A Guest Post by David Marlor
David Marlor was raised in the UK and is currently a regional planning manager working on the coast of British Columbia, Canada in a coastal rural setting. He holds a planning degree from the School of Community and Regional Planning at UBC, where his thesis was on an integrated approach to transportation planning in the lower mainland.
In the past couple of years, led by Bob Boutilier (general manager of transportation) the City has been planning the expansion of the LRT using low-floor technology. Edmonton is credited with leading in transit innovation twice in the past. In the mid-1960s, transit superintendent Don MacDonald introduced an early version of a hub and spoke [or pulse] transit system. This is still widely used in Edmonton and in many other North American cities. The second was the introduction of the modern LRT to North America in 1978.
What’s different about the current LRT plans in Edmonton is that instead of fast LRT trains moving commuters from suburbs to the city, the LRT will be a European style system, still in its own right of way, but with stops closer together, smaller and more intimate with the community, low floor vehicles and replacing car lanes with LRT lanes. That last is a paradigm shift for Edmonton. LRT up until now has been about building it without removing road capacity for private automobiles. The new LRT lines, the cost of which is currently pegged at about Cdn$3.4 billion, will see extension of the existing high-floor system to the north-west north-east and south as demand warrants, but the lines will fit the community better than before.
The approved plans includes a completely new low-floor network running on the street, even in the city centre. It is a system designed to support future TOD at the stations, to encourage higher densities. Unlike the existing system, only five stations on the proposed 29 station low floor line have bus stations attached to them and only two have park and ride facilities, both adjacent to freeways. This is about shaping the city, not moving commuters from the suburbs to the city (although that is part of it, it is not the focus). The plan includes future low-floor line linking the downtown with the Old Strathcona business district on the Southside of the river and a line out to the eastern suburbs.
In fact, the proposed the proposed Edmonton system may remind some readers of Strasbourg. Like Strasbourg, Edmonton is envisioning completely remodeling the streets the trams run on – in many cases removing lanes of traffic, restricting turn movements, closing or redesigning intersections, and where possible, widening and improving the pedestrian infrastructure. Like Strasbourg, the stations (stops) will be located every 3-4 blocks (300-400 metres) in the city centre and further apart outside the core. The aim is around every 800 m, but in reality the stops will be placed at convenient nodes or logical locations that best fit the fabric of the city.
The City of Edmonton website has extensive information, including design details, routing, and illustrations. Unfortunately, the project is not funded yet, but City Council and the Mayor are keen to see it happen and want to get it built in the next 6-8 years. The plans are ambitious, and it's exciting to see a car-oriented oil producing city like Edmonton be thinking and supportive of this direction.
Illustration: Simulated image by City of Edmonton
Hi David,
I guess a common question would be “why didn’t you just choose BRT?”
Perhaps LRT allows for potential for integration with the exist rail system in a way BRT wouldn’t?
Edmonton actually had a $1 billion plan for BRT to complement the existing LRT. However, thinking changed to “if we’re going to spend that kind of money, might as well go with rail”. Generally, in Canada at least, trains are more appealing than buses. Rail never fails to draw more riders than enticipated, especially people who prefer to drive than take a bus. Rail gets them out of their cars.
Edmonton current LRT when the line ran from the University to the northwest, carried about 40,000 per day. That jumped to over 100,000 per day when the four station extension south opened. And suddenly the city saw the types of numbers than LRT to all sectors of the city could generate.
I guess the concern was that BRT is expensive to do right (rights of way, etc). Ottawa is now converting their BRT to LRT and yet there appears no saving to that city in construction cost. Finally, buses are more costly to run – labour and fuel.
Unfortunately Edmonton seems unwilling to introduce a higher-order bus service on the future LRT lines while construction waits.
Model it on Vancouver’s B-Line: Put in an all-day service at 10 or 15 minute headway; put the stop spacing and location roughly to match the future LRT routes; use distinctive buses and shelters. This will build ridership and start to encourage the same TOD development that will support the future LRT route.
Thank’s for your response David M.
What are the per-km costs of LRT in Edmonton?
Obviously I object to a claim that "buses cost more to run." Unit costs of LRT operations are usually higher than for buses, but of course they have a much larger capacity. The real measurement, I think, should be passenger capacity per driver, where rail excels so long as it has no more than 1 employee on board.
As for fuel, just compare LRT to electric trolleybuses. The rail/bus distinction is different from the propulsion distinction.
None of this is an objection to Edmonton's LRT, but there's a need for clarity around why.
Electric Trolleybus BRT, I hadn’t thought about that one. Brisbane’s buses have carbon offset so even though they are run on diesel and CNG, we can still be green.
Sorry if I have a lot of questions (I live on the other side of the planet unfortunately), but its an interesting system, as I think Edmonton was a city that built LRT before it had 1 million people (correct me if I am wrong here).
What is the peak hour frequency of the LRT and the off-peak frequency, and would you have figures for the passengers/hour/direction, theoretical pphd capacity of the LRT and how many people can fit into one LRT vehicle (I am asking as it looks like you use up to four LRT vehicles connected together at once)?
I’m also interested in how the system gains ridership- how do people access the LRT, are there park and rides or feeder buses or just high density TOD around stations?
Was there problems with people transferring to LRT- i’m guessing there might have been a previous direct bus system before the LRT was built.
Edmonton’s existing line has 5 minute frequency at peak times, 10 minute midday and Saturdays, and 15 minutes evenings and Sunday. There are over 100,000 riders per day, and the existing line is just 20km,connecting downtown, the university, two major sports venues, and the northeast and South Suburbs.
Before the proposed low-floor line is built, a 3km+ branch off the high-floor line wit h3 station, serving the community college/planned arena, a major hospital and the tech school will be built, at a cost of about $200m/km.
“Unit costs of LRT operations are usually higher than for buses”
I would be interested in knowing why the unit cost is usually higher, what contributes to this? Is it the maintainence? Lower ridership? More expensive vehicles?
LRT has maintenance of way to deal with, as well as power and signals, stations, and dispatching. And trains are typically maintained to somewhat higher standards than buses, partly because of regulations and partly because a broken down bus inconveniences the passengers on it, while a broken down train can shut down the whole line.
Brisbane. The cars are Siemens U2s and SD-160s. There are 37 U2s and 37 SD-160s. Currently, power supply allows the U2s to run in four-car formations and the SD-160s in three-car formations at a maximum peak frequency of 5 minutes (the SD-160 draws more peak power on acceleration but less in cruise).
Both vehicles hold 260 passengers (64 seated). The city is upgrading the power supply along the line so that 5-car trains can operate (125 metres) at 5 minute interals, plus allow power for additioal trains that will run in the tunnel section when the new north line opens.
Nice thing about the LRT is the length of trains can be adjusted for the expected ridership. In quite times a one car train can operate, saving power consumption and wear and tear.
Edmonton has no TODs to speak of. Access to the LRT is via integrated bus connections at the stations. Buses all feed to the LRT. Fares are integrated (bus ticket can be used on the LRT and LRT ticket can be used on the bus). Buses use the pulse system to connect with each other. In the evenign, when LRT is every 15 minutes, the buses are timed when possible to the LRT too to reduce waiting. Some stations have park and ride lots too.
The LRT replaced a lot of direct buses from many neighbourhoods. I don’t think there was any problems with the change as the conneciton offered a much faster service than the bus could.
On the idea of using trolley buses for BRT that would not happen in Edmonton at present. Maily because the city made short-sighted decision, in my view, to dismantle an existing trolley bus network. Administration nor council would suggest building a trolley BRT because it would undo the arguments used for dismantling the trolley network. It just wouldn’t fly at present.
I suppose another consideration is LRT cars have longer life than buses, especially diesel buses. Edmonton’s first generation U2s are now 33 years old and are still running well. They’re being refurbished for another 30 years of service. Buses are usually retired after 15-18 years, maybe less.
Edmonton is similar to Calgary, which also has a small but very well run light rail system: http://www.calgarytransit.com/route_maps/lrt_stop.html
It has surprising high ridership, especially for North American standards:
http://www.calgarytransit.com/html/technical_information.html
In 2005 the light rail system had 180,000 weekday trips, in a Metro area of just over 1 million people: http://www.lightrailnow.org/news/n_newslog002.htm#CAL_20050126
There are some major problems with the new scheme, the biggest of which is the cost is not expected to be substantially lower than the old-style rapid LRT.
Some others:
-On-street operation downtown will limit train length, due to short block length. This could be a problem in the long term.
-Other than a few bus transfer station in the farther suburbs, there has been little consideration for integration with the bus system. This is important because we are a sprawly city and we cannot all have LRT within walking distance.
-The longer term plan, with links to an eastern suburb etc. is very poorly thought out. It tries to provide one seat rides at the expense of better service, presumably out of a fear of connections (transfers). It also replicates the existing LRT connection between downtown and the university across the river, presumably for the same reason.
-The criteria weighted heaviest when administration evaluated routes was spurring redevelopment. Seriously. It was above mobility, access, ridership, anything. Which is why this west line was chosen over an option that attracted higher ridership, and offered trip times 5+ minutes faster to downtown and 15+ minutes faster to the university, for all riders arriving at the 5 farthest stations, including every single bus transfer on the west line.
For all that, the downtown section appears to be very well designed, and other than one small detour and a handful of infill stations the SE line is not very different that what would be built if the intent were to build a rapid line, and service should be reasonably quick on that leg.
Hi David, thank you for your response, just a clarification-
“Both vehicles hold 260 passengers (64 seated)”.
Is that just one LRT vehicle by itself, or the entire consist of vehicles?
Example: Would I be correct to say 3 LRT vehicles tied together is 260 pax x 3 vehicles = 780 passengers per ‘train’?
And 5 car ‘trains’ = 1300 passengers?
Was there any discussion or doubts over the potential success or failure over the construction of LRT given the density of Edmonton (not sure of the figure- are you a high density city?) or the population size at the time (1 million or less as I am guessing here, at the time of construction?).
I think something very important is missing from the cost comparison between Bus Rapid Transit and light rail, the underground utilities.
My understanding is that on a route where the municipality just replaced major trunk sewer and storm drains (without planning for future rapid transit), BRT can have a very significant but hidden cost advantage. This is because even trolley buses can drive around excavations for repair work, making it acceptable to leave the utilities in place.
I would like to find some examples of costs for utility re-location for rapid transit, and examples where creative thinking has reduced costs. For example, on some major roads the sewer and storm drains run down one side, and putting both light rail tracks to other side might eliminate the need to replace the utilities.
Also, it would be interesting to find out if any cities have started identifying future transit right of ways so utilities can be re-located when they need to be replaced (every 50 to 100 years or so). I would guess that some cities have kept their original streetcar rights of ways free of major utilities and planned for future LRT routes, and others have not.
David, do you know if utility re-location is a major cost factor in the Edmonton LRT project?
Brisbane – your numbers are correct. I think the 260 pax per car is crush (ie major events), more reasonable would probably be around 180 or so per car. But generally according to manufacture specified capacities, your numbers per train is correct.
Eric – the cost estimate for the LRT lines includes everything – road reconstruction, utility relocation, land acquisition where needed, tunneling, bridges (part of the west line is elevated, one major river crossing, one major tunnel and some shorter structures to grade separate freeways, railways and major roads),rolling stock, everything involved.
Edmonton started planning for LRT in the 60s and opened the first line in the late 70s. There’s always debate on funding – with freeways and road construction largely winning out over the last 30 years. In fact LRT construction all but stalled due to costs and limited funds. Edmonton,like Canadian cities, has relative high transit ridership for North America so I don’t think there was any doubt the lines would be successful, just whether it might be better to spend funds on just providing more buses to more parts of the city (or build bigger roads).
If this goes ahead as planned, I see it ushering in a new regime of LRT planning in North America – a switch from the suburban rapid transit LRTs with park and rides (like Calgary) to more urban TOD supporting LRT that displaces auto traffic.
Normally this is an excellent and readable blog. Maybe for future posts, you could make it a rule to explain the TLA’s. As far as LRT, BRT and TOD, I have NFI.
“Edmonton,like Canadian cities, has relative high transit ridership for North America.”
Yes, Jarrett had a graph up a while back (below) and all the Canadian cities were consistently showing unusually higher PT mode share, generally higher than US and Australian cities. We are all scratching our heads as to why this is the case.
As you don’t seem to have high density TODs around stations (and I constantly hear how LRT MUST have high density to be support it- for the record I don’t believe that anymore), I’m guessing that most of your patronage therefore comes from park and ride and feeder buses networking the suburbs and then concentrating passengers at stations to levels so high that they support LRT operation.
So I would hypothesise that maybe the reason why Canadian cities show higher Journey to work mode share is possibly because they have better integration between modes?
What proportion of LRT passengers in Edmonton have transferred from bus, vs walking, cycling, car etc?
https://www.humantransit.org/2010/10/further-cause-for-canadian-triumphalism.html
Edward – sorry about the jargon
LRT = Light Rail Transit – a form of rail rapid transit using lighter vehicles, partway between streetcar and metro
TOD = Transit Oriented Development – multi-use development oriented around a transit station
BRT = Bus Rapid Transit – a fast express bus system making limited stops, usually, but not always, on own right-of-way
Two comments:
1. In Alberta, the wages are very high, which tilts the playing field in favor of LRT over BRT. Calgary accelerated BRT-to-LRT conversion because it realized paying for tracks was cheaper than paying bus operators high wages.
2. I don’t care what Edmonton thinks of itself; Calgary’s far ahead of it in light rail use and transit mode share.
@In Brisbane: Edmonton is not a dense city. Neither is Calgary, though Calgary’s downtown has densified as a result of LRT and parking restrictions.
For all technical data on LRT, see the Calgary Transit link provided by Joseph E.
@Alon – Calgary is far ahead in LRT ridership numbers because of the downtown office worker population and the parking restrictions (which have equaled very expensive parking). When the number of workers are 3 times that of another city downtown (due to a large O&G head office population), these things have an impact on the statistics.
@Dave: yes, but those restrictions and the downtown job density are both results of deliberate city policy. While Edmonton just built LRT, Calgary built LRT (at much lower cost) and also encouraged TOD in the downtown in order to increase use.
Humm. Why can’t we ever have a discussion about transit Alberta that doesn’t turn into Calgary/Edmonton is better argument. I used to live in Edmonton and I found this Alberta attitude quite “interesting and unique”. Both systems are well used. Calgary’s LRT is more complete than Edmonton’s, serving three main areas of the city, so the ridership numbers are high. Edmonton is a single 20km line, but that also has high ridership numbers. It is true the Calgary city centre is much larger with much more office space than Edmonton. Think of Calgary as the corporate head office and Edmonton as the blue collar working city with many refineries and industries distributed outside the city centre.
@Brisbane, regarding densities required, depends. Lower density environment obviously tend have more space and wider roads so it is easier to just reserve car lanes for track. Just laying track on street doesn’t cost that much, and in less crowded environment you get better speeds, that is lower OPEX. So LRT can be economically sound even with relatively low passenger numbers.
German rule of thumb, and IIRC govenment funding criteria, is 2 000 people within 400 meter of a stop per kilometer of track. With 800m stop spacing that is about 3 000 people per square kilometer.
3000 persons/km seems extreme. There are some small “niche” applications of LRT in Australia (Melbourne/Adelaide/Sydney), and as I understand it, and none of our cities come even close to that figure I would think.
I’m forming the view that the “density required” is actually nothing to do with the LRT itself, but simply the density required to run the feeder buses to support the service.
After all, density really seems to be a proxy measure for trying to gauge how many people might access the station, (and there might even be a small unspoken value judgment there that they all must prefer walking to access it).
If you can build patronage to high levels by bussing people in or getting them to park their cars or ride a bike, then I think that the numbers to support LRT operation can be built up, despite what the density immediately around the station might be.
@Brisbane, feeder busses are of course a different thing. Assuming three people per house 3000 ppl./km^2 is 1000 houses and 1000m^2 per house ignoring space taken by roads and such. I don’t think that is particullary dense, cities can have ten times the density easily.
There’s nothing quite like measuring the wrong thing.
I agree with In Brisbane that density is a proxy measure for how many people might use a line, and a bad one at that.
What matters is ridership per unit distance of track or route, or what one might call lineal density of ridership. That’s all that matters in determining if a line or route will be economic. It’s similar to the carloads per mile measure used by the railways to determine if a particular railway line is economic.
How passengers get to the line is less important than that they do so. Country- and city-specific considerations come into play here. A higher propensity to walk can have as much of an effect as land use density. Or perhaps your city has a roller-blading craze, or maybe you’re really lucky and are in a city with high cycling rates. Since bicycles effectively extend the “reach” of a station from several hundred metres to a couple of kilometres, differences in cycling rates can have a major impact. High land use densities around the line might well increase the ridership, but it might not, as well. Evidence from a combined 30-storey condo tower and townhouse development near Westboro Station in Ottawa indicates that the lower-density townhouse development is a far greater boon to ridership. Things like Park & Ride lots and feeder buses have their roles in altering the lineal density of ridership number. From a transit planning perspective, you want to be looking at everything that increases ridership, not just getting fixated on density. Having mixed uses in the immediate proximity of a station that serves the surrounding area, along with mixed-use high streets leading away from the station, may well be more useful than a single mono-use but higher density development. It seems that we hope that putting a lot of people around a transit station will automatically result in more ridership – a sort of brute force rationale in that a small percentage of a very big number is still a big number.
Our streetcar systems used to be economic. Few of them experienced density reductions in the surrounding lands (in fact, many probably experienced increases) but what they did experience was a reduction in ridership. Ridership dropped below the threshold for them to remain economic. Density was not the issue – a reduced propensity to use transit was.
I suspect that in Edmonton and Calgary, LRT wins out over BRT or expanded bus service because a big LRT project can attract provincial and federal contributions.
Expanded bus service or BRT-lite would come largely from the municipal and transit authority budgets
@ David in Ottawa
“How passengers get to the line is less important than that they do so.”
Yes. I am starting to think that the “direct transfer free trip” will be brought into question here. There are two logically separate components to this- the trip to the station (whether the station be served by rail or bus) and the transfer.
Clearly for the people of Edmonton, transfer does not seem a big issue (and I have seen presentations claiming 12-25 minutes “Transfer penalty” for the Edmonton System, which I would have expected to discourage ridership), the important thing is that people can access the station from their street stop.
On a open BRT system, like Brisbane’s, the bus continues directly into the busway system. On a closed BRT system, the bus goes to the busway station and you transfer (Bogota?). In both cases you have access to the station from places that are far away from it.
So I do not see any reason why you cannot replicate this system on rail based modes. It is not necessary to construct “branches” off a rail system when you can just run BRT to the rail station. And you get the flexibility to change, or express or alter routes on the suburban feeder bit if you like.
Perhaps the dislike of transfer is being overstated somewhat, and the main issue is that people just need access to the station (whichever mode might serve it) from the street. And on rail systems that might mean catching the bus first.
Brett wrote: “LRT wins out over BRT or expanded bus service because a big LRT project can attract provincial and federal contributions.”
Which is all fine unless local politicians commit to something that needs big provincial and federal contributions, but the money from higher levels of government never quite materializes. If you want rapid transit within a reasonable time scale, it pays to stay flexible and go with low cost bus rapid transit lines (e.g. re-marking existing lanes) if the cash is not there for rail or high end busways.
One way to look at this is to think about how many years, or even decades, you might have to wait to save 5 minutes travel time.
@ Eric Doherty
I think Calgary did this, with the blue arrow (?) express bus network before they built LRT. While costs are going to be limiting on any project, I also feel that improved service with the LRT or high end busways would also increase the benefits, so both costs AND benefits need to be considered here.
So often I hear arguments about centred around a fixation on costs, particularly up-front capital costs, and little regard to total lifecycle cost (building + operating) and almost zero discussion about possible differences in the magnitude of benefits of each option.
For example, sending an e-mail to every resident in the city extolling the virtues of public transport might be cheap, but also have less benefits both absolutely and proportionally, than a more expensive transit promoting project.
I feel that there is a possibility that a “stop gap” solution can sometimes be seen as the “end” solution or a cheap substitute for the “end” solution, when it really should be the means to the “end” solution- high end busway or LRT. This is not an argument for or against stop-gap solutions, just something to consider.
Bottom line: you have to pay for quality.
There is always a need to have “while we wait” projects to cover us while more expensive projects are taking shape. This is a good idea, in this context.
The question for me is this:
How much of an improvement is over existing service?
The material presented sounded like desires of urban form trumped any empirical notions of improving mobility. Re: the access/mobility debate.
Also: it’s not important how many passengers use the line as to the passenger-miles on the line.
CroMagnon: I’ll admit to being more familiar with Calgary than Edmonton. In Calgary, the transit mode share has noticeably increased as LRT has opened and expanded, going up 5 percentage points from 1996 to 2006. If you Google “Commuting Patterns and Places of Work of Canadians” and the year 2006, you’ll be able to see similar statistics for Edmonton, I’m sure.
And passengers are a lot more important than passenger-km in both Calgary and Edmonton, because the fares are flat.
So it may have increased, but that’s a separate issue to mobility improvement.
Sorry, my comment about passenger-miles was referencing the importance of utilization (per an above comment) as a rationale for mode justification, not fare structure.
This sounds very similar to the LRT that Toronto was proposing (on hold thanks to the new mayor). To replace bus routes with LRTs which make frequent stops every quarter mile or so.
Personally, I find calling these projects “LRT” somewhat misleading. While it technically means “light rail transit,” it might as well be called “light rapid transit.” Even the previous mayor of Toronto who put forth our plan mislead many by referring to it as “rapid transit” while not disclosing the actual stop spacing until much later.
Personally, I prefer to call these local rail systems “tramways”, since they share more in common with European trams/streetcars rather than metros/subways like traditional LRTs.
Rapid streetcar?
As a professional who formerly participated in TOD Design Plans around two potential stations in Edmonton, the main challenge is to get the developers on board to ensure ridership via sustainable and responsible development. To understand what I meant, please read this position statement by the Edmonton Chapter of the Urban Development Institute here: http://www2.canada.com/edmontonjournal/news/ideas/story.html?id=a933b653-78c7-442f-aa1b-ae844d40378a
To clarify a few points.
1. The original Edmonton LRT was modeled after three systems that were built in the 1950’s – Montreal for the underground pedways, Toronto for the infrastructure, and Cleveland for the rail r.o.w. alignment with bus feeders and parking. These original influences were modified by influences from Germany and historical North American streetcar and interurban practices. It was renamed Light Rail in 1977, about a year before it opened. [Edmonton’s largest ethnic group during the approval process for the first line was German and it and Calgary had over-the-pole air competition to the Continent before most U.S. cities, so a large part of the non-transit professional part of the public knew about the general idea.]
2. What was different from the North American ancestors was that in 1962 Edmonton had begun developing the timed-transfer focal point concept for its bus and trolley coach lines, which permitted a lower density Western city to offer convenient transfers — “civilised” connections, as Llew Lawrence of Edmonton Transit.
3. In regard to Canadian higher ridership, it’s a result of numerous policy and legislative differences that I won’t take space here for, but it has long been noticed that there is a tendency for northern U.S. cities to do better, too. The biggest difference is the lack of Federal intervention in urban planning on the U.S. scale.
4. Although trolley BRT in a city where natural gas generated power is available might make sense, the current management spent a lot of effort in getting rid of the residual trolley coach system, in spite of city policies to the contrary. This is somewhat similar to West Berlin, when money was flowing in to build U-Bahn lines to replace tram lines in the 1960’s and 70’s. In both cities, the management became fixated on NOT considering the forbidden alternative further.
I came along in 1976 and had the privilege of working with the people in Edmonton Transit who set all this into motion. They were also responsible for selecting the Siemens-DueWag car, which was to become the standard North American LRV.
Edmonton also pioneered self-serve / Proof-of-Payment fare collection in North American rail practices and adopted numerous other technical experiments and innovations during that era, until it went into hibernation with the 1982-83 energy crash. Calgary, the more risk-taking of the two cities, continued its progress and overtook Edmonton in that era.