Active reservoir management starts with the observation of the dynamic state of reservoirs and subsequent implementation of an optimal control strategy based on that state. To this end, ever more fields are equipped with downhole gauges, inflow control valves and other smart or intelligent completions: so-called smart fields, intelligent fields or e-fields. Current use of the capabilities of these smart fields is quite limited.
In the short term, only those observations that directly translate into an appropriate response and lead to "actionable events" (e.g. water breakthrough, start-up failure) tend to be used while the period between updating reservoir models (and subsequently the control strategy) is often several years despite the constant availability of new observations. Solving such problems can add some value to reservoir management but the ultimate goal is to get an integrated solution through closed-loop or real-time reservoir management. To provide this integrated solution, TNO has developed a "Reservoir Management Toolbox" that includes state-of-the-art assimilation and optimisation modules that may be used as stand-alone history-matching or field development planning tools or can be combined into a closed-loop reservoir management workflow. The toolbox also includes advanced up- and down-scaling methods that help convert detailed geological models into manageable reservoir models.
To gain experience with implementing closed-loop reservoir management tools in a realistic field and to test different workflows, TNO organised a unique benchmark study for the SPE ATW on Closed-Loop Reservoir Management in Brugge in 2008. Nine groups from both universities and industry participated, their goal being to 1) optimise a reservoir model using the data provided over a 10-year production period, and 2) come up with an optimal waterflooding strategy for the next 10 years. This production strategy was tested on the truth model and the resulting production data sent back to the participants to update their models and determine a strategy for the remaining lifetime of the reservoir. The geological model was upscaled to a reservoir model on which all the different water flooding strategies provided by the participants could be tested. The addition of 100 reservoir realisations allowed participants to either build their own reservoir models or use one of these 100 models.
The results from the workshop showed a spread of the Net Present Value (NPV) obtained by the different participants in the order of 10%, with the highest result only 3% below the optimised case determined for the known truth field. Although not an objective of this exercise, it was shown that the increase in NPV as a result of having three control intervals per well instead of one was significant (around 20%). The results also showed that the NPV achieved when additional production data became available was consistently higher than before the data became available. The Brugge field is currently still held by TNO and is not available for third parties. However, to make the Brugge field more accessible and allow closed-loop studies with a higher level of interaction, we are currently working on making the model available for running via web-services. Initially only for our partners NTNU and Stanford University, but in a later stage probably also for a wider public.