For the first time since the start of the crude-by-rail (CBR) boom a few years ago, just as much crude oil is being transported by rail to PADD 5—that is, to states in the western U.S.—as to the Eastern Seaboard states in PADD 1. This primarily reflects the facts that 1) CBR deliveries from the Williston Basin/Bakken to PADD 1 continue to plummet and 2) refineries in the West remain reliable buyers of railed-in crude from the Bakken and Western Canada. Will CBR shipments to the East Coast continue to fall, or have we seen the worst of the decline? Today we take a look at recent trends in crude movements by tank car, and a look ahead.
As we’ve discussed often in the RBN blogosphere, the volumes of U.S. and Western Canadian crude oil moving out of production areas in tank cars via railroads rose sharply in 2011-12, maintained high levels through 2013-14, and declined through most of 2015 and year-to-date 2016. There are a number of reasons for both the rise and fall of crude-by-rail (CBR). The rise was spurred in large part by the lack of sufficient pipeline infrastructure, primarily out of the Williston Basin/Bakken, and to some extent in the Permian Basin, the Denver-Julesburg and other tight-oil and shale plays where crude production was soaring. Building rail-loading terminals represented a logical, near-term fix—they could be constructed quickly and at relatively modest cost (filling a transportation-capacity gap until pipelines were developed), and using the rails gave shippers destination flexibility (allowing oil to be moved to wherever the netbacks were highest). As we said in our Slow Train Coming Drill Down report on CBR a while back, railed shipments of crude within the U.S. averaged only 55 Mb/d in 2010 and 121 Mb/d in 2011, but rose to an average of 394 Mb/d in 2012, 709 Mb/d in 2013, and 867 Mb/d in 2014 before falling back to 754 Mb/d in 2015. U.S. CBR averaged only 439 Mb/d (on average) in the first six months of 2016 and had declined to only 363 Mb/d by June 2016.
From the beginning, the Bakken was the big dog in the CBR world. Bakken production started to move higher in the early years of the Shale Revolution, really took off in late 2011, and in that year output outstripped available pipeline capacity, much of which had to be shared with crude heading south/southeast from Western Canada. The result was pipeline congestion and significant price discounting while producers scrambled to develop alternative routes to market. That spurred the development of a number of rail-loading terminals that enabled producers, shippers and crude-consuming refineries to move increasing volumes out of the Bakken in tank cars while proposals for new pipeline capacity worked their way through the regulatory, right-of-way acquisition, and construction process. As production rose sharply, CBR’s share of Bakken barrels (purple line/left axis in Figure 1) rocketed from less than 20% in 2011 to well over 60% through most of 2013-14—peaking at more than 75% in April 2013, according to the North Dakota Pipeline Authority (NDPA). In the early years most of the volumes went to the crude hub at Cushing, OK and markets along the Gulf Coast––Petroleum Administration for Defense District (PADD) 3. But the market evolved rapidly, with most rail movements eventually headed for either the East Coast (PADD 1) or the West Coast (PADD 5). The access to markets that CBR provided, along with new pipelines from Cushing down to the Gulf Coast, helped reduced the price differential (blue dotted line/right axis) between West Texas Intermediate (WTI, the U.S. light-oil benchmark) and Brent (the international standard), but the spread remained sufficient to allow Bakken producers to earn reasonable returns on crude railed to distant buyers, even after factoring in the cost of shipping by tank car.