The October 2013 issue of Trains magazine carried an article on a 1979 study by Conrail to expand their electrification system, adding an additional 342 route-miles of overhead catenary from Harrisburg to Conway Yard, 19 miles west of Pittsburgh as well as create an electric freight alternative to the Northeast Corridor between Philadelphia and Newark, New Jersey. I'm certainly not going to quote the entire article, but rather simply comment on particular items of it.
The first and most obvious issue is that the decision was made not to electrify, despite the benefits cited in the report (according to the article, a 9.9 year payback period and savings of $84 million per year in operating costs in the first 8 years alone). This was due to a lack of available financing and this was absolutely the right thing to do. Had it gone forward, it likely would have been seen as a debacle and potentially set back Amtrak's already delayed plans to extend electrification north from New Haven to Boston.
The proposed operating savings are, presumably, based upon reduced locomotive maintenance costs and reduced fuel costs thanks to using electricity rather than oil. The reduction in locomotive maintenance costs is probably somewhat overstated, especially compared to using best practices for diesel locomotives. In 1977, Conrail was paying $257,637 (in inflation adjusted 2010 dollars) per diesel locomotive per year; Norfolk Southern, who split Conrail between CSX and themselves, paid $124,720 in 2010 and the highest listed peer $203,840 in 2010 (Slide 3). I do admit that this may be due to improvements in the locomotives themselves. With the crash in the price of oil starting in 1986, I also strongly suspect that the savings in fuel costs would have been greatly diminished, though not dropping to the point of electric traction being more expensive.
The real killer, however, is the cost of installing the electrification itself. While the article refers to a net investment of $767 million ($657 million after accounting for expenses present with diesel traction), much of this is due to various credits and the remaining value of the locomotives. This is a radically different number from the actual investment dollars required. The fixed plant investment, for the catenary, substations, etc. amounted to $432 million; at then-present interest rates, with a thirty year loan, this would have resulted in financial costs of $53.3 million. The additional locomotives amounted to $1.26 billion over the period of 1980-2010, including a $550 million credit. While, to be fair, a diesel only case required $3.72 billion over the same period, including a $2.143 billion credit, the costs would have rapidly eliminated the operational savings that electrification presents. This resulted in sufficient skepticism from the private markets that the funds could not be raised for electrification and so the project was shelved.
Electrification can never justify itself, in the private sector, simply on cost savings. Those savings simply are not there once accounted for. The only way electrification ends up justifying itself is through superior operational characteristics and the resulting increased revenue.