Feature Articles—October 2009 Issue
Modern Geophysical Technology Aids In Finding Bypassed Oiland Gas Reserves
Operators in the Gulf of Mexico Find Seismic Tool Useful in Evaluating Opportunities for Drilling
By Roger Young
Chief Technology Officer and Co-Founder
eSeis Inc.
Houston, Texas
The Gulf of Mexico continental shelf is a mature oil and gas province that has been producing hydrocarbons since the early 1950s. Technology advances over the last five decades have led to the identification of new exploration prospects on deep structures, subtle traps and the direct detection of hydrocarbons via seismic amplitude anomalies and AVO (amplitude variation with offset) studies. However, there are good opportunities remaining in old fields that require another “step change” in technology application.
Multidisciplined asset teams benefit from advanced seismic technologies by using an integrated rock-property-based solution to maximize results in exploration, well planning, drilling, production and reservoir management. The key to success is to understand the rocks using tools to interpret and define lithology, porosity, fluids, pore pressure and reservoir quality so as to provide seismic data in a geologically meaningful and multidisciplinarily useable format.
Conventionally, seismic inversion produces compressional velocity, shear velocity and density. Log analysis starts with velocity and density, as well as other measurements, and finishes with lithology, porosity and fluids. Seismic petrophysics bridges the gap between these two workflows by starting with seismic data and resulting in 3D seismic volumes of lithology, porosity and fluids.
Advantages of Seismic Petrophysics
For decades, the rudimentary process by which geologists determine optimal oil and gas drilling locations for exploration companies has remained static. Exploration companies have come to expect that many of the wells they drill will come up empty, despite preliminary research indications to the contrary.
However, the evolution of seismic petrophysics, combined with new proprietary technologies, provides geologists and engineers with dynamic, three-dimensional models of the subsurface environment that more accurately reveal the reservoir properties and the likely locations of gas or oil.
The integration of well logs with seismic data is challenging because they are presented in different domains. Well logs are displayed in units of microseconds per foot, grams per cubic centimeter, gapi, ohm-meters, etc., while seismic is displayed in terms of reflectivity. When inversions are carried out to put seismic into a more log-similar domain, acoustic impedance is often chosen. Well logs are inverted to lithology, porosity and fluids. Why then take the seismic data to the starting well log domain yet not to the final point? All of the measurements have one thing in common: the rocks. So ideally all of the measurements should be placed into the rock domain—that is, lithology, porosity and fluids. The process can then be thought of in terms of rock-based integration.
Traditional Method
Conventional seismic interpretation is conducted by taking well logs and applying their results to make synthetics. Then horizons are followed on the seismic from well to well, but the results hardly provide a complete portrait of the underground rock structure and the actual location of natural gas.
A person looking at a simple stack may ask: “Where is the gas?” “Where is the sand?” “Where is the shale?” Gas appears on these images as bright spots and dim spots, which occur because the stacking process emphasizes Class 3 anomalies, which have flat-to-increasing amplitude with offset distance, and suppresses Class 1 anomalies, which dim with offset.
Gas sometimes makes the stack bright, sometimes dim and sometimes does not do anything at all. That is because seismic stacks respond to lithology, porosities and fluids. As an analog, well logs also respond to these same factors.
Reviewed separately, a neutron log or a density log would not accurately identify sand, shale or gas zones. By overlapping the density log and the neutron logs in compatible scales on the same track, pieces of information come together to fill in the puzzle. When the two log curves cross, gas is indicated. When the two log curves are separate, shale is likely present. When the two log curves are on top of each other, sand can be found. The porosity is the average of the two—what is known as cross-plot porosity.
A cross-plot can also be developed by putting the neutron points on the x-axis and the density points on the y-axis, resulting in a scatter plot of data. The meaning of the scatter plot is revealed by referencing a chart book. The result is shale, sand, gas and porosity.
By taking a simple petrophysical workflow, it is possible to cross-plot a neutron log and a density log to calculate lithology, porosities and fluids.
New Method
An advanced seismic analysis technique called LithSeis® combines well log analysis concepts with prestack seismic data to determine volume-based lithology, porosity and fluids. By extracting lithologic measurements from the seismic data before stacking or averaging, the result is a more accurate rock-based description of the subsurface.
In petrophysical analysis, the neutron and density combination allows for the separation of lithology and fluids from porosity. This works because neutron and density logs are each independent from the other, and both respond to lithology, porosities and fluids.
A similar technique can be applied to the seismic data; however, two independent seismic sections will be needed instead of one. That problem can be solved by using AVO technology.
The acquisition of seismic data results in the sampling of each surface point at many different angles. The AVO can thus be analyzed, resulting in zero offset (P) and AVO gradient (G) sections, two independent seismic measurements. Conventionally, all the gathers are added up or averaged to create the full offset stack, resulting in just one measurement. The two sections provide the equivalent of two well logs in the logging world. A cross-plot is created by putting P on the x-axis and G on the y-axis and plotting all of the points. The resulting cross-plot scatter is equivalent to that created by the neutron and density logs. To understand the scatter plot in terms of lithology, porosities and fluids, a lot of modeling was required.
Conventional seismic modeling was not sufficient when trying to understand the relationship between rocks and synthetics/seismic. Log analysis is about the relationship between logs and rocks. Seismic models relate logs to seismic. Therefore, the math already exists to relate rocks to synthetics/seismic. Studying how rocks influence seismic reveals what seismic is telling us about rocks.
In modeling, lithology derived from the well logs is used to create the synthetic seismic gathers. As the lithology changes, the modeled seismic gathers also change. This allows for the creation of thousands of modeled responses of different rocks.
From these responses, gathers are created, then P and G sections are created, leading to the final result that all rock responses are plotted into P/G cross-plot space.
The location that a given shale/sand interface point will plot depends on the properties of the sand and shale. After studying many models, it becomes evident that as the sand becomes cleaner, the point plots further from the origin. Add gas to the sand, and the distance increases in generally a southwesterly direction. Change the porosity, and the direction changes. In other words, porosity acts in an orthogonal way to lithology and fluids.
After studying thousands of different models and hundreds of different data sets, it is easy to figure out what is important in cross-plot space. First is elliptical distance from the origin; the cross-plot is elliptical by nature, and its elliptical distance from the origin is a function of lithology, fluids and thickness. Next is direction; direction is a function of porosity and whether the sand is blocky or laminated. Together, these help to infer the depositional facies. Direction is called AVO type and is divided into 10 different types from 5 to -5.
Usefulness in the Gulf of Mexico
Fairways Offshore Exploration Inc. (Houston, Texas) was looking for a tool to help it identify bypassed pay and better understand stratigraphic variability in an old field in the Gulf of Mexico. The field was very structurally simple, with a lot of stacked pay and much stratigraphic variability.
The main zone of interest was oil sands located in depths of less than 9,000 feet.
This AVO technology was run on the 3D dataset to obtain lithology, porosities, fluids and AVO types. Log analysis was run on approximately 70 wells in the project.
Good petrophysical analysis goes beyond doing net pay counts. Well logs are subject to error, just like any data set; therefore, these errors must be corrected. Things such as borehole washout, fluid invasion and swelling shales account for a large portion of the errors which must be corrected.
In a field study such as this, the logging data issues become even more difficult. Different suites of logs are run in each of the wells, making some sort of normalization process necessary. Most of these wells were deviated and had only resistivity and spontaneous potential curves.
There were a few straight holes that had sonic logs available, and these few wells were the first to be run through log analysis and tied into the seismic and lithology. Once a suitable tie was made for these wells, one was chosen as an approximation for the time/depth relationship on the wells that did not have sonics.
Since the log analysis performed came up with a rock model, sonics were created from the rock model in the wells that did not have them. The time/depth function mentioned earlier was used as a starting point for the synthetic tie. Small adjustments were made to the wells in order to get the best ties.
After all 70 wells were tied into the lithology/fluid volume, the volume was checked against the wells for accuracy. All of the wells matched the lithology/fluid volume, so this and the other output volumes were delivered to Fairways for loading into their interpretation system to begin their search for bypassed pay.
A total of approximately 500,000 barrels of bypassed reserves were located by using this technology. Unfortunately, the location of the bypassed pay was scattered and would take too many wells and cost too much to develop.
In future projects, it will be helpful to see the results of this type of work before deciding to buy fields or deciding a field’s worth.
Conclusions
The LithSeis technology will allow companies to more easily and accurately find and develop reserves with tools that are useful throughout the life cycle of a project.
The marriage of petrophysics with geophysics results in a science referred to as seismic petrophysics. The goal is to understand the petrophysical information that seismic data brings to the table. This starts with a good understanding of the rocks. Having seismic data in terms of lithology, porosities and fluids means it can be used by all members of the asset team and provide exploration companies with a more strategic and precise model for finding hidden oil and gas reservoirs.
Roger Young is chief technology officer and co-founder of eSeis Inc. He has 27 years of industry experience, including 14 years of LithSeis development. Young holds a Master of Science in petroleum engineering from the University of Houston and a Bachelor of Science in physics from Clarkson College of Technology.
Sea Technology is read worldwide in more than 110 countries by management, engineers, scientists and technical personnel working in industry, government and educational research institutions. Readers are involved with oceanographic research, fisheries management, offshore oil and gas exploration and production, undersea defense including antisubmarine warfare, ocean mining and commercial diving.
