Tuesday, January 8, 2019

The Promise of Elastic Anisotropy chapter 2

In an anisotropic rock, it is debatable whether the fast or slow S-wave velocity should be used - a slow velocity would give a higher closure stress, therefore a higher volume of pumped fluids. The DSI tool indicated about 8% anisotropy in the shale. Amoco engineers designed a fracture job around the fast shear-wave velocity, predicting lower closure stress, and reducing pumped fluid costs from $100,000 to $35,000 per well. Pump-in closure stress test confirmed the lower stress value indicated by the faster S-wave velocity from the DSI tool. Amoco anticipates saving $10,000 to $65,000 per well on the remaining 300 infill wells to be drilled in the field.

 At the slightly larger scale of a few feet to meters, crosswell seismic surveys also sense elastic anisotropy. But while most oilfield experiments employ vertically traveling waves to study elastic properties, crosswell seismic surveys harness horizontally traveling waves. In such a survey at the British Petroleum test site in Devine, Texas, USA, a seismic source was fired in one well to 56 receiver positions in a wwell 100 m [ 330 ft] away.  Then the experiment was repeated at 55 other source positions. Data processing called tomography divided the area between the wells into 56x56 cells and solved for the P-wave velocity in each square, to create a tomogram. Typical tomography, solving for isotropic velocities, reconstructed an image with layer boundaries that correspond to boundaries seen in gamma ray logs. However, allowing the velocities to be anisotropic enhances the results with a clearer tomographic image between wells.

At the Tree and Forest Scales

Most of the experiments designed to capture in-situ elastic properties have been vertical seismic profiles (VSPs) , at the 10-m [33 ft] wavelength scale. Specially planned VSPs reveal elastic anisotropy of both types, TIV and TIH, but mostly fracture-related TIH anisotropy via shear-wave splitting. These studies show a good correlation between fracture azimuth inferred from VSPs and from other measurements, such as borehole imager tools, regional stress data, surface mapping and experiment on cores. Conducting such studies in the marine setting offers a special challenge, because shear waves can not be generated in nor propagate through water. VSPs can, however, record waves that have been converted from P to S by reflection or refraction. Such vertically propagating shear waves then behave predictably by splitting into fast and slow shear waves when they propagate through fractured rock to borehole receivers. 

As desirable as fractures may be for enhancing fluid flow, they are undesirable in caprock shales, where vertical fractures could diminish their integrity as reservoir seals. Geophysicists are looking into ways to identify fractured and unfractured shale caprock, hoping not to see fracture-related anisotropy in them.

More sophisticated walkaway VSPs, called walkaway for short, can measure elastic properties of layer-anisotropic TIV rock in a way that no others can. Most VSPs rely on near-vertical wave propagation. But without nonvertical travel paths, the elastic properties of TIV materials, such as shales, cannot be measured in situ. The walkaway , with its large source-receiver offset and horizontal travel paths, is able to deliver vital information about shale properties.

 A walkaway survey in the South China Sea sampled a compacting shale sequence through more than 180 degree of propagation angles, usually impossible in all but laboratory experiments. The data revealed fine-scale layering induced anisotropy with horizontal P-wave velocities 12% greater than vertical.

The elastic properties of this highly aneliptic anisotropic shale were used to understand the effects of anisotropy on seismic reflection amplitude variation with offset (AVO) analysis. Surface seismic surveys and VSPs typically involve reflections of waves that propagate within 30 degree of vertical. Even in TIV-anisotropic shales, these near-vertical waves would not sense much anisotropy. But in surveys designed to highlight AVO effects, waves often travel at larger reflection angles. Reflection amplitude depends on the angle of reflection , or offset between source and receiver, and the contrast between P- and S-wave velocities on either side of the reflector. In isotropic rocks, some reflectors - especially those where hydrocarbons are involved - have amplitudes that vary with angle of reflection. 

 In anisotropic rocks, there is the additional complication that the P- and S-wave velocities themselves may vary with angle of propagation, again causing AVO. If a propitious AVO signature is encountered , it is vital to know how much is due to hydrocarbon and how much to anisotropy. This dilemma can be resolved by modeling, which simulates the seismic response to a given rock or fluid contrast. Modeling requires knowledge of elastic properties, and correct modeling should include anisotropy. But anisotropy is a scale-dependent effect, and it is best measured at a scale similiar to the VSP or surface seismic experiment being modeled, such as with a VSP. Most  examples of AVO modeling use sonic-scale elastic parameters - sonic log data. But it is possible to envision an anisotropy , especially if it is fracture-related, at a scale larger than the sonic wavelength but smaller than the VSP wavelength. In this case, the anisotropy may be felt by seismic waves but not by sonic waves. 

Another walkaway, by British Petroleum in the North Sea, measured anisotropic properties in a shale overlying a reservoir with an anomalous AVO signature. The elastic properties were used to model the AVO response at the interface between the shale caprock and the oil sand reservoir. The AVO signature seen in the walkaway data fits the anisotropic model. If the caprock had been assumed to be isotropic, a different AVO response to the oil sand would have been seen, and the sand might not have been identified as oil-bearing. The effect of anisotropy on the interpretation of the AVO anomaly had an important bearing on conclusions drawn from a concurrrent study based on 3D surface seismic data in the area. 

Velocity anisotropy is also beginning to find a home in another corner of the surface seismic world, that of processing surveys to obtain images. This process, called migration, requires knowledge of the velocities of the seismic waves to assign a correct spatial position to reflections recorded in time. In the absence of measurements of anisotropic elastic properties, conventional migration schmes include a 5% fudge factor and assume elliptical anisotropy to convert stacking velocities - results from a prior processing step - to migration velocities. A knowledge of velocity anisotropy beyond the 5% fudge factor, essentially knowing the anellipticity , will become more important in turning ray seismics, as seismic waves spend more time in horizontal travel paths.

Harvesting the Forest

These two types of elastic anisotropy, TIV and TIH, impact the oilfield geoscientist as well as the anisotropist. Measurements of layer-induced anisotropic elastic properties are used to refine processing and produce clearer images or to create better models that lead to more accurate interpretation. In the long run, measuring TIV elastic anisotropy improves reservoir description, which in turn promotes efficient hydrocarbon recovery. 

Measuring fracture- and stress-induced TIH anisotropy may have a more direct and far-reaching impact. Just as elastic waves are bound to travel in the direction of maximum stress or open fractures, so are reservoir fluids. The same forces that induce elastic anisotropy give rise to permeability anisotropy. But the tie between these two is not made routinely, nor is it full understood. 

Establishing the elastic-permeability tie for anisotropy requires geophyscisit, reservoir engineers, geologist and petrophysicst to experiment with such a tie, documenting successes and failures. Today, the most basic level of anisotropic description involves only the geophysicist. The description comprises the azimuth of fracture or stress anisotropy, the degree of anisotropy in relative velocity difference, and the velocities of the two shear waves.

At a more sophisticated level the geologist and petrophysicist add the following information to make further links in the rock-fluid tie: lithology from core or logs; age of the reservoir; history of hydrocarbon maturation; azimuth and aperture of fractures seen in image logs; stress direction in the vicinity of the borehole from caliper logs or hydraulic fractures; and the effect of fluid saturation on resistivity anisotropy.

No comments:

Post a Comment