US refinery closures, stocks smack VLCC rates from Middle East

March 22, 2004
US refinery turnaround for spring maintenance will throw a larger than usual kink into crude oil imports, with throughputs declining toward 14 million b/d. Complicating that eventuality for the world's tanker owners is that US crude oil inventories, in January, were hovering at 270 million bbl.

US refinery turnaround for spring maintenance will throw a larger than usual kink into crude oil imports, with throughputs declining toward 14 million b/d. Complicating that eventuality for the world's tanker owners is that US crude oil inventories, in January, were hovering at 270 million bbl.

The result, said shipping consultant Poten & Partners in an analysis released last month, is that the near-term tanker market will likely weaken.

Once refineries begin to increase throughput towards 15.5 million b/d, however, tanker rates (expressed in worldscale, or WS, terms) will firm "just in time to produce gasoline for all those summer vacationers," said the analysis.

Refinery shutdowns

Poten & Partners's analysis cited a report issued in January by the International Energy Agency that said US refineries' closures for spring maintenance were going to make for "a tight squeeze between facilitating supply and satisfying demand."

IEA reported that US refiners were likely to cut back on capacity utilization by 6%, not the customary 2.6%.

Click here to enlarge image

Fig. 1 shows that refinery throughput has been falling since Spring 2003 and was down to about 14.7 million b/d in early January 2004.

Assuming 14 million b/d as the low point, said the Poten analysis, throughput had another 700,000 b/d to fall. "This would mean a drop of as much as 1.4 million b/d in refinery throughput since" December 2003.

"The worst of all possible worlds for tanker owners would be for this entire falloff in refinery throughput to be compensated with Middle East exports," said the analysis.

Assuming one very large crude carrier (VLCC: >200,000 dwt) can handle 25,000 b/d between the Middle East and the US, then the falloff in vessel demand from peak refinery throughput would equal about 55 vessels, or more than 10% in VLCC demand. (Total world VLCC fleet is about 440 vessels.)

And what about rates?

Poten & Partners said that after peaking in first-half 2003 at WS 145, VLCC rates from the Middle East to the West, including the US, had leveled off at WS 125 by early January 2004.

"Surprisingly," said the shipping consultant, "VLCC rates are not that clearly correlated to US refinery throughput even after allowing for a lag between the load and arrival dates."

Poten & Partners explained to OGJ: "We thought if throughput rates went up, so would demand for crude, hence the rise in need for VLCCs followed by a rise in VLCC rates. Clearly that is not the case in [Fig. 1]. We staggered load and arrival dates to see if the rates caught up, and they did not."

Crude imports

Fig. 2, compiled by Poten & Partners from data collected by the US Energy Information Administration of the Department of Energy, shows a rolling average of US crude imports that has been steadily trending downward for 2 years.

Click here to enlarge image

Average imports in January were down to 9.3 million b/d from the 10 million b/d needed to satisfy expected summer demand.

Fig. 3 shows that US crude inventories had fallen to about 265 million bbl in January, the lowest point during the past 3 years.

Click here to enlarge image

Although refineries can do little to bring these levels back up right now, said Poten & Partners' analysis, "eventually inventories will need to be replenished and may send [tanker] rates back up once refineries resume running at full capacity.

"With the forward price lower than the current price for crude, [Poten & Partners does] not expect inventory levels to significantly increase."

Correction: 2003 Pipeline

Economics table
Table 5 in the 2003 Pipeline Economics Report (Sept. 8, 2003, p. 73) incorrectly included a row for 1,000 hp of compression addition in Tennessee. The column totals, however, excluded the row's data and are therefore correct.