Although LRT is often proposed as a way to offer the service speed of a subway, at the cost of a streetcar, the infrastructure costs are still substantial and in some cases not much less than for a full subway.


Sheppard East LRT

The Sheppard East LRT will run 13.6 km, almost entirely in the centre of Sheppard Avenue, from Don Mills Road to Morningside Avenue. TTC is planning a short underground section at Don Mills Road, for interchange with the subway.

Sheppard is currently 7 lanes wide from Don Mills to Pharmacy Road, then 5 lanes to Markham Street, then 4 lanes. Provision of LRT in the median will mean reconstructing the road, and in most cases eliminating the left-turn lane. Metrolinx quotes a price of $1 billion or about $75 million per km.[1]

Scarborough LRT

Metrolinx issued a Benefits Case Analysis for the Scarborough RT replacement and extension project, dated January 2009. It examined four scenarios, including upgrading the line with existing technology for operation of four-car trains, and conversion to LRT technology with extension to Malvern. Unfortunately, none of the four scenarios that are evaluated matches the scheme actually included in Metrolinx's plans.

The BCA gives an estimated capital cost of $1.4 billion to rebuild the line with LRT technology, and to extend it on an exclusive right of way to Malvern Centre. For slightly less, $1.2 billion, the line could be extended to Malvern but with surface right of way beyond McCowan. TTC has more recently said it would cost $1.8 billion to rebuild and extend the line only as far as Sheppard and Markham Road.

At $1.8 billion for 10 km, the Scarborough LRT line would be considerably more expensive than the Sheppard Line,[2] or about $180 million per km. About half the cost is for conversion of the existing 6.5-km RT to accommodate low-floor LRT cars, with overhead power collection. This involves substantial reconstruction of six intermediate stations, and complete reconstruction of Kennedy Station to provide a larger underground loop, and track connection with the Eglinton LRT so TTC can exchange cars for maintenance purposes (but not for through-running with passengers). The balance is for construction of 4 km of new line, mostly elevated, from McCowan to Sheppard Avenue.

Note that at $180 million per km, the cost per km for the Scarborough RT is about 30% higher than the cost of the Evergreen Line, a fully grade-separated ALRT line in Vancouver,[3] even though the Scarborough line uses mostly existing infrastructure, and otherwise operates through a broadly similar corridor.

In the Scarborough LRT BCA, Metrolinx accepts TTC's position that expenditure of $452 million is "required" in the base case, simply to replace aging rolling stock and increase capacity to meet projected demand. This expenditure has the effect of reducing the incremental costs of other options, including conversion to LRT. The logic is sound, but we question whether the upgrading needs to be so expensive, especially if alternatives such as upgrading the GO Rail System can take passengers who would otherwise ride the Scarborough RT.

We question some of the assumptions in the Scarborough LRT BCA:

  • It is stated that as the ALRT Mark 1 vehicles "are no longer manufactured," the line must be rebuilt to accommodate the larger Mark 2 vehicles. This is not correct. Transit cars are all "built to order," and while there can be cost synergies in combining with another order, trains are not build on continuous assembly lines like cars or planes. Toronto's replacement streetcars are a special design, to match the geometry of Toronto's tracks. Certainly, TTC could procure replacement cars of the size and performance of the Mark 1 cars, if it wished to avoid the cost of rebuilding stations and curves. The cost per car would probably be higher, but total project cost might well be less. There is no evidence TTC or Metrolinx has considered this option, or enquired with Bombardier whether they would be prepared to build additional cars to the Mark 1 dimensions. The BCA notes that the line would need to be closed for at least 8 months for rebuilding. While it may make sense to upgrade the line for the larger Mark 2 cars, for other reasons, replacement with new cars built to the Mark 1 size could avoid any need to close the line down.
  • It is stated in the summary that the existing line has capacity of 4,500 passengers per hour per direction (PPHPD), and that this would be increased in the Base Case to 7,000 PPHPD. In Part B of the BCA, it states that 3-car trains operating at 2-minute headways would provide the "capacity requirement of 5,400 [PPHPD] by 2031." Actually, Mark 2 cars have a capacity of 130 passengers; so 3-car trains at 2-minute headways represents a capacity of about 11,700 PPHPD. Elsewhere, the BCA refers to expected demand of 10,000 PPHPD, about double current levels but easily carried with the existing technology, with additional cars but without rebuilding. Vancouver actually operates the Expo Line at 1.8-minute headways in the peak, with capacity of about 16,000 PPHPD.
  • It is stated that the interchange at Eglinton requires passengers to "travel three levels between platforms." While it is correct, the distance is not far and is partly served by escalators. It is already a very good transfer. We question whether it is necessary to rebuild the station, in the way proposed by TTC, or whether it might be sufficient to add another escalator or two (which could also be much cheaper).
  • TTC seems to be requiring a fairly elaborate and expensive yard. The BCA (which was prepared by consultants) notes that "The cost of a Vancouver facility with comparable capacity was roughly $200m lower, although the yard alignment and maintenance practices differ from the TTC's." If Metrolinx thinks there may be the opportunity to save $200 million, surely it should give this more attention than a short footnote?

In our evaluation, we do not "net out" the $452 million "base case" costs, because we do not accept the need to rebuild the line as the "base case." TTC might need to spend $200 million replacing the train fleet and making other expenditures to restore the line to a "state of good repair," bringing the incremental capital cost of the TTC scheme (converting to LRT and extending to Sheppard) to $1.6 billion, or $1.4 billion excluding rolling stock. Taking 70% of this gives an NPV of $980 million. We use this figure in our evaluation (it has the effect of substantially reducing the Benefit:Cost ratio, actually below 1.0).

Eglinton Crosstown LRT

The Eglinton Line is far more expensive, with capital costs of $4.9 billion, according to Metrolinx.[4]

The less expensive sections are from Laird Drive to Kennedy in the east and from Keele to Jane in the west, on the surface in the median of Eglinton Avenue for about 11 km altogether. Eglinton Avenue will need to be rebuilt, similar to Sheppard, so the cost of this might be $75 million per km or $825 million, similar to Sheppard West.

The expensive parts are the tunnels, 11.25 km in deep bore from Laird to Weston Road, and for 12 or 13 underground stations. The tunnels will cost about $1 billion, or $100 million per km. The stations add a further $150 to $200 million each, or about $2.4 billion. The power system, a maintenance yard at Mount Dennis, design and project management fees, and contingency for overruns make up the remaining $700 million.

Finch West LRT

Metrolinx estimates the cost of this line at about $1 billion, or $100 million per km,[5] the same total as the Sheppard West LRT. It is not clear why the costs are the same, as the Finch line is only 11 km long, and Sheppard line is 13.6 km long.

Rolling stock

When it proposed Transit City, TTC was already procuring new low-floor streetcars, to operate on its existing surface routes. An $851-million contract was awarded in 2009 to Bombardier for 204 "Flexity Outlook" cars, or about $4.17 million per car.[6]

In 2010, Metrolinx signed an order for a further 182 cars to operate the Transit City routes. The price was $770 million or $4.23 million per car.[7] While the initial order is for single-ended units that are designed to operate on Toronto's non-standard track, the additional "Flexity Freedom" cars will be larger, double-ended, 2.65 metres wide, and operate on standard 1435mm gauge track. It seems that, by making the Transit City order an "add-on," Metrolinx could avoid an international competition, as required by World Trade Organization rules, ensuring political support for the procurement because it will create local jobs.

Although it is difficult to make direct comparisons, Bombardier's price seems high for such a large order. For example, Siemens recently won a $153-million contract to supply 41 low-floor LRT cars to Minneapolis,[8] at an estimated price of $3.73 million per car.

Operating costs

Metrolinx has provided estimates of incremental transit operating costs for the Eglinton Crosstown line, the Scarborough LRT schemes, and for the combined Sheppard-Finch scheme, in its BCAs. The Metrolinx forecasts are much less than ours, with the Eglinton Crosstown scheme actually costing $417 million NPV less to operate than the bus services it would replace. Without further detailed information from TTC, it is not possible to verify whether this estimate is reasonable.

Our estimates for the Scarborough LRT scheme are very similar to Metrolinx's; we estimate incremental operating costs of $12.5 million per year, offset by bus costs savings of $1.5 million per year for a net additional annual O&M cost of $11 million per year or about $250 million NPV. Metrolinx costs are slightly higher: $14 million per year O&M costs additional to the Base Case, offset by $1.8 million bus cost savings.

[1] Data for Transit City Light Rail schemes is mostly from Metrolinx project fact sheets,, and from the TTC Commission presentation "Transit City Update" of September 30, 2010. According to TTC, the Metrolinx funding for the Sheppard line is $1.13 billion including escalation. Metrolinx figures are in 2010 dollars, while the TTC prefers to use the expected "outturn" cost, including inflation during the construction period.
[2] This is the Metrolinx figure for a 10-km line ending at Sheppard. The TTC figure is $2.465 billion for a 12-km line extending to Malvern Centre, and also appears to include inflation.
[3] According to the project website, capital cost for the 11-km line is $1.4 billion. Rolling stock is a further $90 million (see, bringing the total cost to about $1.5 billion or $136 million per km.
[4] The TTC quotes a figure of $6.065 billion for the 10-km line from Jane to Kennedy, including inflation. The figure in the Table shown here is lower, because it is a Present Value.
[5] The TTC Transit City Update gives a figure of $1.28 billion, again including inflation.
[6] See The "headline" price was $851 million, but the total contract price was estimated at around $1.2 billion including taxes, spares, and maintenance.