(524f) A Supply Chain Optimization Framework for Distributed Renewable Ammonia Production

Ammonia is a critically important chemical for global food production which can be either used directly as or processed further into fertilizer. The Haber-Bosch process revolutionized global food production and greatly increased mankind's ability to support a larger population. However, the way ammonia is produced and transported currently is unsustainable. Ammonia is produced using hydrogen obtained from fossil fuel sources, ultimately polluting the environment. It is then transported over long distances by train, pipeline, or even ship: over one-third of ammonia consumed in the United States is imported. It would be more desirable from a sustainability standpoint to have a cleaner, smaller scale, and more local ammonia supply chain. Motivated by this goal, a first of its kind plant has been built in Morris, MN, using the wind-powered electrolysis of water to produce hydrogen without the carbon emissions inherent in fossil fuel gasification. This plant has been demonstrated to produce renewable ammonia with minimal carbon emissions at a scale of 26 t/y. Using data from this plant, it is reasonable to ask how, and at what scale, renewable ammonia plants would be economically feasible with respect to the overall ammonia supply chain.

To address this question, this work aims to create a framework for the analysis of renewable plants in a general region's ammonia supply chain. Candidate sites for renewable plants are placed in towns with nearby wind turbines with maximum plant capacity determined by wind turbine energy output. A supply chain optimization problem is fomulated to decide where and how large to build new renewable plants, as well as how much ammonia should be transported between conventional plants, renewable plants, distribution centers, and demand centers to meet local demand. The problem is a nonlinear program due to economies of scale when scaling up the Haber-Bosch reactor and related separation units. The problem can be set up to minimize economic or environmental objectives. When doing so, a clear tradeoff emerges: a minimal cost supply chain obtains most of its ammonia from a few large centralized plants, while a minimal emissions supply chain obtains its ammonia from many small distributed plants. A multiobjective optimization problem is formulated to analyze these tradeoffs via a Pareto curve of annual cost vs. annual carbon dioxide emissions.

The developed framework is then applied in two separate case studies to two upper midwestern states, Minnesota and Iowa, respectively. Despite the fact that these states are neighbors, they have various key differences that make separate analysis instructive. For example, Iowa has more distribution centers, more pipeline connections, more overall demand, more wind potential, and a more geospatially homogeneous ammonia demand distribution than does Minnesota. Despite these differences, we establish that for both states, a renewable plant scaled up to the order of 100,000 t/y ammonia production is selected for inclusion in the base case optimal supply chain. We then further explore how the differences between the two states affects the solution sensitivity, the impact of a carbon tax on the supply chain, and the improvements required for smaller scale plants to be economically feasible.