Thursday, December 30, 2010

Choosing a model car fleet, Part 2

As mentioned in the previous post with this title (, I adhere to the idea of proportioning a model car fleet to prototype proportions. In the previous post, I used the example of PFE cars. Today I would like to comment on a much broader issue, the proportions of cars other than SP and PFE (called “foreign” cars by railroaders).
     The basic approach nowadays to this problem is called the Gilbert-Nelson approach, based on ideas first developed by Tim Gilbert and Dave Nelson. They had looked at extensive prototype information, particularly conductor’s time books, and realized that there was a pattern: many cars present in trains were numerically in proportion to the size of the owning road’s fleet, regardless of where in the country the data originated. Of course, as they fully recognized, this can only be true of free-running cars like box cars, flat cars and gondolas which are not specially equipped, and is likely true only on main lines. A coal branch, for example, will obviously be quite different.
     They also recognized that certain factors can distort the general pattern. For example, interchange requirements or pool agreements can change the data; so can “hostile” or competitive relations among railroads. It’s well established, for example, that the intense rivalry between PFE and Santa Fe’s SFRD meant that empty cars were always sent back to the other company and not reloaded as foreigns. On the other hand, the legally required preference of SP interchange with UP at Ogden would increase each road’s presence on the other. But absent such considerations, free-running foreign cars would be expected to follow Gilbert-Nelson.
     This means that one can collect data on the larger fleets of freight cars in the U.S. and use it as a rough guide to foreign cars on an SP layout. These would include the following major players:

     These data are for 1950, and are modified from total car fleets by removal of ore cars, hopper cars, and ballast cars. As an SP modeler, the likelihood of coal-road hopper cars on my layout is pretty small and I neglect it. I have also separated out the PFE cars from SP and UP. So this figure shows the major railroad ownerships I need to model, and once beyond the “top six” in this diagram, numbers don’t vary greatly.
     How might these railroads be affected by the special considerations mentioned above? It’s known that SP had friendly interchange relations with NP, RI and IC (at far ends of the system) in addition to the UP relationship mentioned above. Conversely, ATSF and WP were rivals in the far west, while MP was a strong rival of T&NO in that territory.
     The final question I’ll raise today is the proportion of home road cars, vs. foreign cars. This does not seem to have a universal answer, and Gilbert-Nelson found different proportions in different places. There is an old rule of thumb, that home road cars may be a third to a half of all cars, but that seems too high for SP in California, based both on photos and conductor time books. I will probably settle on one third, which is perhaps a credible compromise.
     Some discussion of this issue, predating Gilbert-Nelson, can be found in an article I published some time ago, copies of which can be obtained from the NMRA library. It is the source of data for the graph shown above. Here’s the citation:
Anthony Thompson, “Railroad Freight Car Fleets," in Symposium on Railroad History, published by A.C. Kalmbach Memorial Library, NMRA, Chattanooga, TN, 1990, pp. 27–44.
     In future posts, I will explore this topic in more detail.
Tony Thompson


  1. Hey Tony,

    The MP shipped a lot more traffic west on the SP than I think you're giving it credit for. The T&P's (MP) western end from Fort Worth to Sierra Blanca was mostly involved with moving interchange traffic to and from the Sunset Route. I can't vouch specifically for the 1950s but from the ’60s onward it wasn't uncommon to see good-sized blocks of MP family cars moving between Sierra Blanca and California.

  2. You're right about the T&P connection, and I think there WAS a substantial presence of MP (and T&P) cars in the far west as a result. Photos support this idea. I was only referring to the attitude of T&NO toward MP and T&P elsewhere in Texas and Louisiana. What was at one time the Gulf Coast Lines competed strongly in several of T&NO's territories. As a modeler of SP's Coast Division, of course, the T&NO side is of less importance to me.
    Tony Thompson

  3. Tony

    Forgive me, but I have only just come across your blog following a link from the MCFL group.

    I, too, model the Coast Division albeit a little later in time. I find your discourse above really helpful. I was surprised to read that the proportion of 'local' cars is as low as 33% on the Coast line, but will follow your advice.

    Using the Gilbert-Nelson formula as a rough guide to foreign cars on your SP layout, moderated by your comments on local cars which I take it includes both SP and PFE cars on the Coast line, what is the balance of the cars in your train orders? Whilst I accept there is a high liklihood of PRR and NYC box cars turning up on the Coast Line in the 1930's do you surmise that there were c20% of cars in say the Surf Turn or the reefers in the 'Smoky' from these two roads so far west? Using John Signor's and your excellent books on the Coast Line as my guide this does not seem to be the case. But I will be guided by your views.

    Best wishes


    Peter Harris
    Nottinghamshire UK

  4. Peter, I'm glad you find the commentary useful. I think it's vital in thinking about car fleets to think separately about different car types. Gilbert and Nelson themselves only proposed that their idea would work for free-running foreign cars, so it's restricted to box, gondola and flat cars. It would definitely NOT work for reefers, which did not run freely around the country like box cars.

    For free-running foreigns, I think box cars are by far the most relevant in the far western U.S., because gondola fleets in that region were dominated by GS or drop-bottom cars, often owned by the home road, and flat cars (in the case of SP) were dominated by SP's own fleet of cars for lumber traffic.

    I do think that plain box cars (separating them from automobile cars and specially-equipped box cars such as auto parts cars) would obey the Gilbert-Nelson hypothesis. The nagging question is the proportion of home road (SP) cars in this category. I don't yet have a firm grip on this point. The conductor book I've been analyzing in my February and March posts does not have that many box cars, being dominated by reefers and sugar-beet cars. I mean to go through and analyze the box car content, though, and will report it in a future post.
    Tony Thompson