Breadcrumbs Section. Click here to navigate to respective pages.
Chapter
Chapter
reservoirs. Even with only two joint patterns and two different regional stress states, a number of mechanisms and corresponding flow regimes have been revealed. Typically, fluctuating injection pressures, anisotropy of hydraulic conductivities (both initial and induced), and irreversible changes of stress and joint dilation are predicted. In addition, the magnitude of the injection pressures and the net effect of the injection period on the final distribution of fracture conductivity are quite different in each case. This response is qualitatively similar to some field obser-vations (for example, Shuck and Komar, 1979). Thus a very idealised model has prompted a better understanding of mechanisms that may explain the observed characteristics of fluid injection into jointed rock. Both examples fall into the data-limited category as defined by Star-field and Cundall (1988). It would be only too easy to dismiss the modelling attempts on the basis of the lack of information available; there is uncertainty concerning geometry and material behaviour in both cases. However, the implementation of an idealised numerical model which can then be used as an experiment to probe the behaviour of the system has, in both cases, led to a better understanding and a characterisation of the system response. The numerical models described here, based on the explicit finite-difference technique, have proved to be extremely valuable tools in this context.
DOI link for reservoirs. Even with only two joint patterns and two different regional stress states, a number of mechanisms and corresponding flow regimes have been revealed. Typically, fluctuating injection pressures, anisotropy of hydraulic conductivities (both initial and induced), and irreversible changes of stress and joint dilation are predicted. In addition, the magnitude of the injection pressures and the net effect of the injection period on the final distribution of fracture conductivity are quite different in each case. This response is qualitatively similar to some field obser-vations (for example, Shuck and Komar, 1979). Thus a very idealised model has prompted a better understanding of mechanisms that may explain the observed characteristics of fluid injection into jointed rock. Both examples fall into the data-limited category as defined by Star-field and Cundall (1988). It would be only too easy to dismiss the modelling attempts on the basis of the lack of information available; there is uncertainty concerning geometry and material behaviour in both cases. However, the implementation of an idealised numerical model which can then be used as an experiment to probe the behaviour of the system has, in both cases, led to a better understanding and a characterisation of the system response. The numerical models described here, based on the explicit finite-difference technique, have proved to be extremely valuable tools in this context.
reservoirs. Even with only two joint patterns and two different regional stress states, a number of mechanisms and corresponding flow regimes have been revealed. Typically, fluctuating injection pressures, anisotropy of hydraulic conductivities (both initial and induced), and irreversible changes of stress and joint dilation are predicted. In addition, the magnitude of the injection pressures and the net effect of the injection period on the final distribution of fracture conductivity are quite different in each case. This response is qualitatively similar to some field obser-vations (for example, Shuck and Komar, 1979). Thus a very idealised model has prompted a better understanding of mechanisms that may explain the observed characteristics of fluid injection into jointed rock. Both examples fall into the data-limited category as defined by Star-field and Cundall (1988). It would be only too easy to dismiss the modelling attempts on the basis of the lack of information available; there is uncertainty concerning geometry and material behaviour in both cases. However, the implementation of an idealised numerical model which can then be used as an experiment to probe the behaviour of the system has, in both cases, led to a better understanding and a characterisation of the system response. The numerical models described here, based on the explicit finite-difference technique, have proved to be extremely valuable tools in this context.
ABSTRACT