Application of artificial neural networks to predict the downhole inclination in directionally drilled geothermal wells

Download

index.pdf

Date

2018

Author

Burak, Tunç

Metadata

Show full item record

Item Usage Stats

273
views

90
downloads

Drilling directionally through naturally fractured geothermal reservoirs is a challenging task due to unexpected changes in inclination and azimuth of the well axis, which causes inefficient weight on bit transfer, decrease in penetration rate, increasing the risk of stuck pipe and problems in while running casings. To predict the sudden changes in inclination while drilling, a back propagation, feed forwarded multi layered artificial neural network (ANN) model, which uses drilling data collected from 12 J-type directionally drilled geothermal wells from Büyük Menderes Graben was developed. The training dataset consisted of 7600 individual drilling data. During the training process, effects of each drilling parameter on inclination were investigated with different scenarios for different hole sizes. Moreover, inclination predictions were carried out for a field case in which kick off point to the target depth with 30 meters survey intervals and results were compared. It has been found that developed ANN model provided satisfactory results based on the mean-square-error (MSE) value which was measured to check accuracy and quality of each training. The MSE of the training data set is 0.42% and the neural networks predicts the testing data with 1.19% MSE value. According to the sensitivity analysis, it has been found out that as WOB, Bit Revolution Per Minute (RPM) and Stand Pipe Pressure (SPP) increase, inclination increases. On the other hand, increment in flow rate leads to drop in inclination. Moreover, the result of the case study was 0.59% MSE which concludes that network is not memorizing the data. In addition, different ANN’s were created by omitting some drilling parameters to analyze individual effects of each parameter on network accuracy. The results indicated that, Total Flow Area (TFA), International Association of Drilling Contractors (IADC) code and Weight on Bit (WOB) have the highest impact on network dataset when compared to other drilling parameters.

Subject Keywords

Geothermal well drilling., Geothermal engineering., Neural networks (Computer science).

URI

http://etd.lib.metu.edu.tr/upload/12622101/index.pdf
https://hdl.handle.net/11511/27306

Collections

Graduate School of Natural and Applied Sciences, Thesis

Suggestions

OpenMETU
Core

A Comparative study of cuttings transport performance of water versus polymer-based fluids in horizontal well drilling Allahvirdizadeh, Payam; Parlaktuna, Mahmut; Kuru, Ergun; Department of Petroleum and Natural Gas Engineering (2015) High drilling fluid circulation rate is often needed for effective transportation of cuttings in horizontal and extended reach wells, which may not be always achievable due to the risk of fracturing the rock by increased bottom hole dynamic pressure and also limit of pumps capacity. Keeping the bottom hole pressure low enough while increasing the flow rate is, therefore, a major challenge in horizontal well drilling operations. A potential solution to this problem would be to use drag reducing additives in ...
Comprehensive approach to torque and lost circulation problems in geothermal wells in terms of drilling fluid Sönmez, Ahmet; Kök, Mustafa Verşan; Bal, Berk; Bagatır, Gökhan; Gücüyener, I.Hakkı (2021-09-01) Drilling in a lost circulation zone has long been a challenge in geothermal wells due to its strong potential for high torque and wellbore instability. Particularly in deviated, extended reach and horizontal wells, frictional forces between drill string and wellbore or casing may be high enough to interfere with running the pipe in and out of the hole. These forces may also increase the torque and the stress on drill string which may end up with severe problems like stuck pipe and lost in hole. High-tempera...
Comprehensive approach to torque and lost circulation problems in geothermal wells in terms of drilling fluid Sönmez, Ahmet; Kök, Mustafa Verşan; Department of Petroleum and Natural Gas Engineering (2020-7) Drilling in a lost circulation zone has long been a challenge in geothermal wells due to its strong potential for high torque and wellbore instability. Excessive torque and friction values cause overpull, pipe stuck and severe problems ending up with losing the well in some cases. High temperature-high pressure (HTHP) drilling fluids design; which minimize loss rates and friction values, is critical for the success of these challenging drilling practices. Known fact is that, oil/synthetic based drilling flu...
3D Modeling of Ripping Process Basarir, Hakan; Karpuz, Celal; Tutluoglu, Levent (American Society of Civil Engineers (ASCE), 2008-01-01) Due to environmental constraints and limitations on blasting, ripping as a ground loosening and breaking method has become more popular than drilling and blasting method in both mining and civil engineering applications. The best way of estimating the rippability of rocks is to conduct direct ripping runs in the field. However, it is not possible to conduct direct ripping runs in all sites using different dozer types. Therefore, the utilization of numerical modeling of ripping systems becomes unavoidable. A...
Application of fluorescent melamine resin microspheres for use as a geothermal tracer Salar, Mustafa; Akın, Serhat; Department of Petroleum and Natural Gas Engineering (2018) Tracer testing of flow between injection and production wells is an effective tool to map fluid flow pathways in a geothermal reservoir. Tracer concentration curves can be used to provide insight into projected thermal drawdown in the reservoir or thermal breakthrough of re-injected fluids, estimating inter-well volumes and flow geometries. Although radioactive tracers have been used in the past, many tracer tests now use chemical tracers that are less hazardous to handle. Fluorescent organic compounds such...

Citation Formats

T. Burak, “Application of artificial neural networks to predict the downhole inclination in directionally drilled geothermal wells,” M.S. - Master of Science, Middle East Technical University, 2018.