Safety – USA Experience
THE US EXPERIENCE
Many people think that light rail is safer than urban driving, but the statistics from U.S. Department of Transportation on light rail fatalities per passenger mile do not bear out that claim, as shown in the chart shown next. All sources are documented in the spreadsheet showing calculations posted here:
Note on above chart: As of September 27, 2010 PITF has been unable to locate U.S. light rail fatality data for 2008 and 2009.
Sound Transit claims that Seattle’s Central Link Light Rail Initial Segment is safe, because “the probability of a single system failure resulting in a critical chargeable accident is once in 131 to 13,000 years.” However, the 1999 Environmental Impact Statement forecast that there might be a collision between a train and a motor vehicle, bicycle, or pedestrian every twelve days! What’s going on here?
Thirteen people were injured in August 2006 when a light rail train and a transit bus collided in Weehawken, New Jersey. Picture from WABC-TV coverage at http://abclocal.go.com/wabc/slideShow?section=news&id=4454919&slide=6
Chargeable means that an accident could have been prevented by Sound Transit action. Critical means fatalities or serious injuries. According to Sound Transit and (apparently) Federal Transit Administration, fatalities from collisions between light rail trains and motor vehicles that are caused by driver carelessness, confusion, or recklessness are not chargeable, even though such fatalities would be made impossible by designing a grade-separated right of way in the Rainier Valley, and by avoiding the mixing of buses and trains in the downtown Seattle Bus Tunnel.
Instead, the Sound Transit design for light rail makes collisions not only possible, but likely.
In the Rainier Valley, with four miles of surface right-of-way providing numerous legal and illegal places for track crossings, fatal collisions are inevitable. Historical light rail fatality rate data extrapolated to Link Light Rail in the Rainier Valley operating with 272 trains per day as described in the 2002 Environmental Assessment indicates a forecast of approximately eight fatalities every ten years from inter-modal collisions. (Inter-modal means between trains on tracks and vehicles with rubber tires.)
So far, as of April 2013, there have been three accidental fatalities associated with Central Link Light Rail, all pedestrians and not classified as suicides, according to reporting in The Seattle Times. The line opened on July 18, 2009, so this train has experienced nearly exactly the same rate of fatalities as expected in PITF’s calculation, 0.8 per year, only from pedestrians being hit instead of vehicles. There have also been at least a dozen non-fatal pedestrian and vehicle collisions since opening, some with injuries.
|Does Link Light Rail at-grade design contradict Safe Routes to School?
Niles’ testimony to Washington State Transportation Commission.
The train tracks and crossings are intended to support four-car trains, where each car is 95 feet long. A four-car train is about 380 feet long. A photograph of a four-car Sound Transit Link Light Rail Train is shown here:
Picture posted at http://www.soundtransit.org/x78.xml?curID=x7357
In a Seattle Times story on the occasion of the light rail construction groundbreaking on November 8, 2003, reporter Mike Lindblom wrote about the Rainier Valley alignment:
The route also crosses 18 streets at grade, perhaps the most controversial aspect of the project.
Surface stations are easy for pedestrians to reach. And construction is cheaper than elevated or underground lines.
But opponents ask: Why have Washington taxpayers invested millions to separate freight trains from cars with overpasses and underpasses elsewhere while Sound Transit isadding more at-grade crossings?
Sound Transit’s environmental studies predict 29 accidents per year between trains and motor vehicles.
The Inspector General of USDOT audited Sound Transit’s plans for separating buses from trains in the Downtown Seattle Transit Tunnel and found them “adequate.”
However, the Inspector General has not audited the safety of the Link Initial Segment at-grade design in the Rainier Valley, where 272 trains per day traveling at up to 35 miles per hour will cross the path of tens of thousands of vehicles at 18 ungated grade crossings.
|The result of a collision between a car and a light rail vehicle in Houston, Texas, June 2004. There were more than 40 collisions of trains and cars in the first six months of operation of Houston Metro Light Rail. From Houston TV Channel 13, KTRK, http://www.abc13.com.|
Coalition for Effective Transportation Alternatives (CETA) takes the position that all future collisions between trains and motor vehicles along the Central Link right of way would be chargeable to Sound Transit, because these collisions would be a direct, inevitable result of Sound Transit’s decision to implement a light rail right of way design that makes this type of collision possible. The Link Light Rail Initial Segment Record of Decision of May 2002 in conjunction with the FTA Hazard Analysis Guidelines of January 2000 requires Sound Transit to show that the expected mean time between fatal collisions is one million operating hours or more, without regard to fault. Given the history of light rail in America to date, this standard is impossible for Sound Transit to meet with its present design that includes many at-grade crossings traversed by trains moving at 35 mph.
Does the Federal Transit Administration agree with CETA’s position? It’s hard to tell.
Here is how the Federal Transit Administration explains their position, as reported by The Seattle Times on April 12, 2004: “FTA regulations do not refer to a threshold number of collisions that is acceptable. Instead, FTA requires grantees to evaluate their systems to ensure that all potential hazards are mitigated to acceptable levels.”
Hey, what does this mean?
Experience with street-running light rail trains in Houston and Denver has shown that many accidents are the result of mixing street trains with cars.. The Rocky Mountain News in Denver on July 3, 2004 published an extensive historical review of light rail collisions along Denver’s T-REX light rail, and found that ungated crossings are the most dangerous locations. The full story is posted here. As is typical, every single accident, over 300 of them, were blamed on car drivers who did not obey traffic control signals, or who otherwise did not provide the train with its legal right-of-way..
To repeat, for Sound Transit to call a train collision with a motor vehicle not chargeable because of driver behavior is to blame the victim. With a different, safer light rail design, blaming the victim would not be necessary because inter-modal collisions would be impossible.
In a related policy initiative, U.S. Department of Transportation is now advocating Safety Conscious Planning (SCP). Here is the meaning: “SCP implies a proactive approach to the prevention of accidents and unsafe transportation conditions by establishing inherently safe transportation networks…. European research indicates that to achieve further large reductions in crash levels, it will be necessary to change the focus from driver behavior initiatives to ones that will make it more difficult for the driver to have a crash. SCP is one way to accomplish that goal, i.e. one of the next generation of road safety strategies. TEA-21 requires ‘Each statewide and metropolitan planning process shall provide for consideration of projects and strategies that will increase the safety and security of the transportation system for motorized and non-motorized users.'” [from the SCP website of USDOT]
Seattle’s Link Light Rail Initial Segment is NOT “safety conscious” and NOT safe.
It is inherently dangerous, will inevitably result in fatal collisions, and should not be built.
However, it was built, and has been operating since July 18, 2009.
Here is an example of a spectacular collision of a train in June 2012 with a truck where nobody was hurt. Picture taken by Sound Transit.
Following picture is a composite of a KIRO TV news photo shot from its helicopter overlaid on Google maps satellite view of the accident site. Commentary later from the Sound Transit Deputy CEO confirmed that the truck was making a U-Turn when it was struck.
Review some interesting and relevant information from “outside the box” on the safety of complex systems (like mixing trains with cars, buses, and trucks):“Learning From Accidents and a Terrorist Attack” by Dan Bricklin
Detailed commentary comes next. Click here for Resource Links at the bottom of this page.
On February 27, March 27, April 10, and May 22, 2003, I presented testimony on behalf of the Coalition for Effective Transportation Alternatives (CETA) to the Sound Transit Board on the problems Sound Transit has in claiming to be certifying the safety of the at-grade road crossings of the Central Link Initial Segment.
An archive of this testimony is at the bottom of this page.
Because of the role of Congress in reviewing Federal funding for Link Initial Segment, CETA prepared a Report to Congress on May 15, 2003 that among other topics describes the intermodal collision issue. That report is posted full text in pdf at the CETA web site, http://www.effectivetransportation.org.
Because trains are so much heavier than cars or even buses, using taxpayer dollars to build new opportunities for trains and motor vehicles to collide by accident is a bad idea. Also, even ignoring occasional collisions, trains can block roads, and road vehicles can block light rail train tracks, at points where the tracks meet the road.
The emerging focus for Link Light Rail Safety lies in whether Sound Transit is considered to be responsible for accidents and fatalities that are inevitably going to occur because of the design chosen by Sound Transit for the system. Sound Transit and Federal Transit Administration claim that collisions caused by driver error in the face of warning signs and signals are not “chargeable” to Sound Transit, even though grade separation and operational separation of rail and motor vehicle modes would completely prevent accidents of this type by making intermodal collisions impossible.
The Central Link design problems lie in the 18 ungated motor vehicle grade crossings planned for the Rainier Valley (map from Sound Transit), and also in the legal and illegal opportunities for pedestrians to cross the grade-level twin tracks. The operating plan in the Environmental Assessment shows that there will be 272 trains per day moving through these crossings. The system will start with two-car trains, but four-car trains approximately 360 feet in length are planned. There is a Quicktime video simulated animation of four car trains on the Sound Transit web site accessible by clicking here.
Potential problems also arise in the crossings and shared guideway in the Downtown Seattle Transit Tunnel, or Bus Tunnel, where the multi-mode mixture of buses and trains is a unique design creating some uncertainty about safety (diagrams of Tunnel merge points for buses and trains, from Sound Transit, pdf format). The light rail design for the Tunnel includes a shared right of way for buses and trains, a facility that in 2010 peak periods is planned to see 60 buses and 10 trains per hour intermixed in each direction under signal-assisted human control. A merge point at each end of the tunnel provides an opportunity for the buses to cross the path of oncoming trains. DSTT safety for mixed trains and buses has been audited by the US DOT Office of Inspector General, with a report made on July 7, 2003. Here is what the Inspector General reported on the collision hazard in the DSTT:
In August 2002, Sound Transit completed a hazards analysis using FTA’s “Hazard Analysis Guidelines for Transit Projects” to identify risks associated with joint bus and light rail operations in the Downtown Seattle Transit Tunnel. Sound Transit’s analysis evaluated the tunnel facilities, equipment, and operating plan to determine the type, frequency, and severity of potential hazards. The analysis identified fires and vehicle collisions as major hazards, but determined that the probability of these events occurring was very low.
To address these hazards, Sound Transit developed plans to retrofit the transit tunnel with specific safety improvements. These improvements include fire suppression, collision avoidance, ventilation and evacuation features. Our review found that Sound Transit’s plans, which were 60 percent complete at the time of this review, reasonably addressed safety hazards and other risks associated with tunnel operations….
With respect to vehicle collisions, Sound Transit plans to incorporate traffic control mechanisms in the joint operations plan to maintain safe separation between trains and buses. These traffic controls include an integrated signal system designed to allow only one mode (bus or rail) in a tunnel segment or station at any given time. The provisions to prevent vehicle collisions seem adequate for joint operations and are consistent with transit industry guidelines.
Nonetheless, the reliability of the collision prevention measures will ultimately depend on thorough testing of the controls and signal system prior to revenue operations. [Italics added.]
The ability of “traffic control mechanisms” to prevent grade crossing collisions does, however, require that sufficient time is available for a train to stop once a bus or other motor vehicle blocking the tracks is detected!
The Inspector General audit did not have safety outside the DSTT within its scope of work, so the audit report made no comment on grade-crossing safety in the Rainier Valley.
I was especially motivated to explore the safety issue after a ride I took on a relatively new street-running light rail train in Salt Lake City, Utah, summer 2002. During this trip on a bright summer day, I was looking over the train operator’s shoulder out the front. Watching motorists make left turns in front of this train under what I presume is state-of-the-art signalized control did not seem to me to be a safe addition to that urban environment. In fact, that light rail system has killed several people in collisions in the early years of operation. The dramatic effects of a train hitting a car are shown in photos available from a 2003 (fortunately, non-fatal) collision on the Utah light rail, posted on the CETA web, here.
I am also motivated by the long standing concerns of the Save Our Valley community group about the safety of the grade level Central Link route through the Rainier Valley in Seattle. A commentary on the hazards by George Curtis, President of Save Our Valley, is posted in pdf, here. It’s important to remember that the 100 year vision for light rail in this region includes at-grade tracks with street crossings in many communities surrounding Seattle.
In response to my testimony of February 27, Sound Transit staff made a presentation to the Board of Directors on March 27, 2003. The Sound Transit safety manager Hamid Qaasim asserted that the Federally mandated hazard analysis process has so far revealed that “the probability of a single system failure resulting in a critical chargeable accident is once in 131 to 13,000 years.” Mr. Qaasim further asserts, “This meets the FTA Guidelines.” The details of the hazard analysis are not published at this time.
Sound Transit regards some collisions as “not chargeable” to the agency. Given the official right-of-way design decisions made by the Sound Transit Board, the distinction between chargeable and non-chargeable is spurious. It is clear that if the light rail tracks were designed by Sound Transit to be grade separated from motor vehicle roadways, collisions between cars and trains would be impossible. It is further clear that if Sound Transit chose not to mix trains with buses in the Bus Tunnel, then collisions between buses and trains in that Tunnel would be impossible. Therefore, collisions between trains and motor vehicles of any type are certainly “chargeable” to Sound Transit and its light rail design.
Collisions between trains and motor vehicles are hazardous, and may seriously injure and kill people, whether they are “chargeable” or not. When it comes to collisions between trains and other vehicles, the only distinction that counts is between “possible” and “impossible.”
Joni Earl, Sound Transit Executive Director, sent an electronic mail message following the Board Meeting that summed up the light rail safety situation in Southeast Seattle, where light rail will run in an urban street median: She wrote, “In the Rainier Valley where light rail will run at grade, Sound Transit is increasing the signalized pedestrian-only crossings on Martin Luther King Jr. Way South from two to 10, increasing the signalized traffic and pedestrian intersections from 12 to 21, and providing 17 dedicated, signalized left-turn pockets. Warning bells, a safety zone, and protective railings will also be provided to increase safety at pedestrian crossings. The improvements for light rail will actually reduce the current number of accidents on Martin Luther King Jr. Way South and will make that corridor safer.”
Director Earl’s prediction of safety has no support in the environmental record. Safety was last covered in the 1999 Final Environmental Impact Statement. That document states that the addition of new traffic signals and street geometry in association with the light rail trains will reduce the number of vehicle-to-vehicle collisions by 44 per year, and the number of collisions between motor vehicles and pedestrians or cyclists will be reduced by seven annually. The 44 fewer vehicle-to-vehicle collisions is claimed by Sound Transit to be more than ample compensation for the 29 new collisions between trains and motor vehicles. The seven fewer times when vehicles hit pedestrians or cyclists is said to justify the three new annual cases of trains hitting people or cyclists.
However, there is a documented fatality rate for train collisions that exceeds the documented Rainier Valley fatality rate for collisions that only involve motor vehicles. CETA calculates that the historical experience of light rail trains and grade crossings at train speeds planned in the Rainier Valley indicates an expectation of eight fatalities per decade.
This logic of substituting one kind of accident for another kind is a transit industry invention embraced by Sound Transit. The common-sense idea that getting hit by a train may be more serious than being hit by a car is not discussed. The fatality and serious injury rates from the various types of accidents traded-off are not discussed in the EIS.
Furthermore, this kind of trade-off between accident types across modes is not covered in Federal Hazard Analysis Guidelines. There is no authority for adding new dangers to an urban street that are justified by reducing an existing form of danger.
King County Councilman and Sound Transit Board Member Dwight Pelz of Seattle supports the logic that any danger to careless motorists and pedestrians from adding light rail to the Rainier Valley is more than compensated by the forecast of reduced automobile and pedestrian accidents that don’t involve trains. He wrote in July 2004 in the Beacon Hill News & South District Journal:
Light rail is projected to decrease the number of accidents along Martin Luther King Jr. Way by six percent, with 18 fewer pedestrian and vehicle accidents annually than between 1994 and 1999.
All of light rail system’s vehicle and pedestrian crossings in the Rainier Valley are signalized. Sound Transit has increased the number of signalized pedestrian crossings from two to 10 and the number of signalized traffic intersections from 10 to 21, while providing 17 new left-turn pockets. The light rail tracks are on a raised median separated from other traffic by a curb, allowing pedestrians to pause in the middle of the street. Light rail signals will be integrated with the City’s traffic signal system to maximize safety. All these investments will serve to “calm” the frantic traffic on MLK.
Mr. Pelz’s calculation of 18 fewer accidents is probably based on the 1999 EIS statistics mentioned above: 44 fewer auto-to-auto collisions and 7 fewer pedestrian or bicycle collisions involving cars, offset by 32 new collisions involving light rail trains. (The arithmetic of 44 plus 7 minus 32 yields 19 fewer, not 18.)
To repeat, however, new train collisions are not compensated by reducing non-train collisions — not logically, and not legally.
And counter to Mr. Pelz’s conclusion about traffic calming, there is nothing “calm” about 272 trains per day at 35 mph down the middle of the street.
The new risks to human life stemming from the light rail grade crossing design in the Rainier Valley and in the planned joint bus-rail operations in the downtown Seattle Bus Tunnel would clearly be rated “unacceptable” in a safety certification process carried out by objective analysts. This finding would come from considering the difference in weight between trains and rubber-tire vehicles, the nationwide historical experience of inevitable collisions at grade crossings in urban settings, and the results of those collisions.
“Unacceptable” versus “acceptable” has nothing to do with whether the public thinks the risk is worth the benefit of having trains. There are absolute government and professional standards that are not related to system benefits. The Federal standard applying to Seattle’s light rail is to build a system sufficiently safe to create an expectation that there will be no more than one death in one million operating hours. (Sound Transit translates this many operating hours to 131 years.) There are no design compromises authorized to cause this standard to be relaxed.
At right is the cover of the Federal guidelines that set the hazard standards for light rail systems. Click the cover to access them in full text (pdf).
The word “chargeable” — as in “chargeable accident” — is nowhere to be found in this publication. These guidelines do state, however: “Unacceptable and undesirable risk will be reduced to an acceptable level before design acceptance, or a decision must be made to dispose of the system.”
As I said in my testimony to the Sound Transit Board on February 27, 2003, the danger from light rail trains is not to the people riding the trains, but rather to the entire community. Nationwide, the fatality rate per vehicle mile for light rail operations in recent years has been five to ten times higher than for bus transit operations, as shown in this chart:
Sound Transit staff presentation on safety to Board of Directors, March 27, 2003 (pdf file) State rail safety oversight program information from FTA (external link) Recommendation from CETA on safety oversight to Washington State Auditor, October 2002 (pdf file) Second letter to Federal Transit Administration on Link Safety Certification, October 2002 (pdf file) Initial letter to Federal Transit Administration on Link Light Rail Safety Concerns, June 2002 (pdf file) Paper from June 2002 describing the safety analysis of light rail added to the Seattle Bus Tunnel by Hamid Qaasim, Sound Transit System Safety & Quality Assurance Manager (pdf file) “Transit’s Safety Challenges” by John Semmens, presented at Transportation Research Board, January 2003 (pdf file) Report of the USDOT Highway/Rail Crossing Grade Crossing Technical Working Group, November 2002, including a discussion of grade separation of railroad crossings. Web site for USDOT Initiative on Safety Conscious Planning of transportation