Case Study III

Beam: 54'

Depth: Buffalo 292: 12', Buffalo 286: 11'-6''

Full Load Draft: Buffalo 292: 10', Buffalo 286: 9'-9''

Capacity: Buffalo 292: July 1968, Buffalo 286: Dec 1969

Shipyard: Nashville Bridge Co.

Class: ABS

Framing Convention: Longitudinal

Construction date: Buffalo 292: 12', Buffalo 286: 11'-6''

Vessel Type: Inland Tank Barge

Vessel Description:

Both barges are of similar construction and arrangements: Flush deck, single skin, 6 tanks port and starboard,voids at bow and stern. Design plate thicknesses for BUFFALO 292 were; 7/16"bottom, 3/8" sides, and 5/16"deck. Scantlings for BUFFALO 286 are not available, but are presumed similar to BUFFALO 292.

Back to Quick Link Index

On March 18, 1996, the barge was being towed outbound in the Houston Ship Channel to be offloaded in Port Arthur, Texas. Winds were blowing at 35 knots and waves were 2 to 3 feet high. As the vessel was transiting, the bow and stern began rising significantly and rose to form an angle of about 30 degrees with the horizontal position. The deck had buckled about one foot aft of the transverse bulkhead separating the #2 and #3 cargo tanks. The hull was breached and nearly 5000 barrels of oil were spilled into the channel.

In May of 1996, the barge was being towed outbound in the Houston Ship Channel when the same type of failure occurred as exhibited by the BUFFALO 292. The weather and wave conditions were very mild, compared to that of the 292, when this casualty occurred.

Back to Quick Link Index

Events Leading to Failure

The barge began loading March 17, 1996 and completed loading March 18, 1996. Tanks 2, 3, 4, and 5 port and starboard were fully loaded with Grade 380 oil. The voids, 1 and 6 port and starboard tanks were kept empty. The loading configuration placed the barge in a sagging condition.

There were no loading/unloading procedures established for this barge, which is common practice for the inland barge fleet.

After the barge was loaded, a towboat took the barge in tow for the transit to Port Arthur, Texas. About 3 hours into the voyage, the towboat captain felt a lurch and saw the bow and stern of the vessel rising out of the water. The captain was eventually able to ground the barge on a soft sandy bottom. Oil had been spilling out of the #3 cargo tanks since the failure and eventually most of their contents of nearly 5000 barrels spilled into the Houston Ship Channel.

Factors Relative to the Failure

The barge was designed to meet the ABS Rules for Building and Classing Steel Vessels for Service on Rivers and Intracoastal Waterways. The main deck was supported by intermittently welded, serrated, longitudinal stiffeners. Only 3 inches out of every 12 inches was welded to the main deck.

BUFFALO 286

The specific events for this failure cannot be found, but as the barge was built in the same yard, and to the same ABS Rules as BUFFALO 292, it is presumed that similar design and construction methods were employed.

Back to Quick Link Index

The deck buckled about one foot aft of the transverse watertight bulkhead separating the #2 and #3 cargo tanks. The buckle extended across the entire breadth of the barge and the hull was severely torn where the deck met the side shell. The side shell on both sides and the centerline bulkhead were buckled in the upper half. The bottom plate appeared largely unaffected. The structure in the #2 cargo tanks exhibited no apparent damage. Gaugings of the deck plating showed an average of only 10% wastage.

The barge also buckled across the #3 port and starboard cargo tanks.

Various methods for predicting the compressive strength of a longitudinally framed deck were researched. The method presented by Owen F. Hughes in Ship Structural Design (SNAME, 1988) was adopted by the MSC to compute the ultimate strength of inland tank barges. This method has been verified through testing and was adopted by Lloyd's Register of Shipping. It is also used in MAESTRO, a software package developed specifically for ship structures.

Using this method, the ultimate strength was computed for these tank barges. Assuming "as built" thicknesses, a compressive transverse stress of 1 ksi in the deck, and a stiffened panel deflection of 1" (based upon testimony of a pre-existing upward deflection), the ultimate strength for the 292 was 13.6 ksi (in compression), less than 50% of the required material yield strength. The still water bending stress in the BUFFALO 292 under the cargo load at the time was 13.06 ksi. These results support the imminent failure of the barge and the barge actually failed at the location of the pre-existing deflection.

The ultimate strength for the BUFFALO 286 was calculated to be 18.9 ksi. The still water bending stress was calculated to be 15.88 ksi, which is slightly less than the ultimate. The Naval Surface Warfare Center, Carderock Division (NSWCCD) conducted a finite element study, Reference [2], and identified factors which may reduce the calculated value for ultimate strength. The effectiveness of the serrated stiffener and intermittent welds was questioned. A study, Reference [3], was conducted at the U.S. Naval Academy in order to determine the effect that the intermittently welded serrated longitudinal deck stiffeners had on the ultimate compressive strength of the barge deck. Results from the study showed that this construction and design methods can reduce the ultimate compressive strength by approximately 11 percent.

NSWCCD also suggested that inland barges may be experiencing "progressive damage," where the vessel is loaded and unloaded in such a way that the bending stresses exceed the elastic limit and permanent deflections are created while degrading the strength.

The U.S. Coast Guard, in partnership with the Towing Safety Advisory Committee (TSAC), created Navigation and Vessel Inspection Circular (NVIC) 1-98.

This NVIC presents a step-by-step methodology for calculating the ultimate strength of barges. This value can then be compared to the maximum stresses obtained during loading and unloading practices to identify at risk barges.

Back to Quick Link Index

References:
[1] Cameron, J., Nadeau, J., LoSciuto, J., "Ultimate Strength Analysis of Inland Tank Barges," USCG Marine Safety Center Report, dated June 16, 1997, NTIS Pub. No. PB98-126600INZ.
Download the report in PDF format at: http://www.uscg.mil/hq/msc/article/buckle.pdf

[2] Hess III, Paul E., Adamchak, John C., Falls, Jaye, 1997, "Failure Analysis of an Inland Waterway, Oil Bunker Barge Collapse," Naval Surface Warfare Center Carderock Division Report No. U-SSM-65-97/01.

[3] Miller, M., Nadeau, J., White, G. "Longitudinally Stiffened Panels - A Comparative Analysis of the Compressive Strength of Three Common Construction Methods," 1999 SNAME Transactions.

Author:
Lieutenant Matthew Miller, USCG Marine Safety Center, graduated from the Coast Guard Academy in 1993 with a B.S. in Naval Architecture and Marine Engineering. He spent two years as the Damage Control Assistant on the cutters VENTUROUS and STEADFAST. He graduated from the University of California, Berkeley in 1996 with a M.E. degree in Naval Architecture. He is currently working as a naval architect and salvage engineer at the Marine Safety Center.

He is the author of "Longitudinally Stiffened Panels - A Comparative Analysis of the Compressive Strength of Three Common Construction Methods", Reference [3], presented at the 1999 SNAME annual meeting.

Back to Quick Link Index