Metro calamity to face scru­tiny

NTSB public hear­ings will dis­sect deadly smoke episode at L’En­fant Plaza

The Washington Post Sunday - - METRO - BY PAUL DUG­GAN AND MARY PAT FLA­HERTY

On that rain-soaked win­ter Mon­day, shortly be­fore the start of the evening com­muter rush, Metro’s Yel­low Line train No. 302 rolled out of the L’En­fant Plaza sta­tion and en­tered a south­bound tun­nel to­ward Vir­ginia. It was Jan. 12, about 3:15 p.m., and dis­as­ter soon struck, a fa­tal calamity that has yet to be fully ex­plained.

The six-car train, with about 250 pas­sen­gers, stopped in the dark­ness af­ter en­coun­ter­ing heavy smoke. Its lead car was 836 feet into the tun­nel, fed­eral safety in­ves­ti­ga­tors later said, and the rear of the train was 386 feet be­yond the sta­tion’s plat­form.

This week, as the Na­tional Trans­porta­tion Safety Board con­ducts its first public hear­ings on the in­ci­dent, it will be pointed out that as smoke per­me­ated the six cars — sick­en­ing at least 86 riders, one of whom died — the train stayed put in the tun­nel, never budg­ing from the nox­ious haze de­spite the op­er­a­tor’s ap­par­ent at­tempts to back up.

Why train No. 302 re­mained sta­tion­ary while pas­sen­gers gasped for air is one of many ques­tions lin­ger­ing since Jan. 12. Now, five months into an NTSB in­ves­ti­ga­tion that is a long way from be­ing fin­ished, some an­swers could emerge when the safety board of­fers the public its first close look at what the in­quiry has found so far.

The hear­ings — set to take place 9 a.m. to 5 p.m. Tues­day and Wed­nes­day at the NTSB’s head­quar­ters in South­west Washington — will in­clude pre­sen­ta­tions by safety in­ves­ti­ga­tors and dis­cus­sions about ev­i­dence gath­ered in the tun­nel and else­where. Vo­lu­mi­nous records, in­clud­ing tech­ni­cal re­ports and in­ter­view tran­scripts, are due to be re­leased by the board.

The NTSB in­quiry is among sev­eral re­views of the be­lea­guered transit au­thor­ity by out­side agen­cies and ex­perts since the deadly calamity. The Fed­eral Transit Ad­min­is­tra­tion on Wed­nes­day cited ma­jor flaws in Metro’s safety-man­age­ment pro­grams, es­pe­cially in­volv­ing the Lan­dover fa­cil­ity where train traf­fic is mon­i­tored in real time.

Other re­ports — in­clud­ing one by the Gov­ern­ment Ac­count­abil­ity Of­fice, re­quested by mem­bers of Congress — are ex­pected to be pub­lished in com­ing weeks.

The Jan. 12 cri­sis, it seems, re­sulted from a cas­cade of break­downs and fail­ures en­tail­ing not just equip­ment and in­fra­struc­ture. Metro per­son­nel re­spon­si­ble for the up­keep and op­er­a­tion of the na­tion’s sec­ond-busiest sub­way also are im­pli­cated.

Ex­cept for oc­ca­sional bul­letins re­gard­ing ur­gent safety prob­lems dis­cov­ered dur­ing the in­ves­ti­ga­tion, the NTSB has been tightlipped about the in­ci­dent, say­ing its fi­nal re­port will be is­sued early next year. Metro has been cir­cum­spect, as well, cit­ing fed­eral rules that bar the transit au­thor­ity from dis­clos­ing de­tails of an on­go­ing NTSB in­quiry.

For two days, how­ever, the si­lence will be bro­ken, at least to a de­gree. Here are some of the still-murky is­sues that could come into clearer fo­cus:

Why did the smoke build up?

Gen­er­ally speak­ing, there was “an elec­tri­cal arc­ing event,” the NTSB said in a safety bul­letin, mean­ing that elec­tri­cal cur­rent es­caped from heav­ily in­su­lated ca­bles along the track bed and be­gan flow­ing dan­ger­ously around the tun­nel.

The arc­ing, which gen­er­ated tremen­dous heat, oc­curred about 1,100 feet in front of where the lead car of train No. 302 came to a halt. The re­sult­ing ther­mal dam­age to tun­nel in­fra­struc­ture cre­ated a huge amount of smoke, ac­cord­ing to in­ves­ti­ga­tors.

But how did the elec­tri­cal cur­rent get loose? Although “ad­di­tional anal­y­sis is needed be­fore a cause can be de­ter­mined,” the safety board said re­cently, an NTSB doc­u­ment re­leased June 8 con­tained a pos­si­ble ex­pla­na­tion. It cited Metro’s fail­ure to en­sure that all power-ca­ble con­nec­tions in the sub­way are pro­tected from con­tam­i­nants.

The con­tam­i­nants (wa­ter, me­tal dust and other sub­stances) can seep into a ca­ble con­nec­tion and form a path­way for elec­tric­ity to flow out. Whether the hear­ings will shed more light on the is­sue of im­prop­erly con­structed ca­ble con­nec­tions, along with other po­ten­tial rea­sons for the arc­ing, re­mains to be seen.

“The smoke in the tun­nel,” the NTSB doc­u­ment said, “was gen­er­ated by ther­mal dam­age to about 16 feet of elec­tri­cal power ca­bles and in­su­la­tion, por­tions of 4 fiber­glass ca­ble con­nec­tor cov­ers” and about five feet of fiber­glass that was cov­er­ing the elec­tri­fied third rail, from which draw propul­sion power.

The hear­ings could pro­duce de­tails about the chem­i­cal makeup of the smoke from the smol­der­ing in­su­la­tion. That’s a ques­tion of par­tic­u­lar in­ter­est to pas­sen­gers who choked on the fumes for more than 30 min­utes, some of them cling­ing to con­scious­ness, as they waited for fire­fight­ers to reach the train and be­gin an evac­u­a­tion.

A few of the riders left the cars and reached safety on their own. The rest in­haled an ar­ray of gases. Of­fi­cials haven’t spec­i­fied what the gases were. But an in­di­ca­tion might be found in the au­topsy re­port on the de­ceased pas­sen­ger, Carol I. Glover, 61, of Alexandria, who suf­fered acute res­pi­ra­tory fail­ure due to smoke ex­po­sure, ac­cord­ing to the D.C. med­i­cal ex­am­iner’s of­fice.

It’s not known, how­ever, whether de­tails of the post­mortem re­port will be in­cluded in the in­for­ma­tion made public this week.

Not long af­ter the in­ci­dent, Metro an­nounced a plan to re­place many sub­way ca­bles with new, “low smoke/low halo­gen” ca­bles that are de­signed to re­duce emis­sions of ir­ri­tat­ing, even poi­sonous fumes dur­ing an elec­tri­cal melt­down.

Why did the train en­ter the tun­nel?

In what ap­par­ently was the first in­di­ca­tion of a prob­lem the near the L’En­fant Plaza sta­tion that day, ac­cord­ing to the NTSB, a tun­nel smoke de­tec­tor ac­ti­vated at 3:04 p.m. — about 10 min­utes be­fore train No. 302 pulled out of the sta­tion, headed to­ward the mass of fumes.

“Recorded data shows that at about 3:06 p.m., an elec­tri­cal breaker at one end of a sec­tion of third rail tripped,” an NTSB safety bul­letin later said.

That was another sign of trou­ble.

Metro’s sub­way oper­a­tions are mon­i­tored in real time by train con­trollers in the Rail Oper­a­tions Con­trol Cen­ter, called the ROCC, in Lan­dover. Ideally, ex­perts said, con­trollers at the ROCC would have re­acted to the warn­ings, es­pe­cially the smoke-de­tec­tor alert, by im­me­di­ately halt­ing train traf­fic into the tun­nel. But they didn’t.

A re­view of transit au­thor­ity records by The Washington Post found that months be­fore the smoke in­ci­dent, Metro knew that train con­trollers were poorly served by com­puter soft­ware that dates to 2002. The ROCC’s com­put­ers gen­er­ate so many need­less alarms, about smoke and other is­sues, that the warn­ings of­ten go un­heeded.

Although the agency has long planned to buy and in­stall new soft­ware, the pro­cure­ment process, still in the pa­per­work stage, has moved at a gla­cial pace.

In its re­port on Metro’s safe­t­y­man­age­ment sys­tems, the Fed­eral Transit Ad­min­is­tra­tion also cited the in­ad­e­quate soft­ware, along with many other prob­lems at the oper­a­tions cen­ter, in­clud­ing un­der­staffing, in­suf­fi­cient train­ing, poor com­mu­ni­ca­tions among train con­trollers and un­nec­es­sary noise and dis­trac­tions.

But the FTA did not specif­i­cally look at the events of Jan. 12. Part of the NTSB in­quiry is fo- cused on how the con­trol cen­ter’s short­com­ings fac­tored into the L’En­fant Plaza cri­sis by al­low­ing train No. 302 to pro­ceed into the smoke-filled tun­nel. And some of what the safety board has learned could be re­vealed in the public hear­ings.

Why didn’t the train back up?

The ma­neu­ver seems sim­ple. On Feb. 11, a month af­ter the L’En­fant Plaza calamity, a sim­i­lar but less se­vere in­ci­dent oc­curred in Vir­ginia, start­ing at 1:29 p.m. An Or­ange Line train “trav­el­ing from the Court House sta­tion to the Ross­lyn sta­tion re­ported smoke in the tun­nel as it ap­proached Ross­lyn sta­tion,” ac­cord­ing to an NTSB safety bul­letin. The train driver no­ti­fied the con­trol cen­ter.

A Metro train has an op­er­a­tor’s booth at ei­ther end. “The train op­er­a­tor was in­structed to re­verse ends and move the train back to Court House sta­tion, which did oc­cur,” the safety board said. “All smoke was cleared” by 2:50 p.m. “and nor­mal train ser­vice re­sumed.”

By con­trast, the Jan. 12 in­ci­dent was chaotic. In­ves­ti­ga­tors said they have gath­ered all avail­able records of com­mu­ni­ca­tions be­tween the ROCC and train No. 302 dur­ing the cri­sis. But what those records show has yet to be made public.

With the train stopped in the tun­nel, en­veloped by fumes that were rapidly fill­ing the six cars, sev­eral pas­sen­gers later said, they heard the op­er­a­tor speak­ing into his ra­dio, re­peat­edly ask­ing for per­mis­sion to re­turn to L’En­fant Plaza. Over the in­ter­com, the op­er­a­tor urged riders to be calm, say­ing the train would be back­ing out of the smoke.

It’s un­clear whether a power prob­lem kept the train in place, the con­trol cen­ter didn’t au­tho­rize a re­turn to L’En­fant Plaza or some other com­pli­ca­tion arose. Pas­sen­gers said the op­er­a­tor walked from the front of the train to­ward the rear booth. And they said the train lurched a few times, as if it were about to start rolling.

But it stayed put, as gasp­ing riders, some on the verge of panic, shielded their eyes and mouths with nap­kins and scarfs. Within 10 min­utes — be­cause Metro had failed to halt rail traf­fic in the vicin­ity of the L’En­fant Plaza sta­tion — back­ing up ceased to be an op­tion for the op­er­a­tor of train No. 302. By then, the sta­tion was oc­cu­pied by another train.

The “fol­low­ing train,” No. 510, “stopped at the L’En­fant Plaza sta­tion at 3:25 p.m,” the NTSB said. “This train stopped about 100 feet short of the south end of the plat­form, but its cars were en­tirely within the sta­tion.”

Could train No. 510 have been moved back­ward, to make room in the sta­tion? Ap­par­ently not. Be­cause its op­er­a­tor was no longer on board.

Due to the fumes, the NTSB said, “Train 510 was evac­u­ated.”

Why did the smoke blan­ket the train?

Near the spot where the elec­tri­cal melt­down oc­curred, a gi­ant ven­ti­la­tion shaft rises six sto­ries from the tun­nel to the street. In the shaft are four fans, each five feet in di­am­e­ter. Depend­ing on how the blades are set to spin, clock­wise or coun­ter­clock­wise, the fans can pull in or push out mas­sive amounts of air per minute.

The shaft, like the source of the smoke, was about 1,100 feet in front of train No. 302’s lead car. Be­hind the train, in the L’En­fant Plaza sta­tion, two large fans in­stalled be­low the plat­form level also are ca­pa­ble of mov­ing enor­mous vol­umes of air in and out of the tun­nel.

In a safety bul­letin, the NTSB de­scribed a se­ries of mis­steps in which ROCC con­trollers botched the re­mote op­er­a­tion of the fans, ex­ac­er­bat­ing the plight of the pas­sen­gers who were stuck on the train. In ad­di­tion to the mis­takes, two of the fans in the shaft ex­pe­ri­enced me­chan­i­cal prob­lems, in­ves­ti­ga­tors said.

This week’s hear­ings could fill gaps in the public record re­gard­ing the main­te­nance of the fans, what went wrong with their op­er­a­tion and the ef­fects on the riders.

If co­or­di­nated prop­erly, one set of fans would have worked in “sup­ply mode,” spin­ning clock­wise and push­ing fresh air into the tun­nel, while the other set of fans, run­ning in “ex­haust mode,” or coun­ter­clock­wise, pulled out the smoke.

In­stead, the NTSB said, con­trollers in the op­er­a­tion cen­ter ac­ti­vated the fans at crosspur­poses, with both sets pulling the nox­ious fumes. The plat­form fans pulled the smoke in the di­rec­tion of the sta­tion, mean­ing the fumes moved to­ward the train from deep in the tun­nel. Si­mul­ta­ne­ously, the shaft fans pulled the smoke the op­po­site way.

Be­cause of this tug of war, the fumes stayed in the tun­nel, en­gulf­ing the train. “There was not a sup­ply of fresh air to aid in mov­ing the smoke,” the NTSB said.

Mak­ing mat­ters worse, the safety board said, “the train ven­ti­la­tion sys­tem that draws air from the out­side into the cars was not shut off by the train op­er­a­tor.”

Un­der a Metro rule at the time, an op­er­a­tor was not al­lowed to turn off a train’s air-in­take with­out au­tho­riza­tion from the ROCC. Whether the op­er­a­tor of train No. 302 failed to re­quest per­mis­sion or did not re­ceive it re­mains un­clear. Metro has since changed the reg­u­la­tion, leav­ing the air-in­take de­ci­sion to the op­er­a­tor.

As smoke en­veloped the train, and poured into the cars through the in­take sys­tem, sick­ened pas­sen­gers waited more than a halfhour for help to ar­rive, some of them shar­ing wa­ter bot­tles, oth­ers kneel­ing and sit­ting on the floors, seek­ing low clean air.

When D.C. fire­fight­ers fi­nally ar­rived, Lt. Stephen Kuhn later said, the fumes were so thick that they could not see the tail­lights of the train “un­til we were right up next to it.”

JOHN BOAL/EURO­PEAN PRESSPHOTO AGENCY

The L’En­fant Plaza sta­tion was evac­u­ated andMetro ser­vice halted for sev­eral hours af­ter smoke filled a sub­way tun­nel Jan. 12. One pas­sen­ger aboard Yel­low Line train No. 302 died in the in­ci­dent. The NTSB will con­duct its first public hear­ings on the event this week.

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.