ROCK PHYSICS AND MINERALOGY RELATIONS FROM WELL LOGGING IN THE BARNETT SHALE

Abstract:

In this work, we build volumetric models from well logs to relate rock properties like porosity and water saturation to rock constituents. We build and compare volumetric models from logs in two wells in the Barnett shale to those from lab measurements on core samples. Well 1 (location 1) has both a deviated and a horizontal part within the Lower Barnett while logs of Well 2 (location 2, 3 miles of location 1) are from its vertical/deviated part.

 Well 1 (deviated): Well 1 is a horizontal well logged in the producing Lower Barnett. Parts of the well less than 90 degrees in inclination are considered deviated. Effects of Calcite, Montmorillonite (Mont.), Illite and Quartz fractional volumes in addition to TOC and Kerogen weight fractions were investigated. Kerogen and TOC correlate negatively with Quartz (not co-locational in the model) and positively with porosity (Fig. 1). Water saturation correlates positively with Mont. and negatively with Kerogen and TOC.

Well 1 (horizontal): Kerogen or TOC and Quartz anti-correlate. Porosity seems to decrease with Quartz but increase with Kerogen or TOC (Fig. 2). Water saturation decreases with Kerogen and increases with Mont.

Well 2: For this well, there are no water saturation, Kerogen, TOC or resistivity measurements. Porosity here correlates positively with clay and negatively with Calcite. We used Schmoker’s method to calculate TOC from density logs. TOC correlates positively with porosity and clay content (Fig. 3 and 4).

Discussion: Organic matter seems to be a source of porosity in both locations in accordance with previous literature (Loucks et al., 2009; Curtis et al., 2010). Pores in location 1 are not water-filled, which we deduced from the negative correlation with water saturation. A model of rock from location 1 is in Fig. 5. In location 2, the calculated TOC correlates positively with porosity, as in the volumetric model in Fig. 6. Key positive correlations are clay-porosity and clay-TOC/Kerogen, in agreement with the findings of Sone and Zoback (2013a,b). An interesting result is the comparability between models created from well logs and models from direct lab measurements on rock samples. This provides confidence in using well logs for understanding generalized rock composition and linking macrostructure and microstructure.