Rock Fabric is a Better Predictor of Well Performance than TOC in the Marcellus Shale

Madren, Jonathan; Walker, Kristin

For the Marcellus shale (Devonian) in central Pennsylvania, a direct comparison between areas of higher productivity and lower productivity is very instructive. Counter to work done by others, the Total Organic Carbon (TOC) does not correlate to better well performance, implying that the nanometer scale porosity seen in some analyses are not always effective porosity. Instead, well performance is better predicted by a more thorough understanding of the rock fabric, in particular the nature of the grains making up different intervals in the reservoir. Analytical techniques such as QEMSCAN are useful for characterizing both the mineralogy of the rock and the sizes of the pore systems associated with them in a two dimensional sense. Importantly, this method uses a field of observation much larger than other techniques such as Ar-milled electron microscopy, allowing for a more meaningful understanding of rock fabric and larger scale pores (>5 um), which are more likely to be indicative of an effective porosity system. As a general trend, wells with more ductile clay grains tend to have smaller pores and poorer performance relative to wells with more abundant quartz grains, which are more competent and have larger scale pores. This model predicts that wells drilled proximal to the sediment supply, in this case to the northeast, should perform better than wells drilled more distally, which is what is observed. Critically, it means that there may be a negative relationship between the amount of clay seen in an area and well performance. This work would indicate it is not due to issues with completions, such as not being able to fracture the rock, but instead is dependent on the porosity systems seen within the rock. The comparison made here indicates that understanding the facies within an unconventional play is critical to being able to tie log response to reservoir properties, and allows for the creation of geologic models which can predict well performance.

AAPG Search and Discovery Article #90163©2013AAPG 2013 Annual Convention and Exhibition, Pittsburgh, Pennsylvania, May 19-22, 2013