Metrolinx Board Meeting: November 22, 2019

The Metrolinx Board met on November 22, 2019 with its usual mixed agenda of private session and public items. This article deals with the following public reports and mainly with the Kitchener and Niagara Falls business cases.

Ridership Report

The ridership report notes that travel is increasing across the GO network, and that short trips have gone up 12.4% over the previous year thanks to cheaper fares for short trips.

There is a chart on page 3 of the report detailing the year-over-year changes for all stations on the rail network. The version in the report is soft, but there is a separate version here that scales up nicely.

I have extracted the information from the Metrolinx chart into a more digestible format. First, there is the current daily ridership by corridor. There is a problem with these numbers because they do not sum to the reported total GO daily ridership. The reason for this is that the “daily” numbers are averaged over the full six months including weekends and holidays on the Lakeshore and Barrie corridors as well as at Union Station. This understates the actual weekday demand because 2/7 of the days included dilute the averages.

[For those skeptics who claim I simply make up complaints about Metrolinx, just do the math. Divide the daily ridership into the six-month figures for stations on the map. Union plus the LSW, LSE and Barrie corridors give values around 183 (the number of days from April 1 to September 30), while other corridors give above 131 (the number of weekdays). The actual weekday numbers at stations with 7-day service are higher than shown in the map above. Fundamental errors like this undermine one’s confidence in Metrolinx’ work.

[I have sent a request for clarification to Metrolinx to determine if they have separate weekend and weekday figures.]

Note that these count the originating trips which are primarily inbound. The outbound trips from Union are not included (a) as they would swamp the chart and (b) are not allocated to specific corridors in the Metrolinx presentation. Most of these trips are inbound to Union, and on an all-day basis it is reasonable to assume that total ridership per corridor is roughly double the numbers shown here. However, this would not tell the whole story because the values for LSW, LSE and Barrie are probably understated by about 25% due to the weekday/weekend problem noted above.

Broken down by station, the lightweights and the heavy hitters are immediately obvious. I have included count for the UPX in this chart to show its relative position for common stops on the Kitchener corridor. Of particular note, this is a busy service in terms of ridership per station compared to most of the GO network. Because it has two major built-in destinations – downtown and the airport – it is not dependent on feeder services or parking lots that fill up early in the day, and the “last mile problem” is not an issue at either end of the route.

Usage at some stations on corridors that have only a few trains/day is quite respectable by comparison with the frequent, all-day Lakeshore service. Other stations, notably on the politically motivated Niagara Falls extension, do not fare well. Also, the values for stations with 7-day service on LSW, LSE and Barrie would be higher if they were calculated against weekday usage rather than seven-day averages. UPX is the odd one out here because its demand pattern is different from the traditional commuter line.


The Metrolinx map includes percentage increases over the past year, but these are meaningless out of context of actual station usage. The next two charts show first the change in numbers of passengers, and then the percentage changes these represent. Although Bloor (Kitchener service) stands out as the highest percentage change, it is on a low base and the largest rise in passenger counts is at Exhibition. It is noteworthy that these are both stations well inside Toronto, not on the outskirts of the network. They are easily accessed on foot and by the local transit system.

Again, the change in daily ridership for stations with 7-day service cannot be directly compared to those with only weekday or peak service because the numbers are diluted with the inclusion of weekends and holidays. Peak vs offpeak breakdowns should be easy for GO to produce as all fares are collected by Presto.

GO Transit is strongly marketing its off peak services as an easy way to get downtown, and this shows up in off-peak ridership growth to attractions in the core and at Exhibition Place. There is a lot of surplus weekend and evening capacity on GO, and this makes the marginal cost of handling extra riders low.

GO makes use of its capacity to carry more trips (not to mention making possible converts for commute-trip services), but there is a limit on the type of attraction it can serve simply because the rail network does not go everywhere.

Comparison between the GO network and the TTC subway is important for a sense of scale. The total of trips originating at Union GO Rail station every weekday is about 76,000 (subject to the caveat about including weekends in the averaging process) and a reasonable assumption is that the majority of these are the outbound legs of commute trips. For the sake of argument, we could say that 140-150k trips are taken to and from the core each day by train (more if the weekend dilution effect were corrected). Although GO plans to substantially increase service, there is a limit to how many riders and trains will fit through even the “new” Union Station once the Bay concourse opens early next year. There are plans to reconfigure platforms and train paths to increase throughput at the station, but even this has its limitations. There is talk in the Ontario Line plans of offloading GO traffic onto that rapid transit line at East Harbour and Exhibition Stations to relieve pressure on Union.

All of this begs the question of the degree to which GO Transit can handle future demand especially if it takes on more of the inside-416 travel role. GO has blown hot and cold on this over the years. Originally GO feared to lose needed capacity to short haul riders and making extensions of dubious merit if there were not going to be enough room overall. More recently, GO is touted as a “relief” mechanism for the TTC rapid transit network, but that is not as easy to achieve as it seems because of service patterns, station locations and a tenuous link to local transit services. There is also the delicate matter of trade-offs between cost (fares) and speed to downtown, and the fact that not everyone wants to go to Union Station.

There is a large and growing demand for travel that is not oriented to Toronto’s core both within the city itself and beyond in the regions. This shows up in the Initial Business Cases for GO expansion below, but there is no discussion of the role of expanded local transit nor of the absence of good intercity bus service to deal with the last mile problem beyond the reach of existing rail corridors.

Presto Quarterly Report / Customer Service Report

The update on Presto’s status contains the usual statistics about the rise in usage as TTC riders migrate from “legacy” media to the standard fare cards. The report is silent on:

Ironically, more Presto issues are addressed in the Customer Service report.

Although the official line is that Presto customer satisfaction is roughly unchanged, there is a downward trend since 2015.

The GO/UPX ticketing problem was resolved by

“Removal of all GO tap devices on the platform to reduce customer confusion.” [p 7]

Other changes at Union is to place the ticket vending machines (TVMs) in the Skywalk directly on the path to the UPX station, the provision of additional machines, and updating TVMs to accept all forms of payment including coins and US currency.

Presto is working on new fare machines and will test them in 2020 in a limited market. Among other things, these will support “open payment” which, as defined by Metrolinx, means accepting credit and debit cards for single, undiscounted adult fares.

About two thirds of Presto users employ some form of non-web self-service to load their cards (vending machines, Android NFC, Autoload). The problem remains that web-based loads face a delay of up to 24 hours before the transaction propagates through the Presto network and is available to pay fares. [I have sent a request to Metrolinx asking for the breakdown of these three categories to find out how much activity still remains on autoload as opposed to other update channels.]

This fundamental constraint of the system design – having the fare “purse” on the card rather than in a back-end system – will remain Presto’s Achilles’ Heel until it is replaced with an account-based system. This will not happen until Metrolinx has a new contract, potentially with a new supplier, in place for Presto IT support.

Another pending change, now that Apple will enable Metrolinx to run its application on Apple devices and access the NFC communication function, is that more riders will be able to use their phones to reload their Presto “cards” in real time. The launch date for this has not been announced. [This information came from Anne Marie Aikins, Senior Manager, Media & Issues, in a discussion after the meeting.]

Presto still cannot run as an app in its own right replacing the physical Presto card with a rider’s own mobile device. [Updated November 26, 2019 at 12:15 am.]

Initial Business Cases

Before I review the Kitchener and Niagara expansion proposals, I must say that the idea of a “Business Case” is deeply misleading, to the point of being fraudulent, in the sense that no transit proposal we are ever going to see will break even on a balance sheet basis. There will always be money flowing out of public hands to provide private benefit – that is what a transit system does. We build and operate transit in order to move large numbers of people:

  • at a lower cost to society (avoiding the need to invest in roads, for example, or to suffer their direct effects on land use and the environment),
  • with the economic benefit (allowing far more workers or students to access a major node in the city such as a concentration of offices or a university campus),
  • avoiding the cost of a personal vehicle, and
  • reaping the social benefit of a relatively care-free journey (TTC riders might dispute this characterization).

Many of those benefits never appear as revenue directly on the government’s books, and certainly not in a form that could be directly used to pay down investment costs. It  is ironic that the flavour-of-the-day is Transit Oriented Development with an attempt to recoup the investment in new facilities, or at least to defray some of the cost, showing that investing real dollars is something the government would prefer to avoid. Meanwhile we evaluate possible transit “investment” on the basis of notional benefits that are largely beyond the government’s reach as revenue sources.

Furthermore, on the premise that an investment has a 60 year lifespan, future benefits out to that horizon are included in the calculation. Fares and benefits to be paid and reaped by children not yet born are thrown into the pot, but they do nothing for the relatively high short-to-medium term cost of paying for and operating new assets. Even with very generous assumptions about payback periods, some transit projects simply will not break even.

Recently, bowing to the political outlook of the current government, Metrolinx has taken a stance that facilities such as new stations will not be built unless there is associated development that could pay for them. This ignores the very real need people have to travel today whether or not a developer is sufficiently interested in a station site. Moreover, a lot of land near the rail corridors is neither close to stations, nor is it necessarily in an attractive location for development. Lands that are in attractive, built-up areas are not necessarily in public hands where profits can be funneled back to transit, and in parts of the GTA developers already complain about the tithe they pay in development charges.

The “economic case” for improved infrastructure and service rests on imputed savings to existing riders (shorter travel times, more frequent service), to new riders (reduced cost relative to auto travel) and to road users through a presumed reduction in congestion. Anyone who drives in the KW corridor today knows that it is packed full and getting worse. Road widening will “fix” problems for a time, but traffic expands to fill the space available. Moreover, the proportion of regional travel that Metrolinx expects to shift off the road network is a small portion of current, let alone future demand. Even in corridors where GO now provides a major alternative commuting route (the Lakeshore corridors east and west), the roads are not empty, and the real benefit is that far more people can travel to and from Toronto than would be possible if GO were not there. However, GO serves almost exclusively that market, and Highway 401 and the QEW, for example, are full of traffic that is not bound to or from Union Station.

Finally, “Business Cases” are presented on the basis of a single line or facility, or possibly alternatives within a scheme, but never as a wider review of alternative futures for the transit system as a network, not as free-standing projects.

Kitchener-Waterloo Initial Business Case

[Note that page numbers cited here are within the report, not the PDF. “Page 1” is actually page 8 of the PDF because cover and index pages preceded it.]

The IBC speaks of the academic and job/technical hubs along the KW-Toronto corridor, but does not address the problem of how riders are supposed to make trips that are not destined for downtown Toronto where they can walk or take good transit service to their destination. Since its inception, GO’s service model has been for Toronto commuters with most riders arriving at outlying stations and parking their own cars. Feeder transit services are comparatively unimportant, or in some cases, completely absent.

Kitchener-Waterloo and Toronto have been identified as major high tech hubs, with the number of tech jobs increasing by 66% and 32% respectively between 2011 and 20161. Collectively, the Toronto-Waterloo Innovation Corridor has been identified as an emerging technology cluster. In 2018, the federal government announced funding to establish an Advanced Manufacturing Supercluster centred on Toronto, Kitchener-Waterloo and Hamilton.

The expansion of rail service between Kitchener and Toronto has been a key aspiration for communities on the corridor. All day service was identified as a catalyst to support the development of the corridor as a tech employment hub. The transit service would provide new mobility options, strengthen the connections between Waterloo Region and the Greater Toronto and Hamilton Area, and support economic development along the corridor. Despite the benefits, efforts to increase service have been limited since the portion of the Kitchener corridor between Bramalea and Georgetown GO stations is a key freight rail corridor owned by CN, and the available capacity must be shared between passenger and freight rail services. While the challenges continue to be explored and worked through in partnership with CN, the extension of two-way all-day service to Kitchener was not included in the initial commitments of the GO Expansion program. [p 1]

Both Kitchener-Waterloo and Toronto have been identified as major high tech hubs, with the number of tech jobs increasing by 66% and 32% respectively between 2011 and 20162. Collectively, the Waterloo-Toronto Innovation Corridor has been identified as an emerging technology cluster. In 2018, the federal government announced funding to establish an Advanced Manufacturing Supercluster centred on Toronto, Kitchener-Waterloo and Hamilton. Part of this economic growth can be attributed to the concentration of post-secondary educational institutions along the corridor. Thirteen institutions have campuses located within 5 km of the Kitchener corridor … [p 9]

The problem is not simply to  get those pesky freights out of the way, but to provide transport between GO stations and homes or destinations. This is particularly important for all-day travel when parking lots may fill up early, and for trips that require local transit at the destination end (a job or school that is not sitting beside a GO station). Local transit is not well-supported by Ontario, and planned subsidy increases have been cancelled. Moreover, a GO station may or may not be ideally located as part of the local travel demand if GO sits in an old industrial area far from the main local concentration of jobs or academic activity. Which node should local transit focus on? Who will pay for additional service as demand to and from GO rises, and conversely can GO ridership really grow if there is no local transit for the “last mile” of trips?

This is part of a broader problem of the absence of good intercity transit in the GTHA and beyond.

While Kitchener and Guelph are both served by GO bus for off-peak and counter-peak trips, there are currently no direct bus connections between the two cities. GO bus services route through Highway 401, and require a transfer at the Aberfoyle Park and Ride lot. The average travel time by bus varies between 75 to 120 minutes, depending on the transfer time.

In addition to GO Transit services, VIA Rail and private intercity bus operators, such as Greyhound and Coach Canada, also provide transit options for Kitchener, Waterloo, Guelph and Toronto. [p 11]

One might argue that the service provided by other bus companies is on a profit-making basis, not for the economic benefit to the region nor for personal subsidy to travellers. As for VIA, its service is a sad remnant of what was once a busy rail corridor whose trains connected London to Toronto via many cities and towns along the way on convenient schedules. The dormant proposal for High Speed Rail showed an extreme example of the focus on a handful of stations between Toronto and London, not on the fine-grained distribution of travel along the route or how would-be riders would access the service.

The local service grids are shown in this map. They look good, at least where they exist, but the lines give no indication of service levels or how the local routes might serve GO as a regional spine network and connect outward to destinations.

Travel patterns between parts of the region are summarized in the following charts. The green lines are transit while the blue lines are auto trips. Transit has a presence, although not the lion’s share of the market, for trips to and from Toronto. However, there are many trips between regional centres where transit has a minuscule presence. This will not change if we only have a few more trains on the rail corridor without frequent local feeder/distributor service at both ends of riders’ journeys.

This is particularly true in the case of the proposed station at Breslau which is projected to have a large demand either from parking or from development, but which now sits in the empty space on the transit map.

The report unexpectedly contains an indictment of the Toronto-centric service plan Metrolinx has followed through its history which I reproduce in full here. These are Metrolinx’ words, not mine.

Travel between Brampton / Halton Hills and Toronto

Trips between Brampton / Halton Hills and Toronto are the largest travel market on the Kitchener corridor, with an estimated 188,000 daily trips. Nearly half of the trips are peak period, peak direction trips made by commuters traveling to Toronto in the morning and returning in the afternoon. Transit is very competitive for peak period peak direction trips (32%) and somewhat competitive during off-peak periods (19%). Transit is least competitive for counter-peak trips, with only a 9% mode share.

Travel between Kitchener / Waterloo and Guelph

Travel between Kitchener / Waterloo and Guelph is the second largest market on the corridor. An estimated 32,000 trips are made each day between these cities. The direction of travel is relatively even during both peak periods, with an approximately 55%-45% directional split favouring eastbound travel in the morning and westbound travel in the afternoon.

Nearly all trips between Kitchener-Waterloo and Guelph are made by car, with transit representing less than 1% of the total mode share, due in part to the low level of transit service provided. The service areas of the local transit agencies do not intersect, and transfers are not possible. GO rail service has limited frequencies and is not offered at times that are conducive to a typical working day in Guelph (last a.m. train arrives at 07:34 and the first p.m. train departs at 18:13). There is also no counter-peak service for commutes from Guelph to Kitchener. GO bus service requires a circuitous route via Highway 401 that is not time competitive compared to driving between Guelph and Kitchener on Highway 7.

Travel between the Outer Ring Municipalities and Toronto

The third largest market is the long distance inter-regional trips between the outer ring municipalities (Kitchener, Waterloo and Guelph) and Toronto, totalling approximately 18,000 daily trips. Approximately 60% of peak period travel is towards Toronto during the morning peak, and returning west during the afternoon peak.

Similar to the Suburban GTHA-Toronto market, transit is competitive for peak period, peak direction trips, representing 33% of the mode share; however, the mode share for counter-peak (10%) and off-peak (10%) travel is low.

Travel between the Outer Ring Municipalities and Brampton / Halton Hills

The smallest of the travel markets analyzed is travel between the outer ring municipalities and the municipalities of Brampton and Halton Hills. Approximately 11,000 daily trips are made between these municipalities. The travel patterns are oriented towards a westward commute, with a 55%-45% directional split during the peak periods.

Transit represents a 15% mode share for peak period peak direction trips, 6% for off-peak trips, and only 1% for counter-peak trips. [pp 15-16]

These are markets where GO plucked easy trips, the peak period, peak direction travel into Toronto that could be served by rail while ignoring just about everything else. What does Metrolinx propose as the solution? More rail service.

Through previous business case analyses, Metrolinx determined that two-way all-day service on the Kitchener corridor is the preferred solution to improving transportation options. This business case evaluates the optimal approach to deliver enhanced rail service. [p 17]

There will be a demand, but without looking beyond the rail corridor, the ability to compete with driving will be limited by those trips for which some kind of easy access is available at both ends of the trip. Moreover, if that local component presents an added fare on top of the rail journey, this will harm market share due to last mile issues.

Non-Toronto Trips

The economic and financial model utilized in this business case was designed to estimate ridership to and from Union Station. With the introduction of two-way all-day service on the Kitchener corridor, there is a potential to increase transit use between other urban centres on the corridor.

In particular, travel between Kitchener to Guelph generates approximately 23,000 daily trips; however, transit currently represents only 0.6% of the mode share (approximately 150 daily trips) due, in part, to the lack of viable transit alternatives. A review of the 2016 TTS data for trips across local transit agency service boundaries, excluding Toronto (e.g. between Milton and Mississauga, or York Region and Brampton), found that approximately 3.4% of these trips are made by transit. A comparable mode share applied to Kitchener-Guelph trips could generate an additional 650 daily trips between these stations. [p 39]

One does not use a demand model designed around Union Station when there is a clear recognition that the untapped markets lie well beyond Toronto’s core. One does not undertake a multi-million dollar expansion project to gain 650 trips per day when tens of thousands would still drive (with the possible exception of certain subway and GO stations that had political blessings at their births). A demand under 1,000 per day between two regions is territory for an infrequent and ineffective bus service. Metrolinx should be aiming much higher.

Far too much of the discussion about service deals with how to reach Guelph and Kitchener, and not enough with how to truly link cities and towns, economies and academic institutions, along the corridor.

Option for the Corridor

The evolution of options for the corridor is described here:

The implementation of two-way all-day service between Union Station and Kitchener GO station was previously studied through the 2015 IBC, the 2016 Economic and Financial Assessment and the 2016 Feasibility Study. These studies recommended the construction of a freight bypass corridor that would divert CN freight traffic off the Halton Subdivision between Bramalea and Georgetown GO stations, and allow for an increase in passenger service frequencies.

Through ongoing project development work, Metrolinx identified a potential option to address rail capacity constraints on the Halton Subdivision through incremental infrastructure improvements, without the need to build a dedicated freight bypass. Metrolinx initiated this business case update to compare this new option against the previously approved freight bypass corridor. [p 22]

The first option removes a bottleneck and conflicts with CN, but costs more and is challenging to build. The second retains a potential for bottleneck because freight service must cross over the GO corridor from the south (at Silver, the junction just west of Georgetown) to the north (at Halwest, the junction just east of Bramalea).

Here is the layout of the corridor and other rail lines in the area:

Here is the freight bypass option:

There are several considerations for both options that affect their “deliverability”. For the new freight corridor these include the combined property requirements for utilities, the railway and potential future expansion. In particular there are issues for the Hydro One lines in this corridor both for maintenance access and for electromagnetic interference in train signalling systems.

Along the existing corridor, there are some constraints where they pass through the older towns along the way with limited room for additional tracks and station expansion.

Any work to expand capacity in the KW corridor must be coordinated with other GO expansion plans both for the effects on common portions of the network and to avoid building in constraints that would limit furthur expansion of service. The desire to have all GO expansions in place by 2025 could preclude the 407 corridor because of its complexity. Advocates of that option would point of, of course, that the delay in starting the project and the now-compressed timetable for service improvement are the culprits.

The study is silent on the effect of an HSR project on this corridor, or indeed on any significant increase in passenger rail service.

For all of the details, please see the IBC starting at p 45.

Service patterns

The diagram below shows the “Business As Usual” service plan. The report acknowledges that this is already out of date due to recent service changes, and the diagram will be corrected in the next iteration, the “Preliminary Business Case”.

Note that the number of inbound trains/hour in the AM peak does not change between the two charts, and there nine in either case. The changes lie in the provision of two-way peak service and the substantial increase in off-peak two way service.

It is heartwarming to see that a commitment to electrification, at least of the inner part of the corridor, remains on Metrolinx’ books, but this is confined to the portion of the line that is not under CN’s control. Two thirds of the trains on the corridor from Bramalea to Union would be electric. Electric UPX trains would add to this east of the airport, but the whole future of UPX and its connection to the airport is, so to speak, “up in the air” as the severe constraints on its fleet and operations have finally been acknowledged by Metrolinx. It is conceivable that the UPX would be completely replaced by the new corridor service plan, but the IBC does not discuss this option which has major implications for the corridor’s configuration near Pearson Airport.

The next diagram shows the service level planned

Projected Ridership

Existing ridership in the corridor is detailed in the chart below. It is no surprise that the busy stations are those with the most service.

The report projects an addition of about 8,000 daily riders by 2031. These numbers are highly suspect because of limitations in the demand model which does not address regional travel. In the paragraph below, “FBC” refers to the model used in the Full Business Case for GO Expansion.

The FBC model is a direct-demand, elasticity-based model which relies on current ridership levels as a base for predicting future growth in demand in response to service changes. The current Kitchener service levels and ridership are very low in the off-peak, which limits the ability of the model to estimate the potential demand uplift from significant service increases. In the peak period, the FBC model is limited in its ability to estimate demand uplift due to significant improvements in travel time, and due to introducing significant new counter-peak service. A regional travel demand model would be able to more accurately estimate the potential ridership growth from introducing large travel time or service improvements to corridors with limited existing service and ridership. [pp 36-37]

On top of this, a new model is needed to address the potential effect of a transformative network change with a rail spine coupled to a robust local collector/distributor network.

With a model based on current demand, the growth occurs at locations where demand is already strong, but also where the primary access mode is by car, not by transit. Even “Transit Oriented Development” will not necessarily solve this problem because people living in or working at the new TOD nodes will not necessarily have origins and destinations on the transit network.

Particularly astounding is the projected loss of ridership east and south of Bramalea. This implies that Metrolinx, or at least its demand model, forecasts no change in that portion of the corridor over the coming decade even with better service than it has today. This is simply not credible.

Conversely, if the demand projections are off, what does this say for the capacity of service to handle future demand? In turn, this has implications for Union Station where, as the Ontario Line proposal revealed, there is hoped-for diversion of traffic from the Lakeshore corridors east and west onto the OL as a downtown distributor/collector.

Niagara Falls Corridor Initial Business Case

The Niagara Falls IBC follows the same general structure of the KW IBC, and shares many of its shortcomings.

In particular, although the concept of “regional” travel is important, there is no discussion of travel between cities along the corridor nor the recognition of the importance of Hamilton or St. Catharines as destinations in their own rights. Tourist travel is important to Niagara Falls, but there is no discussion of how someone arriving by rail would actually visit the many attractions of the region, not just in walking distance of the railway station, but further north in Niagara on the Lake and the many wineries outside of urban areas. It is not clear how much of the total travel market this scheme addresses because the report, unlike the KW IBC, does not quantify existing and potential travel to, from and within the corridor.

Any service east of Hamilton must deal with the fact that the railway is a major CN asset and passenger traffic will have to co-exist on a line Metrolinx is unlikely to ever own. This is very different from the KW situation where much of the corridor is already in public hands, and one of the options would bring it fully under Metrolinx control. There is also an operational issue at the Welland Canal (between St. Catharines and Niagara Falls) where shipping traffic will interfere with any frequent scheduled passenger service.

All that said, the options actually studied give the impression that Metrolinx really does not want to offer much service well east of Hamilton’s suburbs even though demand projections show that there is a market.

There is also a political dynamic at work here where “service” to Niagara Falls was implemented well before original projections, and the fact that it is one train a day each way carrying very few riders at an inconvenient time is swept under the carpet. There is a train, and so there is a “promise kept”.

Options for the Corridor

There are four options including the “Business As Usual” case which perpetuates the existing limited train and bus services expanding the latter only as necessary to meet demand.

  • Option 1 would add a second train each way to the peak service to Niagara Falls, and the off-peak service now operated in the summer would extend to the full year.
  • Option 2 would have the same Niagara Falls service as Option 1, but would add trains between Confederation and Union Stations at “up to” an hourly frequency.
  • Option 3 would provide half-hourly service to St. Catharines and hourly service beyond to Niagara Falls all day.

The service plan for the three options is diagrammed below. Note that this is different from the corresponding chart in the KW study which shows trains/hour. The Niagara chart shows trains per period of the day.

The projected demand varies as service and coverage increase. However, in spite of the strong showing of Option 3, it is discarded because of the Welland Canal issue. As the station-by-station demand projection shows, there would be strong demand at Grimsby if only the trains went there, and continuing to St. Catharines would not be a stretch. What appears to be at work is a desire to minimize the infrastructure requirements (and potential freight traffic conflicts) beyond Confederation Station.

Given potential problems with frequent service crossing the canal, it is surprising that the study did not at least evaluate an option to run regular service to St. Catharines with occasional trains and a bus service beyond to Niagara Falls. Option 3 has enough service crossing the canal (two trains/hour all day) to doom the proposal.

The evaluation summary shows how this falls out.

Economic Case Summary

Results indicate that Option 2 generates the greatest return on investment per dollar spent. For every dollar spent $1.2 of benefits are returned to society. This BCR of 1.2 is the highest of the three options analyzed.

Option 1 has a BCR close to Option 2 of 1.1. This option has the lowest costs of all three, but in turn generates the lowest net benefits to society at $66M. Lower benefit totals in Option 1 are a result of the reduced service pattern provided compared to Option 2 and 3.

Total benefits for Option 3 more than double totals for Option 1 and 2. However, the costs of Option 3 are much higher than Option 1 and 2. This is due to the operation and corresponding maintenance requirements of half hourly rail service to St. Catharines GO Station and hourly service to Niagara Falls GO Station. While having the highest net benefit total of all options at $172M, the substantial costs of this option bring into question the viability of the proposed service and give the option a BCR of 1.1, lower than Option 2. [p 47]

This is a bizarre way to analyze benefits even if we could believe in the methodology and the 60-year timeframe. All of the proposed options have a benefit cost ratio (BCR) above 1 indicating that they return more benefits than their cost. Option 3 is expensive, but does more for the transit network. It has a known problem at Welland Canal, but this is used to sabotage the viability of service to points west of the canal that could attract ridership and support regional growth plans.

Capital costs for Option 1 and 2 comprise of infrastructure at stations and track work in select areas of CN’s Grimsby Subdivision to meet the tested service patterns. Option 3 has the highest capital costs. Operating half hourly rail services on the corridor will require sections of net new tracks, extensive re-signalling, on-corridor work across CN’s Grimsby Subdivision to enable increased services, larger stations to handle increased demand and expansion to the existing Lewis Road Layover facility. [p 51]

Yes, of course, if we run more service we will need more infrastructure. This is not normally considered to be an impediment to GO expansion. In fact, the capital costs associated with the three options are $312 million for options 1 and 2, and $374 million for option 3 [see p 51]. This is not a huge premium compared to the additional ridership available with option 3. Some of this cost could be avoided with a scenario where frequent service did not operate through to Niagara Falls. Such an option was not part of the evaluation even with the higher potential demand.

The big difference comes in Operating and Maintenance costs which are much higher for option 3 because many more trains operate between St. Catharines and Toronto in that scenario. When this cost is considered on its own without offsetting fare revenue and economic benefits to the region, it is a bigger number, but option 3 accomplishes more.

The operating cost recovery for option 2 is highest, but is only 38%, while it is lowest, at 20%, for option 3. The BCRs for all options are above 1.0 because the wider benefits to riders and to the regional economy have an imputed value that outweighs these costs. However from the government’s point of view, those benefits cannot be counted on the balance sheet, and any of these options will represent a net new subsidy cost.

One must wonder whether another agenda is at work here to limit commitment to this project to a relatively minor extension of service east of Hamilton.

Projected Ridership

These tables do not indicate how many riders would shift their boarding points from existing stations in Hamilton to points east when the alternative is available.

The “business analysis” cites savings in auto operating costs for riders, but does not break these down by new and existing GO trips which would benefit differently. Moreover, the revenue estimates do not distinguish between full new fares to Union versus marginal fare increases for riders who now board in Hamilton shifting to stations further east. Based on the relative dollar amounts below, far more of the imputed benefit flows to transit users than to auto users implying that the contribution to reducing road congestion is relatively small. Note that these values are calculated over a sixty year life cycle reduced to Net Present Value (NPV) in 2018.

The proposed Grimsby Station does not exist, and in line with the new economic orthodoxy at Queen’s Park, Metrolinx would seek to built it in conjunction with development.

The Ministry of Transportation has asked Metrolinx to assess the status of all current transit projects and determine the feasibility of applying a market-driven approach that leverages third party investment in transit to help reduce the cost to provincial taxpayers. All scenario costs assume that the proposed GO station in Grimsby along Casablanca Boulevard will be paid for by third parties under this market-driven strategy approach.

If no alternative funding source can be secured to deliver this station, then alternatives in Grimsby should be considered. One such alternative solution calls for stopping at Grimsby`s existing VIA station. The proximity of this station to the town centre remedies the constricted parking supply at this site and makes the case for a station that would see similar ridership and benefits to the Casablanca Boulevard site. [pp 51-52]

The reference to parking shows that Metrolinx is thinking more of commuters to Toronto than of local, all day benefits, in its design. However, the actual construction proposed for a new Grimsby Station would provide only a 221 space lot [p 57] which is clearly incapable of supporting the level of demand that service beyond a few trains a day would bring.

The stations are seen as key to municipal plans. These plans cannot be achieved, however, if GO will only provide occasional service to stations. There is a disconnect between the local desires to support development along the corridor and the service Metrolinx is prepared to build and operate.

Municipal secondary plans and site plans for proposed station sites such as Grimsby will be compromised without GO Rail stations; the core of several of these plans. [p 8]

Creating new connections between areas that are proposed for new residential and commercial development, including station sites such as Grimsby GO Station, near which the municipality has plans for medium and high density mixed-use developments, as well as existing economic and tourist activity centres. [p 10]

New stations at Confederation in Hamilton and Grimsby (Casablanca Boulevard) would see a minimal build of access infrastructure to support this service while St. Catharines and Niagara Falls GO stations see minimal investment to existing facilities. [p 19]

All the options would serve these station locations, however, Option 3 performs particularly well in this regard as it would provide all-day service. [p 23]

Local Transit and Access to Stations

As with the KW IBC, the Niagara study does not contemplate how riders would get to and from stations, or the nature of trips the corridor might serve beyond commuting trips to Toronto and a relatively small tourist trade to Niagara Falls. The existing local services are shown in the maps below. Of particular note, Grimsby has no transit at all but gets a high projected demand from the model through a combination of parking and proposed development around the station.

The report also mentions the possibility of extending the Hamilton LRT east to Confederation Station, although its function would be quite different if the GO trains continued further east rather than terminating here.

The IBC projects travel time savings for riders, but this does not include the local access which is dependent on local transit unless one assumes everyone will be driving to stations. For a new service territory, the comparison should be with auto travel. Indeed, that is a fundamental part of the business analysis, the imputed value of travel time saved by switching riders to transit.

For those who now drive to a GO station from the Niagara Region, the projected total travel times do not change.

The travel times for passengers boarding at Niagara Falls, St. Catharines, and Grimsby GO stations under all options would be comparable to the travel time of individuals driving to a GO station on the Lakeshore West GO Line. An example of this trip pattern can be seen with commuters who currently drive from Niagara Region to Burlington GO Station and board the Lakeshore West GO train from this station to continue eastbound to Union Station during morning peak hours. Expanding GO Rail services year round to the Niagara Region would provide these commuters with comparable travel times and eliminate transfer and wait periods by providing a continuous trip option between Union Station and Hamilton and Niagara Region. [p 34]

One issue that is not addressed in detail is the effect on users of the existing bus service. Replacing bus trips with rail trips reduces overall operating costs, but also removes service from stops that do not match the new rail stations. Note that the savings quoted below are for the sixty year project lifecycle, not annually.

With the expansion of rail services between Hamilton and Niagara Regions, GO Bus route 12 will see reductions in service and operating pattern under all three options. Under Option 1 and 2 GO route 12 would reduce service by 11 trips per day as these timeslots in the peak and off-peak would be serviced by rail. This leads to an $84M reduction in operating and maintenance costs over the project lifecycle for Option 1 and 2.

Option 3 sees a far greater reduction in route 12 operations, approximately 40 daily trips, as rail services will operate all-day at hourly or shorter intervals. This reduces operating and maintenance costs to the route by approximately $204M. This potential reduction in service would be a disadvantage to users of GO Bus stops in Beamsville, St. Catharines, Fairview Mall and Niagara College which are on the bus route but not along the rail corridor. [pp 52-53]

Regional Travel Patterns

The Niagara Falls IBC is less detailed in its examination of travel along the corridor than the KW study, but does mention this in the context of calculating benefits. Depending on where a would-be rider will board and where they are going, the benefit of a shift to transit will vary.

A key metric which impacts project benefits are the proportion of Niagara Region riders who make trips to Toronto. It is unreasonable to assume that all Niagara Region boardings would travel to and from Union Station. With this in mind, 2016 Transportation Tomorrow Survey (TTS) data was pulled to determine which upper-tier municipalities Niagara Region transit users travel to. Only those upper-tier municipalities along the Lakeshore West Line were examined, excluding Niagara Region so to not capture intra-region travel.

Survey results show that for the AM peak period 79 per cent of Niagara Region local transit users are destined for Toronto, 10 per cent for Peel Region, one per cent for Halton Region and 10 per cent for the City of Hamilton. These percentages were applied to results to properly attribute benefits to those users travelling the full distance of the Lakeshore West Line and others only utilizing partial segments of it. [p 43]

This breakdown shows that travel will predominantly for commuters to downtown Toronto. What it does not address is whether the proportion to other locations along the route is constrained by the lack of “last mile” connections from the GO corridor to job and academic centres.

Tourist/Seasonal Demand

The projected tourism market, one of the targets foreseen by politicians in Niagara, is quite small even with all day service beyond St. Catharines in Option 3. Assuming that the model is accurate, this implies that the transit service, considered as a network, does not provide what tourists want.

The existing seasonal tourism demand on GO is shown below, followed by the projected 2031 numbers for the three options. One is presented as an annual total while the other as a daily value. Given seasonal fluctuations, a daily average does not show how well the service would be used in high season, or how poorly in the off months. If we assume that the 360 tourists for Niagara Falls are an all year average, and using a 250-day year, this translates to 90,000 riders, roughly double the 2018 demand. It is intriguing that St. Catharines would have close to the same tourist demand as Niagara Falls, and one wonders just what is going on in that model.

A further assumption is that a high proportion of demand in Niagara Falls will arrive by walking and cycling, not by car. Notable by its absence is any mention of arrivals by local transit.

A map of tourism destinations in Niagara Falls further shows how the IBC does not look at the wider market for travel to the Niagara region, only to Niagara Falls itself.

The Need For an Inter City Bus Network

Metrolinx’ clear goal for expansion of its train service is to replace buses.

The service expansion will replace the existing GO bus service with rail service that will operate within a separate right-of-way from road vehicles. [KW IBC, p 32]

This may be true for travel along the network, but it does nothing for travel to and from the network. Metrolinx is astoundingly short-sighted in ignoring this part of the overall travel problem. After decades of running a network where riders provided their own transport to outlying stations, and either walked or used the TTC (primarily the subway) to reach destinations within Toronto, Metrolinx does not conceive of serving trips that cannot fit this pattern.

At best, there is talk of last mile travel with ride sharing or even autonomous vehicles, but without any discussion of the size of fleet or the terminal and road capacity needed to make this the way GO stations are accessed. It is as if the TTC subway would exist without its supporting bus and streetcar network.

This is a bankrupt approach to a truly regional network design. The Ontario government may not want to get into the business of supporting substantially better local bus service, but simply running more trains will not divert much of the growing regional demand away from cars, local roads and highways.

The objects of the Metrolinx corporation set by statute include:

… to provide leadership in the co-ordination, planning, financing, development and implementation of an integrated transit network in the regional transportation area … [Metrolinx Act S.O. 2006, c. 16 s. 5 (1)(a).]

The Act does not say anything about running a rail only network, nor of focusing this network on Union Station.

For years Metrolinx has abdicated its responsibility to plan and build a truly regional system. In the early days, this might have been excused given the backlog of capacity required for commuting, but this problem extends well beyond trips to and from the old City of Toronto.

The crew now in charge at Queen’s Park believes that the private sector will provide anything we require. The history of bus service in Ontario, indeed in Canada, is that the private sector is walking away from intercity services because they cannot make a profit at that business.

This does not eliminate the need to travel, and it is long overdue for the government to determine how that travel demand will be met with a public network.

Metrolinx could lead on this file, but both its management and board are silent.

12 thoughts on “Metrolinx Board Meeting: November 22, 2019

  1. This is exceptional analysis with insight further to the following sub-topic alone. It immediately flagged interest, and Steve follows on it:

    [I have included count for the UPX in this chart to show its relative position for common stops on the Kitchener corridor. Of particular note, this is a busy service in terms of ridership per station compared to most of the GO network. Because it has two major built-in destinations – downtown and the airport – it is not dependent on feeder services or parking lots that fill up early in the day, and the “last mile problem” is not an issue at either end of the route.
    UPX is the odd one out here because its demand pattern is different from the traditional commuter line.
    Two thirds of the trains on the corridor from Bramalea to Union would be electric. Electric UPX trains would add to this east of the airport, but the whole future of UPX and its connection to the airport is, so to speak, “up in the air” as the severe constraints on its fleet and operations have finally been acknowledged by Metrolinx. It is conceivable that the UPX would be completely replaced by the new corridor service plan, but the IBC does not discuss this option which has major implications for the corridor’s configuration near Pearson Airport.

    I have only read the first few paragraphs intently, but scanned through the rest, my point in posting without fully digesting it all is the importance of the UPX story, and it being a harbinger.

    I look forward to Steve writing more on this. It’s interesting that Metrolinx have divulged the retirement of the Nippon Sharyo DMUs, a belated admission that it never was a good choice.

    I’m sure others will add more comment on this. The ridership figures are off the chart for Union and UPX.


  2. If GO Transit can operate 3-4 trains each way between Union Station and Niagara Falls on weekends, why is it only operating 1 train each way on weekdays? It seems counter-intuitive to have more service on weekends.

    Steve: This is probably a combination of equipment availability and track time on weekends compared to weekdays.


  3. “Another pending change, now that Apple will enable Metrolinx to run its application on Apple devices, is that more riders will be able to use their phones as their Presto “cards” and they will be able to reload in real time. The launch date for this has not been announced.”

    Currently AFAIK, no one may use their phone as their Presto “card”. The Android app only permits one to instantly load their Presto card.

    Steve: Yes. The Andoid device can only be used to instantly load value onto your Presto card, but the card itself remains the “device” that one taps to pay fares. I will correct the text.


  4. And 3 stations in Toronto (Bloor, Exhibition, Danforth have all increased in trips but they don’t have any parking lots. Ex is probably a destination stop but Bloor & Danforth have both way demands all day that I have seen.


  5. “Tourist travel is important to Niagara Falls, but there is no discussion of how someone arriving by rail would actually visit the many attractions of the region, not just in walking distance of the railway station, but further north in Niagara on the Lake”

    The Niagara Falls WeGo bus system does a good job at providing transit service to area attractions, including the train/bus station and they have a shuttle to Niagara-on-the-Lake. They have been slowly improving service over the years.

    Steve: That may be, but the projected GO ridership to Niagara Falls does not constitute a tourist bonanza.


  6. Any update on the much needed Scarborough subway and Ontario Line? These two projects need to be expedited as the SRT and the Yonge Line are dangerously overcrowded including the Yonge-Bloor station. SRT also frequently breaks down.

    Steve: Nothing new. Metrolinx has taken over these projects from the TTC and consolidated project staffing into their own operation. Public consultation on the Ontario Line proposal is supposed to start later this fall, but there are no details on this yet, and time is running out to begin the process this year. The Scarborough subway is over a decade away, and the TTC is studying how they can keep the SRT alive in the interim.


  7. Fantastic work Steve and so interesting. Feel like I should be paying for this analysis…

    Took UPX yesterday evening and it was practically empty. What are they replacing those cars with? Is the electrification idea still in the works or is Queen’s Park walking away from this idea? When will UPX be integrated in to Metrolinx and available for ordinary folks?

    Steve: By “integrated” I assume you mean the fare structure. That’s hard to say, but it will be an inevitable side effect of the project to, eventually, make the airport a stop on the KW corridor, although I’m sure Metrolinx will try to figure out a way to soak air travellers for more than the standard fare. They are big on “integration” except when it will cost them money.

    The electrification depends on the bidders to build and operate the GO Expansion Network. If someone wants to run it on hot air, and puts in a competitive bid, who knows what we will get. Metrolinx has offloaded all of the significant technical decisions to the future private sector partner.

    Oh, and thank you!


  8. Of course it’ll increase the capital cost substantially, but they could get around the issue of train/ship conflicts at the Welland Canal by building a taller bridge, or tunneling under the canal. I seem to recall there is a rail tunnel under the canal somewhere. Possibly the ex-TH&B line?

    Steve: It’s the CPR line (ex TH&B) and it is much further south (see Google Maps). This rail line goes to Fort Erie, not to Niagara Falls.

    A bridge would be very expensive because of the height needed to clear shipping and the maximum gradient of 1% (preferably 0.5%) for freight traffic which dictates the length of the approach structures.


  9. In 2013 I was travelling down the Welland canal in my boat when I heard the Seaway authority tell a freighter to pull over and tie up because there was a GO train coming. The ship captain couldn’t believe it, neither could I. I doubt it has happened much since then. I later learned there was also a problem with the bridge and they decided to wait for the GO train before fixing it.


  10. The GO ridership numbers are interesting. I don’t think there’s much that is surprising in them; many lines show a “higher ridership” inner region followed by a “lower ridership” outer region, but the breakpoint is also where the service levels drop, so it indicates nothing more than that.

    I will say that the Oshawa passenger numbers indicate to me that the line should probably be extended past Oshawa. I believe there was a plan to go to downtown Bowmanville parallel to the CP corridor, but the Doug Ford maladministration has apparently sabotaged all work. Wait for 2023…sigh…


  11. I’m surprised the GO Niagara Falls analysis doesn’t contain any analysis of GO going a couple of blocks further down the rail and hooking up with the American Niagara Falls train station. The American senators and congresspeople regularly ask Canada to improve this connection, so there might even be some leeway to have that be considered still “within” Canada. The Niagara Falls US train service is very unreliable, but it has a decent number of trains per day heading to NY. I don’t know what the ridership would be like, but it could open up lots of weird possibilities: cross-border commuters, people from Hamilton using Buffalo as their airport for southern trips, tourists from upstate New York coming to Toronto, tying the upstate NY economy into being more dependent on Toronto.


Comments are closed.