Reservoir Study of the Largest Oil Field in Argentina – A Two Reservoir 2200 Well Simulation Model”, Claudio Marcelo Fonseca, Gustavo Fernandez, Repsol-YPF, Tuba Firincioglu, Chet Ozgen, Alejandro Albertoni, NITEC LLC, SPE ATCE, September 2004. SPE 90952

Abstract

Reservoir Study of the Largest Oil Field in Argentina – A Two Reservoir 2200 Well Simulation Model

This paper describes the processes incorporated during the simulation study of the Chihuido de la Sierra Negra Field in Argentina, which is presently the largest oil field in the country. The field was discovered in 1968, and an aggressive water injection program was started in 1995. So far the oil production from this field is 65 million cubic meters and the water injection is 158 million cubic meters.

The field has two separate stacked reservoirs that are structurally complex with different fault systems and fluids. Both reservoirs are highly compartmentalized with each compartment having different water-oil contact. The upper reservoir is composed of 6 isolated geologic layers, with each being different depositional environments (sand bars, dunes, fluvial channels, etc.). The presence of unconformities and volcanic intrusions further complicate the reservoir communication and plays an important role in flow behavior.

Currently, the field has 734 production wells with commingled production from both reservoirs and the isolated zones with in the upper reservoir. Approximately 1250 injection strings are currently injecting water.

The purpose of this study was to provide a detailed reservoir characterization to optimize recovery and to create a simulation model with predictive capability that can be used in improving field management. To achieve this purpose a 900,000 cell simulation model was constructed.

This paper will discuss some of the challenges encountered during the history matching of the field model. The identification of the compartment boundaries and their associated water oil contacts (57 contacts) required the development of a new consistent approach. This method provided significant accuracy and time savings over the traditional approach of iterating between history matching and reservoir characterization. The overwhelming volume of data and the volume of simulation results for 2200 well strings required special considerations for pre- and post-processing. New tools were needed to quickly modify the simulation arrays and review all the wells in an efficient and timely manner. As the history matching progressed many additional practical tools were developed. This paper will discuss the significance of these innovations and tools to achieve a successful history match in a timely manner.