Comprehensive Quantification of Uncertainty Parameters and its Role in Integrated Reservoir Characterization of Carbonate Environment

Abstract

Quantification of uncertainty in input parameters to build a robust 3D geological model is an integral and perhaps the most crucial requirement in high-risk exploration areas. This demands more innovative and effective management of uncertainties for optimizing reserve portfolios and better formulation of exploration and exploitation strategies for oil and gas fields. The present area of study pertains to Mumbai/Western Offshore Basin of India. The reservoirs of the study area are challenging due to their high spatio-temporal heterogeneity and discrete fluid distribution. Wells drilled during the field development plan are devoid of hydrocarbon from the lower zone of the reservoir (Middle Eocene/lower Bassein pay) while upper zone (Late Eocene/upper Bassein pay) produced significant amount of gas despite same lithological composition and structural setup, which reduced the utility of pressure-performance based or conventional modeling approach as it couldn't explain the complex geological set up of the deposition. In this background, a thorough evaluation of critical aspect of most complex and anisotropic carbonate reservoir of Bassein Formation of Middle to Late Eocene age has been taken up to delineate the trends of favorable locales in the area. Inputs from micro-facies analysis, fluid transmissibility of Formations and diagenetic imprint analysis were considered to start the present study. An integrated methodology was designed incorporating seismic, well/logs, core samples, sedimentological, bio-stratigraphic & reservoir data to estimate petrophysical properties and necessary modifications in conventional approach were introduced for capturing the reservoir heterogeneity and stochasticity. Hi-frequency digenetic cycle mapping at log scale and pre-stack inversion results (P & S impedance, Vp/Vs ratio) were incorporated to build a robust geo cellular model and characterize the reservoir. Uncertainty analysis presented in this study is mainly focused on structural and petrophysical parameters. The effect of each parameter/factor and their interaction effect (response) with other parameters are analyzed through Optimization Algorithms, to quantify the uncertainties and its impact on reservoir characterization. Sensitivity analysis indicated that Oil Initially In Place (OIIP) exhibits significant sensitivity to effective porosity and water saturation. Therefore, distribution pattern of these uncertainty parameters are derived from Probability Density Function (PDF) and used to restrict the variability of the volumetric estimates to retain the P10/P90 ratios within the acceptable ranges. Quantification of structural parameter was performed using non-linear multiple regression technique, constrained by statistically Maximum Allowable Error (+Standard Deviation). Present analysis enabled us to reduce the uncertainty associated with various reservoir characterization elements. Further, it enhanced robustness of velocity modeling, petrophysical and lithological interpretation through determination of uncertainties with high degree of accuracy and provided their role in estimation of final hydrocarbon-in-place volumes. The parameterization of the uncertainties deliberated could be used as a template in other fields sharing similar structural and depositional characteristics to mitigate the risks associated with Field Development Plan.

Uncertainty Assessment of Reservoir Modeling for Oilfield in the South of Iraq

“…The most influential parameters will be transferred to uncertainty quantification while the parameters with negligible effect will be ignored. Figure (1) shows the Sensitivity analysis of parameters on STOIIP for the geological model. The results illustrate that the oil-water contact's depth is the most predominant parameter influence on the STOIIP then comes net to gross (NTGseed), water saturation (SwSeed), and porosity (PHIseed) respectively.…”

“…Each uncertainty parameters have a special probability distribution which is considered more realistic to describe uncertainty in variables of a risk analysis. Also, the probability density function (PDF) is used to limit variation of the objective prediction to maintain the P10/P90 ratios within the appropriate ranges [1]. A new Bayesian approach was used for history matching and uncertainty assessment.…”