Transit City, Paris, Reviewed

Last night, I had the immense pleasure of attenting the RATP’s presentation about the use of LRT rather than subways.  I’m not going to attempt to reproduce the information here, but am hopeful that the illustrations will show up on the TTC’s website fairly soon.

Toronto has needed this sort of presentation for a long time, and if only scheduling problems had allowed it other than on a Friday evening, there might even have been media coverage and more representation from senior staff and politicians outside of the City.

The Mayor of Paris decided that he wanted to reduce car use and green the city, and that transit was a key to regeneration of the inner suburbs.  ‘Tramways” (LRT in our terms) were the solution both for their lower cost (why build “five times the capacity at five to eight times the price”) and for their ability to stimulate the neighbourhoods through which they passed because of the pedestrian activity along the route.

Major street redesign was integral to their plans.  They knew perfectly well that the tramway would reduce road capacity, and the lower traffic volume combined with the lowered road speed converted semi-arterials into calmer, walkable neighbourhoods.

The bus service to be replaced had reached the maximum it could handle, and substantial additional riding came with the conversion to LRT.  They are now running peak headways of 4 minutes (15 trains/hour) of cars with a capacity of 300.  This is on a street with short blocks and much local demand.  Indeed, stop service time is a considerable part of the trip time even with all-door loading.  This makes the trip slightly slower, but avoids the need for passengers to access stations.

The construction projects were co-ordinated between all utilities and agencies, and a liaison committee met monthly with people and businesses in the affected areas.  A standard method of compensation for business interruption handled the vast majority of complaints in that department.  Construction co-ordination was vital to avoid the sort of cock-ups we have seen on St. Clair where each city agency rearranges its priorities without regard for the impact on overall project plans.

I could not help noticing the absence of centre poles to hold up the overhead even though the streets were a good six lanes wide.  Poles are considered visual polution in Paris and their use is minimized.  Where one pole can do the work of two or three, it does.  Transit City urban design team please take note.

This is not to say that the Paris Tramways and street geometries are a model for everything we do in Toronto, but it is so refreshing to have a city say “this is what we can do” rather than endless reasons for delay.

As and when the presentation is available online, I will update this post with descriptive comments.

17 thoughts on “Transit City, Paris, Reviewed

  1. I had to leave right at the start of the Q+A period, but I also noticed the lack of centre poles, as well as the use of side poles to accommodate catenary supports and street lighting. The streetscape images looked very uncluttered, and notable for their lack of massive clusters of traffic signals at intersections (a worse visual blight than overhead wire, in my opinion).

    Did anyone ask the RATP representatives about centre poles?

    Steve: No. Unlike Mitch Stambler, I didn’t undertake to ask questions with an agenda of discrediting positions taken by local agencies like the Fire Department. I was actually a bit annoyed with his asking leading questions of a visiting delegation. The TTC did a terrible job of fighting the worst of the design elements, and indeed they have been very supportive of centre poles. They completely missed the urban design issues and repeatedly defended that configuration.


  2. Steve,

    Thanks for this precis of the talk. It sounds like it was very interesting and I am sorry I missed it. But isn’t there a bit of a disconnect here?

    I mean, surely Paris has some lessons learned for us about LRT.

    But there, LRT has become the mode of choice for expansion because…there are already ±250 metro stations in an area the size of the old city of Toronto. There’s basically nowhere in the city of Paris you could stick a new metro line that wouldn’t duplicate one that already exists. The baseline in Paris is, therefore, 100% not comparable to that in Toronto.

    LRT for them is a sort of curious transit frill, not the basis of citywide transportation policy–a decision we may or may not live to regret in Toronto. I think we should remember that context when we celebrate mayor Delanoe’s LRT accomplishments.

    Steve: As I said in my post, we cannot apply what Paris has done directly to Toronto. However, to view their LRT as a “curious transit frill” when it is seen as an important way of revitalizing the city without the expense of subways, and when their new T3 line is carrying roughly double the ridership of any of our busy streetcar lines, this is not a “frill”.

    We need to see this in the context of other cities who are moving away from subways to LRT such as Madrid. Their LRT is a lot less “streetcar-ish” than the Paris lines, but Madrid like Paris recognizes that they don’t need full subways everywhere.

    The issue in any city, in any corridor, is density and distribution of demand. For some patterns, subways are the best answer. In Toronto, until quite recently, they were the only answer with the result that we built almost nothing.


  3. Your comments are accurate and perceptive on the Paris presentation. You might have added that the new T3 trams are almost the length of half a football field and move over 100,000 passengers daily. Indeed, the inner rim line is soon going to be expanded to more than double the current length.

    It is perhaps reasonable to suggest that Councillors and TTC Staff should do more travel for inspection of other transit properties to become more aware of new options and technology. Indeed, only Mitch Stambler engaged the guests in the dialogue following the presentation!


  4. Steve, I think you are missing Matt’s point: Paris simply doesn’t need to build any more subways. With 16 lines, 300 stations, and more than 200km of tracks, Paris is indeed a true City of Subways. If you already have a spiderweb network of subways and you still have a big traffic problem, it doesn’t take much to convince people that it’s probably time to look at replacing surface bus routes with LRTs. That’s hardly the case in Toronto. Can you imagine what traffic in Paris would be like if they only had 2 subway lines and 1 RT line like we do in Toronto?

    Steve: I understand exactly the point Matt is making. There is a big difference in the density and layout of Paris compared to Toronto. The network of subway lines is concentrated in the centre of the city, a ring roughly 11km across, about the size of the old City of Toronto, but with a much higher population. A few of the lines extend beyond into the next tier of the city corresponding to the 416 suburbs.

    The tramsways run parallel to the Périphique, the ring road around this area. The RATP presentation included a reference to their network of 120 streetcar lines dismantled in the name of modernization.

    The subway system is much older than ours, and it’s hard to say how much of it would be built that way if Paris were starting from scratch today. In a way, it’s a chicken-and-egg situation because the city supports its subway by being so dense, but can be so dense because of the subway.

    Building a spiderweb of subways would not turn Toronto into Paris any more than the CN Tower will turn the Gardiner Expressway into the Champs Elysées.

    Transit lines need to be designed to serve the demand patterns that exist and that will evolve as the city rejuvenates itself, and LRT is much more suitable for many corridors in Toronto than subways.


  5. I was keen on this presentation but messed up and remembered at Metro Hall that it was at City Hall, so I only caught the QandA.

    Paris is doing an awful lot more for bike than we are, including allowing the bikes to be in the centre busways and co-existence seems to be occurring.

    An answer to my question about the Velib helping or harming transit brought out a significant philsophical difference entre Paris en Caronto: the Velib parking is in the place of car parking.

    And part of the reason I missed the presentation was the blocking of Queen St.! along with some of John for some TIFF stuff I guess at City Tv. It defies belief that the City can sabotage Queen St. transit for fluff, while they admit the service can be trans*it. Why can’t the local councillor and the TTC insist that City TV use its own parking lot for fluff, and possibly John St. and squeeze Richmond in an absolute pinch, but get the frick off of Queen St.

    Does the local councillor have a pension or something from City, or is he still doing some programming there?

    Steve: Strictly speaking, that was not a CITY event, it was one of the new CTV-owned channels. CITY is still in that building, but will decamp to Dundas Square when their new studios are finished in the “Olympic Spirit” building. They are now owned by Rogers.

    The issue is the same: disruption of transit service to enable a media event. Very annoying, especially for people trying to get to the Scotiabank cinema whose closest transit stop is at Queen and John.

    As for cycles in Paris, I think the question of whether they harm transit was answered more from the point of view that they carry such a small fraction of total demand as to have no impact, from a ridership point of view. As you say, the real issue is whether to use roadspace for parking or for cycling, and that’s a political decision.


  6. How much would it cost to upgrade the Eglinton and Sheppard LRT lines to subways?

    Steve: The current rule-of-thumb for subway construction is at least $200-million/km, but this varies depending on how closely spaced the stations are. Underground LRT is slightly cheaper than full subways because the stations are smaller, but the main cost is the hole in the ground. Surface LRT is quite another matter and here the cost comes down into the tens-of-millions per km. There can be a lot of variation in this depending on the amount of civil work (utilities, road realignment, bridge adaptation) needed as well as any urban design elements.

    Then we have rolling stock and maintenance yards. The lower speed of LRT versus subway means that you need more cars for a given capacity, but subways have such high intrinsic capacities that we would probably see a minimum service level higher on the subway than the LRT. Then there are operating costs. Just owning and maintaining a subway structure and stations is quite expensive compared to surface trackage.

    The cost comparison gets tricky because it would have to be calculated for each line. We do know that a figure of over $6-billion has been cited for the Eglinton “Metro”, but there are no detailed breakdowns yet for the Metrolinx proposals.


  7. I think you’ve overdone the rhetorical flourish. Torontonians don’t expect (or want) to turn the city into Paris. People do have a need to get around the city.

    Just looking at the Wikipidia and other websites, only the very short T-3 tramline in Paris is akin to what is being proposed in Transit City. The T-1 line is mostly in the median of a dual carriageway. T-2 is on a converted SNCF roight or way. T-4 is a actually operated by the SNCF – and is know as a ‘tram-train’.

    The available literature highlights the success of the T-2 line (i.e. the one on the converted rail line) as:

    “Because of the success of this line (65,000 people use it daily) the trams were doubled in length in 2005, raising the capacity of each tram to 440 passengers.”

    From relying on your posts over the last months, we would not believe it possible that an LRT on a converted rail line would be successful – after all it doesn’t run on the streets where the would-be passengers live. Yet – apparently this is a success despite the average inter-station distance of 942 metres – and not being built in the middle of an urban street!

    If we’re going to look at examples from other cities, surely we shouldn’t be so selective. We need to look at what these lines really are – and how similar concepts might apply in Toronto.

    When we look at the specs for the Parisien LRT lines, we see:

    T-1 – 5 connections to Metro / RER
    T-2 – 2 connections to Metro / RER
    T-3 – 7 connections to Metro / RER

    Hence, it’s not unreasonable to assume that connectivity to a good network of rapid transit lines is a key success factor in the ridership on the LRT feeders.

    I should note that the Metro itself is quite central – but Paris also has an extensive rapid transit system (the RER) built on ROWs taken over from the SNCF. All in all there are 21 rapid transit lines. I don’t think Toronto needs 21 rapid transit lines – but I don’t think 5 or 6 is unreasonable. Let’s built the network first and get to the addons later as needed.

    Steve: There are times I am astounded at the way some posters here just want to twist what I write. Using a rail corridor is a perfectly good application of LRT and in fact I have repeatedly advocated this for the Weston line as a far superior way of getting service to the airport. I have also advocated that the SRT be an LRT line (as it was supposed to be originally), and it would run largely on a private right-of-way, including some of the extension in, wait for it, a rail corridor.

    The active rail corridors in Toronto are already operated by GO or have proposals for service on them, and this service will address long-haul passengers, not the market of a street-based LRT line.

    LRT also has a place on streets, and this was explicitly mentioned in the RATP presentation. Moreover, the variety of implementations in other cities shows just how flexible the mode is.

    The problem with your analogy is that it ignores the population density of Paris, and it also treats the Transit City lines as “add ons”. Sorry, but they are major parts of the future network. I agree that we should have more RER-like lines, and if only GO Transit would get off its ass (or more accurately GO’s funders) we would have had more lines and much better service decades ago. One thing this would have done would be to mute demands for the TTC to extend the subway into territory that could have had good “RER” service at much lower cost.


  8. For those who want Parisian transit density in Toronto – the city of Paris has about the same number of people as the city of Toronto (2.5 million) on a land area less than 1/7th of this city’s. With 25,000 people per sq. km. you can justify a lot of metros AND surface transit. The purpose of the visit was to show off how Paris had built T3 rather than their entire system and entirely coincidentally to show off swishy looking French-built tramcars.


  9. The other examples that would be wonderful to hear from would be the cities in Poland. Warsaw and others have extensive, exclusve rights-of-way, LRT systems covering the entire city. At one time the World Bank turned down the idea of supporting a Metro (subway) in Warsaw because it wasn’t worth the expense given the LRT network. Economic benefit-cost analysis can work for or against megaprojects, as one would expect, depending on the specifics of the proposal. Warsaw went ahead with the Metro anyway.


  10. It was a real pleasure to attend the RATP presentation and listen to planners and engineers who not only understand transit fundamentals but were able to successfully implement an outstanding project. The contrast with the on-going St. Clair comedy-of-errors could not be more apparent, especially in terms of urban design and traffic engineering.

    But, the enthusiastic Parisian visitors made it very clear that sound transit objectives essentially neutralized traffic and parking concerns. Improving transit was the prime goal and was overwhelmingly achieved, even at the expense of the automobile. Sadly, on St. Clair, it’s obvious that suburban traffic and parking methodologies from the 1960’s have hi-jacked the project. I wonder what the Parisians really thought of St. Clair when they toured it earlier on Friday.


  11. In terms of population density, I would argue that the urban area densities are not that different.

    In Paris, the Urban area (as defined as the Paris Agglolmeration) has a density of 3,542 per sq km. In Toronto, if we consider the city + contiguous suburbs (So Toronto + Mississauga + Brampton + Richmond Hill + Markham) we get roughly 2,712 per sq km. Since the GTA is growing somewhat faster than Paris, we’ll be approaching the density of urban Paris within the planning horizon.

    Steve: You miss the point. The area served by the subway lines in Paris is much smaller than the Agglomeration, and that area has a density higher than Toronto. The City of Paris proper has a population of over 2 million in an area of 105 square km. The old City of Toronto had a population of well under 1 million in a slightly smaller area. Most of the subways serve the old City of Paris.

    We are nowhere near reaching the population density of Paris, and your use of the figures is inaccurate and misleading.


  12. Your point seems to be that since Toronto’s population density is not distributed exactly the same as Paris’s, we shouldn’t consider their transit system for comparison. (Oh yes – other than the LRT part.)

    Paris has the 5 very high capacity RER rapid transit lines that serve an areas much larger than the center. Furthermore, many regular Metro lines extend outisde the core. 60 stations of the metro are outside the old core – roughly the size of our entire subway system.)

    Steve: You cannot cherry-pick comparisons for any mode. I am well aware of the differences between Paris and Toronto, and am not suggesting that the T3 can be duplicated here. It would be wonderful if our two million population existed in a less auto-oriented suburbia, but that’s a change we won’t make overnight, if ever.

    The important threads are the recognition that even in Paris, subways are not the answer to every problem, and that building a transit line of that type involves more than just putting tracks down on an existing street.


  13. Sorry, I don’t mean to be overly argumentative. It’s the exactly the opposite of ‘cherry picking’ that is my intent. Granted that we can learn tactical lessons from how transit projects are implemented in other cities. However, it’s important to present the overall context in which the service is provided.

    It makes no sense to place an imaginary loop around the center of a city and say it’s dense – so it needs subways – and outside of none. Were this true, Paris wouldn’t have extended the older Metro lines outside – and wouldn’t have created the RER when it became impractical to extend the original lines because of stop spacing.

    The RER is nothing like GO – the lines are similar in deployment to London’s Central and Metroplitain lines – which extend far into the not so dense suburbs.


  14. Unfortunately I had to miss the presentation, so I don’t know if this was covered or not, but there appear to be some minor but critical differences between the design of the T3 line and recent history in Toronto.

    The first is in the traffic lane width. Paris accepts a narrower traffic lane than Toronto. The lanes adjacent to the existing T3 line are only 2.75 m (9 feet). In Toronto, I believe the typical “absolute minimum” width would be 3.2 m. A 4-lane road would have a difference of 2 metres, which may not sound like much but could make a big difference in something like, say, sidewalk width (St. Clair, anyone?). There are 2.75-metre left turn lanes in some areas of Metro where the objective was to retrofit a left turn lane within a legacy 48-foot road (e.g., Victoria Park), but applying this width to general traffic lanes would be a major departure for Toronto.

    (Another example: Road designers here like to point to the Champs-Elysées as an example of how a 10-lane boulevard can be compatible with a vibrant streetscape, but fail to note that the 27.5-metre width would only fit 8 North American lanes, if that.)

    Secondly, there are very few left turn lanes at intersections along the T3 line, which narrows the street cross-section. Presumably left turns work like on University Avenue, with vehicles waiting within the median (the Paris signals work like Queens Quay — tram phase, main street phase including left turns, side street phase).

    Thirdly, there are pedestrian refuge islands at most street crossings (main street and many side streets); from what I can glean from RATP background documents about the T3 extension, these are actually intended as such (in contrast with typical North American practice).

    Fourthly, the corner radii are generally smaller (around 5-6 metres, rather than 10 metres or more here in the burbs).

    These three items reduce pedestrian crossing times at intersections to just a few seconds, which makes it much easier to implement signal priority for trams. Side street green can’t be cut off until pedestrian times have been met. In Paris, pedestrians are crossing narrow lanes, fewer lanes (no left turn lanes), can wait on the refuge islands, and don’t have to cross wide corners, so the side street phase can be cut off with only 5-6 seconds or so of warning time. On, say, Don Mills Road, you would need 24 or 25 seconds. This means there is a lot more flexibility in Paris, which allows them to make the claim that trams (at 4-minute headways!) can operate without having to stop at signals… even though they operate on a short separate phase like on Queens Quay, not during the main phase like on Spadina.

    Steve: Yes, the design aspects are very important and have been part of the ongoing debates about street layouts here in Toronto. One other point I noted from the presentation was that the average speed of the T3 is in the same range, even a bit lower, than the hoped-for speeds on the Transit City lines. This is very much an urban route akin to our downtown streetcar routes, even though it is on a private right-of-way.


  15. one of the things that people always comment on when seeing pictures of T3 is the grass. If TTC wanted to do something to distract people with “oooh, pretty!” there is that open trackbed on the Queensway… although my choice would be FieldTurf given the climate.

    Unfortunately my confidence that we will ever have the courage to do transit right took another dent this evening while waiting for an EB 504 at 6.25pm. I observed the line of parked cabs across the road outside First Canadian stretch even further than usual – the trunk of the last cab (from Beck, and I have the cab number) was protruding into the pedestrian crossing area at Bay. A TPS vehicle passed this spectacle without so much as a direction to the rearmost 4-5 cabs to find another rank, let alone write a ticket. For once I wished my Blackberry had a camera.

    Where are the measures we were promised to relieve congestion on King last year? Now that the media is concentrating on the streetcar contract it seems the routes they will run on have decreased in importance.


  16. I wonder if anyone from the presentation talked about the countless problems Paris is having with these tram lines, due to overcrowding and the issue that these lines really should have been subway extensions.

    Just google the Paris tram site and you can find a number of websites with videos showing how bad the conditions are on the Paris tram network, and how these “LRT” lines are not providing the needed service required.


Comments are closed.