Directly to ...
Motivation and background
Numerical optimization is seen as a valuable technology for decision support in various stages of the life cycle of hydrocarbon fields. Its potential has been demonstrated in previous benchmark studies such as the 2008 Brugge study on Closed-Loop Reservoir Management. Here we review the learnings from that study and motivate the choice for the set-up of a new benchmark study.
The objective of the Brugge case study was to demonstrate the value of managing reservoirs using a near-continuous application of system control concepts and of advanced data assimilation and optimization techniques in a closed-loop framework. Most participating parties were able to perform two history matching and optimization steps at 10-year intervals, thus completing the full loop twice. A number of conclusions could be drawn from this study (Peters et al., 2010):
- - Participants using advanced workflows were generally able to outperform participants working with industry-standard workflows as implemented in commercial software tools available at that time
- - Taking uncertainty into account during all stages of the closed loop has clear benefits towards creating increased realized value
- - Going through the loop more frequently leads to increased realized value.
There were a number of problems identified with the design of the study (Peters et al., 2012):
- - It turned out to be difficult to draw firm conclusions on the relative performance of different algorithms used in the individual steps (i.e. data assimilation or optimization) since only the data mismatch and the final outcome were evaluated and compared.
- - It was difficult to set up a fully automated workflow for (online) evaluation of model updates and proposed operating strategies, thus requiring a significant amount of time for manual evaluation by TNO staff.
- - This latter point, combined with the labor-intensity for the participants of applying their workflows, also meant that is was not possible to complete the closed-loop multiple times with all participants.
- - The development of the field in terms of number, type and placement of wells was not part of the optimization while this may be expected to have a major impact on the value that can be realized
Taking these points into account, an improved closed-loop exercise would at the very least involve a fully-automated online system accessible to all parties that is able to evaluate proposed strategies and generate and return measurements. While in principle this is technically possible, comparison of individual algorithms would remain difficult - if not impossible - since every participant would be confronted with different measurements. Secondly, the first step of any reservoir management workflow would ideally be an optimization of the field development plan. There is therefore clear value in comparison benchmarks for history matching and/or optimization separately.
Over the past 10 years many studies have appeared on numerical optimization of well controls such as rates, pressures and Interval Control Valve (ICV) settings. These studies have investigated optimization algorithms, handling of uncertainty, constraints and multiple objective functions, and most recently also various measures of risk. The types of controls that have been considered have been extended to also include various field development parameters such as well positions and trajectories and, more recently, drilling schedule. Some of these controls have even been considered jointly. While for typical well controls a number of approaches have emerged as more promising than others, for field development-type problems no clear consensus on best practices has emerged yet. We believe that further experience with, and comparison of, different approaches would therefore be very useful. Therefore, on these pages an optimization benchmark challenge is proposed aimed at developing further insight into the value of field development optimization under geological uncertainty and into the potential of different methods to perform this optimization.
The optimization benchmark challenge is aimed at field development (FD) optimization under uncertainty. History matching, e.g. as part of a closed-loop workflow, will not be considered. Questions that we aim to address in this benchmark challenge are:
- - What added value can be expected if optimization methods were applied to make field development decisions?
- - What are good workflows to arrive at optimal development plans?
- - Which controls should be considered to construct an optimal development plan?
- - Which methods are best suited for field development optimization?
- - Should well placement and control be considered jointly?
These questions can be answered by formulating and addressing a number of well-defined challenges that are described on the problem statement page.
The scope as defined above introduces the need to address a number of technical issues that have not been part of previous benchmark studies:
- - Some controls may naturally appear as integer or binary variables. For example, well positions and drilling ordering are often thought to require treatment as integer controls.
- - The number of wells may not be constant throughout an optimization process.
- - Incorporating drilling costs into the objective function, and considering the well drilling order, will lead to non-smoothness of the objective function.
- - Evaluating different well trajectories requires frequent re-computation of well-reservoir connectivity.
- - Joint well placement and control optimization is a mixed-control problem that is expected to be challenging for some workflows and algorithms.
- -The FD problem will involve time-dependent nonlinear input and/or output constraints that could be handled by the simulator or, more formally, by an optimizer.
- - There is typically high uncertainty involved during the early FD stage. The risk associated with uncertainty is probably a more important driver for field development decisions than for decisions on well control strategies.