As crews labored yesterday to stanch the flow of water so they could patch the break, investigators linked the train derailment Wednesday under Howard Street with the rupture of a 40-inch water main above it on Lombard Street.
Because the main sits atop the rail tunnel, it was suspected from the start that the broken pipe and rail fire were related.
"It's looking more and more like it's the fire," said Robert H. Murrow, spokesman for the Department of Public Works.
The broken main - normally a mundane affair - added to the sense of chaos created by the crippling freight train fire in the Howard Street Tunnel about 30 feet underground.
The sudden gush of water at Howard and Lombard streets - through a 6-by-8-foot hole - made a horrific traffic jam even worse, carved earthquake-like fissures into the pavement and disrupted water service to some businesses.
The area was ground zero in a disaster that has become national news: Late yesterday, as repair crews hovered over a gaping 4-foot-deep hole at the intersection, other workers about 70 feet away pumped corrosive hydrochloric acid from one of the rail cars to waiting tanker trucks.
Crews worked into the night to stop the flow of water into the main, a major east-west conduit in the city's 3,200-mile network of pipes.
More than 24 hours after the break, crews continued to work last night to shut the nearly three dozen valves that feed the main with water that originates at Gunpowder Falls in Baltimore County. Some valves require up to 300 turns to close.
The last open valve, at Paca and Lombard streets, proved stubborn because of its age and condition. By late last night, crews had dug a 12-foot-deep hole to reach and replace the valve. They were expected to install a new one this morning.
Once the water is out of their way, crews can install a patch, or sleeve, made of ductile iron that officials say is stronger than cast iron and should last another 75 years.
The repairs will not happen overnight.
"Not knowing the extent of the problem, it may take several days," said Public Works Director George L. Winfield, adding that workers again would be on the scene through the night.
Winfield said there might have been damage elsewhere along the pipe, which cuts across downtown from Martin Luther King Jr. Boulevard to South Street.
The breach in the main occurred about 6:15 p.m., roughly three hours after CSX officials say the train derailed. Public works officials let the water flow unimpeded until about 9 p.m., some of it flowing into the tunnel and helping firefighters who could not get close to the burning rail cars.
As water raced down Pratt Street on Wednesday, an entrepreneurial spirit flowed as well. Stranded Orioles fans paid panhandlers up to a few dollars apiece to climb on their backs so they could cross the street without getting wet.
Reliable information was hard to find at times yesterday.
At one point, officials at the University of Maryland Medical Center heard that repairs could disturb the hospital's water supply. Officials assured the hospital nothing would affect its operations without advance notice, said Ellen Beth Levitt, spokeswoman for the hospital.
On Wednesday, the hospital had brown water and a temporary loss in water pressure, as did homes as far south as Otterbein and Federal Hill. The Holiday Inn at Howard and Lombard had similar problems, but they were cleared up yesterday.
Water samples from the area yesterday indicated it was safe to drink, said Edward Huff, the city's water system manager.
One theory for the rupture considered by public works officials was that the fire heated the pipe to the point that water inside boiled and pressure caused the pipe to explode.
But Amy L. Rechenmacher, an assistant professor of civil engineering at the Johns Hopkins University, called that "unlikely" in part because fast-moving water would carry away the heat.
Other theories held that vibration from the crash was the culprit, or that heat caused movement in the ground and a concrete slab thought to cradle the pipe.
Another theory - that the heat caused the pipe to expand and crack - sounded more plausible to Rechenmacher, who stressed that she knew little about specifics of the break.
"If the [area] was heated enough," she said, "it could have caused the iron to expand" and burst.



