CO2 miscible injection is generally one of the most efficient enhanced oil recovery (EOR) methods and widely used in the conventional oil reservoirs. The applicability of CO2 EOR technology for unlocking the resources from unconventional tight and shale formations and the mechanisms of miscible flooding in these reservoirs still remain unclear. An important parameter used to evaluate the feasibility of CO2 miscible flooding is the minimum miscibility pressure (MMP). Even though experimental approaches, empirical correlations and theoretical methods have performed well in measuring or predicting MMP between CO2 and crude oil in conventional reservoirs, they may not be suitable for unconventional formations as phase behavior and MMP can be significantly affected by confinement effect in small pores (e.g., nanopores) in such formations.
In this study, a new MMP prediction model based on the modified Parachor Model associated with the Perturbed-Chain Statistical Associating Fluid Theory (PC-SAFT) is developed to determine CO2 MMP both in the bulk phase and nanopores. The Parachor Model is modified to account for the confinement effect of nanopore walls on the equilibrium interfacial tension (IFT). The Equilibrium IFT reduction in nanopores is related to a temperature-dependent and slit pore width-dependent modification term. The parameters of the new Parachor Model are determined by matching the vapor-liquid surface tension values for CH4, C2H6, C3H8,
The newly developed model successfully reproduces MMP in bulk phase as compared with both other methods and experimental data. The overall average absolute relative deviation (AARD) for MMP is within 8 %. The calculated equilibrium IFT for liquid-vapor phase has a good agreement with molecular simulation results. For Bakken oil-CO2 system, if the slit pore width is larger than 10 nm, MMP is independent on pore width; otherwise, it decreases significantly with the decrease of the pore width. If pore width decreases to 3 nm, 67.5 % decrease in the IFT is observed and 23.5% reduction is achieved for MMP between Bakken oil and CO2 stream, indicating that it is easier to reach miscibility in nanopores, and CO2 miscible flooding might be a promising enhanced oil recovery (EOR) technology for tight oil and shale oil reservoirs. Furthermore, MMP increases with an increase of temperature in bulk phase, whereas IFT and MMP decrease with an increase of temperature in nanopores.
As one of the unconventional resources, tight oil has become one of the most important contributor of oil reserves and production growth. The successful commercial production of tight oil is mainly reliant on the advancement in horizontal drilling and multistage hydraulic fracturing technique. Development of tight oil reservoirs remains in an early stage. Primary oil recovery factor in these reservoirs is very low, leaving substantial volume of oil trapped underground due to the low porosity, low permeability characteristic of tight oil reservoirs. Thus, investigation of enhanced oil recovery methods is more than imperative in tight oil reservoirs. CO2 Huff-and-Puff technology has been effectively applied in conventional reservoirs and can be tailored to adapt for the characteristics of tight oil reservoirs.
In this study, the performance of water flooding in tight oil reservoir is studied and compared with that of the CO2 Huff-and-Puff process. Sensitivity analysis demonstrates that the performance of CO2 Huff-and-Puff is more sensitive to the length of gas injection and production step in each cycle, compared to the soaking time. The CO2 Huff-and-Puff process is optimized and an adaptive CO2 Huff-and-Puff process is conducted for tight oil reservoirs after primary production. Simulation results show that the adaptive cycle length CO2 Huff-and-Puff process can improve the incremental oil recovery by 11.1% over a fixed cycle length process. Finally, the inter-well interference during CO2 Huff-and-Puff is studied, and it is found that a multi-well asynchronous CO2 Huff-and-Puff pattern can improve the incremental oil recovery by 31.6% over that of a synchronous pattern.