A Multimodal Network Approach to the Inland and Coastal Waterway System

https://doi.org/10.5281/zenodo.26169662018Final Research Report; 07/01/2017-11/30/2018PDFTech Report69A3551747130Waterway facility operationsDredgingStochastic programmingMultimodal transportationFreight transportationTraffic congestionWaterwaysMarine TransportationPlanning and ForecastingUnited StatesMaritime Transportation Research and Education Center (MarTREC)Wang, BruceMahmoudzadeh, AhmadrezaMaritime Transportation Research and Education Center (MarTREC)University of Arkansas, Fayetteville. Dept. of Civil EngineeringUnited States. Dept. of Transportation. University Transportation Centers (UTC) ProgramUS Transportation CollectionThe inland waterway system carries a significant percentage of the national freight. Maintenance operations including dredging and dam repair are important to maintaining the effective and efficient operation of the system. Dredging projects are for recovery of the navigational channel draft from the shoaling effect while lock/dam repair is about maintaining a maximum possible operational hours to reduce the waiting and delay of vessels therein. The special feature in this study is that the shoaling effect is random, as is subject to weather and other effects. This study specially deals with maintenance fund allocation to these maintenance requests by first proposing a multimodal approach for formulating the waterway maintenance problem in a connected network, which considers rivers, locks/dams, and highways and railways. The random shoaling effect essentially renders a two-stage stochastic problem, which is in our case solved by a deterministic approximation. The solution identifies maintenance projects to fund for the most system benefit in terms of serving the most OD demand at the least cost. Improving the locks and dams has a random effect on the delay function, so a probability distribution is considered for the delay. The unwanted interruption delays, which is not related to the traffic, are detected from the data set; and two functions, a linear and a non-linear respectively, formulate the amount of improvement gained by the number of hours delay reduced. The historical origin-destination tonnage data and associated flows along the routes of each commodity OD are used to optimize the maintenance project selection based on the needed dredging depth of each project, the future dredging needs, and the lock and dam rehabilitation needs. Another distinct feature of the model is that it considers interdependence of the maintenance projects, which means the benefit along a route does not realize if not all the improvement projects along a route are conducted. The model is applied to the Ohio River basin network which includes the land side routes. The results show the importance of considering the lock and dam rehabilitation costs in different budget scenarios for each linear or non-linear delay cost approximation. The optimal decision heavily favors lock/dam repair over dredging along the navigable river channels.78