I haven't had much modeling time over the last couple of days, but I have been working through more of the details of my roster and operations, particularly the associated paperwork. For those who haven't seen it, I highly recommend Tony Thompson's blog as he has a number of very detailed posts on both subjects, among others.
I'm generally following the Nelson-Gilbert hypothesis that free-roaming cars like box cars should be represented on my layout in roughly the same ratios as the national roster. To that end I'm planning on roughly 1 box car per every 5,000 rostered. So using a spreadsheet that was posted on the Steam Era Freight Car Yahoo Group that compared the national box car rosters to the cars represented in the NMRA Charles Winter collection of photos, I compared the number of cars in my planned roster by road. As it turns out, with the models available on the market today, I'm right about where I should be with my planned roster. I have a number of extra cars planned, but the basic ratios match very well. While this does not address the mix of types of cars (single-sheathed vs. all steel), nor does it mean that the most plentiful classes of cars for a given railroad are always represented, it does show that I can develop a very representative roster using only the highly detailed models that are currently available without a lot of stand-ins.
In conjunction with this I've been looking at the operations paperwork. Tony's approach is very much along the lines of what I want to do. On the Model Railcast Show, Craig Bisgeier has also been talking about an operations application he's working on called 'Here to There' and his descriptions are also very close to what I'm thinking.
What I'd like to have is:
•Paperwork with a prototypical appearance
•Waybills properly generated by the loading road, not based on the car.
•Paperwork is generated based on shipments, not a specific car.
•I will select appropriate cars to account for ICC rules, rarity of cars, etc.
I'm finding that not only is this acheivable, it actually simplifies things. I have tried a number of applications, some quite expensive, to accomplish what I'd like. But they all require an enormous amount of input to approximate an accurate operational scheme. While some applications do feature loads, and fill 'orders' based on shippers and consignees, they don't account for ICC loading rules, or maintain a proper mix of cars other than your general roster.
Most of our rosters aren't balanced well as a whole. For example, some relatively rare cars are over represented - Atlas 1932 ARA, Proto 2000 50' Box and Auto Cars, Proto 2000 Mather Box Cars, etc. Other more common cars may be underrepresented, such as the NYC USRA Design steel box car (formerly available only as a Westerfield model, soon to be available from Broadway Limited.
Writing a program to account for the general ICC loading rules, how rare a car is, etc. is very complicated. In addition, all of the waybill programs I am aware of base the waybill on the loaded car. While this is accurate some of the time, the reality is that the waybill was generated by the loading railroad. Since the ICC Interchange rules requires a road to load an available foreign car first, it's just as likely that the New Haven would be loading a car bound for Texas in an ATSF box car as a NH box car. But the waybill will still be a NH waybill.
So right now I'm working on creating my potential waybills in Excel, and using a Word document with the waybill images set up using a Mail Merge to populate the data. Although it has taken quite a bit of thinking through, it's coming together nicely. Now the challenge I have is identifying what the various industries in New Britain would receive, and more importantly, what they would ship and where.
By doing it manually, all I have to do is populate relatively repetative information - The Household Fuel company receives coal of various types and grades. Once I've identified that, I simply duplicate it for the various railroads that might have shipped coal to New Britain, CT, and voila! I have all of the potential waybills I need for that industry. I don't need enter all of my roster, trains, interchanges, etc. nor do I need a complicated program to fulfill the order. I just need the waybill and then take a look at the available cars in storage to select an appropriate car to put on a train in staging.
What's even better is that I'm gaining a much better understanding of how all of this worked, and it actually makes sense, and gives a purpose to the layout.