BRT FAQ #005: What’s this dang thing cost to build?

Oh, it’s easy, just divide the total cost by the number of miles and you get…a potentially really misleading number!

Photo by: Shanghai Daddy

Building a new bus line through the center of a city sounds like a time- and money-intensive task. But just how money-intensive? Like, a billion bucks, or what? Let’s wade into some numbers!

The total estimated cost of the 7.6 mile GRTC Pulse bus rapid transit line is $53.8 million. With a couple million already spent on preliminary engineering, we’ve got $49.8 million left in projected design and construction costs. That significant chunk of change is funded through four different sources:

Funding the Pulse (in millions)

BRT005--Pie

You can see how all that money gets allocated in the following table taken from the Broad Street Rapid Transit Study (Apendix A-13 (PDF)), which was commissioned by the GRTC, the Federal Transit Administration, and the Virginia Department of Rail and Public Transportation. Stations, sitework1, systems2, and vehicles make up the bulk of the costs.

6C89D615-73DE-42BF-91AB-8203227A2F8A

So for about $50 million in funding, with 84% of that coming from federal and state sources, Richmond gets the first major upgrade to our public transportation system in years? Decades? A long, long time.

The mysterious and misleading cost per mile

So how does the projected cost of the Pulse compare to other BRTs around the country? Turns out, that’s a really challenging question to answer. So much of any given transportation system is designed specifically to work with the existing geography and infrastructure of its locality that it’s hard to get a direct comparison. But that doesn’t stop armchair analysts (like myself) from mashing some numbers together and seeing what comes out the other side.

The most straightforward way to compare BRTs (or any route-based system) is cost per mile–it’s so easy! Just take the total cost of a project and divide it by length of the route, and boom, you’re done. The Pulse’s cost per mile is: $7.08 million. This is super cheap when you compare it to similar systems (which the Broad Street Rapid Transit Study does in Appendix 13 on page 4). But, cost per mile can be very misleading. Stay with me as I run through a few examples:

Norfolk’s Tide Light Rail

Our neighbors to the southeast built a light rail! Maybe we should be jealous! But maybe not, since the rail is built at sea level, and that place is slowly sinking into the sea.

  • 7.4 miles • $318 million
  • $42.97 million per mile
  • Why that’s misleading: Norfolk had to build not one but two bridges for its light rail system.3

Cleveland’s Healthline

Cleveland’s system is the gold standard for BRTs in America4, we probably should be legit jealous of this one.

  • 7.1 miles • $197.2 million
  • $27.77 million per mile
  • Why that’s misleading: Over four miles of the route had to be reconstructed, including roads, utilities, and sidewalks. The FTA estimates (PDF) that the cost per mile varies massively from $2 – $27 million (but it’s probably closer to $17 million) depending on what exactly had to be done to the streets.

Eugene’s EmX

Eugene is not Portland, but they are in Oregon, so they probably still have some of that West Coast transit magic.

  • 4 miles • $25 million
  • $6.25 million per mile
  • Why that’s misleading: I’m…not sure! There are only 10 stations, which is far fewer than the 26 that will line the Pulse’s route. They also did custom-design their buses, and build some pretty fantastic end-point stations (PDF)–but you’d think both of those would increase the cost per mile, not decrease it.

Richmond’s Pulse

That’s us!

  • 7.6 miles • $53.8 million
  • $7.08 million per mile

So why is that misleading?

Part of the concern with the current Pulse plan is that, when compared to other systems around the country, $7.08 million per mile just seems so dang cheap. Surely that’s too good to be true, right? You don’t have to Google too hard to come across articles skeptical of the purported low cost of bus rapid transit. Here’s one from The Transport Politic in 2011:

Nor is the difference in costs between rail lines and BRT nearly as great as some would argue. The Journal article quotes Dennis Hinebaugh, head of a transportation center at the University of South Florida, saying “You can build up to 10 BRT lines for the cost of one light-rail line.” That might be true if you’re comparing a train operating entirely in its own right-of-way with a bus running in a lane painted on the street. [emphasis mine]

For Richmond, the part in bold is the case for significant portions of the current Pulse Plan. Only 3.1 miles of the entire (current) Pulse route have dedicated BRT lanes. Large sections on either end of the route are “mixed-traffic operations”–aka just a bus riding around as it normally would with some signal priority technology thrown in.

I’m not sure what the cost breakdown for each and every mile of the Pulse is, although I have a pretty good feeling that that data is buried somewhere in the massive stack of PDFs sitting in my Dropbox. Even without that information, we can guess that the 40% of it requiring ripping up roads and building dedicated lanes will be much more expensive than the 60% that does not. So you can see how cost per mile doesn’t always tell the whole story.

Bad analysis alert!

Here’s another way to look at it. And, please, keep in mind that some of these numbers are made up (and indicated as such) to illustrate a point that data can be grossly mishandled. Don’t you dare quote this out of context. That said:

The only section of the Pulse in Henrico is the western terminus and about 1/3 mile of the route. The county is paying $400,000 for that, or $1.2 million per mile. If we say that’s the cost of the mixed-traffic operations sections of the route (which is a bad and stupid thing to say) that gives us a cost per mile for the rest of the route of $15.61 million. That number is way more in line with what folks generally think a BRT costs per mile.5

See! Cost per mile! You’ve got to read a lot of information to really see what’s going on.

More BRT FAQs


  1. Demolition, utility relocation, pedestrian and bike access, vehicle accessways. 
  2. Traffic signals, fare collection systems, communication equipment. 
  3. I’m totally guilty of using this misleading stat back in the very first BRT FAQ
  4. Technically they’re an ITDP silver BRT corridor, but no BRT system in America has brought home a gold…yet! 
  5. Check out page 4 of this 2000 study by the federal General Accounting Office, which says the average cost per mile of a BRT is $13.5 million. 
  • error

    Report an error

Ross Catrow

Founder and publisher of RVANews.

Notice: Comments that are not conducive to an interesting and thoughtful conversation may be removed at the editor’s discretion.

  1. Scott Burger on said:

    The initial, implementation costs are one thing, the ongoing longterm costs are another.
    We need more commitment from the universities and counties. Richmond taxpayers deserve better.
    I am not against BRT, but there needs to be more scrutiny. Too much of this so far is about serving developers’ interests and not the public.

  2. When you have a downtown that’s half empty theres a need to help developers. The number 1 priority of the city should be to revitalize downtown. Until that hole is filled we have nothing. You love ugly surface parking lots don’t you.

  3. Appreciate this info!
    Wiki says the Cleveland system has 59 stations (every 3 or 4 blocks), so that is a big difference.
    Other things to emulate from the Cleveland example – diesel/hybrid engines, approx 7 minute wait between buses during most of the day and dedicated bike lanes leading to bus line.
    Wiki says the Eugene system offered free riding at first. There are/were lawsuits relating to the expansion.

    Our new northside/hermitage lane is great but there is no way that riding on Boulevard is a good idea like was suggested in the RTD. I would like to see the northside system extend down Laburnum to Saunders/Malvern (or the road parallel to Laburnum down to the 95 bridge then Laburnum.) Old Hermitage is great but it leads to that congested spot on Broad and then narrow Meadow.
    Checking GRTC site – to drive from my house on Northside to VCU is 8 minutes, to take the bus is 40 minutes because it goes to 3rd Street first then switch buses to one down Broad.

  4. Scott Burger on said:

    Joe, there’s been over $2 billion invested in downtown over the last twenty years.
    A lot of the public investment will never be recouped, while citizens and other neighborhoods are going without basic services. There needs to a revolution in civic leadership away from the corporate welfare mavens and their development schemes.

    If you don’t like ugly surface parking lots, then go after the owners who are sitting on them. Venture Richmond showed these old parking lot photos to try to bully everyone to follow their Shockoe Bottom stadium scheme. The find the ones with the most cracks and weeds, yet they don’t mention who old these lots and who is allowing them to get into this condition.

    Did anyone go after Trani and VCU for parking lots? The one right at the corner of Belvidere and Broad sat there for years before they finally started the ICA building. In the meantime they lied and told the City and state they had to use the Cary Street site for their student rec center because they did not have anywhere else.

    As I stated, I am for the BRT and better public mass transit, but we need to make sure it serves primarily public interests without draining the City’s finances. If VCU students are going to use it more than anyone else, then VCU should do the right thing, end their private Groome shuttle contract, and make a financial commitment to BRT.

Leave a comment

Your email address will not be published. Required fields are marked with an asterisk (*).

Or report an error instead