Essential Components of an Integrated Data Mining Tool for the Oil and Gas Industry with an Example Application in the DJ Basin

Author

Department of Petroleum and Natural Gas Engineering,West Virginia University

Abstract

Data mining seems to be the new buzz word. During the past several years many industries
other than the oil and gas industry have realized the potential bene ts of da5ta mining and
have established sophisticated operations in order to implement this exciting technology in their
respective organizations. Data mining is not new. It has been around for many years. What
is new about its current implementation is the incorporation of machine learning techniques.
The oil and gas industry has become familiar with machine learning techniques since the early
1990s. Neural networks, genetic optimization and fuzzy logic have been used in numerous
applications, from well log interpretations to hydraulic fracturing optimization. Therefore, the
new interest in data mining in this industry is not surprising. The industry is at its peak state
for bene ting from what data mining has to o er, thanks to an abundance of digital data. A
word of caution is in order, which is the main motivation behind writing this paper. As with
many other new tools and technologies, the term \Data Mining" can be, and is currently being,
misused on several occasions. In this paper, an attempt has been made to answer questions such
as; what is Data Mining? How can it be accomplished? What are the essential components of
an integrated data mining process and what would be the bene ts of such a process? In addition
to answering questions such as those mentioned above, this paper will provide a road map (a
set of guidelines) for a successful data mining project. Finally, the paper concludes by applying
the presented guidelines to a hydraulic fracturing data set in the DJ basin of the United States
Rockies for a data mining study.