Recovery-Impacting Petrophysical Alterations of CO2-EOR Application in Carbonate Formations: Case Study on North Dakota's Upper Red River

Abstract

Abstract Carbonate-based mineral dissolution and precipitation, driven by carbon dioxide (CO2) injection, introduces complexities into carbonate reservoir systems, triggering interactions that are different than those seen in traditional CO2 enhanced oil recovery (CO2-EOR) applications in siliciclastic/sandstone reservoirs. The thrust of this paper is to couple experimental (laboratory-scale) and numerical (computationally-assisted) analyses in order to assess how CO2-induced petrophysical alterations impact the resultant hydrocarbon recovery from CO2-EOR application in carbonate reservoirs. The Upper Red River Formation, located in North Dakota's Cedar Creek Anticline (CCA) Field presents significant remaining oil-in-place, albeit a high-water saturation from waterflood operations undergoing since the 1960s. The residual (post-waterflooding), oil saturation, makes the Red River Formation a good target for modern-day CO2-EOR technology. The first part of this study involves a core-scale investigation of dynamic-permeability variations triggered by CO2 injection into three primary-productive zones, designated as "Red River Units" (RRU2, RRU4, and RRU6). The second part involves a compositional reservoir model built in SLB's Petrel software to perform numerical simulations of CO2 injection incorporating pre-established permeability variations that honor our laboratory-obtained results. Correlations between differential-pressure variations observed during carbonated-brine (CO2/brine) injection were assessed against pore volumes injected (PVI). These pressure fluctuations were induced by dynamic-permeability variations resulting from carbonate-based mineral dissolutions/precipitations. Baseline-permeability variations were established a-priori using nitrogenated-brine (N2/brine) injection in order to correct for physico-chemical effects from the brine. During CO2/brine injection, the recorded permeability increased significantly compared to its original value, peaking before gradually recovering. The history-matched compositional reservoir model was used to project the production from CO2-EOR through a section incorporating four existing wells, integrating laboratory-derived dynamic-permeability variations into the model, yielding different results due to incorporating laboratory-derived dynamic-permeability variations compared to simulations performed at constant permeability. Reduced reservoir permeability correlated with decreased oil recovery, emphasizing the significant impact of dynamic-permeability variations on CO2-EOR performance.