Abstract
Horizontal wells have altered the ability to deliver production from marginal reservoirs. Although reservoirs with permeability values in the 0.1 to 0.01 millidarcy range had been identified in many producing basins, these reservoirs could not be exploited until horizontal drilling became a commercial reality. Drilling these reservoirs became an engineering process that could be applied with great success in many different basins and rock types.
The completion of these resources proved to be a different matter, however. Wells that were drilled with excellent vertical control within the reservoir did not exhibit a consistent response to stimulation treatment along the entire traverse of the horizontal sections. Although reservoirs that are considerably different lithologically were always expected to provide a diversity of completion challenges. More troubling, however, are the reservoirs that appear to be homogeneous that do not respond well to similar treatment designs along the length of the horizontal section. Pump-in tests were used in an attempt to understand this lack of consistency. In some cases, pump-in tests followed by treatment designed from that data failed to yield the expected production results. Sometimes, breakdown could not be achieved in the reservoir.
This paper describes how to apply data derived from dipole sonic logs in horizontal sections of a well to establish the anisotropy along the length of the horizontal section. This information is then used to define the brittleness properties of each section of the wellbore. Perforations are selected to take advantage of this brittleness, and effective fracture treatments are designed and pumped.
These case studies are examples of this application in difficult and troubling reservoirs. In each well, the fracture treatment success increased as much as 100% and had reduced breakdown pressures. Production rates were also increased by 20% when compared to production from the nearest offset wells in the same reservoir and of similar horizontal length.