The MBTA board has been presented with six options for more frequent service on our commuter rail lines, ranging up to a “full transformation” option that provides 15 minute service to most existing stations at a cost of almost $30 billion. In my view, we need some better options.
Without diving into each of the options that have been presented, one central observation needs to be made about all of them: Whatever higher frequency of service they propose, they propose the same frequency of service on each of the 13 existing commuter rail lines.
This across-the-board approach would make sense if all of the lines had the same potential to attract riders. However, the detailed data from the modeling show that many more new riders will be attracted by increased service on some lines than on others. Some lines are expected to attract as few as 10 new riders per added trip. Adding service that will be under-used makes no sense from any perspective, including an environmental perspective – running large rail coaches empty is very wasteful.
One might argue that the potential to attract riders has been underestimated. And the projections do fail to fully account for the possibility that over the long term “if you build it they will come.” Certainly, it would be desirable if new rail service could attract new affordable housing development around it. On the other hand, it is unwise to assume that the development will materialize in every instance.
The MBTA board should not endorse any of the proposed options, but should ask staff to develop a set of options that maximize the ridership returns on investment over the next ten years and support a pathway to more improvement over the long term. Unlike the current set of options, the options presented should target the lines that are most likely to be well used.
Transportation is a field that is changing very rapidly under the influence of technology and we should be in no rush to lock in our multi-decade plans. The better approach is to choose an aggressive, high-yield plan for improved service over the next 5 or 10 years and keep thinking about our long term options. Unfortunately, no such plan has been formally presented to the MBTA board.
The Boston area is blessed with a rich set of railroad links, mostly running radially in and out of the city. A wide variety of rail service configurations could be run on those links. The dimensions of variation are outlined in the boxed text below.
The service model: The Rail Vision team has defined and explored the feasibility of a number of different models for service. A service model is defined by which stations are served and at what frequency. One model, for example, would be “clock face” pulsed service to all stations — trains would move at fixed intervals along the line stopping at every station. Other models include “key station” service that provides frequent service to heavily used stations (bypassing others), expanded “regional” express service from the furthest out stations or subway like “urban” rail service for stations inside 128.
A service model can include more or less complex definitions of frequency of service, for example, half hourly during rush hour and hourly outside rush hour. Frequency of service could be different for different groups of stations.
Whether a given model is feasible on a particular commuter rail line depends most heavily on the number of tracks in service, which varies from 1 to 4 around the area. It is more or less impossible to mix express regional service with high frequency urban rail on a line with a single track and difficult on a line with only two tracks. In some areas, adding a track is a realistic prospect. In other areas, adding tracks would require politically infeasible property acquisitions and/or disruption of adjacent roadways. Adding track only in a short stretch of a line may be sufficient to allow additional service models if trains can be timed to pass in that stretch — whether or not the necessary timing is possible is a mathematical question that the team had the software to answer.
The vehicle power source: Currently, the MBTA relies on diesel powered locomotives. Going electric would reduce local air pollution, although modern diesel locomotives are much cleaner than our oldest locomotives. It would not change carbon emissions much today, since fossil generation plants supply most of the power our power grid. But, in the long term, if we also had a fully renewable power grid, going electric would reduce greenhouse emissions.
Electric engines are intrinsically more reliable than diesel engines. However, when the possibility of regional or local power outages is factored in, diesel locomotives may be more reliable.
Electric locomotives accelerate faster than diesel locomotives so they save time in service models involving more frequent stops. However, one finding that came out of the Rail Vision study was that given our system configuration, the acceleration time savings are too modest to make a dramatic difference in projected ridership (compare the ridership increases in alternatives 4 and 5 below).
The train model: Currently, the MBTA uses traditional “consists” — a locomotive pulling unpowered cars. We can gain service flexibility by using electric or diesel “multiple units” — self-propelled coaches. Multiple units can be easily assembled or split, allowing more flexibility in consist length. Electric multiple units can be combined with diesel generator cars to allow them to work on routes that do not have overhead power.
The North South Rail Link and South Station Expansion: These two major projects compete with each other as ways to manage more frequent service coming into Boston. Incoming trains have to dwell for a certain amount of time before they can be turned in the other direction at a terminus like North or South Station. The dwell times needed to turn the trains, together with the number of tracks and other logistical factors, limit the number of trains per hour that a terminus can accommodate. Connecting North and South station would allow many trains to run straight through, reducing dwell time and switching time and perhaps providing sufficient capacity to allow more trains without expanding South Station.
Detailed study of either major project was mostly outside the scope of the Rail Vision process, but in the maximal service models that were explored at least one of them was included because South Station capacity may become a constraint as service frequency increases. Some analysts have argued that South Station could handle more trains without expansion through faster turning of trains.
Stations: The modeling exercise did not contemplate the closure of low volume stations, but the “Key Stations” service model recognizes that some stations do not have the surrounding density of housing or jobs needed to justify increased service. A key priority for accelerating service is to allow level boarding at at all key stations — currently, at many stations passengers board slowly by climbing up from the ground.
Parking: Parking is a constraint on commuter rail use in some locations today. The Rail Vision process did not design specific new parking locations. However, when modeling future ridership increases as a result of service increases, the team assumed that parking problems can and will be solved — in other words parking was unconstrained.
Parking needs to be expanded at selected existing stops where it is already or might become a constraint. Additionally, in an “urban rail” service model, it would make sense to create new garages at the end of the urban segment of the routes — likely somewhere near Route 128.
Fares: Fare reductions were outside of the scope of the modeling generally, but the ridership increase from fare reductions was explored in one scenario.
The variables above create dozens of possibilities. The first phase of the Rail Vision process was to winnow out models that were not feasible or did not represent important variations on other models. The team settled on the following major models to explore in depth:
- Alternative 1: Optimize Current System
- Alternative 2: Regional Rail to Key Stations (Diesel)
- Alternative 3: Regional Rail to Key Stations (Electric)
- Alternative 4: Urban Rail (Diesel)
- Alternative 5: Urban Rail (Electric)
- Alternative 6: Full Transformation
Without diving into each model, one central observation needs to be made about all of the defined alternatives: They assume more or less the same level of service on each of the 13 existing commuter rail lines. For example, Alternative Six, the “Full Transformation”, delivers 15 minute service to all inner core stations and all key stations and all other stations where possible.
The need for more refined options
The detailed data from the modeling show that ridership responds much more on some lines than on others as additional train trips are added. For example, the board presentation on Models 1-3 shows that in Model 3 (page 20), adding 65 daily trips on the Franklin line would add roughly 500 riders, whereas adding only 16 daily trips to the Needham line would add roughly 2500 riders — in other words, in Model 3, the Franklin line would add under 10 riders per trip, but the Needham line would add 160 riders per trip.
Serving ten riders on a train (or adding a train trip that only nets a gain of 10 riders) makes no sense from any perspective, including an environmental perspective. Both the energy efficiency and the economic efficiency of any mode of travel depend on the average loading of the mode. Loading has to be averaged across all trips and may vary widely across trips — if a train heads into town full, but cycles back out, the return leg may be near empty. Given average loading factors, the current per-passenger energy efficiency of our current diesel-based commuter rail system is about twice the efficiency of a single occupancy motor vehicle (based on emissions per passenger-mile).
The rail vision team’s environmental analysis of the six proposed models shows that all of the diesel variations — models 1, 2, and 4 (see page 35 of presentation) — increase net carbon emissions. That means that they are less carbon efficient than the automobile trips that they are displacing. This is not just because these variations involve non-electric trains. Rather, it reflects that the diesel trains are inadequately loaded. If they were loaded even half as heavily as our current service, their impact should be roughly neutral. We can conclude that models 1, 2 and 4 are environmentally and economically inefficient. It also follows that parallel electric options (3 and 5 and likely 6) are also very lightly loaded on average and therefore economically inefficient, even if they may involve lower emissions.
The inefficiency of all of the presented options should be no surprise given that they fail to differentiate among the commuter lines each of which responds differently to service improvements and some of which respond very poorly.
One might argue that the ridership projections are low. And they do fail to fully account for the possibility that new service will induce long-term transit-oriented density increases that will increase ridership. Certainly, there is a strong case for supporting more distributed affordable housing development. On the other hand, it is unwise to assume that the necessary density increases will occur in every instance — many communities are geographically unsuited for dense development and/or adamantly opposed to density increases that might alter their community “character”.
Recommended Next Steps
The board should not endorse any of the proposed options, but should send staff back to develop a set of alternative migration paths that maximize the ridership returns on investment over the next ten years and sketch a pathway to maximal gains over the long term. Unlike the current set of options, the migration paths presented will necessarily identify particular lines for staged change.
Ridership increase is a good proxy for environmental, congestion and economic development benefits and should be used as a preliminary screen. A final decision should be based on more careful analysis of costs and benefits.
At least one migration path that ultimately achieves full electrification with its potential environmental benefits should be among the alternatives presented. The major project alternatives — North South Rail Link and South Station Expansion are both long term projects and do not need to be included in medium term migration paths, although they should remain on the table.
Many transportation advocates, including me, have been limited in our thinking by the compelling vision of “regional rail”. Our default right answer has from the start been to just do it all — to choose option 6, “full transformation”, regardless of its economic and environmental return on investment. However, there is nothing desirable about large vehicles running around near empty. Even if powered by renewable energy, large vehicles have large environmental impacts in their fabrication and operation. We should not think of Option 6 as a progressive vision.
Ultimately, our focus on “full transformation” is an historically limited vision. A 21-st century vision of how to use our extremely valuable right of ways might look completely different. Commuter rail service, meaning rush hour service from the suburbs to the core, is an early 20-th century vision. Regional rail, meaning all day frequent service which supports reverse commutes and a variety of commuting schedules and non-commuting trips, is actually a late 20-th century vision. Rail service, since it requires running large vehicles to handle even a small load, is often intrinsically inefficient.
A 21-st century vision might involve using the rail right-of-ways to support non-rail forms of mass transit that we have yet to fully envision. As John Moavenzadeh has pointed out, a paved railroad right of way might be a perfect conduit for somewhat smaller, autonomous shared vehicles that could pack together tightly to achieve the efficiency of rail, but depart from the right of way to complete local trips. As the Governor’s Transportation Commission on the Future of Transportation recommended: “As today’s roadways and travel corridors are maintained and modernized, MassDOT, municipalities, and other owners of roadway and travel infrastructure should update and redesign them to accommodate mobility of all kinds.”
Transportation is a field that is changing very rapidly under the influence of technology and we should be in no rush to lock in our multi-decade plans. The better approach is to choose an aggressive, high-yield migration path for our rail system for the next 5 or 10 years and keep thinking about our long term options. Unfortunately, no such migration path has been formally presented to the MBTA board.
For me, the most compelling aspect of the full transformation vision has been that we could connect the affordable housing distributed across our gateway cities to the vibrant job market of Boston. This is especially important as rising rents push more and more people out of the core MBTA service area. As we consider our migration path, this consideration should remain front and center. In particular, consideration of the fare structure should be part of our analysis of alternatives.