Skip to main content

Hydro-fracture in the laboratory: matching diagnostic seismic signals to fracture networks via new rock physics experiments

Final Report Summary - FRACSEIS (Hydro-fracture in the laboratory: matching diagnostic seismic signals to fracture networks via new rock physics experiments)

Hydrofracturing is a key process in many areas of pure and applied geosciences, such as the intentional hydraulic fracturing of impermeable rock formations in the hydrocarbon and geothermal energy industries, as well as natural processes in volcanology and tectonics. However, previous work investigating such phenomena were either limited to high pressure and high flow ‘pumping tests’ or various ‘trial-and-error’ approaches (particularly in the USA) which monitored only the fluid injection rate and resulting microseismicity. In both cases, little control or attenuation was paid to the stresses on the rock mass, the type of fluids being used, and with respect to the fracture onset.

These are important questions, as active fluids (those at high temperature and/or containing elevated acidic components) are known to promote the formation of fractures through a process known as stress corrosion. In addition, the majority of rocks have an inherent anisotropy, manifested as changing physical properties as a function of measurement direction, including the tensile fracture propagation so important to hydraulic fracture. To date these questions have yet to be thoroughly addressed in a well controlled laboratory environment and especially with reference to the important and topical issue of ‘fracking’: the deliberate fracture of anisotropic shales in order to develop unconventional hydrocarbon resources. This project has addressed some of these shortfalls by conducting a comprehensive new suite of experiments on anisotropic rocks in a well-controlled laboratory environment utilizing different pore fluids and stress regimes.

The project sought to address three key aspects of the scientific problem:
1) Firstly, to explore the dependence and fracture mechanics behaviour of the fluid driven mechanical fracture process without using a conventional rubber lining in the borehole, so as to assess the competition between rock permeability and the fluid overpressure needed to fracture it.
2) To test the importance of inherent and induced anisotropy on the fracture pattern. By using a small seismic sensor network, the resulting data is then used both for detecting the fracture network, and as a monitoring/safety feature capable of being scaled up to field environments.
3) Finally, to conduct preliminary experiments on the effect of the fluid “activity” level, either though temperature or acidity (pH), in order to assess resulting mechanical effects on permeability via the fracture pattern.

Initially, the measurement of the base permeability of two test rock types was conducted: 1) Crab Orchard Sandstone (COS) and 2) Nash Point Shale (NPS) using simulated reservoir conditions. This has been complimented by an extensive dataset of indirect tensile strength data to provide baseline comparison fracture data. These measurements show that permeability is significantly (factor of ten) higher along the bedding planes than normal to them, but the tensile strength is only a factor of 2-3 lower parallel to this same direction. However, this method is only usable at room pressure conditions.

To address this limitation and fracture the cylindrical samples in the tensile regime with confining pressure (hence, to simulate burial depth), a novel insert made of steel and sealed with rubber O-ring seals was designed specifically for the project, known as a TRX fracker. This device slots into the sample assembly inside a standard high pressure triaxial testing apparatus and allows a pore pressure to drive tensile fracture whilst also being in direct contact with the rock. This allows us to test the interplay of fluid over-pressure as a function of the rock base permeability, in addition to measuring the fracture strength. To understand the speed of the process, laboratory seismic data, known as Acoustic Emission (AE), is recorded, and compared to the pore fluid and radial strain data. The final fracture patterns are also imaged post-test using X-Ray CT methods, as a calibration metric to the AE data.

Our results compare the baseline datasets for permeability of the un-fractured COS and NPS rock types to the fracture onset as a function of breakdown pressure and with respect to the known confining pressure. This is a critical property that is related to the frackability of the formation – a geo-engineering parameter used to determine how well a given formation maybe exploited for unconventional resource extraction. Our results suggest an increase in breakdown pressure with increasing burial depth, but with a static fracture toughness value. Taken together, these new data have elucidated the conditions needed to generate fracture sets of a known size (and permeability) and as a function of seismicity, which is routinely measured in the field as a safety measure. By measuring these parameters under well imposed conditions of pressure and temperature we additionally seek to better use the seismicity (which cannot be avoided) as a tool to reduce the inherent hazards of the process as well as to improve exploitation of the resource by guiding the fracture sets and optimizing their generation.
To summarise, the FRACSEIS project has successfully developed a new standard method for deriving the fracture intensity from breakdown pressure, the common field engineering parameter in in unconventional gas/oil exploration. When combined with Acoustic Emission, the laboratory analogue of field seismic events, we have discovered that the onset of earthquakes is first detected some seconds before the development of the main through-going fracture. This has important implications for developing new safety feature in hydraulic fracturing installations, assisting policy makers in assessing risk and hazards. Continued research aims to link the fracture sets with new estimates for the generation of permeably networks using X-Ray CT imaging, and with respect to the significant anisotropy exhibited in this rock type. Future efforts to link fundamental micromechanical parameters to the breakdown pressure are planned.