My pictures of Chama to Alamosa trains mirrors Olaf's, the number of cars was quite irregular. The only "unit train" of tank cars up the hill that I have any knowledge of is the one photographed by Tom Gildersleeve in his well know winter pictures with the flanger and spreader on the rear of the tanks. But Everett raises an interesting point, there had to be some kind of "management" of the flow of tank cars. And like so many things about the ng., no detail is too trivial to not be worth detailed discussion. A key variable is obviously the irregular operation of Chama to Alamosa trains, especially after around 1960. Clearly there was a holding tank in Chama, and presumably the refinery also had a holding tank to allow take the slack for some irregularity at both ends. But i would opine that somebody at Alamosa would keep in eye on the holding tank there, and then instruct the Gramps folks in Chama as to how many cars to load. And once the cars were loaded they would be billed and the railroad would move whatever cars were billed. But clearly the number of cars loaded had to include some kind of "buffer" amound to reflect the uncertainty about when the next train would run. That would be my theory, but it would be more interesting to learn more about how this was actually implemented, who talked to who when, etc. For example it would be interesting if from time to time Alamosa would be running low and special trains might be called, etc. It would be interesting to compare how this was done in the 40's and 50's (pre computer, pre email, pre almost everything) compared to a similar process that must be in place for today's broad gauge unit trains of crude. It's not just railroad history it is industrial history.
JBWX