4D tolkning av gassmigrasjon i grunne sandlag: sammenlignet med reservoar simulering
MetadataVis full innførsel
The work done in this Master thesis is focused on an area influenced by an underground blow-out from Saga Petroleum ASAs 2/4-14 well. 2D seismic site surveys acquired before, during and after the blow-out took place have been used in order to perform 4D amplitude and time shift analyses. These 4D analyses have been performed along the seismic 2D line 602, which run through the relief well and pass southeast of the blowing well. The work has been focused on two sandbodies situated at 492 and 828 meters, which are of Pleistocene and Early Pliocene age respectively. The change in amplitude has been computed along both reflectors, while the change in travel time has been computed along the lower reflector. The first 3D seismic survey over the area was acquired in 1991. The reflectors corresponding to the two sandbodies were interpreted, and maps of the two-way travel time and the variation in amplitude were made. In addition reservoir simulations were performed in order to investigate the probability of the observations done on seismic. The results from the seismic investigations and reservoir simulations were compared with the observations and conclusions made by Saga during and in the period following the blow-out from well 2/4-14. ProMAX® was used to display the 2D data and to pick out information about amplitudes and travel times. The 2D data was oriented by CDP-numbers. The distance between two adjacent CDPs is 12.5 meters, making the CDP-interval from 1050 to 2020 cover a distance of over 12 kilometres. The investigations of the 3D data were performed in Petrel, and ECLIPSE 100 was used to build the reservoir model and to run the simulations. All plots displayed in this thesis are made in MATLAB. The amplitude analyses of the seismic data gave clear indication that the Pleistocene and the Early Pliocene sands became charged with gas following the blow-out. The reservoir simulation confirmed that the likely reason for the observed amplitude increases and positive time shifts were due to gas charging. There was good match between the spread of gas saturation found during reservoir simulation, the areas where increase in amplitude was found and the areas where time shifts could be observed. The formation pressures found during reservoir simulation were relatively high, and had a spatial distribution that deviated from the gas saturation and the observed 4D effects. From comparison with 4D observations it is concluded that pressure effects does not seem to be detectable on seismic in this area. It is also a possibility that simplifications in the reservoir model led to incorrect pressures. Ideas for further work is to do more thorough analyses of the 3D seismic data in order to compare the amplitude and travel time response with the response on the 2D seismic data.