crust in many areas of Switzerland, anomalous hydrogeological conditions may develop in low permeability rocks such as Opalinus Clay as a result of stress-or strain-induced pore pressure changes.

The general picture which emerges from our calculations on the overconsolidation mechanism is one in which fracture flow fairly rapidly re-equilibrates after exhumation but re-adjustments of the water content of the intact mudrock may occur over a much longer time period, constrained somewhat by diagenetic bonding of clay minerals and the presence of cements. Thus fracture flow might be regarded as being essentially out of phase with flow in the intact clay (matrix flow).

The low porosity and generally high clay content of Opalinus Clay suggest that it might also act as an efficient semi-permeable membrane supporting osmotically-driven flow not only across the stratum but also within it. Our calculations show that osmosis can have large effects on hydraulic head measurements and these must be considered when testing in mudrocks. Coupled flow may also be extremely important in modelling groundwater movement in mudrock environments. If we acknowledge osmotic flow to be a significant component of total flow in mudrocks, then long-term transient behaviour must be viewed as a response to hydraulic and chemical disequilibria.

Deep burial and subsequent exhumation of the Opalinus Clay have endowed it with a marked "thirst" for water. This propensity to draw in water can be viewed in simple mechanical terms as a consequence of rebound from overconsolidation or in chemical terms as a suction associated with the processes of osmosis and/or hydration operating at the microscopic scale. Large suctions (negative pore pressures) may develop in near surface material. The presence of these suctions is entirely consistent with the observed high swelling capacity of the clay. It is also fundamental to our discussions on long-term transients and osmotically-driven (coupled) flow within mudrocks.

Thus we see that in mudrocks, particularly highly compacted clay-shales like the Opalinus Clay, there are a considerable number of reasons why the hydraulic head (determined in the usual way) might not be interpretable in terms of gravitational flow under steady state conditions. In fact, it is difficult to argue the case that the head should not be "anomalous" in many mudrock environments.

Borehole effects associated with the mechanical, thermal and chemical changes occurring in the formation during drilling and testing can have a significant influence on the measured hydraulic response during testing. Although swelling due to the introduction of fresh water is probably the most significant effect, in deep boreholes perturbation of the in situ stress field and plastic deformation of the rock may have a considerable effect on the pore pressure distribution around the borehole. If the measuring system has low compliance, the measurements may also be affected by time-dependent borehole closure.

When we examine the particular case of the anomalous hydraulic heads in borehole RB26B at the site of the proposed Wisenberg Tunnel in the Homburger Tal (see Section 6.2), then we cannot totally exclude very simple explanations for the response such as the presence of air in the test zones. However, these anomalous heads can also be explained (semi-quantitatively) by assuming that the total head at each point in this borehole is given by:

h = z + h_pp + h_sp

where z is the elevation head, h_pp is the pressure head in the macropores (associated with hydrostatic pore pressure), and h_sp is the solute head in the macropores. If this scenario is correct, then it has very important implications to the design and interpretation of hydraulic tests in mudrock formations. The most obvious consequence is that the magnitude of the hydraulic head determined in such tests is likely to be sensitive to the solute chemistry of the test fluid. This calls into question the common practice of using fresh water as the test fluid. Non-polar, non-reactive liquids might be more suitable for this purpose. Alternatively, the chemistry of the test fluid might be regarded as a variable in the development of a testing methodology for mudrocks.

An interesting relationship emerges in this study between chemico-osmotic (coupled) flow in a mudrock and the physico-chemical process of interparticle swelling. In the vicinity of a borehole, the two processes are virtually inseparable. Both provide a mechanism by which water in the test zone is drawn into the formation and both are sensitive to the chemistry of the test fluid. This suggests that both processes might be describable in terms of a single unified theory. It is probably no coincidence that the characteristics of a clay which render it an efficient osmotic membrane are identical to those which endow it with a high capacity to swell.

The issues raised in this report are so important to hydrogeological site characterisation in mudrock environments that additional field and laboratory studies are clearly demanded, backed up by a parallel development of the theoretical framework. We distinguish three priority areas: (a) Development of a borehole testing methodology specific to mudrocks, together with the necessary theoretical models for reduction of the test data; (b) Development of a relatively simple, possibly one-dimensional, numerical modelling capability to allow sensitivity analyses to be performed on transient flow under hydraulic and chemical gradients; and (c) Further examination of the implications of tectonic deformation to the regional hydrogeology of the Opalinus Clay. Specific recommendations are made in Chapter 9.