The TTC commissioned a report from Dr. Richard Soberman on the economics of trolley bus operation in Toronto. Cutting to the chase, the conclusion is that creating a new system from scratch is uneconomic, and we should wait for coming improvements in electric vehicles.
Soberman’s report makes a strong case against trolley buses on its basic economic arguments, and that’s a debate worth having. However, electric vehicles have yet to make a substantial dent in the personal car market, let alone for vehicles the size of a city bus.
I have one simple reply: Remember CNG? The saviour of the enviroment for the TTC? We lost the old system through neglect and through belief in an unproven technology, not to mention political machinations.
For your reading pleasure:
Updated February 15:
A detailed review has been added to this post. Some of the document is reasonably accurate, but there are enough outright mistakes and misdirections to cast the whole thing in an unsavoury light. This is a report that tries to sound balanced while hoping we won’t notice what it gets wrong either by accident or by design.
Update 2, February 16: My long-time Vancouver friend Angus McIntyre pointed out two issues with the Soberman report.
- Vancouver will order an additional 34 articulated trolley buses funded from the Federal gas tax. This adds to the new fleet of 188 standard and 40 articulated buses. These plans are not reflected in the TTC report even though the press release is over a month old.
- The substation spacing of 1.5 to 2km is a measure used on “feederless” systems such as Seattle’s where small local stations feed directly into the contact wires rather than the Toronto or Vancouver model with large substations feeding a local network of services. Has the report used Seattle’s close spacing, but Toronto’s costs for larger substations?
The report proposes a network based at Wilson Garage because power is already available from the nearby subway complex. Whether this is entirely germane is dubious considering that modern trolley bus garages tend to be largely without overhead and vehicles use their auxiliary power units to move around the yard. Some power is necessary, but whether enough to run a fleet of 150 to 250 buses is another matter. Similar considerations apply for overhead wiring which no longer needs to be the spiderweb of specialwork found in classic garages like Lansdowne at its height.
Substations are required at intervals from 1.5 to 2 km, although the lower bound is used in the calculated costs. This is quite amusing considering the scope of the former trolley bus and existing streetcar systems. The existing streetcar system (ignoring non-revenue trackage) has about 80 route-km of overhead, and this would imply a substation count on the order of 40 (at 2km spacing) or 52 (at 1.5km). The TTC has many, many fewer substations and of course these are shared by nearby routes.
Vancouver’s system in 1993 had 13 routes, 305 km of wire and 20 substations. (See T2000 BC’s overview of the network.) The ratio is ten times lower than the substation spacing claimed in the TTC study and shows what happens when resources are sized for a network rather than a hypothetical, stand-alone kilometre of construction.
One major issue is substation design. The TTC opts for large buildings including land where their service vehicles can park. If a substation will feed a considerable area, that may make sense, but not at the fine-grained spacing under consideration here. A good counterexample can be found in Vancouver where the UBC extension, in service since 1988 to a bustling campus bus and trolley bus terminal, uses “substations” small enough that they can hide behind hedges at the side of the road. This approach was used in recognition of the sensitivity of the UBC campus and of the fact that the extension would be an outlier of the overall system requiring only power for the local segments, not for other nearby lines fed from a larger, traditional substation.
Both the number and design of substations is important in the context of $2.7-million each, equivalent to $1.8-million per route kilometre on a 1.5km spacing.
Updated: As noted above, the close spacing cited in the TTC report may be based on Seattle’s feederless design which uses small substations and thicker-than-average contact wires to avoid the need for a feeder system. It is possible that the report has mixed this design with the larger cost of traditional Toronto-style substations.
Finally, overhead construction costs include an allowance of $50,000 per intersection for “special treatment”, and there are assumed to be over 5 intersections/km (220 spread over 42 route km). The last time I looked, trolley bus routes go through intersections with no special treatment at all because the pole spacing at these locations (typically the intersecting road’s width) is generally closer than the spacing on regular parts of the route. If these costs are adapted from Transit City, they may be taking into account special costs for different pole and suspension arrangements built into the LRT right-of-way design (see St. Clair for examples) that are not applicable to a trolleybus network.
TTC Engineering and Management Overhead
All of the base costs used for the power distribution system are factored up to include TTC management overhead at the following rates:
- contract changes (10%)
- design and engineering (25%)
- contingencies (30%)
yielding a cumulative total of 78.5%. I cannot find a polite word to describe such a blatant distortion of cost estimates. A trolley bus overhead system is not something that requires detailed inch-by-inch design and management. Once you establish the general parameters, you go out on the street, install poles, string feeder wire and install the overhead. The cost of design does not go up just because you extend a route another kilometre.
The report notes that recent cost estimates for Vancouver are “somewhat lower” but that Vancouver has a considerable sunk investment in electrical infrastructure. It would have been nice had the author bothered to find out why Vancouver’s costs were actually lower rather than suggesting that Toronto doesn’t have the foundation of electrical infrastructure on which to build. I will try to remember this the next time I hear the subway rumble across the Bloor Viaduct (in view outside my window as I write this) or ride the King car down Broadview.
The combined effect of overstating the cost of substations and larding the entire estimate with overhead charges (pun intended) that should be distributed (you can groan now) over a larger network is to substantially inflate the projected cost of the power distribution infrastructure. This is addressed briefly by a chart showing the effect of changing the assumed ratio electrification costs and future diesel prices, but the overall impression left with the casual reader is of a wide gulf between diesel and TB options.
Fleet Size and Network Layout
The TTC’s own studies have shown that a fleet should contain at least 150 vehicles and that 250 would be ideal. However, there is no attempt to sketch out a network that would yield this fleet nor to determine the economies of scale this would bring.
As others have already commented, the proposed network of 52 Lawrence, 29 Dufferin, 63 Ossington and 90 Vaughan is an odd one that includes no common running between the routes and may not even be ideal from the viewpoint of network density, service levels and possible expansion. Lawrence, for example, shares running with 59 Maple Laef and 58 Malton, but these services would remain diesel.
Notable for their absence are 96 Wilson, 41 Keele and 7 Bathurst all of which would fit into the target service area. These routes also lie in an area that is or will be well populated with subway, streetcar and LRT routes with which they could share substations.
The current report cites a 1992 TTC study prepared when the clear intent was to justify trolley bus elimination. Some statements cited from the earlier report were questionable in 1992 and remain so today.
- “Overhead wire support systems involve some degree of objectionable visual intrusion”. The question here is the trade off between other effects including lower noise and better vehicle performance and the so-called intrusion of the wires. “Visual pollution” was a term coined by TB opponents in an attempt to make vice of the virtue of electric propulsion.
- “Routing for trolley buses is far less flexible when compared to non-electric buses”. Ah yes, all those bus routes that are changed so frequently that the map can hardly keep up. Any TB network would be designed around routes that are well-established and unlikely to need restructuring or rerouting just as, indeed, we design LRT and, dare I say it, subway lines. Off-wire capability for emergency situations was not considered seriously by the TTC.
The telling line in today’s report is this:
At that time, CNG buses received considerable attention …
That’s quite an understatement considering the cabal assembled to promote this now-abandoned technology.
The Hamilton Peer Review Group report is also cited without mentioning that the HSR system, like the TTC’s, was run down and that its management was part of the group pushing CNG buses.
GHG Emission Reductions
Here, the TTC report goes right off the rails, or more aptly, dewires. (I am reminded of the occasional Yonge 97 trolley bus operator who forgetfully thought he was driving North Yonge, and wound up in Hogg’s Hollow far from the nearest overhead at Glen Echo Loop.)
The TTC compares the cost of GHG reduction per tonne calculated at $1,840 with the Metrolinx Benefits Case Analysis value of $40. This comparison is faulty on three counts:
- The cost of the TB system is inflated as discussed above.
- Metrolinx is calculating the cost per tonne saved by diverting auto users onto transit. This conversion reaps a large benefit per user because of the relatively high GHG output of auto commuting. Here we are replacing a transit bus with an already-lower emission level per passenger km with a trolley bus.
- GHG savings are only one of several factors considered by Metrolinx studies, and they do not assume that the entire cost of a project should be assessed against emission reduction. Pollution reduction is only one of many benefits conferred by construction of a transit project and the project cost is properly charged against all of the benefits it brings.
The report concludes that there are less expensive ways to address GHG production by transit vehicles including:
- Improved hybrid bus performance: This is dubious considering that performance problems relative to expectations are already traced to the operating characteristics of routes, not to inherent problems with the technology.
- Improved cost effectiveness of fuel cell buses: I have already written at length about this boondoggle. Does anyone beyond a handful of desperate promoters expect this technology to scale up to bus fleet demands?
- Development of an electric bus. Such a vehicle requires power from somewhere. Either the vehicle carries an enormous battery pack to carry a long-term charge, or the vehicle is constantly going out of service to top up its batteries. Note that this means that charging would not all occur on an overnight basis as many electric bus advocates claim, but throughout the day. Moreover, charging stations would have to exist all over the network where buses could pull out of service to charge up.
This report, like so many I have seen before, tries to hard to make its case and in the end shows the preconceptions and anticipated outcome on which it was based.
I could have written a paper that said, yes, trolley buses might look good under certain assumtions, but it’s a leap. That would at least have been honest.
Reintroducing trolley buses in Toronto isn’t an easy decision, and there will be costs, certainly in the short term, well in excess of the benefits. We could say the same for Transit City’s LRT network or various subway expansion projects.
Some projects have sponsors, some don’t, and reports about them reflect the political markets for which they are written.