New Tools for New Technology – OASIS: Multiphysics Engineering Simulation Software for “What-If Physics”
Paper Number: WHOC15-167
Source: This paper was prepared for presentation at the World Heavy Oil Congress held in Edmonton, Alberta, Canada, March 23 – 27, 2015
Authors: Carolina Diaz-Goano, Achara Tiong, Heather Herring – Laricina Energy Ltd.
Reservoir simulation models are used extensively for comparative studies and forecasting performance. As more complex reservoirs are developed, the understanding of the physical phenomena occurring in situ needs to be reconsidered. Incorporating new physics in commercial simulators is very time consuming and software companies cannot provide the average user flexibility on the model formulation at a reasonable cost. Therefore, Laricina developed LASIS, a multiphysics engineering simulation software that enables physics equations to be used as plug and play components. It is a living system that evolves incrementally, using interchangeable and extensible models to develop and modify complex formulations with minimal effort, making the studies of “what-if physics” possible. The OASIS Modeling Language (OML) and compiler translate equations into a working model and provide analytic tools to check the consistency of the system and examine the problem formulation. The model formulation, numerical engine, user interfaces and other components are each defined by an abstract programming interface. Incorporated features ensure that only physically plausible answers are calculated, boundary conditions can be set and any condition can be defined to stop the simulation. Different linear and non-linear solvers are integrated including open source code developed at the University of Florida and the French Institute for Research in Computer Science and Automation (INRIA) with the flexibility to incorporate others. The choice of equations is transparent and changes are auditable. Input data can be loaded through the OML script or an external ASCII file. The ability to selectively record any or all computational output is incredibly powerful for debugging. All results are saved into a database thus enabling cross plotting and analysis during or after the simulation. Database queries are done through a filtering mechanism. The cloud based capability allows scalability without additional infrastructure investment, unlimited storage, system integrity backup and recovery.