ABSTRACT

With the advent of horizontal wells and the use of wellbore microscanners, the ability to analyze naturally fractured reservoirs has increased markedly.The microscanner provides an estimate of aperture size, fracture orientation, and the nominal depth for each intersected fracture. Depending on the length ofthe well, up to several thousand fractures can be identified and measured in asingle well. We report on the statistics of over 13,000 fractures from severalhorizontal wells in a single field. Of particular interest is the scalingcharacteristics of the fractures.

The fractures have a single preferred strike orientation (N74E) with astandard deviation of less than 10°. In all wells there is a high probabilityoffending a fracture within any 5 ft interval. This demonstrates a homogeneousfracture distribution along the wellbore and implies similar homogeneity in thereservoir. However, the probability decreases when considering differentfracture subsets characterized by an aperture cutoff.

Aperture size appears to have a log-normal distribution, but the smallervalues approach the tools' resolution. This suggests that the log-normaldistribution is probably an artifact resulting from resolution bias and theunderlying distribution is a power law. Resolution also affects averagefracture spacing. By including increasingly smaller aperture fractures, averagefracture spacing decreases with power law behavior. Spacing variance alsodecays in a hyperbolic manner. One objective in analyzing these statistics isto create a basis for generating stochastic simulations of fracture networks.The intent is to provide probabilistic answers to questions that arise duringthe evaluation of naturally fractured reservoirs. One possible approach isdiscussed.

INTRODUCTION

One objective of drilling horizontal wellbores in naturally fracturedreservoirs is to attempt to intersect a greater number of fractures than couldotherwise be intersected by a vertical well. Presumably, this leads to higherproduction rates than for a vertical well. Often overlooked in this idealisticapproach is the necessity of characterizing fracture distribution in thereservoir. Differences in fracture density, aperture and connectivity translateinto varying production rates, drainage radius and recovery.

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