The ability of induced fractures to improve well production is dependent upon three primary characteristics of the fracture geometry: fracture height, fracture width, and fracture length. It is therefore very important to develop methods which can determine each of these characteristics. The evaluation of a new nuclear measurement for fracture width determination is the subject of this paper.

The use of non-radioactive tracer (NRT) tagged proppant with a pulsed neutron capture (PNC) tool or a compensated neutron tool (CNT) has been successfully employed in over 200 wells to determine induced fracture height in cased downhole formations. In several of those wells, independent rock-properties based computations of estimated fracture width and height were also available. Excellent qualitative agreement was observed between the magnitudes of the NRT proppant signals and the estimated fracture widths (estimated fracture heights also compared favorably). To further investigate the capabilities and limitations of using pulsed neutron tools to evaluate the widths of NRT tagged fractures, MCNP5 Monte Carlo software (Los Alamos National Laboratory, 2003) was used to simulate the NRT fracture width responses of several PNC measurements (e.g. formation sigma, capture gamma ray count rate, and gadolinium yield), and extensive modeling data were analyzed.

First, for each simulated PNC measurement, the relationship between fracture width and NRT signal was determined and evaluated in an ideal cased borehole environment. Based on these fracture width signals, an investigation was then conducted to study the depths of investigation (DOI) of fracture detection into the formation of selected PNC measurements. The modeling results show that there are monotonic relationships between fracture width and the NRT signal for each measurement, but there are significant differences among the measurements in fracture width signal magnitude and measurement DOI.

To further evaluate how these fracture width signals would be affected by changes in the borehole environment, and the possibility of compensating/correcting for these non-fracture related interference effects, additional MCNP studies were conducted to evaluate the effects of:

  1. borehole fluid salinity,

  2. azimuthal nonalignment of the fracture plane and the logging tool, and

  3. the presence of tagged proppant in the borehole (cement) region.

In some cases, and especially when there is significant tagged proppant in the borehole region, the corrections were determined to be too large for effective compensation.This paper briefly describes the NRT method and shows log examples comparing NRT and rock properties based indications of fracture width and height. Extensive MCNP modeling data are then presented to further evaluate the capability of using NRT tagged proppant to detect fracture width. Conclusions are drawn regarding:

  1. which PNC fracture width measurement is least-affected by near-wellbore variations,

  2. situations when it is possible to obtain good indications of relative fracture width, and

  3. situations when fracture width estimations are impractical.

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