Comprehensive Design and Diagnostic Approach for Horizontal Completions in Carbonate Environment

Abstract

Abstract The development of tight carbonate reservoirs is moving towards drilling and completing wells with longer laterals. This leads to challenges of longer completion time, high number of fracturing stages, longer interventions, and eventually higher costs. Design cycle implementation is required to devise an engineered strategy to mitigate these challenges. Lateral landing was conducted based on the cross-section grid consisting of two offset horizontal wells completed with up to 13 fracturing stages. A longer lateral greater than 6,000 ft was drilled compared to 4,000 ft in offset wells to get the production potential. With a strategic design involving engineered chemistry and numerical simulation models, a cluster design was devised to reduce to stages. A mathematical algorithm employing tube wave velocity calculations was used as a diagnostic to ensure diversion success after each stage. The horizontal lateral was landed traversing the prolific layer. Stage reduction sensitivity simulations were conducted using multiphysics numerical models and novel beta factor workflows to evaluate the extent of stage reduction. The design was extended to plan for five stages only, with increased number of perforation clusters per stage. The reliable diversion chemistry utilized was accompanied by a revised perforation length as dictated by the beta factor workflow. A total of 39 clusters, 2-ft each, were distributed across 6,000 ft with four mechanical isolation plugs. A novel nonintrusive diagnostic model built on mathematical fundamentals of wave travel time was used with a Bayesian statistical approach after each diversion pill placement to ensure fracture fluid entry points and enough coverage in each stage. The high fluid viscosity and operating pumps during the water hammer events resulted in low signal-to-noise ratio in the input data. To overcome these limitations, the water hammer events were processed with a combination of two newly developed algorithms: predictive deconvolution and comb filter, which produced more robust results than the traditional approach. Consequently, the well production was analyzed to show equivalent or higher productivity index compared to the offset laterals with up to two times higher stage count. The paper presents a unique example in which an experiment was fully engineered from design to evaluation and monitored with reliable diagnostics. This example gives a blueprint for future completion designs.