Completion Optimization and Validation in the Williston Basin Through Integrated Datasets

Abstract

AbstractWith the surge in activity in the Williston Basin, the ability to optimize completions has mainly been an expensive trial and error system. With the advent of sliding sleeve technology, operational efficiency could be gained; however, the effectiveness of the completion remained unclear. This paper details the necessary datasets to gather and a workflow to analyze them in order to optimize completion efficiency. A field trial of the optimized completion design, along with similar data gathering, could then be done to confirm the results from the previous dataset.By varying completion and isolation styles on a select set of wells and obtaining a wide range of data, an optimized design can be obtained at a fraction of the cost, including isolation type, job size, stage length, number of entry points, entry type, perforation scheme, and rate. In Phase 1, an s-shaped pilot well was drilled for core, specialty logs, DFITs, and microseismic monitoring. A combination of sliding sleeves, rapid frac sleeves, and plug and perf were used across two open-hole adjacent laterals. Two cemented laterals were used initially as horizontal monitors and then to test different proppant volumes and number of clusters. The tracer program was designed to examine both stage isolation and well-to-well communication. Hydraulic Fracture Pressure History Matching (FPHM) was done on all stages; trends of propped fracture area and proppant concentration could be broken out by completion and isolation type.Cemented, three entry point plug and perf was the clear winner. The FPHM and microseismic analysis demonstrated that this design reduced out-of-zone height growth and tracer results showed it minimized stage-to-stage and well-to-well communication. Multivariate analysis suggested an optimal stage length and pounds of proppant per foot. Using the results from all the datasets, the optimized completion design could be input into the calibrated fracture model to forward model the predicted fracture geometry. Phase 2 was implemented to test the effectiveness of the new design and test the applicability of the calibrated fracture model. A microseismic and tracer (water, proppant and oil) program was executed and FPHMs were conducted. The FPHMs showed that actual fracture geomoetries were very close to what the model was predicting. There was also greater containment with the new design from a microseismic and tracer perspective. Finally, production exceeded the predictive production models from the multivariate analysis.This project provided an optimized completion design that could be used across an area of interest, but there was valuable insight beyond the completion design itself. Numerous lessons were learned about data acquisition and how to improve data quality with future projects. Fracture azimuth was confirmed with microseismic and tracer data. The effects of depletion could be seen by these datasets as well. The FPHM outputs also provided insight into height growth out of the pay zone being attributed to higher rates and fewer entry points. The importance of WOR was also evident with the multivariate analysis.Knowing what type of completion to pump and knowing where to apply the design will always be a topic of concern. This methodology demonstrates a cost effective solution to this complex problem and verifies the results with field application.