Quantification of Phase Behaviour and Physical Properties of Solvents-Heavy Oil/Bitumen-Water Systems at High Pressures and Elevated Temperatures

Due to the excess heat loss of steam assisted gravity drainage (SAGD) processes and low oil production rate of solvent-based processes, the expanding solvent SAGD (ES-SAGD) process has been considered as a promising technique for enhancing heavy oil/bitumen recovery. The main ES-SAGD mechanisms include the heat transferred and dissolution of solvents into the heavy oil/bitumen to swell it and reduce its viscosity, which is closely related to the phase behaviour of solvents-heavy oil/bitumen-water systems. Thus, it is of fundamental and practical importance to accurately quantify the phase behaviour and physical properties of the aforementioned systems. A pragmatic technique has been developed to optimize the reduced temperature for acentric factor for the Peng-Robinson equation of state (PR-EOS) and Soave-Redlich- Kwong equation of state (SRK-EOS) by minimizing the deviation between the measured and calculated vapour pressures. The reduced temperature has its optimum value of 0.59 for the two EOSs, while 0.60 is recommended for practical use. The mutual solubility for n-alkanes/n-alkylbenzenes-water pairs is correlated using the PR-EOS together with the two newly modified alpha functions. The binary interaction parameters (BIPs) for both aqueous phase and liquid hydrocarbon phase are generalized as functions of reduced temperatures and carbon numbers of hydrocarbons, reproducing the experimental measurements well. Then, the modified PR-EOS model is successfully applied to predict the multi-phase compositions and three-phase upper critical ending points (UCEPs) for n-alkane-CO2-water mixtures. A new correlation has been developed to calculate the redefined acentric factor for pseudocomponents (PCs), while new BIP correlations are proposed respectively for ii toluene-water pair and heavy oil/bitumen-water pairs. The BIP correlation for heavy oil/bitumen-water pairs is validated by the measured water solubility in other oils. The newly developed model is found to accurately predict the measured ALV/AL (A is the aqueous phase, L represents the oleic phase, and V denotes the vapour phase) and LV/L boundaries with an overall average absolute relative deviation (AARD) of 4.5% and solvent solubility in the oleic phase with an overall AARD of 9.4%, respectively. Two new methods have been proposed to predict the density/swelling factor for solvents-heavy oil/bitumen/water mixtures, i.e., one is a new volume translation (VT) strategy for PR-EOS, while the other is the ideal mixing rule with effective density (IME) calculated using a newly developed tangent-line method. It is found that both of these two methods are accurate enough, while the IM-E is better than the VT PR-EOS. Experiments for C3H8/CO2-Lloydminster heavy oil/water systems have been performed in a temperature range of 328.7-432.3 K. A dynamic volume analysis method is proposed to simultaneously simulate the total volume and height of vapour/oleic phase interface, while a new framework incorporated with the modified PR-EOS can be used to accurately predict the solvent solubility, phase boundary, and phase density for the aforementioned systems. Also, six widely used mixing rules have been respectively evaluated, while water is incorporated using the ideal mixing rule. The order of the best ones in their accuracy is the volume-based power law > the weight-based power law > the weight-based Cragoe’s mixing rule. The effective density rather than real density of dissolved gas should be used for all the volume-based mixing rules.

Name thesis:degree_name

Doctor of Philosophy (PhD)

Level thesis:degree_level

Doctoral -- first

Discipline thesis:degree_discipline

Engineering - Petroleum Systems

Grantor dc:publisher

Faculty of Graduate Studies and Research, University of Regina

Year dc:date.issued

2019

Author dc:creator

• Chen, Zehua

Advisor dc:contributor.advisor

• Yang, Daoyong

Committee members dc:contributor.committeemember

• Gu, Yongan
• Henni, Amr
• Salad Hersi, Osman

Language dc:language.iso

en

DOI

https://doi.org/10.82465/4045

OAI identifier oai:identifier

oai:uregina.scholaris.ca:10294/9016