Sanjay Srinivasan (ssriniva@mail.utexas.edu) is the Program Manager of the Integrated Reservoir Characterization research program.
The focus of the Integrated Reservoir Characterization program is on:
Providing practical approaches for assessing and managing geologic and flow related uncertainty using improved physical and stochastic models
Providing insights into the physics of fluid flow through multi-scale heterogeneous media
Formation Evaluation Research Program
Principal Investigator: Carlos Torres-Verdin
Broadly speaking, the goals of the Formation Evaluation (FE) Program can be summarized as follows:
- Develop algorithms for the numerical simulation of wireline logging, logging-while-drilling, and cased-hole logging measurements.
- Develop numerical algorithms for the simulation of multi-phase fluid flow phenomena in the near-borehole region.
- Develop numerical algorithms for the inversion of near-borehole measurements into physical properties and petrophysical variables.
- Develop algorithms for the numerical simulation and interpretation of measurements acquired with in-situ permanent sensors.
- Develop numerical algorithms for the joint interpretation of a multitude of single-well and multi-well logging measurements.
- Develop numerical algorithms for the integration of 3D seismic and well-log data.
- Develop numerical algorithms for the quantitative integration of well-log, 3D seismic, well-test, and production data.
- Undertake innovative data interpretation projects that call for the integrated us of well logs, geological information, 3D seismic data, and production measurements.
- Undertake laboratory measurements of fluids and porous rocks aimed at estimation petrophysical properties.
Please see the Formation Evaluation page for more details.
Current DOE Projects
Combined Borehole Seismic and Electromagnetic Inversion for High-Resolution Petrophysical Assessment of Hydrocarbon Reservoirs
Carlos Torres-Verdin (cverdin@mail.utexas.edu)Funding amount: $800,000 for the period of Jan. 2005 - Dec. 2008
This research will investigate a new generation of deep-sensing borehole instruments that can provide three-dimensional images of the subsurface far from oil and gas wells. Development of such improved characterization methods is vital for the efficient production from domestic oil reservoirs.
Integrated, Multi-scale Characterization of Imbibition and Wettability Phenomena Using Magnetic Resonance and Wide-Band Dielectric Measurements
Mukul M. Sharma, Carlos Torres-Verdin, Steven L. Byant (in collaboration with Professor George Hirasaki at Rice Unversity) (cverdin@mail.utexas.edu)Funding amount: $800,000 for the period of Oct. 2004 - Sep. 2007
This research will use advanced dielectric and NMR measurements to understand and predict the flow of oil and water in reservoir rocks. Such improved characterization methods are vital for the continued production from existing domestic oil reservoirs at low cost and high efficiency.
Interwell Connectivity and Diagnosis Using Correlation of Production and Injection Rate Data in Hydrocarbon Reservoirs
Jerry Jensen (Texas A&M University), Larry W. Lake (Larry_Lake@mail.utexas.edu)Funding amount: $179,000 for the period of Sept. 2003 - Aug. 2006
This work aims to develop a new approach to evaluating the flow paths between injection and production wells. The procedure will use injection and production rates and target three different production scenarios: fields with wells shut in for extended periods; fields with non-uniform compressibility; and very heterogeneous reservoirs.
Other Research Projects
Field-Scale Reservoir Modeling - Sanjay Srinivasan
Developing an Improved Methodology for Quantification of Geological Information
The approach consists of developing a digital repository of reservoir models
classified on the basis of reservoir depositional environments. These analog
reservoir models will be constructed using rock outcrop data interpreted by
expert geologists. The resultant 3-D analog reservoir models can be processed
through statistical pattern recognition schemes to extract the multiple point
information specific to that type of depositional environment. The extracted
information can then be applied in conjunction with reservoir specific
information for an unknown reservoir, to develop geostatistical reservoir
models for that reservoir. The constraint is that the target reservoir exhibits
similar depositional features as the analog. More info available from the Field Scale Modeling of Fracture Networks public abstract and Gamma-Team web pages.
Multi-Scale Reservoir Modeling - Steven L. Bryant
Collaborators: Todd Arbogast (Dept. of Mathematics and Center for Subsurface Modeling) and Jim Jennings (Bureau of Economic Geology)
Vuggy rocks like the one pictured on the left present a challenge to traditional
models of flow and transport in porous media. In this formation, the vug lengths are
10s of cm, too big for standard core analysis to capture. This research project is
seeking to model large scale behavior (effective permeability, contaminant transport)
from fine-scale models that account for the geometry of the vugs and the matrix
surrounding them. Sub-mm resolution CT scans of a large chunk of rock (25 cm diameter
by 35 cm height) provide the detailed geometry. The vugs are so large that we are
forced to consider multiphysics models (Stokes flow in the vugs, Darcy flow in the
matrix). An example of the extreme heterogeneity of simulated tracer transport
appears in the right panel. The green contour corresponds to a tracer concentration
of 80% of the injected value after 0.95 PV of fluid have been injected. The tracer
samples only a small part of the sample, and the "swiss cheese" appearance of the
contour indicates that much of the pore space will not see the injected fluid until
long after the vug-dominated volumes have been swept. Efforts to validate the mathematical multi-scale and multi-physics models under development present their own challenges. Traditional core floods are not possible because the vugs are so large, so we are conducting "chunk-floods" on samples 25 cm in diameter. Such samples fit conveniently in a five-gallon bucket. The sample came from a creek bed near the town of Pipe Creek, Texas. In the experiments pictured here we discovered that modern sediment has infiltrated the vugs, significantly reducing the permeability and complicating our comparison with model predictions. More experiments are currently underway in CPGE laboratories.
More information is avaiable from the pore-scale modeling web page.
Characterization of Fractured Reservoir - Jon E. Olson
Predicting the performance of fractured reservoirs requires characterization of
fracture length, spacing, height and aperture distributions. Because none of these
parameters are typically well constrained by available subsurface data because of
small sample size (wellbores) or indirectness of the measurement (seismic),
theoretical models are required to fully populate fracture networks for flow
simulation. Our approach is to combine insights from the limited geologic
observations and core mechanical properties with a fracture mechanics-based model
to predict fracture network attributes. We import these fracture networks into a
reservoir simulator to discover the key attributes that control production behavior
to help further guide characterization efforts. The development of the
geomechanical model is linked to fracture mechanics measurements on core, fracture
network characterization at outcrop analogs, and the influence of diagenetic history
on rock mechanical properties and fracture aperture. Flow simulation work has
discovered fundamental relationships between permeability and fracture attributes
such as length distributions and connectivity, providing more realistic estimates
than the idealized parallel-plate model. Additional information is available on Dr. Olson's web page.
Analysis of Production and Injection Rate Data for Diagnosing Interwell Connectivity - Larry W. Lake
The exploration and production industry has had a long history of purely
statistical models, though this practice has almost entirely given way to the
algorithmic models now referred to as numerical simulators. On the other end of
the spectrum t he ultimate simplification will be a simple model that does
empirically quantifies the flow physics in the form of simple to evaluate
statistical correlation(s). We propose a radically different approach utilizing
well rate fluctuations to predict interwell connectivity. They expressed the
total fluid production at a producer as a linear combination of the injection
rates at different injectors located in the reservoir. The key conclusions are
that injector-producer weights are (a) independent of average well rates, (b)
appear to reflect geologic features, (c) require some filtering in the presence
of dissipation (indeed, with excessive dissipation the method fails entirely),
and (d) can be calculated with several statistical procedures. The best thing
about the method is that it can be used on virtually any injection process
because injection-production rate data exists on every type of mature project. Currently, research efforts are focused on demonstrating the technique on a truly definitive data set. The Magnus data (based on daily rate data) was obtained some months ago and application of the technique on a dataset so large would require substantial modifications to the approach for calculating inter-well correlation.
Core-Scale Petrophysics - Mukul Sharma