Immiscible displacement of trapped oil through experimental and data mining techniques.

Abunumah, Ofasa

Abstract

Extensive experimental and data mining techniques have been applied to investigate the potential and competitiveness of gases used in immiscible gas-enhanced oil recovery (EOR) processes. Methane (CH4), Nitrogen (N2), Air (21%O2/N2) and Carbon Dioxide (CO2) are some of the gases injected in reservoirs to displace trapped oil from reservoir pores. The EOR screening process has been well-documented in the literature. However, for immiscible gas EOR technology, very few resources are available for evaluating the selection and performance criteria for commonly-injected EOR gases; immiscible EOR gases are usually lumped up in published screening models, and the gases are reportedly selected based on availability and accessibility, rather than on technical criteria such as displacement efficiency. Furthermore, available experimental studies have investigated EOR gases only separately. This research has been able to fill these gaps and more, through rigorous data mining and gas experiments processes. The methodology utilised empirical approaches set in three phases. Phase I applied data mining techniques to 10,850 data from 484 EOR field projects, to identify twenty-four EOR geological and engineering quantities, and objective functions. Phase II utilised Phase I outcomes to design and execute a set of rigorous gas experiments, involving 1,920 experimental runs (comprising five reservoir analogous core samples, eight gases, eight isobars and six isotherms), to generate and analyse 15,360 experimental data points. Several established and modified constitutive equations were used to model gas responses to EOR geological and engineering quantities. In Phase III, Phase I and Phase II results were coupled for the purpose of knowledge validation and application. This research's outcomes have contributed to reservoir engineering practice and knowledge in providing useful information on EOR gases' competitiveness. Results from Phase I indicate that immiscible gas EOR can be unbundled through data mining and clustering techniques. A novel screening model has been developed for immiscible gas EOR that incorporates sensitivity and criticality markers for each petrophysical quantity investigated. It has been demonstrated in Phase II that, in a heterogeneous system, CH4 is the most competitive gas for ten geological and engineering quantities and objective functions, such as Volumetric Rate, Interstitial Velocity, and Well Density. Similarly, CO2 is most competitive for ten other quantities investigated, such as Mobility and Interstitial Momentum. N2 is the most competitive for the cost of injected gas per area coverage. Air is second-best for several objective functions. Suffice to state that at some structural settings and operational conditions (such as porosity, pore size, surface area and temperature), the competitiveness ranking of the gases switches position. Such was observed between N2 and CO2 in low porosity (4% and 3%) core samples. EOR gas mixtures and non EOR gases - such as 20% CH4/N2, He, and Ar - were added to the experiments to investigate the relationship between gas flow and gas properties. It was observed that the structural variability (heterogeneity) of the system distorts the correlation between gas properties, such as molecular weight, and the performance criteria of the respective gases. The results from Phase I and II couple significantly in Phase III. Based on well number and placement, it has been demonstrated that the well placement of CH4, CO2, and Air favours a negative pore size gradient, while N2 favours a positive gradient. The economic analysis demonstrates that CO2 incurs the least cumulative injectant cost and the highest capital expenditure cost (CAPAX). The three Phases validate the field and laboratory well density profile. CH4 requires the least well density (0.2 well/acre, 1.0 well/cm2) compared to CO2 (0.7 well/acre, 2.0 well/cm2). In some analyses, it was discovered that gas mixture, such as 20%CH4/N2, performs better than when the individual component gas acted alone. Single-phase and two-phase relationships have been analytically and experimentally coupled. The experimental findings at low pressure could also lend utility to the gas separation, fluidised bed, and catalytic reaction processes and industry.

Citation

ABUNUMAH, O. 2021. Immiscible displacement of trapped oil through experimental and data mining techniques. Robert Gordon University, PhD thesis. Hosted on OpenAIR [online]. Available from: https://doi.org/10.48526/rgu-wt-1987936

• Thesis Type: Thesis
• Deposit Date: Jun 15, 2023
• Publicly Available Date: Jun 15, 2023
• DOI: https://doi.org/10.48526/rgu-wt-1987936
• Keywords: Petroleum engineering; Oil recovery; Immiscible gases; Oil wells
• Public URL: https://rgu-repository.worktribe.com/output/1987936
• Award Date: Oct 31, 2021