No abstract

Optimal exploitation of hydrocarbon reservoirs has always been a challenge due to uncertainties posed by subsurface heterogeneities that are often not factored into field development plans. Secondary and tertiary recovery mechanisms, such as waterflooding and enhanced oil recovery (EOR), are used to enhance the oilfield recovery beyond primary recovery. However, as the field development transitions to secondary/tertiary mechanisms, the challenges in monitoring these mechanisms further increase the uncertainty in field development. If these uncertainties are not reduced or incorporated properly, the field development may easily become uneconomic. This work presents a workflow that addresses the limitation of regular waterflood surveillance while characterizing the reservoir for optimal exploitation. The current technologies for waterflood surveillance are limited either to local surveillance methods, such as tracers, crosswell seismic and crosswell electromagnetics (EM), or to uncalibrated global realizations, such as full-field streamline simulation, with no validation between the wells (It is to be noted that a full-field reservoir simulation calibrated with production-injection data in defined time-interval is stated as a global-surveillance method in this paper). This workflow devises integration of an effective local waterflood monitoring method, crosswell EM, and a global waterflood modeling method, streamline simulation. The process of validating the parameters of a geological model and a dynamic model with time-lapse crosswell EM data significantly reduces reservoir characterization uncertainty and helps in the preparation of a precise dynamic model.

Reservoir-fluid properties play a key role in exploration and field development planning as accurate fluid characterization is important for designing reservoir development strategy, optimizing well completion, optimizingthe production system and efficient reservoir management. Characterization of reservoir fluids with more complex behaviors, such as gas condensates and near-critical fluids can be technically challenging, especially in a deepwater environment as reservoir development planning and production facility design is contingent on getting an accurate description of the reservoir fluid. Fluid characterization begins with the collection of representative formation fluid samples during initial wireline formation testing, bottomhole sampling and during conventional well testing operations. Traditionally these fluid samples are sent to offsite laboratories for sample analysis. However, characterizing a gas condensate fluid system based on a single sample set can be potentially misleading since PVT properties of samples acquired across a reservoir may be different due to spatial variation in their components and compositional grading. In this technical contribution we present a case study to demonstrate that characterizing gas and gas-condensate systems using only a basic set of measurements and from analysis of pressure gradients alone; could lead to potentially ambiguous results, an inappropriate fluid model or misinterpretation. In our study, we describe the practical application of advanced wireline formation sampling and testing techniques in combination with downhole fluid analysis, and their integration with laboratory PVT studies and equation-of-state (EOS) models. We describe the importance of a comprehensive data collection and fluid analysis plan early in the exploration/ appraisal process, and illustrate how high-quality fluid sample data and property measurements from advanced formation sampling and testing tools in combination with conventional well testing techniques can add significant value by helping to reduce uncertainty and aiding better technical decision making.

Oil India Limited's (OIL) operational areas, in Upper Assam-Arakan Basin, are located in a seismically active thrust fault zone (Bora et al., 2010). Multiple stacked layers, highly faulted anticlines and large number of compartments make the structural setting of these fields very complex. In terms of lithology, some of these reservoirs are low resistivity pays, leading to ambiguities in interpretation due to fresh water environment and complexities in evaluation of hydrocarbon-bearing and water-bearing sands (Koithara et al., 1973, Borah et al., 1998). Greater Nahorkatiya and Greater Jorajan, since the inception of commercial production in the 1950s, have been intensively studied to find prospective sweet spots, perforation intervals for new well locations and potential workover candidates. These forecasts, guided only by dynamic numerical model results, have had mixed results when implemented in the field. A validation of the dynamic model forecasts with near-wellbore saturation logs, can help to reduce uncertainty. This paper describes the success stories in field implementation of workovers, guided by dynamic reservoir model results and cross-validated with Pulse Neutron Tool (Roscoe et al., 1991, Schnorr, 1996) log recordings. The intricacy of delivering a precise dynamic reservoir model was managed by state-of-the-art seismic-to-simulation workflows, an integrated approach to improve the quality of the geological model and specific analytical techniques to fill in the data gaps. The calibrated model was analyzed for workover opportunities of zone transfer. In zones with high confidence, (i.e. high Hydrocarbon Pore Volume (HCPV), high porosity, permeability, etc., perforation intervals were recommended for hydrocarbon saturation monitoring to confirm the near-wellbore saturation predicted by the model. This workflow was followed in 8 wells which added immense value both technically and economically. The validation of model predictions with near-wellbore saturation was carried out in old wells which helped in making informed decision about tapping bypassed hydrocarbon pockets. It helped to avoid non-hydrocarbon bearing zones, which were removed from the existing workover plan. Moreover, it introduced confidence in the dynamic model which will be used in future for more aggressive economic development of the fields. This approach resulted in better understanding of the reservoir characteristics which led to identification of some potential reserves which could be characterized as "Reserve Growth".

For planning the operations of Oil and Natural Gas Corporation Limited (ONGC) in the complex Heera field, it was estimated that over one hundred simulation runs would be needed to complete the history match of the field and almost the same number of simulations would be needed for production forecasting. Heera is a large field, with multiple faults and seven stacked carbonate formations. There are significant variations in petrophysical properties, and variable degrees of communication between reservoir zones. The simulation models include 479 wells with commingled production or injection. Well trajectories are complex and include multilateral and horizontal configurations. Field development options include use of simultaneous water alternating gas (SWAG) for enhanced oil recovery. Combining all these features, it would be difficult to run all the necessary sensitivity cases within the required project timeline, using a conventional reservoir simulator. Therefore, it was decided to test the applicability of a new generation simulation tool to address the challenges of the study. To ensure that the change of simulator would not impact the integrity of the model, rigorous quality checks were performed on the input data. After successful evaluation, the new software was used for the reservoir engineering study. The decision to apply the new simulator significantly reduced the elapsed time, with some realizations over 20 times faster compared to the original base case. As a result of this speed-up, numerous runs could be carried out to refine the history match. Multiple sensitivities could be used to help understand and reduce the uncertainties in a more comprehensive manner. Moreover, the prediction cases could be optimized to identify the best recovery strategy. This study has demonstrated the value of reducing simulation run times, to complete the project with greater efficiency and more confidence in the results. In future studies, high performance software tools can also enable use of fine resolution models, to capture detailed heterogeneities and optimize areal and vertical sweep.