Saint Louis v. Kansas City (and Chicago) - Trains Magazine

Railway Man

[snip] Very few hump yards that generate trans-Mississippi cars have the kind of volume that can justify a St. Louis train as opposed to a Chicago train, unless the yard is going to accumulate cars for two or more days, which has rather expensive effects on car-hire and yard infrastructure costs.  Suppose you go to one of the big western hump yards and look at it closely.  Say the yard processes 2,400 cars per day.  Half of those will go east, half west.  That breaks down to 12 trains each way daily.  Now subtract all the cars that are not destined for Chicago and St. Louis, and the long-easts that could go through Chicago and St. Louis.  That will be roughly one-third to two-thirds of the cars, leaving us 6 trains per day worth of Chicago's, St. Louis's, and long-easts.  Now subtract out the Chicagos, and we have maybe 3 trains per day left of St. Louis and long-easts.  Now subtract the long-easts and we have maybe 1 train per day left of St. Louis cars.  Now think about having to have another 2 bowl tracks in the yard to hold the long-easts for another St. Louis train every day, that we could otherwise just lump into the Chicagos and get rid of them.  In other words, instead of averaging 4 hours dwell time on a long-east by putting it into a Chicago train, we've mandated 12 hours average dwell time.  Now multiply 8 hours dwell time, times 100 cars per day, times 365 days per year, times $4/hour, and we've got a cost of about $1.2 million/year extra to avoid Chicago.  Add onto that the allocated cost of $10 million for two bowl tracks, amoritized in (rough numbers) over 10 years at $3 million/year.  Add onto that the loss of traffic because we've put a 10% hit into the travel times of the long-easts.

RWM

RWM, whatever the mathematical disease is that I've got, it looks like it's contagious . . . spreads via close forum contact, perhaps. 

I see and concur with your point above about the first St. Louis bridge not being determinative in attracting railroads - even with an invitation from on high - due to the disadvantage of the intervening territory beyond, in comparison with alternative routes around that territory to the north directly from Chicago. 

Next, with respect – because I might begin to look “nit-picky” here - a couple of clarifications, questions, and quibbles with regard to the above:

1.  The "delta" / increase in the dwell for the long-east cars via St. Louis is more like 1.6 hours average, not 8, because those trains will leave more often than just the St. Louis cars (only) trains, as follows: 

My understanding of your hypothetical eastward train volume is 6 per day, of which 3 go to Chicago only (6 - 3), 2 are long-easts only (3 - 1), and 1 to St. Louis only.  Keeping the same implied (unstated) assumptions / basis, such as inbound cars arriving at a uniform or constant rate per day with a proportionate distribution of destinations, and outbound trains also leaving at a uniform rate with no unbalanced destinations:

Long-easts via Chicago scenario:  If we route all the 2 trains of long-easts via CHI in addition to the 3 CHI trains, then there's 5 trains a day (24 hrs.) going that-away at a 4.8 hours average interval, or 2.4 hours average dwell for those cars.   St. Louis gets its 1 train (only) per day, with 12 hour average dwell.  The weighted average dwell is then [(5 x 2.4 = 12) + (1 x 12 = 12) = 24] / 6 = 4 hrs. per car.

Long-easts via St. Louis scenario:  If we route all the 2 trains of long-easts via St. Louis instead, then there's 3 trains (1 St. Louis + 2 long-easts) a day (24 hrs.) going that way at a 8 hours average interval, or 4 hours average dwell for those cars.  Chicago now also gets its 3 trains (only) per day, likewise with 4 hour average dwell.  The weighted average dwell is then [(3 x 4 = 12) + (3 x 4) = 12) = 24] / 6 = 4 hrs. per car again.  But the dwell for the 3 trains of Chicago (only) cars has increased from 2.4 hrs. to 4 hrs. (+ 1.6 hrs.), as has the dwell for the 2 trains of long-east (only) cars increased from 2.4 hrs. to 4 hrs. (+ 1.6 hrs.), while the 1 train of St. Louis (only) cars have benefited from a decreased dwell from 12 hours to 4 hours (-8.0 hrs.).  [Note that this checks: the aggregate of the increases in dwell times = 5 trains (3 Chicago + 2 long-east) x 1.6 hours increase = 8.0 hrs. equals the sum of the decrease in dwell times for the 1 train of St. Louis (only) cars at 8.0 hrs.] 

2.  2 trains of long-east cars per day would be 200 cars per day that are affected by the increased dwell, not 100 cars.

3.  $4 per hour for car hire ?  That’s $96 to $100 per day, around $3,000 per month or $35,000 per year.  Seems high to me unless there are other components or aspects there that aren’t stated, it’s a special car, or has unusually high daily expenses - such as a refrigerator car that uses fuel all the time as well, etc.  I thought about half that - $2 per hour, $40 to $50 per day was more typical ?

4.  Redoing the multiplication, but with my figures instead:  Multiply 1.6 hours increase in average dwell time, times 200 cars per day, times 365 days per year, times $2/hour, is about $235,000/ year extra for the long-east cars (only) to avoid Chicago.  Breaking it down a little: 1.6 hours increase in per-car dwell at $2 per hour = $3.20 per car, times 200 cars per day is $640 per day, times 365 days = $235,000 per year in increased car-hire costs.  Note that the $640 per day in increased car hire isn’t near enough to justify even another crew/ train start and locomotives to avoid it by sending out a train sooner.

If we also look at the 3 trains of Chicago cars (300, of 50 % more), whose average dwell is also increased by 1.6 hours, the annual increase in car hire cost from that is about $350,000.

Balanced against that is the savings in car hire for the 100 St. Louis (only) cars – 8 hours x $2/ hr. = $16 per car x 100 cars per day = $1,600 per day x 365 = $584,000 – exactly the same as the total of the increased car hire costs for the Chicago (only) and long-east cars (only) !  (Or, as my father-in-law likes to say, “Save on the bread – lose on the cheese.”)  So, it’s really more of a question as to which traffic route gains, and which one loses.

5.  The $10 million for the 2 bowl tracks and the amortization of same at $3 million per year sounds way high to me, as follows: 

For a 6,500 ft. long (roughly 100 cars at 60 ft. each = 1 of these train lengths, plus a couple hundred feet back to each of the clearance points, etc.) bowl track, the track (only) at $100 per foot is $650,000.  Add a pair of turnouts (switches) at $50,000 each and allow $100,000 for the retarder to serve that track plus something for signals and controls takes it up to $850,000. 

For land, use a 20 ft. wide average strip x 6,500 ft. long including the turnouts and distance to the clearance point = 130,000 sq. ft. = 3.0 acres.  At $200,000 per acre (informed guess), that’s $600,000.

The capital recovery factor for the amortization over 10 years at an internal rate of return (“IRR”) of 10 % per annum is 0.16275, or $162,750 per $1 million; the stated $3 million per year for $10 million over 10 years would be an IRR in the range of 27.5 %.  Even accepting that, the $4 million of allocated yard infrastructure investment will cost about $1.2 million additional; if the 10 % IRR is used instead, it would be more like $650,000 added per year for those 2 additional bowl tracks.

Certainly if I have misunderstood or made a mistake in any of the above, please feel free to point that out.

- Paul North.

"This Fascinating Railroad Business" (title of 1943 book by Robert Selph Henry of the AAR)