In the oil and gas industry, an optimal approach to field development planning and asset value optimization usually requires complex Integrated Asset Models (IAMs) where reservoirs and gathering network simulators work together coordinated by an orchestrator program. IAMs are developed to answer a wide range of questions posed by the integrated reservoir engineering and petroleum/production team. In general, reservoir engineering evaluations target long-term forecast with emphasis on subsurface uncertainty, which may require Monte-Carlo type analysis to be performed: simplified network solvers based on pre-computed Vertical Lift Performance tables can then be used to run batch of simulations where scalability and computational performance are critical. On the other hand, production/petroleum engineering focus is on network layout, de-bottlenecking and flow assurance issues, and this usually requires running a limited number of IAM simulations using complex, (typically Windows-based) network solvers where multiphase correlations are used on-line to compute pressure drop and fluid temperature changes along the network. Then, for a given asset it is not unusual to have many IAMs available, where not only the network, but even the reservoir simulators and the controller logic may be different according to the specific purpose. This is suboptimal in terms of resources and it may lead to inconsistent answers.
In this work a concept of flexible Integrated Asset Model is presented. The key idea is to use the logic of a modern, high-resolution reservoir simulator (HRRS) to develop a unique IAM where different network models and, if needed, even different reservoir simulators could be hooked up according to the purpose of the simulation. The implementation hinges on HRRS modularity, where two processes, the Field Manager (FM) and the Reservoir Simulator (RS) are integrated, the former one providing boundary conditions and well allocations according to a given strategy, the latter one solving reservoir equations. By means of proper software connectors based on socket technology, the FM acts within the IAM as an orchestrator for a variety of reservoir and network simulation instances, with the possibility to change reservoir and/or network simulator type without modifying the underling development strategy. Notably, the degree of coupling between reservoir instances and network solutions may span from iteratively lagged to periodic coupling depending on system dynamic and simulation efficiency.
The benefits given by this flexible IAM approach are proven first throughout key synthetic problems, where trade-off between process detail and computational efficiency is addressed. Then, application to a real deep-water asset, consisting of six reservoirs connected by complex network which includes manifold, flow lines, risers, lifting systems and top-side floating facilities, are used to show how this flexible solution can be used for real purposes.
Often, components of an oil & gas production asset are treated in a loosely connected manner, with key components like subsurface, gathering network and process plant simulated as stand-alone systems by different professionals. This in principle may prevent the possibility to address interdependencies across components, e.g. global constraints acting on many reservoirs, backpressure amongst well streams in the gathering network, presence of bottlenecks of various type in the flow of the fluids from reservoir to selling point. Interdependencies are critical in deep-water assets.
