Stratigraphic Forward Modelling comparison with Eclipse for SW HubPhase Two - Dynamic Model

[摘要] In a previous project "Stratigraphic forward modelling for South West Collie Hub: Phase One - Static Model. ANLEC RD Targeted Project 129" a stratigraphic forward model of the time period from 250 Ma to 182 Ma (Olenekian to Pliensbachian - Base Wonnerup equivalent to top Eneabba equivalent) was constructed in that part of the South Perth basin which is a candidate area for geological sequestration of carbon dioxide. In that project we used the CSIRO Sedsim stratigraphic forward modelling software (constrained by available well and seismic data and previous published geological studies) to produce a ‘static reservoir model’, or in other words a predicted three-dimensional distribution of porosity and permeability, of the potential carbon sequestration site in the area around the Harvey-1 well. The research targeted the significant current risk involved in applying traditional oil-field reservoir characterisation tools to the construction of static geological models of saline formations. As stated in the latest US DoE document on CCS,“…. The successful characterization of a site is the most important step in ensuring the safe and economic operation of a CO2 GS site.” [2010 DoE Site Screening, Selection, and Initial Characterization for Storage ofCO2 in Deep Geologic Formations. DOE/NETL-401/090808, November 2010].Phase One demonstrated an alternate workflow for producing a static reservoir model for geological sequestration ofCO2. The workflow was based on limited well and seismic data constraints, and used the Sedsim stratigraphic forward modelling software to translate a conceptual geological model into a cellular geological volume. That volume formed the basis of a Phase Two study which consisted of a reservoir flow simulation of CO2 injection into the Lesueur Sandstone using the nested 500 m grid over the Collie Hub dynamic model area of 25 km x 15 km.The initial UNSW Eclipse E300 CO2STORE runs on the unfaulted deposition model produced similar results to the Schlumberger Eclipse model based on a ‘conventional’ static model.However an incorrect fault juxtaposition format caused delays in the faulted static model runs and valid comparisons were not possible within the framework of the project.The results of the Phase Two dynamic modelling study support the general applicability of stratigraphic forward modelling as an alternative route to static model generation, but also point to areas where further work is needed.Incorrect conversion of faulted Sedsim output to grdecl format, and in particular the fault definition capability meant that the correct fault juxtapositions were not present in the grid. This led to convergence errors in the dynamic model. Although there was insufficient time to re-run the dynamic model with the correct fault juxtapositions, there is a clear path to remediation.

[发布日期] 2014-03-12 [发布机构] CSIRO

[学科分类] 地球科学(综合)