“Is it BRT if it doesn’t have dedicated lanes?” That was a hot topic of conversation in 2013 and I can’t finish out my series on Pike Premium Transit without digging back into the most controversial aspect of the Columbia Pike Streetcar debate.
If you need a refresher, Part 1 introduced what was envisioned for the Pike Premium Transit Network. Part 2 looked at progress on features to improve bus travel time. Part 3 examined the planned convenience & dependability features. Today, I’ll take a look at the most controversial feature that was ever discussed for Columbia Pike’s transit system.
What’s the point of dedicated transit lanes?
Dedicated lanes are a core feature of the best transit systems, no matter the mode. Given that your bus or train needs to stop places that aren’t your personal destination, transit systems can only be time-competitive with cars if they have a way to make up some or all of that time – like by not having to sit in the traffic created by those cars.
Why weren’t they planned for the Streetcar System?
Why wasn’t a dedicated lane recommended for the Columbia Pike Streetcar or the Premium Bus network that is replacing it? The first reason, is that it was seen as impossible at the time. During the first Alternative’s Analysis, VDOT was in control of Columbia Pike and made it clear they would never approve a reduction in the number of available through travel lanes of traffic. During the second Alternative’s Analysis, Arlington had gained control of Columbia Pike (except for the intersection with Glebe Rd and the interchanges with Route 27), but the agreement between Arlington and Virginia stipulated that Virginia would withhold maintenance funds for the entirety of Columbia Pike if it were reduced to fewer than two through travel lanes in each direction except during temporary lane closures related to construction, repair and maintenance.
That agreement, however, is no longer in effect. It was amended in 2017 and would now allow a lane in each direction to be dedicated to only transit, or transit and HOV, or some other prioritization. So now that it technically possible, should we do it?
One of the earliest transportation analyses that was done for the Pike (in 2003) actually looked at several possible configurations for transit on the Pike:
- Curb Shared – transit operates in the curb lane and shares space with cars.
- Median Shared – transit operates in the median lanes (moving transit stops to the median) and shares space with cars.
- Curb Varies – transit operates in the curb lanes. It shares space with cars east of Taylor Street and new dedicated transit lanes that would be built in addition to the existing car lanes West of Taylor.
- Median Varies – same as Curb varies, but transit in the median lanes.
- Curb Exclusive – the existing curb lanes are dedicated to transit, all car traffic consolidated to the median lanes.
- Median Exclusive – the existing median lanes are dedicated to transit, all car traffic consolidated to the curb lanes.
A small improvement in travel time for transit over the “no-build” scenario (which already accounts for the rest of the improvements we’ve looked at – off-vehicle fare collection, all-door boarding, etc.) and a major increase in travel time for cars. Unfortunately, the study in question doesn’t ever calculate “person-throughput” or “per-person travel time” which are metrics that would clearly quantify the overall effect on travelers in the corridor, but with the effect on car traffic so much higher than the effect on transit and knowing transit ridership, while high on the Pike, is not more than 50% of travelers, it’s clear the overall effect is negative.
So are dedicated lanes on Columbia Pike a bad idea then?
While this study is likely one of the reasons Arlington has not seized its newfound ability to dedicate lanes on the Pike to transit, it doesn’t tell the whole story and it’s certainly not time to close the book on dedicated lanes on the Pike.
First, the 2003 study was modeling what they thought traffic conditions would be in 2020. Now that we’re actually living in 2020, we can sanity check those assumptions. They assumed that traffic volumes in 2020 would be 15.4% higher than in 2002. Looking at 2019 traffic data (to avoid COVID impacts) daily 2019 traffic on Columbia Pike between Fairfax and Glebe Road is completely unchanged from 2002 (26,000 per weekday vs 26,000 per weekday) and traffic between Glebe and Washington Blvd has fallen by 13% (31,000 per weekday down to 27,000) so the results of the modeling are, it turns out, extremely flawed.
The lack of travel time savings in this study was because, at least in 2003, buses weren’t spending much time sitting in traffic. As that changes, the benefits to transit improve, but it also becomes more and more politically difficult to make the change. The more congested the road is for cars, the more natural resistance there is to dedicating a lane to transit.
The major impact to travel time for cars could be alleviated to some extent by allowing cars to use the dedicated curb transit lane for right turn movements so they wouldn’t slow other cars going straight. The dedicated transit lane could also allow HOV traffic – this would both speed carpools and lessen the impact on travel times for cars.
Dedicated transit lanes don’t just improve travel time, they also improve trip time reliability which is, in many ways, more important than a fast trip. The 2003 study fails to capture or quantify that improve travel time reliability for transit.
Faster travel times and improved reliability for transit are also likely to increase transit ridership which would further concentrate the benefits of dedicated transit lanes.
Given our climate goals, and the flexibility that Arlington gained in 2017, it may be time to take another look at lane allocations on Columbia Pike. Can we speed transit and speed carpools in a way that increases overall person throughput in the corridor? In a way that improves travel times for the majority of Pike travelers and especially those who rely on transit? It’s worth another look.