NETL researchers are measuring fluid flow at pressures and temperatures that are representative of conditions deep underground, and determining the effect that changing the amount of CO₂ or other fluids has on ultrasonic seismic signal velocities in the rock.

NETL researchers and collaborators at the University of Pittsburgh have developed a correlation between acoustic wave velocities measured in the laboratory and relative carbon dioxide (CO₂) saturation, which allows researchers to calibrate and refine the interpretation of seismic reflection surveys. The researchers are able to simulate the flow of CO₂ in a core sample of reservoir rock at the temperatures and pressures that are encountered deep underground, and then measure how the amount of CO₂ present affects how seismic energy moves through the rock. Recent laboratory tests showed that there is a marked decrease in signal velocity when CO₂ is introduced into the pore space, and that this decrease is large enough to be detectable in a seismic survey. Although it was previously possible to observe apparent gas plumes underground, interpretation of the data was very complicated because of all of the unknowns, such as how different types of ultrasonic seismic signals are affected by the concentrations of various fluids at the temperatures and pressures encountered at various depths. So, it was assumed that the image changes were due to migration of the gas, but the margin of the migration was hard to define and the changes observed could possibly have reflected other changes caused by the injection, such as movement of other fluids, like brine. However, by conducting laboratory tests on actual samples of the rock that the seismic signals are passing through, the effects of CO₂ can be more easily and more precisely measured, which in turn makes it possible to much more easily track the movement of the CO₂ deep underground using surface instruments. In addition, subtle changes in the rock matrix, caused by the injection process and the CO₂, can be much better understood after modeling these experiments.

CLICK ON GRAPHIC TO ENLARGE   This image, created by computerized tomography (CT), shows connected pores in white and unconnected pores in red, within a sample of a rock type that may eventually be used to sequester carbon dioxide; the rectangular cube shown is 1.54 x 1.40 x 1.12 mm.  

Associated laboratory tests indicate that it may also be possible to use changes in the seismic images, which symbolize reflection strength, variations in this strength between surveys, and other changes observed over time, to identify changes in fluid phases deep underground. Laboratory experiments conducted at NETL have examined how rock porosity can be changed by the dissolution or precipitation of minerals by the CO₂ that dissolves in the water that is present, but it is important to examine the large-scale effects of this over time to predict the long term fate of the injected CO₂, since changes in how rock pores are connected can dramatically affect the rate at which the CO₂ flows underground. The image below shows how some pores are connected, which allows fluids, such as CO₂, to flow through the rock.

Associated laboratory tests indicate that it may also be possible to use changes in the seismic images, which symbolize reflection strength, variations in this strength between surveys, and other changes observed over time, to identify changes in fluid phases deep underground. Laboratory experiments conducted at NETL have examined how rock porosity can be changed by the dissolution or precipitation of minerals by the CO₂ that dissolves in the water that is present, but it is important to examine the large-scale effects of this over time to predict the long term fate of the injected CO₂, since changes in how rock pores are connected can dramatically affect the rate at which the CO₂ flows underground. The image below shows how some pores are connected, which allows fluids, such as CO₂, to flow through the rock.

---

MVA Technology Proves Useful in Pilot-scale Field Test

CLICK ON GRAPHIC TO ENLARGE     Figure 1. A seismic image of a 9 square mile area of the lower Fruitland coal seam at a sequestration test site in New Mexico; the thickest areas of the coal seam are shown in red. The dots represent the injection well and the monitoring wells.

Accurate monitoring, verification, and accounting (MVA) requires knowing where the CO₂ will go once it is injected deep underground. After all, to determine if CO₂ is leaking, one must know where to look. Geological sequestration field tests are being conducted so that we can learn what works and what does not. However, we are still learning; this LabNote provides an example of how complex such predictions can be and the importance of conducting such research now, well before CO₂ sequestration becomes standard practice.
 
Models of how the CO₂ would flow at the San Juan Basin pilot test site (being conducted by the Southwest Regional Partnership (SWP) on Carbon Sequestration) were created using information from wells drilled down through the strata, geophysical logs, and rock samples collected during drilling of the wells. Movement of injected CO₂ into the Fruitland coal seam was expected to follow the prominent face cleat trend. Face cleats are fairly continuous fracture systems within coal, and it was expected that the injected CO₂ would migrate through these fracture systems. Because the face cleats there have a northeast-southwest orientation, it was anticipated that the injected CO₂ would first be detected at a well about 1500 feet to the southwest; however, the PFC tracer, and later, the CO₂, was first observed in the east offset well, opposite to what was expected.

The SWP obtained seismic data to help researchers better understand the subsurface geological environments deep underground and why the CO₂ moved the way it did. Seismic data is based on analysis of how energy associated with vibrations, in this case, locally imposed shock waves, passes through and reflects off of subsurface features. Seismic images can provide us with a more comprehensive picture of subsurface CO₂ storage environments than is possible from sparsely distributed well data, but can also be tricky to interpret because so many factors affect the path and speed of the reflected signals. The 3D seismic interpretations of the subsurface in the vicinity of the San Juan Basin pilot site revealed additional complexity in the coal depositional environments and their deformation. The seismic imagery indicated that the coal apparently thickened northeast of the injection well, and that this thickened zone has a northwest-southeast trend. The injection well happens to be located near the southwestern end of this thickened coal zone (see Figure 1).

CLICK ON GRAPHIC TO ENLARGE   Figure 2. 3D seismic imagery reveals discontin-uities that cut through laterally coherent reflection events. This image represents a line about 3.4 miles in length and oriented in the NE-SW dip direction through the injection well (see Figure 1). Neighboring wells are located about 1500 feet from the injection well. Arrows point to the tips of linear features cross-cutting reflection events, which represent possible faults and fracture zones. MVA efforts should be concentrated near and above such zones.  

The researchers working with these data speculate that the injected CO₂ that entered the thicker coal section migrated more rapidly, which would explain why the injected PFC tracers were first observed in the well east of the injection well, near the edge of the thickened coal zone. Researchers working with the 3D seismic data also observed that the area is more complexly deformed than expected (Figure 2). These subsurface structures may provide additional pathways along which injected CO₂ might escape from the reservoir interval (the Fruitland in Figure 2) and migrate to higher strata, and possibly to the surface.

Thus, MVA activities were critical to this pilot scale test. They revealed unexpected behavior in the subsurface CO₂ flood. This kind of information is critical in determining whether a given site would be appropriate for larger-scale sequestration and, if so, where monitoring wells should be concentrated.  Second generation flow simulations will be developed in the coming year to incorporate the results of ongoing seismic research.