By Ted Prince
As originally published in the Journal of Commerce.
October 21, 2015, was “Back to the Future Day.” It was the date in the future that Marty McFly and Doc Brown traveled to in “Back to The Future: Part II” from 1989. As a new year dawns for intermodal, there is a definitive “back to the future” feel as railroads look to maintain their franchises in the face of a “new normal” of less coal and the desire for lower operating ratios.
The year 1989 was also a jumping-off point for intermodal. Earlier in the decade, achieved dramatic growth from international trade. APL and Sea-Land introduced “liner trains” enabling them to substitute rail service from the west coast to east coast ports. The introduction of double-stack several years later super-charged the industry. It was a win-win situation. Although the ocean carriers cannibalized all-water service through the the faster service, accompanied by an explosion in global trade, grew overall volume. Railroads enabled the ocean carriers to cannibalize westbound domestic business so that the lines could balance equipment flows.
The railroads’ focus on domestic intermodal remained about moving trailers — if they were 40 or 45 feet — primarily for intermodal marketing companies, the Post Office, and Motor carriers, such as Schneider National, refused the repeated entreaties of railroads to implement intermodal solutions based on their belief that the advanced truckload model developed after deregulation was the one — and only — answer to meet customer expectations.
That changed, when, in 1989, the Santa Fe Railway Co. (ATSF) and JB Hunt formed Quantum, a joint venture to move 48-foot trailers between the Midwest and California. This was cannibalization too. JB Hunt (the man and the company) realized that his long-haul network was not sustainable. Ironically, part of his concern was driver supply. Rather than wait to have his competitors cannibalize his market, he took the bold step of doing it himself. The ATSF also cannibalized its market, although it is unclear if it truly understood the magnitude and ramifications of what it was giving away.
Since 1989, the introduction of 53-foot double-stack sparked a 25-year explosion of growth that transformed the railroad and intermodal industries. These were the good old days. With coal providing high returns — and cash flows — railroads could invest to grow intermodal. These investments transformed the networks.
In the 1970s, intermodal was handled in an intricate many-to-many network. It was not unusual for an intermodal terminal to serve (and be served by) 25 to 40 percent of the ramps in that railroad. This was supported by complex, enroute switching where cars were “block swapped” among many trains. Although such an arrangement maximized network scope, it created diseconomies of scale.
Over time, railroads built an array of large, greenfield terminals that supported a network of high-volume trains assembled at origin and run intact to destination. These point-to-point trains eliminated the cost and uncertainty of intermediate rehandling. Service reliability improved because trains no longer needed to “wait” for connecting traffic.
Rail networks are driven by economies of scale. As the technology of double-stack and locomotive traction changed, the desired premium train length has increased by almost 30 percent. Given sidings of sufficient length, railroads will now run trains of up to 10,000 feet (and more). However, since there are very few origin-destination pairs that have such daily volume and balance, the result has been “back to the future” operating strategies.
In the absence of point-to-point corridors, railroads implemented other approaches:
- Traffic from a single inbound train is rehandled in a car-to-car transfer terminal, which connects to multiple outbound trains Non-intermodal carload volume may be added to “fill out” an intermodal train (see Figure 1)
- "Hub and Spoke”: Traffic from a single inbound train is rehandled in a car-to-car transfer terminal, which connects to multiple outbound trains (see Figure 2)
- Non-intermodal carload volume may be added to “fill out” an intermodal train. When fill is comprised of reliable, high-service traffic (e.g., automotive multi-levels) premium service can frequently be maintained. However, as the fill varies by corridors, service level, and day-of-the week volumes, the intermodal schedule may be significantly degraded.
Rail networks are generally viewed as the most complex of all asset-based network-operating systems. The train planning described above is further complicated by the characteristics of intermodal terminals. The major challenge occurs when the desired train length is greater than the available ramp track space.
- This requires trains being handled in different pieces (see Figure 3). This poses a scheduling dilemma. Should availability be based on the 75 percent that can be spotted on arrival — which would misstate the last 25 percent; or, should availability be based on the absolute last unloaded unit — which might cause the first 75 percent to delay an achievable delivery for one day.
- Intermodal traffic is not uniform; trailers, containers, and rail cars come in different configurations. The (unloaded) inbound cars might not be in the desired location for (loading) outbound cars.
- The extra rehandling increases switching expense, elapsed time, and service uncertainty.
Figure 4 demonstrates, the need to aggregate volume may increase transit time — delaying delivery by up to three days.
From 1980 to 2010, it is estimated that the number of intermodal terminals shrunk by 90 percent — even though volume grew. As mega-terminals were built, small ramps were frequently closed. This improved railroad linehaul efficiency, although it reduced intermodal competitiveness versus truck. Intermodal competitive position to truck can be quantified by two metrics. In both cases, the lower the number, the more competitive intermodal is with truck.
- Intermodal miles (pickup, delivery, and rail) versus truck door-to-door miles
- (Roundtrip) drayage miles as a percent of intermodal miles
Figures 5 to 8 analyze intermodal versus truck from points in Washington to Bolingbrook, Illinois
- The move from Kent over the Seattle ramp is very competitive to truck because it has very little circuitry (see Figure 5)
- The move from Prescott over the Seattle ramp is not competitive with truck because it is very circuitous (see Figure 6)
- However, the move from Prescott over a theoretical Burbank ramp could be very competitive to truck because it has very little circuitry (see Figure 7)
The theoretical Burbank ramp could represent the future of intermodal. Although intermodal will benefit from electronic logging device enforcement on trucks, intermodal drayage could also be subject to the same enforcement. The obvious solution to this challenge is to create more intermodal terminals to reduce the drayage distance and likelihood of reverse circuitry (see Figure 8).
Terminals built in the 1980s were frequently built “on the cheap.” These “pocket ramps” could be built in the same manner. The network dilution could be minimized by using stub-end trains to serve them to/from a select number of major markets and gateways (e.g., Chicago.)
The challenge is that such an approach may fly in the face of current doctrine of longer trains and fewer ramps. It might even appear as cannibalizing operating margins; however, the past 40 years have proven that the intermodal innovator willing to dismantle an “established” methodology has emerged as the market leader. Perhaps this “Back to the Future” approach would serve the industry well.