Posts about Metro Safety

While this standard argument is compelling, there has been little engagement with the counterargument to federal oversight of urban transit. Consider the following concerns.

Urban rail is very safe: Subways and light rail are already very safe, safer by far than other modes of transportation that are regulated by the DOT including air travel. One wonders then if improving on an already very low fatality rate should be a priority for federal dollars given the other more dangerous modes regulated by the DOT.

The TOC can be improved easily without federal intervention: The criticism leveled against the TOC is not directed at their competence, but at their lack of enforcement powers and funding. So, instead of building a new federal agency, why not give the TOC enforcement powers and increased funding?

TOC audit was actually better than the FTA audit of Metro: While it received little press attention, the TOC audit released earlier this month was more detailed and actionable than either the NTSB or FTA audits concerning the systemic safety hazards at Metro.

Federal urban rail regulation may be unconstitutional: Federal regulation of urban transit systems may ultimately be overturned by the courts. The Commerce Clause of the Constitution limits federal regulation to interstate commerce, and most urban transit systems don't cross state lines like Metro does.

NTSB previously opposed FTA oversight of urban rail: Every urban transit system is very different, despite appearances to the contrary. Unlike other transit modes regulated by DOT which share a common network, urban transit systems develop independently according to unique needs and constraints. The NTSB argued in the 90s that this was reason enough to support the regional system of safety oversight in place today.

For these reasons, I would strongly oppose FTA regulation of Metro and other urban transit agencies if not for one prominent benefit that would result from FTA regulation:

FTA can balance NTSB: While the NTSB serves a valuable role in transportation safety, they are an exclusively reactive organization by statute. Unfortunately, the political pressure to implement any and all NTSB recommendations is overwhelming. This undermines attempts to create a proactive safety organization.

The USDOT, which requires transportation providers to take a more proactive approach to safety, balances the NTSB in the transport modes that it regulates. This balance will never be provided by the TOC or other regional safety oversight bodies.

I am honestly on the fence on this critical issue. While the answer to this issue seems obvious to many, I suspect that the damning of all things Metro since the Red Line Crash is undermining the healthy debate that this issue deserves.

The Obama administration supports a bill that would give the FTA this power, but Senator Tom Coburn (R-OK) has put a hold on the bill in the Senate for many of the reasons listed here, as well as the lack of offsetting spending cuts or taxes in the legislation.

What do you think? Should the FTA regulate urban transit agencies?

Metro also provides information about the operating divisions that have the most days without a lost time injury, highlighting their best performance. It is a good idea to commend those divisions for maintaining safe work practices and keeping their workers safe. But for a board oversight role, it would be more instructive to show the operating divisions with the worst safety record, and show what are the findings from safety assessments or the results of incident investigations into the injuries.

Metro should look at their best performing divisions as a source of best practices to transfer over to other divisions where workers have been injured. Rather than highlight their best performance, Metro and its oversight board should be looking into their worst performance in order to improve it.

In my opinion, the least valuable part of this presentation is a series of 10 bar charts showing the number of various types of incidents in 2009 compared to 2010. Metro shows that the "2009 Average" rail passenger injury rate was 0.47 per million passenger trips, and that the 2010 injury rate to date is much lower, at 0.14.

This allows the casual reader to think that Metro is somehow improving. But the previous year had the worst accident in Metrorail history, and should not be used as a sort of "average." Instead, Metro should treat every accident as undesirable, and avoid showing averages unless they are based on the long-term average rate over the past decade or a similar time period.

Given these charts showing incident and injury rates varying over 16 months in the past, what is the board supposed to do? There isn't any analysis given, only data. Are a lot of the injuries preventable? Are they related to the NTSB findings and therefore the money to be spent on fixes would help the incidents Metro is experiencing? Is there some other fix that would help to prevent these incidents?

Metro is showing data about the injury rate, but there is no explanation about why the rates are increasing or decreasing. Maybe this will be part of the discussion, but so far it doesn't look like Metro has dug into what the problems really are.

Compare this with the discussion about Elevator and Escalator repair. Metro had to call in an assessment consultant to find its problems, and plans to re-hire the same consultant to figure out whether the corrections have stuck. This indicates that at least in some areas, Metro has lost the ability to assess its own performance, and must rely on outside organizations like consultants, the NTSB and the TOC to find its problems and recommend fixes. One of Metro's goals should be to develop a culture of self-assessment, so that the organization can better find problems itself without relying on outside help.

The TOC audit portrays the safety management within Metro (known as SAFE for System Safety and Environmental Management) as an island, isolated from the departments whose processes and procedures SAFE is charged with continuously improving.

For example, SAFE's manual of operating procedures (Safety System Program Plan, or SSPP) apparently bears little relation to what workers actually do, because other departments aren't included in writing or revising these procedures.

p 64. Area of Concern 4-1. Non-SAFE departments and the ELT are not engaged in updates to the SSPP. WMATA did not solicit the review and comments of the other WMATA departments to which the SSPP applies, per the lessons learned from the December 2009 Internal Safety Audit conducted by APTA. Thus, the descriptions provided within each element of the SSPP do not fully represent the processes and documentation used by the non-SAFE departments in implementing the SSPP.

The island move further from the shore when hazard analysis is conducted, as SAFE ignores reports of hazards from most sources.

p 78. Area of Concern 6-6. Too few sources provide input regarding hazardous conditions. Primary input of hazardous conditions to SAFE comes from the OCC [Operations Control Center]. SAFE needs to expand the sources of hazardous condition reporting to include inspections, audits, investigations, observations. hotlines, etc.

This is very consistent with the complaint amongst Metro workers that the organization won't do anything about reported safety hazards.

SAFE is primarily reacting to major accidents, not getting out ahead of the next accident. This reactive position is only hardened by the enormous pressure to respond to NTSB recommendations, all of which are reactive preventions of the causes of previous major accidents. TOC criticizes this reactive posture, as have I on multiple occasions.

p 79. Area of Concern 6-9. Hazard management does not include smaller incidents. The trending and analysis of multiple, less serious, incidents or near misses is not currently being accomplished.

Some portion of these issues are being addressed by Metro, with new hires and training in SAFE, all of which are mentioned by TOC.

Thanks to Metro now releasing a Vital Signs Report each month, we know that escalator availability has been hovering around 90% and elevator availability around 96%.

What level of availability should riders reasonably expect once these maintenance procedures are fixed? Are the targets set in the Vital Signs report of 93% escalator and 97.5% elevator availability the right targets, and will these changes get us to these targets?

1 out of 14 escalators out of service may be a significant shift in expectations for riders. And the target for elevator availability would actually mean Metrorail availability of 90% for riders with wheelchairs and strollers, since each trip requires 4 elevators and these riders can't take escalators or stairs.

Once targets are agreed upon, there needs to be a clear link between these targets and the investments required to meet them. There's lots of good ideas for improving availability, like the "team-building initiative to enhance communications among staff and improve the team's effectiveness" that the current Vital Signs report commits to.

But we need more than good ideas. What will it take to actually reach availability targets? And how do we know?

The main proposal put forward by the audit requires increasing scheduled maintenance intervals. But maintenance requires taking escalators out of service, and Kubicek says in Vital Signs that this is the cause of some lack of availability. At what point does the downtime of increased maintenance outweigh the unscheduled downtime that this maintenance prevents? Can any of the inspections conducted during routine maintenance be done with remote sensors (detecting water, vibration, heat, etc)?

Also, as David pointed out in an interview on Fox 5 (embedded below), there's another factor behind escalator problems: resources. The system was new in the 1970s and 1980s, and escalators and elevators needed little repair. Now they need more repair, but budgets still are underfunding ongoing maintenance like this. And at least so far, the Board has been showing an unlimited willingness to spend money on safety fixes without regard to cost, but may continue to shortchange other needs.

A smart General Manager would announce to the Board and, by extension, to the media: We're at 90% now. We've found these internal issues and will fix them. These contribute to (say) 20% of downtime so that will get us to 92%. If we allocate some of our capital dollars to escalator repair, we can get to (say) 96% by addressing additional causes of 40% of downtime.

Soon, the public will be invited to comment on WMATA's Capital Improvement Plan, a multi-year priority list of where to spend limited capital dollars. Much of that rightly ought to go to safety, but there will be a danger of putting too many eggs in that one basket. If we really want escalators to be working, which by the way has safety implications of its own, we will need to send a message to the Board to be sure to allocate some capital dollars to repairing and replacing escalators.

That means that 42% of workers (70% of 60%) have reported a safety hazard in the past year. That's alot of safety hazards reported. That also means that 30% of safety hazards noticed by workers aren't being reported. Sarles then addresses the latter issue of fear of retaliation.

There is a strong concern about retaliation but in somewhat unexpected ways. Employees' primary concerns are not that they would be fired or demoted. The strongest concerns that were cited were that it would be difficult for them to work among their peers, that the organization wouldn't do anything about their report, and that the organization would not protect them against retaliation in their immediate working environment.

So, imagine that you are the Chair of the Board at this point. The NTSB has excoriated your Board for its lack of safety oversight following the Red Line crash that killed 9 people. Everyone is pushing Congress to let the FTA regulate transit agencies because of your Board's failure in oversight.

You have responded to the NTSB and FTA by saying that you are now up to the task of safety oversight, and are modifying your mission statement in the very same meeting to place more focus on safety.

You would probably ask if the workers are correct that the organization wouldn't do anything about their report. What does the organization do with reports of safety hazards, and is this process documented and audited? What has it done with the safety hazards reported by 42% of workers?

Board Chair Benjamin:

My reaction is that your report is impressive, it's thorough, it's in-depth, it's exhaustive, and more than anything else it shows your leadership as a General Manager, and the combined efforts of a really dedicated senior staff, middle management and staff throughout this authority. It's an extremely, extremely good report.

Huhh? What is "extremely, extremely good" about 60% of workers noticing safety hazards and 30% of workers fearing retaliation if they report safety hazards?

I would like to comment very briefly on the employee survey because it's such an interesting set of results which reflect not necessarily reality but perception. And in this particular case, perception is so important. Because if what you are trying to do is revamp your culture, you have to know where you're starting.

So, after lauding praise on management, Benjamin dismisses the workers' perceptions as "not necessarily reality" and shows absolutely no interest in what has been done with the safety hazards noticed by 70% of workers and reported by 42% of workers.

This meeting exemplified the dangerous rut that Metro is in: (a) writing blank checks to demonstrate its response to NTSB recommendations designed to prevent a crash like the Red Line crash, while (b) placing their heads in the sand when anyone suggests that more safety hazards exist.

What exactly is the Board doing any differently than they did before the Red Line crash to demonstrate safety oversight?

Other than reviewing statistics of injuries and fatalities by month (which we all knew by reading the newspaper anyway), forming a Safety committee and changing the mission statement, it's unclear what is happening that is supposed to resemble safety oversight.

Here's what has happened so far. Metro announced in July that it has set aside $30 million over three years to implement any NTSB and FTA recommendations following the June 2009 red line crash that killed 9 people.

However, when Senator Mikulski (D-MD) asked in August for cost estimates of each recommendation, the total provided by Board chair Peter Benjamin was $100 million. And that's just for recommendations for which Metro has cost estimates.

When Chief Safety Officer Jim Dougherty met with the Metro Board on Sept 16, not a single question was asked about the skyrocketing costs and trade-offs of implementing federal recommendations.

Actually, not a single question was asked about the details or trade-offs of any of the recommendations, from the $55 million replacement of Gen 2 track circuit modules to the $25 million safety analysis of the automated train control system.

The oversight meeting with Dougherty lasted for only 45 minutes, and consisted primarily of a self-congratulatory presentation on the progress made by WMATA, which included the new logo seen here.

To exercise safety oversight, the Metro Board must ask about safety trade-offs in every meeting: Why are the current safety actions, whether they originate from the NTSB or not, the highest safety priorities?

The FTA asked this question during their audit and was told that no prioritized list of safety actions exists. The answer to the Board should look something like the table below. In fact, this should just be added to the monthly Vital Signs report.

This table is a Hazard Tracking Log (HTL). It's based on a similar table from a booklet called Hazard Analysis Guidelines for Transit Projects, published 10 years ago by the FTA. Lots of safety actions are prioritized based on the severity and likelihood of the identified hazard causing injury or fatality. Hazards and their corresponding safety actions are generated by 2 types of hazard analysis, reactive and proactive, which I describe elsewhere.

The non-NTSB recommendations in the table are empty because the Metro Safety Office has yet to conduct proactive hazard analysis for any critical system, as I've discussed elsewhere, and integrate the resulting safety actions into a prioritized list.

Most of the FTA's recommendations are focused on putting a Hazard Management System in place (basically, doing what the aforementioned booklet says to do) that consists of hazard analyses that continuously update the prioritized Hazard Tracking Log table. Metro's responses to FTA and NTSB recommendations, however, raise two serious concerns about its ability to put this System in place.

Metro is outsourcing hazard analysis of the Automatic Train Control system.

This $25 million, 3-year project, which is in response to an NTSB recommendation, was announced by Benjamin in his August letter to the Congressional delegation. That's a lot of money. $25 million would employ 75-100 engineers and analysts full-time for 3 years. One wonders what the WMATA safety office does if we are paying $25 million to contractors to do hazard analysis.

And what happens when the analysis ends, and we upgrade the automated train control system? Do we pay several million dollars again to a contractor to conduct another safety analysis? It seems like a good idea for the contractor to train and transition the safety analysis to WMATA's own safety office.

However, when asked if this would happen, a WMATA spokesperson responded, "The task will not specifically train Metro employees in how to conduct safety analysis, but will identify proper response and prioritization to safety concerns, particularly in an integrated environment."

Metro touts Hazard Management success without actually doing hazard analysis.

In Metro's August reply to the FTA audit, Metro merely copied the FTA recommendations (e.g. identify skills required for hazard analysis; train employees in these skills; etc) and pasted them into the HTL table shown above as a demonstration of progress.

Metro then announced triumphantly, "By evaluating the FTA recommendations in this manner, WMATA demonstrates that it has established a true hazard management program that incorporates a risk-based approach to evaluate and mitigate hazards".

This misplaced concern for the presentation of the results of hazard analysis, over the actual analysis itself, is even aired by WMATA's own IT department in the very same letter to the FTA. After discussing changes to the IT architecture being made to support hazard analysis, the following concern is said to be a "threat" to the entire project:

The System Safety and Environmental Management Department is awed by product suite success stories, dynamite product demonstrations and industry colleagues' evaluation of technology.

The FTA should not accept the responses of WMATA to its recommendations until WMATA has demonstrated its ability to actually do a hazard analysis of a complex system, which would enable it to then prioritize hazards in a system. It doesn't really matter which system it is—the elevators, the train doors, even the payroll system would be fine.

Metro can do this. It's my hope that, when the FTA begins regulating transit agencies, they will hold up Metro as an example for the rest of the country of world-class safety management.

But Metro can't do this and hand out blank checks for responding to NTSB recommendations regardless of the safety trade-offs. They are simply incompatible approaches to safety. The latter, reactive approach leads to budget shortfalls requiring fare increases, and to injuries and fatalities. The former, systematic approach leads to improved safety at the most efficient and rapid pace possible.

But Metro can only do this with leadership in oversight, particularly from Board chair Benjamin and Safety & Security Committee chair Mort Downey.

Kenneth Hawkins, brother of one of the killed passengers from the Red Line crash, asked following the NTSB hearing, "Who's going to hold WMATA accountable?" I still have the same question.

The current statement reads,

Metro provides the nation's best transit service to our customers and improves the quality of life in the Washington metropolitan area.

This is the proposed new language:

Metro operates and maintains a safe, reliable and effective transit system that enhances mobility, improves the quality of life, and stimulates economic development in the Washington metropolitan area.

The new committee seems to be a good idea, and Downey is a good choice to run it. I doubt the mission statement change will make any difference, but it's also an easy way to comply with an NTSB recommendation.