Statistical regression model for estimating the rate of penetration of horizontal well in Al-Halfaya oil field

Rate of penetration models are essential to obtain a maximum drilling rate and minimum drilling cost. Traditional rate of penetration models has limited field application because they build on specific condition and did not incorporated logging data to develop models for dynamic drilling processes....

Full description

Saved in:
Bibliographic Details
Main Author: AL-Bahadly, Kadhim Hmood Mnaty
Format: Conference Proceeding
Language:English
Subjects:
Correlation coefficients
Data analysis
Data logging
Data processing
Dependent variables
Drilling muds
Horizontal wells
Independent variables
Modelling
Oil fields
Outliers (statistics)
Regression analysis
Regression models
Root-mean-square errors
Statistical analysis
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Rate of penetration models are essential to obtain a maximum drilling rate and minimum drilling cost. Traditional rate of penetration models has limited field application because they build on specific condition and did not incorporated logging data to develop models for dynamic drilling processes. The aim of this study is to construct rate of penetration models for AL-halfaya oil field by using routinely available mud logging data from a range of drilling conditions. The data used in this study are provided by PetroChina Company Limited Iraq Branch (from Contractor Bohai Mudlogging), that working in AL-Halfaya oil field south of Iraq, for the horizontal well called (HF119-M119 H) for Mishraf formation. This formation represents group data. A Statistical software called (Statistical Analysis System (SAS)) was used to perform statistical calculation and regression analysis of modeling data. For regression purpose, the ROP was considered as a dependent variable, while the (WOB, RPM, HSI and Torque) were considered as independent variables. A data set are considered in the modeling which represented formation. Systematic statistical approaches have been developed in this study for handling the logging data. These approaches include examination of data, removing outliers, make moving-average and grouping data. Data processing analysis were performed in various methods according to the completed building model. The values of ROP were predicted based on grouping data. The results of R2 were calculated to be around (0.842) and RMSE (0.74) for a group set data. The Root Mean Squared Error (RMSE) and the correlation coefficient (R2) were indicating the strength of the model. It was a comparison between the actual values and the predict values of ROP of the model.
ISSN:0094-243X
1551-7616
DOI:10.1063/5.0143937