CT2030 plan will help too little and cost too much

Thank your lucky stars the CT 2030 plan hasn’t passed. It needs significant revision and a sharp eye placed on cost control.

Kerri Provost has a great piece at Real Hartford explaining why dedicating the majority of the money to highway projects is a terrible idea, so I won’t dwell on that. I focus here on a different, equally important criticism: the rail projects that are included are severely overpriced. As many American leaders continue to shortchange rail, it is tempting to charge ahead with any plan, but the legislature must resist. CT 2030 will, if approved in its current form and entrusted to CTDOT’s current managers, send the state back into the same debt hole out of which it has begun to climb, for crumbs’ worth of transit improvements.

First, the trip time reductions promised in exchange for proposed investment into the New Haven Line trunk are unambitious. The fastest express trains between New Haven and New York made the journey in 1h20 in the 1960s, and zone local trains making every stop between New Haven and Stamford took 1h50 a few years ago. Those services take 1h45 and 2h10 today, respectively. Even though 1h05 express and 1h25 zone local trip times have been shown to be feasible on the existing alignment with existing trains, CT 2030 promises to merely match historical trip times.

There is little reason to believe that today’s slow travel times increase safety. Last year, American passenger railroads (many of which have slowed their operations like Metro-North recently) had 1,505 significant accidents per billion train miles, an increase from earlier in the decade and well above the UIC average.

Although the speed improvements are modest, the costs, totaling $20 million per route mile, are not. The CT 2030 plan estimates $7.6 million per mile to upgrade signals for the promised trip times. In Denmark, by contrast, $2.2 million a mile is buying the entire network new signal hardware, communications-based train control (the technology used on the New York City Subway), and positive train control (PTC) to facilitate modern speeds and acceleration/braking rates.

Moreover, whereas track work is pegged at $3.9 million per track mile, $1.2 million per track mile is buying a total reconstruction of the double-track legacy line between Sagunto and Castellon in eastern Spain. Once rebuilt, the line, in place since the 1860s, will allow a maximum speed of 135 mi/h. The work is being done at night or on weekends to avoid disrupting the 150 trains that run each weekday. One 4-to-5-hour night window is enough to replace one-third mile of ties; all the tie replacement planned on the New Haven Line for 2019 would have taken two weeks of nighttime work at that rate.

Worse still, bridge construction costs are completely out of control. The Walk Bridge replacement is now estimated at $715.5 million for a 550-foot span. That is $6.5 billion a mile; the tunnels and systems on the highly overpriced Second Avenue Subway cost $500 million a mile. Moreover, the chosen design bakes in an increased maintenance cost for a movable span all for one customer upstream. SEPTA, no paragon of international efficiency, replaced the Crum Creek viaduct for $400 million a mile; at that cost, there would be a generous cushion for relocating any firm that relies on over-height boats.

Conspicuously absent from CT2030 is platform access for the center tracks at Bridgeport. The current setup forces all trains calling there to the local tracks, slowing them down and greatly reducing resilience. To accommodate island platforms serving all four tracks, the embankment needs minor widening into Water Street, which is oversized, and the headhouse needs to be replaced. All of that is standard-issue station reconstruction worth tens of millions of dollars. This way, a local train and an express train traveling in the same direction could board simultaneously out of each other’s way, and Waterbury trains could reverse at a platform out of the way of mainline traffic. And yes, the express service should stop, where the bus terminal and most of the jobs are, not along Barnum Avenue, where there are no concrete plans to build anything that comes close to justifying any rail service.

Even for simpler station builds, CTDOT estimates are excessive. For new stations at Enfield and Windsor Locks, along non-electrified track, the plan estimates $50 million and $65 million respectively. New stations in the Barcelona suburbs, where crews have to work around electrification, cost around $1.5 million each.

Finally, the plan budgets too much for new trains. CT 2030 calls for $985 million for 132 coaches and 30 locomotives. That sum should buy about 300 M8s or about 400 European specification diesel or electric (or dual mode) multiple units (MUs). These are legal in the US since the FRA published updated crashworthiness regulations a year ago. Locomotive-hauled trains are lucky to accelerate or brake at 1 mi/h per second, while MUs routinely achieve double that, translating to 1-2 minutes saved off the time needed to make each stop. There is positively no excuse to pay double for laggard trains.

The CT 2030 plan anticipates $300-400 million of yearly toll revenue. Fortunately, that much money with federal matching funds will be enough to maintain existing highways and build the mass transit projects if they are done at internationally acceptable costs. At least until 2021, increased federal funding is unlikely, and more debt is a nonstarter for this state. The mismatch between inflow and outflow in the current CT 2030 iteration should be a cue to clean house at CTDOT. Its current leaders are saying, through absurdly high cost estimates and a huge bias in favor of highway expansion, that they are incapable of taking the state where it needs to go. Who in power is listening?

Robert Hale lives in New Haven.