Experimental and Simulation Based Interpretation of Characteristic Behavior During Forced and Spontaneous Imbibition in Strongly Water-Wet Sandstones

Abstract

Abstract In this work we investigate forced and spontaneous imbibition of water to displace oil from strongly water-wet Gray Berea (~130 mD) and Bentheimer (~1900 mD) sandstone core plugs. Two nonpolar, nonvolatile oils (n-heptane and Marcol-82) and their mixtures were used as non-wetting phase, giving oil viscosities between 0.4 and 31 cP between experiments. Brine (1 M NaCl) was used as wetting phase with viscosity 1.1 cP. Recovery was measured for both imbibition modes, and pressure drop was also measured during forced imbibition. Forced imbibition (five tests) was conducted with same viscosities at low and high injection rate using two different viscosities. 17 spontaneous imbibition experiments were performed at four different oil viscosities, and on the two rock types, including tests at same conditions. By varying the oil viscosity, injection rate and imbibition modes we measured the system's response to displacing oil by water under different conditions where both capillary and advective forces were allowed to dominate. Our hypothesis is that such a combination of experiments allows us to determine some characteristics of water-wet systems. Transient analytical solutions were derived accounting for low water mobility and inlet end effects, allowing theoretical predictions consistent with the observations. Full numerical simulations were also run to consistently match all the experimental observations. We find that, consistent with the literature, water has low mobility associated with its relative permeability. Thus, complete oil recovery was achieved at water breakthrough during the forced imbibition both at low and high oil viscosity tests. For the same reason, increasing oil viscosity by a factor of almost 100 did not increase the spontaneous imbibition time scale by more than 5 compared to the lowest oil viscosity. This was consistently matched by our models. Theoretical analysis indicates that pressure drop increases linearly with time until water breakthrough if capillary pressure is negligible and that the initial pressure drop correspond to the oil relative permeability end point. Positive capillary forces assist water in entering the core, and the pressure drop is reduced and possibly nonlinear with time. Using a high injection rate we could a linear trend more clear than at low rate, consistent with our predictions.