Improving cement wellbore integrity with nanomaterials: Design of experiments and machine learning techniques

Date

2020-12

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

The cementing process is an integral part of drilling a wellbore which ensures wellbore integrity throughout the life of the well. Failure of the cement sheath can lead to hazardous rig operations and deleterious environmental issues. Embedding strength enhancing nanomaterials into the cement slurry is a promising approach for creating multifunctional, durable composites capable of withstanding stressful wellbore conditions. In recent years carbon nanotubes (CNTs) and carbon nanofibers (CNFs) have received ubiquitous attention in oil well cement research. However, due to their high hydrophobicity and tendency for graphene material to agglomerate, obtaining an adequate dispersion is an arduous task and the cost is substantially high. This study focuses on embedding alumina nanofibers (ANFs), a relatively newly discovered material, in the cement slurry. The dispersibility of ANFs were assessed before implementation into the cement matrix. Afterwards, cement specimens were singly-reinforced with ANFs and cured in simulated wellbore conditions with various properties tested upon removal. Also, since cement failure is a multistep process, hybrid-reinforcement was implemented on the nanoscale and microscale levels. Sol-gel treated micro-synthetic polypropylene (PP) fibers were added as micro reinforcement and the mechanical properties of the cement composite was assessed. Considering the mixture design is a multi-variable and multi-objective optimization formulation, the response surface method (RSM) through the design of experiment (DOE) was utilized. The applicability of using supervised machine learning to predict the unconfined compressive strength (UCS) of cement samples was also assessed. 195 cement samples were embedded with varying dosages of strength enhancing pre-dispersed nanoparticles consisting of nanosilica (nano-SiO2), nanoalumina (nano-Al2O3), and nanotitanium dioxide (nano-TiO2) at various simulated wellbore temperatures. The developed model can replace, or be used in combination with, destructive UCS tests which can save the petroleum industry time, resources, and capital.

Description

Keywords

Petroleum engineering, Machine learning, Nanomaterials, Wellbore integrity

Citation