The influence of hydrothermal alteration and lithology on rock properties from different geothermal fields with relation to drilling.
Degree GrantorUniversity of Canterbury
Degree NameDoctor of Philosophy
Deep drilling is required to reach the geothermal fluids extracted for generation of electricity; therefore, the different rock properties and the hydrothermal alteration of the lithologies being drilled become an important factor to a conventional geothermal industry. If the correct equipment required to complete drilling is not selected, the rate of penetration (ROP) can be suboptimal, potentially increasing the cost of the project. Mechanical characterisation of hydrothermally altered rocks from geothermal reservoirs will lead to an improved understanding of rock mechanics in a geothermal environment. Core samples obtained from the Ngatamariki, Rotokawa and Kawerau Geothermal Fields covered a wide range of lithologies (ignimbrite, rhyolite lava, sandstone, mudstone, andesite lava/breccia and tonalite) encountered during drilling. A suite of non-destructive and destructive laboratory tests along with petrographical analysis were conducted on the samples. Some key findings are that samples that originated from the shallow and low temperature section of the Kawerau geothermal field had higher porosity (15 – 56%), lower density (1222 – 2114 kg/m3) and slower ultrasonic wave velocities (1925 – 3512 m/s (vp) and 818 – 1980 m/s (vs)), than the samples from a deeper and higher temperature section of the field (1.5 – 20%, 2072 – 2837 kg/m3, 2639 – 4593 m/s (vp) and 1476 – 2752 m/s (vs), respectively). The shallow lithologies had uniaxial compressive strengths (UCS) of 2 – 75 MPa, and the deep lithologies had strengths of 23 – 211 MPa. Typically samples of the same lithologies that originate from multiple wells across a field have variable rock properties because of the different alteration zones from which each sample originates. To obtain a way to relate this rock property data back to the geomechanical model, we developed a method - Alteration Strength Index (ASI) - to address the effect of hydrothermal alteration on mechanical rock properties. The index constitutes three components; the mineralogy parameter, derived from petrological analysis, alteration index (degree of alteration) and an assessment of mineral hardness; the fracture parameter, assigned based on an assessment of structural damage; and the porosity parameter, which accounts for the effect of voids. This method can be used to estimate a range of rock strengths comparable to UCS, and the ASI calibrated against measured UCS for the samples produced a strong correlation (R2 of 0.86). From this correlation an equation was derived to convert ASI to UCS. Because the ASI–UCS relationship is based on an empirical fit, the UCS value that is obtained from conversion of the ASI includes an error of 7 MPa for the 50th percentile and 25 MPa for the 90th percentile with a mean error of 11 MPa. A sensitivity analysis showed that the mineralogy parameter is the dominant characteristic in this equation, and the ASI equation using only mineralogy can be used to provide an estimated UCS range, although the uncertainty becomes greater. This provides the ability to estimate strength even when either fracture or porosity information are not available, for example in the case of logging drill cuttings. To determine the usefulness of the ASI method with drill cuttings and drilling data we compared it to two methods; mechanical specific energy (MSE) and R/N-W/D chart, both developed for the oil and gas industry, in a geothermal context. We demonstrated how they can be used to estimate a range of rock strengths for hydrothermally altered lithologies for the 800 metre long 17 inch (432 mm) diameter section of well NM8 in the Ngatamariki Geothermal Field, New Zealand. We found that MSE and the R/N-W/D charts correctly ranked relative strength to ROP for three of six lithologies, while ASI correctly ranked all six lithologies. We also show that the strength values predicted by ASI correlate to ROP better than those based on MSE or R/N-W/D. We argue that ASI is more comprehensive than these methods because it provides a range of rock strength indices for a given hydrothermally altered lithology, is based on the geology, and does not require drilling parameters (ROP, WOB, RPM, and Torque) to estimate rock strength. This is particularly important in geothermal systems where lithologies can exhibit high variability in their physical characteristics and geothermal fields tend to have widely spaced wells. Using ASI we show how hydrothermal alteration affects drilling, and when used in conjunction with a predictive geologic model, how it will aid with optimisation of drilling practices through drill bit selection. Rock failure modes are difficult to predict, and are important to rock engineering environments, which include drilling. By using rock property and mineralogy information, four modes of failure were identified – axial splitting, single plane shearing, y shaped failure and multiple fracturing - in this research. The results of this study indicate that these easily measured rock properties can be inferred to have some control over the failure mode of a sample under uniaxial loading; however it would be useful to examine these samples further at the microstructural level to determine the role of microfracturing in the occurrence of failure modes. Further research in this field has the potential to aid in drilling optimisation through the utilisation of drill bits designed to fracture rocks in the ways that they are predisposed to fail.