Alternative Financing for Harbor Infrastructure using Big Data Analytics in the Great Lakes Waterway

Abstract

Decades of under-investment into aging infrastructure have resulted in uncertain reliability and systemic under-performance. The infrastructure spending gap in the U.S has grown to $2.6 trillion, and estimates suggest half of that is necessary within the next five years to avoid major impact to GDP. Yet spending levels remain below needs and policymakers seek more efficient allocation models for public funds and alternative financing mechanisms to accelerate the pace of investment to meet society’s needs. There is substantial private capital ready to enter the infrastructure sector along with innovations in contractual public-private partnership models. Financing mechanisms, such as infrastructure banking, show promise in extending the value of federal spending. However, a gap exists in the modeling of revenue streams and risk exposures for private entities which are necessary for the integration of public and private capital. Big data analytics are applied in this research to reveal opportunity costs and risk exposures which we apply to model revenue streams and assess infrastructure funding decisions. This dissertation investigated the waterway infrastructure of the Great Lakes, which comprises a network of deep-draft ports and connecting channels that serve a prominent role for commerce and manufacturing in North America. The waterway system requires annual funding to maintain navigable depths and functional port and lock infrastructure. An obstacle to funding decisions is the uncertainty surrounding financial returns on investment from improved maritime efficiency, in part because transportation and logistics metrics or benchmarks are lacking. Iron ore, the primary commodity in the Great Lakes, serves as the use case in this work to assess performance metrics for the waterway infrastructure that enables efficient and sustainable transport from mines to steel mills. This dissertation integrates new data analytics across traditional disciplinary silos to gain new insight into the risks, performance, and funding mechanisms for harbor infrastructure. Corporate financial metrics are used to map and quantify interdependencies within the value chain from iron ore production to finished goods. These interdependencies are further applied to assess financial risk exposures to infrastructure disruption using analytic tools such as input-output modeling. We applied big data analytic tools to assess the performance of maritime shipping with highly granular spatial and temporal datasets, including vessel draft, transit time and cargo. Vessel position information from historic Automatic Identification System (AIS) was used to develop a novel Maritime Transportation Efficiency (MTE) metric, defined as mass per time and directly applicable to bulk carriers. Regression analysis of vessel performance to hydrologic conditions in the waterway provided a means to predict changes in logistics performance resulting from infrastructure investment. We use Monte Carlo simulation to calculate expected MTE for vessels in the waterway under varying conditions which are correlated to transportation costs. Analytics techniques, like those applied in this dissertation, are useful to model revenue streams and reveal potential for new funding mechanisms and market-driven financing models. We suggest a new funding model for harbor infrastructure based on user demand with a fee structure adaptive to actual vessel requirements, attainable through existing data sources and new analytical tools. Demand-driven funding decisions for harbor maintenance can maximize value returns for users. A fee structure, outside of the Congressional appropriations processes, is more responsive to user needs and provides a means to deploy alternative financing models such as infrastructure banking for waterway maintenance and port depth construction dredging.