Regular Article
Factorial two-stage analyses of parameters affecting the oil–gas interface and miscibility in bulk phase and nanopores

https://doi.org/10.1016/j.jcis.2019.07.109Get rights and content

Abstract

In this paper, a factorial analysis approach is applied to characterize the potential single and interactive factors as well as their effects on the interface and miscibility of three light oil–CO2 systems under 32 different conditions. First, a modified Peng–Robinson equation of state coupled with the parachor model is applied to calculate the vapour–liquid equilibrium and interfacial tensions (IFTs) at a variation of pore radii and different pressures, based on which the MMPs are determined from the diminishing interface method. Second, by means of the factorial-analysis approach and calculated IFTs and minimum miscibility pressures (MMPs), the following five factors are specifically studied to evaluate their main and interactive effects on the IFTs and MMPs: temperature, initial oil and gas compositions, feed gas to oil ratio (feed GOR), and pore radius. It is found that the main and interactive effects of the five factors on the IFTs are inconsistent at different pressures. The effects of the five factors on the MMPs are evaluated quantitatively, which contribute to screen out significant factors, analyze interactions, and identify schemes for the miscible CO2 enhanced oil recovery. The most positive significant main and interactive effects on the MMPs are Factors C (gas composition) and AB (temperature and oil composition), whereas the most negative results are Factors E (pore radius) and AC (temperature and gas compositions). A three-factor analysis indicates that the MMP is substantially reduced in small pores by controlling the percentage of the CH4-dominated gas in the impure CO2 sample and lowering the feed GOR.

Introduction

In the past half century, CO2 injection has been extensively applied to enhance oil recovery, especially light and medium oils, due to its unique features, such as miscible displacement and interfacial tension (IFT) reduction [1]. The IFT is a property of interface, which exists when two phases are present and depends on the mass transfer interactions between the phases [2], [3]. The IFTs between the oil and gas phases can be measured at the reservoir temperature and different pressures by applying, for instance, the axisymmetric drop shape analysis (ADSA) technique for the pendant drop case [4]. On the other hand, the IFTs can be calculated by using, for example, the parachor model [5]. A zero-IFT condition is always considered to be an important characterization of the miscibility, which is a distinct oil–gas state that can be developed with the continuing mass transfer and its induced similar physicochemical properties [6], [7]. Hence, the definition of the minimum miscibility pressure (MMP) is the lowest operating pressure at which the oil and gas phases can become miscible in any portions through a dynamic multi-contact miscibility process [8], [9]. A comprehensive analysis of factors affecting the IFTs and MMPs of the light oil–CO2 systems are required to ensure a successful miscible CO2 flooding project with high oil productions.

The IFT and MMP of the light oil–CO2 system strongly depend on the pressure and temperature, as well as initial overall fluid composition [10]. In general, the oil–CO2 IFT is increased with temperature if pressure keeps constant, whereas it is decreased with pressure at a constant temperature [11]. On the other hand, the temperature increase always results in a higher MMP [12]. Additionally, the initial overall fluid composition affects the IFT and MMP to a large extent. In the literature, the CH4-dominated live oil–CO2 IFTs/MMPs were higher, while the IFTs/MMPs of the intermediate HCs pre-saturated case were lower when compared with those of the dead oil–CO2 system. Also, similar results were obtained from the effect of injection gas composition studies [13]. The feed gas to oil ratio (feed GOR) effect on the IFTs/MMPs cannot be ignored even though no general consensus on it. In a recent study, the measured IFTs/MMPs were found to reach a minimum value at the feed GOR of 1:1 and slightly increase with increasing the injection gas concentration [4]. A recent study was conducted with different temperatures, initial oil and gas compositions, and feed GORs and their effects on the IFTs and MMPs of different light oil–gas systems in nanopores were specifically investigated, whose results agree well with the above-mentioned conclusions [14], [15]. However, no studies so far has been found to evaluate the pore radius effect on the IFTs/MMPs. Moreover, all existing studies only analyze the effects of main factor, while the interactive effects on the IFTs/MMPs have never been studied.

Factorial analysis is an effective method to evaluate different-level interactions of multiple factors and their influences on the system performance [16], through which the significant effects can be accurately determined. This method has been widely applied in various academic research and practical applications. For example, a factorial-analysis based stochastic modeling system (FSMS) was proposed to study the uncertainties associated with hydrocarbon-contamination transport impacts in subsurface [17]. In 2013, a factorial two-stage stochastic programming (FTSP) approach was developed to manage the water resource uncertainty [18]. Later, sequential factorial analyses were used to understand the impacts of uncertainties in the management of air quality [19]. Recently, a two-stage factorial optimization model was proposed for the environmental system [20]. However, the factorial analysis has never been used to evaluate the main and interactive effects of several typical IFT/MMP influential factors (e.g., temperature, initial overall compositions, and pore radius).

In this work, on a basis of five important factors, i.e., temperature, initial oil and gas compositions, feed GOR, and pore radius, a total of 32 different test conditions are designed through the factorial-analysis approach. A modified Peng–Robinson equation of state (PR-EOS) coupled with the parachor model is applied to calculate the vapour–liquid equilibrium and IFTs at different pressures and a variation of pore radii, based on which the MMPs of the dead and live light oil–pure and impure CO2 systems are determined from the diminishing interface method (DIM). The main and interactive effects of the five factors on the IFTs and MMPs are specifically studied through the factorial-analysis approach based on the calculated IFTs and MMPs at different conditions.

Section snippets

Experimental

A dead light oil sample from the Pembina oilfield in Alberta was applied in this study. The detailed statements with respect to the oil physiochemical properties and compositional results were recorded in the previous study [9]. A live oil sample with the GOR of 15:1 m3/m3 was obtained by adding the produced hydrocarbon gas, which composes 66.50 mol% CH4 + 11.41 mol% C2H6 + 11.39 mol% C3H8 + 10.70 mol% n-C4H10, into the dead oil sample. In addition, pure CO2 and a mixture consisting of

Factorial design

The effects of the temperature, oil and gas composition, feed GOR, and pore radius on the IFTs and MMPs are complicated. In particular, some interactive effects, which are different from the effect of single factor, can be significant but difficult to be predicted. Hence, it is necessary to understand the complex interrelationships of a variety of factors and their effects on the IFTs and MMPs in the conventional and unconventional reservoir production processes.

In general, 2k factorial design

IFT analysis

In Table 4, the sums of squares, standardized effects, and contributions for factors and interactions in terms of the IFTs at P = 1 MPa are calculated and listed. It is found from the table that Factor C owns the most positive effect, 1.201, and contributes 6.930% to the overall effect, which is followed by Factor D with the positive effect of 0.045 and the contribution of 0.010%. The positive effect means the IFT is increased when the factor is changed from low to high level. More

Conclusions

The following six major conclusions can be drawn from this work:

  • A factorial analysis approach is applied to characterize the potential single and interactive factors as well as their effects on the interfacial tensions (IFTs) and minimum miscibility pressures (MMPs) of different light oil–CO2 systems under different conditions.

  • The effects table, half-normal plot, and Pareto chart are used to screen the significance and positivity/negativity of the main and interactive effects on the IFTs. For

Acknowledgements

The authors would like to acknowledge the Petroleum Systems Engineering at the University of Regina.

References (29)

Cited by (5)

  • Revealing the impact of an energy–water–carbon nexus–based joint tax management policy on the environ-economic system

    2023, Applied Energy
    Citation Excerpt :

    Thus, the two-way interactions are investigated and high-order interactions are arranged as error items [44]. Considering the model output to be deterministic, a single-replication algorithm is adopted [43]. In this study, the carbon tax on coal and wastewater tax on four industrial sectors are regarded as five factors.

  • An integrated Bayesian least-squares-support-vector-machine factorial-analysis (B-LSVM-FA) method for inferring inflow from the Amu Darya to the Aral Sea under ensemble prediction

    2021, Journal of Hydrology
    Citation Excerpt :

    Besides, the interactive effects of input factors should not be neglected because they could significantly influence system performance (Jiang et al., 2017). Factorial analysis (FA), viewed as a multivariate inference method, can efficiently reveal the main effect of a single factor and different-level interactions of multiple input factors on the system performance (Zhang et al., 2019). FA has been widely used in various experimental researches and practical applications (Bourgeois et al., 2015; Saleh et al., 2018; Lin et al., 2019).

View full text