A progressive decrease in matrix permeability occurs with burial in the Paleogene limestones of west-central Florida, yet grainstones with permeabilities > 400 md persist below burial depths of 300 m. Why some grainstones retain high permeability during passage through the shallow-burial realm was investigated using 85 dominantly foraminifera and micritized-grain grainstones from the Eocene Avon Park Formation. Twenty-seven variables representing diagenesis, stratigraphic packaging, depositional energy, grain types, and pore-network attributes were quantified. Permeabilities range from 23 to 7,160 md and thin-section porosities range from 14% to 36%. Grain size (lower coarse to upper fine sand) covaries positively with sorting (poor to very good). The samples are from three different systems tracts within the Avon Park, and sampled grainstone bodies ranged in thickness from 0.5 to 3.5 m. Evidence of diagenesis include mechanical and chemical burial compaction, pre-compaction cements, and dissolution. Original interparticle porosity loss due to compaction (COPL) and cementation (CEPL) range from 0 to 35.3% and 1.6 to 26.2% respectively. Most original intraparticle porosity is occluded, but most moldic porosity, which ranges from 1.2% to 12%, remains open.
Multivariate regression reveals that 61% of the variance in permeability is attributed to grain size, CEPL, COPL, and systems tract (a proxy for relative abundance of compactable foraminifera versus robust micritized grains). Large grain size means a large initial permeability; all other regression components reflect diagenetic reductions in permeability. Observed ranges in CEPL and COPL values mean they can reduce permeability by more than three orders of magnitude whereas the other regression parameters reduce it by only one order of magnitude. Retention of high (> 400 md) permeability can occur by (1) cementation of ~ 40% of the original interparticle pore space, which inhibits subsequent compaction yet preserves high permeability, or (2) a combination of cementation and compaction that destroys less than two thirds of the original interparticle pore space. These relations, derived from originally calcite-rich sediments, provide a quantitative reference by which to evaluate permeability preservation during shallow burial in numerical simulation of diagenesis and reservoir development.
Capillary-pressure data also support the hypothesis that pore geometries, and hence permeability, respond differently to compaction versus cementation. Cement-dominated samples exhibit a dominance of porosity behind intermediate-sized pores at higher permeabilities than compaction-dominated samples. Compaction also does not yield a large change in the amount of porosity behind pore throats smaller than 1 μ at any permeability above 20 md, whereas cementation-dominated diagenesis generates significant amounts of pore space behind those small throats once permeability goes below 100 md. These differences arise because cementation converts the original interparticle pores to a system of interparticle polyhedral and sheet-like pores connected by progressively smaller pore throats between cement crystals, whereas compaction merely constricts the original interparticle porosity behind smaller intergranular pore throats.