Intelligent cement design: Utilizing machine learning algorithms to assure effective long-term well integrity

Arash Shadravan, Mohammadali Tarrahi, Mahmood Amani

Research output: Chapter in Book/Report/Conference proceedingConference contribution

5 Citations (Scopus)

Abstract

Designing drilling fluids, spacers and cement slurries are all often done by trial and error in the laboratory. There are several test data that would not be used in this method and it is hard to digest a plethora of information for users and take intelligent and cost-effective decision to design a fluid with the desired properties. Therefore, trial and error method is considered to be time consuming, very costly and misleading. Today, there is a need for an intelligent system which uses all the available fluid design data stored in a database by which we can benefit from its insights for smart fluid designs. This predictive tool suggest a composition for drilling fluids, spacer fluid or cement slurry by implementing a machine learning algorithm on imported experimental data. This paper investigates and implements a data-driven predictive tool which uses Gaussian Process Regression (GPR). GPR as a novel machine learning method considerably reduces the costs of testing, optimizes the material use, integrates available experimental data and eliminates the user bias. This practical nonlinear regression method fosters an efficient and fast prediction analysis which do not require including complex physics of the underlying intricate chemical fluid behavior while integrating all available data from different databases. GPR has exceptional advantages over traditional regression methods since it does not require a known form for regression function. Also its capability of determining estimation error and confidence interval is unique. This machine learning based tool offers comprehensive insights for intelligent fluid design and considerably reduces the experiment cost. This study showcases an example through which of GPR predicts rheological properties and helped engineers to maintain the fluid rheological hierarchy for better cement jobs and well integrity. Zonal isolation has the utmost importance for carbon sequestration projects.

Original languageEnglish
Title of host publicationCarbon Management Technology Conference 2015: Sustainable and Economical CCUS Options, CMTC 2015
PublisherAIChE
Pages1102-1113
Number of pages12
Volume2
ISBN (Electronic)9781510818866
Publication statusPublished - 2015
Externally publishedYes
EventCarbon Management Technology Conference 2015: Sustainable and Economical CCUS Options, CMTC 2015 - Sugar Land, United States
Duration: 17 Nov 201519 Nov 2015

Other

OtherCarbon Management Technology Conference 2015: Sustainable and Economical CCUS Options, CMTC 2015
CountryUnited States
CitySugar Land
Period17/11/1519/11/15

Fingerprint

Learning algorithms
Learning systems
Cements
cement
Fluids
fluid
Drilling fluids
drilling fluid
cost
Costs
tool use
Slurries
Intelligent systems
machine learning
carbon sequestration
Error analysis
confidence interval
slurry
physics
Physics

ASJC Scopus subject areas

  • Media Technology
  • Industrial and Manufacturing Engineering
  • Environmental Science(all)

Cite this

Shadravan, A., Tarrahi, M., & Amani, M. (2015). Intelligent cement design: Utilizing machine learning algorithms to assure effective long-term well integrity. In Carbon Management Technology Conference 2015: Sustainable and Economical CCUS Options, CMTC 2015 (Vol. 2, pp. 1102-1113). AIChE.

Intelligent cement design : Utilizing machine learning algorithms to assure effective long-term well integrity. / Shadravan, Arash; Tarrahi, Mohammadali; Amani, Mahmood.

Carbon Management Technology Conference 2015: Sustainable and Economical CCUS Options, CMTC 2015. Vol. 2 AIChE, 2015. p. 1102-1113.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Shadravan, A, Tarrahi, M & Amani, M 2015, Intelligent cement design: Utilizing machine learning algorithms to assure effective long-term well integrity. in Carbon Management Technology Conference 2015: Sustainable and Economical CCUS Options, CMTC 2015. vol. 2, AIChE, pp. 1102-1113, Carbon Management Technology Conference 2015: Sustainable and Economical CCUS Options, CMTC 2015, Sugar Land, United States, 17/11/15.
Shadravan A, Tarrahi M, Amani M. Intelligent cement design: Utilizing machine learning algorithms to assure effective long-term well integrity. In Carbon Management Technology Conference 2015: Sustainable and Economical CCUS Options, CMTC 2015. Vol. 2. AIChE. 2015. p. 1102-1113
Shadravan, Arash ; Tarrahi, Mohammadali ; Amani, Mahmood. / Intelligent cement design : Utilizing machine learning algorithms to assure effective long-term well integrity. Carbon Management Technology Conference 2015: Sustainable and Economical CCUS Options, CMTC 2015. Vol. 2 AIChE, 2015. pp. 1102-1113
@inproceedings{9963ed2600e646ef81962c163a47ceb0,
title = "Intelligent cement design: Utilizing machine learning algorithms to assure effective long-term well integrity",
abstract = "Designing drilling fluids, spacers and cement slurries are all often done by trial and error in the laboratory. There are several test data that would not be used in this method and it is hard to digest a plethora of information for users and take intelligent and cost-effective decision to design a fluid with the desired properties. Therefore, trial and error method is considered to be time consuming, very costly and misleading. Today, there is a need for an intelligent system which uses all the available fluid design data stored in a database by which we can benefit from its insights for smart fluid designs. This predictive tool suggest a composition for drilling fluids, spacer fluid or cement slurry by implementing a machine learning algorithm on imported experimental data. This paper investigates and implements a data-driven predictive tool which uses Gaussian Process Regression (GPR). GPR as a novel machine learning method considerably reduces the costs of testing, optimizes the material use, integrates available experimental data and eliminates the user bias. This practical nonlinear regression method fosters an efficient and fast prediction analysis which do not require including complex physics of the underlying intricate chemical fluid behavior while integrating all available data from different databases. GPR has exceptional advantages over traditional regression methods since it does not require a known form for regression function. Also its capability of determining estimation error and confidence interval is unique. This machine learning based tool offers comprehensive insights for intelligent fluid design and considerably reduces the experiment cost. This study showcases an example through which of GPR predicts rheological properties and helped engineers to maintain the fluid rheological hierarchy for better cement jobs and well integrity. Zonal isolation has the utmost importance for carbon sequestration projects.",
author = "Arash Shadravan and Mohammadali Tarrahi and Mahmood Amani",
year = "2015",
language = "English",
volume = "2",
pages = "1102--1113",
booktitle = "Carbon Management Technology Conference 2015: Sustainable and Economical CCUS Options, CMTC 2015",
publisher = "AIChE",

}

TY - GEN

T1 - Intelligent cement design

T2 - Utilizing machine learning algorithms to assure effective long-term well integrity

AU - Shadravan, Arash

AU - Tarrahi, Mohammadali

AU - Amani, Mahmood

PY - 2015

Y1 - 2015

N2 - Designing drilling fluids, spacers and cement slurries are all often done by trial and error in the laboratory. There are several test data that would not be used in this method and it is hard to digest a plethora of information for users and take intelligent and cost-effective decision to design a fluid with the desired properties. Therefore, trial and error method is considered to be time consuming, very costly and misleading. Today, there is a need for an intelligent system which uses all the available fluid design data stored in a database by which we can benefit from its insights for smart fluid designs. This predictive tool suggest a composition for drilling fluids, spacer fluid or cement slurry by implementing a machine learning algorithm on imported experimental data. This paper investigates and implements a data-driven predictive tool which uses Gaussian Process Regression (GPR). GPR as a novel machine learning method considerably reduces the costs of testing, optimizes the material use, integrates available experimental data and eliminates the user bias. This practical nonlinear regression method fosters an efficient and fast prediction analysis which do not require including complex physics of the underlying intricate chemical fluid behavior while integrating all available data from different databases. GPR has exceptional advantages over traditional regression methods since it does not require a known form for regression function. Also its capability of determining estimation error and confidence interval is unique. This machine learning based tool offers comprehensive insights for intelligent fluid design and considerably reduces the experiment cost. This study showcases an example through which of GPR predicts rheological properties and helped engineers to maintain the fluid rheological hierarchy for better cement jobs and well integrity. Zonal isolation has the utmost importance for carbon sequestration projects.

AB - Designing drilling fluids, spacers and cement slurries are all often done by trial and error in the laboratory. There are several test data that would not be used in this method and it is hard to digest a plethora of information for users and take intelligent and cost-effective decision to design a fluid with the desired properties. Therefore, trial and error method is considered to be time consuming, very costly and misleading. Today, there is a need for an intelligent system which uses all the available fluid design data stored in a database by which we can benefit from its insights for smart fluid designs. This predictive tool suggest a composition for drilling fluids, spacer fluid or cement slurry by implementing a machine learning algorithm on imported experimental data. This paper investigates and implements a data-driven predictive tool which uses Gaussian Process Regression (GPR). GPR as a novel machine learning method considerably reduces the costs of testing, optimizes the material use, integrates available experimental data and eliminates the user bias. This practical nonlinear regression method fosters an efficient and fast prediction analysis which do not require including complex physics of the underlying intricate chemical fluid behavior while integrating all available data from different databases. GPR has exceptional advantages over traditional regression methods since it does not require a known form for regression function. Also its capability of determining estimation error and confidence interval is unique. This machine learning based tool offers comprehensive insights for intelligent fluid design and considerably reduces the experiment cost. This study showcases an example through which of GPR predicts rheological properties and helped engineers to maintain the fluid rheological hierarchy for better cement jobs and well integrity. Zonal isolation has the utmost importance for carbon sequestration projects.

UR - http://www.scopus.com/inward/record.url?scp=84981207011&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84981207011&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84981207011

VL - 2

SP - 1102

EP - 1113

BT - Carbon Management Technology Conference 2015: Sustainable and Economical CCUS Options, CMTC 2015

PB - AIChE

ER -