1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 Engineering Safety Improvements for the Aurora Bridge Trevor Daviscourt University of Washington 1410 NE Campus Parkway, Seattle WA 98195 Tel: 425-213-2697 Email: trevorad@uw.edu Word count: 5487 words text + 10 tables/figures x 250 words (each) = 7987 words May 26th, 2018 Daviscourt 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 2 ABSTRACT The Aurora Bridge in Seattle, WA is a vital link in the transportation system for the region. However, the current lane configuration on the bridge deck is out of date and unsafe. Traffic on the bridge has increased six-fold since it opened in the 1930s, yet the design of the roadway has seen little change. The lanes are too narrow to meet current guidelines and there is no median safety barrier to prevent crossover collisions. Sadly, this archaic design may have contributed to the deaths of five students in 2015. With the goal of improving safety on the bridge, lane geometry and safety barrier guidelines have been consulted to construct four design alternatives. These alternatives were analyzed and compared. A 2003 WSDOT plan to reconfigure the bridge deck completely and a plan to reduce the number of travel lanes and install a median safety barrier were found to be the best solutions. Keywords: Safety Barriers, Lane Geometry, Urban Highway, Aurora Bridge, Seattle Daviscourt 3 1 2 3 4 5 6 7 INTRODUCTION Seattle's Aurora bridge, a crucial lifeline for the movement of people and goods through the city, is in need of a safety upgrade. The bridge was designed and built in the 1930s and while safety standards have been updated since then, the bridge has not, and injuries and fatalities continue to happen. The objective of this document is to investigate lane geometry and safety barrier improvements to the bridge deck to improve safety for motorists and passengers. 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 FIGURE 1 Little has changed on the Aurora Bridge between 1933 and 2015 (1,2). Officially named the George Washington Memorial Bridge, the bridge was opened on what would have been its namesake's 200th birthday, in 1932. The mayor of Seattle made sure to have representatives from Canada and Mexico join him during the opening ceremony as he hailed the bridge as "One more link in the concrete chain between Canada and Mexico" (3). The Aurora Bridge carries State Route 99, the first continuous, high-order road to connect all three countries. While State Route 99 has since been fragmented and overshadowed by Interstate-5, the bridge is now working harder than ever. When it opened there were around 11,000 vehicles per day crossing it, but today it carries over 70,000 per day (3, 4). The structural integrity of the bridge has deteriorated but it is still usable, and due to funding constraints and other priorities the bridge likely will not be replaced any time soon (5). In 2013 the bridge underwent a $10 million seismic retrofit (6). Sadly, for most of its history the majority of the deaths on the bridge have been suicides. The Aurora Bridge ranked second only to the Golden Gate Bridge for suicides until 2006 when a $6 million safety fence and other measures were put in place to reduce these incidents (7,8). After these increased safety measures were implemented, the real safety focus has become the configuration of the bridge deck. Currently, there are three lanes in each direction, ranging from 9-10 feet in width, and there is no median barrier dividing opposing travel lanes. This has resulted in a high level of crashes and fatalities, many of which could be avoided using existing research and technology. Daviscourt 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 4 CRASH HISTORY FIGURE 2 Aurora Bridge Crash Data 2013-2017 (9). The most recent available crash data obtained from the Washington Department of Transportation (WSDOT) from 2013-2017 shows consistent but upsetting data. Except for 2015, there were around 30 vehicles involved in collisions each year which resulted in around 10 injuries per year (9). The year 2015, however, saw the worst tragedy in the history of the bridge. On the morning of September 24, 2015, a northbound amphibious passenger vehicle, known as a duck boat, crossed into opposing traffic striking a southbound charter bus. The collision resulted in 5 fatalities, 30 serious injuries, and 41 minor injuries (10). Two other vehicles were not able to get out of the way and crashed as well. This was one of the worst traffic accidents in the city of Seattle. While the cause of the accident was determined by the NTSB to be mechanical failure of an axel on the duck boat operated by Ride the Ducks of Seattle, the median crossover collision would not have happened if there had been a median safety barrier in place. There likely would have been some injuries, but those five college students would still have been alive today. This is the motivation for this research. Daviscourt 5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 FIGURE 3 Scene from 2015 crash Note: Duck boat was originally traveling the opposite direction (10). Washington State and the Seattle Department of Transportation (SDOT) both have "Vision Zero" plans (11, 12). The city has a goal of ending all traffic deaths and serious injuries by 2030. This goal will likely never be obtained without safety improvements to the Aurora Bridge. Even without the Ride the Ducks crash, there are still several serious injuries on the bridge each year. Nothing has been done to improve the safety of the bridge since the crash, but fortunately there is a wealth of government-approved engineering solutions and several funding sources to finance such projects. ENGINEERING SOLUTIONS: RECHANNELIZATION AND SAFETY BARRIERS The safety record shows that something needs to change. Fortunately, there are many practical solutions to the problem. This paper will focus on four alternatives that utilize existing technology and proven research to improve safety on the bridge. All of the alternatives focus on realistic, low-to-moderate cost solutions that are politically feasible. The first alternative is a nobuild scenario. In order to measure whether or not subsequent alternatives are worthwhile, they must be evaluated against current conditions. The second alternative is a WSDOT proposal from 2003 that involves widening lanes, installing a median barrier, and moving pedestrian walkways below the bridge deck. The third alternative involves removing one of the northbound lanes, widening all lanes, and installing a median barrier. Finally, the last alternative involves reducing the number of lanes to five and installing a movable "zipper" barrier. These alternatives will be discussed in greater detail later. First, research into the effectiveness of lane widening and safety barriers will be outlined. Safety research is indeed an evolving field, yet everything being proposed here is from government-approved design guidelines and manuals. Most research into rechannelization and safety barriers has been known for years and most new research just confirms what is already in design manuals. Additionally, it is not legal, practical, cost effective, or efficient to experiment Daviscourt 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 6 with unapproved safety technology on the roadway. The need for this project is urgent and will be best achieved by using technology that is already DOT-approved. Lane width and median barrier recommendations are covered in great detail in just a handful of texts. This limits the amount of source data, but also reinforces the validity of the information as it is utilized nationwide and reflects the current state of the field. FIGURE 4 Approximation of current road geometry on the Aurora Bridge (6). Rechannelization in this context refers to widening and reconfiguring lanes on the bridge deck. This is a necessary and beneficial step to not only improve safety, but to improve traffic flow as well. Currently there are 3 lanes in each direction, ranging from 9-10 feet wide. These are not wide enough to meet the requirements of any relevant jurisdiction and are especially too narrow for the large volume of freight and transit vehicles that cross the bridge. The E Line, a pseudoBRT route connecting North Seattle to the Central Business District, is the busiest route in the city and crosses the bridge every 10 minutes during peak periods (13). This bus, combined with other popular routes, results in a nearly constant stream of buses on the bridge. King County Metro, the local transit agency, prefers that its buses have 11-foot lanes and most bus drivers straddle two lanes for safety as they cross the bridge. This shows that widening lanes will not only increase safety but traffic flow as well. Highway Capacity Manual (HCM) Exhibits 14-8 and 14-10 show that increasing lane width and adding a median barrier will increase free flow speeds by over 8 MPH (14). According to HSM, Highway Safety Manual Table 13-4, increasing lane width from 9 to 12 feet results in an average crash reduction of 25% alone (15). The HSM also states that there is no measurable difference in safety between 11 and 12-foot lanes, so 11foot lanes will be evaluated here. When the bridge was completed in the 1930s, transportation engineering guidelines were vastly different than they are today. City, state, and federal regulations have undergone vast changes Daviscourt 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 7 since then. In its Right of Way Improvement Manual, the Seattle Department of Transportation (SDOT) states that the through-travel lanes on urban city connector, freight, and transit roads need to be at least 11 feet wide (16). Similarly, WSDOT states that lane widths on this type of road should be 11 to 12 feet Federal regulations also show a preference for wider lanes. The Highway Safety Manual suggests 12-foot lanes as well, while 11-foot lanes may be acceptable (17). Section 7.3.3 that states that lane widths may vary from 10-12 feet in constrained areas. This is still greater than current lane widths on the bridge (17). The NTSB report on the Ride the Ducks crash ruled that road geometry was not a factor because of a loophole in the Road Side Design Guide. The loophole states that narrower lane widths may be allowed in some on roads with speeds of less than 45 MPH, saying that a narrower lane width is usually adequate in those cases and may have some advantages including increased capacity. This interpretation is misguided because that section is referring to 11-foot rather than 9-foot lanes (15). Wider lanes alone facilitate higher speeds and reduce crashes, and by dividing opposing lanes we can see even greater benefits. In addition to rechannelization, a median safety barrier should be installed to prevent crossover collisions like the Ride the Ducks crash. Median barriers are reinforced walls that physically prevent vehicles from crossing into oncoming traffic and are very effective. Barriers come in many varieties, but only reinforced concrete and steel barriers will be evaluated here due to their superior performance. The American Association of State Highway and Transportation Officials (AAHSTO) has shown that undivided multilane highways, such as the one carried by the Aurora Bridge, have much higher crash rates that those that are divided (18). It goes on to say barriers should be used on roads with high volumes and "a history of cross median collisions"(18). All median barriers should have a crash cushion at each end to mitigate vehicle strikes. As stated above, there are many types of median barriers, but only the concrete and steel barriers will be considered here due to the rate of heavy vehicles and trucks that cross the bridge. Median barriers are rated at different test levels, or TLs. A TL of three is sufficient to redirect cars and light trucks, but a TL of at least four is required for heavy vehicles (19). Zone of Intrusion (ZOI) is a measure of how much a barrier deflects when struck. This is an important measurement since a barrier may be forced into opposing traffic when struck hard enough, which is another reason why only the strongest barriers should be selected. While safety barriers do reduce the risk of a crossover collision, they increase the risk of collision with the barrier itself. This can be considered a net gain though, as these types of collisions tend to be less severe. FUNDING SOURCES Research is essentially meaningless unless it can be put into use. Nothing gets planned, designed, or built without capital. Fortunately, there are ample federal, state, and local funding sources available for road safety improvements. A selection of the most relevant funding sources is presented here. Federal money for safety improvements comes from the FAST Act through the Federal Highway Administration (FHWA). The FAST Act, which raises money from document fees rather than gas taxes, has a pool of $305 billion and is available from 2016-2020 (20). This program is focused on safety improvements, and funding is available to cities, states, and MPOs. There is also federal money doled out to states through the Strategic Highway Safety Plan. This is a Daviscourt 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 8 statewide plan for identifying and evaluating safety problems and finding solutions (20). Funding from the state level comes from WSDOT. One problem with funding for this project is that it is not entirely clear who should have to pay for it. The bridge is owned by WSDOT but operated by SDOT. Currently, the city and state are in court debating who should pay for safety improvements to the bridge (21). That being said, cities can apply for safety funding from WSDOT though the City Safety Program (22). The purpose of the City Safety Program is to "reduce fatal and serious crashes on city streets using engineering improvements and countermeasures". This project would be an ideal candidate for funds under this program. The next call for proposals is January 2020, which would give the city ample opportunity to plan and prepare. The program will pay for initial design, engineering, and construction. Projects require a 10% local match for each of the three phases, but projects authorized by 2021 are eligible for 100% construction funding with no local match. There is no maximum amount of funding, so all of the alternatives listed could be potential candidates. SDOT provides funding for road safety improvements as well. Move Seattle is a program which started in 2015 and holds five values: to provide a "Safe, interconnected, vibrant, affordable, and innovative city" (23). In order to fund these goals, the Transportation Levy to Move Seattle was established. This is a 9-year, $930 million program to increase safety, improve bridges, and provide affordable transportation (23). To reach the goal of a safe city, their goal is to "keep Seattle travelers safe by working to eliminate serious and fatal crashes and seismically reinforcing vulnerable bridges" (23). The program also strives to improve important corridors and modernize and repave 35% of the city’s busiest arterials for freight and passenger movement (23). Improving safety on the Aurora Bridge ticks all of these boxes as it is an outdated bridge, a deadly road, and is part of one of the city's most important passenger and freight corridors. With funding available from all levels of government, the alternatives presented here can all be considered financially feasible. Daviscourt 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 9 DESIGN ALTERNATIVES There are countless solutions for improving safety on the Aurora Bridge. These have been narrowed down to the most cost effective, realistic, effective, and politically feasible alternatives and they are briefly described below. Each alternative is then compared and evaluated later. Alternative 1: No-build FIGURE 5 Cross section of no-build alternative (24). The first alternative, as with any project, is a no-build alternative. If the bridge is left in its current configuration, it can be expected that crash rates will continue at current levels or increase over time. Nationally, traffic safety is getting worse as distracted driving continues to be a growing problem. This may be an explanation for the slight uptick in crashes on the bridge in 2017 (9). Seattle is one of the fastest-growing cities in North America, and with increased population comes congestion, accidents, and new drivers who may not be used to such an unusually narrow lane design. If the bridge is not modified, expanding bus service along State Route 99 will be delayed as the lanes are not configured properly for buses, and the City of Seattle will likely never reach its vision zero goal. Money may be saved in construction and maintenance, but it will be lost down the line, as the state values the societal and actual cost of a traffic death at several million dollars. Daviscourt 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 10 Alternative 2: 2003 WSDOT Plan FIGURE 6 Proposed cross section from 2003 WSDOT plan (5). In 2003, WSDOT had a plan for improving the bridge. This was before the suicide barrier fence was installed and the highlight of the plan involved placing pedestrian and bicycle paths under the bridge deck and fully fencing them in. Once the walkways were removed, the right-of-way for traffic would open up and they planned on placing 3 full size lanes in each direction on the bridge, with a median barrier dividing opposing traffic. The outside lanes would be specially designed with freight and transit vehicles in mind. The median barrier would be a standard, reinforced concrete design. The plan also included intersection improvements in nearby neighborhoods to prevent traffic from backing up onto the bridge deck. This is not included in other alternatives listed here for funding qualification reasons but should be considered as part of a holistic solution to congestion on the bridge. This plan meets all local, state, and federal design guidelines for lane widths, which increases safety and throughput. A highlight of the plan is that 3 travel lanes in each direction can be retained so that throughput is maximized. This alternative is by far the costliest at an estimated $28 million in 2003, and today the cost would likely be substantially higher. Daviscourt 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 11 Alternative 3: Conventional Barrier FIGURE 7 Cross section of conventional barrier alternative (24). The least costly alternative, other than the no-build, is installing a concrete barrier. These barriers come in many forms, but all have among the highest crash test levels of any barrier, at a four to five rating. To make room for a barrier, wider lanes would be added, and existing pedestrian and bicycle facilities would be retained. Adding a concrete barrier would require the lane configuration to change from three lanes in each direction to having three southbound lanes and two northbound lanes. Tidal flow and directionality are discussed later in this document, but suffice it to say that this lane configuration may be sufficient. As discussed above, throughput and speed can be increased with wider lanes, so congestion may not increase as much as some experts might fear. According to WSDOT, these barriers cost between $42-72 per foot which works out to $124,000-212,000 for the length of the bridge (25). Crash cushions will need to be installed on either end as well. Installing a concrete barrier and restriping the roadway could also be an interim alternative to protect travelers while a more complex solution is being planned. Daviscourt 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 12 Alternative 4: Movable Barrier FIGURE 8 Cross section of movable barrier alternative (24). A movable safety barrier, know under the trade name Road Zipper, is a series of connected steel and concrete barrier segments that can be moved the width of a lane in live traffic with a barrier transfer machine. The barrier transfer machine is run by two operators at a speed of about 10 mph. This allows for fully protected, reversible lanes to assist in tidal traffic flows. The machines have been used all over the world for traffic management and road construction. They have been proven to be most advantageous in lane management for bridges and tunnels. Notable examples including: The Golden Gate bridge in San Francisco, the Lions Gate Bridge in Vancouver, and the Auckland Harbor Bridge in Auckland, New Zealand (pictured below) have shown this system can be successful in climates similar to Seattle. While the cost of these systems is high, they have great safety performance and have the effect of adding capacity to facilities where adding lanes is impracticable or impossible. According to the manufacturer, Lindsay Transportation Solutions of California, these barriers perform better than a conventional reinforced concrete barrier (26). They have seen no crossover fatalities on any facility with their barriers. The movable barrier has other advantages as well. There is no harm to the bridge deck because there is no need to anchor each segment. Additionally, the barriers have very little deflection when struck, only around 2 feet, so encroachment into oncoming lanes is minimized during a collision. This is advantageous for the Aurora Bridge because there is not a lot of extra room for a median around the barrier. Daviscourt 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 13 Costs for this type of barrier, however, are much higher than with a conventional concrete barrier. According to the manufacturer, an 18-inch wide movable barrier costs $351 per foot (26). At 2,946 feet long, the barrier alone would cost $1,034,046, plus the cost of crash cushions on each end. Barrier transfer machines cost around $1,800,000 plus operating costs. The manufacturer estimates the costs of two operators at $250/hour, 5 days per week, 52 weeks per year is around $400,000 each year. A big labor concern is what to do with the operators between movements. Labor cost could possibly be reduced if bridge tenders from nearby draw bridges were trained to operate the machine. Draw bridges do not open during peak periods in Seattle, so this may be worth considering if this option is used. The main consideration in whether to use a movable barrier system is if the traffic is tidal enough to justify a lane management system or not. The only publicly available traffic data for the bridge shows total daily vehicle counts. A proper traffic study would need to be completed to determine directionality during peak periods. From casual daily observation, it appears that the traffic seems to be consistently congested in the southern direction and generally free flowing northbound throughout the day. There is a 7-way stop sign-controlled intersection that has a queue that can extend out onto the bridge deck in the southbound direction during the PM peak. This would probably make a reduction to two lanes in that direction unfeasible. Research by Rathbone et al. shows that there needs to be around a 64/46 split in directionality to justify implementing a contraflow system with this lane configuration (27). FIGURE 9 Movable barrier system on the Auckland Harbor Bridge (27). Daviscourt 1 2 3 4 5 6 7 8 9 10 14 ANALYSIS OF ALTERNATIVES FIGURE 10 Alternatives analysis table. The alternative analysis table was created using an amalgam of advice and examples from WSDOT and the FHWA. Four categories are considered for each alternative: safety, cost, mobility, and societal factors. Each subcategory was given a number on a scale between 0 and 10. These numbers reflect the performance of each alternative in accordance with guidelines and Daviscourt 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 15 recommendations shown in this document and in relation to one another. Each section has a weighted score on par with its importance and the number of subsections. The results of the alternatives analysis table are discussed below. As with any analysis table, this is subject to the judgment of the author. As stated above, weights and values were assigned to reflect the relative qualities and importance of each alternative. These can be very subjective. It is quite possible for another person to interpret the research differently and come up with quite different results. The table should be used as a tool for comparing the alternatives using the evidence in this text, not as the final measurement for the alternatives. An advantage of the table is that it can be easily modified to reflect new priorities and evaluations as the process moves forward. This table could be brought to meetings as a presentation tool to demonstrate the evaluation process, and is designed to be easily interpreted by the general population. Thus, while it is clear and easily understood, it is not very robust. That being said, the results of the table are supported by the research cited in the text of this document. It is up to the reader to determine if they believe that weights and values are correct given this information. Safety is the first and most important consideration in every transportation engineering project. The purpose of a safety barrier is to prevent head-on collisions. All the build alternatives include barriers of the highest safety rating, which should eliminate nearly all crossover collisions if implemented. These are given a score of 10. Other collision types include side and rear collisions for vehicles traveling in the same direction. Adding a median barrier does little or nothing to prevent these types of collisions, but the build alternatives all include widening lanes which can reduce these types of collisions significantly. Fixed object collisions increase somewhat when a median barrier is put in place without a wide median. All of the build alternatives increase this risk, but the movable barrier option is penalized the most because of possible collisions while the barrier transfer machine is in operation. Cost is the second most important factor when evaluating this project. The original WSDOT plan from 2003 would have prevented the Ride the Ducks collision, but it was not implemented due to high costs. Even the best solutions are often not considered for this reason. Cost factors are scored with a 0 being the most expensive and a 10 being the least expensive. Planning costs are nothing for a no-build scenario, and not too high for WSDOT’s plan because a study was already completed in 2003. Planning costs for the conventional and movable barrier plans should not be too high as these are relatively minor projects. Construction costs are by far the highest with the WSDOT plan, and costs for the other alternatives are outlined above. Maintenance and operational costs increase for the WSDOT plan because of the added structure under the bridge deck, while the movable barrier alternative has quite high daily costs for maintaining and operating the equipment. The mobility section of the table considers how each alternative affects movement and traffic. Throughput will possibly stay the same between the no-build and conventional barrier alternatives as gains from wider lanes may be offset by the loss of a northbound lane. Currently, the bridge does not support transit, freight, and emergency vehicles well because the lanes are too narrow for them to safely travel in only one lane. The build alternatives all address this. Pedestrians and cyclists make up a very small proportion of traffic on the Aurora Bridge. There Daviscourt 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 16 are already six footpaths on either side of the bridge with guard rails and a suicide prevention fence. However, the WSDOT plan would make things much safer for them, with 10-foot paths that are grade-separated and fenced in. Free flow speed is increased as lane width increases. The accident/construction resilience subcategory evaluates how traffic is impacted when there is an incident or work being done on the bridge. Currently, this is very poor due narrow lanes and the lack of a shoulder. The movable barrier alternative is the best for addressing this as channelization can be adjusted on the fly. Movable barriers are popular for road construction for this reason. Congestion relief is best with six lanes, and pretty good with a movable lane option. Comfort is a subcategory that evaluates how comfortable people feel when driving across the bridge. Currently, many people slow down, and merge right, for fear of accidents. If a median barrier is added, driver comfort should increase. The societal factors category evaluates less tangible elements with any of the alternatives. Vision Zero Attainment measures how effective the alternatives are at helping the city and state reach their Vision Zero target. The only fatalities on the bridge in the last five years have been from a head on lane-crossover accident, so adding a median barrier is essential for reaching the goal (9). Environmental impact is difficult to determine with a project like this. In this context environmental impact is really a measure of the reduction of idling emissions from lowering accident rates and resulting congestion. Currently, public approval of the bridge is very low, and there have been countless complaints and articles written about the safety of the bridge both before and after the Ride the Ducks crash. Any safety improvement should garner public support, though the conventional barrier alternative might face criticism for losing a lane, and some may be opposed to the costs of any build plan. Political feasibility was a latent factor that needed to be included. This ties in heavily with cost but also accounts for public perception. Finally, the weighted totals were combined. Interestingly, there is a tie for the best alternative and a virtual tie for the worst alternative. Both the WSDOT plan and the conventional barrier alternative scored highest. The WSDOT plan excels in performance, but is hindered by cost. Conversely, the conventional barrier alternative is relatively low cost, but there is a risk for northbound congestion. The movable barrier and no-build alternatives are not recommended. Movable barriers have too high of capital and operating costs, and the traffic may not be tidal enough to justify reversible lanes. The no-build alternative is naturally the easiest and least expensive option, but it should not be considered a final plan. The continued risk for danger and the outdated design are too high of a threat to motorists to be ignored. CONCLUSION The Aurora Bridge is a Seattle icon that has been marred by tragedy for too long. By using existing technologies, a median safety barrier can be put in place to save lives. The opportunity is also ripe for rechannelization, and the bridge could be made more efficient and safer at the same time. There are several federal, state, and local funding sources for safety improvement and this project appears to meet the criteria for all of them. The WSDOT plan and the conventional barrier alternative both scored well during analysis. These two plans represent a high cost/high reward and a low cost/medium reward option for the government to implement. Regardless of which plan is chosen, it is clear that something needs to be done soon to prevent further tragedies from occurring on the bridge. REFERENCES Daviscourt 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 17 1. Sherrard, J. R. (2015, March 28). SEATTLE NOW & THEN: “MURDER” ON AURORA. Retrieved April 20, 2018, from https://pauldorpat.com/tag/aurora-bridge/ 2. Kroman, D., Davis, B., Oliver, N., Chasan, D. J., Carlson, J., Upchurch, M., & Capozziello, N. (2015, September 24). NTSB team leading investigation into Aurora Bridge crash. Retrieved May 4, 2018, from https://crosscut.com/2015/09/2-die-in-aurora-bridge-crash-of-duck-vehicle-charterbus-with-college-students 3. Long, P. (2015, September 24). Seattle's George Washington Memorial Bridge (Aurora Bridge) is dedicated on February 22, 1932. Retrieved March 30, 2018, from http://www.historylink.org/File/5418 4. WSDOT - Travel Data. (2018, March 15). Retrieved March 15, 2018, from http://www.wsdot.wa.gov/mapsdata/travel/traveldata.htm 5. WSDOT. (2003). 2003 WSDOT Corridor Study. Retrieved March 10, 2018, from http://www.wsdot.wa.gov/NR/rdonlyres/17DF1C94-C244-48EA-978E0F91153C815A/0/SR_99_NorthFinalReport.pdf 6. Lindblom, M., & Baker, M. (2015, September 24). Span's narrow lanes a longtime safety concern. Retrieved May 22, 2018, from https://www.seattletimes.com/seattlenews/transportation/spans-narrow-lanes-a-longtime-safety-concern/ 7. Seattle FRIENDS. (n.d.). Seattle FRIENDS Aurora Bridge Suicides. Retrieved May 11, 2018, from http://seattlefriends.org/ 8. O’Sullivan, J., & Lindblom, M. (2015, September 26). Lawmakers seek ways to improve Aurora Bridge safety. Retrieved May 12, 2018, from https://www.seattletimes.com/seattlenews/transportation/lawmakers-take-action-to-improve-safety-on-aurora-bridge/ 9. Joblonski, T. (2018). Standard Crash Summary Report Aurora Bridge (Rep. No. PDR-18-1359). Tumwater, WA: WSDOT Crash Data Reporting Branch. 10. NTSB. (2018). Amphibious Passenger Vehicle DUCK 6 Lane Crossover Collision With Motorcoach on State Route 99, Aurora Bridge Seattle, Washington September 24, 2015 (Rep. No. Accident Report NTSB/HAR-16/02 PB2017-100408). Washington D.C.: NTSB. 11. WSDOT. (n.d.). Target Zero - Washington State Strategic Highway Safety Plan. Retrieved from http://www.targetzero.com/ 12. Vision Zero. (n.d.). Retrieved from https://www.seattle.gov/visionzero Daviscourt 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 18 13. McNichols, J. (2018, April 14). What it's like to commute every day on Seattle's most crowded bus. Retrieved from http://kuow.org/post/what-it-s-commute-every-day-seattle-s-mostcrowded-bus 14. (2010; 2014). Highway Safety Manual (1st Edition) with Supplement 2014. American Association of State Highway and Transportation Officials (AASHTO). 15. (2010). Highway Capacity Manual 2010, Volumes 1 - 4, including 2014 Errata. Transportation Research Board. 16. SDOT. (2018). Right-of-Way Improvements Manual (Seattle, Seattle Department of Transportation). Seattle, WA: SDOT. 17. WSDOT. (2018). Design Manual (01st ed., Vol. 22, Ser. 14) (Washington State, Washington State Department of Transportation, Development Division, Design Office). Olympia, WA: WSDOT 18. (2011; 2012; 2013). Policy on Geometric Design of Highways and Streets (6th Edition) with 2012 and 2013 Errata. American Association of State Highway and Transportation Officials (AASHTO). 19. (2011; 2015) Roadside Design Guide (4th Edition). American Association of State Highway and Transportation Officials (AASHTO). 20. FHWA (Ed.). (n.d.). The FAST Act. Retrieved from https://www.fhwa.dot.gov/fastact/ 21. Cohen, G. (2018, May 24). Judge denies Seattle motion to avoid Ride the Ducks lawsuit. Retrieved from http://komonews.com/news/local/judge-denies-seattle-motion-to-avoid-ridethe-ducks-lawsuit 22. WSDOT (Ed.). (2018, April 18). 2018 City Safety Program: Call for Projects. Retrieved from https://www.wsdot.wa.gov/LocalPrograms/Traffic/2018CitySafetyProgram.htm 23. SDOT (Ed.). (2015, March). Levy to Move Seattle FAQ. Retrieved from https://www.seattle.gov/transportation/about-sdot/funding/levy-to-moveseattle/faq#howisthislevydifferentfromthebridgingthegaptransportationlevy 24. The Northwest Urbanist. (2015, September 29). Seattle's Aurora Bridge Needs a Safety Redesign. Retrieved from https://thenorthwesturbanist.com/2015/09/24/seattles-aurora-bridge-needs-asafety-redesign/ 25. WSDOT (Ed.). (n.d.). Traffic Median Protecting Yourself Brochure. Retrieved from http://www.wsdot.wa.gov/publications/fulltext/design/RoadsideSafety/TrafficMedian.pdf 26. Sanders, C. (2018, May 5). Aurora Bridge Paper [E-mail to the author]. Email correspondence with Chris Sanders, Senior Vice President of Lindsay Transportation Solutions Daviscourt 1 2 27. Lindsay Transportation Solutions. (2017). Improving Cost-Effectiveness of Urban Freeways[Brochure]. Rio Vista, CA: Author. 19