The Bakken boom in North Dakota is currently focused on the Central Basin area where around half of the drilling rigs in North Dakota are now operating. What makes this area different compared to previous areas of Bakken development is that there is only minor structural variation and significantly less naturally induced fracturing as compared to the higher permeability rock facies that exist in the sub-reservoirs of earlier Bakken development such as at the Elm Coulee field in Montana or the Sanish and Parshall fields in North Dakota. As a consequence, the role of the well's completion and stimulation design has a greater significance and impact on well productivity and ultimate recovery.
Different companies have taken very different approaches to well design using either plug and perf or ball and sleeve completions, and a variety of fracture designs with slickwater, hybrid or cross-linked gel fluids and a variety of proppants from 100% natural sand to 100% ceramics. As a consequence, it is not uncommon for different operators to have over a 2 million dollar difference in their AFE's solely because of the differences in approach to the well's completion and stimulation design. The authors have chosen to apply "advanced completion and stimulation designs" which are designed to maximize the reservoir contact area (slickwater and plug and perf) and optimize the conductivity (ceramic proppant at relatively high volumes).1,2
In order to benchmark performance of its completion and stimulation program the authors in 2010 developed a production and completion database of all wells completed in the Central Basin using publicly available information from the North Dakota Industrial Commission's records augmented by additional completions information obtained directly from the operators. From an initial dataset of ∼30 wells the database has been updated monthly and has now grown to over 1100 wells in the Central Basin from 28 operators. Benchmarking of completion performance has been performed using the above database together with a Petra geological database developed from all publicly available logs in the Central Basin (∼500 vertical wells which had been drilled and logged prior to the first Bakken horizontal well).
Benchmarking of performance is somewhat subjective in the Bakken (as well as most reservoirs) due to variations in reservoir quality. Without geological input (reservoir quality), the acquired data were too scattered to achieve meaningful correlations to completion methods unless the area was limited to ensure similar reservoir quality for the wells being evaluated. By narrowing the analysis to very limited areas; this also reduced the input as to the number and type of completion methods being compared.
Multivariate analysis methods, that included geological input, were used to benchmark performance over most of the Williston Central Basin, allowing comparison of the varied completion methods for 28 operators and over a thousand wells. Using multiple parameters an excellent correlation for completion methodology versus reservoir quality was obtained.
While early analysis focused on 30, 60 and 90-day cumulative production; in the past year there have been enough wells to generate performance metrics based on 180 and 365-day cumulative production since over 600 wells have now been on production for at least a year.
The analysis shows that companies with a greater completion spend are typically paying out the increased spend in 1 to 8 months depending on the area and these wells are still producing at significantly higher rates after one year of production. The declines of these wells also project a significantly greater EUR and hence have lower F&D costs per barrel of oil. This higher performance continues to hold after 365-days of production.