Mar 29, 2022 | 09:00 AM CST
March 29, 2022 | 9:00 a.m. - 10:00 a.m. CST (UTC-05:00)
3D reservoir models are now mainstream tools for forecasting both production and injection (storage) projects. Despite this, many models still disappoint, and we maintain a systematic tendency towards optimistic forecasts. There are ‘good models’ and ‘bad models’ and it is argued that the difference rests primarily on the choice of workflows: the central issue of model design.
This talk draws on experiences accumulated over the last 30 years from working with international and national oil companies, and draws out three common themes which tend to distinguish ‘good’ models from ‘bad’:
Model purpose, and the balance of simplicity vs. complexity in modelling
Imperfect data and the requirement for concept-based modelling – “If you can sketch it, you can model it”
Taking a fluid-centric view of the need for detail: reservoir modelling, rather than geological modelling
The scale transition, restoring 'lost heterogeneity'
The points above will be illustrated with a case study from a mature oil field undergoing waterflood, where a decision for infill drilling is to be made. The talk will close on the application of these themes to the (generally more difficult) issue of modelling for storage.
TRACS International & Heriot-Watt University