You will receive 2 emails. One confirming registration and one to register at gotowebinar. There is no fee for this but you must register at gotowebinar to receive the email with the personal link. The approximate speaker start time is 12pm, but an update will be provided at that time on the webinar.
We present a general technology for simultaneous optimization of a broad range of field development decisions with subsurface uncertainties.
We optimize discrete and continuous decision variables such as the number of production or injection wells, their locations, perforation intervals, drilling schedules, well rates, etc. As a novel approach, we include additional categorical variables such as depletion strategy, well pattern, or facility size in the optimization process. We consider a limited number of discrete scenarios for each categorical variable (e.g., primary depletion, gas injection, or water injection as three development scenarios).
Field development constraints on well locations, rig schedules, economic risks etc. are incorporated in the optimization. Hydrocarbon recovery or some economic indicator can be used as the objective function for the optimization and applied for ranking the field development options.
Subsurface uncertainties are represented by incorporating multiple reservoir models in the optimization process. Ideally, all reservoir models in the ensemble should be evaluated for every considered field development option to define cumulative probability functions. However, this would make CPU demands very large in some cases. We propose two effective approaches to reduce CPU requirements: (1) one reservoir model is run to test the optimization criterion, and the remaining models are only run if the objective function is significantly improved; or (2) a novel application of a statistical proxy procedure to define a subset of the reservoir model ensemble that is run during the optimization cycle. The efficiencies gained with these techniques allow us to incorporate the additional decision variables in the full optimization process.
Our results indicate that the proposed algorithms sufficiently reduce CPU requirements to effectively handle field development optimization problems with many reservoir models representing subsurface uncertainties.
The developed methodology have been effectively applied in many fields for simultaneous optimization of well placement, drilling schedule, well production/injection rates, perforation strategy, injection strategy, and facility modifications. They have also been successfully applied in giant oil/gas fields optimizing general field development scenarios. Previously published applications in oil fields in North Sea, Gulf of Mexico, Russia, and Azerbaijan have illustrated the following:
Field development scenarios have been identified (from evaluations of thousands of options) with significant benefits in oil recovery and economic indicators as compared to “manually” derived reference cases.
>Impacts of subsurface uncertainties on field development decisions have been estimated.
>Powerful capabilities of the developed field development optimization technology have been demonstrated.