Data Mining: Applications in the Petroleum Industry

Rating:
Summary: The First Thorough Treatment for E&P
Review: For years engineers have struggled to describe nature using closed form, analytic expressions. Now data mining gives us the ability to observe patterns in nature without regard to modeling the underlying physical system. The upstream oil and gas industry has been late to adopt artificial intelligence techniques for data mining, partly because no one source demystifies the topic. Here at last is a book that treats the subject from a petroleum engineering perspective, complete with case studies.

The authors give an overview of data mining, talk about some of the tools available, and some simpler models before launching into a thorough examination of data mining in reservoir surveillance and reservoir modeling and forecasting. An excellent chapter on data preparation and quality control emphasizes the importance of this often overlooked aspect. Two case studies, one on permeability prediction from well logs and another on re-stimulation candidate selection, round out the book. The appendices treat data and databases as well as available data mining software.

This book is an important addition to the body of knowledge available to petroleum engineers on the topic of data mining using artificial intelligence techniques, and should be in the library of anyone interested in the topic. The sections on fuzzy logic, self-organizing maps, and genetic algorithm optimization alone are worth the purchase price.

Although the authors' company produces and markets one of the best purpose built software packages available for the E&P industry their treatment of the topics covered here is fair and balanced, and concentrates on concepts rather than specific tools.

I highly recommend this book for anyone in the oil patch.

Rating:
Summary: Fairly mediocre
Review: I found this book to be fairly mediocre - not very deep in its content and skipping by the hard problems. Typical book written by a software vendor who would like to seem like a domain expert, while in reality they are only looking to promote their product. On the bright side it's a book about applications of data mining, which is always interesting - I only wish it was worthwhile to read, especially at the ludicrous price of $99.