Practical Advances in Tight Gas Sands Core-Based Shaley Sands Water Saturation Analysis

John Dacy
Core Laboratories, Houston, TX

One of the core-based approaches to reservoir water saturation in shaley sands is the Waxman-Smits-Thomas (WST) model where water saturation is dependent on: porosity, true resistivity (Rt), formation brine resistivity (Rw), counterion conductance (B), intrinsic shaliness (effective Qv, Qve), intrinsic Archie parameters (m*, n*), and water saturation. New methods and equations are presented for Rw, B, and Qve. These methods are particularly well-suited for tight gas sands and encourage use of the WST model. These advances facilitate a practical assessment of the water saturation dependency of excess conductivity in the resistivity index (I) versus water saturation relationship at reservoir conditions over anticipated saturation ranges. This can lead to reliable application of reservoir conditions m and n in a normal Archie equation for Sw from Rt that takes excess conductivity into account. For some situations, this application may be preferred to use of the full WST model.

High quality produced formation water samples can provide Rw data through direct measurement. Such samples are rare in tight gas sands as many of these formations produce no water. Further, when water samples are available, often the source (interval) is uncertain and/or the samples tend to be contaminated by dilution with an uncertain quantity of condensation water. Cores cut under conditions of minimal or quantifiable disturbance to reservoir Sw provide a source for improved Rw data. This is sometimes accomplished by mechanical removal of whole brine and direct Rw measurement although this approach is seldom feasible for tight gas sands. Rather, total core water is removed by distillation and quantified and the resulting precipitated salts are removed by crushing the core sample and dissolving the salts in a known volume of distilled water. Various chemical methods can be used to determine the salt quantity which can then be related to the core water volume. The key consideration for Rw in shaley sand cores is recognition that the core total water volume includes a fresh water component associated with clay-bound water and a component associated with the formation brine. The salts should be recombined to the formation brine water component. The new method includes a determination of the clay-bound water component in the fresh core samples by NMR T2 experiments using brine-saturated companions. When this information is integrated with the traditional core water salinity data the computed formation brine concentrations are more representative.

The WST term B when combined with Rw contains all the temperature and salinity dependence of shaley sand conductivity. The B term is of equal importance to Qve in relating Sw to resistivity. An empirical B equation for the lab environment was provided by WST. For the reservoir environment Waxman and Thomas provided a cross plot displaying B isotherms as functions of Rw. Subsequent authors have published empirical equations for B in the reservoir environment. These equations compare poorly to parts of the Waxman-Thomas cross plot and do not provide a room temperature (25 C) isotherm consistent with the lab environment equation. An improved and unifying empirical equation is presented that compares more favorably with the Waxman-Thomas reservoir environment cross plot and perfectly with the lab environment equation.

Effective Qv is the intrinsic rock shaliness term in the WST equation. Traditional methods for Qv are limited in tight gas sands. The wet chemistry cation exchange capacity (CEC) approach to Qv frequently over or under predicts the intrinsic property. The multiple-salinity core conductivity test (Co-Cw) is the preferred traditional method for effective Qv. However, as it is a flow-through technique, test times in tight gas sand samples tend to be lengthy and impractical. A new technique for effective Qv provides results consistent with the Co-Cw method. This technique has neither the limitations nor restrictions of the traditional methods. The new method is based on core measurement of stressed total porosity and results of a fully brine-saturated NMR T2 experiment. This test is rapid, accurate, repeatable, non-destructive, cost-effective, and ideal for tight gas sands.

Application of core-derived Archie m and n values is generally straight forward and understood by most log analysts. Application of the terms B, Qv, m*, and n* in the WST model is often poorly understood. The data analysis and model process are open to many pitfalls and misapplications. A simplified approach is sought that will facilitate application of the core test results and include the correct effects of excess conductivity. Determining the intrinsic rock properties and using the equations described above permits rapid transformation of the lab results, F vs porosity and I vs Sw, to reservoir equivalent results that include the temperature and salinity dependence of formation Rw and B. These results will display the water saturation dependence of excess conductivity in the I vs Sw relationship. Limited scatter in the transformed data permits selection of saturation ranges believed to represent the reservoir and simple n values for the reservoir conditions case. The resulting reservoir conditions m and n values can be used in a standard Archie equation and will include the correct excess conductivity effects. On the other hand, the transformed data may dictate use of a shaley sand model to account for the observed variation. Advantages and limitations of this simplified approach are discussed.