NellieBly wrote: In any case, a fairly exhaustive study of "PTC-preventable accidents" by FRA using 20 years of data demonstrated conclusively that improvements in safety will not pay for the cost of PTC.
Ok, I'm gonna go fold a tin foil hat, put it on, and say:
"Well, of COURSE the FRA's study says it's not worth it - the Class Is would freak out if they did, and the last thing the FRA wants to do is upset the industry."
That said, I have a hard time believing the FRA's study was unbiased - the rest of the world apparently thinks some form of PTC is worth it, and let's face it, determining it's not worth it means you don't have to attack the sticky issue of getting a dozen railroads to agree on a single standard
That having been said, I have two responses to Mr. Norman's comments about "investor owned" railroads having to bear the costs of PTC. First, in the 1920s the ICC required all Class I railroads to equip one division each with automatic train stop technology. Railroads had to bear the cost. So there is a clear legal precedent. FRA has the legal authority to compel the railroads to install PTC, and there is nothing in law required that the railroads be compensated for any costs incurred.
There's beyond plenty of precedent - Airliners have to bear the cost of A/Ds, nuke plants have to bear the cost of upgrades (The new rules after TMI-2 and Browns Ferry 1 closed a number of older, smaller nukes), etc etc etc. Railroads aren't unique in this regard. In fact, if anything, they're unique in that the regulating agency's used implementation cost as an excuse to drag their feet on the issue. It'd be like a post-TMI NRC letting Dresden 1 keep on running without an emergency cooling system*, arguing that the installation cost exceeded the benefit of preventing that small risk** of Illinois becoming uninhabitable.
The cost of PTC is estimated at about $4.5 billion. However, if PTC is installed, significant parts of the existing signal system (all the intermediate signals, for example) will not need to be replaced. They can be removed. Signals and other equipment at interlockings and ends of controlled sidings will have to be renewed, of course. The total cost of PTC plus replacement of those signal system components that will still be required with PTC will be around $7.5 billion ($4.5 billion for PTC, where most of the equipment is on locomotives or in the control center, plus about $3 billion for wayside equipment that will have to remain). Thus the *net* cost to the industry of PTC will be a savings of about $1.5 billion over simply renewing the conventional signal system.
It depends on the system implemented, and whatever cost scaling happens. Which is why I think ETCS would be a decent choice, though the industry fears anything with 'European' in its name. but creating an instant huge scale market for an international standard would push prices down a lot, IMHO. And, from a technical standpoint, I'm not convinced GPS-based systems can do the job - I see the desire to use GPS here more of a hammer/nail thing than a practical one.
And of course PTC will provide greatly enhanced functionality. In addition to eliminating human factors accidents, the real-time train position data produced by PTC will yield a number of economic benefits to the railroad industry. I previously estimated those benefits to be in the range of $2 billion per year.
Which begs the question - what's the cost of the average human factors accident?
But that's...just my opinion, even if I did say it in a report to Congress.
Congress never listens to reasoning. We'd be a lot better off if they did...
*I'm guessing here, actually, that it never had one. What I've read says 'no' on it, Indian Point 1, Humbolt Bay, and Big Rock (Shippingport I don't think did either, but it was basically on the way out by '79 anyway), though it WAS a feature at Yankee Rowe and SONGS-1, both of which were first generation nukes also. Or it might have had a limited one, I know sodium pentaborate injection, a big feature in GE BWRs, was used on it.
**I'm not aware of any GE BWR that ever had a LOCA.