Abstract
Poor production results in the subject area was determined to be caused by poor reservoir contact along the laterals. In a series of wells using multi-stage plug-and-perf completions, poor perforation efficiency was diagnosed. A solution to improve formation contact was employed which made absolute contact with the reservoir by using sliding sleeves and coil tubing technology. However this system demonstrated issues mainly around lower estimated ultimate recovery (EUR) values due to wide fracture spacing limitations and also suffered occasional human related, not tool quality, failures which caused costly penalties in completing wells. Returning back to the original completion scheme re-presented the original problem that very few of the perforation clusters were opening and thus using plug-and-perf and also obtaining better perforation efficiency was re-sought.
This solution step developed in this paper uses drilling data which through the use of neural networks, can generate a synthetic log. The log can yield rock properties which can generate stresses along the wellbore. These stresses are then used to select perforation spacing to better take advantage of ‘like’ stresses so that perforation opening can improve. Minimum pressure drops between perforations can then be designed. Additional information pertaining to reservoir quality was generated by adding gas chromatograph data to determine the highest permeability sections of the lateral as well as the liquid-rich hydrocarbon bearing sections. This analysis allows for a solution matrix of job and cluster spacing types to optimize fracture spacing (frequency of contact) and conductivity requirements to optimize the completion based on stress and relative or contrasting permeability.
This paper will demonstrate how the problem was determined, the solution was implemented, how some alternatives were explored and how fracture geometry was greatly improved using these workflows.
