| Popis: |
Abstract The interpretation of pressure transients by the use of analytical models generally leads to a simplified description of the geological heterogeneities around the well. Although this approach works when the geology is simple, clearly, more complex geological environments will prove difficult to analyses. In channelized systems, such as alluvial and deep sea fans, the reservoir geometry may be poorly represented by simplistic geometrical arrangements that exist in the current analytical models. For example, the calculation of the distance to the no-flow boundaries is disturbed by the presence of levees. This aspect of the problem is clearly important as such calculations lead to an estimation of channel width, which may later be used as an input parameter for stochastic modelling. Results of flow simulations of well tests from such complex geometries are presented in this paper. Using finely-gridded models generated from outcrop description and stochastic modelling, the influence of different geological parameters on pressure response is evaluated. The study is performed in four steps. Firstly, a field example is used to define the problem: a pressure transient test is interpreted in terms of complex geology by the use of a combination of analytical solutions. Secondly, a known reservoir geometry (from outcrop study) is input into a well-test flow simulation model. Thirdly, a more complex reservoir is generated by stochastic (Boolean) techniques which is also simulated as in step 2. Fourthly, a new analytical solution allows for the correct interpretation of the original field example. When reservoirs are characterized by channel type features, and particularly stacked channels, the no-flow boundaries identified on the derivative curve represent the limits of the system, rather than the limits of individual features. When channels are connected by levees, the levees influence the calculation of the distance to no-flow boundaries. The presence of levees increases the apparent channel width. All well-test interpretation results must be checked by comparison with the best available geological description of the sedimentary environment before quantitative data are supplied for geostatistical modelling. Introduction Interdependence between geology and well-test interpretation has been well recognized. Due to non-uniqueness of interpretation, geology plays a leading role in the selection of the best possible analytical model for a given well-test. At the same time, the results coming from well-test interpretation contribute to the improvement of the geological understanding and modelling. As the well-test physically 'up-scales' the incorporated reservoir geology it is used to characterize the reservoir geometry at the appropriate scale, and specifically:–reservoir connectivity, and–sand body dimensions. In sedimentary environments typified by channelized sandbodies (fluvial and deep-sea fans), reservoirs are often characterized by laterally non-continuous reservoir units. The evaluation of such reservoirs by the appropriate use of well-test and petrophysical data is critical to the determination of lateral continuity and channel connectivity. Furthermore, sand-body geometries calculated from welltest interpretation are used as quantitative input data for stochastic reservoir modelling. However, it is accepted that geometries and dimensions derived from well-test data are simplistic (due to the volume-averaging problem), hence, it is also necessary to use data from outcrop studies. To improve our ability in the interpretation of well-tests in such complex geological environments we investigate the well-test response from a known reservoir geometry. P. 501^ |
