Abstract

Carbonate reservoirs can be complex and difficult to understand from conventional logs. Low porosity from density/neutron logs in the 3 to 4% range is not unusual in many dense carbonates and is quite typical in some. Even in these low-porosity conditions, secondary enhancement features can have a dramatically positive impact on production. An adequate description of these characteristics is necessary to determine completion candidates.

Nuclear magnetic resonance (NMR) responses have been proposed as a method that can measure and predict the results of these alterations. One analysis method of this data is the Bray-Smith permeability equation. This paper examines the application of the Bray-Smith permeability equation for determining the productive capacity of several carbonate reservoirs. This calculation uses relaxation (T2) NMR responses to provide direct permeability without any external inputs. It is not necessary to adjust for rock type, and the equation corrects for any rock texture differences.

Responses from a tight lime with natural fractures, an alluvial system with a carbonate base, and a tight carbonate reservoir with secondary porosity and isolated molds are included. This diversity provides a good overview of the possible variations in carbonate reservoirs. Consistent characterization of the primary and secondary porosity conditions of every reservoir is provided by the NMR response. The Bray-Smith equation for permeability provides an approximation of core results (where available) and of actual production results in every case.

Introduction

Carbonate reservoirs comprise many of the productive intervals throughout the world. These carbonate formations can exist in various forms. The consistent characteristic of these reservoirs is that any attempt to measure and quantify the ability of the formations to produce hydrocarbons is a difficult exercise. The most productive of these reservoirs consist of horizons with extensive alteration of the basic carbonate fabric. There are, however, some excellent producers from apparently very tight carbonate reservoirs that defy log attempts to define their character.

Fig. 1 shows an example of one of these very tight reservoirs with porosity logged on a limestone matrix. The gamma ray has very little character and varies between 15 and 30 API units. Resistivity is near 100 Ohm-meters through the section. The unusual characteristic of this resistivity log is that the deep resistivity does not overlay the shallow and medium. This could be an artifact of the log or of the setup of the log. Regardless, the resistivity is sufficient to be caused either by a hydrocarbon filled reservoir or one that is tight.

The density/neutron log indicates a very tight reservoir. The neutron measurement varies from 4 to 6 porosity units (p.u.), with most of the measurement at 4 p.u. The area with the apparent higher porosity does not have a similar response from any of the other curves. There is some movement in the spontaneous potential (SP) that appears to reflect the neutron response, but in some areas, there is not a relative response. The density response varies from -2 to -6 p.u. A traditional crossplot of density and neutron porosity yields a porosity of 0%. The photoelectric cross-section (Pe) provides clarity because it is approximately 3.5 through the entire section. This would lead to the conclusion that this is an almost pure dolomite section with very low porosity. This log is from a productive reservoir.

The density can provide some clue for the ability of this reservoir to produce. The movement from -6 to -2 p.u. through the section might indicate the presence of fractures. A focused density measurement is obtained as the pad is pulled across a formation. Any fracture would appear to be a large, infinite event and would reflect as high spikes on the log. In this case, the indication might or might not be accurate, as sections of this reservoir with this log signature are productive, and other sections with virtually identical log responses are not. Therefore, the solution must be found using another log characteristic.

This content is only available via PDF.
You can access this article if you purchase or spend a download.