2021 Virtual Undergraduate Research Symposium
2021 Virtual Undergraduate Research Symposium
Apache Drill Data Collection and Analysis
Apache Drill Data Collection and Analysis
PROJECT NUMBER: 82 | AUTHOR: Edward Perea, Petroleum Engineering
MENTOR: Alfred Eustes III, Petroleum Engineering
GRADUATE STUDENT MENTOR: Kirtland McKenna, Petroleum Engineering
ABSTRACT
The purpose behind this project was to drill and measure the geo-mechanical properties of core at the Edgar Mine. This would be used to help improve the autonomy of drilling and aid when drilling for oil and gas, geothermal energy, and help with space resources exploration. Hence, there would be a minimization of instrumentation/logging needs and optimize the overall drilling performance in respect to the mechanical specific energy vs unconfined stress. Moreover, drilling was done using an Apache horizontal drilling rig, the drilled core was collected, and prepared. In terms of preparation, the core was cut with a saw in the core prep lab on campus to be approximately 6-8 mm. This core was then taken to the geomechanics lab where it would be inserted into the MTS press. The core had its dimensions measured, it was weighed, photographed, and then inserted into the MTS press where it was pressed until failure. It was important to take note of any failures due to natural fractures because this data would be discarded. The press would simultaneously measure displacement and force making it possible to determine stress/strain and calculate Young’s Modulus and the uniaxial compressive strength (UCS). The core was made up of various rock types, but the four predominant ones were biotite, quartz-plagioclase gneiss, hornblende, and biotite-microcline-pegmatite. Young’s modulus ranged from 2.612 GPa to 11.239 GPa and the UCS ranged from 11.7 MPa to 132.7 MPa. The densities ranged from 2.022 g/cc to 2.933 g/cc. Ideally, the next steps would be to use this drilling data and train artificial intelligence (AI) with it to create a neural network. Ultimately, the AI would be able to characterize geomechanics from drilling data and optimize the drilling process.
PRESENTATION
AUTHOR BIOGRAPHY
Edward Perea is a third year at the Colorado School of Mines where he is majoring in Petroleum Engineering with a minor in Data Analytics. The research he worked around was primarily composed of drilling core at the Edgar mine and measuring the geo-mechanical properties of the rock. Ideally, in the future, he’d like to create a neural network and train the data so that it could characterize different geomechanics. This would help optimize the autonomy of drilling for oil/gas, geothermal energy, and help with space resources exploration.
Some types of rock had very consistent stress/strain relationships, and some were quite variable. What would account for some of this variability, and how would this impact the drilling process?
Tyrone, the variability in the rock’s stress/strain relationship depend on a number of different factors. For instance, rock type is a huge factor as quartz-plagioclase gneiss was one of the harder rock types. This type of rock took longer to fail and was also a lot harder to drill. When at the rig, it took a longer time to drill a single foot of this rock type. Additionally, there were some rocks that were difficult to differentiate as they had a mixture of rock types and minerals. A possible source of error was incorrectly categorizing a core as we really just eyeballed it and drew our conclusions from some prior research and image searches. Moreover, hopefully in the mere future, we are able to drill more, attain more core samples, and become more familiar with the different categories of present rocks so we can see less variability within each rock types stress/strain relationship.