Technology Strategy Board
This project will produce software for modelling three-dimensional sand-body distributions in oil reservoirs. This new tool and technology, by providing additional understanding and characterisation of the reservoir, will increase recovery from existing fields and provide better recovery estimates in new discoveries.
The project will develop software to enhance understanding of turbidite hydrocarbon-reservoirs such as those found in mature areas of the North Sea (e.g. the Outer Moray Firth) and frontier areas of the UKCS (e.g. West of Shetland). Such reservoirs are formed from sand-bodies deposited by turbidity currents (undersea “avalanches” of sand and mud). These currents recur at any particular location on timescales of years to centuries and hence, over geological time, build up substantial thicknesses of sand bodies which may be connected to form high quality, large-volume, potential oil reservoirs. The quality of the resultant turbidite reservoirs is strongly controlled by the vertical connectivity and horizontal extent of the bodies deposited by successive flow events. Unfortunately, 3D seismic techniques are rarely able to image individual sand-bodies in sufficient detail to provide an input for modelling of hydrocarbon flow during extraction and so, until now, assumptions must be made concerning sand-body distributions based upon 1D sampling via wells. Hence, a major stumbling block to effective prediction of reservoir potential is uncertainty over how to convert vertical measurements of sand-body geometries into realistic models of lateral extent and connectivity. Computer forward modelling, of the geological processes which formed the reservoir, is an innovative approach to this wide-spread problem.
The consortium members recently released “4DSediment”; software for modelling of single turbidite flows. This has been used commercially by a number of oil companies (Nexen, Conoco, Shell and BHP) and figures 1 and 2 show this software in use. Further details are given in Waltham et al (Petroleum Geoscience 14, 273-280, 2007).
The next stage is to develop a multi-flow program to allow modelling of connectivity between successive sand deposits. This requires modification of the existing code to allow it to generate multiple flows with each new deposit settling upon a seafloor modified by the preceding flows. The technically risky part of this project is to then develop efficient “inversion” procedures, i.e. methods for finding input parameters (flow-size distributions, concentration distributions etc) which reproduce the sand-body statistics observed in wells. This will be achieved through a combination of basic research into efficient inversion-algorithms together with applied research looking at the use of novel processor architectures (Graphics Processing Units, GPUs and Multi-core Processors, CPUs) allowing cost-effective high-power computing.