Updated Tuesday, November 16, 2010 at 5:10 pm: Metrolinx today announced that it will be ordering DMUs from Sumitomo, piggy backing on the Sonoma-Marin order. The statement, which is available in full on the Metrolinx site, includes:
Metrolinx will be entering into formal negotiations with Sumitomo Corporation of America to exercise an option from the Sonoma-Marin Area Rail Transit (California) procurement contract to purchase up to eighteen (18) highly efficient Diesel Multiple Units (DMU’s). These vehicles will meet stringent Tier 4 emissions standards and will be convertible to electric for the Air Rail Link.
Updated Sunday, November 14, 2010 at 3:30 pm: Information on the proposed Sonoma-Marin “SMART” Diesel Multiple Unit (DMU) acquisition has been linked from this article and the price per unit cited by me in the original text has been corrected. See the section on the ARL for updates.
The original article (as amended) from November 12 follows below.
The Electrification Study has been underway through 2010 and it has produced a number of background reports. I will leave the truly keen readers to plough through all of this, but a few high points deserve mention.
- Electric locomotives are the most cost-effective option for GO services
- The most value-for-money comes from electrifying entire corridors
That electric operations are better for GO is no surprise to anyone who has watched the growth of electric railways worldwide. Sadly, GO has decades of saying “no” to electrics on the grounds that investment in better service trumped investment in technology at the service levels then in effect. With the proposals found in The Big Move, this position is no longer valid.
The study workshops have seen vigourous debate on the issue of locomotives vs a fleet of electric multiple units (EMUs). It is cheaper to haul longer trains of coaches with one electric locomotive than to power each car in a train. However, this places a limitation on acceleration and speed between stations because the locomotive must do all of the work. (Only the locomotive’s wheels provide the power for acceleration, and there are limits to the forces that can be transmitted in this manner.)
The finding that full corridor electrification is most cost-effective comes from the high cost of dual-mode locomotives and the operational constraints that would probably exist if only some units had this capability. Only trains with “off-wire” capability could be dispatched to outer, peak-only parts of corridors. The study does not review a configuration with a mix of pure diesel-hauled trains with electric trains, although these would have effectively the same operational constraints.
Rolling Stock and Motive Power
Options for rolling stock and motive power are reviewed and short-listed in a comprehensive report by LTK Engineering Services, a respected consulting firm. To those who are familiar with railway practices and available technologies, there are few surprises here. What is important, however, is that the report takes as a basic premise the need for proven, commercially viable technology that is compatible with the infrastructure and service level likely to be provided on a “base case” which is, essentially, the existing and planned GO network.
This filters out a number of technologies whose future is much brighter in the starry eyes of their proponents than in the hard-nosed world of providing reliable, cost-effective transit service for decades to come. GO and other transit systems should not be showcases for expensive technology that may not even exist.
Four options remain on the table:
- Diesel locomotives hauling bi-level coaches
- Electric locomotives hauling bi-level coaches
- Electric multiple-unit bi-level coaches
- Dual-mode (diesel/electric) locomotives hauling bi-level coaches
Any of the electric modes requires infrastructure, and the cost of this should be recovered either in direct savings or in the ability to provide better, higher-capacity service than with diesels. Additional savings and effects of electrification are harder to express monetarily, but they include:
- Reduced noise and pollution along the corridors where electric trains operate in place of equivalent diesel service.
- Secondary benefits of lower pollution levels immediately adjacent to rail corridors, and the associated cost of health effects.
- Reduced or eliminated need for noise barrier construction and the visual effect this could have on affected neighbourhoods. This is an offset to the visual effect of overhead power supply supports.
One major problem with the many related studies is the inconsistency in assumed levels of service. The Big Move projected a very large increase in train operations on many corridors by 2030, and the demand projections show a 4-fold increase in GO passenger volumes at Union Station. However, the current station expansion project will only double capacity for GO. Part of the constraint lies in the physical limit on passenger flow to and from platforms, and part lies in the ability of the rail corridor to handle additional traffic.
During workshop sessions on the Electrification Study itself, GO has repeatedly maintained that service levels projected by The Big Move would never actually be operated, even though the higher level is assumed both for Union Station itself, and for regional projections of traffic diversions from roads to GO Transit. If the higher projections are accurate, then the Electrification Study is low-balling the benefits of converting from diesel. If the lower projections are accurate, then The Big Move’s projections of traffic diversions are wildly overstated, as are the projected demands at Union Station.
This is a fundamental problem with Metrolinx’ work.
A study of power supply options recommends implementation of a 2×25 kV 60 hz AC system commonly found on other electrified railways. This technology minimizes the number of points where substations must be located while staying within constraints of bridge clearances (higher voltages require clearances not available at some locations), and is compatible with Hydro One’s preferences for power supply.
The study notes that Hydro One proposes to connect to the network at two points, both of which are on the Lakeshore corridor. This implies that the choice of a first corridor may already have been taken.
One issue not mentioned in the power supply study is the question of the rail shed at Union Station. During the Electrification Study, and especially as this relates to the Air Rail Link, this has often been raised as a potential barrier due to clearance problems with the existing rail shed roof.
However, the City’s Union Station Revitalization Public Advisory Group (of which I was a member until it became clear that City staff had no further use for our input) was repeatedly told that the new train shed roof design allows sufficient clearance for electrification. Note that this is not just a matter of enough height for the wires, but also of ensuring that the distance between electrified components and the station structure are sufficient to prevent arcing.
This is another case where the claims of one GO study/project conflict with assumptions and statements made by the Electrification Study’s team.
The situation is made even more interesting by the ARL update report which claims that trains will stop outside of the train shed itself. Therefore, the question of conflict between an electrified ARL and the train shed is moot, even though this has come up often in discussions about this option.
When GO electrifies, they will do the most cost-effective parts of the network first. These will be the sections where trains run frequently, where implementation does not trigger expensive side-projects to eliminate obstacles.
The Network Option Report shows 18 possible configurations of an electrified GO system ranging from only Georgetown or Lakeshore up to the full network and many combinations in between. By the time we reach the update for the Metrolinx Board, this list is cut down to:
- Georgetown & ARL (the ARL was not included in the Network Option Report which had assumed it was a separate, non-Metrolinx operation)
- Lakeshore (to Hamilton James Street Station in the west)
- Georgetown, ARL & Lakeshore
- Georgetown, ARL, Lakeshore & Milton
- Georgetown, ARL, Lakeshore, Milton & Barrie
- Entire network (including Hamilton Hunter Street Station and St. Catharines extension)
What is not mentioned is a “prebuild” of the ARL segment. During workshop meetings where this question was raised, the issue of the train shed at Union was raised as a problem even though now the ARL update says that these trains will not operate into the roofed area of the station.
GO’s announcement that rail service to Kitchener will begin operation in late 2011 adds a small wrinkle to the Georgetown corridor in that the two trains each way daily to Kitchener will almost certainly remain diesel-hauled and they will be mixed in with the much larger number of trains on other routes sharing the corridor.
Implementation and Operations
Still to come are reports on the actual implementation of electrification including signalling changes (both for compatibility with electric trains and for closer headways), construction of the power distribution system, fleet planning and maintenance facilities.
Air Rail Link Update
The ARL update confirms that the stations planned for this service are Pearson, Weston, Bloor and Union. Provision will be made for a connection with the Eglinton LRT, although the current plans for the Weston Station on that line make such a connection inconvenient (not unlike the connection at Bloor to Dundas West Station). Indeed, if someone wants to reach the airport, they would be better to stay on the Eglinton line rather than transferring the ARL at Weston. However, it is unclear just when the Eglinton/Airport connection will actually be built.
The one-way trip time will be 25 minutes, service will operate 20 hours per day and the headway will be 15 minutes at all times. That’s four trips/hour each way times 20 hours, or 160 trips per day. Other Metrolinx reports have used 140 trips per day.
Projected ridership by year 5 is 5,000 passengers per day, and this implies an average of 31 riders per trip. Obviously there will be lightly used trips and peak loads will be well above 31/train. The light average loading begs the question of emissions per trip. The trains moving back and forth produce pollution with or without passengers, and it would be worthwhile knowing what ridership is needed to reach emissions lower than those caused simply by putting everyone in taxis.
This was originally a PPP scheme, but that idea fell apart in June. Although the presentation says that “financing arrangements became unacceptable” to the private partner, my understanding is that Queen’s Park refused to underwrite their risk. Indeed why, after investment of hundreds of millions in new infrastructure for Georgetown GO and ARL service should the government further guarantee the “private partner’s” business plan? When this scheme was concocted by the then Liberal government in Ottawa, it was spoken of as an all-private undertaking, but things didn’t quite work out that way.
The lounge for the ARL will be located west of the train shed at Union. The shed ends just west of York Street, but the Skywalk runs over to Simcoe, one block west, providing easy access to an ARL platform on the northernmost track. This arrangement, as I noted above, eliminates the requirement for an electrified ARL to enter the train shed and decouples any work on that structure from an ARL implementation.
Most interesting in the report is confirmation that the proposed vehicles are not reconditioned RDCs (the private proponent’s original choice), but are similar to vehicles to be built for the Sonoma-Marin “SMART” proposal in California by Sumitomo. These will be built as Tier 4 diesels with the capability of future electrification. How, exactly, this provision will be made remains to be seen. Indeed, at workshops, such cars have been discussed rather as equipment that would be eventually deployed on outer parts of the GO network such as Niagara Falls, and it would make much more sense to buy purpose-built electric equipment for the ARL.
Updated November 14:
Although the cost is not in the update report, I understand from other sources that the price of these cars will be $8.6-million each, vastly higher than prices for electric MU cars.
The price to be paid by SMART is US$56.85-million for nine 2-car trainsets. The ARL will require four 2-car sets for service plus one spare. On a proportionate basis, this would cost US$31.586-million. The cost per car would be US$3.159-million.
These trains are subject to Buy America provisions for the SMART order, and Sumitomo proposes to establish a plant in Illinois where final assembly will be performed using components manufactured in Japan. What arrangements, if any, would be made to meet Ontario’s Canadian content procurement rules is unknown.
The project itself is in political trouble as described by many articles in the Marin Independent Journal. (Use the search “marin independent journal smart” in Google.) Whether it will actually proceed, and therefore whether there will be an order onto which GO/Metrolinx can piggyback, remains to be seen.
End of Update
The ARL is still presented as a premium service; however, the fare has yet to be determined. When this was a PPP, fares above $20 were mooted.
This line will be a separate business unit within Metrolinx with its own President. This is complete nonsense. The ARL will have a fleet of, maybe, a dozen vehicles and eventually will be part of a much larger network. Maybe we should set up each of the GO lines as a separate company with its own President and associated bureaucracy.
The concept of a separate operation implies a desire to eventually sell it off once the capital costs of creating the service have been absorbed in Ontario’s budget. Can you say “Highway 407”?
Airport service should be provided by a variety of carriers and routes including an express rail service, the LRT lines proposed for Eglinton and Finch (and possibly a connection from Mississauga/Brampton as well), and regional bus services. Service to the airport, like any other major node, should be provided by the region’s public transit systems.
Particularly galling is the comment that this line will not serve commuters. Those who work at the airport and those who travel in the Weston corridor can look elsewhere for their travel needs. For $300-million (the cost of the spur, airport station and vehicles) plus the ARL’s share of infrastructure upgrades in the Weston Corridor (two of eight tracks), the cost of infrastructure to serve 5,000 tourists and business travellers per day is getting rather high.
Union Station Alternatives
In addition to the reports going to the public meeting of the Metrolinx board, there is a study on capacity issues at Union Station. This study has not yet reported and its reports are not publicly available, unlike the Electrification Study.
Five options were discussed at a recent workshop (which I did not attend):
- The existing and committed facilities at Union Station
- Satellite stations
- The Summerhill Corridor and North Toronto Station
- GO/Downtown Rapid Transit Integration
- New GO Tunnel through the Central Business District
An important part of this review is a proper understanding of passenger flows and demands in the GTA, as well as the capacities of various existing and planned components of the network. For example, offloading passengers to the TTC is counterproductive if their network is already full, of if the additional load would trigger the need for significant new infrastructure. Similarly, a “satellite” station is of little use unless it is close to someone’s destination either by walking, or by a very convenient transit link.
This study uses the service levels contemplated by The Big Move as its starting point for the GO network. These services imply a demand at Union much larger than that used for the Electrification Study and, of course, make any new infrastructure more financially attractive because it must serve a larger potential market.
Riding on GO rail has grown 11% from 2006 to 2009 with the largest increases in the Barrie and Stouffville corridors. I cannot help pointing out that the AM peak ridership on every corridor except Richmond Hill is already higher than 5,000, the projected all-day demand on the Air Rail Link. This gives some idea of just how small potatoes the ARL really is despite the money lavished on it.
About 10% of GO’s riding originates within the City of Toronto, and 60% of that is on the Lakeshore corridor.
Almost half of the trips destined for the core area (“Planning District 1”) use GO although the proportion is higher in Halton (75%) than in Durham (67%). Clearly driving to work in PD1 is not the choice for parts of GO’s service territory, but the market penetration varies, no doubt because of variations in service levels.
Total GO Rail growth from 2009 to 2031 is forecast to rise by about 120%.
However, there is a conundrum in the projections. The demand at Union Station is going to quadruple but the peak period demand is going up only by a factor of two. This implies a very tight, brief peak at Union made possible only if many trains arrive in a short period. Put another way, a demand/time curve will have a peak four times as high as current operations, but the area under the curve will only double. This suggests a much worse problem with peaking and unidirectional riding in complete contrast to the hope for more flexible commuting times and bi-directional traffic.
This conundrum needs to be examined in detail to understand whether the projections are valid and what contributes to them. Other studies have projected increases at various levels, and until everyone can agree which of these, or even what range of values, is the definitive one, we will continue to have studies that choose ridership projections as it suits their desired outcomes.
Those who want to build lots of infrastructure will use the highest numbers. Those who want to run the fewest trains (presumably with diesels as long as it is financially and operationally possible) will use the lowest numbers. The public is ill-served by cherry-picked demand estimates.
Far too much time has been wasted trying to avoid the need to electrify GO Transit. Without question, this will be a complex, lengthy and expensive process. Wild claims have been made that the option is unaffordable — in the billions of dollars — without addressing the basics.
Other commuter rail systems have electrified and, indeed, have seen this as the only responsible way to move forward. GO service levels proposed in The Big Move cannot be operated with diesel-hauled trains. If these service levels are fictional and planning continues on the assumption of much less service, then commuters are in for a big surprise. Claims made for traffic diversion off of highways, reduction of pollution, noise and associated health effects will simply not be attainable because far less service will actually operate.
I will leave a detailed discussion of Union Station and various alternatives to another post in the hope that more material on related studies becomes public soon. Yes, this is a broad hint to Metrolinx that conducting studies without keeping the public informed is a very bad idea.
We cannot make intelligent decisions about technologies, about staging of system expansion, about the priority and effects of various investments, without sound background material and studies that are consistent and defensible. A small army of consultants is working on a vital part of our region’s future plans. They should have coffee now and then, and get their stories straight.
The Metrolinx Board and the Ontario Government need to wake up and find a way to bring electrification to GO and to the ARL as soon as possible, not in some indefinite future many elections away.