Reservoir characterization, porosity, and recovery efficiency of deeply-buried paleozoic carbonates: Examples from Oklahoma, Texas and New Mexico

Abstract

Capillary-pressure data from the Early Ordovician Ellenburger Dolomite (west Texas and New Mexico) and the Late Ordovician-Early Devonian Hunton Group carbonates (Oklahoma) are used to calculate or infer petrophysical characteristics, such as median pore-throat size, pore-throat size distribution, effective porosity, and recovery efficiency (RE). For both data sets, porosity and RE are inversely related. A positive relationship between RE and porosity has been reported by other workers, but the relative importance of these opposed trends is unknown. The ability to accurately predict which relationship will hold in a given reservoir unit would be of great value for predicting reservoir performance.

RE is also inversely related to median throat size. This is a consequence of two controlling factors: rock characteristics and experimental procedure. Drainage (recovery) from small throats is more efficient than from large throats, and hysteresis limits recovery from large throats because throats filled at very low pressures early in the intrusion process remain filled at comparable pressures upon extrusion. The experimental procedure also suppresses extrusion from large throats because the minimum pressure attained on extrusion is greater than the initial intrusion pressure, due to limitations of the apparatus.

Reservoir rocks are classified in terms of their capillary-pressure curve form, because curve form is controlled by a variety of petrophysical factors which can be measured, and because curve form is strongly correlated with recovery efficiency. Steep-convex capillary-pressure curves correspond to samples with high REs, low porosities, small median throats, and high entry pressures. Steep-concave curves correlate with low REs, high porosities, large median throats, and low entry pressures. Gently-sloping curves correspond to samples with moderate REs, intermediate median throat sizes, poorly-defined entry pressures, platykurtic throat-size distributions, and variable porosities. Polymodal curves result from polymodal throat-size distributions, and exhibit variable REs and porosities.

Steep-concave and steep-convex curves are interpreted in two quite different ways, as follows. Steep-convex curves indicate reservoir rocks with high REs, but low porosities and small throats, so that production is likely to be economical only under high pressures (or thick oil columns) or from very large hydrocarbon pools. Conversely, steep-concave curves indicate porous reservoir rocks with large throats but probably poor primary recovery efficiency. These reservoirs will be economical even at low pressures and with short oil columns and small total reserves, but will probably need enhanced recovery to produce a significant proportion of reserves. This classification may allow characterization of an “ideal” capillary-pressure curve, which is characterized by a moderate entry pressure, intermediate median throat size, good RE, moderate porosity, and leptokurtic throat-size distribution.

Capillary-pressure plots showing both cumulative and incremental mercury intrusion are more useful than the traditional graphs which show only cumulative intrusion. The incremental intrusion histograms used here highlight modal throat sizes (or modal capillary pressures) which are not well-displayed on cumulative plots. The existence of multiple modes is significant because it affects the relationship between Rwa (water saturation) and oil recovery efficiency, as well as overall nonwetting-phase recovery efficiency.