Formation grain size distribution is important in reservoir sand management, perforation, optimizing fracturing strategy, and predicting depositional and diagenetic facies and hydraulic units. This paper describes a new method of using acoustic velocities and NMR relaxation time spectrum to determine parameters of formation grain size distribution.
Grain size distribution is often approximated by normal, log-normal, binary, or more general types of distribution, e.g., Weibull distribution. In most cases, however, it can be characterized by two numbers: mean grain size and sorting, which describe, respectively, the measures of the center and the spread of a grain size population. The influence of sorting on acoustic properties of porous media is well known; however, there is no direct relationship between acoustic properties and mean grain size. NMR T1 and T2 relaxation time spectra of wetting phase are directly related to the distribution of pore volume-to-surface ratio, which, in turn, depends on the mean grain size. On the other hand, there is no straightforward correlation between NMR relaxation time and sorting. Moreover, neither NMR nor acoustic data alone can resolve effects due to heterogeneity in mineralogy and heterogeneity in pore geometry (sorting, compaction, cementation, etc).
The approach presented in this paper is based on the predictions of NMR T2 spectrum and acoustic velocities in numerical model rocks having different grain size distribution. Pore geometry of these model rocks is defined by geologic and mineralogical information about the formation. This information is supplied by a mineral reading from a logging tool, from drilling cuttings, or from prior knowledge based on core or logging data of a similar field. Based on the results of these predictions, we show that NMR relaxation time spectrum allows estimating mean grain size, and that compressional and shear acoustic velocities provide means to compute sorting parameter. The results are compared with the sandstone core measurements data and are found to be consistent with them.
Determination of grain size distribution is important to sand management of unconsolidated formations [Oyeneyin et al., 2005; Nouri et al., 2006], for predicting depositional and diagenetic facies and hydraulic units [Altunbay et al., 1994], and for estimating penetration depth in perforation [Brooks et al., 1998]. Traditional methods of measuring grain size distribution include sieve analysis or laser particle size analysis for unconsolidated sediments or crushed rocks, and image analysis of thin sections for consolidated rocks. However, all these methods are laboratory analysis methods and require core samples, which are not readily available and may not be representative of the whole intervals of interest. Therefore, there is a need for a reliable technique to calculate grain size distribution of formation rocks from the downhole formation evaluation data.
The idea of using NMR T1 or T2 relaxation time distribution to determine grain size is well known in the literature. Recently, we have developed an approach of utilizing NMR T2 spectrum to calculate grain size distribution based on pore scale modeling [Chen et al., 2007].