Yes, these costs are high. However, note the presentation is specific about the scope in that these are full accessibility upgrades, specifically noting appropriate vertical circulation. In some cases this could mean at least 2 elevators, extensive ramps and extensive stairs squeezed within the ROW, complex topography, and MBTA property, potential neighboring infrastructure adjustments and upgrades, and contractors charging a high premium when working around the T's operating schedule. I'm sure alongside all of that they're penning contingency in for some community requests, like Winchester's full canopy system, etc.
Inflated? Yes. Am I as shocked as everyone here? Not entirely..
It does seem strange that the solution then is to pick the really simple ones and have Keolis put cheap mini-highs in. Those are the easy ones where you won't have these high "complexity" costs.
I'll give every CR station without full high platforms a score between 1-4 for complexity. The rough scoring metrics are 1 is just build the platforms, 2 is add a couple ramps and a new crossing, 3 is adding a new footbridge/elevators and 4 is a full rebuild.
Fitchburg Line:
Fitchburg - 1
N. Leominster - 3
Shirley - 2-3, depending on freight traffic
Ayer - 2
West Concord - 1
Concord - 1-2, depending on how to treat the station building
Lincoln - 1
Kendall Green - 1
Brandeis/Roberts - 1
Waltham - 1 but the station needs to be moved to west of Moody St
Waverley - 2-3
Belmont - 2-3
Porter - 1-2
(4 difficult ones, 9 easy ones)
Lowell Line:
Lowell - 1
N. Billerica - 2
Wilmington - 3
Mishawum - 2 if it's ever reopened
Wedgemere - 2
W. Medford - 1
(1 difficult one, 4-5 easy ones)
Haverhill Line:
Haverhill - 2
Bradford - 1
Andover -1
Ballardvale - 1
N. Wilmington - 1
Reading - 1
Wakefield - 2 since it needs to be either North or South
Greenwood - 1
Melrose Highlands - 1
Melrose/Cedar Park - 1
Wyoming Hill - 1
(11 easy ones, no difficult ones)
Newburyport/Rockport Line:
Ipswich - 1
Hamilton/Wenham - 1
N. Beverly - 1
Rockport - 1
Gloucester - 1
W. Gloucester - 1-2
Manchester-by-the-Sea - 2
Beverly Farms - 1
Montserrat - 1 but the platforms need to be moved west of Spring St
Beverly - 3
Swapscott - 1-2
Lynn - 4
River Works - Doesn't count
(2 difficult ones, 10 easy ones)
Greenbush, Kingston, & Middleborough/Lakeville Lines:
No upgrades needed
Providence/Stoughton Line:
S. Attleboro - 4
Attleboro - 2
Mansfield - 3 (No foot crossings on the NEC, if this isn't a rule then it should be.)
Stoughton - 1 but the platform needs to be moved south of Wyman St
Canton Center - 1
Canton Junction - 3
Readville - 4
Hyde Park - 2 if the tracks aren't modified, 4 if they are
(4-5 difficult ones, 3-4 easy ones)
Franklin Line:
Forge Park/495 - 1
Franklin/Dean College - 1
Norfolk - 1
Foxboro - 1
Walpole - 2-3
Windsor Gardens - 1
Norwood Central - 1
Norwood Depot - 1 (Do we actually need both of these?)
Islington - 1
Dedham Corporate Center - 1
Endicott - 1-2
(0-1 difficult ones, 10-11 easy ones depending on how you configure Walpole,
Needham Line:
Should just be converted to rapid transit
Fairmount Line:
Readville is a 4, everything else is fine
Framingham/Worcester Line:
Grafton - 3
Westborough - 3 (1 if the giant metal ramp of doom isn't touched or level crossings are allowed)
Southborough - 1
Ashland - 3 (1 if the giant metal ramp of doom isn't touched or level crossings are allowed)
Framingham - 1
West Natick - 3 (1 if level crossings are allowed)
Wellesly Sq. - 3 (1 if level crossings are allowed
Wellesly Hills - 3 (1 if level crossings are allowed
Wellesly Farms. - 3 (1 if level crossings are allowed
Auburndale - 3 (1 if level crossings are allowed
W. Newton - 3
Newtonville - 3
(10 difficult ones, 2 easy ones, or 3/9 if level crossings are okay)
So about 19-20 stations where there is some major complexity factor, and a whopping 45ish ones that range from very easy to "It literally cannot be easier than this." Nowhere near enough to drive the average cost per station up to $80 million, or probably even $50 million.