Return to search

Development of an Integrity Evaluation System for Wells in Carbon Sequestration Fields

<p> Carbon sequestration is a promising solution to mitigate the accumulation of greenhouse gases. Depleted oil and gas reservoirs are desirable vessels for carbon sequestration. It is crucial to maintain the sealing ability of carbon sequestration fields with high concentrations of CO<sub>2</sub>.</p><p> A systematic well integrity evaluation system has been developed and validated for carbon sequestration fields. The system constitutes 1) a newly developed analytical model for assessing cement sheath integrity under various operating conditions, 2) quantifications of well parameters contributing to the probability of well leakage, and 3) genetic-neural network algorithm for data analysis and well-leakage probability assessment.</p><p> A wellbore system consists of well casing, cement sheath, and formation rock. A new analytical stress model was developed. The new analytical model solves for the stresses in the casing-cement sheath formation system loaded by the isotropic and anisotropic horizontal in-situ stresses. Further analyses with the analytical model reveal that Young&rsquo;s modulus of cement sheath is a major factor that contributes to the sealing ability of the cement sheath, while Poisson&rsquo;s ratio and cohesion play less important roles in the cement sheath sealing ability. The cement sheath in the shale formation exhibits higher sealing ability than that in the sandstone formation. The sealing ability of weak cement is higher than that of strong cement.</p><p> Descriptive quantifications of well parameters were made in this study for analyzing their effect on the probability of well leakage. These parameters include well cement placement relative to aquifers and fluid reservoir zones, cement type, cement sheath integrity in operating conditions, well aging, and well plugging conditions. It is the combination of these parameters that controls the probability of well leakage. A significant proportion of wells were identified as risky wells in these two fields. It is concluded that the well trained neural network model can be used to predict the well leakage risk over the CO<sub>2</sub> sequestration lifespan, which can promote prevention activities and mitigations to the CO<sub>2</sub> leakage risky wells.</p>

Identiferoai:union.ndltd.org:PROQUEST/oai:pqdtoai.proquest.com:10003769
Date03 February 2016
CreatorsLi, Ben
PublisherUniversity of Louisiana at Lafayette
Source SetsProQuest.com
LanguageEnglish
Detected LanguageEnglish
Typethesis

Page generated in 0.0015 seconds