Designing neural networks for the prediction of the drilling parameters for Kuwait oil and gas fields

In this study a new methodology was developed to predict the drilling parameters using the Artificial Neural Network. Three models were developed to predict bit type, rate of penetration (ROP), and cost-per-foot (cost/ft), respectively.;The prediction of bit type and other drilling parameters from the current available data is an important criterion in selecting the most cost efficient bit. History of bit runs plays an important factor in bit selection and bit design. Based on field data, the selection of bit type can be accomplished by the use of a neural network as an alternative bit selection method.;Three drilling parameters were modeled with data from different fields located in Kuwait. Results show that the drilling parameters of the new well can be predicted with the neural network models developed from the previous wells, a cost efficient alternative.

Name thesis:degree_name

MS

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Petroleum and Natural Gas Engineering

Year dc:date.available

2000

Author dc:creator

• Al-Rashidi, Abdulrahman F.

Contributors dc:contributor

• H. I. Bilgesu.

• Petroleum engineering
• Artificial intelligence

Identifier

https://researchrepository.wvu.edu/etd/1101

OAI identifier oai:identifier

oai:researchrepository.wvu.edu:etd-2104