A view of the L'Enfant Plaza Metro station, which was evacuated as smoke filled a Metro train on Jan. 12. (John Boal/EPA)
As experts continue trying to determine the precise reason for an electrical malfunction that filled a Metro tunnel with smoke Jan. 12, federal safety officials have disclosed what might be the best clue yet regarding the cause of the deadly calamity.
In a 10-page document laden with transit-engineering terminology, the National Transportation Safety Board, which is investigating the January incident, implicitly criticized Metro last week for failing to ensure that all power-cable connections near its subway tracks are protected from contaminants.
And transit engineers who specialize in electricity, speaking on the condition of anonymity because they are not involved in the NTSB inquiry, said in interviews that the intrusion of contaminants into power-cable connections could create conditions for an electrical meltdown such as the one that occurred near Metro’s L’Enfant Plaza station.
The NTSB said “a number” of power-cable connections throughout the rail system lack “sealing sleeves” that are designed to keep water, grime, metal dust and other substances away from electrical current. The problem is significant, the safety board said, because these “improperly constructed power-cable connector assemblies . . . can lead to short circuits that can generate fire and smoke in tunnels.”
Whether the cable connector assemblies involved in the Jan. 12 crisis were fitted with sealing sleeves remains unclear. The electrical meltdown, in a Yellow Line tunnel just south of the L’Enfant station, caused so much “thermal damage” that the scorched power cables from the tunnel are still undergoing laboratory analysis, the NTSB said.
However, the densely worded NTSB document — written by and for experts in subway electrical systems — cites circumstantial evidence that the cable connector assemblies at ground zero of the L’Enfant incident lacked sealing sleeves.
Noxious fumes generated by the malfunction enveloped a six-car train that had stopped in the tunnel, sickening more than 80 riders, one of whom died.
[Choking riders gasped for air on train filled with noxious fumes.]
After the document was released last week, Metro acknowledged that about 80 percent of the approximately 6,400 power-cable connector assemblies throughout the subway lack sealing sleeves. Rectifying the situation will require months of track work, resulting in service disruptions for many riders, the transit agency said.
Metro officials have been publicly tight-lipped about the L’Enfant incident, citing federal rules that limit how much information the agency can disclose while the NTSB inquiry is underway. For that reason, Metro’s top managers said, they could not offer a full explanation to the public for why only about 20 percent of subway power-cable connections have the sealing-sleeve protection that the NTSB recommends.
Even members of Metro’s board of directors couldn’t get a detailed answer when Deputy General Manager Rob Troup was asked about the issue at a meeting last week.
“I am under the NTSB confidentiality agreement,” Troup later told reporters, echoing his response to Metro board members. “I can’t speak to this.”
The problem described in the NTSB document has to do with the way electricity is distributed in the subway — and the importance of keeping that electrical current contained within heavy insulation.
Along Metro’s 230-plus miles of tracks, there are thousands of power supply cables, each carrying electricity to the rail system from outside transmission stations. In tunnels, each insulated supply cable enters through a conduit beside the tracks. The supply cable then connects to another insulated cable, which delivers the power to the third rail. The electrified third rail is what propels the trains.
The point where the two cables meet is critical. Because the insulation has been severed to allow for the connection, the electricity is exposed. Think of splicing two small power cords in a home and wrapping the connection with electrical tape. The difference is, the subway cables are nearly two inches in diameter, and the connection is protected by a heavy, elbow-shaped fiberglass cover called a “boot.”
The two openings in the boot — one for each cable to go in — need to be sealed against contaminants, including water. Here’s why:
Moisture is commonplace in subway tunnels, a result of underground water “intrusion.” Tunnels also are loaded with what the NTSB document called “particulate contaminants.” These include “brake dust; rust particles; rail grinding swarf [fine metallic filings or shavings]; rail and wheel wear debris”; and on and on. Power cables and other tunnel infrastructure are coated with the stuff.
In electrical-speak, the particulates and moisture are “conductive substances,” meaning they have a low resistance to electricity. They offer a path for electrical current to jump dangerously all over a tunnel if the electricity escapes from its insulated containment. And for that to happen, all it takes is for the path to be completed — for the trail of moisture and particulates to seep inside the protective boot and reach the exposed current.
When electricity manages to escape, following that path of low resistance, it can generate tremendous heat as it makes its way to ground. Electrical current flowing into the conductive steel walls of a tunnel, for example, can turn the walls into giant hot plates, creating an oven effect. Depending on the severity of the incident, the thermal damage to surrounding infrastructure can create a huge volume of smoke.
The phenomenon is called “arcing.” Within hours of the Jan. 12 incident, the NTSB said that the heavy smoke in the L’Enfant tunnel had resulted from “an electrical arcing event.” In the document delivered to Metro last week, the safety board specified that the arcing involved uncontained current flowing “between the electrical power cables and the steel wall of the tunnel.”
“The smoke in the tunnel was generated by thermal damage to about 16 feet of electrical power cables and insulation, portions of 4 fiberglass cable connector covers” — meaning four boots — “and about 5 feet of fiberglass third rail cover,” the document said. As for why the arcing occurred, the NTSB said investigators have yet to determine whether the electricity escaped because of missing sealing sleeves on the boots.
“Extensive thermal damage to the electrical power cables and connector covers in the L’Enfant Plaza incident consumed evidence,” the document said, adding that “additional analysis is needed before the cause can be determined.”
[Burning cable insulation aggravated the L’Enfant smoke crisis.]
Still, there is circumstantial evidence.
On Feb. 11, a few weeks after the L’Enfant calamity, a similar but less severe smoke incident occurred in a tunnel between the Court House and Rosslyn stations in Northern Virginia. The arcing in that case appeared to have been caused by moisture and particulates seeping into a cable-connection boot, according to the NTSB document. It said investigators “found that the failed power-cable connector assembly . . . was missing its sealing sleeve.”
This apparently prompted the NTSB to conduct a wider examination of power-cable connections in the subway. “In what should be identical assemblies throughout the system, NTSB investigators found numerous power-cable connector assemblies that are not in accordance with [Metro’s] engineering design specifications,” the document said.
“These assemblies are frequently missing sealing sleeves. . . . Even the post-accident repairs made to the power-cable connector assembly at L’Enfant Plaza did not include the sealing sleeves indicated in the [Metro] engineering design specifications.” In some locations, the document said, investigators found that “heat-shrink tubing or electrical tape” had been used instead of sealing sleeves.
The document, signed by the NTSB’s acting administrator, Christopher A. Hart, urged Metro to immediately begin “a program to ensure that power-cable connector assemblies are installed in accordance with its engineering design specifications.” The recommendation was intended “to prevent accidents and save lives,” the document said.
Metro’s public response was clear in one respect, yet vague in another.
“We fully embrace this recommendation from the NTSB and will apply the resources needed to complete the work in a timely manner,” said Troup, the transit agency’s No. 2 manager and top engineering official. “We recognize this as being an improvement to our construction methodology and also an important safety initiative.”
At Thursday’s meeting of Metro board members , Troup explained that “prior practice going back 25 years . . . was to not install the sleeves” on cable-connection boots. That was before Metro, about three years ago, adopted a “design directive drawing” calling for the boots to be protected against contaminants, he said. However, the directive did not specify that sealing sleeves had to be used in every instance.
Troup said the directive allowed work crews “some latitude” in choosing what type of protection to install on the boots. Although some boots were fitted with sealing sleeves, he said, “there could have been tape that they could have put on” others. “They could have put heat-shrinking around there. Different things like that.”
He said: “The issue of concern in this particular case is the potential for inconsistent application represented by our design directive drawing. And this was recognized by the NTSB. Moving forward, we will have what is termed a ‘design standard drawing,’ which does not allow for deviation.”
In other words, henceforth, sealing sleeves will be required on all boots.
So the question arose: Since the transit agency acknowledges that installing the sleeves will be “an improvement” and “an important safety initiative,” why didn’t Metro require sleeves in the past, instead of allowing work crews “some latitude”?
When a board member raised that issue, Troup declined to answer, citing the ongoing safety investigation. “I need to be very, very clear,” he said. “I’m under the NTSB confidentiality agreement. I can’t speak to that.”
The NTSB has said that its final report on the Jan. 12 incident won’t be issued until early next year. However, more details about the sleeves, as well as other aspects of the L’Enfant electrical meltdown, could emerge June 23 and 24, when the safety board holds public hearings related to the investigation.
As for installing sleeves on 80 percent of the subway’s 6,400 or so boots, Troup said it will take months of labor at a yet-to-be-determined cost. And in stretches of the subway where work crews are on the tracks, trains will run slower and less frequently.
“I really don’t want to give a time frame now,” Troup said when asked when the disruptive work will get underway. “What I can tell you is that we will [decide] as quickly as possible, to be able to transmit that information out to the public.”