ALI DANESHY, Daneshy Consultants International
During the last several years, there has been media speculation about the connection between earthquakes and oil and gas operations in general, and hydraulic fracturing in particular. For people who live near oil and gas producing areas, this can cause understandable anxiety and discomfort. Unfortunately, much of the discussion has been based on insinuation and not on credible scientific observation. This article reviews the subject from a purely technical point-of-view and offers a summary of the types of rock failure usually occurring as a result of drilling, completion and production activities, and existing theories connecting oil and gas operations with formation failure in general, and earthquakes in particular.
Technology background. Every formation is under the influence of underground stresses. These stresses result from weight of the overlying rock (overburden), fluid pressure, and geologic and tectonic activities that the formation has experienced during the millions of years of its life. Stresses in the formation are divided into two basic types: normal and shear. Normal stresses tend to compress the material or pull it apart in tension. Shear stresses tend to tear the material, either by sliding or twisting action. These stresses have different magnitudes in different locations and directions, and generally increase with depth. An important mechanical property of the formation is its strength. This is usually measured in the laboratory. We usually measure two types of formation strength: compressive and tensile. Rocks are usually under much stronger compression than they are under tension.
Formation failure occurs whenever the stresses acting on any part of the rock exceed its strength. Such failure is limited only to areas where failure conditions have been met. Failure in hydraulic fractruring occurs mainly as a result of tensile stresses induced by fluid pressure inside the fracture. There is very little shear failure associated with fracturing, and even then on a very small and local scale and due to natural fractures or planes of weakness along the path of the fracture. Because of formation geological and material heterogeneity (planes of weakness), its strength varies in different locations and orientations. This results in local (rather than global) and directional rock failure. In engineering, failure is usually analyzed using the Mohr envelope, Fig. 1. The circle in this figure represents the state of stress, and its location is based on the magnitudes of the two normal stresses in that plane. The outside curve represents strength, and its location is determined experimentally. As long as the stresses acting in any given plane are such that the corresponding circle does not touch the Mohr envelope, the formation is intact. If the stresses increase, or rock becomes weaker, such that the circle touches the envelope, then formation failure will occur. The side effect of formation failure depends on dimensions of the failed volume. If these dimensions are small, then failure will be very local and often unnoticeable. When the planes of weakness are large, then the failure can be sensed and measured at the surface, sometimes even reaching the earthquake designation.
 |
| Fig. 1. Mohr envelope of failure |
|
In the vast majority of situations, the strength of the formation rock is higher than the stresses acting on it. The result is an equilibrium condition, and the formation remains intact. In tectonically active areas, these stresses change gradually over time. If and when they exceed the strength of the formation, then the stress circle grows and touches the Mohr envelope, causing failure. Thus, every rock failure requires either a changing stress regime, or weakening of the rock. When the failure occurs over a very large area under large stresses, the effect is an earthquake.
Formation failure is a common occurrence in oil and gas operations. The vast majority of these failures are extremely small in magnitude and limited to the very near-wellbore region. These failures are often so minor that they can not even be detected at the surface. In severe instances, they may cause wellbore collapse or casing shear failure. Even these types of failure are local and often not even detectable at the surface. Their main cause is a change in stress state created by drilling of the well, or local reduction in formation strength. The influence of drilling-induced failure is very limited and does not cause any harm to the surface installation.
Removal (production) of oil and gas from the formation can also cause a change in the state of stress in the overlying formation, which can result in surface subsidence. Usually this subsidence is extremely small and detectable only with very sensitive instrumentation at the surface. The depth of the formation and stiffness of the overlying rock substantially reduce its impact at the ground level. In rare situations, subsidence can cause obvious effects at the surface. A well-known example of extreme subsidence was observed in the Ekofisk reservoir in the Norwegian North Sea during the mid-1980s that caused partial sinking of the platform. The cause of the problem was traced back to formation mechanical properties. The very soft, deformable chalk formation was compacting unusually large amounts as oil and gas were extracted from it. The problem was solved by modifications of the platform structure.
Another manifestation of subsidence may be slippage along the bedding planes between the layers above the reservoir. In some instances, this can cause shear failure of the casing. Examples of this situation have been observed while producing from the shallow diatomaceous formation in California. Again, the area influenced by this subsidence has been very small. Aside from these rare, well-documented examples, the ground subsidence has not caused a disruption for the people or installations above oil and gas reservoirs.
In order for large-scale ground failure to occur, there needs to be a combination of two pre-existing conditions: large, naturally existing stresses that are close to the formation mechanical strength and naturally existing weaknesses that are susceptible to failure with small perturbations. The most common of these weaknesses are natural fault planes. In the vast majority of oil and gas operations, the perturbations in the stress state or formation strength are small and not capable of causing formation failure. Exploration and development activities usually take account of existing faults, and generally try to stay clear of them. In those situations where oil and gas perturbations are known to have caused earthquakes, the magnitude of these earthquakes has usually been small.
The first known systematic study of earthquakes being trigged by human activity was done in the Rangely field, in Rio Blanco County of northwestern Colorado.1 The motivation for this study was the speculation that small earthquakes near the U.S. Army’s disposal well at the Rocky Mountain Arsenal were triggered by wastewater injection. The fact that these earthquakes ceased when injection was stopped, strengthened the possibility of a connection. After extensive instrumentation and monitoring of the field seismic events, together with geological mapping, in-situ stress measurement and rock mechanical studies, it was determined that the cause of earthquakes was slippage along an existing strike-slip fault which was, in turn, triggered by an increase in pore pressure of the reservoir. The engineering explanation was that the increase in pore pressure caused two side effects: reduction in frictional resistance along the fault plane by lubrication and a decrease in the effective stress normal to the fault plane caused by increase in pore pressure. Through rigorous measurements and computation, the study determined the critical pressure above which earthquakes would occur.2 Later manipulation of the fluid pressure in the field (through fluid injection and withdrawal) verified the connection between fluid injection and earthquake occurrence.
A simple explanation of the mechanism is presented in Fig. 2. During injection, the increase in pressure (ΔP) pushes the circle to the left, bringing it closer to the Mohr failure envelope. At the same time, reduction in frictional resistance of the plane of weakness lowers the Mohr envelope. Failure occurs if and when the moving stress circle touches the lowered Mohr envelope.
 |
| Fig. 2. Failure mechanism associated with fluid injection |
|
Occurrence of an earthquake actually relieves the stresses and prevents their accumulation, thus reducing the intensity of the next earthquake. As such, the question remains as to which option is a safer alternative—accumulation of stress without failure to a point of major catastrophic failure, or gradual periodic relief of lower-magnitude stresses that may prevent the more catastrophic earthquake. In the report following the Rangely field study, the authors note that injection-induced earthquakes had usually been of magnitudes less than 4.5 and had not caused surface damage. They, therefore, propose that in areas with a known high risk of earthquake occurrence, one way of limiting the damage would be to artificially and periodically induce lower-magnitude earthquakes in selected locations along the fault plane, thus relieving the stresses and avoiding their build-up before they cause a more catastrophic failure.1 A careful examination of this mechanism leads us to conclude that oil and gas production should actually inhibit earthquakes, simply because it causes a reduction in pore pressure, thus pushing the circle away from the failure envelope. In fact, as stated earlier, in the Rangely field experiments, earthquakes stopped when pore pressure along the active fault was artificially reduced by fluid withdrawal.
Under very special circumstances, there are some studies that have also linked ground seismicity with production operations. Production of fluid causes formation shrinkage. If the reservoir is in close contact with a pre-existing fault, and if due to formation heterogeneity the fluid pressures are different on the two sides of the fault, then differential shrinkage of the rock may cause a stress build-up along the fault plane. Failure can occur if the stresses on the fault plane exceed the shear strength. Failure of this type has been reported to have occurred in South Texas.3
Whether in connection with fluid injection or withdrawal, the common denominator of all reported ground seismicity has been existence of two conditions:
-
Pre-disposition to failure, due to existing natural planes of weakness (faults) in the formation
-
Narrow gap between formation strength and existing stress state
Under these conditions, the production or injection of fluid can cause earlier formation failure at lower stress levels. This will actually reduce the severity of possible surface damage when compared with failure at later stages at higher stress levels.
Impact of hydraulic fracturing. Some recent news reports have insinuated a connection between hydraulic fracturing and occurrence of earthquakes. Although hydraulic fracturing is the result of formation failure, the nature of this failure is very different than those discussed earlier. The main failure mode in earthquakes is shear, while the dominant mechanism for hydraulic fracturing is tensile. This is an important distinction. The orientation of the fracture is usually vertical and perpendicular to the minimum in-situ principal stress. There is very little, if any, shear stress acting along this plane. Thus, chances of shear failure along the fracture plane are very slim. Furthermore, hydraulic fractures extend a very short distance away from the wellbore (relative to the size of faults or other naturally existing weaknesses in the formation that can trigger earthquakes). Therefore, at worst, their impact is very local. Presence of multiple fractures in horizontal wells creates a fractured zone and may reduce effective bulk strength of the formation. But even if these fractures are within an area already pre-disposed to earthquakes, the reduced strength of the rock causes the earthquake to occur at lower stress levels, thus reducing its intensity and damage. Due to higher pressures created during fluid injection, such earthquakes are more likely to occur while fracturing. The author is not aware of any reports of such occurrence.
After the fracturing operations, production of reservoir fluid reduces fluid pressure within the reservoir, and therefore, chances of any later earthquake.
Another important operational aspect of oil and gas production is the urge to avoid major faults while planning a fracturing operation. The reason is the uncertainty in reservoir properties on the two sides of the faults and the fear of operational problems while fracturing that may be caused by these faults.
Conclusion. If a formation is already predisposed to shear failure, then injection of fluid may trigger earlier occurrence of that failure. When the existing stresses within the formation are large enough to cause an earthquake, then injection of fluid may trigger it to occur earlier. This should be tempered with two other effects; earlier triggering reduces intensity of the earthquake and its damage, and the redistribution of the stresses after the earthquake reduces the intensity of any subsequent failures.
There have been no reported cases of earthquakes occurring during or as a result of hydraulic fracturing operations. The technical reason for this is that hydraulic fracturing is caused by tensile failure of the formation. Earthquakes are caused by shear failure along existing massive planes of weakness in the formation. The magnitude of shear stresses along the fracture plane will be extremely small (if any at all) and definitely insufficient to cause an earthquake. 
LITERATURE CITED
1. Raleigh, C. B., J. H. Healy and J.D. Bredehoeft, “An experiment in earthquake control at Rangely, Colorado,” Science, vol. 191, pp. 1230–1237.
2. Haimson. B. C., “Earthquake related stresses at Rangely, Colorado,” Proceedings of 14th Symposium on Rock Mechanics, (1972), p. 689.
3. Pennington, W. D., S. D. Davis, S. M. Carlson, J. Dupree and T. E. Ewing, “The evolution of seismic barriers and asperities caused by the depressuring of fault planes in the oil and gas of south Texas,” Bulletin of the Seismological Society of America, vol. 76, pp. 939–948.
|
