Abstract
The Bakken/Three Forks oil play is extremely active, with more than 2,500 horizontal wells drilled in Montana, North Dakota, and Saskatchewan during the past nine years. Typical well costs range between $4 -$8 million per well, suggesting the industry has invested more than $15 billion in this resource, in addition to lease, infrastructure, and operation expenses. Despite this enormous investment, the industry has struggled to identify best practices and many operators currently pursue significantly different strategies throughout the play.
This paper will summarize the development philosophy and key learnings of one operator in completing this low permeability reservoir with hydraulically fractured horizontal wells. Significant production gains have been achieved through advancements in completion methods, improving the overall economics of the Bakken/Three Forks oil play.
This paper describes the evolution of completion strategy, including:
- –
Lateral length and number of stages
- –
Method of isolation
- –
Completion and fracturing techniques
- –
Proppant type and concentration
Through careful investigation and continual improvement, the authors have been able to significantly increase well production compared to their previous wells. Wells with improved completions have also experienced notably higher production rates than offset wells completed by operators with differing strategies. It is clear that lateral length and number of stages, coupled with proppant type and concentration, are key variables that have significantly improved well productivity and profitability. Included economic and financial analyses demonstrate that Bakken/Three Forks profitability can be dramatically improved by altering completion designs. Although specific treatment parameters should be adapted as formation characteristics vary across the play, it is believed this discussion of current design strategy will be useful to the SPE audience and accelerate optimization of Bakken/Three Forks completions. Future innovation will occur, and it is hoped that the publication of these learnings will encourage additional experimentation and optimization.
