A Simulation Model to Evaluate Decision Rules for Lock Operations on the Upper Mississippi River L. Douglas Smith, Donald C. Sweeney II, James F. Campbell College of Business Administration, University of Missouri-St. Louis ldsmith@umsl.edu ,, sweeneyd@msx.umsl.edu , campbell@umsl.edu Abstract We describe a discrete-event simulation model used to investigate the impact of alternative decision rules for operations in a congested section of the Upper Mississippi River navigation system. The model accommodates commercial vessels with different barge-tow configurations and recreational vessels. Activities vary by time of year, day of week and time of day. We reveal potential benefits of alternative priority dispatching rules and show their differential impact on vessels with different tow configurations. 1. Introduction The Upper Mississippi River (UMR) navigation system extends 660 miles from Minneapolis, MN, southward to the confluence of the Mississippi and Missouri Rivers near St. Louis, MO. In 2004, approximately 73.3 million tons of cargos were transported in “tows” of barges [21] through a series of 29 lock and dam facilities that accommodate both commercial tows and other non-commercial vessels. The dams create a series of 29 “pools” of sufficient depth for reliable navigation and the locks enable passage between adjacent pools in the water staircase (Figure 1). Most of the locks were constructed with main chambers 600 feet in length designed to accommodate the largest commercial tows in the mid 20 th century. Exceptions are three 1200-foot locks numbered 19, 26, and 27 in Figure 1. Most fully assembled tows on the river now exceed 600 feet in length and require that a group of barges be decoupled from the fully assembled flotilla in order for the tow to pass through the 600-foot long locks in two separated sections. The two segments of the tow are subsequently re-coupled after passage through the lock and the tow then continues its journey. Such lock operations are called “double lockages” and require considerably more time to complete than do the “single lockages” required to process flotillas able to fit in the lock chamber in their fully assembled configurations. Figure 1. The UMR navigation system The most congested and heavily used section of the river involves a network of five locks numbered 20, 21, 22, 24, and 25 in Fig. 1 (there is no lock numbered 23). Queues of recreational and commercial vessels form periodically upstream and downstream at each of these locks. Each lock is surrounded by different terrain and river conditions that affect the efficiency with which they can process tows of different configurations and also have slightly different mixes of vessel traffic. Vessels are processed using a first- come, first-served (FCFS or FIFO) policy, but with some priority granted to commercial vessels without barges and to recreational vessels as they can be processed quickly without significantly delaying the processing of commercial vessels with barges. Under extreme operating conditions, the FIFO policy is waived with the agreement of commercial tow operators, and tows are processed through a congested lock in sequences that allow greater local efficiencies to be achieved. In years when traffic has been heavy, or when locks have been out of service for lengthy periods, vessels have experienced prolonged delays (sometimes as long as 100 hours) as they waited for their turn at these congested locks in the UMR system. To relieve such congestion and to increase the carrying capacity in this section of the river, the U.S. Army Corps of Engineers (USACE) has proposed expanding the five locks Proceedings of the 40th Hawaii International Conference on System Sciences - 2007 1 © 1530-1605/07 $20.00 2007 IEEE