Addressing Capacitance-Resistance Modeling Limitations and Introducing a New Practical Formulation

Open Access
Author:
Altaheini, Suleiman
Graduate Program:
Energy and Mineral Engineering
Degree:
Master of Science
Document Type:
Master Thesis
Date of Defense:
July 17, 2015
Committee Members:
  • Turgay Ertekin, Thesis Advisor
Keywords:
  • crm
  • crmip
  • crmp
  • crmid
  • Capacitance-Resistance
  • injection stream
  • capacitance resistance modeling
  • png
  • petroleum.
Abstract:
Capacitance-resistance modeling (CRM) has become a popular and convenient tool that is used to model production rates and assist in reservoir characterization. Upon successful modeling, engineers would be able to link reservoir injected volumes to respective production wells, quantify injection signal dissipation times, and qualitatively infer presence of no-flow barriers and fractures. Requiring only well rates and bottom-hole pressures as inputs, CRM is able to mimic streamline simulation when seeking injected fluid allocation, and help guide reservoir engineers to educated decision making. However, this nonlinear semi-analytical tool, due to its simplicity, comes with a few documented limitations. This study aims to build a robust capacitance-resistance model by both addressing modeling limitations, and introducing a new modeling equation. CRM limitations of extended producer shut-in times and new well introduction have been addressed by employing pseudo-injection rates. Aquifer influx cases are also successfully modeled by mapping influx rates to a new injector. If the aquifer influx cannot be modeled, the rates are inferred from the CRM's match quality, but are not a guarantee representation of actual aquifer influx rates. The limitation of highly compressible fluids, however, remains to be a modeling hurdle. The new modeling equation encapsulates the compound heterogeneity effect on injection streams from near-producer and far injection-well volumes. The model is validated against a numerical reservoir simulator, and compared to another documented equation. The model produced acceptable results using fewer variables with physically-accepted values. Cases used for modeling include an inverted 9-spot injection pattern, a compartmentalized reservoir, and dual stacked reservoirs with cross-flow communication.