Updated 10:15 am, June 27: Metrolinx has decided to release GO’s electrification studies without a formal Freedom of Information request. They will be available online sometime in the next two weeks once they are converted to a suitable format.
Updated 12:30 am, June 23: A list of existing commuter rail operations including those with electrified operation has been added at the end in response to a bogus claim in a Metrolinx FAQ.
Updated 9:50 am, June 23: Another Caltrain newsletter shows the benefits of electrification and the benefits of EMU operation.
Much of the debate on the Weston corridor study, formally known as the Georgetown South Service Expansion (GSSE) Environmental Assessment, focuses on noise, vibration and pollution effects from the substantial addition to train traffic in the corridor. One major option, electrification, was not included in the EA on the premise that this conversion will, possibly, be done sometime in the future, but not now.
Responding to public pressure, Metrolinx will launch a detailed study of system-wide electrification for GO. The first step will be to strike a consultative committee of various interested members of the public to advise on the terms of reference for the study. The committee should be appointed by the Metrolinx Board in July 2009. Once the terms are nailed down, the study itself is expected to take until the end of 2010.
Alas, this is far enough in the future that it will have little outcome on what is actually built in the short term. Also, by looking at the full network, and having cited very high figures for a complete conversion, I can’t help wondering if Metrolinx hopes to derail support for a movement to electrify “now”.
Unfortunately for Metrolinx, GO has already studied electrification of the Lake Shore corridor first in 1992, then an update in 2001, and, I understand, another update in mid-2008. Do we really need another study for this most important of GO corridors? Can we estimate, broadly, the cost of converting the Georgetown corridor based on the Lake Shore study?
Metrolinx was asked to release the Lake Shore studies, but in a splendid example of contempt for the public, they require a Freedom of Information request to release documents we all know to exist in the first place. As of June 26, the requirement for an FOI request was dropped, and Metrolinx will post all of the GO studies online within the next two weeks once they have been formatted for that medium.
The FAQ for the electrification study (linked above) states:
Q. Hasn’t a study already been done?
A. Yes. A smaller study was done for the Lakeshore West line only.
Well, no, actually the April 2001 update covers Oshawa to Hamilton. Moreover, this may not be the entire system, but it is certainly the heart of GO’s network and information here gives order-of-magnitude values that can be used when looking at other lines.
Realistically, the only lines that will be considered for electrification in the short term are Lakeshore East/West and Georgetown. Nobody expects us to electrify the CPR to Bolton for a few trains a day, and the service proposed to Richmond Hill in the Metrolinx Regional Plan requires major changes in track capacity, never mind electrification, to implement.
Any study of the system should proceed along a fairly simple path:
- Validate the 2008 Lake Shore study and update it to adjust for current conditions.
- Separate costs associated with Union Station that would be shared with other corridors (this allows realistic cost per kilometre comparisons)
- Study the Georgetown corridor to a similar level of detail as already done for Lake Shore.
- Consider any network implications (operational strategies, equipment requirements) for a configuration in which part of the network is electrified and part is not.
- Review the tradeoffs between pure electric and dual mode locomotives, as well as the option of a fleet of self-propelled electric cars (EMUs)
Detailed review of other corridors is complicated both by the light level of service (and low return in a financial or environmental sense for electrification), and by the likelihood that some corridors would be significantly reconfigured to handle more intensive service with or without electrification.
In the interest of getting some numbers on the table, it is worth reviewing the 2001 update of the Lake Shore study of which I have a copy (sorry, not in electronic format). One of the most interesting statements appears quite early on:
An implementation scehdule of almost three years was determined for the project including a 14-month planning and approval period before any on-site work can commence. On this basis, revenue service for the electrified line could be targeted for the first quarter of 2004 [three years after the date of this document]. [Page ii]
The summary goes on to conclude:
In the economic analyses … the existing diesel operation demonstrates an economic advantage over the introduction of an electrified railway.
The vital word here is “existing diesel operations” as the study appears to have been based on the level of service existing in 2001 with some modest additions in future decades. Service levels now planned for the Lake Shore and other corridors substantially exceed those operated or contemplated in 2001. The modelled electrical requirements were based on actual 2001 schedules.
In the current context, one of the study objectives is worth noting:
Given the significant environmental benefits of an electrified railway, such as reduced levels of air and noise pollution and the potential for improved service levels due to the better performance of electric locomotives, the opportunity to implement electrification at reduced cost must be fully examined and pursued if considered feasible.
The “reduced cost” mentioned here refers to then-available second-hand electric locomotives whose purchase could lower the capital cost of conversion. Whether a similar situation with available equipment exists today is worth reviewing as part of the overall study.
The capital cost buildup for electrification has the following components over the period 2001-2025, the timeframe of the study:
- Electric locomotives: 20 in 2001-4, 3 in 2007, 3 in 2021, 20 replacements in 2025. $232-million (2001$).
- Locomotive maintenance facility: $15-million
- Subtotal: $247-million
- Diesel locomotives: 10 in 2006, 14 in 2007 and 4 in 2021. $145.6-million.
- 24 new bilevel cars (12 in 2007, 12 in 2021) that are needed only for diesel option because of better fleet utilization with electric operation. $62.4-million.
- Subtotal: $208-million
Overhead Catenary System (OCS) and Related Work
These costs are unique to the electric scenario:
- Overhead catenary: $56.9-million
- Traction power substations (3): $24.5-million
- Hydro connections: $35-million
- Signal modifications: $47.8-million
- Overhead clearance program: $39-million
- OCS maintenance facility: $3-million
- Property approvals/permits: $1-million
- Subtotal: $207.2-million
Note that this is for about 114 route-km or a cost of about $2-million/km. This seems rather low compared to other projects, and this may be due to assumptions about the availability of second hand equipment. Conversely, this total includes signal and trainshed modifications at Union which have, in practice, already been incorporated in the reconstruction projects now underway. Therefore these are not net costs chargeable to an electrification decision, but part of the ongoing provision that has been made for this option.
Comparison of the Options
The totals for the two options are:
- Electric: $454.2-million (NPV $326-million)
- Diesel: $208.0-million (NPV $149-million)
One important note here is that the Net Present Value calculations are not shown in detail. In particular, it is unclear whether the residual capital value of assets is considered. Those new electric locomotives purchase right at the end of the study period would have a very large residual value that should not be included in the NPV for the electric rolling stock.
It is likely that this is not accounted for because the ratio of total capital costs above (2.18:1) is also the same for the NPV values. This implies that there is no allowance for residual value. Indeed a worst-case scenario has been presented by placing a locomotive fleet replacement right at the end of the study period and failing to recognize that most of its value will be consumed in years beyond the end of the study. This is not just bad accounting, but a serious skewing of the financial comparison.
I will not go into the details of the operating costs, but they have been broken down by year showing the changes as service improvements are factored in and allowing infrequent costs (major overhauls) that appear only in certain years.
Power costs are consistently less than half of the cost of diesel fuel. Maintenance and overhaul costs are also lower.
This leads to NPV values on the operating side of the accounts of:
- Electric: $119-million
- Diesel: $199-million
Total Net Present Value
Combining the capital and operating figures gives a base case for the 25-year period of:
- Electric: $445-million
- Diesel: $348-million
Various alternative scenarios were considered as a sensitivity analysis.
- Cases 1 and 2: Energy costs go up by 10% and 20% for both modes. Since the electric costs are roughly half of diesel costs, both of these have a higher effect on the diesel option raising the latter by about $12-million relative to electric for case 2.
- Case 3: Escalation rate changed from 3% to 4%. Again, because the base case for diesel is already higher, this affects diesel disproportionately adding $11-million to the NPV relative to the base case.
- Case 4: The cost of signal and trainshed roof modifications at Union Station are assumed to be borne by the Union Station project. This is what actually happened, and so this cost has already been incurred for either option. This knocks $50-million off of the NPV for the capital cost of electrification.
None of these changes, by itself, overcomes the $103-million difference between electric and diesel.
For all cases, the electric option remains the more expensive, but to this I raise caveats:
- The treatment of residual value in the net present value calculation artificially inflates the NPV for the electric option.
- The relative and absolute prices for electric and diesel power need to be adjusted to current and likely future values and trends. In particular, we cannot assume that both types of energy will rise in cost in the same manner over time.
- The cost of infrastructure changes at Union is no longer in the equation, and this effectively makes Case 4 the new base.
- The degree to which assumptions about the availability of second hand equipment and parts might still be valid, and the degree to which this affects the overall comparison, needs to be verified and factored into the discussion.
- The claim that this would be a three-year project from planning to implementation must be verified. We need a realistic construction projection to work backward from any “hard” dates for actual operation. In particular, we need to know whether this plan would be subject to a full-blown two-year environmental assessment.
- If the estimates in past studies are seriously flawed, why? Did GO’s consultants not understand what they were doing? If they are substantially correct, why are current claims for electrification costs so high?
Even if I take the Lake Shore electrification projections on a raw basis from the 2001 study, the marginal capital cost (with all its faults) between electric and diesel options is $246.2-million (2001$), and this would generate an ongoing operational saving. How much depends on the degree to which actual service improvements exceed those used in the model because electric operations are inherently cheaper.
For the Weston/Georgetown corridor, implementation is not as straightforward as on the Lake Shore because of the branched service design. In the short term, it is likely that we would see only the Georgetown and Airport services electrified (assuming we could ever get SNC-Lavalin on board with that), with expanded service on the Milton line to follow. Service to Bolton is planned only for peak periods and could never offset the capital cost. Service to Barrie may be a candidate eventally, but it only runs for a short distance in the Weston corridor branching off at Dundas Street.
GO/Metrolinx has a lot of work to do in updating past studies, but they need to recognize that starting from scratch isn’t an option. Moreover, they must avoid burdening the overall review with a great deal of work on corridors that are low priority or hopeless candidates for electrification. Certainly we don’t need to know how much electrification to Bolton or Peterborough would cost as a pre-requisite to meaningful information about the Weston corridor.
Update: List of North American Commuter Rail Operations
In the FAQ concerning electrification, Metrolinx states:
Q. Is GO Transit the only commuter operator in Canada and North America still buying diesel locomotives? What about the rest of the world?
A. No it is not. 99 per cent of the North American commuter rail network operates diesel engines. Many transit authorities use similar locomotives and cars as GO Transit. Some examples: Florida’s Tri-Rail, Seattle’s Sounder, California’s Metrolink and Caltrain, Vancouver’s West Coast Express and Montreal’s AMT.
The following information was supplied to me by Greg Gormick, a reputable source of information about the railway industry.
The U.S. and Canadian commuter rail operators are:
- West Coast Express
- Connecticut DOT (Shore Line East and Metro-North New Haven Line)
- Virginia Rail Express
- Northern Indiana Commuter Transportation District (South Shore Line)
- North Star (the Minneapolis service that begins operation later this year or next)
- Trinity Rail Express (Dallas-Fort Worth)
- Front Runner (Salt Lake City)
- Rail Runner (Albuquerque)
- Coaster (San Diego)
- Metrolink (Los Angeles)
- Altamont Commuter Express (Stockton-San Jose)
- Sounder (Seattle).
Portland’s Tri-Met is now operating one DMU line and Austin is about to launch one later this year. I guess you could throw Ottawa’s O-Train in there, too, if we’re going to include DMU service, although it’s more LRT than commuter rail.
Of those operators, nine have significant electrified operations: AMT, Metro-North New Haven Line, Metro-North, LIRR, NJT, SEPTA, MARC, NICTD and Metra.
Metrolinx cleverly tries to give the impression that everyone is running diesel when in fact many operators run a mixture of both technologies, each taking its appropriate place. It is true, only in the most narrow sense, to say that “99% of the industry runs diesel”, but that is not the same as saying “99% of commuter rail operations are diesel powered”, the impression Metrolinx is clearly trying to achieve.
The GO Transit study is quite clear that there are operational and environmental benefits of electrification, but Metrolinx would have us believe that electric trains are a minor part of the rail industry. I can’t help thinking back to the 1970s when the Ontario Government bamboozled everyone into believing that LRT didn’t exist and we had to invent it in the form of ICTS. People travel. They know what transit and commuter systems look like in other cities.
Metrolinx needs to embrace honesty and openness if they expect public credibility.