Hi guys, this is my first post, I hope I follow all the procedures.

Could anybody out there tell me how you would calculate how much coal a locomotive would burn and how much water it would consume over a certain length of track?

Many thanks!

Welcome!


I recall in reading a first-hand account about the N&W turbine loco that they would dump 50 lb. buckets of coal into the auger as needed, and would keep track of how many buckets were used. I recall him saying that the same was done for the other classes of locos on the N&W.

I can't recall the name of the book off the top of my head, but will post it when I find it sometime tonight.

As for water, I think they could just measure how much was left at the end of the trip, and subtract it from what they put in the tank at the beginning, giving how much they used during the trip.

Others will correct me if this wasn't standard practice.

Well, there's a lot of variables to be considered. Load, grade of coal, condition of locomotive, weather, expertise of crew........ I think if you need accuracy, you have to answer these kinds of questions before you try to do any calculations. If you're being purely theoretical, you could determine an average train weight, average speed and steam consumption, average efficiency of the boiler and running gear - how much of the heat of the burning coal actually provides power at the rail / driver interface, the the effect of curvature and curve VS tangent ratio, and grades up or down, BTU content of coal used, potential rate of consumption to provide X amount of steam to move the weight at the speed desired......... it's pretty complicated.

Best wishes. I'd expect you'd be better off running a lot of trips with the same locomotive under the same loads on the same track and averaging the consumption. Real life doesn't act like theoretical life is supposed to with steam a lot of the time.

"Real life doesn't act like theoretical life is supposed to"

Well.... in theory, there is no difference between theory and practice, but in practice there is....

Dave is correct, very hard to calculate coal and water consumption very accurately, way too many variables.

You could do some rough approximations, but I would expect them to be +/- 50% from reality.

Cheers, Kevin.

I think the best option is to find an operating locomotive of similar proportions, doing similar work, and see if its owners would be able/ willing to tell you how much coal and water it uses per run, mile, month, etc... and how many miles are included... then do the math.

Then, take a wild a** guess to take into account differences between their operation and your theoretical one... double it for insurance... well, you get the picture... just be sure to include a risk factor in your estimated costs.

Steve Hunter

Book was by Louis M. Newton, Rails Remembered, vol 4, "Tale of a Turbine"

I thought it was bags, not buckets, but that's a minor nit to pick.

If I'm not mistaken (I will check in a bit) Ralph Johnson covers this in some detail in his book on the steam locomotive. I remember a fold-out drawing of a measuring stick that gauges the drop in tender cistern level to give the water consumption between two times, with some correction for slosh and other factors that shift the level.

Another potential data source for calculating historical fuel consumption at coarse resolution are state railroad commissioners' reports, which depending on the state and the year in question will frequently show fuel burned per train mile. These are averages calculated across an entire company over the course of a year, but in the absence of actual measurements for a particular locomotive or class of locomotives they may be better than nothing.

I know these fuel statistics were included in the annual railroad commissioners' reports for the State of Maine in the 1890s (and probably later as well but I'm not 100% sure), and have been used by historians of the several Maine Two-Foot railroads. It seems likely that other states collected this kind of data as well.

-Philip Marshall

The PRR did thousands of exhaustive tests on these factors in the Altoona Works' test plant. I'm pretty sure that many of these reports were archived.....I know the Altoona Model RR Club has an extensive archive of them, and I'd bet that the RR Museum of Pa in Strasburg has some, and probably the PA State Museum in Harrisburg, maybe Penn State's Pattee Library and the Altoona City Library as well.

In the book Set-Up Running, which covers the author's father's career as an engineer on the PRR in central PA, he gives extensive discussion of performance tests the father was involved in where the PRR was recording fuel and water consumption for different trains/locos/weather conditions/grades and crew differences. (Fabulous book on what life on the RR was back in the 1920's and 1930's, and enlightening for me as that is the area my grandfather railroaded on....)

The big RR's were well aware of the consumption factors and used such in determining distances between fuel and water stops, tonnage ratings on various grades, need for helpers, and etc.

I'd bet other RR's did this testing ( as described above RE the N&W) but the Altoona records are most likely the best documented/preserved, unless historical societies from the other RR's have archived the test reports.......

How realistic test plant data is compared to the real world is a matter of discussion. Besides PRR's famous test stand, the University of Illinois (Urbana) and Purdue University (West Lafayette) had test plants.

Is the OP looking for how to calculate, which is what was asked, or some actual numbers?

The tests in the Book "Set Up Running" mentioned by fairmontdave above were actual running tests with a train in normal operation. I suspect the PRR used this data to validate what they were observing at the test plant.

Even if you eliminate most variables, you still have the crew as a large and uncertain factor.

A tourist railroad eliminates many of the usual variables. The route, speed and consist are usually the same, or close. Passenger load may vary a bit, but in the total weight of the train isn't much change.


The obvious factors are the skill of both the engineer and the fireman. This is especially true for an oil burner. Just like driving a car, how you handle the throttle matters, and where you run the cutoff also influences consumption. The same for the fireman. Smoke is un-burned fuel, and popping off wastes steam.

Another factor I've noticed include weather, cold weather draws heat from the boiler, wet weather lubricates the rails.

It may be more consistent on a relatively flat railroad, but on a line with lots of curves and grades, it can vary a lot, even with identical consists and routes.

The test plant could vary the load to some extent, but provided what was a means of benchmarking within proscribed conditions - which is a great way to compare things since you can control things pretty well, but not perfectly reflective of the vagaries of real world service. That's OK....... provided you know that's what you are getting in the first place.

I wonder if the simulator programs have developed some realistic algorithms that could be adapted for this purpose, or are they only based in an artificial world?

I also wonder about that. Microsoft Train Simulator was terrible for simulating steam locomotives. The cutoff seemed to simply make you go faster, and wasn't at all realistic. I actually spoke with the developer about that, and he said they couldn't get it programmed the way it really should work.

I recall one of the newer programs, probably one of the versions of "Train Simulator" (Not the MS one) and it seemed to work pretty well. They even had a training program that showed you how a steam locomotive worked, and covered using the Johnson Bar, but to be honest I don't play it long enough to see if it really felt right. Initial impressions were good though.

Remember that another factor is how bad the management and/or accounting team wants to end the steam program.....

We got called into audit a particular program by the railroad VP of operations which shall remain nameless, and no matter how I looked at it, the fuel purchase records in particular were just...nuts. No historical basis with anything I could find that would explain how a particular mid-size locomotive could be burning coal at 2X to 3X any known standard compared to what I was seeing here. Made no sense at all given the boiler days and mileage on the locomotive.

This particular program also had several excursions operating from various points on the given system rather than a fixed location. Have you guessed it yet?

Yup, invoiced, and piled it on the ground. What they bought never necessarily hit the tender. Whatever was left at the end of the year turned to 'shrinkage', as there was no coal inventory - anywhere - recorded on the books at year end. Some was found, which helped somewhat, but it was obvious to me that the accountant involved wasn't exactly sympathetic to the cause and had no interest in determining 'actual' usage as opposed to what got purchased, piled, and...well, liberated. Even back-of-the envelope calculations of how much you bought, how much the tender would hold, and how many times you reloaded the tender on a regular basis proved as much.

Yup, steam is expensive to operate, certainly that way.

In comparison, I was somewhat amused to see that the 2014 21st Century Steam Program not only pulled their equipment, but their coal inventory, along with them as they went from location to location in the east last year.

Some real-life figures for coal and water consumption of British express passenger steam locos here: http://homepage.ntlworld.com/andrew.her ... m/kgv.html - scroll down to the end
also an article on the same topic at
http://semaphore.avonvalleyrailway.org/ ... loco3.html

These date from late 1940s/early 1950s and were established by comparative 'real-life' testing of different types over a range of routes. The 4-6-2's and 4-6-0's involved are of modest size compared with their North American equivalents of course, and the speeds probably higher than would be attempted today with the surviving elderly locomotives of the same class.
The conclusions appear to show that the way the loco is driven is so important that it's dangerous to generalise. But figures of around 40 pounds of coal per mile run appear to have been the norm.

nb British ton = 2240 pounds, US ton 2000 pounds
1 British gallon =1.20095 US gallons

A more modern discussion of coal consumption and driving techniques on much smaller narrow-gauge locos on a 9-mile line is here:
http://www.steam-loco-design.co.uk/zb_article_1.html
Some figures are, alas, in metric. 1 kg = 2.20462 lb.
Coal consumption after 'tweaks' equates to about 10 lbs per mile.

All sources emphasise that not all coal is the same ...

I hope that is of some help!

Chris Cook
London UK