Crucial Factors for in-Situ CO2 Monitoring in CO2-EOR Projects Using Pulsed Neutron Well Logging

Abstract

Carbon dioxide (CO2) injection for enhanced oil recovery (EOR), also known as CO2-EOR, has become increasingly important due to the growing need for CO2 utilization and sequestration. CO2 monitoring is an integral part of the CO2- EOR process. Pulsed neutron (PN) well logging is an efficient and effective technology for understanding subsurface CO2 propagation and quantifying multiphase saturation. This paper discusses the critical factors—well conditions and reservoir properties—in designing a PN well logging program and analyzing PN data. The Monte Carlo N-Particle (MCNP) simulation is a stochastic forward modeling method that generates PN tool responses under diverse well and formation conditions. With a series of MCNP models, including perturbations of various well logging environments, the characteristics of key PN measurements were delineated. This enabled the establishment of best practices in PN well logging operations and data analysis for in-situ CO2 profiling. The requirement for the PN tool is to be slim in terms of the outer diameter (i.e., 1.69 inches), allowing through-tubing deployment, and to have three scintillation gamma-ray detectors, enhancing the formation sensitivity compared to traditional dual detector-based PN tools. We constructed MCNP models of time-spectra-based well logs; inelastic and thermal neutron capture logs were simulated considering several vital parameters—wellbore fluid, formation lithology, annular space materials, in-situ oil and CO2 densities, and reactions between CO2 and formation minerals and resident fluid. The type of wellbore fluid is water, CO2, or a mixture of the two depending on the well type—injection, monitoring, or production well. Although a water-filled wellbore is optimal for PN well logging, a CO2- filled wellbore does not adversely impact CO2 monitoring to a high degree if a sleeved-PN tool is used. As formation lithology types and water salinity influence inelastic and capture PN measurements differently, determining which PN log should be used for CO2 saturation analysis is essential. The impact of the shale volume and properties on the PN data is nonuniform, so this factor must also be carefully reviewed in conjunction with the lithology type. Furthermore, accurate estimation of oil and CO2 densities under downhole conditions minimizes the uncertainty of multiphase fluid saturation characterization. Finally, considering the effects of mineral alteration and formation dry-out is required when evaluating saturation in post-injection stages. It is crucial to consider well- and formation- specific factors to ensure accurate monitoring of in-situ CO2 propagation and multiphase fluid volume variation with PN well logging. The best practices for PN well logging and data analysis were evaluated using a customized MCNP modeling technique. Furthermore, surveilling the behavior of injected CO2 on a well basis enables the optimization of CO2 injection parameters and the updating of reservoir models on a field-scale level.