Wait long enough at the roadside pull-off in southeastern Washington’s Wallula Gap, overlooking the nearly mile-wide Columbia River, and you’ll see one: A unit train of Bakken crude heading west. The string of 100-plus tank cars, with locomotives at both ends, stretches out more than a mile, yet from our vantage point it looks like a mere millepede on a landscape dominated by water and wide open spaces.
Knowing that some of these oil trains turn left near Wishram, Wash., and head south through Oregon to reach terminals in California makes you wonder. Could such trains be filled with water, and could that water be delivered to some of the severely drought-stricken areas along the West Coast? That question was posed in a Railway Age blog back in early 2014 (“If Crude by Rail, Why Not Water?”), and in the days that followed, there was much discussion among the Railway Age staff. Water-by-rail (WBR) was considered either a sure-deal or a no-win, depending on whom you asked.
While that debate resurfaces occasionally at various rail-oriented chat sites, WBR has finally reached the mainstream. In a May 2015 story, “California Drought: Can Railroads Come to the Rescue?,” MSNBC and NBC interviewed a BNSF spokesman who said, “We certainly have that capability today.” The story made a couple of questionable points, including the idea of shipping water from the rain-soaked East Coast (a tremendous distance from California compared to other water sources), but it helped bring the concept of WBR to a wider audience.
In August 2014, BNSF Chairman Matt Rose told Railway Age, “We actually have an initiative on water by rail. We have looked at it seriously several times and haven’t quite seen the economics work but it might in the future.” In September 2014, Union Pacific Senior Vice President-Corporate Relations Robert Turner, responding on behalf of UP Chairman and CEO Jack Koraleski, said, “There are multiple variables to be considered regarding shipping water by rail. As with any new shipment option, a thorough assessment is required prior to providing general comment on the logistical feasibility.”
At first glance, WBR seems as economically feasible as crude by rail. The going rate for moving crude from North Dakota to California sits near $14 per barrel, which translates to roughly 32 cents per gallon. (In U.S. measures, one barrel equals 31.5 gallons. One petroleum barrel, however, equals 44 gallons. This article applies petroleum barrels to translate CBR shipping rates into theoretical WBR rates.) Now, consider that WBR from the Columbia River to California would travel roughly half the distance of that North Dakota crude; hence, roughly half the transport cost, one might expect. But then factor in the costs to procure, load, unload, and possibly treat water. Could the net cost be kept on par with the $1.20 (or more) per gallon that Californians pay for water off the store shelf, or the half-penny to one-cent-per-gallon they pay through their utility? If WBR ends up costing more than that, is it a price some customers will be willing to pay? Can federal or state funding help offset the overage?
Those questions and many others need to be considered. Where will the water come from? How much WBR traffic (empties as well as loads) can railroads take on? Can enough tank cars be bought or built? How much water can each car actually carry, and how does water behave while in transit? Would sub-freezing temperatures present challenges during loading or shipment? We’ll examine some of those issues in a moment. First, let’s look where WBR has already made a positive impact.
It’s Not As Outlandish As You Think
The use of railways to transport water holds historic precedence both in the U.S. and abroad. Beginning in May 2000, the Indian Railways were asked to provide emergency WBR service to several drought-stricken regions, with a reported 262 trips totaling 4,535 tank cars (an average 17 cars per trip) delivering water to nine towns and 200 villages. Between September 2002 and April 2003, India’s water shipments increased to 1,277 trips, handling 65,512 tank cars (an average 51 cars per trip). News of such operations has gone virtually unnoticed in the Western world. To see Indian WBR in action, visit Youtube.com and watch this video posted in 2009.
In 2008, Queensland Rail Freight delivered water to the town of Cloncurry (population 2,400) in north-central Queensland, Australia. A regularly scheduled train picked up as many as six carloads of water each day from a reservoir outside the town of Mt. Isa and forwarded them 75 miles to Cloncurry. Australia’s reliance on WBR goes back as far as the late 1800s. Inter-Basin Water Transfer: Case Studies from Australia, United States, Canada, China and India by Fereidoun Ghassemi and Ian White (2007, Cambridge University Press) is a must-read for understanding the geopolitical issues surrounding WBR and water distribution in general. In past years, the authors state, “Attempts were made to ease [Australia’s] water shortages by dispatching water trains. One water train could haul 30 wagons, carrying a total of 163.5m3 [5,774 cubic feet, or 43,192 gallons].” In Struggle Country: The Rural Ideal in Twentieth Century Australia (2005, Monash University Press), author Graeme Davison says, “Only the water trains to small towns like Tempy, Walpeup, and Piangil during drought years ensured survival.” The mining town of Broken Hill in New South Wales required drought-relief water shipments by rail during 1888-92, in 1903, and again during 1951-52.
WBR has also been implemented in Israel, Africa, South America, and elsewhere around the world. As for the U.S., tank cars were used until 1960 for the routine delivery of water from Arrowhead Springs in the foothills of the San Bernardino Mountains to bottling plants in or around Los Angeles. On a more critical and widespread basis, railroads in some parts of the American West shipped water to not only re-stock their trackside supply tanks (most notably during the era of steam locomotives) but also to support isolated communities where railway employees and other residents lived close to the tracks but far from any consistent, naturally-occurring water sources. Among the more famous was Grand Canyon Village, Ariz., whose growing tourist population relied almost entirely on water delivered by thr Santa Fe Railway until the site’s first pipeline was completed in 1932.
And who would have expected drought relief by rail in, of all places, Illinois? According to Water Hauling by Trucks in Illinois, published in 1971 under the Illinois State Water Survey, that state had occasionally been forced to “obtain supplemental water” via both truck and train. The report said, “Mount Vernon had to import drinking water by railroad tank cars in 1905, 1925, and 1945. Starting in January 1945, 100 tank cars were used to ship 4.5 million gallons of treated water by rail from Evansville, Indiana. This operation extended through 45 days and cost more than $50,000.”
Assembling the Water Car Fleet
With most of North America’s 371,000 tank cars already dedicated to crude, chemicals and other commodities, and orders for new cars running a year or more behind schedule, is it even possible to field an effective number of unit trains for water? Right after Railway Age posted “If Crude by Rail, Why Not Water?,” Editor-in-Chief William C. Vantuono suggested this: “The 165,00-plus DOT-111 tank cars that do not meet the latest voluntary AAR safety standards (CPC-1232) for hazmat and that could be forced out of service by a pending PHMSA rulemaking could be purchased by the feds and re-purposed into water cars (i.e. clean them up, install new linings, new fittings, etc.). That would help solve the rail industry problem of what to do with all these suddenly obsolete but serviceable tank cars that the feds say we can’t use for the purpose for which they were originally built.”
Vantuono ran the idea past Robert Pickel, Senior Vice President-Marketing and Sales at National Steel Car. “Blasting a car can cost $1,500 to $2,000 per car,” Pickel said. “Lining, depending on the material, would likely cost around $5,000 to $6,000 per car.” That’s compared to the $20,000 to $40,000 estimated cost of refitting a DOT-111 tank car for continued crude service.
On May 1, 2015, the same day that Canada and the U.S. announced new design requirements for crude-carrying tank cars, Pickel told Railway Age, “We may have 100,000 plus cars available for such service [water],” his number reflecting the DOT-111s, jacketed or non-jacketed, that will drop out of crude service if they’re not retrofitted. Among those DOT-111s currently moving hazmat, Pickel says there are 80,000 “legacy” cars built before 2011, which will be obsolete for crude service by 2017.
DOT-111s and other tank cars used in CBR service are mostly in the 30,000 to 32,000-gallon-capacity range. But with water being heavier than crude, tank cars used in WBR service could only be filled to around 20,000 to 22,000 gallons so as not to exceed gross axle load limits. Ideally, this would mean using tank cars in that medium size range for WBR. But if thousands of DOT-111s became more readily available, could they be used, safely, if filled to only two-thirds of their capacity with water? Pickel says, “Yes, but sloshing could be problematic. Baffles might work as they are used in other transport modes. Partitions are another possibility. However, train handling and operating dynamics would need to be modeled to determine the feasibility of unit train service.”
Logical Sourcing and Routing for WBR
TV and movie star William Shatner re-ignited the war over Western water with his recent suggestion for a multi-billion-dollar pipeline that would move water from the Pacific Northwest to California. The media mocked the idea, but Shatner’s overland pipeline, which the actor envisioned as similar to the 800-mile Alaskan Pipeline, seems no more audacious than other projects that have been proposed or even built. The U.S. Bureau of Reclamation initiated studies in 1971 for an undersea pipeline from the Columbia and other Northwest rivers southward to California. In 1990, Los Angeles County Supervisor Kenneth Hahn submitted a motion to build two aqueducts, one channeling Columbia River water down to the Shasta Lake reservoir in northern California, the other carrying Snake River water from Idaho south to the Colorado River. In 1991, Alaska Governor Walter Hickel proposed a pipeline linking his state with northern California.
Space does not allow a thorough explanation of Western U.S. water policy here, but with regard to WBR, several key points deserve mention. The Pacific Northwest currently holds the most abundant supply of free-flowing surface water in proximity to California. Using rudimentary pumps and hoses, unit trains could be loaded at any number of sidings adjacent to major rivers as far inland as Montana. But to shorten the haul and minimize the impact on often-congested terminals such as Hauser, Idaho; Yardley (Spokane) and Pasco, Wash.; and Hinkle, Ore.; loading sites would be best confined westward and downstream, where the Columbia River lies between south central Washington and north central Oregon. BNSF and UP have main lines on opposite sides of the river here, and while both are working to alleviate existing congestion in the Columbia River Gorge, there are sidings and bulk commodity loop tracks along the river capable of holding unit trains. From the Columbia River southward, BNSF and UP have routes that already handle unit crude into California, so there should be nothing to prevent unit tank trains of water from traveling in the same direction.
Despite the federal government’s role in managing dams, hydroelectric power, navigation and salmon migration on the Columbia River, Northwest states have some say in the use of its water. Governors of Washington and Oregon have so far turned down every attempt to export Columbia River water beyond their collective borders. But in 2007, Oregon Senator David Nelson submitted an alternative view. In a paper titled “Columbia River Diversion as a Public Revenue Source,” Nelson cited studies showing that seasonally timed extraction of Columbia River water would pose no risk to salmon migration, and he argued how water could, and should, be exported like any other natural resource—comparing it with timber, oil and gas shipped from other states—in order to help fund Oregon’s public services. The senator concluded with words that hold serious implication for all water-rich states. “Oregon can either engage the federal government now on its own terms and divert water under amicable and highly profitable circumstances, or it can run the risk of fighting off a water grab by its much more powerful neighbors using a constitutionally suspect law. The federal government will facilitate either way, but ultimately diversion from the Columbia River will probably happen.”
The enormous quantity of unused water the Columbia River releases into the ocean was described by L.A. County official Kenneth Hahn in 1990 as “… 3.7 billion gallons an hour, 61 million gallons a minute, and 1 million gallons a second.” The Washington Department of Ecology says the Columbia’s average annual discharge into the Pacific can actually be as much as 116 million gallons per minute during high-water years. Take a unit train of 100 cars filled with 20,000 gallons of water per car, and WBR would draw roughly two million gallons from the Columbia each trip. Multiply that to the tune of eight trains per day (four on BNSF, four on UP), and it still means WBR would extract only a tiny fraction of the river’s total outflow.
Other regions might be better positioned to supply water to southernmost California. Railway Age has spoken with a company located in the Southern U.S. who says its facility is being readied to load unit water trains, with interested buyers already knocking on its door. This company expects to announce details of the operation later this summer.
Who Pays for WBR? Can it Pay for Itself?
Water-by-rail could never move enough trains on enough tracks to supply all of California’s needs. Crude-by-rail currently delivers only 10% or less of the oil that’s processed in California, yet those rail shipments are considered vital and cost-effective to the refining industry. What makes WBR different is that it can reach parts of California where there are currently no other comparable means for delivering substantial volumes of water at all.
One way to make WBR work as efficiently as CBR would be to send water trains to the same places crude trains are headed. From Bakersfield to Long Beach to the Bay Area, terminals that receive unit crude could also receive unit water, and with a bit of ingenuity and investment, those water trains could be off-loaded into municipal systems or holding tanks. Valero, Shell, Tesoro and other refiners that have faced public opposition in their efforts to expand CBR business in California might win some favors if they accommodated the delivery of one unit water train to their terminals for every one or two crude trains that arrive there.
Agriculture accounts for 80% of California’s water consumption. According to the U.S. EPA, “Of the ten most productive agricultural counties in the U.S., nine are in California, and the San Joaquin Valley is the single-richest agricultural region in the world.” Farms could utilize WBR in places where branch lines or sidings provide acceptable parking space for tank cars that could feed directly into irrigation systems. Would the federal government subsidize the cost of WBR for farming in order to maintain inventory and affordability at the grocery store, or will that food production instead be shifted to states possessing adequate water?
WBR’s start-up costs would be comparatively minimal. As was the case with CBR, most of the components necessary for WBR (track, cars, motive power) already exist. Meanwhile, the new billion-dollar Poseidon Water desalination plant in Carlsbad, Calif., is scheduled to begin operating in November. It will provide the San Diego County Water Authority with about 54 million gallons of water per day, which covers less than 10% of that county’s average water consumption. Desalinated water delivered to San Diego County will reportedly work out to just over one cent per gallon to residential users. WBR, if priced anything like CBR, could cost anywhere from 10 to 32 cents per gallon, or more.
To match the Poseidon plant’s output would require 2,700 tank cars (if loaded to 20,000 gallons each) rolling into San Diego County every day. That’s at least 27 loaded trains arriving, plus empties departing, every day, on a corridor that handles two-dozen intercity passenger trains, every day, plus another 50 or more commuter trains Monday-Friday. The delivery cost alone for WBR would still be higher than the wholesale cost of desalinated water, based on today’s CBR rates. Desalination, however, presents its own costs in terms of energy consumption and environmental impact. Still, WBR would likely find its niche among California’s smaller inland communities rather than major cities located on the coast.
It could come down to WBR being delivered to communities or businesses for which price is not the issue, or government agencies stepping in to fund water shipments to those who otherwise can’t pay for it themselves. Where lies the tipping point that separates a profit-driven operation from a program of humanitarian relief? FEMA and the Red Cross don’t typically charge citizens for water, food and other aid being handed out following a natural disaster. The extent of California’s drought is, of course, much broader than that.
If California is indeed entering a long-term megadrought, similar to those that struck the region several times over the past thousand years or more, wholly new sources of imported water could be the only way of preventing a mass migration, as well as a major loss to America’s food supply. Pipelines and aqueducts would ultimately be the movers of greatest volume. Until those conveyances are built, railways could be delivering appreciable amounts of water to at least a few of California’s hardest-hit areas. Conventional metrics suggest that water-by-rail may simply cost too much. But the Western drought has thrown convention out the window. The millions who live in California, and the millions more who eat or earn income from its productivity, might soon be asking whether we can afford not to put WBR into action.