Computers & Geosciences 28 (2002) 261–269 Application of artificial neural networks to optimum bit selection Serkan Yilmaz, Cem Demircioglu, Serhat Akin* Petroleum and Natural Gas Engineering Department, Middle East Technical University, Inonu Bulvari 06531, Ankara, Turkey Received 1 August 2000; received in revised form 1 May 2001; accepted 18 June 2001 Abstract Optimum bit selection is one of the important issues in drilling engineering. Usually, optimum bit selection is determined by the lowest cost per foot and is a function of bit cost and performance as well as penetration rate. Conventional optimum rock bit selection program involves development of computer programs created from mathematical models along with information from previously drilled wells in the same area. Based on the data gathered on a daily basis for each well drilled, the optimum drilling program may be modified and revised as unexpected problems arose. The approach in this study uses the power of Artificial Neural Networks (ANN) and fractal geostatistics to solve the optimum bit selection problem. In order to achieve this goal a back-propagation ANN model was developed by training the model using real rock bit data for several wells in a carbonate field. The training and fine- tuning of the basic model involved use of both gamma ray and sonic log data. After that the model was tested using various drilling scenarios in different lithologic units. It was observed that the model provided satisfactory results. r 2002 Elsevier Science Ltd. All rights reserved. Keywords: Neural networks; Rock bit; Optimization; Fractals; Petroleum engineering I. Introduction As the general trend in the oil industry is shifting towards cutting expenditure due to fluctuating oil prices, challenges to optimize drilling performance is getting greater. One of the main challenges in drilling is optimizing bit performance. Although bit cost comprises a relatively small fraction in a well’s budget (+5%), the impact of bit performance on overall well cost can be significant. Selecting the most appropriate bit for any well section is a key feature in achieving superior drilling performance, thus cutting expenditures (Estes, 1973; Lummus, 1970). Several different methods are intro- duced for optimum bit selection program. The classical way to select a drilling bit is based on all existing data related to cost per foot (CPF), specific energy (SE), bit dullness, offset-well bit records and geological informa- tion (Rabia, 1985). CPF is function of bit cost, conditions under which operating costs are assigned, area where bit is run, operational environment, and drilling parameters. The following equation can be used to obtain CPF ($/bbl). CPF ¼ B þðT þ tÞR F ; ð1Þ where B is bit cost ($), T is trip time in hours, t is rotating time in hours, R is rig cost per hour ($/h) and length of section drilled (footage), ft. CPF directly affects drilling economics, but it cannot establish a direct correlation between technological advancements and performance. Moreover, the CPF concept cannot be used for directional and horizontal drilling programs. Specific energy establishes a relationship between bit performance and bit energy requirements. It is defined as *Corresponding author. Tel.: 90-312-210-4892; fax: 90-312- 210-1271. E-mail address: serhat@metu.edu.tr (S. Akin). 0098-3004/02/$-see front matter r 2002 Elsevier Science Ltd. All rights reserved. PII:S0098-3004(01)00071-1