NTSB: Fans pushed smoke toward passengers on Washington subway train

In its most detailed revelations yet about the Jan.

杭州桑拿

12 incident, the National Transportation Safety Board outlined a sequence of missteps in which Metro controllers, 11 miles away from the scene near the L’Enfant Plaza station, activated two sets of giant fans at cross-purposes with devastating consequences.

The two sets of fans, on opposite ends of the train, were both pulling the smoke instead of one set pushing while the other pulled, the NTSB said. As a result, the mass of fumes settled over the stationary train and stayed there.

If the fans had been working in proper coordination, creating brisk air circulation, the smoke would have been expelled from the tunnel, the NTSB said.

The mistake occurred because Metro, unlike some transit agencies, “does not have a means to determine the exact location of a source of smoke in their tunnel network,” the NTSB’s acting director, Christopher Hart, said in letters made public Wednesday.

In the letters, to Metro and other agencies, Hart also said that two of the six fans involved in the failed ventilation effort did not work.

The absence of a procedure for pinpointing the origin of tunnel smoke indicates a dangerous blind spot for the transit agency operating the nation’s second-busiest subway, with 46 underground stations and many miles of subterranean tracks.

Making matters worse Jan. 12, Hart said, “the train ventilation system that draws air from the outside into the cars was not shut off by the train operator.” After the Yellow Line train stopped in the tunnel, the surrounding smoke permeated the six cars largely because it was pulled in by the train’s heating, ventilation and air-conditioning system.

In making Hart’s letters public, the NTSB issued several “urgent recommendations” to Metro for tunnel ventilation. The NTSB said the transit agency should draft detailed written procedures for activating its fans “that take into account the probable source location of smoke and fire, the location of the train” and other factors.

Hart has said that the NTSB is a year away from publishing a final report on the incident. Metro has declined to comment on specifics of the emergency, saying it is barred from doing so by federal regulations while the safety board’s investigation continues.

“We welcome the NTSB recommendations issued today and have begun to address these at Metro,” the transit agency said in a statement Wednesday.

After the smoke crisis, “we conducted systemwide inspections and tests of all tunnel fan shafts and found them to be in good working condition,” Metro said. “We also completed familiarization training for all 39 controllers in our rail control center and we are reviewing additional protocols and training associated with tunnel fan use.”

At 3:15 p.m. Jan. 12, moments after southbound train No. 302 departed the L’Enfant Plaza station, it encountered smoke in the Yellow Line tunnel and stopped, according to the NTSB. Much more smoke was ahead, the result of an electrical malfunction on the tracks about 1,100 feet in front of the train’s lead car.

At 3:16 p.m., Hart said, Metro controllers in Landover reacted to the train operator’s report of smoke by turning on giant fans inside the L’Enfant Plaza station — behind the stationary train in the tunnel. The fans were activated in “exhaust mode,” Hart said, meaning they were sucking massive volumes of air in the direction of the station.

“This action pulled smoke” toward the station from the spot of the electrical meltdown deep in the tunnel, Hart said. As a result, the smoke was also moving in the direction of the stopped train, which was soon enveloped.

Then, at 3:24 p.m., according to Hart, the Landover controllers switched on another set of fans — inside a huge ventilation shaft about 1,100 feet in front of the train, near the source of the smoke. The shaft rises from the tunnel to the street.

But the fans in the shaft also were activated in exhaust mode, Hart said. This meant that the two sets of powerful fans, at both ends of the train, were sucking air in opposite directions, causing the smoke to linger in place, surrounding the train.

At least 200 passengers — most of them choking, many sickened and some growing panicked — waited more than 30 minutes to be evacuated by rescuers. One of the riders, Carol I. Glover, 61, of Alexandria, Virginia, died of smoke inhalation, an autopsy showed.

The critical problem was the absence of air circulation, Hart said.

“We know that using the fans in the station sucked the smoke from the source into the train, which was in between the source and those fans,” Hart told the Metropolitan Washington Council of Governments at a meeting Wednesday.

The smoke did not pass over the train and flow entirely into the station because it began to be pulled from the other direction by the fans in the vent shaft. “You need a push and a pull,” Hart said at the COG meeting.

If used properly, one set of fans would have been activated in “supply mode,” pushing in fresh air, while the other set, running in exhaust mode, pulled smoke from the tunnel, Hart said. However, “because both station and vent shaft fans were all activated in exhaust mode, there was not a supply of fresh air to aid in moving the smoke.” In dealing with situations such as the Jan. 12 incident, Metro’s Landover controllers are guided by a standard operating procedure that “contains a number of key actions that must be taken when a train encounters smoke in a tunnel,” Hart said in his letters. But “this SOP does not address tunnel ventilation strategies.”

“Other transit agencies . . . have developed detailed ventilation procedures for addressing train fires and smoke events in tunnels,” he wrote.

“A common approach . . . is (1) to identify the most likely location of the smoke or fire, (2) to start the ventilation fans on one side of the smoke or fire in supply mode, and (3) to start the ventilation fans on the other side in exhaust mode. This strategy is designed to move smoke away from the passengers and the evacuation route.”

On Capitol Hill, several members of the area’s congressional delegation reacted angrily to the ventilation missteps outlined by Hart. The House Committee on Oversight and Government Reform scheduled a hearing about Metro’s overall safety for Friday.

“Frankly, it is stunning that [Metro] would need NTSB to remind it that the Metrorail system should have a ventilation system in good working order, specific ventilation procedures, and appropriate training for its employees on those procedures,” Rep. Gerald Connolly, D-Va., said in a statement.

Although Metro has declined to comment on what went wrong Jan. 12, the agency’s director of plant maintenance, Randall Grooman, met with Washington Post journalists this week and described in detail how the subway’s ventilation system is designed to work.

Each underground station, including L’Enfant Plaza’s, has two big fans under a platform. In exhaust mode, with the blades spinning counterclockwise, the fans are capable of pulling a total of 60,000 cubic feet of air per minute out of the tunnels. In supply mode, with the blades turning clockwise, they can push in 42,000 cubic feet of air per minute.

Inside the tunnels, between each station, there is at least one large, concrete ventilation shaft rising several stories from the tracks to the street. In some long stretches of tunnel, there are two shafts. The shafts, about 50 feet by 50 feet, double as emergency evacuation routes and have staircases leading from the tunnels to the surface.

At the top of each shaft are ventilation fans, usually two to six, depending on the tunnel’s length and the depth of the shaft. In the Yellow Line tunnel south of L’Enfant Plaza, there is one ventilation shaft between the station and where the tracks emerge from underground at the Yellow Line bridge across the Potomac River.

That shaft is about 1,950 feet beyond the station. There are four fans, each five feet in diameter, at the top of the shaft. Combined, the fans can pull air out of the tunnel at a rate of 200,000 cubic feet per minute in exhaust mode. In supply mode, they can push in 140,000 cubic feet per minute.

Grooman said that in an emergency, with passengers stuck in a smoky tunnel, the procedure would be to first decide in which direction to evacuate them — toward the station or toward the ventilation shaft — to get them away from the smoke. That decision dictates how the fans should be coordinated.

Based on the general procedures Grooman described, the passengers in the Yellow Line incident should have been evacuated toward the station, because the origin of the smoke was in front of the train. And that was how they eventually were taken out.

As the passengers were led back to the station, the shaft fans were operating in exhaust mode, pulling the smoke in the other direction, away from the train. The station fans should have been working in supply mode, pushing the smoke toward the shaft, creating a circulation of air. But they also were set to exhaust.

In his letters, Hart said: “NTSB investigators found during post-accident inspection that two of the four [shaft] fan. . . . were non-operational.”

That faulty setup left the smoke caught between two forces — two fans in the station pulling air at 60,000 cubic feet per minute, and two functioning fans in the shaft pulling air at 100,000 cubic feet per minute.

“So that doesn’t move anything anywhere,” Hart said.