Correlation Techniques, Perforation Strategies, and Recovery Factors: An Integrated 3-D Reservoir Modeling Study, Sirikit Field, Thailand

Abstract

Production in the Sirikit oil field of the Phitsanulok basin, Thailand, commenced in 1982. The field has an STOIIP (stock tank oil initially in place) of approximately 800 MMbbl (million barrels). To date, more than 130 wells have been drilled and over 100 MMbbl produced. Reservoir management studies into optimizing recovery and identifying unswept oil volumes are ongoing so as to maximize the venture's profitability. This work includes 3-D (three-dimensional) static and dynamic modeling. This study is one such exercise whereby different geological correlation techniques and perforation strategies were analyzed for their potential impact on recovery factors. The Sirikit field and adjacent satellite fields are composed of predominantly lacustrine mouth-bar and fluvial deposits. Two correlation techniques, lithostratigraphy and chronostratigraphy, were employed to generate two 3-D reservoir architectures using the Shell proprietary 3-D static reservoir modeling system GEOCAP. Significantly different recovery factors are derived from 3-D dynamic flow simulation of these architectures (no vertical upscaling, 2 X 2 areal upscaling) within the Shell proprietary reservoir simulator MORES. When good-quality sands observed on well logs are perforated in both models, the lithostratigraphic correlation model results in an absolute 3% higher recovery factor than that of the chronostratigraphic correlation model. In relative terms, this represents a 43% higher recovery factor. This suggests that if a lithostratigraphic correlation model is assumed, then for this perforation strategy, the recovery factor would be overestimated by 3% absolute and 43% relative, if the chronostratigraphic interpretation represents the actual subsurface architecture. In addition, the architecture of the chronostratigraphic model results in the lateral correlation of the thicker upper mouth-bar sands with good-quality, thin-bedded toeset sands and shales (heterolithics). These thin-bedded sands are only marginally resolvable on wireline logs and have not generally been perforated in the Sirikit field. When these thin-bedded heterolithics are additionally perforated in the 3-D model, an absolute increase in recovery factor of 2% is observed over the same model that has only the upper mouth-bar sands perforated. In relative terms, this represents a 28% improvement in the recovery factor. This work confirms and partially quantifies the impact that the choice of mouth-bar correlation style can have on recovery factors as predicted by 3-D dynamic reservoir models. However, the chronostratigraphic model described in this paper represents an end-member connectivity scenario whereby sand-body continuity is relatively low due to continuous dipping shale barriers. Further work is required to quantify the recovery factors for the whole spectrum of potential chronostratigraphic model connectivity scenarios.