Sushil Acharya
About
Sushil Acharya is a data science researcher from Nepal. He completed his undergraduate degree in Physics from Tribhuvan University Nepal and proceeded to obtain a Master's degree in Applied Computer and Information Technology with a specialization in Data Science from Oslomet University Norway. At present, Sushil is pursuing a Ph.D. in the Department of Geoscience and Petroleum at the Norwegian University of Science and Technology (NTNU) in Norway. Given his solid foundation in both physics and data science, Sushil is well-prepared to tackle the challenges of using machine learning for dynamic well-log depth alignment and makes valuable contributions to the fields of petrophysics, geophysics, and machine learning.
Competencies
Research
Well log depth alignment is an important process in the oil and gas industry. It involves comparing the depth readings of two different types of well logs, one collected by measuring while drilling (MWD) and the other by electrical wireline logging (EWL), and ensuring that they are properly aligned to a same reference depth in order to accurately interpret the petrophysical data.
One way to align the well logs is through the use of bulk depth shift, which involves shifting the MWD well log by a certain amount in order to match it with the EWL well log. This is done by comparing the depth readings of the two logs at specific points and determining the amount of shift needed to align them. However, in some cases, bulk shift may not be sufficient to accurately align the well logs. In these cases, dynamic depth shifts may be needed.
Dynamic depth shifts involve stretching or squeezing the MWD well log in order to align it perfectly with the EWL well log. This is done by comparing the depth readings of the two logs over a longer distance and determining the amount of stretch or squeeze needed to align them. Dynamic depth shifts are more complex than bulk shifts, but they can provide a more accurate alignment of the well logs.
In this research, we will be using several tools to align the MWD and EWL well logs, including cross correlation, dynamic time warping, machine learning, and deep learning.
Cross correlation is a statistical technique that measures the similarity between two sets of data, and can be used to align the MWD and EWL well logs by comparing the depth readings of the two logs at specific points.
Dynamic time warping is a method that compares the shape of two time series, and can be used to align the MWD and EWL well logs by comparing the depth readings of the two logs over a longer distance.
Machine learning and deep learning are both forms of artificial intelligence that involve training algorithms to analyze data and make predictions or decisions. In this research, we will be using machine learning and deep learning to analyze the MWD and EWL well logs and determine the best way to align them.
Similar tasks can be found in other fields, such as geology and geophysics, where different types of data are collected and need to be aligned in order to accurately interpret the results. For example, in geology, different types of rock samples may be collected from different depths and need to be aligned in order to accurately interpret the rock layers. In geophysics, different types of data, such as seismic data and gravity data, may be collected from different locations and need to be aligned in order to accurately interpret the subsurface structure.
In summary, well log depth alignment between MWD and EWL well logs is an important process in the oil and gas industry, and can be achieved through the use of bulk depth shift or dynamic depth shifts. Similar tasks can be found in other fields, such as geology and geophysics, where different types of data are collected and need to be aligned in order to accurately interpret the results.