SEPTEMBER-OCTOBER 2001
Vol 2 No 5

 

The Application of Image Analysis to Improve Permeability Prediction

Kate Hatfield and Jenny Garnham

 

ABSTRACT

Key to the understanding of fluid flow within a hydrocarbon reservoir is the acquisition of core, and the measurements of core porosity and permeability. Therefore, in regions where coring is not viable a local pre-established permeability-porosity relationship may be used for the prediction of permeability from wireline-derived porosity. This paper investigates a poorly correlated (for prediction purposes) permeability-porosity relationship of a single reservoir. Image analysis techniques, utilizing back scattered scanning electron microscope images, have been employed to understand and better constrain the relationship.

A stronger relationship, statistically and for prediction purposes, is obtained if image porosity, rather than plug porosity, is used in the permeability-porosity plot. Further investigation to determine the cause of this improved relationship was undertaken by considering pore size distributions. It was concluded that the image porosity did not include micro-porosity in its measurement and that micro-porosity does not contribute to fluid flow, whereas, micro-porosity is included in the plug porosity measurement in core analysis.

 

Lithology of Paleovalley-Fills from Petrophysical Data: Mannville Group, Lower Cretaceous, Southern Alberta, Canada

John Hopkins and Rudi Meyer

 

ABSTRACT

Reservoirs in the middle part of the Mannville Group in southern Alberta are hosted in permeable (10–500 mD) quartzose sandstones deposited in a variety of environments. The sandstones have commonly been incised by Upper Mannville paleovalleys filled with lithic sandstone and shale. Lithic sandstone is relatively impermeable (5–60 mD) so that Upper Mannville paleovalleys can form lateral seals to Middle Mannville reservoirs. Particularly where the sections adjacent to paleovalleys comprise siltstone or shale, difficulties are encountered in mapping the valley courses and knowledge of the lithologic variation and succession within valleys becomes critical.

Lithology is determined from density-versus-neutron porosity and gamma-versus-“neutron-minus-density” porosity cross plots which resolve porous quartzose sandstone, cemented sandstone, lithic sandstone, calcareous shale and shale. Lithic sandstone modes are modified by the presence of shale intraclasts, occurring as pebbles, easily recognized in core, but also as sand-size intraclasts that form part of the detrital grain framework. The presence of large and small intraclasts is consistent with the analysis of the total-gamma radiation spectrum. In addition, shale intraclasts within lithic sandstone are characterized by relatively high thorium counts.

In the modal lithic sandstone the spread between density and neutron porosity is about 10 porosity percentage units. A few wells have produced oil from Upper Mannville lithic sandstone, invariably from the most permeable medium-grained lithic sandstone. Use of the techniques illustrated herein help to evaluate the potential of lithic sandstone paleovalleys aiming to identify the location of medium and coarse-grained lithic sandstones.

 

Velocity to Porosity Transform in Marine Sediments

Manika Prasad and Jack Dvorkin

 

ABSTRACT

We apply a new effective medium model for calculating elastic moduli of high-porosity ocean-bottom sediments to several Ocean Drilling Program (ODP) well log data sets from offshore Amazon delta, Peru margin, and Bengal Bay. The model relates the bulk and shear elastic moduli of the sediment to porosity, differential pressure, elastic constants of the grains, and pore-fluid compressibility. As a result, we can accurately match sonic well log data using porosity and differential pressure as the main inputs. Therefore, velocity data can be used to infer porosity and pressure.