Assessment of Blue Line extension into Somerville [emphasis added]
Overall, this project is rated medium priority. It would provide rail transit service to densely developed sections of Somerville and Medford that are currently served by bus routes connecting with the Green, Red, or Orange lines.
It would serve more total riders and new transit riders than a Green Line extension to West Medford, but would also have a much higher capital cost per new transit rider. In absolute terms, it would be one of the most costly rapid transit projects examined.
Travel times between extension stations and destinations in downtown Boston would be a few minutes faster via the Blue Line than via the Green Line. Air quality improvements would be about twice as great for a Blue Line extension as for a Green Line extension.
It is rated high in economic and land use impacts. The majority of the stations would be located in state-designated revitalization areas where transit-oriented development is planned. This would include a mixture of high-density residential, commercial, and industrial development.
The above cost assessment was to use the BLUE Line as the means for the 'heavy rail' service to Somerville, from which it summarized that it would be a costly project. However, a Blue Line extension would have necessitated (as devised) extensive infrastructure to bring it from its end at Bowdoin Square up to the existing rail corridor in Somerville. It is that portion which made it costly.
The associated costs were summed up with this:
This would be a 5.3-mile extension, including a new subway between Bowdoin and Lechmere, six new stations in Somerville and Medford, a relocated Lechmere Station, and a new underground Science Park Station. (Bowdoin Station itself is scheduled to be closed in conjunction with implementation of six-car train service on the Blue Line.)
A variation, adding about one half mile, would run closer to Union Square in Somerville, via a new subway under Prospect Hill. This variation is not reflected in the capital cost estimate.
Capital Cost $696.5 million (Based on 1994 PMT, adjusted to 2003)
Operating Cost $76,800 per weekday
Daily Ridership Increase on Mode 13,500
Net Increase in Daily Transit Ridership 5,800
Capital Cost per New Transit Rider $119,500
Operating Cost per Wkday/New Transit Rider $13.20
Capital Cost/Travel Time Benefit $343,300 per hour
Operating Cost/Travel Time Benefit $37.90 per hour
Travel Time Savings 2,029 hours per weekday
Cost estimates made for the BLX proposal and other transit related proposals can be used to estimate a cost for the GLR's proposal to branch the Orange Line.
Capital Costs of an Orange Line Branch
The Blue Line extension into Somerville, whose summary is above, serves as a sound basis for providing much of what is needed to estimate an Orange Line extension into the corridor. The BLX was planned to end with a station in West Medford, as is the GLR's OLB, and would require the same basic infrastructure as an Orange Line branch. Not needed in an Orange Line branch is the extensive effort on the south end to connect into the heavy rail trunk/mainline.
The BLX costs include the expense to tunnel from Bowdoin Square to Lechmere Station, with new underground stations for Science Park and Lechmere. The costs of the BLX were not broken down in the summary, but we can use the cost estimate of another proposed project to aide in our adjustments.
Included in the report was a proposal to tunnel from Bowdoin to the Red Line at Charles/MGH Station, with a Blue Line station below ground at Charles/MGH. We'll use the cost for this project, of $174.6 Million, to stand in for the expense of taking the Blue Line from Bowdoin to an underground Science Park Station, although the latter would likely be considerably more due to the routing.
Therefore, we can deduct from the total BLX costs $174.6 Million.
Now, we're at the river. There isn't any other proposal in the report which included a rail transit tunnel beneath a river.
Unlike the Fort Point Channel and the Red Line tunnel, this one would most likely be placed below the bottom of the waterway. The Orange Line tunnel near this same point would be a comparable expense, but without that data we'll have to ball park it.
The more extensive planning and material needs of this portion would likely make it much higher on a cost/foot basis than the last segment from Bowdoin. They both include an underground station. Using a doubling of the first section costs is not unreasonable for this portion, especially given that the first section cost is under reported. Thus, we can estimate that taking the Blue Line under the river and then to a new underground station at Lechmere is $349.2 Million; further downward adjusting the BLX cost:
The remaining route expense is for taking the Blue Line northward from Lechmere, crossing over the Fitchburg Commuter Line on its way up to the Lowell Line rail corridor. At this point the costs should be comparable to an Orange Line branch. The Orange Line would leave Community College Station, have one track pass under the other branch's tracks and then both tracks immediately pass under the Haverhill Commuter tracks on its way to the same corridor.
These two segments would likely have similar costs, though it could be said that the latter has to make one additional underpass, however, it could be added that the former has to bring the BLX line to the surface from a below ground Lechmere Station. Thus, there are differences which likely balance each other out.
Removing those costs that do not pertain to the OLB brings the total project down to $172.7 Million. That total includes extending service to West Medford, north of the Mystic River, as is proposed by the GLR/OLB.
Before making other dollar adjustments, we need to bring the cost estimate figure forward to match the public works 'creep' we've seen occur in the GLX cost estimates. Close in time to the BLX estimates were the early GLX cost projections. The $375 Million GLX estimate of that day has grown to $954 Million, giving us a 2.544 multiplier. When applied to our working number of $172.7 Million gives $439.35 Million.
Adjusting for Additional Rail Cars
To implement the GLR/OLB plan without detrimentally impacting the Oak Grove (now) Branch, additional trains, more than that planned for the BLX, will need to be purchased. There are several scenarios that could be priced out here. The balancing of frequency, capacity, safety and costs is a subjective process.
If we seek to avoid any reduction to the Oak Grove Branch it will require the MBTA to go with 3 minute headways (during peak use) on the trunk line. A 3 minute headway should be within the capability of the OL (the MBTA will likely disagree); indeed, other systems with similar signaling and even higher speeds make use of 2 minute headways during peak use hours.
With 90 second headways in use on the trunk of the Green Line, it is disingenuous to suggest 180 second headways with signaling equipment is not doable on the OL.
A 3 minute headway would allow for a 27 train scenario, with 18 allocated to Oak Grove and 9 to West Medford. Currently there are 17 OL trains, so it will be a one train increase for Oak Grove. The West Medford allocation will produce a train frequency of 9.5 minutes, while frequencies for the Oak Grove Branch will alternate between 6.25 and 3.125 minutes.
While another scenario with fewer trains or at least fewer coaches/cars could be implemented, for pricing purposes the 27 train (6 cars each) option will be used. That nets a need for 10 new OL trains (60 cars) to be purchased, as there now exists 17 trains in service. These expenses have been in part factored into the BLX number which has been used as our base number, and in part are ameliorated by a separate project already planned and budgeted.
The BLX would have necessitated at least the purchase of 4 trains (24 cars) to implement service, thus we can consider their costs built into our current estimate. Additionally, the MBTA has already budgeted for a 3 train addition for the Orange Line in its fleet replacement project. That leaves 3 trains (18 cars) that have to be factored into our cost estimate. The cars are straight forward, being approximately $2 Million each for Blue or Orange Line cars, thus we'll add in $36 Million (18 x $2), to bring our cost estimate to $475.35 M (439.35 + 36).
Two additional deductions can be made, an adjustment for the cost of a service yard facility, and one for the cost of adding tracks to the Somerville rail corridor.
Both the GLX and the BLX factored in costs for service facilities, especially since neither had any repair or storage capability in Somerville or Medford. However, for the GLR/OLB proposal an OL service facility is conveniently located on the northern portion of its line, and the MBTA has already budgeted for an upgrade of that yard facility in conjunction with its fleet replacement.
Adjusting our cost estimate is not aided by a specific number in the BLX figures, thus we'll cautiously use the GLX estimate for its yard facility, and reduce it by half: $70M/2 = $35M. This deduction brings the whole GLR/OLB project total to $440.35M.
Diverting the Lowell Line
The MBTA should also not spend money on adding tracks, right next to a perfectly usable set of tracks. Diverting the Lowell Line commuter service to the Malden rail corridor (Haverhill Line tracks), as proposed in the GLR, could net a $120M savings, estimating the cost of the diversion at $30M and subtracting it from $150M cost of adding tracks as detailed in the GLX proposal.
That cost was also included in the BLX estimate, allowing us to deduct it from our working number, $440.35 - $120M = $320.35. Those savings could be better used to eliminate all of the dangerous grade intersections on the Haverhill Line between Reading and Oak Grove and to prepare it for a possible future extension of the Orange Line northward.
Thus, using numbers comparable to the revised estimates of the GLX ($954 Million), the GLR's cost of providing Somerville/Medford urban rail service sought by public transit proponents is $320,350,000 beyond expenses which the MBTA already intends to incur.
The above calculations are presented again below.
It would take several long paragraphs to even attempt to scratch the service on problems with the factors and ratios offered as comparative data in the summations of the MPO report. A thorough explanation would aide in promoting the benefits of this proposal, but for now a simple, not thoroughly explained computation, will have to suffice.
As the BLX might be substituted, operating cost wise, for this site's Orange Line Branch proposal, I find it necessary to issue a caveat on the numbers offered in the MPO report when being compared to the same factors and ratios provided for the GLX. A simple adjustment needs to be made, total operating costs should be compared on a per rider/mile basis.
Without giving the long explanation, there is a flaw in raw data used for hours saved relative to the cost basis used in the ratio, thus a more objective method of comparing operating costs of two proposals which offer different lengths of service is by costs/rider/mile. Using only 'new transit rider' is attempting to measure on a basis of additional revenue, but the same basis has not been used on the costs side, to deduct operating expenses on other modes displaced by a shift in riders. There are many angles there and too much to elaborate upon here.
Operating expense for a weekday on
the GLX is reported as $41,700 to transport 8,400 riders up to 4.2 miles, or $1.18/rider/mile, and
the BLX is reported as $76,800 to transport 13,500 riders up to 5.3 miles, or $1.07/rider/mile.