Forecasting the Productivity of Thinly Laminated Sands with a Single Well Predictive Model

Abstract

As the cost of exploration wells continue to escalate, we need more than ever to evaluate each well quickly and efficiently to improve the appraisal process and avoid unnecessary expenditure. At the same time, an accurate reservoir characterization is the key to successful reservoir development. This is especially true in thinly laminated reservoirs which exhibit vertical heterogeneity and a wide range of flow properties. Therefore, it is critical to combine high resolution formation evaluation logs and formation tests to predict the well performance prior to the production test. We present an integrated and structured approach for calculating the productivity of a laminated clastic reservoir and we illustrate the method with a field example from Malaysia. A single well predictive model incorporates logs, rock and PVT data, and formation tests to build a flow simulation model at the resolution of the petrophysical analysis. By calibrating the high resolution flow model with dynamic test data from a formation tester Interval Pressure Transient Test (IPTT), the model can be used to predict the well performance. We investigate several key characteristics of thinly laminated reservoirs that affect the well productivity, such as vertical communication between layers. In particular, we examine the effects of clay, silt and sand laminations geometry on the reservoirs productivity. For that purpose, we comment on the information from borehole electrical images, NMR logs, single probe and dual packer wireline formation testers, and production well tests. The workflow is fast to implement as it can be accomplished quickly and efficiently after the well is drilled, in time for planning the well completion and production tests. The high resolution simulation model permits to conduct further engineering studies, whenever required, such as designing the injection and production test for multi-layer reservoirs and water or gas coning studies. Introduction These Deepwater turbidite shallow marine to lower coastal plain reservoirs are composed of interbedded porous/permeable sands with varying percentages of interbedded silt and clay beds. These reservoir sands vary in thickness from millimeter to meters. The reservoirs sands may be highly permeable, but the silt and clay laminations affect the reservoir vertical permeability in each layer. As a result, there are significant vertical heterogeneities in theseis types of reservoirs. It is widely known that the conventional logs may not be able to detect these thinly bedded reservoirs due to their limited insufficient vertical resolution. Therefore, new technology logging services, such as tri-axial resistivity, high resolution oil-base borehole images and nuclear magnetic resonance (NMR) logs have been increasingly used over recent years to help determine an accurate reservoir pay thickness (Ref. 1). In addition to these new logging techniques, interpretation methods such as Log Enhanced Resolution using Borehole Image (SHARP analysis) have been developed to improve the reservoir characterization of these thinly bedded reservoirs (Ref. 2 and 3). After reservoir characterization, other frequently asked questions for thinly bedded reservoir are:What is the productivity of a well drilled in this type of reservoir?What is the connectivity between wells drilled in this type of reservoir? Answers of these questions allow us to evaluate reservoir recoverable reserves. Traditionally, a full scale well test and an interference test are conducted to determine well productivity and well-to-well connectivity, respectively.