Metrolinx has published a set of documents containing the “Initial Business Case” for the GO Transit Regional Express Rail (GO/RER) network.
- Full Report
- Appendices A-J
- A: Corridor Specifications
- B: Corridor and System Schematics
- C: Model Assumptions and Results
- D: Record of Assumptions – Direct Demand Model
- E: Financial Performance of RER Systems
- F: Sensitivity Analysis
- G: Wider Economic Benefits
- H: Line Speed Analysis
- I: Environmental Assessment Program
- J: Fare Structure Issues and Solutions
- Appendix K: Station Access Analysis
[Note that except for the Summary, the documents are large PDFs.]
This article begins a review of these documents and of the various RER proposals examined in the Metrolinx studies.
Work on this review of GO/RER began in April 2014 following the announcement by Queen’s Park of its commitment to the RER concept. Unlike previous reports, this study looks in depth at all of the GO corridors, and reviews the technical issues associated with both increased service and electrification. This is not a final review, and much engineering work remains to be done, but there is a great deal more information now publicly available as the basis for discussions.
These documents were completed sometime in 2015 as is clear from references to future events that will occur later in the year, notably reports from the City of Toronto on SmartTrack. That scheme gets only passing mention, some of it the usual political cover story, because the specifics had yet to be decided. Exactly what the incremental effect of ST will be beyond the proposed GO/RER configuration is not yet known. Preliminary information in City reports implies that ST will amount to considerably less than was foreseen by the Tory election campaign, possibly as little as a few more stations and some sort of TTC/GO fare integration.
Five scenarios were reviewed to compare the effects, benefits, costs and technical issues associated with various possible future networks.
- The “Do Minimum” scenario provides only marginal peak period improvements to the existing system in response to projected demand growth, but with no electrification. This is effectively a “business as usual” model for the base case.
- The “Two-Way All-Day” scenario expands off peak service, but with diesel operation and no electrification. This is a minimal level of service expansion.
- The “10-Year Plan” would provide frequent service on the inner parts of some corridors, but with limited electrification.
- The “Full Build” extends beyond the 10-Year Plan to provide frequent service on the inner parts of all corridors, and with full electrification.
- The “10-Year Plan Optimized” extends the scope of electrification beyond that contemplated in scenario 3.
This progression implies a certain sequence of events during the study where a full build is impractical and the original 10-year plan was not aggressive enough with electrification, a key component of the announced government direction.
The estimated capital costs rise from $5 billion for scenario 1, through $10b, $12b and $19b for scenarios 2 to 4. The price tag for the latter is well above what Queen’s Park has available, and scenario 5 was developed with a projected cost of $13.5b. All but scenario 4 are said to be achievable by 2024. Given that it is now 2016, and this is a 10 year plan, that date probably requires some adjustment.
Scenario 5 is the 10-Year Plan Optimized, it represents significant progress towards implementing the service levels of Scenario 4. It goes beyond the investments and service included in Scenario 3 (10-Year Plan), with electrification also to Bramalea, Barrie, Stouffville and to Pearson Airport. This scenario and the resulting recommended RER program has been defined to maximize return on investment while mitigating risks. Depending on resolving various challenges, it can be delivered over 10 years for approximately $13.5 billion. It does not preclude, but rather prepares for, services to Milton and Kitchener to be eventually electrified and frequent all-day services introduced when agreement is reached on co-existence of GO and freight on these privately-owned corridors. [p. iv, Full Report]
Annual ridership is expected to go up by a factor of 2.5 over the coming 15 years, but operation costs will not rise at the same rate. The study postulates that an operating profit would be possible, eventually, but that will depend a lot on future fare policies, and on the evolution of trip patterns (length, direction, average fare). The ridership model foresees that “hundreds of thousands” of auto trips would be replaced by GO ridership each weekday comparing scenario 5 to scenario 1. The proportion of trips and its relationship to expected growth is not specified in the Executive Summary. (Possibly in the demand modelling later.)
The rate of demand increase on GO overall is projected at 2.3% which is lower than recent levels, but allows for some leveling off in a more mature service.
One big issue is the problem of getting riders physically to and from the GO trains. Either this will be done with substantially improved local transit services (an option that brings many issues associated with fare integration and cross-system subsidies), or with parking. The cost estimates include $750m for 15,000 new parking spaces, or $50k per space. At that scale, simply paving empty lots is not an option. The study notes the possibility that some of this cost “may not be necessary if service integration and fare integration with local transit services can be improved”. [p. v]
Those 15,000 spaces represent nowhere near the ratio of new parking spaces to existing facilities that the projected ridership growth would entail if everyone arrived by car. Parking charges are listed as a way of raising additional capital for the RER project, and of encouraging a shift to ride sharing and public transit feeder services.
It is amusing to read about the benefits of proven technology, something for which Ontario has not been noted in past endeavours.
Virtually all of the works are within existing rail corridors, so environmental and community impacts are limited mostly to noise and vibration. RER will use proven technology that is working around the world. [p. v]
Descriptions of RER cite similar operations in more than 50 city regions worldwide [page 6], and list a number of factors that simplify implementation [p. 4]. I cannot help thinking of how badly past studies have downplayed the benefits of LRT which bears a family resemblance, but at a local rather than a regional level.
The first electric railway opened in 1883 (the Volks Tourist Railway on the Brighton seafront in the U.K.). Ever since that time, electric traction has increasingly become the default source of power for the world’s more intensively used rail systems. [p. 14]
Finding this statement in a Metrolinx report is quite amusing considering some of the remarks made during community meetings on electrification before Metrolinx and GO “got religion” on the subject. The report skirts that debate by observing that GO is now at the threshold where electrification makes sense:
Until recently, diesel traction has been the appropriate mode of traction for the GO rail operation. However, the service enhancements envisaged in the near future will take GO rail beyond the threshold of service intensity appropriate for electrification. Continued use of diesel traction will become a source of financial and economic inefficiency. [p. 14]
Metrolinx intends to pursue discussions with the railways regarding the upgrades needed on their trackage, and also intends to review “modern, proven technology” with Transport Canada and the railways.
This is an “initial” analysis, and changes are likely depending on the evolution of expectations, changes in provincial funding, and who knows what political meddling that could arise.
A decade is a long time in politics, and the likelihood that the current governing parties or councillors will still be in place at that distant time is minuscule. Moreover, changes could come at any level part way through the project, and only a very strong, unshakeable commitment (i.e. very popular and difficult to derail) is likely to survive. This is not simply a case of showing up for a photo op or two with a gigantic prop cheque, but of supporting the plan for the long haul, including building a constituency that can survive beyond current governments. The arrival of a Ford-equivalent who simply wanted to start over with his own plan would be disastrous.
The Vision of Regional Express Rail
This section of the report begins by recounting the history of transit in the GTHA, of the formerly high transit mode share when more jobs were concentrated in central Toronto, of the rise and subsequent fall in subway construction, and of the growth of GO Transit for core-bound commuting trips.
At one point, there is even a claim that GO carries as many people into the core in the peak as the subway does, but this is hard to credit given that total one-way GO ridership is only about 90k. Possibly a restricted definition of the “core” has been used that omits a wide area served well by the subway, but less so by GO. In any event, growth of GO capacity will increase the commuting load it can bring to downtown Toronto.
Ridership into the core area has been growing by transit, at least during the peak period.
While “Active Transport” (walking, cycling) is a factor with all of the new housing downtown, it is small compared to overall figures during the peak. Their role is greater during the off-peak period, possibly reflecting a different pattern of trip origins more conducive to these modes. (Beware of differences in the vertical scales used in these charts.)
(The Transportation Tomorrow Survey updates these data every five years, but the 2016 numbers will not be available until 2017. The evolution of Active Transportation numbers will be interesting to see.)
Auto trips are growing outside the peak as anyone now dealing with off-peak congestion will know. GO, a peak-only service, has not seen much of the overall increase.
Between 2006 and 2011, average daily car trips to the downtown outside the morning peak period grew by 16,000, compared with less than 4,000 by GO Rail. Meanwhile, over 55,000 new trips were made by walking, cycling and TTC. [p. 10]
Trips from “downtown” are also growing strongly after a long period when the numbers were almost flat. Again, this will bear watching in the 2016 data when it is available. However, we must be careful about the definition of the “downtown” area which could very well exclude a good chunk of the central city such as the University of Toronto campus. It will be important to distinguish between “outward” trips that are simply among various locations in the old City of Toronto, and those to suburbs where GO/RER would make a difference.
Finally, there is the question of commuting that is not focussed on the City of Toronto. The total trip count is very large, and it is almost entirely served by private autos with little sign of growth in other modes. This is a huge challenge for both the regional rail network, and for transit systems within the communities outside of Toronto.
The challenge for transit in the overall travel market shows up in the proportion of trips by each mode taken within Toronto (the 416) and in the region beyond (the 905). GO rail services only have a tiny sliver of the travel market, lower even than walking. The problem here is that this is a chart of all trips from the most trivial journeys to the corner store to a long regional commute. A later chart looks only at trips over 10km, and this shows a different picture. All the same, transit has a lot of work to catch up with auto travel, and its credibility is hampered by decades in which transit simply was not an option for much travel throughout the city and region.
GO’s fleet plans include their current diesel-hauled trains with 10 or 12 bilevel cars, but also 4-car EMU sets running in 4, 8 or 12-car consists. The fleet required to operate the basic off-peak service would be EMUs, with electric locomotives hauling existing bilevels for peak service on routes that are electrified. [p. 20]
The comparative operating costs of different train consists are quite striking.
Although there will be additional costs once RER is operating, the report claims that additional revenue will more than offset this because fare revenue goes up more than operating costs, especially considering the savings available with EMU operations over the current diesel-hauled trains.
Many service options were tested, and the permutations are too extensive for an article here. I recommend that interested readers download the full reports for all of the details. As an overview, here is a table of the principal options studied (click to enlarge):
For the Lakeshore corridor, as an example, there is a much more detailed look at the options:
These figures show a very high return on investment assuming that one agrees with the methodology. Some points here are worth noting:
- Although the base 2014 ridership is not shown, it can be derived from the 2029 ridership and the percentage increases this represents. For Lakeshore West, the 2014 number is 17m, and for Lakeshore East it is about 14.5m. The scale of ridership increase has profound implications for the capacity that will be required to and from feeder stations.
- The revenues and operating costs are Net Present Value (NPV) figures over the life of the model, not one-year data. However, the ratios show that different service models produce different riding and revenue patterns. For example, for LSW scenario 5 has 33m riders/year by 2029, or 57% more than scenario 1. However, the revenues are only 31% higher suggesting that proportionately more riders in scenario 5 make shorter trips.
- The operating cost increment is small beyond the base case for LSW, and for LSE it is actually cheaper to operate the scenario 5 service than the base case.
- Capital investments are higher for scenarios with more service, but these are more than offset by the benefits imputed from shifting so many riders onto the RER network from autos.
Similar tables are included in the report for all of the other corridors and service options.
A common thread in all of the analyses is that the more aggressive service options tend to do best on the Business Case Analysis, probably because the additional ridership, revenue and benefits (trip diversion to transit) for well-used infrastructure more than offsets the capital investment required. Each of the lines reaches a point where better service produces proportionately more demand. If this is what might happen looking only to 2029, this has important implications for network growth in the decades beyond when RER really will operate as a “regional subway” rather than an occasional commuter train.
Growth to the full scenario 4 is constrained by the available provincial funding, and the tactic is to get as much value as possible for the $13.5b envelope Queen’s Park has announced.
The service plans for scenario 5 are summarized in the charts below.
The colours indicate, generally, the level of service with red being 15′ or better through blue and green to the less frequent services. The Milton and Richmond Hill corridors remain with only peak period, peak direction service because of constraints on operation over CP and CN trackage, as well as the need to floodproof tracks in the Don Valley for the Richmond Hill service.
As a point of comparison, here is the “full build” scenario 4 service.
Whether any of us will ever see service at this level (let along just getting to scenario 4) is quite another matter. This will require a sustained commitment and funding, and it will not be paid for with fairy dust schemes such as “tax increment financing”. The Metrolinx report stays away from the “how to pay for it” problem and concentrates on the “what can we build” issue. The obvious variables in the financial picture are:
- Fare structure: should a higher or lower overall farebox cost recovery be attempted?
- Subsidies: is the provision of good transit considered a “social good” and the cost seen as a preferable alternative to continued expansion of road capacity?
- Parking: should GO parking lots charge for this service, and how much can this revenue source contribute to overall financing? The report estimates that parking could contribute $100m annually from 2020 onward, enough to finance $3.8b in capital spending.
Ridership projections for the corridors vary considerably with the largest gains, no surprise, coming in corridors that receive more service.
Projections for Lakeshore West, Barrie and Stouffville corridors are noteworthy because off-peak values are similar to peak (albeit they are spread over more service hours). This would put GO/RER in the same situation as TTC where off-peak ridership is an important part of overall demand, and many trips have nothing to do with conventional commuting patterns. As with other parts of the evolving regional travel patterns, this change cannot be achieved without substantial investment in local feeder/distributor services.
The projected growth for each scenario is shown in more detail in the following chart.
“Business Case” analysis is something of a black art, and much depends on assumptions used in the model such as the value of travel time, presumed changes in road congestion, reduction of pollution, and secondary effects such as reduced health care costs. For the five scenarios, the analysis produces the following result:
The business-as-usual scenario 1 produces a surplus, but only by virtue of minimal investment. There are no “transport benefits” because this scenario is the base against which others are measured. Other scenarios require varying degrees of subsidy, but the elimination of the Milton and Richmond Hill corridors reduces this quite markedly. However, the large decline in transport benefits between scenarios 4 and 5 shows how much network benefit goes untapped in the recommended scenario.
The electrification cost is a relatively small part of scenarios 3, 4 and 5, much smaller than the fleet, infrastructure and property costs. Although scenarios 4 and 5 provide considerably more service, their operating and maintenance costs do not skyrocket because of the substantial savings with an electrified (mainly EMU) operation.
Another way to look at the costs and benefits is to compare the base case with the four scenarios for improvement.
Again, the much more substantial benefits available with the full build scenario 4 are evident, but the cost is not at a level Queen’s Park wishes to undertake. “Transport benefits” are broken down here by type, and this shows an important distinction in the distribution of notional savings. By far the largest benefit accrues to to transit users whose trips become faster, and to former auto users who save on the cost of driving their cars. Benefits to road users in reduced congestion are comparatively small as are the “safety” benefits that flow from supposedly lower traffic volumes. These can easily be understood in the context that roads are full beyond capacity already, and the improvement of transit merely “buys” room for other drivers (and future growth) by shifting many trips onto transit. The idea that roads will ever be uncongested is a pipedream. In that sense, although GO/RER will make life much better for trips that fit well on the rail corridors, its benefit for off-corridor journeys (orbital rather than radial) will be considerably lower.
Journey times are projected to fall with the new network through a combination of frequent service (and thus shorter waits for a train) and improved speed with electrification. This will make transit more attractive for various trips, although the degree will vary from location to location. The report includes a chart of comparative travel times for multi-corridor journeys such as Richmond Hill to Hamilton where transit times will be substantially reduced. However, it is unclear whether the times used include access times at both ends of the journey, an area where autos have a distinct advantage for many trips. Transit’s ability to serve this type of trip will be constrained by the dispersal of origins and destinations across the region even if frequent trunk services were to fill in an east-west grid.
The uphill battle faced by GO/RER is illustrated below in a chart of mode share today and in 2029.
This chart would have been more useful if it included other transit modes and autos, but even as it is, this shows the relative scale of GO/RER growth versus transportation demand overall. Although GO/RER will roughly triple in the medium-to-long trip market, a rate much higher than overall trips, it will still be roughly one sixth of the total travel. The breakdown needs to be much better understood by market segment, and the importance of short trips must not be forgotten. The average TTC journey is under 10km, and so many TTC trips do not contribute to the “other modes” values above. Moreover, many of the “last mile” trips required for riders to access the GO system are short journeys on local transit systems.
On the financial side, the report presents a rosy picture of future profitability for the network. This is possible due to various factors including:
- GO’s existing fare-by-distance structure will generate more revenue as service capacity increases and infrastructure is better-utilized.
- Operating costs for the electrified network do not rise proportionally to the change in service level.
- The cost of capital debt service is not included as a charge against operations.
- Last mile services are borne by local municipal operations and is not counted as part of the GO/RER cost base.
Whether this unexpected state, a transit service that actually shows a profit, is achievable will depend on many factors including political decisions about service, fares and subsidy levels, not to mention the basic task of simply building out the network to support the proposed level of service. As for local services, the report is somewhat evasive on additional local costs suggesting that all of these new GO/RER riders represent a market that might be tapped, and possibly even at no extra cost, by local systems. That really takes us into the realm of fantasy and shows the degree to which Metrolinx continues to avoid the delicate political question of subsidies to local systems. A further report on this issue is supposed to be in the works.
In future articles, I will turn to more details behind this study.