(187g) Quantitative Analysis of the Influence of Capillary Pressure on Geologic Carbon Storage Forecasts. Case Study: CO2-EOR in the Anadarko Basin, Texas

Numerical models are critical for forecasting subsurface multi-phase flow associated with geologic carbon storage. Uncertainty of model results stems from many factors, including and especially uncertainty in multi-phase flow parameters. Specifically, relative permeability and capillary pressure relationships depend on both the rock properties and fluid properties, and the latter may be highly nonlinear as fluid temperature and pressure conditions change. Forecasts of trapping mechanisms, phase behavior, and plume movement are impacted by the choice of relative permeability and capillary pressure functions and how those functions are calibrated and constrained. In particular, one of the most neglected aspects of such model simulations is meaningful capillary pressure processes, most likely due to lack of measured data. A primary goal of this study is quantify the difference in forecasts or results for models that utilize capillary pressure functions calibrated with measured data from results of models without such. Additionally, the relative permeability models developed here were derived from measured capillary pressure data. Those data were used to constrain saturation end points in the relative permeability curves and dictate how that relative permeability was distributed spatially.

The case study area for the research presented is the Farnsworth Unit (FWU), an active CO₂ – enhanced oil recovery field in the Anadarko Basin of northern Texas. Correlation among porosity and permeability data within the FWU suggest a series of definitive hydrostratigraphic units, zones of the reservoir that exhibit similar flow properties. In a separate study, capillary pressure was measured by mercury intrusion testing of 46 core samples from two different wells in the FWU. Thirteen of those 46 samples were correlated to specific hydrostratigraphic units within the principal reservoir, the Morrow ‘B’ Sandstone. Five of the 13 samples were chosen for this study to represent each of the five hydrostratigraphic units identified in the reservoir. The relative permeability curve assigned to each of these flow units included saturation end-points determined from the capillary pressure data for that particular flow unit, in effect a heterogeneous parameter assignment. Results of models parameterized with this approach were compared to models that utilize a more conventional approach of assigning a single relative permeability and capillary pressure relationship to a reservoir, typically based on lithology type or geologic formation. Also compared were results of models parameterized with other synthetic relative permeability relationships paired to measured capillary pressure relationships and simulations run without capillary pressure.

The main conclusions drawn from this analysis include (1) heterogeneity in relative permeability plays a major role in simulated forecasts of CO₂ migration, trapping mechanisms and storage capacity, as well as oil and water production, and (2) capillary pressure, and in particular the magnitude of capillarity, also plays a major role in these processes; conversely, if the magnitude of capillarity is small relative to in situ fluid pressures, it imparts insignificant effects on these processes.