The REM Proposal is the Best Transit Project Montreal has seen in 30 Years

April 28th, 2016 by ant6n

Last Friday, the Caisse de Dépôt presented their project to build the ‘REM’, short for ‘réseau électrique métropolitain’. It’s an electrified rapid transit line connecting Montreal with its suburbs in the West Island and Brossard, which includes a branch to the Trudeau international airport.

Rendering of a REM line station

Rendering of a REM line station

The proposal is exciting: 67 kilometres and 24 stations, replacing the existing Deux-Montagnes line (30 km) and the bus corridor to Brossard with one large line.

There’s a lot to like about this project:

Electrified, Frequent, All-Day, Rapid Transit

Why it’s good:

Electrified: electric trains accelerate faster, which allows for more frequent stops while maintaining high overall speed. Moreover, electric trains are much cheaper to operate than diesel ones, so transit agencies can afford to run them all day.
Frequent: Frequency is freedom. Taking a train that comes every 30 minutes requires planning ahead: show up too early or too late, and you risk waiting half an hour for the next train. Frequent service means travelers can just show up at the station, knowing there will be a train within a reasonable time.
All-day: Commuter rail provides a very specific service: shuttling office workers between their homes in the suburbs and their 9-to-5 jobs downtown. All-day transit allows anyone to take any trip along the line at any time of the day. Frequent all-day service increases flexibility and reliability. If you can take transit for any trip you need to make during the day, rather than just the commuting trip, you can rely on it more, so you don’t have to rely on owning a car.

Previous proposals to improve transit only focused on 1 or 2 of these criteria. For example:

  • Previous proposals to improve transit to the West Island included adding tracks to the existing diesel commuter rail line, allowing more trains during rush hour without providing frequent, all-day transit.
  • Before the Caisse’s REM project, the most recent proposals for the new Champlain bridge did not include a rapid transit line, but would instead have replaced the existing bus corridor with just another bus corridor.
  • The Mascouche line, the only transit project completed in almost ten years, is served by only 10 trains a day per direction, and looks more like an exercise in drawing lines on a map than providing a transit line that can move around a significant number of people

Integrated Regional Network Using Through-Routing

Through-routing is where a single line goes from one end of a region, through downtown, to the other end. Such regional lines facilitate trips other than between downtown and its suburbs. By making these trips faster and more convenient, more people are encouraged to take transit. In many places in the world, this is how transit systems are designed.

In Quebec, however, such regional interconnected systems have never been a priority, as the practice was to solve each problem with its own technology, its own line, without much concern for designing a network. Need more capacity on the Deux-Montagnes line? Improve the Deux-Montagnes line. Need an airport connection? Build an express airport line. Need to improve transit to the West Island? Add more tracks to the existing diesel commuter rail line. What about the South Shore? Light rail on the Champlain bridge!

The Caisse’s project is the first one that shows a much more integrated vision for regional transit, folding all these projects into a single line. Now you could take transit from suburb to suburb, or from one suburb to places not quite downtown. Imagine going directly from Brossard to UdeM!

Repurposing Existing Infrastructure

Reusing existing infrastructure makes projects more capital efficient, which means more transit for our money.

The REM network is to be built on the Deux-Montagnes line, which is already electrified and connects the West Island to downtown through the Mount-Royal tunnel. Using the tunnel makes sense: it provides a direct connection from the north of Montreal through the mountain to Gare Centrale in just 6 minutes, basically like a subway. Given how costly the Blue line extension is supposed to be, it’s a good thing we have the tunnel already built.

Connections with Existing Lines

A good transit network requires good transfers between lines, to allow users to reach more destinations quicker. That’s because people want to travel between all sorts of places, which are not necessarily on the same line. Consequently, good connections will allow making these trips quicker and with fewer transfers.

Connections between REM-line and existing lines.

Connections between REM-line and existing transit lines.

For example, the station at Edouard-Montpetit will allow riders on the Blue line to directly connect downtown, without having to transfer twice (to the Orange line and then the Green line).

Also, consider that currently, the Saint-Jérôme line crosses the Deux-Montagnes line at Canora, but there is no stop or transfer possible. Instead, St-Jérôme line travelers are stuck on their train for another 25 minutes while Deux-Montagnes trains will be downtown in 6 minutes. The transfer station at Canora will help reduce their travel times a lot.

Planning for good transfer stations to existing lines for new transit lines seems obvious, but it’s not something that’s usually being considered much in Montreal. Consider the following examples:

  • The Mascouche line has one inconvenient connection at Sauvé, with a long walk outside through a cemetery.
  • The line was built without transfer to the St-Jérôme line, even though their stations are 200m apart near Marché Central, because both lines were independently designed to only shuttle commuters to and from work.
  • The proposal by Aeroport de Montréal for the airport train envisioned an express that would’ve gone straight downtown, with no connection to any other line, even though adding a single connection at Vendôme would have increased the convenience for many users.

It is therefore refreshing to see the Caisse’s REM proposal give consideration to how their system will connect to the lines crossing it.

Overall, I’m excited that the Caisse wants to get this line operational in only four years, a line that is truly regional, a line that will benefit more people than just the typical commuters. In the last 30 years, the only rapid transit we’ve built is the 4 Km Laval extension of the Metro line, that means a bit more than a kilometre per decade. Now we’re building almost seventy kilometres all at once, built by an organization that has proven it can get a transit line built. Montreal is dreaming transit again.

Thanks to JC for her help in writing the article.

The Lac-Megantic Incident, It’s All About The Brakes

July 9th, 2013 by ant6n

The Lac-Megantic incident was caused by an apparently unmanned train of the Montreal, Maine and Atlantic Railway (MMA) loaded with crude blasting into a small town at high speed. The train consisted of 72 “DOT-111” tanker cars, which have been criticized in the past to “have a high incidence of tank failures during accidents” (NTSB report report, page 75).


There’s a sharp curve in downtown, and that’s where an explosion and subsequent fire destroyed a large number of buildings, and killed at least 13 people. This happened during Friday night, bad timing because there were some festivities in a bar that was engulfed in the fire (video).

A Timeline

The investigation is still ongoing, so explaining causes is basically just speculation. But we have some information, let’s look at a timeline leading up to the explosion:

  • 23:25 The engineer parked the train with 72 crude-carrying tanker cars at a siding near Nantes, QC. This is about 10~11km away from downtown Lac-Megantic. The engineer shut down four of the five engines and left the lead engine (#5017) running to supply the air brakes.
  • 23:30 A resident called 911 to report a locomotive engine on fire at the Nantes siding.
  • 23:42 Firefighters arrive on the scene. As part of their operations, They shut down the lead engine. The Nantes Fire Chief Patrick Lambert told Reuters that the crew had switched off the locomotive a “good-sized” blaze in the motor, possibly caused by a fuel or oil leak in the engine. “We shut down the engine before fighting the fire.” “Our protocol calls for us to shut down an engine because it is the only way to stop the fuel from circulating into the fire.”
  • 0:12 The fire is extinguished
  • 0:13-12:15 Two emplyes of the (MMA) show up at the scene
  • 1:30 The train derails at high speed at Lac-Megantics Rue Frontenac crossing.

According to MMA’s chairman Ed Burkhardt, the brakes will not work if a train is switched off: “If the operating locomotive is shut down, there’s nothing left to keep the brakes charged up, and the brake pressure will drop finally to the point where they can’t be held in place any longer”.

How many brakes in a train?

A train relies on air brakes to keep the cars from moving. There’s a high pressure air line running through the whole train, supplying power to the brakes and activating the brakes. The brakes are fail-safe, that is they release at high pressure, lower pressure means ‘stop’. If the air pressure fails, it will go into emergency brake mode and stop. The energy for that braking action is supplied by an air pressure reservoir on each car. However, these reservoirs can deplete after a while and will release the brakes. That’s why the lead engine of the runaway train was kept on, so that the onboard air-compressor could maintain the air brake pressure.

Since the air brakes can stop working, every car also has hand brakes, operated by a large steering wheel. The Canadian Rail Operating Rules (CROR) specifies the following:

(a) When equipment is left at any point a sufficient number of hand brakes must be applied to prevent it from moving. Special instructions will indicate the minimum hand brake requirements for all locations where equipment is left. (…)

Equipment here is defined as “One or more engines and/or cars which can be handled on their own wheels in a movement.” The rules also specify that sufficient application of handbrakes has to be tested:

(b) Before relying on the retarding force of the hand brake(s), whether leaving equipment or riding equipment to rest, the effectiveness of the hand brake(s) must be tested by fully applying the hand brake(s) and moving the cut of cars slightly to ensure sufficient retarding force is present to prevent the equipment from moving. When leaving a cut of cars secured, and after completion of this test, the cut should be observed while pulling away to ensure slack action has settled and that the cars remain in place.

So the air brakes may have been made ineffective related to the fire on the lead engine and the fire department operations. But the hand-brakes should have still prevented the trains from moving, if properly applied according to the operating rules. There may not have been enough hand brake force. The MMA claimed that hand brakes were engaged on all five engines, but it is unclear on how many cars, and whether the brake test was performed.

Even if the magnitude of the incident at Lac-Megantic is unusual, air brakes failing and insufficient hand brake force resulting in run-away equipment and a subsequent collision is not. This is exactly what happened in a January 2012 collision near Hanlon, Alberta.

One detail that is unusual is that there was only a single engineer. Usually, freight trains are staffed by an engineer (as the ‘driver’) and a conductor, who among other things has the responsibility to operate the hand brakes. One could speculate that without the conductor, it is too much work to walk along the train and operate a set of hand brakes, then get back into the engine to perform the brake tests. It turns out that in 2012 Transport Canada actually allowed the MMA to operate with reduced staffing levels.

A Runaway Train Rolling Down the Hill

In the title image one can see the path from the Nantes siding to the point where the train derailed, a distance of a bit more than 10km. From the geoprofile one can see that path is basically all on a downward slope, for a total elevation difference of about a 100m. That elevation difference in 10km is a gradient of only about 1%. But steel rail on steel track has extremely low rolling resistance, and air resistance will only come into play at higher speeds.

It doesn’t matter very much whether whether there’s a 1% grade, a 10% grade or a roller-coaster falling down 100m, most of the potential energy will be transferred into the kinetic energy of the train moving forward. A simple calculation (E = mgh = mv²/2 => v = √2gh) shows that this energy can speed the train up to 160km/h, assuming no rolling and air resistance. Even half that is enough to derail the train at the downtown Lac-Megantic curve, and cause the disaster.

The train was travelling from from New Town, North Dakota, over 3000km to it’s destination at the Irving Oil Refinery in Saint John, New Brunswick; it also passed through Toronto.

The investigation will most likely center around all the brake failures. But I hope we will re-visit the safety of leaving trains full of dangerous substances unattended, on inclined sidings with grades targeting populated areas, or the purpose in general of moving crude oil over thousands of miles, whether it be by train or by pipeline.

Montreal Tram Study – Going About it the Wrong Way?

May 9th, 2013 by ant6n

A Montreal tram report studies the viability of a starter line. However, the proposed line includes a downtown loop, a section with questionable utility. Also, the proposed construction costs are too high. If we want a sensible network of trams that improves rides for as many people as possible, we need to focus on utility, and we need to aggressively contain costs.

tram route

Last week the city of Montreal finally released their tram study, which the city received 18 months ago. The 1095-page study details the feasibility of a starter line. Here are the highlights:

  • line length: 13.2km
  • number of stations: 32
  • average distance between stations: 425m
  • average speed 18.1km/h
  • projected ridership: 26.6 M/year (70K a day)
  • trams: 26 trains, 30-35m length, 2.65m width
  • cost: 850million, without tax and contingencies (about a billion with)

As you can see from the above map, the line consists of a corridor along Côte-des-Neiges and a loop around the Old Port. The line is on separated lanes on its entire length. The base service frequency is intended to be around 8 minutes. During rush hour, the Côte-des-Neiges corridor would see the frequency double to every 4 minutes. During the “peak downtown period”, the pattern would reverse, with the downtown loop now having its frequency doubled to every 4 minutes.

It is great that this study was finally released. It would be nice if the AMT did this too. Its also good news that the basic viability of trams in Montreal is shown. But looking over the documents, there are some issues that I would like to raise.

The Downtown Loop

The first thing to notice is the loop around Old Montreal, a section that seems to be a rather bad piece of planning. Actually it’s not a complete loop with trains travelling in a complete circle, because it’s disconnected at the north-west corner, where trams terminate and turn around. The construction of this loop seems to be mostly driven by political interest, and from developers. In general, downtown circulators are a bad idea, they underperform. They are not useful for enough people. They don’t let you travel the shortest distance to where you need to go. Many travels along small loops are faster by walking directly, rather than waiting for a train and taking an indirect route to a nearby location.

In 2008, The STM established the 515 bus (now 715) as a precursor for the planned tram (except that it actually runs in a complete circle). Already by Decemeber the failure of that line was obvious, with only 1200 daily boardings rather than the projected 6000. According to ridership data from 2011 that I extracted from Opus card data, there were less than 800 daily boardings on that bus. That is not nearly enough for a tram. and given that there are many bus corridors with fifty times the daily boardings like Pie-IX, Sauvé, Saint-Michel or Henri-Bourassa, focusing on this section is misguided at best.

Naturally, a tram may attract more potential riders, and more development, which should bring more riders. The study considers this induced demand, and gives us this graph of what the STM projected hourly ridership during the peak period would be like:


We see that according to the projections the Côte-des-Neiges section shows much more potential than the downtown loop, which simply cannot attract riders. And those numbers may be optimistic, just like the projections for the 515 bus. Many riders may also only be seasonal, and that’s not just tourists. And the Côte-des-Neiges and downtown loop section appear to have non-connected ridership. Trams are nearly empty somewhere along the middle of the line. There are few potential trips that cross those points.

Overall, the downtown loop is simply a waste of money considering the transportation needs of the greater Montreal region. The only section along the the loop that has some potential is the Peel corridor along the West of the loop. That portion overlaps with the another transit plan along the Champlain bridge light rail. This is a transit project to connect the Shore suburb of Brossard to downtown, via the replacement Champlain through Nun’s island and Griffintown. That project is very worthwhile and would make the less than 2km track along Peel redundant.

The study also compares two possible corridors for the non-downtown-loop portion, Côte-des-Neiges and Ave due Parc:

Côte-des-Neiges vs Parc Ave

The above table shows that the the Côte-des-Neiges corridor is a better than the one along Parc, which appears reasonable. But why were the two corridors compared for their potential, but the downtown loop was included in both cases seemingly without question?

The two compared corridors have similar ridership, so it appears natural to just connect them into one line which will have fairly balanced ridership. This would almost exactly follow the successful rush-hour-only 435 bus. It would be possible to “unroll” the downtown loop, and instead build it along the Parc corridor. The length of that line would be about the same as the one proposed by the study, ~13km. This line would replace the 80, 165 and 435 buses, which together have about 60K daily boardings (in 2011) – when one considers the possibility of induced demand, it should be obvious that this is a much more worthwhile starting segment than including the proposed loop with only one of the useful corridors.

Another, if smaller, issue is that the downtown loop includes the steepest grade along the whole line – the ruling gradient:


Because of only the small section of 12%, all the trains have to be able to climb that grade. If excluded, the ruling grade is 10%, which is generally considered the maximum for trams. This may result in less technical complexity, more bidders on the rolling stock, and less cost for vehicle acquisition.


The cost is currently budgeted at 850$ million for 13.2km, without taxes and contingencies, a billion with those included. That means about 65$ million/km without taxes and contingencies, and 75$ million/km with. This appears pretty high. For example, the French city of Besancon is building its initial 14.5km line for 22 million$/km. This is of course only one data point. the Montreal tram study compares to five different French data points to show that the cost are not too high.

To get a more thorough comparison, I compiled a list of about a hundred French tram construction projects. It covers projects from the beginning of the tram renaissance at the end of the eighties, to projects being planned now. The numbers are all without taxes, adjusted for inflation relative to 2010 (to compare to the Montreal study) and converted to dollars using the average rate of that year (1.37$/€). According to these numbers, the construction costs in France for a kilometer of tram are about 35$ million/km on average. In comparison, the Montreal costs are estimated 85% higher than that. (I will publish the data and more details in another blog post soon).

French tram projects aren’t actually considered cheap. They tend to include a lot of urban improvement and renovation of the streetscape, and are built to a high visual standard. They are used as drivers for development not just via transportation, but also urban beautification. The tram systems tend to be a little over-designed. Some of the systems include short tunneling sections. The French also like to avoid overhead wiring in their historic city centers, using a system called l’alimentation par le sol (APS), a ground level power supply, which adds to the cost.

Compare to Germany, where trams are used as a form of transit to bridge the gap between buses and subway, rather than a chance to beautify the city. There, tram construction costs are generally considered to be at around 10 million €/km (14 million$/km) and 15million € when built in downtown (21 million $/km). A more utilitarian approach is also what allowed Besancon to keep its tram construction low.

There are some budget items in the Montreal project that are inherently more expensive than elsewhere, for example the maintenance center and garage. It has to be completely indoors, due the climate. But overall, the construciton costs should be lower, because the project is comparably simple. The tram can be completely contained inside the existing streetscape. There are no heavy works required like rebuilding bridges. It shouldn’t need any of the features that drive up cost, for example tunnels or ground level power supply. And while nice, it doesn’t need urban beautification along its entire length.

Pine Tunnel option.

The study includes a 100$M option for a 500m tunnel under Pine, providing better access to the Montreal General Hospital via an underground station. This is probably not a worthwhile investment. But note how over-designed this tunnel station is: wider than some metro stations, three elevators, 12m deep, and a tunnel height of 6.2m. It could be designed with a 6~8m island platform, directly under the street, with a single elevator and stairs, no mezzanine, and tunnels that are barely higher than the trains themselves.

What is the Role of a Tram?

All trams are not created equal. There are different paradigms of how trams are built and what exact purpose they have in terms of transportation and urban development.

In France, the tram systems built over the last twenty years have been urban development and renewal projects as much as they are transportation projects. There has been great care to integrate those trams into historic centers with its pedestrian areas, and a lot of money spent on beautification and new technology that avoids overhead wires in certain sections. At the same time, they are the major mode of transport in some medium size cities (i.e. around and less than a million). Some of those metropolitan areas are pretty compact, so these systems have high ridership.

In the US, the development potential of trams has been recognized. Now there are a few new projects underway to build streetcars. Many of these projects have little transportation value, using single, short lines, sharing space with cars. Some are only single track for portions and have very low frequency. These projects appear to be built solely for their development potential.

In Germany, trams tend have more utilitarian transport character. They are built generally as pure transportation projects, in corridors where ridership is expected to be between about 4K and 30K~40k per day, above which rapid transit is favored. Towards the high end, they tend to build subway-tram hybrid systems (Stadtbahn), which attempt to combine advantages of subway (high capacity/speed downtown) with trams (cheaper construction or possibility of re-using existing systems to get good coverage).

So what Role Should Trams Have in Montreal?

The released study appears to orient itself very closely to the French model of tram building. That’s evident not just in the use of pictures and comparisons for cost and rolling stock, but also how the development of the line is envisioned, the complete re-configuring of streets. The loop around the old port may also be an attempt to replicate the image of the French tram through a historic city center.

Given the high cost of the project relative to its transportation value, especially in comparison to other tram systems, I’m concerned that we may end up with a project that is largely a development stimulus one. The chosen paradigm may not lead to transportation improvements throughout the city, simply because building a large network at this cost is not economically viable. This is problematic for our transportation starved city.

Montreal is currently a major bus city, with many people spending large amounts of time in crowded buses feeding into metros. Some of our bus corridors have more ridership than the 40K/day that Germans consider good enough for rapid transit, many more have lower ridership which can still be more economical to operate as trams. For many of these riders that live further away form the metro, trams could improve the quality of their commute, and thus the quality of life. A large network could spread the development and gentrification pressure, and allow less affluent people better access to jobs, again rising quality of live for everybody.

Trams in Montreal should primarily be used to improve existing bus commutes. Tram projects should be about transportation, not helping development interests downtown.

Costs Need to be Contained

If we want to build as many tram lines as possible, we have to focus more on keeping the costs down. Construction should be simplified. For example, we don’t need to reconfigure entire streets. For the tram line in the study, we could do that just for the René-Lévesque Levesque corridor, and only few parts of Côte-des-Neiges.

Beautifully rebuilt portion of Côte-des-Neiges where nobody lives and development potential is low.

Beautifully rebuilt portion of Côte-des-Neiges where nobody lives and development potential is low.

The downtown loop, with its relatively small transportation value, should be scrapped. Instead, the Parc Ave corridor should be preferred. If we aim for good cost/benefit in terms of ridership and improved commutes, one could also pick a completely different initial segment. why not start the tram outside of downtown, for example along the Pie-IX or Gouin/Henri-Bourassa corridors?

Pie-ix currently has a expensive BRT project, at 20$M/km. It suffers high cost (for a BRT project) due to a similar issue as the proposed tram: rebuilding the street. The project includes widening its right of way from 30.5 to 33.5m. With a tram, the transit right of way could be narrowed from the planned 8.40m (for two bus lanes) to as little as 6.35m (using 2.4m wide trams). Steal some more space from wide car lanes, and it will be unnecessary to buy land and cut down trees, and the higher cost of the track could be partly offset by reducing the street-widening costs. The maintenance center and garage may be cheaper further away from downtown. And the potential to improve people’s commutes is greater along that corridor, that sees long, crowded bus commutes.

Other ways to keep construction costs down for any tram project may be to partner up with other Quebec cities that envision building trams, like Quebec City, (or the AMT for the Champlain bridge corridor), or cities that think about building trolleybuses, like Laval. The rolling stock and related expertise could be pooled and large orders be made, which result in lower per-unit prices. The development expertise and management may be shared and done in-house, potentially reducing costs (right now budgeted at 100$M before tax, nearly as much as the rolling stock).

I really want trams to come to Montreal. But if we cannot keep the costs down, and the city has to plunk down a cool billion for 13 km, and the possibilities of building an actual network are slim, then this will just become another project in the region that envisions one short line and yet another technology. In that case, maybe it makes more sense to invest in much cheaper improvements to our bus corridors.

Thanks to Kamal Marhubi and Julia Evans for reading drafts.

The New Yorker: Income Inequality Visualized along New York’s Subway

April 17th, 2013 by ant6n

The New Yorker just released this nifty visualization, showing the median household income near New York Subway stations. It’s interactive, letting you click through the system’s lines, showing the the numbers for every stop. The accompanying article explains it as a way to show how the United States, and in particular New York City, has a problem with income inequality:

(…) if the borough of Manhattan were a country, the income gap between the richest twenty per cent and the poorest twenty per cent would be on par with countries like Sierra Leone, Namibia, and Lesotho.

Looking at it from more of a transit angle, for Benjamin Kabak at the 2nd Avenue Sagas the infographic presents a visual case against zone fares. New York’s very large subway system is under a single fare zone, from end to end you can travel more than 50km on a single ride ticket. That would seem unusual from a European point of view, where transit is viewed as something that ideally pays for itself, and distance-based fares are used to bring fares closer in line with the actual cost.

But in New York, as we can see from the interactive infographic, the median household income drops off rapidly as we go away from the center. So distance based fares would disproportionally affect less affluent people, who travel long distances to get to work in the city center, while more affluent people live nearby and only have short rides. The single fare zone is thus seen as a tool in helping the reduce or at least alleviate the income gap problem.

Looking at it from a more urban development perspective, I find the apparent concentration of income brackets problematic. Both concentrating poverty and concentrating wealth are bad for the city, its health, economic activity and liveliness. The city may not be able to improve the wealth distribution of its citizens by much (upper levels of government are generally more effective for that), but maybe it can find more ways to better integrate people of different income levels, give less affluent people access to more neighborhoods and promote mixed-income neighborhoods.

Either way, I was interested to get a better overview of all the data, instead of having to click through the lines one by one. So I went into the code that generates the infographic, and hacked at it a bit to get all of the lines shown on a single graph. All lines are aligned in the center with their wealthiest stop, which also approximately aligns them at the center of the city. The graph goes left to right, generally from Brooklyn/Queens via Manhattan to Bronx/Queens. By showing all the data on a single graph, the problematic distribution of wealth can be seen in a single view:


Ottawa’s O-train (Part II): A Cost-Effective Project

April 16th, 2013 by ant6n

In the first part of this series, I introduced the O-train and how it works as a transit line. The second part is how it got running, and how cost effective the project was.

O-Train over Rideau River

O-Train over Rideau River (in public domain, via wiki commons)

The O-Train was built as an evaluation project for light right technologies, to test whether it would make sense to build it all over the city. The project was realized pretty quickly and cheaply: After a yearlong study the pilot project was approved in September 1999, and public service started on October 15, 2001. The initial budget of 16$ Million had grown to about 21$ Million by the time the train started running. It is a very cost-effective project, both in cost/rider and cost/km. Compare to a couple of other transit projects (not necessarily light rail):

rider/km cost/km
O-Train (lrt) 8 27 14 1750 3.3 1901
C-Train (lrt) 29.3 582 187 6382 19.8 3110
Edmonton lrt 12.2 404 70 5738 33.1 5774
Pie-IX Busway (brt) 15 316 70 4667 21.1 4514
Laval Extension (metro) 5.2 829 60 11538 159.5 13825
Spadina Extension (metro) 8.6 2400 100 11628 279.1 24000
Train de l’Ouest (commuter) 33 1000 41 1231 30.3 24608
Train de l’Est (commuter) 44 671 11 250 15.2 61000

Naturally, these projects are not directly comparable, but generally one should expect cheaper projects to get less riders, so that the ratios work out to be at least in the same order of magnitude. But the O-train is a bit of an out-lier, even compared to the initial segments of the Calgary C-Train, which is considered one of the most cost effective transit projects in North America (by attracting a lot of riders to the relatively low cost rail lines).

The low cost of the O-train is due to the construction on an existing freight line, the use of a single track except for a passing point at the center station, the relatively high ridership for the short line thanks to it’s main anchor, Carleton University – the smart transit planning introduced in the previous part of this series. But even then, the capital costs are surprisingly low for what was built. The following shows the capital budget, taken from a report of the O-train, with items listed in thousand dollars:

Expenditure Budget Actual Cost
Corridor Lease $ 515 $ 535
Track Rehabilitation $ 1,525 $ 1,925
Tunnel/Bridge Rehabilitation $ 1,275 $ 1,532
Station Construction $ 4,675 $ 5,700
Fencing Installation $ 405 $ 566
Train Control $ 590 $ 1,088
Operator Training $ 300 $ 718
Vehicle Acquisition $ 4,875 $ 5,369
Vehicle Modifications $ 675 $ 614
Vehicle Maintenance Facility $ 755 $ 1,160
Project Management $ 190 $ 770
Start-up Marketing $ 50
Monitoring Equipment $ 165 $ 299
Other Items $ 17
Noise Fencing $ 447
Total $   15,995 $   20,740

The Stations

Several of the items appear pretty cheap, compared to other projects throughout North America – for example the reconstruction of one commuter rail station near Montreal cost 20$ million by itself. Consider the cost of the stations: less than 6$ million for 6 platforms, that’s less than a million per platform. It certainly helps that the platforms are only 35m long, and built very simply with identical design, with an asphalt or concrete surface and bus-style shelters. Also, no parking.

The stations are pretty spartan

The stations are pretty spartan.

The Trains

The vehicle acquisition is another item on the budget that appears pretty low – around 5$ Million for three trains. Each of them is equivalent to a bit less than two 85 ft passenger cars as they are used in North America on mainlines. That’s equivalent to about 1$ Million per car. Consider other rolling stock purchases for similar non-compliant DMUs, which are all small orders from North American transit agencies:

train Location/Service order year units unit cost capacity
Talent Ottawa O-Train 1999 3 1.8 M$ 280
Desiro NCTD Sprinter 2004 12 4.2 M$ 230
GTW 2/6 Austin Capital MetroRail 2005 6 5.4 M$ 220
GTW 2/6 Denton County A-Train 2009 11 6.7 M$ 220
Lint 41 Ottawa O-train 2011 6 5.7 M$ 260

These numbers are approximations. There’s been inflation over the years, the capacities may not be exactly comparable (i.e. by removing seats, capacity can be increased, and standing capacity can be counted in different ways), just like the trains themselves (e.g. some trains are more powerful than others). Also, the cost of the vehicle depends on the size and kind of the acquisition contract.

Nevertheless, it is obvious, given the tiny order, that the Talents were an amazing deal. They should’ve cost three times as much. And while the budget lists the vehicle acquisition as 5 million, a closer look reveals that they did in fact cost three times as much. Take this quote from the report:

The approach to vehicle acquisition was to buy the three trains for $17.1 million with a guaranteed buy-back of $13,537,500 after the two-year pilot phase for a net cost of $3.563 million.

Basically the trains were a lease for the two year pilot project, and had to be paid off when the project continued. This makes sense, given that the O-train started as a 2-year evaluation project, and it could’ve been discontinued after that. But it results in an unfair comparisons with other projects that use the full capital cost of the vehicles involved, and so that fact should be kept in mind.

It is still a pretty decent deal considering how small the order was, and how quickly the trains were delivered. Apparently they came so quick and affordable because they were surplus/option trains from a larger “BR 643”-order made for Deutsche Bahn (unfortunately I couldn’t find a source for that original order, or what exactly the option/add-on deal looked like). Presumably this explains the similarity with the DB trains, and some of the German signs that exist on the train.

DB actually preferred a more powerful version of the train, the “BR 644”, which uses a diesel-electric rather than a diesel-mechanical drive, which allows faster acceleration and mixes better with electric trains. Ironically, the diesel-electric version would be more appropriate for the O-train with its short interstation distances, Ottawa probably uses the diesel-mechanic Talents in its most rapid transit-like application.

The Corridor

The infrastructure was leased, just like the trains. The contract was with CP (Canadian Pacific Railway), who owned the infrastructure, as a public private partnership. CP provided the corridor, upgraded the infrastructure, built the stations, and provided the maintenance facility. The partnership also included Bombardier for trains, their maintenance and the signalling system. The lease included the provision to purchase the corridor for around 11$million at the end of the contract. There was also an option to extend the lease for two years – which the city exercised, and finally purchased the corridor in 2005 (at which point the city also paid half a million in deferred costs from the original design/build).

CP actually considered the corridor lease too low, so there was a clause in the contract for ‘incentive payments’: for every passenger above 5100 per day, CP was supposed to be paid 70 cents per passenger. By the end of 2005, 380K$ of those incentive payments had accumulated (it seems that CP waived about 200$K of those when the city purchased the line in 2005). This seems like another way in which the capital budget was kept a bit lower, possibly at the expense of future capital costs (or operating costs, depending on how the accounting is done). On the other hand, this may have also provided an actual incentive for CP to make the line attractive for users.

It is clear that the O-train was a cost effective project in terms of capital costs. But it may not be quite as cheap when deferred capital costs are considered. Once the total cost for the vehicles and and corridor are included in the capital cost estimation, the total cost is around 40$million, double the initial capital budget, but still pretty low.

Operating Costs

It should be expected that the operating cost of the O-train are not very favorable compared to buses. There is a lot of infrastructure and vehicles that need to be maintained. The fact that this is such a short line, with relatively few riders per km compared to actual rapid transit lines, and a tiny vehicle fleet doesn’t help, in terms of economies of scale. Plus, with three trains, two of which are always operating, at most two thirds of the fleet is operating. Such a high spare ratio should result in higher operating costs. Additionally, the project seems to have been designed with the desire to keep capital costs low, so costs may have shifted to the operations side.

It certainly helps that the ridership is pretty high given the scope and line length of the project. Nevertheless, the operating cost per passenger was higher than for buses during the initial years: during the first years, only about 30% of the cost was covered by ticket fares, compared to about 55% for the OC Transpo system as a whole.

Nevertheless there are cost-effectiveness measures on the operations side that are interesting and unique. For one, the O-train was the first federally regulated railway using single man operation (i.e. no conductors). Additionally, the drivers where actually taken out of the pool of OC Transpo bus drivers, instead of hired locomotive engineers. When the system started, 28 bus drivers completed the training to operate the O-train. This large reserve pool means that there is never a shortage of drivers, contributing to the service reliability of more than 99% (compared to 70% for the overall system). And when the drivers are not operating the O-train, they are still operating buses. This is an interesting approach not only to reduce the costs of reliably providing vehicle operators, it also addressed the concerns of the labor union, which saw the rail line as a threat to their members’ livelihoods.

Overall it should be obvious that the O-train was a very low-cost project. For a rail project, especially one of this small scale, it seems to have been very cost-effective. However, the costs may not be quite as low as they appear at first sight – one has to be careful to not count the start-up costs for the evaluation project as the capital costs of the transit line. Other projects attempting to emulate the success of the O-train should be mindful of that.

A Bike Rack on the Stuttgart Rack Railway

April 13th, 2013 by ant6n

Stuttgart Rack Railway

The above image that I came across on this tumblr, got me intruiged. The image comes with this short text:

Intermodal transportation involves using two or more forms of transportation in a journey. Bicycles can often increase ridership of public transit. However, limited storage space onboard hinders the promotion of intermodal commutes. This innovative solution from Stuttgart, Germany adds desperately needed bicycle accommodation on this light rail line.

Sure this is looks like an interesting solution, a bike rack on the move. Bicycles can help bridge the last mile (or two), between the transit stop and the home, and make commutes much more pleasant that way. But it immediately raises some questions. If people can get on and off with their bicycles, doesn’t that increase the dwell time and slow down the train a lot? If the vehicle operators observes the bike rack in front of them, what happens when the train goes the other way? This appears to be a terminus, and there’s no loop in sight. Also, doesn’t the rack appear a bit small, how useful could it be?

To answer these questions, one has to look at the actual transit line in question. The line in the picture is actually the “Zahnradbahn Stuttgart” (Stuttgart Rack Railway, the German Wikipedia article is much more informative). A rack railway, or cog railway, is a railway that uses a toothed rail, which can be seen in the picture running between the rails, to be able to climb steep grades. Normal rails can only be operated at grades up to 7~10% for trams, and up to around 4% for electrified main lines. This line has a maximum grade of 17.8%.

The Stuttgart Rack Railway is one of only four cog railways still operating in Germany, and the only one that is integrated into a transit system (i.e. not a tourist line). It started operating in 1884, and it’s really short. Along its 2.2km length there are 9 stations, it’s single track except for a passing station in the center, service is provided every 15 minutes. In its operation it is remarkably similar to the Ottawa O-train, a short single track line with a passing station in the center, and two trains shuttling back and forth every 15 minutes – except it has about the quarter of the length.

Given the short length of the line, it would seem like a line with not a lot of utility. Why wait fifteen minutes, if you can just walk the whole length of the line in that time? And it could be biked even faster! But the key to its utility is that it is a cog railway — along its short length, it climbs 200m in altitude. That’s quite a schlepp, and if you add a bicycle to the mix, the utility of the bike rack reveals itself – the bike rack is only used in the uphill direction, and only from terminus to terminus. This way there are no delays due to bikes when the train is running, and the driver can always observe all the bikes, and make sure none are falling off etc.

The Stuttgart Cog Railway is thus not a system that solve the last mile problem – it’s barely more than a mile long. Instead, it is more like an urban elevator and solves the problem that walking and riding in hilly areas can be a pain.

xkcd: Subways of North America

April 8th, 2013 by ant6n

Subways of North America

The webcomic xkcd put all real the subway systems of North America on one map, joined by imagined lines connecting the systems far away from another. Helpfully the author defines for us what he means by “subway”, for “the pedantic rail enthusiasts”:

a network containing high capacity grade-separated passenger rail transit lines which run frequently, serve an urban core, and are underground or elevated for at least part of their downtown route.

Sounds good enough to me.

Did you know that cities like Baltimore, Santo Domingo and San Juan have subway systems? That one in three subway stops are in NYC? That Springfield is located somewhere between Vancouver, Toronto and San Francisco? Well, now you know.

Walksheds Visualized
Showing Populations near Montreal Rail Stations

April 3rd, 2013 by ant6n

walksheds are the area around a particular point of interest from where people are willing to walk to said point of interest. This point of interest is often a transit station, and the walkshed gives an idea how many people will reach the station on foot. While the walkshed depends on how walkable the area is, whether there are barriers like freeways, the geometry of the street grid, and how attractive the transit station is (people are willing to walk further to metros compared to bus stops), The area is often simplified as a circle around the transit station.


Montreal rail station walksheds – population within 800m of stations. The sizes of the circles and the numbers inside them correspond to the population in 1000 people

How much population lives within the walkshed gives an indication on the ridership of the transit line, the more people live in it – the more ridership can be expected. The corollary is that you want as much population within walksheds as possible.

If there are more people near the transit station, it reduces the reliance on feeder buses and park’n’rides (a reduction in park’n’rides also means more land within the walksheds is available for development). This will reduce both the cost of transportation services that have to be offered, and will increases the chances that people use transit – which will reduce the cost of required infrastructure like roads and parking spots, but will also decrease the burden on the environment.

The above visualization shows the population within 800m (1/2mile) of Montreal transit stations. This is considered a reasonable distance that people are willing to walk to a rail station. Picking this distance for all stations makes the comparison between stations easier, even if the actual walksheds may be different depending on how walkable the areas are. Also, a circle with 800m radius has about 2 square km of area, so by dividing the number by two, we get the population density (in people per square km).

In the image, we can see that many metro stations have more than 20 thousand people living near them, corresponding to more than 10 thousend people per square km. Compare that to the a density of 3.7K people/km² for the whole island of Montreal, 7.7 people/km² for the borough of Côte-des-Neiges–Notre-Dame-de-Grâcee, and 12.3K people/km² for the Plateau. The highest population is near station Cote-Ste-Catherine, the 2nd most populous is Mont-Royal, then Guy-Concordia (which also includes the most densest spot in Montreal).

We see that the commuter rail stations generally don’t have many people living near them, especially off the island. Some stations are directly next to highways, others are surrounded by a sea of parking, others are in the middle of nowhere. There is quite a deficit in Transit Oriented Development near many of the commuter rail stations, a missed opportunity. In many places, development would be a better use of station-adjacent land, compared to parking lots, or very low density development.

The most populous commuter rail stations outside of the inner city are on the Deux-Montagnes line, at least up to Roxboro – the Deux Montagnes line is the the only electrified line, has the most trains, and by far the most ridership. Some of the stations rival metro stations in terms of population density. Other populous stations are on the Blainville line, at least up to de la Concorde – it would seem that this stretch could also be a candidate for electrification, possibly more infill stations and much more frequent service (incidentally that would also help relieve congestion on the orange line).

What is interesting is that some metro stations have surprisingly little population near them. The missed development opportunities may not just exist near the commuter rail stations, but near the metro network itself. While some of those metro stations are also directly next to highways (like de la Savane) and have limited development opportunities, others are in very suburban looking areas that could probably be densified (like Assomption).

The data was created using the Canadian census data (from 2011), the gtfs for the STM and AMT was used to find the station locations. And a disclaimer: the stations are considered completely independently; so population of very close stations may be “counted” more than once.

Ottawa’s O‑Train (Part I):
a Little German Train in Canada

March 29th, 2013 by ant6n

Passengers at Carleton Station.

One a recent(ish) visit to Ottawa I got to visit one of it’s major attractions – the Ottawa O-Train! I had read about it before, a small project that transformed a short stretch of a freight railway into a transit line, using “Talent” Diesel Multiple Unit trains built by Bombardier. Using these trains is unusual in North America, because they do not fulfil the requirements of main line trains with respect to buff strength (basically they are too light). Although these trains are considered main line trains in Europe, in North America they are called “light rail”. The three Talents that OC Transpo uses for the O-Train were acquired out of an order by DB (Deutsche Bahn), which uses these trains on un-electrified regional rail lines.

The interior of the O-Train should look familiar to people who have taken DB Regio trains

The interior of the O-Train should look familiar to people who have taken DB Regio trains. Note how the luggage racks were blocked off.

The arriving train came in a familiar color – the color scheme of DB and OC Transpo are so similar that the trains weren’t even repainted. The interior of the train felt strangely familiar as well, because it is all just the DB design. Having ridden many trains in Germany, I felt like I was sitting in piece of Europe in Canada. The ride is also unusally smooth by North American standards (I haven’t encountered too many rail vehicles with air suspension here).

Even the "Bitte druecken"-Button is original

Even the door opening button is still labelled “Bitte drücken” (please press).

If you like rail and transit, it’s easy to be a fan of the O-Train. But the real innovation of the O-Train is not the specific train that is used, but the transit planning of the line. It is easy to focus on a particular transit technology (in this case the DMUs), and overlook the transit line that is built with it. And some of of that planning was imported from Europe just like the trains themselves.

Overview of the O-train line. The blue lines are OC Transpo busways.

Overview of the O-train line. The blue lines are OC Transpo busways.

The O-Train is a short line established on a segment of an existing, infrequently used freight line corridor. Along it’s 7.8km length there are 5 stations. At both terminals passengers can transfer to busways. The station at the center, Carleton, provides access to the University, which is otherwise awkward to reach. The line is single tracked, except for the Carleton station, which has two tracks and two platforms. This allows two trains to run simultaneously, passing each other there. The trains take about 12 minutes to traverse the length of the line, and with a short turn-around time, this allows service with a headway of exactly 15 minutes.

The O-train is entirely single tracked.

The O-train line is entirely single tracked.

One train every fifteen minutes doesn’t sound very often by rapid transit standards, but one has to consider that the schedule is completely regular, the same every hour. The train leaves at :00, :15, :30, :45 and arrives twelve minutes later. This fixed interval schedule (or “Takftfahrplan” in the original German) allows people to easily remember the complete schedule, making the service more convenient even for less than regular riders.

Trains pass each other at Carleton Station, so two trains can run on the line at the same time

Trains pass each other at Carleton Station, so two trains can run on the line at the same time.

Another way in which the O-Train is more part of the transit network rather than a railroad in the traditional sense is the ticketing. Unlike many commuter railroads, the tickets are integrated with the surrounding rapid transit system, monthly passes are accepted, as well as day passes and transfers from buses. The line is viewed as part of the overall transit network, not some premium service. There are also no conductors aboard, and no turn-stiles either – the tickets are checked randomly, and there’s a fine for traveling without a valid ticket. This system, called Proof-of-Purchase (POP) is heavily used in Europe. It allows operation of the train with the vehicle operator only, no conductors required.

The 35m long platforms at the stations are level with the train entries. There are no steps into the train, and the doors are wide. Not only does this mean the train is easily accessible, it reduces the boarding and alighting time. Together with the reasonably fast acceleration of the DMUs (for example compared to diesel locomotive hauled trains), the resulting short dwell time allows the quick running time which again results in the relatively frequent schedule using only two trains.

Wide doors, Level Boarding

Wide doors, Level Boarding. Note the POP-sticker – that’ll remind ya ;-)

The O-train started service in October 2001 as a pilot project to evaluate the Light Rail technologies for Ottawa. Although the specific technology was not chosen for other rail lines in Ottawa (the Confederation line will use electrified low floor LRTs), the line itself has earned its keep. After the first year, the ridership reached around 6200 passengers per day, 9500 by Fall 2004 and close to 14000 by September 2011. A 2005 report, updated in 2008 (pdf, also includes some of the ridership numbers) noted that in 2007, discontinuing the O-train would have required the purchase of 16 extra buses at more than 7$ million.

Note, however, that the cost to revenue ratio for the O-train is worse than for buses: about 26% in 2002, and 36% by 2007, compared to about 55% for the system as a whole. So while the O-train provides a smoother and more comfortable ride, the ticket revenues of the train cover less of the operating expenses than for buses. It shows that rail, even a system as cost-effective as the O-train, needs good ridership and a decent size before it makes economical sense over buses.

There is now a project underway to increase service on the O-train, which is partly done to mitigate the disruption of the transitways while the Confederation Line is under construction. This summer, OC Transpo will shut down the O-train for 18 weeks to allow the construction of two extra sidings (and some other general maintenance). The purchase of 6 Alstom Lint trains back in 2001 2011 will then allow doubling the frequency of trains to every 8 minutes starting in 2014.

Another historical Map: 1949 Montreal Land Use plan

March 23rd, 2013 by ant6n


This blog wasn’t really meant to become about historical map mashups, but with the city of Montreal (and the Bibliothèque et Archives nationales du Québec) releasing all this interesting imagenary, I made the third map app within as many weeks. This one is a 1949 Land Use plan (source), showing a detailled view of the buildings, many of them named, their uses, street names. This map is not stitched together, and not overlayed with a modern view – it’s the raw scanned pages arranged in a grid. This makes for a primary source, not only showing the data of the map, but also showing the age of the paper it’s on, pencil scribbles, creases and tape fixes. The result looks like this:

Again there’s a full screen version, which can be found here. The 9 principal colors correspond to the following:

  • Jaune: résidentiel : unifamilial
  • Orange: résidentiel : duplex
  • Brun: résidentiel : triplex
  • Rouge: commerce
  • Rose: industries légères (alimentation, entreposage, vêtements…)
  • Violet: industries lourdes (mécanique, chimie, fonderie…)
  • Bleu: chemin de fer
  • Vert: parcs
  • Gris-vert: institutions (administration, écoles, églises…)

Note that some areas are not colored in: Those were not part of the city of Montreal, but independent municipalities. Montreal had quite strange boundaries and a weird shape, reaching the water in the North, but with many cut-outs all around. The city has consolidated much more, but some of those small cities (like Westmount, and TMR) are still holding on to their independence today.

map index

map index

All these map mashups I’ve showcased the last couple of weeks are a work in progress. I hope there will be one master map one day overlaying all the information in one app. This involves the major help of some pixel magicians who’ve been stitching and cleaning images (in particular mare and Luc). I thank them for all their work and hope for the continuation of these projects.