No abstract

No abstract

TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractAt three different wettabilities, two stacked outcrop core plugs, separated by a 1 mm fracture, were waterflooded. During the experiment in-situ fluid saturations were monitored with Magnetic Resonance Imaging (MRI). The sequence of 2D MRI images corroborated earlier lower resolution, larger scale experimental results on the effect of fractures during waterflooding at various wettabilities. The MRI images of oil saturation development in the fracture clearly revealed two distinct transport mechanisms for the wetting phase, water, across the fracture at several wettability conditions. When strongly-water-wet, the first core reached its spontaneous imbibition endpoint before water left the matrix and entered the fracture. The displaced water flowed down the exit face to the bottom of the fracture and displaced the oil upward at the rate of water injection. At less-water-wet conditions water droplets formed on the exit face of the first plug and grew large enough to form individual bridges between the two plugs. This happened well before the first plug reached its spontaneous imbibition endpoint. Under these conditions, the fracture filled slowly, as the bridges increased in diameter and additional bridges formed. Due to the capillary continuity of the wetting phase, a viscous pressure drop was established across the stacked core plugs, providing a viscous component to the total oil recovery.

TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractIterative comparison between experiments and numerical simulation has been used to predict oil recovery mechanisms in fractured chalk as a function of wettability. Selective alteration of wettability, by aging in crude oil at elevated temperature, produced chalk blocks which were stronglywater-wet and moderately-water-wet, but with similar pore geometry and mineralogy. Larger scale, nuclear-tracer, 2Dimaging experiments monitored fluid distributions while waterflooding blocks of chalk, first whole then fractured. This data provided in-situ fluid saturation development for validating numerical simulation and evaluating capillary pressure-and relative permeability input data used in the simulations. Capillary pressure and relative permeability at each given wettability were experimentally measured and used as input for the simulations. Optimization of either Pc-data or kr-curves gave indications of which of these input data could be more trusted. History matching both the production profile and the in-situ saturation distribution development gave higher confidence in the simulation. Labelling the injection water differently from the in-situ water made it possible to determine the degree of water mixing during the waterfloods. Mixing of injection water and in-situ water during waterfloodng was determined for both unfractured and fractured blocks. Reduced water wettability resulted in less oil recovery by spontaneous imbibition. Interconnected fractures did not significantly impact the final oil production when the permeability increase after fracturing was low, for both strongly-water-wet and moderately-water-wet conditions. However, in-situ saturation distributions were significantly affected by the wettability conditions. For higher permeability increase after fracturing significant reduction in oil recovery was experienced at less water wet conditions, while oil recovery at strongly-water-wet conditions was not reduced, even at high permeability increase after fracturing. The overall best match between the simulations and the experiments was obtained using the experimentally obtained capillary pressure curves and optimizing the experimentally measured relative permeabilities.

Summary This work studies the mixing of injected water and in-situ water during waterfloods and demonstrates that the mixing process is sensitive to the initial water saturation. The results illustrate differences between a waterflooded zone and a preflooded zone during, for example, water-based EOR displacement processes. The mixing of in-situ, or connate, water and injected water during laboratory waterfloods in a strongly water-wet chalk core sample was determined at different initial water saturations. Dynamic 1D fluid-saturation profiles were determined with nuclear-tracer imaging (NTI) during waterfloods, distinguishing between the oil phase, connate water, and injected water. The mixing of connate and injected water during waterfloods, with the presence of an oil phase, resulted in a displacement of all connate water from the core plug. During displacement, connate water banked in front of the injecting water, separating (or partially separating) the injected water from the mobile oil phase. This may impact the ability of chemicals dissolved in the injected water to contact the oil during secondary recovery and EOR processes. The effect of the connate-water-bank separation was sensitive to the initial water saturation (Swi). The time difference between breakthrough of connate water and breakthrough of injected water at the outlet showed a linear correlation to the initial water saturation Swi. The results obtained in chalk confirmed earlier findings in sandpacks (Brown 1957) and thus demonstrated the generality in the results.