Authors :
Ijegwa David Acheme; Osemengbe Oyaimare Uddin; Ayodeji Samuel Makindes
Volume/Issue :
Volume 8 - 2023, Issue 10 - October
Google Scholar :
https://tinyurl.com/3fuch93w
Scribd :
https://tinyurl.com/4by5ptzs
DOI :
https://doi.org/10.5281/zenodo.10088393
Abstract :
The drilling phase has been reported to be the
most expensive phase of oil exploration and production,
hence several research efforts have been targeted at
improving its efficiency. The rate of penetration (ROP)
has also been identified as the most important metric for
improving drilling performance, hence, several research
efforts have reported different methods of predicting
ROP optimal values. Recently, artificial intelligence (AI)
and machine learning (ML) models have been reported
for the prediction of ROP. However, the ROP is
influenced by several factors, and the interactions among
these factors introduces a kind of complexity that affects
its accurate prediction. This research work sets out to
achieve two important objectives, firstly, to investigate
and rank the most important factors for the prediction
of the ROP, and secondly, to carry out a comparative
study and ranking of selected machine learning
algorithms for the prediction of ROP. In order to achieve
this, the open source volve dataset which is a complete
set of data from the North Sea oil field was utilized.
Eighteen (18) machine learning models were built using
this dataset and their performances compared. The
result showed the random forest regressor with an
RMSE value of 0.0010 and R2 score of 0.891 as the most
efficient algorithm among the eighteen chosen for this
work. Further experimentation also revealed the most
influential factors for predicting the rate of penetration,
these features in order of importance are; measured
depth, bit rotation per minute, formation porosity, shale
volume, water saturation, log permeability. The output
of this study work offers a blueprint for choosing
algorithms and features when implementing ML
solutions for optimizing oil drilling, and this is helful in
the development of real-time ROP prediction models and
hybridization.
Keywords :
Rate of Penetration Prediction, oil drilling, machine learning, feature selections
The drilling phase has been reported to be the
most expensive phase of oil exploration and production,
hence several research efforts have been targeted at
improving its efficiency. The rate of penetration (ROP)
has also been identified as the most important metric for
improving drilling performance, hence, several research
efforts have reported different methods of predicting
ROP optimal values. Recently, artificial intelligence (AI)
and machine learning (ML) models have been reported
for the prediction of ROP. However, the ROP is
influenced by several factors, and the interactions among
these factors introduces a kind of complexity that affects
its accurate prediction. This research work sets out to
achieve two important objectives, firstly, to investigate
and rank the most important factors for the prediction
of the ROP, and secondly, to carry out a comparative
study and ranking of selected machine learning
algorithms for the prediction of ROP. In order to achieve
this, the open source volve dataset which is a complete
set of data from the North Sea oil field was utilized.
Eighteen (18) machine learning models were built using
this dataset and their performances compared. The
result showed the random forest regressor with an
RMSE value of 0.0010 and R2 score of 0.891 as the most
efficient algorithm among the eighteen chosen for this
work. Further experimentation also revealed the most
influential factors for predicting the rate of penetration,
these features in order of importance are; measured
depth, bit rotation per minute, formation porosity, shale
volume, water saturation, log permeability. The output
of this study work offers a blueprint for choosing
algorithms and features when implementing ML
solutions for optimizing oil drilling, and this is helful in
the development of real-time ROP prediction models and
hybridization.
Keywords :
Rate of Penetration Prediction, oil drilling, machine learning, feature selections
