Abstract
This paper discusses the use of a novel methodology for the petrophysical characterization of carbonate reservoirs. The methodology makes use of nuclear magnetic resonance (NMR) log and core data and provides direct and fast estimates of downhole porosity, pore size, petrophysical rock types, irreducible water saturation and permeability. The novelties of the approach lie on the developed permeability model and the critical role played by effective surface relaxivity. It is shown that if this parameter is properly defined, NMR logs do not need any additional calibration.
The methodology is demonstrated for a Cretaceous carbonate reservoir in the Middle East. A dataset comprised of NMR log acquired on one key well and special core analysis results from seven wells is utilized. Fifteen other wells are used as blind tests, once the model has been calibrated. Generalization to other carbonate reservoirs is then discussed.
In details, a pore size distribution predictor is defined, which makes use of a system built upon mercury injection measurements and is able to discriminate micropores, mesopores and macropores downhole. Making this categorization is fundamental for reservoir modeling purposes, fluid-flow behavior assessment and to understand wettability. Micropores are associated with bound fluids, so the aforementioned porosity partition supplies also an estimation of irreducible water saturation. The established link between NMR transverse relaxation time distributions and pore size distributions provides a direct estimation of NMR effective surface relaxivity, both at laboratory and reservoir conditions. This parameter is fundamental since it represents the main driver for the subsequent computation of permeability, by means of a new and advanced modeling. The overall good match between the entire NMR-based interpretation and core data demonstrates the reliability of the methodology.
Although several approaches exist to aid formation evaluation in carbonates based on quantifying rock texture by NMR, the quantitative use of effective surface relaxivity is relatively new and can shed new light on this topic.
