neroden
Engineer
Anyway, those of us who are looking at real numbers know that the stations aren't going to cost a significant amount.
The real question is what will be required in terms of track upgrades. That's generally far more expensive than stations.
Because of the rural nature of the line, I am going to make the assumption that the effort will be made to keep the price down -- no gold-plating, sandbagging, "throw in the kitchen sink" designs. This is perhaps an overly hopeful assumption...
Someone here who used to work in transportation planning (Wilbridge, IIRC?) pointed out specific ways in which the 2009 study was sandbagged by not even considering expert suggestions from him and others on how to keep costs down. I have heard statements from people inside Amtrak that the 2009 study was not done seriously because they didn't like the project. This shows in certain areas where they simply didn't bother to get cost estimates at all (see below).
That said, I just reviewed the 2009 study and it estimated the TOTAL station costs at $17.6 million -- for more stations than the minimal customer-serving number which I assumed. That's lower than my estimates! Amtrak, like me, assumed that only platform and parking was needed; their station cost estimates are in the $1 million range each.
Stations are minimal and can be funded individually by localities; one or two crewbases are minimal; but track upgrades are going to be a *lot*.
It's inevitable that that the project will require a *lot* of track upgrades. And this is going to be the bulk of the construction cost.
- I don't know the speed limits to which the line is currently maintained... but I do know they're too low in some places, lower than they were when this line was running passenger trains.
- I don't know how much of the line is currently double-tracked, but it appears that the answer is "not much". I don't know how many sidings it has, but I believe it's fewer than when it was running many passenger trains. It's not going to be enough and some new sidings will be required for passenger train to pass slow freight trains.
- All the freight operators really prefer station sidings at this point, and while BNSF is cooperative and likely to not demand them at every single station, there are going to be stations where they want an additional track at the station. In single-track territory, this would be to run opposite-direction freights past passenger trains. In places like Mandan, the area next to the station is a yard, and freight trains are using the tracks closest to the station for yard work; to change those tracks to station tracks, they'll require that their yard tracks be replaced.
The situation has probably changed since 2009 and the 2009 study says outright that there's "significant uncertainty" in the costs, *and* it was done in a fairly slapdash manner because Amtrak wasn't really that interested in it, but they *did* ask the host railroads what would be necessary (the host railroads will probably have changed their mind since then) and had specific project lists. At the time, these were the costs:
-- $44 million for equilateral turnouts between Chicago and St. Paul on the entrance and exit to single-track segments. I suspect that today, with higher traffic, this would be replaced with a call for more double-tracking, at higher costs. On the other hand, some of the double-tracking work may have already been done for freight service so some costs may have disappeared.
-- $24 million for more double-tracking between St. Paul and Fargo. To the extent this hasn't been done, this is probably still an accurate estimate. (I know some additional tracks have already been put in between St Cloud and MSP, so again some of the costs may have disappeared)
-- $307.3 million between Fargo and Jones, Montana. This is mostly for new sidings, plus grade crossing upgrades and powered switches and a major upgrade of the connecting track at Fargo. This is the bulk of the entire capital costs, and from some comments in the study, I suspect they've underestimated the needs.
-- $23 million from Jones, Montana to Helena, Montana for sidings and crossovers
-- an undetermined and unknown amount for a passenger track to bypass Laurel Yard, which isn't included in the study numbers because they didn't bother to price it out (they weren't really serious about the study)
-- $6 million from Helena to Sandpoint
-- $24 million from Sandpoint to Spokane
-- $96 million from Spokane to Pasco for double-tracking
-- $95 million from Pasco to Seattle for signalling and upgrading 221 miles of "dark territory", which really sounds like an underestimate to me. I would consider sending the train to Portland and be done with it; I don't know what the ridership difference would be between sending it to Portland vs. sending it to Seattle, but that question would have to be redone with a new study as all the demographics have changed. Seattle is a 4 million person metro area versus Portland's 2.5 million, however -- a factor of 1.6 -- so sending it to Seattle might be worth the $100 million+ in track upgrades. On the other hand, if this is the deciding factor on whether it's "too expensive" for the politicians' taste, just forget it and send the train to Portland.
-- Unknown amounts for installing PTC. At this point, much of the line already has PTC, but of course it would be needed for the parts which don't.
My summary: to a first-order estimate, the cost depends on how much of the line needs to be double-tracked / needs new sidings.
The other capital cost listed was $330 million for new trainsets, but they assumed bilevels, so let's try again. They somewhat pessimistically assume 6 trainsets, but let's go with that, since fewer trainsets would mean more expensive track upgrades which would probably cost more.
Nowadays, I think we can assume they'll be buying Siemens single-level cars and locomotives. Assume 2 locos per train (total 12) with Siemens's efficiency and fuel capacity; they're about $7 million per loco based on the standalone locomotive-only orders I found prices for, or $84 million. If you had to run 3 locos per train (you shouldn't) then it would be $126 million.
It's harder to estimate how many single-level cars will be needed. Approximately 4 sleepers, 5 coaches, diner, lounge, that's 11 cars per train or about 66. The cars appear to cost $3.5 million each based on the VIA order (though 1/5 of the VIA order is cab cars, which are more expensive, so it's a substantial overestimate) so that's $231 million.
So that would be about $315 million, or $357 million with 3 locos per train.
The real question is what will be required in terms of track upgrades. That's generally far more expensive than stations.
Because of the rural nature of the line, I am going to make the assumption that the effort will be made to keep the price down -- no gold-plating, sandbagging, "throw in the kitchen sink" designs. This is perhaps an overly hopeful assumption...
Someone here who used to work in transportation planning (Wilbridge, IIRC?) pointed out specific ways in which the 2009 study was sandbagged by not even considering expert suggestions from him and others on how to keep costs down. I have heard statements from people inside Amtrak that the 2009 study was not done seriously because they didn't like the project. This shows in certain areas where they simply didn't bother to get cost estimates at all (see below).
That said, I just reviewed the 2009 study and it estimated the TOTAL station costs at $17.6 million -- for more stations than the minimal customer-serving number which I assumed. That's lower than my estimates! Amtrak, like me, assumed that only platform and parking was needed; their station cost estimates are in the $1 million range each.
Stations are minimal and can be funded individually by localities; one or two crewbases are minimal; but track upgrades are going to be a *lot*.
It's inevitable that that the project will require a *lot* of track upgrades. And this is going to be the bulk of the construction cost.
- I don't know the speed limits to which the line is currently maintained... but I do know they're too low in some places, lower than they were when this line was running passenger trains.
- I don't know how much of the line is currently double-tracked, but it appears that the answer is "not much". I don't know how many sidings it has, but I believe it's fewer than when it was running many passenger trains. It's not going to be enough and some new sidings will be required for passenger train to pass slow freight trains.
- All the freight operators really prefer station sidings at this point, and while BNSF is cooperative and likely to not demand them at every single station, there are going to be stations where they want an additional track at the station. In single-track territory, this would be to run opposite-direction freights past passenger trains. In places like Mandan, the area next to the station is a yard, and freight trains are using the tracks closest to the station for yard work; to change those tracks to station tracks, they'll require that their yard tracks be replaced.
The situation has probably changed since 2009 and the 2009 study says outright that there's "significant uncertainty" in the costs, *and* it was done in a fairly slapdash manner because Amtrak wasn't really that interested in it, but they *did* ask the host railroads what would be necessary (the host railroads will probably have changed their mind since then) and had specific project lists. At the time, these were the costs:
-- $44 million for equilateral turnouts between Chicago and St. Paul on the entrance and exit to single-track segments. I suspect that today, with higher traffic, this would be replaced with a call for more double-tracking, at higher costs. On the other hand, some of the double-tracking work may have already been done for freight service so some costs may have disappeared.
-- $24 million for more double-tracking between St. Paul and Fargo. To the extent this hasn't been done, this is probably still an accurate estimate. (I know some additional tracks have already been put in between St Cloud and MSP, so again some of the costs may have disappeared)
-- $307.3 million between Fargo and Jones, Montana. This is mostly for new sidings, plus grade crossing upgrades and powered switches and a major upgrade of the connecting track at Fargo. This is the bulk of the entire capital costs, and from some comments in the study, I suspect they've underestimated the needs.
-- $23 million from Jones, Montana to Helena, Montana for sidings and crossovers
-- an undetermined and unknown amount for a passenger track to bypass Laurel Yard, which isn't included in the study numbers because they didn't bother to price it out (they weren't really serious about the study)
-- $6 million from Helena to Sandpoint
-- $24 million from Sandpoint to Spokane
-- $96 million from Spokane to Pasco for double-tracking
-- $95 million from Pasco to Seattle for signalling and upgrading 221 miles of "dark territory", which really sounds like an underestimate to me. I would consider sending the train to Portland and be done with it; I don't know what the ridership difference would be between sending it to Portland vs. sending it to Seattle, but that question would have to be redone with a new study as all the demographics have changed. Seattle is a 4 million person metro area versus Portland's 2.5 million, however -- a factor of 1.6 -- so sending it to Seattle might be worth the $100 million+ in track upgrades. On the other hand, if this is the deciding factor on whether it's "too expensive" for the politicians' taste, just forget it and send the train to Portland.
-- Unknown amounts for installing PTC. At this point, much of the line already has PTC, but of course it would be needed for the parts which don't.
My summary: to a first-order estimate, the cost depends on how much of the line needs to be double-tracked / needs new sidings.
The other capital cost listed was $330 million for new trainsets, but they assumed bilevels, so let's try again. They somewhat pessimistically assume 6 trainsets, but let's go with that, since fewer trainsets would mean more expensive track upgrades which would probably cost more.
Nowadays, I think we can assume they'll be buying Siemens single-level cars and locomotives. Assume 2 locos per train (total 12) with Siemens's efficiency and fuel capacity; they're about $7 million per loco based on the standalone locomotive-only orders I found prices for, or $84 million. If you had to run 3 locos per train (you shouldn't) then it would be $126 million.
It's harder to estimate how many single-level cars will be needed. Approximately 4 sleepers, 5 coaches, diner, lounge, that's 11 cars per train or about 66. The cars appear to cost $3.5 million each based on the VIA order (though 1/5 of the VIA order is cab cars, which are more expensive, so it's a substantial overestimate) so that's $231 million.
So that would be about $315 million, or $357 million with 3 locos per train.
In general a great series that unfortunately never made it to DVD. I have some episodes that I taped off the TV at the time.
