How to Fix PennDOT’s Fern Hollow Bridge Design

PennDOT and the City of Pittsburgh are rushing to build a new Fern Hollow Bridge to replace the one that collapsed in January 2022. PennDOT released a conceptual rendering of the new bridge design on March 8, 2022. Unfortunately, the rushed process has resulted in some serious design flaws that need to be addressed before construction is complete.

With improvements, the bridge could be safer, have capacity for more travelers, and add beauty to Frick Park.

Conceptual Rendering of the New Fern Hollow Bridge

A Brief History of the Fern Hollow Bridge

The original Fern Hollow Bridge was built in 1901 before the creation of Frick Park. It was built for \$100,000, which is \$3,337,835 adjusted for inflation1This calculation was based on 1901 dollars inflated to 2022 dollars. The Consumer Price Index was not created until 1913, but this website uses a historical study to estimate inflation prior to 1913.. It included 6.5 ft wide sidewalks on each side, and four lanes of motorized traffic including two lanes of streetcar tracks.

Frick park opened in 19272As described by Historic Pittsburgh, Henry Clay Frick bequeathed 151 acres of land for the park in 1919, which opened on June 25, 1927., and the gatehouse on the Southwest side of the bridge was built in the 1930s3Frick Park’s beautiful gatehouses were designed by the famous architect John Russell Pope..

Fern Hollow Bridge from 1901 to 1972

In 1973, the bridge was replaced with the structure that collapsed in 2022. This bridge won an award for the most beautiful medium-span, high-clearance bridge to open in 1973. The bridge was built for \$1,230,0004The cost is documented in this Post Gazette article.. This corresponds to \$8,408,932 adjusted for inflation5The bridge built in 1901 was closed for replacement in April 1972. The inflation adjusted price of \$8,408,932 was based on the Bureau of Labor Statistics CPI Inflation Calculator using April 1972 prices inflated to February 2022 prices.. This is substantially less than the \$25.3 million that has been budgeted in 2022 for the new replacement bridge6The replacement budget can be found here..

Aerial View of 1973 to 2022 Fern Hollow Bridge
Ground level view of the 1973 to 2022 Bridge
Additional view of the 1973-2022 Bridge.

The bridge had sidewalks on both sides with barriers between the motor vehicle lanes and the side walks. A narrow shoulder was provided for cyclists.

Street view of 1973 to 2022 Bridge.

The Bridge Needs to be able to Transport More People

The Fern Hollow Bridge sits at a traffic bottleneck. Frick Park limits the number of access routes to Pittsburgh from the East. There are only three roads South of Penn Avenue and North of the Monongahela River that allow travelers to go from locations East of Frick Park to locations West of Frick Park (or vice versa): the I-376 Parkway, Forward/Commercial Avenue, and Forbes Avenue traversing the Fern Hollow Bridge. People from the Eastern suburbs travelling to Oakland, Downtown or going West of the city either need to detour to Penn Avenue or take one of these three routes. As a result, all three of these approaches are often congested.

Given the importance of the Fern Hollow Bridge route it is important to be able to maximize the number of travelers who are able to cross the bridge.

Sidewalks, Bike Lanes and Bus Lanes can Move More People

Bicycle lanes and bus lanes have the capacity to move far more people than an automotive lane. Imagine standing by the side of the road and counting the number of travelers that can pass in an hour before congestion starts to significantly slow travel speeds. That’s the capacity. A 1 m (3.3 ft) bike lane has a capacity of 2,600 bikes per hour. A bus lane can move 6,000 passengers per hour. The feature with the greatest capacity is the humble sidewalk. An 11 ft wide sidewalk has a capacity of 15,000 people per hour7Lane capacity data is taken from this article in The Urbanist..

In contrast, the capacity of a 10-12 ft wide freeway lane is 2,200 cars per hour, and as will be discussed in the next section traffic lights have a lower capacity per lane.

In order for bike lanes and sidewalks to be used, they need to be connected to a network connecting origins and destinations.

This stretch of Forbes connects Squirrel Hill to Regent Square, both of which have excellent sidewalk networks. While Pittsburgh’s bike network is currently more limited, the Pittsburgh Bike(+) Master Plan will result in a significant bike network expansion. The bike connection along Forbes between Squirrel Hill and Regent Square is a key segment to Pittsburgh’s future bike network. As Pittsburgh’s bike network expands, more riders will take advantage of a bike lane over the Fern Hollow Bridge.

Reducing the number of automobile lanes would allow for better pedestrian and bicycle infrastructure and for the addition of a bus lane. Traffic data shows that two of Forbes automobile lanes are underutilized. Therefore, even if utilization of an additional sidewalk, bike lane or bus lane is low, reallocation of space would result in more people being able to cross the bridge.

Traffic Data Shows that Forbes does not Need Four Lanes

The number of lanes on Forbes is not the limiting factor to the volume of car traffic. The main bottleneck is the Forbes/Braddock intersection, with the Forbes/South Dallas intersection being a secondary bottleneck. The intersection of Forbes Avenue and Braddock Avenue is on the East/Outbound side of the bridge. The intersection of Forbes Avenue and South Dallas is on the West/Inbound side of the bridge. There is currently a cut-through for cars where Beechwood Boulevard touches Forbes just East of South Dallas. Beechwood Boulevard then intersects Forbes West of the South Dallas intersection (see map).

Between the intersection with Beechwood Boulevard and the intersection with Braddock Avenue, Forbes Avenue has four lanes. However, other than this segment Forbes is only two lanes, as are all of the roads that intersect with it. Furthermore, at the intersection of Braddock and Forbes, each green signal phase only allows one lane of cars to enter Forbes inbound at a time.

There is an inbound bus stop just West of the Forbes/Braddock intersection. Having two inbound lanes until this point allows cars to pass stopped buses and prevent a backup from impacting the intersection at Braddock. Having two inbound lanes at the inbound approach to South Dallas also adds capacity. However from the bus stop to the Beechwood cut-through (a distance of approximately 3,500 ft), the second inbound traffic lane adds no capacity for cars.

While this lane does allow cars to pass slower moving vehicles, the time savings this provides to motorists is minimal and this lane is an inefficient use of space. Eliminating this travel lane would provide space for better and safer pedestrian and bicycle infrastructure.

The second westbound traffic lane may provide some additional capacity. Inbound traffic at the South Dallas intersection has a left turn signal during which outbound traffic cannot proceed. This reduces the duration of the green signal for outbound traffic going straight. Having two lanes on the outbound side of Forbes at both sides of the intersection with South Dallas increases the number of cars that can pass through the reduced duration green phase. Therefore, if only one traffic lane is removed it should be the inbound lane with the outbound retaining two motor vehicle lanes.

However, available traffic counts on the Fern Hollow Bridge show that only one lane in each direction is needed to handle the traffic. The PennDOT traffic count database contains a traffic count for the bridge collected on October 30, 20198The applicable traffic monitoring report can be found here..

The Federal Highway Administration provides lane capacities for different types of road. The capacity of a freeway is 2,200 cars per hour per lane. For signalized highways, the capacity is $1,900\times\frac{g}{C}\times\textrm{lanes}$ where $\frac{g}{C}$ is the proportion of signal cycle time that is green.9Capacity numbers are based on the Simplified Highway Capacity Calculation Method for the Highway Performance Monitoring System published by the Federal Highway Administration. The formula for freeways includes an adjustment to reduce the capacity based on the percentage of vehicles that are trucks. The PennDOT traffic count data shows that only 5.1% of the Forbes traffic is from trucks which reduces the capacity by 4.9%.

The PennDOT traffic counts show that most of the outbound traffic at this intersection enters from South Dallas. The traffic counts on Forbes West of the South Dallas intersection are too low to justify two outbound lanes10While there are no traffic counts on the segment of Forbes between Beechwood and South Dallas, the TMS site 8051 located just West of the Forbes/Beechwood intersection counted an hourly peak of only 339 outbound cars on September 13, 2017..

Therefore, the number of outbound general travel lanes could also be reduced to one with minimal impact to car traffic. While this additional space could be used for additional bicycle or pedestrian space, or traffic calming and beatification features such as median street trees, there is space good bicycle/pedestrian infrastructure and three motor vehicle lanes. Therefore the best use of this space would be for a bus lane to improve transit service.

Reducing the number of motor vehicle lanes to three11This could be one inbound and two outbound general traffic lanes. Or it could be one general traffic lane in each direction as well as a single bus lane. would reduce the amount of weight the bridge needs to carry12Or to use more precise terminology, it would reduce the necessary live load rating. The live load is the weight of a structure’s nonpermanent, moveable parts, contents, or “users,” such as the traffic, and people on a bridge.. This should provide substantial cost savings that would ideally be used to make improvements to Forbes Avenue leading up to the bridge, allowing the bridge itself to be more fully utilized.

Safety for Pedestrians and Cyclists is Essential

Before it collapsed, Fern Hollow Bridge and the adjacent segment of Forbes Avenue served fewer pedestrians and cyclists than it had the potential to because it was and felt dangerous to these users. The speed limit is 35 mph, but cars commonly drive at much higher speeds. These speeds are deadly to vulnerable road users, and no protective infrastructure separated cyclists and pedestrians from fast moving vehicles.

The plan for the new bridge provides modest safety improvements for cyclists, but may reduce safety for pedestrians. The city plans to reduce the lane width from 12 ft to 10 ft and allocate the recovered space for a 10 ft wide shared use path13A shared use path is typically used to describe bike trails like the Greater Allegheny Passage that have a high level of separation from motor vehicle traffic. A buffered bike lane may be a better term for the design that has been described. primarily for cyclists on the outbound side of Forbes. The reduction in lane width will calm traffic, improving safety. The shared use path will be separated from cars by flex posts from South Dallas to the bridge, and by a crash worthy barrier over the bridge. There will no longer be any barrier on the bridge between the sidewalk and the car lanes on the inbound side of the bridge14The details in this paragraph are based on a presentation given by Nick Ross from the city’s department of mobility and infrastructure to the Squirrel Hill Urban Coalition Bike and Pedestrian Committee on March 23, 2022. The city is responsible for the improvements on Forbes between South Dallas to 500 ft of the bridge. The city has budgeted \$500k for these improvements..

The shared use path would run until where Forbes Avenue meets Beechwood Boulevard. This connection would be closed to motor vehicle traffic and an enhanced crossing for bikes and pedestrians would be added.

The street space allocation proposed by the City of Pittsburgh is as follows:

These design changes overall represent an improvement compared to the previous bridge. It provides better accommodations to cyclists and pedestrians than most bridges designed by PennDOT. However, the design is still dangerous enough for pedestrians and cyclists that it will deter travelers from using these modes. It does not provide the level of safety that these users have a right to.

A Crash Worth Barrier is Needed to Separate Cars from Vulnerable Users

At speeds above 35 mph, there is a high risk of death or serious injury if a car hits a pedestrian or cyclist15This ProPublica article provides a good overview of the hazard vehicle speed represents to pedestrians.. Without a crash worthy barrier, many pedestrians and cyclists will avoid this route due to safety concerns. It is scary to walk with nothing to separate you from cars flying by at 40 mph.

While flex posts provide a better visual demarcation of the bike lane than paint on the roadway, they don’t provide any actual crash protection. They are essentially vertical paint. Additionally, they are not durable. If they are not promptly replaced by city maintenance after being hit by cars, they stop providing even visual demarcation.

Flex posts that have been knocked over by cars.

Infrastructure that provides actual protection to cyclists and pedestrians includes crash worthy guard rails, bollards, planters or street trees. This quality infrastructure should be installed on both sides of Forbes to protect both the inbound sidewalk and the outbound bike lane. There is not sufficient space for this protective infrastructure and four motor vehicle lanes. Therefore eliminating one inbound travel lane is necessary to provide safe infrastructure for cyclists and pedestrians.

If there is insufficient funding for these items, curb protection or parking block protection would be an improvement over flex posts. Even though a curb is not a crash worthy barrier, a Federal Highway Administration report concludes that adding a curb protected sidewalk reduces the likelihood of a pedestrian being struck by 88% compared with a shoulder marked with paint16The Federal Highway Administration bike lane design guide has a good overview of the forms of separation for a protected bike lane. While I could not find a specific study quantifying the safety benefit of using parking blocks, the FHA study on the benefit of sidewalks is evidence that parking blocks provide better protection than flex posts alone. The sad truth is that sometimes it takes creating a risk of property damage to driver’s vehicles for them to drive with sufficient caution to create a safe environment for pedestrians and cyclists..

A parking block protected bike lane

The bridge design shows an unacceptable difference to the safety standards afforded to motorists compared with pedestrians and cyclists. Bridges are built with factors of safety 17The Engineering Toolbox states that the factor of safety for bridges is typically between 5 and 7. This means that the bridge must be designed so that the failure load is at least 5 times the allowable load for the bridge. that make collapses extremely rare. In the past five years, there have been no motor vehicle bridge collapses in the United States that have resulted in fatalities18Wikipedia has a list of bridge failures which records only four bridge failures in the United States between 2017 and 2022, including the Fern Hollow Bridge collapse. The only collapse that resulted in fatalities was a pedestrian bridge that collapsed while still under construction, killing six people. .

The need to meet safety factors required by standards will add millions of dollars to the cost of the Fern Hollow Bridge replacement. If funding wasn’t available to meet standards the bridge would not be built no matter the delay to motorists. Motorist safety is prioritized over motorist convenience.

By contrast, the standards for infrastructure to protect vulnerable road users from cars is comparatively lax. Car crashes killed 6,700 pedestrians in 202019According to Axios.. However most would consider it unthinkable to close a road to cars until resources are available to build sidewalks or guard rails. Yet allowing cars on roads without adequate bike and pedestrian infrastructure creates a greater risk of fatalities than opening a substandard bridge. Doing so results in thousands of preventable deaths every year.

Pedestrians and cyclists have a right to a safe right of way. If they are not provided one, it effectively puts people under house arrest for the crime of not being able to afford a car, or the crime of not being willing or able to drive. Walking and cycling are a right, driving is a privilege.

Pedestrians need Dedicated Sidewalks without Cyclists on Both Sides of Forbes

While bikes don’t represent the same level of safety hazard to pedestrians that cars do, providing separate facilities for pedestrians and cyclists improves safety. Many pedestrians would avoid a shared use path that is primarily for cyclists for fear of being hit. The outbound side of Forbes is closer to the Regent Square business district and also connects to the Frick Park trail network by the outbound side gatehouse. Outbound bus stops on Forbes by the bridge also require pedestrian access. Providing a dedicated pedestrian facility on the outbound side of Forbes is important.

The new street design should therefore have dedicated sidewalks on both the inbound and outbound side of Forbes. There should be a 10 ft wide two way bicycle lane on he outbound side of Forbes over the bridge to the Beechwood cut-through as proposed by the city. Crash worthy barriers should be provided on both sides of the street between the motor vehicle lanes and the bicycle/pedestrian facilities. Since the bike lane ends at the Beechwood cut-through, there is space to preserve the two inbound motor vehicle lanes approaching the South Dallas intersection.

Bike Lane Placement

Locating both directions of bike lane on the outbound side of the street is the best location with the existing bike lane network. It allows cyclists travelling in both directions good access to the existing bike lanes on Beechwood Ave.

However, longer term placing the 5 ft inbound bicycle lane on the inbound side of the street may be better. The Pittsburgh Bike(+) Master Plan includes extending the bike trail along Forbes to Schenley Park. Ridership on the bike lanes will be much higher when people are able to bike on protected lanes between Regent Square and the Squirrel Hill business district and Oakland. With the bike lanes ending at the Beechwood cut-through, cyclists heading towards Oakland would continue on Beechwood until it intersected with Forbes. At this point they would need to make a left turn to continue on Forbes. Unless an exclusive signal phase is added for cyclists, this turn movement is less safe than being able to proceed straight and has a bigger impact on vehicle capacity. Therefore, the city should evaluate locating the inbound bicycle lane on the inbound side of the street.

The Case for Adding a Bus Lane

The 61A and 61B bus routes run over the Fern Hollow Bridge. These are two of the busiest bus routes in Allegheny County. Prepandemic, they together carried 9,770 daily weekday riders20The Port Authority’s 2020 annual service report documents that the 61A had 5,138 weekday riders and the 61B had 4,632 weekday riders.. Both these routes will receive major upgrades by the end of 2023 as part of the Downtown-to-Oakland Bus Rapid Transit Project that will result in further ridership increases.

Delays at the bottleneck intersections adjacent to the Fern Hollow Bridge propagate to affect riders all along the affected routes. Recurring delays reduce the service frequency that the Port Authority is able to provide to riders.

A dedicated bus lane would speed up buses. With a bus lane there would still be space for one general purpose motor vehicle lane in each direction. The lane should be in the outbound direction over the bridge and approaching the Forbes/Braddock intersection. The lane should be in the inbound direction West of the bridge approaching the South Dallas intersection. Buses traveling outbound would initially drive in mixed traffic on Forbes, but would be able to enter the bus lane ahead of the Forbes/Braddock intersection.

This would allow buses to skip the queue of cars at these bottlenecks, speeding travel for bus riders21The configuration of the bus lane with respect to the bus stops would need to be evaluated. It may be necessary to adjust bus stop locations for buses to fully take advantage of the bus lane.. The BRT development will give the 61 buses signal priority capability which could then be leveraged at both of these intersections.

One possible bus lane configuration is shown, with the bus lane direction changing midway between the Braddock and South Dallas intersections.

At the South Dallas intersection, there would be an inbound bus lane and two general purpose inbound travel lanes22The space for the bus lane would come here from one of the outbound motor vehicle lanes. Since the bike lanes end at the Beechwood cut-through, there would be space here for two inbound motor general purpose motor vehicle lanes.

The inclusion of the bus lane would likely reduce automobile capacity at the Forbes/Braddock Intersection23More evaluation should be performed on the implementation at this intersection. One possibility would be to reroute the 61B bus to go straight rather than right, then follow the route of the 71 bus back to Braddock. The left turn lane could be repurposed as a bus lane and left turn movements prohibited. Left turn movements have lower vehicle capacity so this change may allow for buses to speed up without reducing the overall number of automobiles that pass through this intersection. The bus stops of the 71 have more residences in their walkshed than the 61B stops, so this route adjustment would provide bus service to more people. The bus stops between Braddock and the Fern Hollow Bridge could be eliminated in favor of those between Braddock and East End Avenue..

Counter intuitively, speeding up buses reduces travel time for motorists even if automobile capacity is reduced. Most travelers choose the travel mode that gets them to their destination fastest. When buses run faster, the mode share they capture increases, reducing congestion and improving travel times for cars.

When automotive travel demand at a street feature like the Forbes/Braddock intersection exceeds its capacity, motorists will wait in line to use the feature. This is more commonly referred to as congestion. Capacity is rationed based on people’s willingness to wait. People are less willing to wait the better their alternatives are. This is why speeding buses and ensuring safe pedestrian and cycling options are the best ways to reduce congestion.

This lane allocation would leave one automobile lane in each direction for motorists. The resulting space allocation would like like this24The only crash worthy barrier available in the software tool used to create this drawing was a concrete barrier. Hopefully a barrier that is more aesthetically appropriate for a park would be selected.:

Other Enhancements

The bus stops by the bridge do not have bus shelters. Bus shelters are one of the most cost effective ways to increase transit ridership. The Port Authority should evaluate stop consolidation and stop position, and then the city should add shelters to the remaining stops.

Additionally, the Forbes/Braddock intersection has a slip lane for cars turning from Braddock right onto the bridge. Slip lanes are dangerous for pedestrians because motorists often look left for cars and don’t look ahead for pedestrians. This street feature is inappropriate for an urban context and should be removed.

Make the Bridge Beautiful

The concept rendering for the new bridge is less visually attractive than other bridges in Pittsburgh city parks. It is something that would be fine for a freeway overpass in location where people only see it speeding by in a car. However, the Fern Hollow Bridge passes over a hiking trail of a major, heavily used city park. It also replaces a bridge that won an award for beauty.

The bridge concept uses concrete rather than steel as a primary material, in contrast to the other major bridges in Pittsburgh’s city parks. The reason is that there are currently cost and lead time issues with steel. Use of steel would add 18 months to the reconstruction process and millions of dollars in cost25Steve Cowan, the PennDOT Press officer for the project, informed me of this in an email dated March 14, 2022..

However, it is possible to design concrete bridges that are much more beautiful than the published design concept. More thought should be put into the design to ensure the new bridge adds to the beauty of Frick Park.

Vilas Bridge in Bellows Falls VT – An Example of a Beautiful Concrete Bridge

Let’s Take the Time to Get this Right

The original Fern Hollow Bridge lasted 71 years, and it’s replacement 49 years. The next bridge we build will be with us for a long time. If we need to take an extra few months, or even an extra year to get the design right, it will be worth it.

Correction (4/6/2022): An earlier version of this post incorrectly stated that the flex posts on the Panther Hollow Bridge were knocked over by cars. I was informed they were removed for the Pittsburgh Grand Prix and subsequently re-installed. Flex posts are more susceptible to damage by being hit by cars than other types of barriers, but this was not the cause of the missing posts at the Panther Hollow Bridge.

1 comment
  1. The other problem with putting all the bike infrastructure on that side is that there’s a spring near the gate house that creates a 20×6’ ice rink all winter. The connecting bike lane on Forbes is regularly blocked by debris that rolls downhill from the park. If there isn’t improvement in that infrastructure we’re putting all our eggs in a broken basket.

Leave a Reply