No abstract

Reservoir relative permeability and capillary pressure, as a function of saturation, is important for assessing reservoir hydrocarbon recovery, selecting the well completion method, and determining the production strategy because they are fundamental inputs to reservoir simulation for predicting lifetime production of a well. Estimation of relative permeability and capillary pressure curves at reservoir conditions is also an important task for successful planning of waterflooding and enhanced oil recovery. The relative permeability and capillary pressure data estimated from core analysis might cause concern regarding representativeness, and adjustments are typically necessary for successful production forecasting. This paper proposes a new method to obtain relative permeability and capillary pressure curves with downhole pressure-transient analysis (PTA) of mini-drillstem tests (miniDSTs) and well log-derived saturations. The new approach was based on performing miniDSTs in the free water, oil, and oil-water transition zones. Analyses of the miniDST buildup tests provided absolute formation permeability, endpoints of relative permeability to both oil and water, and curvature of the relative permeability data. Additionally, resistivity, dielectric, and nuclear magnetic resonance (NMR) logs were used to determine irreducible water, residual oil, and transition zone saturations. Combining these downhole measurements provided the relative permeability and capillary pressure curves.

A detailed Geological and Petrophysical characterization was achieved in a stepwise approach as part of full field 3D Reservoir Modeling and Simulation study for Minagish reservoir in the Greater Burgan field in Kuwait.Foundation of Reservoir Rock Types (RRT) was developed in first step based on Mercury Injection Capillary Pressure (MICP) dataset. A combination of Discriminant Analysis and Indexed Self Organizing Map (SOM) was used for rock type classification using hyperbolic tangent method. To improve classification of bimodal Pc curves, additional pressure values at different non-wetting phase saturations were used in conjunction with above mentioned parameters. In second step, the available Routine Core Analysis (RCA) porosity, permeability data was grouped together based on common patterns to generate rock types in RCA domain. Blind tests in two of the cored wells revealed a conformance of 81% between MICP and RCA Petrophysical Groups (PG). In the final step of the process, petrophysical groups were propagated in log domain using available log measurements common in all the wells of the field. It was challenging to establish a high level of accuracy for PG's in log domain mainly due to fine scale heterogeneity and inability of log data to capture rock fabric variation.This porosity estimate, coupled with rock type classification, helped to derive a continuous permeability log with a correlation coefficient of 0.89 validated in key cored wells. The porosity and permeability data in all the wells was incorporated in the 3D geocellular model after up-scaling honoring the unique, per rock type, Phi-K relationship.Modeled capillary pressure curves generated for each rock type in the core domain using MICP data set in 3 wells were used in saturation height modeling. The modeled equation was captured in the 3D geocellular model after populating rock types in the 3D grid to map water saturation for volumetric estimation.

Downhole sampling and laboratory analysis are key complementary techniques offering a step change in fluid characterization. It is generally accepted that fluids in the reservoir are in chemical equilibrium. However this assertion, although convenient, is often invalid, for several reasons from source variations, due to in-reservoir reaction, degradation or precipitation. In a large majority of cases many of these phenomena are difficult to appreciate, because the techniques used for fluid sampling often lack the pin-point acquisition accuracy (geographical, depth and time) required to provide the information at a level of accuracy sufficient to detect subtle variations. Instead only a mixture of different oil is captured providing "averaged" oil composition and characteristics, whereas the reality in the reservoir may be markedly different. The vertical compositional gradient in oil column has been documented in many large oil column reservoirs. The geographical variation has been much less documented. In this paper we will demonstrate how variation at the scale of individual sand bodies in the reservoir, sometimes only tens of feet apart, can be very large (as observed in two sand bodies in the same well) and have a dramatic impact on oil property understanding and modeling. In turn understanding of these property variations and their impact on mobility, is a key factor for understanding fluid flow and for the choice of the correct secondary recovery mechanism for an improved recovery. The example of the Wara channelized sand, within the Greater Burgan complex, will illustrate the paper, where downhole fluid analysis is presented as a necessary complementary tool to improve the selection and construction of an accurate set of samples to ensure the data collection is complete and exhaustive before the well is completed.

A study was designed to confirm the formation properties obtained from available conventional RCA data and inferred from corrected wireline log data using digital rock analysis (digital RCA and SCAL analysis) on cores from the Greater Burgan field. This study was performed for Kuwait Oil Company, Fields Development Group (S&EK) by FEI Digital Rock Services in 2014. As part of this study, 27 feet of whole core, from the Lower Ahmadi (AHL2) to Upper Wara (WU1) formations, were imaged by X-ray computed tomography (CT) imaging, including 1 foot of partially preserved core. 14 plugs were extracted from these cores and imaged in 3D by a high resolution helical micro-CT. Analysis revealed stark differences in mineralogy, grain size and sorting and the presence of severe fracturing in some plugs due to the fragility and friability of the rock. Sub-plugs were extracted from 10 of the 14 plugs (including one sub-plug from the partially preserved section) and imaged in 3D by helical micro-CT. 7 of the sub-plugs proved suitable for digital RCA and SCAL analysis. The 3D images were used to calculate digital RCA properties (porosity, permeability, grain density, grain size distribution and formation factor) and pore network models were built to perform digital SCAL simulations and predict multiphase transport properties such as Pc, kr and resistivity index for primary drainage and imbibition. In addition, the in situ mineralogy of each plug was analysed using 2D SEM-EDS automated, quantified mineral mapping. The mineral maps, combined with BSEM images, contain rich textural information and were registered into perfect geometric alignment with the 3D micro-CT images. A tight rock workflow was used to identify sub-resolution porosity in 3 of the plugs. Experimental MICP curves showed that substantial portions of the pore throats were below the image resolution, caused by large amounts of pore-filling materials. Hence, pore scale information could not be directly extracted from some images. Consequently, process based modelling was carried out on two plugs to generate pore-networks. A quasi-static pore-network model was used to simulate oil/water displacements and predict multiphase transport properties. Detailed imaging of oil-in-place and porosity was performed on a partially preserved plug to create a map of remaining oil which revealed that oil was retained in most porous grains and strongly retained in clay-rich zones. The digital core analysis results are in agreement with available log and core data. The Lower Ahmadi (AHL2) section is good quality in terms of porosity, permeability and flow properties, whereas the Upper Wara (WU1) section is of poorer quality.