Fidler, Scott Richard2006-07-282006-07-2819981998http://hdl.handle.net/10012/240The relative importance of fractures and other macropores in clay-rich glacial deposits depends on the temporal and spatial scales being considered. Hydraulic conductivity of such materials generally depends on the scale at which the measurement is made, as a result of the localised high conductivity associated with the macropores. Small scale measurements tend to be highly variable, whereas tests conducted at a sufficiently large scale yield a relatively constant value for hydraulic conductivity, since they encompass a representative portion of the macropore network. Tests which involve rapidly varying hydraulic conditions will mainly influence the macropore network, since rapid changes in pressure can propagate quickly through these high conductivity, low storage features, while the pressure in the low conductivity, high storage matrix remains unchanged or changes very slowly. On the other hand, if sufficiently slow pressure changes are applied to the system, very little pressure disequilibrium will develop between the macropores and the matrix. In this thesis, field and modelling studies are employed to investigate particular aspects of the spatial and temporal scale dependence of hydraulic behaviour in naturally fractured clay-rich glacial deposits. The overall objective of the thesis is to develop appropriate field methods, and conceptual models for their interpretation, for measuring parameters which control hydraulic behaviour in heavily fractured, near surface clays. Some of these parameters, such as bulk vertical hydraulic conductivity, have been very rarely measured in this type of material. Similarly, the methods used in this thesis to interpret transient hydraulic tests have not commonly been used for tests in fractured clays. Two field experiments were conducted at locations separated by approximately 1 km, at a site in the St. Clair Clay Plain in south-western Ontario. At this site, glacio-lacustrine clay is present from the surface to a depth of approximately 40m. The upper 2-4m of the clay are heavily fractured and pervasively weathered. Below the zone of pervasive weathering the clay is unweathered except in the vicinity of fractures, which are visible to a depth of 6m. Hydraulic testing and mapping of fractures and other macropore features was carried out in the upper weathered zone at the two experimental sites. The results of the mapping provide the context and important parameters for the interpretation of the hydraulic tests, and also indicate the potential variability in fracture patterns between two nearby sites within the same clay plain. Although surface observations at the sites did not reveal any features that suggest they are atypical of the clay plain in general, two very different fracture patterns were observed in the weathered near-surface clay at the two sites considered. Mapping of the preferential flow paths was carried out at one of the two sites by infiltrating a dye and then mapping its distribution in the subsurface. At this site, it was found that loosely infilled, primarily vertical, cylindrical cavities were major conduits for fluid flow. Open root holes, often located along fracture faces, were also identified as important conduits for fluid flow. Dye staining did not indicate distributed flow throughout the plane of major fractures in any instances. Hydraulic testing was carried out at the two sites using an experimental arrangement in which a pond was maintained at the surface, overlying several flat-lying sand-filled induced fractures. This arrangement was used to conduct a variety of transient and steady state hydraulic tests, which were monitored using a network of electronic strain gause pressure transducers. At both sites, the pond and induced fractures were several metres in diameter, and thus influence a relatively large volume of the fractured clay, in proportion to the volume that would be tested with single-well tests. Steady state conditions during pumping tests in which water was pumped from the induced fractures allow the calculation of vertical hydraulic conductivity and its variation with depth, and also indirectly allow horizontal hydraulic conductivity to be estimated. At the first site, it was found that vertical hydraulic conductivity does not change with depth in the weathered clay, whereas vertical hydraulic conductivity decreased by an order of magnitude over the depth of the weathered zone at the second site. It was also found that the ratio of bulk horizontal hydraulic conductivity to bulk vertical hydraulic conductivity was much lower at the second site than at the first site. The use of induced fractures to conduct pumping test s made it possible to observe differences in behaviour between the fracture network and the clay matrix during transient hydraulic tests. The results of pumping tests conducted at the first site indicates that disequilibrium between fractures and the clay matrix can persist for several hours. Thus in general, interpretation of scenarios in which hydraulic conditions vary on the time scale of hours or less will require models which account for the interaction between the fractures and the clay. Such models can reproduce certain aspects of behaviour, such as the rapid transmission of pressure through fractures and the concurrent slow changes in pressure in the clay matrix, which can not be described using equivalent porous medium models. A discrete fracture model was used to interpret the transient hydraulic field tests. The values obtained for parameters such as bulk hydraulic conductivity and specific storage of the clay are consistent with values measured independently in other tests.application/pdf8291922 bytesapplication/pdfenCopyright: 1998, Fidler, Scott Richard. All rights reserved.Harvested from Collections CanadaDoctoral Thesis