Developing an Agent-based Model for the Depot-based Water Allocation System in the Bakken Field in Western North Dakota

Abstract:

The oil production at the Bakken Shale increased more than ten times from 2008 to 2013 due to technological advancement in hydraulic fracturing and North Dakota has become the second largest oil producing state in the U.S. behind only Texas since 2012. On average it requires about 2-4 million gallons of freshwater to complete one oil well in the Bakken field and the number of oil well completions (i.e., hydraulic fracturing) in the Bakken field increased from 500 in 2008 to 2085 in 2013. A large quantity of freshwater used for hydraulic fracturing renders a significant impact on water resource management in the semi-arid region. A novel water allocation system – water depots – was spontaneously created to distribute surface and ground water for industrial uses. A GIS-based multi-agent model is developed to simulate the emergent patterns and dynamics of the water depot-based water allocation system and to explore its economic and environmental consequences. Four different types of water depot are defined as agents and water price, climate condition, water source, geology, and other physical and economic constraints are considered in the model. Decentralized optimization algorithm will be used to determine the agents’ behaviors. The agent-based model for water depots will be coupled with hydrological models to improve the region’s water resources management.