Longer Combination Vehicles on Exclusive Truck Lanes: Interstate 90 Corridor Case Study
4.0 Demand and Utilization
Utilization of the new exclusive truck lane facility by LCVs largely depends on the degree to which the carriers or shippers choose to move by LCV instead of their current form of transport – either semi-trailer truck or rail. This section estimates the potential shift of freight to LCVs on the new facility from the current and future populations of semi-trailer truck and rail moves on the I-90 corridor. The outcomes of these estimates provide the basis for toll revenue projections and the public/private benefits and impacts presented in Section 5.0.
Baseline Corridor Combination Truck Volumes
To estimate the potentially divertible trucks to the exclusive truck lanes, existing truck flow and count data were evaluated to develop estimates of baseline (2010) and target year (2040) no-build average daily combination truck traffic (ADTT) on I-90 and I-271. The foundation of the baseline estimates is the FAF2 truck flow data assigned to a network. Following the assignment of the FAF2 network flows, the following three count-based data sets were used to scale the FAF2 flows, where FAF2 volumes are either higher or lower than count-based sources.
To generate 2040 ADTT, the volume-weighted average FAF2 growth rate for the entire corridor is applied to each segment volume. This value is 3.04 percent annually. (Calculated from FHWA’s FAF2 data.) Figure 4.1 shows the results of this baseline 2010 and 2040 estimation of total trucks on the corridor versus other vehicle traffic. In this case, trucks are five-axle combination units and greater. The volumes shown in Figure 4.1 are indicated at locations of key interchanges. It reflects AADT (average annual daily traffic) counts to illustrate the magnitude of total traffic (noncommercial and commercial vehicles).
Source: FAF2, VTRIS, State Counts.
Figure 4.1 shows that the proportion of truck traffic will grow over time from 12 percent in 2010 to 15 percent in 2040. This growth trend highlights the need for consideration of truck productivity measures, such as LCVs.
Estimation of Combination Trucks Diverted to LCVs
Given the study parameters and highway operating assumptions, the primary source of utilization in the exclusive truck lane in this study are the single-trailer trucks (STT) that would divert to LCVs to reap higher productivity gains. This initial estimate of LCVs includes both double- and triple-trailer combination units; the following section provides a specific estimate for triples portion.
In order to estimate the share of STTs diverting to LCVs, the following high and low scenario estimates were developed based on observed truck classification count data, where LCVs are currently allowed to operate:
Once estimated, the high and low ranges of STTs diverted (see Table 4.1) were converted to LCV trips based on a payload factor reflecting the change in truck weight from a 5-Axle STT to a 9-Axle LCV. To accomplish this, a payload factor of 1.60 was utilized to convert 5-Axle STT (80,000lbs) to 9-Axle LCV (127,000lbs). The following table shows the results of the high and low potential for truck-to-LCV diversion in the base and future years. The estimates are expressed as daily truck trips.
As an alternative to this observation-based method, this study also estimated potential LCV diversions using a second set of assumptions. A description of this method, which uses a combination of distance, weight, and commodities as filters, is summarized in Section 6.0 appendices of this report. The alternative scenario diverts a higher percentage of single-trailer trucks to LCVs –roughly 20 percent. The study team chose the more conservative observation-based methodology to estimate truck-to-LCV diversion because the utilization trends of long-distance moves in western LCV states have persisted for decades.
Estimation of Triple-Trailer Combinations
While the majority of single-trailer trucks trips diverted to LCVs on the corridor would use a double-trailer combination, a portion of the LCVs diverted would utilize triple-trailer combination units. Currently, triples are allowed on the Ohio and Indiana portions of the corridor. For example, the Ohio Turnpike Commission stipulates triples to between 90 to 113 feet long, and having anywhere between 7 to 9 axles. The maximum gross weight of a triple should not exceed 115,000 pounds. Accordingly, this estimate assumes that the range of triples would extend from the Illinois/Indiana state line to the Pennsylvania/New York state line.
To calculate the percentage share of triples among all LCVs, vehicle miles traveled (VMT) shares from the U.S. DOT’s Comprehensive Truck Size and Weight Study by state were evaluated. Across the nation, in states where triples are allowed, triples account for anywhere from 2 to 42 percent of the share of total LCV VMT on the states’ highways. In Indiana and Ohio, where LCVs are restricted to the Ohio Turnpike and I-90, the VMT share for triples averages 29.9 percent.
Applying this factor to the estimated population of LCVs on the corridor in 2010 results an average of 132 triple-trailer combinations on the corridor in 2010, assuming the 10-percent diversion of all single-trailer combination units to LCVs. The following table supplements Table 4.1 by illustrating the potential for triple-trailer combinations under high (10 percent) and low (2 percent) diversion scenarios for 2010 and 2040. This estimate does not consider triples on the corridor’s New York Thruway or Massachusetts Turnpike, where triples are not currently allowed. Also, this estimate does not consider the effect of tolls on triple utilization.
The following figure (Figure 4.2) graphically illustrates the magnitude of resulting doubles and triples under the low and high scenarios for the base year.
With the development of a long-distance LCV route, a share of the freight traffic moving over CSXT and Norfolk Southern (NS) rail lines parallel to the corridor would likely shift to LCVs to realize per-mile cost savings. To estimate the number of railcars diverted to LCVs, the study team identified the LCV cost savings per mile and a cross-elasticity to predict change in rail ton-miles because of changes in trip costs. Using these two factors, high and low diversion estimates were developed. The primary difference between the “high” and “low” estimate is that the methodology for the high estimate does not include costs for assembly and total corridor trip costs. Rail ownership was not considered in this analysis, since both NS and CSXT operations and commodities on this corridor would be similarly affected by any addition of ETLs. The following map (Figure 4.3) illustrates the regional extent of the rail network.
Source: FHWA, The Strategic Multimodal Analysis Task 3: Chicago-New York City Corridor Analysis using 2002 HPMS data.
Corridor Rail Tonnage Estimates
To develop baseline and forecast rail tonnage estimates, the study team applied national FAF2 growth rates to detailed commodity flow data for the CSXT and NS operations at Erie, Pennsylvania. The FAF2 data were also used to identify the share of low-density commodities traveling by rail that would be eligible for diversion to LCV (9.5 percent). The appendices of this report contain a more detailed description of the approach and development of rail tonnage estimates and both scenarios (high and low).
High Rail Diversion Estimate
The high rail diversion estimate is based exclusively on cost savings resulting from expanded eligibility for LCV operations and new ETL tolls in the project corridor. The high diversion effect is estimated by identifying the change in per-mile truck operating costs and by applying a cross elasticity.
Table 4.3 presents the results of this analysis. As demonstrated in Table 4.3, the number of LCVs resulting from rail diversion is relatively low compared to diversion from single-trailer trucks.
Low Rail Diversion Estimate
This estimate takes into account cost difference of operating an STT versus an LCV in the corridor. The corridor evaluated is I-90 from its interchange with I-95 just west of Boston to the Illinois State line. The STT corridor trip assumes current per-mile toll rates and operating costs. A lower cross-elasticity is applied to identify the number of LCVs that would result from rail-to-truck diversion. Table 4.3 demonstrates the results of this low rail diversion scenario.
The result of combined traffic estimates of truck and rail moves diverted to LCVs is summarized below in Table 4.5.
To determine the exclusive truck lane usage and revenue estimates, three tiers of toll rates were developed – high, medium, and low – to test against the high and low diversion estimates from truck and rail. The change in cost per truck-mile with and without tolls reflects the following:
LCV usage of the ETLs varies linearly with per-mile toll estimates (i.e., 100 percent of all LCVs diverted from STTs use the ETL at the minimum $0.30 per mile toll, 0 percent would use the ETL at the maximum $1.20 per mile toll). The high estimate uses the 10-percent diversion scenario; the low estimate uses the 2-percent scenario.
All rail trips that divert use the toll lanes. The justification for this assumption is that, with increased corridor congestion expected in 2040, the time savings and cost savings of operating LCVs in a corridor with free-flow travel speeds is the primary factor guiding rail-to-truck diversion in the absence of corridor increases in rail capacity to meet future demand.
High Toll ETL Usage/Revenue Estimate
The high toll estimate assumes a toll rate three times higher than the existing per-mile rates for LCVs in New York, Ohio, and Indiana. Higher tolls will assist in supporting capital, operations, and maintenance expenses for the new infrastructure. However, the benefit of higher per-mile tolls is offset by low cost savings of only $0.13 per mile (out of a total possible savings of $1.13 per mile); therefore resulting in lower overall volumes and revenues. Additional travel time savings are expected on the ETLs, and are not considered as a factor in the diversion estimates.
Assuming a constant $1.00 per mile toll, from 2010 to 2040 this scenario results in a high cumulative revenue estimate in 2008 dollars of $498 million to a low estimate of $116.3 million.
Medium Toll ETL Usage/Revenue Estimate
The medium toll estimate assumes a 75- to 100-percent higher toll rate than existing per-mile rates for LCVs in New York, Ohio, and Indiana. Higher tolls will assist in supporting capital, operations, and maintenance expenses for the new infrastructure. Cost savings are high enough ($0.53 per mile) to result in significant diversion from rail and truck. Additional travel time savings are expected on the ETLs, and are not considered as a factor in the diversion estimates.
Assuming a constant $0.60 per mile toll, from 2010 to 2040 this scenario results in a high cumulative revenue estimate in 2008 dollars of $913.9 million to a low estimate of $195.9 million.
Low Toll ETL Usage/Revenue Estimate
The low toll estimate assumes a toll rate comparable to existing per mile rates for LCVs in New York, Ohio, and Indiana. Cost savings are $0.83 per mile; therefore resulting in significantly higher volumes, which outweigh lower per-mile revenues. Because these tolls are in line with current corridor rates which truckers are accustomed to paying, trips are experiencing additional time and cost benefits from use of the ETL and there is reduced need for assembly and disassembly, it is assumed that 100 percent of LCVs in the corridor will use the ETL. Additional travel time savings are expected on the ETLs, and are not considered as a factor in the diversion estimates.
Assuming a $0.30 per mile toll adjusted for inflation from 2010 to 2040 this scenario results in a high cumulative revenue estimate in 2008 dollars of $687.9 million to a low estimate of $135.3 million.
Table 4.5 summarizes the results of the three toll scenarios by providing the 30-year revenue stream.
Universal Truck Toll Scenario
Because none of the LCV revenue estimates contributes significantly toward the total cost of construction and operation of the exclusive truck lane, this scenario estimates toll revenues if ALL combination units choose to use the facility, but at a discounted rate. This study assumes that a discounted toll rate would attract more STTs than equivalent per-mile tolls on LCVs because the corridor is relatively uncongested, except on some metropolitan Cleveland segments at peak hours. Over time, as levels of delay and congestion increase, the motivation for STTs to utilize the ETL may increase. The following table illustrates the results of this exercise, assuming that combination units would pay 50 percent of the per-mile toll of LCVs.