A new parallel, black-oil-production reservoir simulator (Powers**) has been developed and fully integrated into the pre-and post-processing graphical environment. Its primary use is to simulate the giant oil and gas reservoirs of the Middle East using millions of cells. The new simulator has been created for parallelism and scalability, with the aim of making megacell simulation a day-to-day reservoir-management tool. Upon its completion, the parallel simulator was validated against published benchmark problems and other industrial simulators. Several giant oilreservoir studies have been conducted with million-cell descriptions. This paper presents the model formulation, parallel linear solver, parallel locally refined grids, and parallel well management. The benefits of using megacell simulation models are illustrated by a real field example used to confirm bypassed oil zones and obtain a history match in a short time period. With the new technology, preprocessing, construction, running, and postprocessing of megacell models is finally practical. A typical history-match run for a field with 30 to 50 years of production takes only a few hours.

Summary Equilibrium ratios play a fundamental role in the understanding of phase behavior of hydrocarbon mixtures. They are important in predicting compositional changes under varying temperatures and pressures conditions in reservoirs, surface separators, production and transportation facilities. In particular they are critical for reliable and successful compositional reservoir simulation. This paper presents a new approach for predicting K-values using Neural Networks (NN). The method is applied to binary and multicomponent mixtures, K-values prediction accuracy is in the order of the tradition methods. However, computing speed is significantly faster. Introduction Equilibrium rations, more commonly known as K-values, relate the vapor mole fractions (yi), to the liquid mole fraction (xi) of a component (i) in a mixture, (1) In a fluid mixture consisting of different chemical components, K-values are dependent on mixture pressure, temperature, and composition of the mixture. There are a number of methods for predicting K-values, basically these methods compute K-values explicitly or iteratively. The explicit methods correlate K-values with components parameters (i.e. critical properties), mixtures parameters (i.e. convergence pressure). Iterative methods are based on the equation of state (EOS) and are, usually, tuned with binary interaction parameters. Literature search and experience in the phase behavior of hydrocarbon systems, have shown that current explicit methods are not accurate because they neglect compositional affects. EOS approach requires extensive amount of computational time, may have convergence problems, and must be supplied with good binary interaction parameters. In compositional reservoir simulation where million of K-values are required, the method becomes time consuming and adds to the complexity of simulation studies making some of them impractical. Neural Networks (NN) are emerging technology that seems to offer two advantages, fast computation and accuracy. The objective of this paper is to show the potential of using NN for predicting K-values. Different NN where trained using the Scaled Conjugate Gradient (SCG), and where used to predict the K-values for binary and multicomponent mixtures.

For the last five years parallel reservoir simulation has enjoyed great interest at Saudi Aramco. Realism in modeling the fields in Arabia requires the use of massive simulation models. These massive models capture detailed geological heterogeneity at relatively fine resolution. In addition, some of these fields have been producing for the last 50 years and are expected to continue to produce for decades to come. Today, there are more than 20 massive simulation models ranging in size from 500,000 to 10 million cells and are routinely used in managing the Saudi fields. Models of this magnitude require sufficiently "massive" computational power in order to carry out simulation studies in efficient and reasonable time. Faced with the challenge of meeting computational requirements, we were motivated to find a cost effective computational platform. In this work, we report on our experience in investigating PC-Clusters for mega-cell reservoir simulation studies. Our investigation demonstrated that commodity off-the-shelf Xeon based PC-Clusterscan deliver an exceptional performance at minimal investment in computing hardware and software. Up to 9.6 million cells model running for 50 years have been tested. Currently Saudi Aramco uses PC-Clusters for its simulation work at not more than 15% of the cost of traditional solutions using supercomputers. Introduction The area of parallel reservoir simulation has been extensively investigated with earliest attempts dating back to the late 1980s (Ref.1 has good coverage of work done). Fundamentally, the objective behind parallelization effort is to drive simulation run time to absolute minimal. The needs and benefits behind such effort are numerous: Mega-Cell Reservoir Simulation. Relatively high-resolution million-cell model provide the necessary realism in modeling large or highly heterogeneous reservoirs. A grid size of 50 meters is common in modeling small to medium scale reservoirs. However, use of same grid size in large fields would easily result in models with more than a million cells. At the extreme; a 50×50 meters areal grid size on the Ghawar field (the largest oil field in the world) would require more than 100 million cells. High-resolution grid block swill also capture details of geological heterogeneity providing us with better understanding and modeling of fluid flow. Modeling Complexity. Compositional and dual-porosity-dual-permeability models, by nature of the mathematical formulations require extensive numeric computations. Jacobian building and solver solution require more processing time as the number of unknowns are more. Long Simulation Time. Simulation models for old-fields would also benefit from parallel reservoir simulation. Fields in the Arabian gulf, especially Saudi fields have been producing for the last 50 years and are expected to continue production for decades to come. Such models would require an extended simulation time in "serial" mode but would be more manageable with parallel simulation. Stochastic Modeling. Statistical approach to geological model building results in numerous simulation-model realizations. Parallel simulation can process the models with minimal or no up-scaling and provide us with capabilities to better utilize stochastic tools and define and evaluate risks and uncertainties in field operation. Automatic and Semi-Automatic History Matching. The success of utilizing history matching tools is highly dependent on the speed of running a single simulation and the evaluation of numerous models. Parallel reservoir simulation is a natural complement that provides the necessary simulation performance. Improved Turn Around Time. The longer a simulation study takes the more resource it requires and the more costly it becomes. Simulation runs that take days or weeks to complete can be improved to few hours if the technology of parallel computing is properly utilized.

This paper was prepared for presentation at the 1999 SPE Reservoir Simulation Symposium held in Houston, Texas, 14-17 February 1999.

No abstract