Skip Ribbon Commands
Skip to main content
Safety Recommendation Details

Safety Recommendation R-16-004
Details
Synopsis: On January 12, 2015, at 3:15 p.m. eastern standard time, Washington Metropolitan Area Transit Authority (WMATA) southbound Yellow Line train 302, with about 380 passengers on board, stopped after encountering heavy smoke in the tunnel between the L’Enfant Plaza station and the Potomac River bridge in Washington, DC. The operator of train 302 told the Rail Operations Control Center (ROCC) that the train was filling with smoke and he needed to return to the station. The WMATA ROCC allowed train 510, following train 302, to enter the L’Enfant Plaza station, which also was filling with smoke. Train 302 was unable to return to the station before power to the electrified third rail, which supplied the train’s propulsion power, was lost. Some passengers on train 302 evacuated the train on their own, and others were assisted in evacuating by first responders from the District of Columbia Fire and Emergency Medical Services Department (FEMS). As a result of the accident, 91 people were injured, including passengers, emergency responders, and WMATA employees, and one passenger died. WMATA estimated the total damages to be $120,000. The National Transportation Safety Board (NTSB) has been concerned with the safety of the WMATA rail system since 1970, when it conducted a special study of the proposed transit rail system while it was still under construction. The resulting report, NTSB/RSS-70/1, Study of Washington Metropolitan Area Transit Authority’s Safety Procedures for the Proposed Metro System, resulted in one safety recommendation to WMATA to “develop the capability within WMATA for system safety engineering and apply system safety principles to all aspects of the proposed [rail] system to identify, assess, and correct those deficiencies identified by the analysis.” This accident is the 13th WMATA rail accident investigated by the NSTB since WMATA rail began operation in 1976. The NTSB has issued 101 safety recommendations to WMATA since 1970.
Recommendation: TO THE DISTRICT OF COLUMBIA OFFICE OF UNIFIED COMMUNICATIONS: Audit your public service answering point (PSAP) to validate compliance with the standards published by the National Emergency Number Association or another similar standards organization. The audit should (1) determine the average length of time that call takers use to process an emergency call and dispatch emergency service, and (2) compare those results with those of other comparable PSAPs.
Original recommendation transmittal letter: PDF
Overall Status: Open - Acceptable Response
Mode: Railroad
Location: Washington, DC, United States
Is Reiterated: No
Is Hazmat: No
Is NPRM: No
Accident #: DCA15FR004
Accident Reports: ​Preliminary Report: WMATA Smoke and Electrical Arcing Accident in Washington, DCWashington Metropolitan Area Transit Authority L’Enfant Plaza Station Electrical Arcing and Smoke Accident
Report #: RAR-16-01
Accident Date: 1/12/2015
Issue Date: 5/23/2016
Date Closed:
Addressee(s) and Addressee Status: District of Columbia, Office of Unified Communications (Open - Acceptable Response)
Keyword(s):

Safety Recommendation History
From: NTSB
To: District of Columbia, Office of Unified Communications
Date: 8/10/2018
Response: Our investigation of this accident found that the call from the WMATA rail operations control center to your PSAP lasted 2 minutes, 11 seconds. Our report points out that National Fire Protection Association Standard 1221 states that 90 percent of emergency calls should be processed within 64 seconds; however, you did not dispatch fire and emergency medical services (FEMS) until 2 minutes, 38 seconds after the conclusion of the call, and there was a lapse of almost 5 minutes from WMATA’s call until the initial dispatch of FEMS to the scene. Our recommendations to you resulted from our conclusion that your call processing delayed the emergency response to this accident. We note that the OUC is subject to annual performance oversight by the Judiciary Committee of the DC Council, and that a primary component of this oversight is a key performance indicator for emergency call processing and dispatching times. As part of the performance report, you audit the average length of time that call takers use to process an emergency call and dispatch emergency service. Because you are already subject to this oversight, you do not believe you need to create a new auditing, oversight, and review process for PSAP call processing. Regarding the second part of Safety Recommendation R-16-4, we note that you believe it is inappropriate to compare the OUC to other PSAPs for the following reasons: 1) The OUC is the only state-level PSAP in the United States. In every other US jurisdiction, the state agency provides guidance, regulation, and oversight, while actual call processing is handled by PSAPs falling under local jurisdiction. The OUC performs all of those functions as one entity. 2) You are responsible for the unified communications of the DC government. Unlike most PSAPs, in addition to your emergency 911 responsibilities, you are responsible for nonemergency government call centers and service requests as well as for nonemergency radio communications. 3) There are 32 distinct and separate police forces within Washington, DC, creating unique jurisdictional disputes that frequently increase the time it takes to dispatch to an incident that falls near a jurisdictional line, sometimes requiring the OUC police dispatcher to negotiate between the different potential police departments that may need to respond. 4) A large portion of potential 911 callers are not DC residents, but commuters who live outside the city and visitors to Washington. Such visitors, in particular, often have difficulty identifying their locations in reference to the city’s quadrants. The same physical address can potentially exist in multiple quadrants, resulting in complexities unique to the DC PSAP. We disagree with you that the response times for emergency calls to your PSAP cannot be compared to those of other cities. The comparison needed is for emergency calls; issues of state-level responsibilities and nonemergency communications are not related to emergency call response times. We believe it is common in most US cities for a large percentage of people to be commuters who live in nearby suburbs or visitors who are not familiar with the city. Although there are many police departments within Washington, DC, other large cities also have numerous police departments and deal with issues of jurisdictional boundaries. We acknowledge that you have been internally auditing the OUC’s performance for many years. We reviewed the summary fiscal year (FY) 2017 OUC Performance Accountability Report (OUC PAR), which reports your audit results for percent of 911 emergency calls in which the call to-queue time is 90 seconds or less, and in which time from queue to dispatch is 60 seconds or less. These two measures are similar to the average length of time that we recommend that call takers use to process an emergency call and dispatch emergency service. Despite your argument that your audit results cannot be compared to those of other cities, we note that the OUC PAR contains performance goals for the measures that you audit. Notably, in FY17, you failed to meet the goals of these two standards, but you have established a call flow taskforce (CFTF) to analyze emergency calls and identify issues that hinder quick call processing. You are using your current internal auditing program (required by the DC Council), your performance goals, and the CFTF to address the problems that we found in our accident investigation. To satisfy this recommendation, after you believe you have identified and resolved problems that prevent your PSAP from meeting your performance goals, you will need to sponsor an independent audit by an appropriate outside organization that will report the results and compare them to established standards published by NENA or another similar standards organization. Pending that, Safety Recommendation R-16-4 is classified OPEN--ACCEPTABLE RESPONSE.

From: NTSB
To: District of Columbia, Office of Unified Communications
Date: 9/15/2016
Response: The National Transportation Safety Board (NTSB) has reviewed the National Highway Traffic Safety Administration (NHTSA) “Request for Information: Nationally Uniform 911 Data Systems,” published at 81 Federal Register 27904 on June 30, 2016. The proposed nationally uniform 911 data system would contain uniform data elements for all computer-aided dispatch (CAD) systems. These self-defining data would facilitate the collection, analysis, and sharing of all CAD data received, collected, processed, and transmitted during 911 calls, including administrative and cost data. These data, in turn, would be made available to all 911 public safety answering points (PSAP) and 911 authorities at the state and local levels. NHTSA poses 15 questions, of which three are applicable to the NTSB. Question 1 addresses significant changes in 911 data systems over the last 10 years, question 2 addresses the usefulness of the data that would be created by implementing a nationally uniform 911 system, and question 3 focuses on the most critical issues facing the current use and interconnection of PSAP CAD systems. Usefulness of Implementing a Nationally Uniform 911 Data System (question 2) The proposed nationally uniform 911 data system would improve the efficiency of NTSB investigations by providing a consistent data format, which would allow us to compare investigative cases, identify trends, and provide more informative evaluations to PSAPs and other stakeholders. Together, these benefits would have a positive effect on our mission to improve transportation safety. The NTSB recently issued safety recommendations to the District of Columbia Office of Unified Communications as a result of our investigation of the January 12, 2015, Washington Metropolitan Area Transit Authority L’Enfant Plaza station electrical arcing and smoke accident: Audit your public service answering point (PSAP) to validate compliance with the standards published by the National Emergency Number Association or another similar standards organization. The audit should (1) determine the average length of time that call takers use to process an emergency call and dispatch emergency service, and (2) compare those results with those of other comparable PSAPs. (R-16-4) Upon completion of action satisfying Safety Recommendation R-16-4, develop call processing standards for the public service answering point (PSAP) to ensure that 911 calls are processed in accordance with those of other comparable PSAPs. (R-16-5) Train call takers for the public service answering point on the standards developed in Safety Recommendation R-16-5, and include the standards in recurrent training. (R-16-6) These three recommendations address system performance relative to comparable PSAPs. Each recommendation is currently classified “Open?Await Response.” Collectively, the intent of the recommendations is to audit PSAP performance, develop call processing standards, and train call takers so that all aspects of public safety response meet a reasonable comparable standard. These recommendations reinforce the necessity for a better understanding of comparable PSAP performance and the integral role of a uniform 911 data system in improving all levels of emergency response.

From: District of Columbia, Office of Unified Communications
To: NTSB
Date: 8/17/2016
Response: -From Karima Holmes, Director: The Office of Unified Communications (OUC) is subject to annual performance oversight by the Judiciary Committee of the District of Columbia Council (D.C. Council), the District of Columbia's legislative body. A primary component of this oversight is the Key Performance Indicator. For OUC, the Key Performance Indicators have included standards for emergency call processing and dispatching times since the creation of the agency. Early in its history, OUC's Key Performance Indicators, included the following: I. Percentage of 9-1-1 Calls Answered within 5 Seconds 2. Percentage of9-l-l Calls (Wire Line and Wireless) Abandoned1 ln Fiscal Year 2009, the D.C. Council added the following Key Performance Indicators, which were based upon the national standards established by the National Emergency Number Association (NENA), in Document 56-005 (2006): I. Percentage of Calls in which Call to Queue is 60 Seconds or Less 2. Percentage of9-1 -1 Police Priority 1 Calls in which Queue to Dispatch is 60 Seconds or Jess 3. Percentage of9-l-1 FEMS Calls in which Queue to Dispatch is 60 Seconds or Less2 Over the years, these Key Performance Indicators have been adjusted by the D.C. Council periodically. Currently, the agency is tasked with meeting the following standards: 1. Percentage of9-1-1 Calls Answered within 5 Seconds 2. Percentage of9- l-1 calls (Wire Line and Wireless) Abandoned 3. Percentage of9-l-l Calls in which Call to Queue is 90 Seconds or Less3 In preparation for each D.C. Council oversight, OUC performs an annual audit of the average length of time that call takers use to process an emergency call and dispatch emergency service. As a result, OUC need not create a new auditing, oversight, and review process for its call processing, as these processes are well-established. OUC does not regularly compare its call processing times with those of other comparable PSAPs because the District of Columbia PSAP is not generally comparable to any other PSAP. First, OUC is the only state-level PSAP in the United States. In every other U.S. jurisdiction, the state agency provides guidance, regulation, and oversight, while actual call processing is handled by PSAPs falling under local jurisdiction. OUC performs all of those functions as one entity. Second, the OUC is responsible for the unified communications of the District of Columbia government. Unlike most PSAPs, OUC is responsible for non-emergency government call centers and service requests, and nonemergency radio communications, in addition to its emergency 9-1 -1 responsibilities. Third, OUC is the only PSAP in the United States with 32 distinct and separate police forces with active jurisdictions as follows: 1. 11th Security Police Squadron, Bolling Air Force Base 2. Bureau of Alcohol, Tobacco, and Firearms 3. Amtrak Police Department 4. Criminal Investigation Division, Internal Revenue Service 5. Defense Protective Services, Department of Defense Washington Headquarters Services 6. Drug Enforcement Administration 7. Bureau of Engraving and Printing Police Force 8. Federal Bureau of Investigation 9. Federal Protective Service, Department of Homeland Security 10. Government Printing Office Police 11. U.S. Immigration and Customs Enforcement 12. Library of Congress Police 13. Marine Corps Law Enforcement 14. National Zoological Park Police 15. United States Postal Police 16. Postal Inspection Service, United States Postal Service 17. Office of Protective Services, National Gallery of Art 18. Office of Protective Services, Smithsonian Institution 19. Department of Protective Services, United States Holocaust Museum 20. Department of State Diplomatic Security 21. Supreme Court Police 22. Department of the Navy Police Division, Naval District Washington 23. Naval Criminal Investigative Service 24. United States Army Criminal Investigative Command, Department of the Army Washington District, 3rd Military Police Group 25. United States Army Military District of Washington 26. United States Capitol Police 27. United States Coast Guard 28. United States Customs Service 29. United States Marshals Service 30. United States Park Police 31. United States Secret Service 32. United States Secret Service Uniformed Division This uniquely creates inherent jurisdictional disputes that frequently increase the time to dispatch an incident that falls near a jurisdictional line. For example, when an incident takes place near the National Mall, a federal park, the OUC police dispatcher may have to negotiate between the Metropolitan Police Department and the United States Park Police as to which organization is the appropriate first responder. Finally, the District of Columbia's population is uniquely transient. A large portion of potential 9- l-1 callers in the District of Columbia at any given time are not D.C. residents, but Maryland and Virginia residents who commute into the District for work, and visitors to the Nation's Capital. Visitors to the Nation's Capital, in particular, often experience significant difficulty identifying their locations in reference to the quadrants of Northwest, Northeast, Southwest, and Southeast. The same physical address can potentially exist in multiple quadrants, resulting in complexities unique to the District's PSAP.