Autonomic oil reservoir optimization on the Grid

Vincent Matossian, Viraj Bhat, Manish Parashar, Ma?gorzata Peszy?ska, Mrinal Sen, Paul Stoffa, Mary F. Wheeler
2004 Concurrency and Computation  
The emerging Grid infrastructure and its support for seamless and secure interactions is enabling a new generation of autonomic applications where the application components, Grid services, resources, and data interact as peers to manage, adapt and optimize themselves and the overall application. In this paper we describe the design, development and operation of a prototype of such an application that uses peer-to-peer interactions between distributed services and data on the Grid to enable the
more » ... autonomic optimization of an oil reservoir. V. MATOSSIAN ET AL. possible for scientists and engineers to conceive a new generation of applications that enable realistic investigation of complex scientific and engineering problems. These applications will symbiotically and opportunistically combine computations, experiments, observations, and data, and will provide important insights into complex systems such as interacting black holes and neutron stars, formations of galaxies, subsurface flows in oil reservoirs and aquifers, and dynamic response of materials to detonations. However, such a global scientific investigation requires continuous, seamless and secure interactions where the application components, Grid services, resources (systems, CPUs, instruments, storage) and data (archives, sensors) can interact as peers. In this paper we focus on one such application: the autonomic oil production management process. Such a process involves (1) sophisticated reservoir simulation components that encapsulate complex mathematical models of the physical interaction in the subsurface, and execute on distributed computing systems on the Grid; (2) Grid services that provide secure and coordinated access to the resources required by the simulations; (3) distributed data archives that store historical, experimental and observed data; (4) sensors embedded in the instrumented oilfield that provide real-time data about the current state of the oil field; (5) external services that provide data relevant to optimization of oil production or of the economic profit such as current weather information or current prices; and (6) the actions of scientists, engineers and other experts in the field, the laboratory, and in management offices. These entities need to dynamically discover one another and interact as peers to achieve the overall application objectives. First, the simulation components interact with Grid services to dynamically obtain necessary resources, detect the current resource state, and negotiate the required quality of service. Next, we recall that the data necessary for reservoir simulation are usually sparse and incomplete; in particular, this concerns the data on the geology of the subsurface and on the resident fluids which are very difficult to obtain. Therefore, the simulation components interact with one another and with data archives and real-time sensor data to enable better characterization of the reservoir through processes of history matching, dynamic data injection, and data driven adaptations. Next, the reservoir simulation components interact with other services on the Grid, for example, with optimization services to optimize well placement, with weather services to control production, and with economic modeling services to detect current oil prices so as to maximize the revenue from the production. Finally, the experts (scientists, engineers, and managers) collaboratively access, monitor, interact with, and steer the simulations and data at runtime to drive the discovery process. The overall oil production process described above is autonomic in that the peers involved automatically detect sub-optimal oil production behaviors at runtime and orchestrate interactions among themselves to correct this behavior. Further, the detection and optimization process is achieved using policies and constraints that minimize human intervention. The interactions between instances of peer services are opportunistic, based on runtime discovery and specified policies, and are not predefined. In this sense, the process is self-managing, self-adapting, and self-optimizing. In this paper we describe the development and operation of a prototype application that uses peer-topeer interactions between applications and services on the Grid which enable autonomic optimization of oil and gas recovery from a subsurface reservoir. Here we assume that the information on the properties of this reservoir is complete and adequate. The paper has three key objectives: (1) to provide a proof-of-concept implementation to evaluate the correctness and feasibility of the formulation and methodology, (2) to use this sample application to demonstrate the feasibility and benefits of such selfoptimizing applications, and (3) to develop a prototype peer-to-peer (P2P) middleware substrate that provides the functionalities required to enable the identified application interactions. The prototype
doi:10.1002/cpe.871 fatcat:wverkbh4qzfwzauekr2gyuyggu