Integrating MTS Commerce Data with Multimodal Freight Transportation Performance Measures to Support MTS Maintenance Investment Decision Making (2014)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

34 Generalized Findings The most obvious issue that surfaced in this research effort is that there is a lack of the kind of data needed for develop- ing a model that can support MTS maintenance investment decision making by being correlated between the modes and almost no accurate data on origins and destinations (in the case of publicly available data). The research team had ini- tially hoped to be able to access both the Corps’ dock-level trip data file and the confidential FAF3 file and investigate whether they might work together to answer some of the questions that remain unanswered. Although the team was not able to access these data, the researchers searched for ways to combine public data from various sources in a way that would provide relatively meaningful, if not totally accurate, information on cargo flows and the interactions between modes. The research team believes that it found enough useful information to facilitate future investigations at a deeper level and with an expanded scope—a scope that involves the entire trade corridor. A combination of publicly available informa- tion from the Corps (primarily CPT data); FAF3; regional, state, and local freight studies; and trade literature makes it possible to gain enough of an understanding of a port’s primary trade flows to be able to provide meaningful input into the model developed for this project. There is another significant deficiency that can only be solved by the Corps. The goal of this research project was to develop a new methodology for evaluating potential O&M expenditures by a port. However, the metrics currently used for prioritizing these expenditures are not based on post-project evaluation of the effects of such spending; rather, they are based on presumed measures of the general importance of the asset: tonnage, sus- tainability of the region, and so forth. In order for a model of the type proposed by this research project to have real value, there must be a way to measure the effect of maintenance projects on freight flows. While this project used tonnage as the maxi- mization metric, tonnage is not the only viable metric. The modeling approach used here can be modified to maximize (or minimize) other meaningful variables. If the proposed model were to be used for new projects, this deficiency would not exist because a detailed evaluation of traffic patterns with and without a proposed new project is typically done. The limited geographic scope of this research does not lend itself to making generalized statements for the entire U.S. sys- tem. The case studies presented in this report are illustrative and provide a foundation for future research, but they cannot be considered definitive across a broad range of projects. Further model development with more complete data and a more comprehensive supply chain will be necessary to make recommendations on overall system performance. Despite several limitations, the general findings from this proof-of-concept level analysis can be helpful to decision makers. It provides a framework within which to evaluate bud- get requests for a set of projects that are directly or indirectly tied into the surface transportation system. This framework enables an analyst to evaluate the freight flows to and from the port and determine the criticality of given corridors or pathways. This work highlights the need for and the impor- tance of considering waterway investments within the context of multimodal systems, connections, and investments. What the Data Show Without the Model The data-gathering and analysis processes for this research project revealed some interesting aspects of the case study ports, even without the model: 1. There is a wide variation in the degree to which the bot- tom stratum of water depth are used. For example, CPT data revealed that 92 percent of Duluth’s shipments used the bottom 3 ft of water depth. In other words, the autho- rized channel depth is critical to operation. Conversely, only 21 percent of the tonnage in Hampton Roads uses S E C T I O N 4 Findings from the Case Studies

35 the bottom 3 ft, and only 10 percent of the tonnage at Plaquemines uses its bottom 3 ft. Table 23 provides the statistics for all case studies. Not surprisingly, where the vessel types are more homo- geneous, the depth utilization is the greatest. For example, both Portland-Shallow Draft (Portland-Inland) and Huntington have high utilization percentages (77.6 per- cent and 93.5 percent). Duluth had the second highest percentage and the second greatest tonnage utilization of the bottom depths. It is a port that handles a high volume of a limited number of commodities, and the vessels that transport those commodities are fairly similar in their design and operating characteristics. In future model- ing efforts, it would be useful to incorporate the relative importance of the water depth that is being dredged as a metric in the prioritization process. 2. Highways are not at all critical in some instances. In Duluth, a high percentage of the commodities that move through the port arrive by rail and leave by water. Trucking plays almost no role in their transport. In Plaquemines, a high percentage of the traffic arrives by barge and leaves by deep draft vessel. Landside transport tends to be for very short distances; most of the cargo remains close to the water. In the case of Huntington, a high percentage of its traffic originates and terminates at waterside facilities, coal being the primary example. In Hampton Roads, the only com- modity that uses the bottom 3 ft of channel depth is export coal, and this commodity arrives exclusively by rail. Of the five case study ports selected for this research project, only one (Portland) was significantly tied to its highway system. 3. Highway congestion caused by port truck traffic is a con- straint only in the immediate vicinity of the port termi- nals. Prior work conducted by the UMRP at TTI revealed that even with a large increase in port throughput, once the traffic leaves the port area, it disperses quickly enough that it does not significantly affect congestion measures any- where else. If trucking companies are selective in the routes that they travel in urban areas, and if they are able to time their urban movements outside of peak hours, the effect on traffic will be negligible. What the Model Shows Appendix F shows the results of the model for each of the scenarios specified in Table 22. Although the model considers water, rail, and highway segments, it only makes decisions regarding the improvement of waterways through dredging and the improvement to locks/dams by reducing delays. In effect, the model treats the current state of landside transpor- tation (railroads and highways) as a constant (constraints will not be removed). The model shows the following: 1. Only a few segments really make a difference. In all sce- narios, many segments are not selected for improvement. Segments that pertain to landside modes are not considered in the maintenance decision. When a waterway segment or lock/dam segment that is associated with a potential main- tenance project is not selected, it means that the improve- ment does not add any additional capacity to the system. This could be due to any of four reasons: • The fixed landside capacity constrains the input and output of the waterway segment. • There is limited demand from all origin-destination pairs that pass through the links, and there is no justifi- cation for spending additional money on maintenance. • The capacity of the segments is not a limiting factor in light of the origin-destination demand; therefore, spend- ing maintenance dollars on these segments would have no effect on system-wide capacity or throughput. • When the budget is limited enough, there is competi tion between different segments for maintenance dollars; accordingly, only the segments that bring the greatest increase to system-wide capacity would be selected for funding. Typically, the zeroed segments would be selected for improvement if the budget were increased; they are not constrained by demand or landside limits. These segments can be identified in the tables shown in Appendix F; these are the segments where the mag- nitude of maintenance is decreasing as the budget decreases. For example, in a scenario where all the ports are included in the model, if the available budget Port Total Average Tonnage (Million Tons, 2006–2010) Tonnage Using Bottom 3 Ft Percent Duluth 40,721 37,479 92.0% Hampton Roads 52,946 11,362 21.5% Plaquemines 61,412 6,429 10.5% Portland-Deep Draft (Portland-Coastal) 24,999 8,981 35.9% Portland-Shallow Draft (Portland-Inland) 10,633 8,249 77.6% Huntington* 71,945 67,267 93.5% *Only the bottom 2 ft was used for Huntington. To remove 3 ft of draft would essentially make it impossible for towboats to operate, and most traffic would shut down. Table 23. Utilization of bottom stratum of channel depth.

36 decreases from 80 percent to 50 percent, then the seg- ment DS1 (Duluth-Superior, see Table 9) improvement drops from 2 ft of dredging to 1 ft of dredging. 2. Not all eligible segments need to be fully maintained— some not at all. The results shown in Appendix F reveal that only rarely must the entire mix of potential projects be accomplished to accommodate expected increases in demand. Furthermore, in many cases, it is not necessary to execute the full maintenance project; only a portion of the project (for example, one unit of improvement, rather than the full potential amount) is required to fully accom- modate demand given system constraints. In no case was it deemed necessary to improve a lock to accommodate expected demand. 3. It is possible for budgets to be set too high. It is also noticeable in the results that many times, even though the system is facing a funding decrease, the results do not change unless the budget decrease becomes significant enough (say 20 percent of the total for the project). This indicates that the budget is far greater than what is needed to accommodate demand. Change occurs when the sys- tem is restricted by high demand or constrained landside capacity, and funding the segment in question would not increase throughput, since there is a landside bottleneck. In other words, improvements are allocated to segments to the degree that they enable the system to use available landside capacity or accommodate demand. Thus, even in situations where there is more money available than needed for the selected slate of improvements, the model does not call for it to be spent on further improvements due to external throughput constraints or the fact that demand is already fully accommodated. However, while the capacity of a lock may be sufficient, it is still impor- tant to be certain that the lock is not in a state of immi- nent failure. While it may not be important to increase a lock’s capacity, it may be of critical importance to prevent its failure. 4. In some cases, it’s all or nothing. In some instances, decreasing the budget by just 10 percent (from 100 per- cent to 90 percent) results in a zero maintenance decision. This occurs in cases where a port has only one water origin- destination corridor. In these cases, when a budget of less than 100 percent of the amount needed for one unit of improvement of that corridor is selected, there are no alternative courses of action. 5. Locks are not a capacity constraint. Neither of the locks (John Day and Emsworth) that represented possible capacity constraints was selected for maintenance under any of the scenarios. While locks may be a cause of con- cern in terms of efficiency, they do not appear to create important bottlenecks in the near future (assuming they continue to function adequately). As mentioned in Item 3, it is important to ensure that even a lock with sufficient capacity is not in a state of imminent failure. The point here is that locks that are operating satisfactorily do not appear to constrain freight flows. Note that the model is artificially constrained and does not incorporate all factors that may have to be considered in a real world situation. For example, shoaling effect is not con- sidered in the model. The model only assumes a static state, which means dredged depth will remain effective for a period of time. Additionally, system reliability often requires mainte- nance of the waterway system to have a capacity higher than needed capacity to hedge against capacity reduction due to weather and other incidents.