Integrated Characterisation
Reservoir characterisation is a complex process in which team members work together to describe reservoir characteristics using a variety of data from a number of disciplines. (3) These studies must incorporate results from all scales including cores, seismic and well tests. The result of the process is a static and dynamic mathematical model of the system. The ultimate goal of the process is to predict future reservoir flow behaviour under a variety of operational methods in order to optimise reservoir development. The models are often validated through comparison with historical performance. Uncertainty analysis is also a very important issue.
Every decade has seen improvements in the individual reservoir characterisation technical sub-topics. Early characterisation improvements that directly impacted the industry’s productivity, range from G E Archie’s initial petrophysical work for estimating rock properties from well logs, (4) to the introduction of seismic algorithms to map subsurface structure and stratigraphy, (5) to the use of reservoir flow simulation for reservoir performance prediction. (6) Rapid reservoir characterisation improvements in the individual technical disciplines continued during the last decade. These technical productivity tools have had major impacts on the industry’s bottom line. Improvements such as neural-net classification of seismic rock properties, seismic inversion, 3-D geostatistical modelling of reservoir connectivity, interactive transient test analysis tied to alternate reservoir models, and dynamic geo-model ranking using streamline technology. (7) The most significant advances in reservoir characterisation are a result of creating integrated 3-D reservoir geo-models that are capable of visualising data and interpretations from different technical disciplines. Use of such shared earth models greatly improves reservoir management decisions. No less important is the effort that has gone into creating integrated databases of petrophysical, geophysical, geologic and engineering data that is used as a foundation for ensuring a common data source for the new shared earth models.
Generally, each characterisation improvement requires more complex algorithms and requires better-educated technical experts to effectively quality control (QC) the usable input data, pick appropriate parameter choices and separate usable results from low-probability scenarios. Unfortunately, today’s reservoir characterisation and modelling workflow choices are poorly documented. A single poor choice of data or algorithm by a team member that is not detected and resolved will rapidly become an earth model constraint that may impact the interpretations of other team members. These choices will ultimately affect the reliability of the earth model predictions that are used for business unit planning. These shared earth models can be easily created and their results are very believable, but they still suffer from the QC of input data integration, appropriate modelling choices and the level of skill of the multidisciplinary modelling team.
Category:
Reservoir Engineering
|
Download PDF
References
