Estimation Model of Permeability for Tight Sandstone Based on NMR T2 Centralized Distribution Method

Abstract

Some applications showed that the Coates model and SDR model acquired unsatisfied permeability results for tight rocks with low or ultra-low permeable properties. This study introduced a novel model with centralized distribution of pore space based on nuclear magnetic resonance T₂ data, which was used to estimate tight sandstone reservoir of Yingcheng and Shahezi Formation in Ordos Basin in China. The petrographic properties of tight sandstone were firstly analyzed based on thin sections and SEM observations. The sandstone generally experienced intensive compaction and cementation, leaving primary pores of small size with partially cement- filled pores. The interparticle pores within clastic debris and mineral cleavage seams were also formed. The secondary pores comprised moldic pores and dissolved pores within feldspar grains. The medium sandstone and fine sandstone characterized by secondary pores and cracks, causing connecting pore and throat of small size, but the pore sizes were relatively uniform, which means the concentration of different pore types. The real data were investigated based on 39 core samples with porosities ranging from 5.33 to 14.98 percent, and permeabilities from 0.01 to 9.97 Micro-Darcy. The results showed that the T2 centralized distribution constant (C) in the model reflects pore concentration, which means the lower value of C, the more concentration of pores could been identified by T₂ spectra. The similar size of pores and channels were likely to formed in medium or fine sandstone, which resulted in relatively good flow capacity depending on small size of connected pores and channels. For the fine sandstone with ultra-low porosity and permeability, the random distribution of pore space caused T₂ spectra with scattered features and lower value of C, which reflected the association of interparticle pore, moldic pore and micro-fractures with different pore size. The novel mathematical model referring pore space concentration succeeded in permeability estimation, with coincidence rate of above 80%. This study provided an effective method of permeability estimation for tight sandstone reservoirs.

AAPG Datapages/Search and Discovery Article #90350 © 2019 AAPG Annual Convention and Exhibition, San Antonio, Texas, May 19-22, 2019