The description and characterization of hydrocarbon reservoirs are critical elements of reservoir management. In the broadest sense, this means finding quantitative means of describing highly variable petrophysical properties. The analysis of wireline log data is essential in this process. However, using wireline data for reservoir description often requires that data from only a few wells be extrapolated to the entire reservoir. This results in large uncertainties in the estimation of the needed petrophysical properties. Nuclear magnetic resonance (NMR) logging, integrated with conventional wireline and core analysis data, can provide information to reduce the aforementioned uncertainties and provide improved reservoir description.

This study demonstrates integrating NMR data with conventional wireline data. By using the concept of hydraulic delineation, it is possible to estimate petrophysical properties in intervals where core or NMR data are not available. NMR log data provide the basis for hydraulic delineation and the derivation of deterministic transform equations for petrophysical properties. The accuracy of every transform equation is statistically enhanced because of the sampling and the coverage attained through NMR logging. Intervals with similar flow properties are identified and grouped. Their unique log signatures link these intervals to conventional wireline data. Conventional log data - on intervals where NMR data are unavailable - are compared to the signatures of the controlled intervals to derive individual profiles. Finally, these profiles combined with the transformation equations, are used to estimate petrophysical properties for the unlogged intervals.

This methodology provides an avenue for enhancing the accuracy of reservoir description studies. It is possible to establish key economic quantifiers such as productivity index, injectivity profile, and water cut. Economic analyses of all potentially productive zones are thereby improved.


Decisions affecting the development of an oil reservoir such as well placement, perforation and recovery methods are based upon the characterization of the reservoir. Its economic value is also based upon the extent to which the reservoir has been characterized. Reservoir properties must be quantifiable for these decisions to be made rationally and the economic value to be realistic. Data collected concerning reservoir properties must, when properly integrated, answer the questions - how much oil is present, where it is located, how fast it can be produced, and how much will ultimately be recovered.

Openhole well-logs are one source of data. These data can be sparse, and are occasionally incomplete for reservoir description purposes. There are several reasons why such data are incomplete. First, well-log data characterize the reservoir near the wellbore and this represents only a minute fraction of the reservoir. Second, data are collected based upon the initial characterization of the reservoir and the perceived value each measurement might add. The initial assumptions and characterization of the reservoir strongly influence the types of measurements are made and over what interval. Finally, new technologies result in more measurement techniques having made available for use on recently drilled wells. Data that might be useful are often missing on older wells as well as on some of the intervals in newly drilled wells.

When this situation occurs, it is useful to reprocess the data from older wells and unlogged intervals using the newer characterization techniques. Hydraulic unitization, when combined with deterministic transform equations, is one method that is available to infer properties in incompletely logged intervals. Examples of applying this method include deriving a permeability profile from measured core permeabilities and wireline logs.

P. 99^

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