A comprehensive study of nano-scale grain boundary channels in fracture cements using scanning electron microscopy, electron backscatter diffraction, and transmission electron microscopy

Natural fractures in shale reservoirs are frequently filled with mineral cement that lack any residual fracture porosity visible under the petrographic microscope and are generally interpreted to be impermeable to fluid flow. Scanning electron microscopy of calcite, dolomite, quartz, and barite fracture cements from a variety of shale and low-permeability sandstone formations, prepared using broad ion beam milling, provides evidence of open to partially healed elongate pores that are on the order of hundreds of nanometers in aperture. In calcite, dolomite, quartz, and barite fracture cements, these pores have apertures of about 10-400 nm. In quartz fracture cements that have experienced low grade metamorphic temperatures in excess of 250°C, they are up to 600 nm. These pores have been previously overlooked because traditional thin sectioning and polishing destroys sub-micron details of the fracture cement pore structure. Ion milling preserves these details. Electron backscatter diffraction shows that these pores occur along most grain boundaries within the blocky or columnar fracture cement. Mineral cement crystal faces are rarely faceted on the nano-to-micrometer scale, but contain varying degrees of roughness with distinct morphologies. Pores tend to increase in aperture with increasing maximum formation temperature, indicating dissolution-precipitation kinetics strongly influences grain boundary structure. Stress relaxation from formation exhumation may also favor wider channel apertures, evidenced by a trend of wider apertures with increased distance of exhumation. HRTEM analysis shows different crystallographic domains with possible amorphous regions bridging across grain boundary channels, demonstrating that these channels are complex structures, contrasting the conventional view of diagenetic cement grain boundaries as simple crystallographic dislocations or discontinuities. I propose a model of dissolution-precipitation along grain boundaries that preserves NGBC ubiquitously in carbonate and quartz fracture cements that have experienced diagenetic and low-grade metamorphic conditions. While partially healed, these pores are frequently well connected and have strong implications for understanding flow through matrix cements of low permeability reservoirs. They may act as conduits for fluid flow along and across fully cemented natural fractures. When their effects are considered, they may universally increase estimates of formation permeability of low permeability formations containing cemented fractures.