Modelling Irreducible Water Saturation in Heterogeneous Rocks From Pore-Size Distribution

Abstract

Irreducible water saturation is a key property for the estimation of original oil and gas-in-place. It is also key to end-point scaling of capillary pressure and relative permeability, with significant impact on simulation results of reservoirs under improved/enhanced oil recovery (IOR/EOR). Several definitions of irreducible water saturation exist, based on different experimental measurements and standard estimation methods. We propose a comprehensive model and a new method for improved estimation of irreducible water saturation. The model considers rock wettability; the thin film of water that coats portions of the rock grains; the pore size distribution; the tortuosity; and the ratio between pore-throat and pore-body sizes (BTR). Different components of the irreducible water saturation are identified for multimodal, heterogeneous rocks: a nano-porosity system completely filled with water and other pore systems with their walls coated by water. The model also considers an additional residual water saturation resulting from laboratory experimental limits as the maximum applied pressure and duration. The method adjusts the model parameters by fitting to a set of irreducible saturation data, obtained from both mercury injection (MICP) and air-brine drainage capillary pressure experiments. The method estimates the irreducible water saturation for the asymptotic ideal condition - very high capillary pressure and reservoir geological times – as well as for other laboratory and reservoir conditions. We applied the proposed method to experimental data from Corelab's worldwide rock catalog. The fraction of nano-porosity not revealed by MICP experiment was estimated by comparing MICP porosity with routine effective porosity. Hydraulic tortuosity and truncated multi-Gaussian decomposition of pore-throat-size distribution were also obtained from MICP data. BTR range was estimated from NMR data, thin sections, and hydraulic tortuosity data. Water thin film thickness range was estimated from the literature. Model parameters were then successfully estimated using data from 49 carbonate and 106 clastic samples from all over the world. The results showed that, in several cases, the asymptotic irreducible water saturation might be significantly smaller than the observed value from the air-brine experiment. Therefore, the corresponding reservoir irreducible water saturation could also be overestimated. The relative importance of the different components of the irreducible water saturation varied from one sample to the other, confirming the relevance and completeness of the proposed method. When compared to traditional methods, the proposed method significantly improves irreducible water saturation estimates, resulting in better saturation-height and end-point scaling functions, and more accurate reserves. It is particularly important for simulation of IOR/EOR processes. The method may also be integrated with dielectric and NMR well log measurements, increasing the resolution of dynamic reservoir characterization, with particular importance to mixed-wet rock environments.

The Brooks and Corey Capillary Pressure Model Revisited from Pore Network Simulations of Capillarity-Controlled Invasion Percolation Process

“…Such observations can be supported by the percolation theory (see, e.g., [35,36]). The residual saturation reported in the current work might be questioned for higher values than those reported by Sok et al [33] and Ferreira et al [37]. This discrepancy can partly be due to the fact that the present work does not account for the film effect.…”

“…This discrepancy can partly be due to the fact that the present work does not account for the film effect. In fact, liquid films play a crucial role in the wetting phase residual saturation [37], because they reduce the resistance for liquid transport and thereby enhance the liquid hydraulic connectivity [38]. Another reason might be that the spread of size distributions in the present network simulations is smaller than in real porous media.…”