General Discussion

MichaelSol wrote:

Oops, must have cut off the bottom of the image when I photoshopped it to reduce the image size. The attribution there reads "from H. F. Brown, "Economic Results of Diesel Electric Motive Power on the Railways of the United States of America," Proceedings of the Institution of Mechanical Engineers, 175:5 (1961)."

! thanks...

And in terms of effeciency, as implied by another poster, you will need to define HOW you measure effeciency...the external combustion engine (classic steam) is more mechanically effecient, but the overall system effeciency in terms of man hours to maintain (man hours labor vs operations hours) may be something entirely different...and dont forget support requirements, such as building, maintaining and resupplying water towers.  (In the east, water towers may not be a large problem, but as the RR's stretched west over the deserts, it became a significant issue. )  

JSGreen wrote:

...and dont forget support requirements, such as building, maintaining and resupplying water towers.  (In the east, water towers may not be a large problem, but as the RR's stretched west over the deserts, it became a significant issue. )

As a "for example" of the kinds of considerations that go into motive power studies, "water" savings is often cited as a benefit, without recognition of corresponding needs of the diesel-electric locomotive.

A rule of thumb in these kinds of studies is that, in addition to fuel costs, a diesel locomotive consumes substantially more lubricant than a steam engine, running between 10% and 12% of the cost of the diesel fuel consumed, while at the same time using considerably less water than the steam engine.

So, to properly account for the motive power changes, savings in water is, to a large extent, offset by increased charges for lubricants.

On the Milwaukee, which represents a good motive power model because it dieselized more rapidly than its competitors, in 1944 it used $861,039 worth of water, which declined to $158,074 by 1959, two years after dieselization was complete, adjusted for ton-miles.

But, lubricant costs increased from $419,687 to $1,324,196 on an equivalent ton-mile basis.

The cost savings in water of $702,965 were more than offset by the increase in lubrication costs of $904,509.

The difference of $201,544 represents a net loss incurred because of dieselization due to increased lubricant costs over the savings in water costs.

As I say, that is a "for example" of the kinds of considerations necessary to come to a realistic conclusion about relative "costs". Another would be the "availability" of the respective motive power types which impacts fleet size needs. Brown estimated for steam "at least 60%" availability, whereas he estimated 90% for a new diesel-electric. That figure was optimistic as EMD's availability estimate for the SD-40-2, advertised as a much more reliable locomotive than its predecessors, was only 84%.

it's 1:34 AM eastern time...

but the blog has been updated.  I have started an actual model based on data one of you gave me (i forget who....but please take credit if you wish).  The blog can be viewed here, it explains the update, and provides a link to the file:  http://steamvsdiesel.blogspot.com/

 Comments/Suggestions are highly appreciated, either here or on the blog.

 - Peter

Two observations:

1) tonnage and carloads declined substantially during the era of dieselization. If railroads had stayed with steam, the decline in needs would have permitted railroads to scrap older models and gain the fleet efficiencies shown by newer models. As of 1950, 40% of the existing steam had been built before 1915. Had the diesel-electric been the primary motive power, as a hypothetical, the same statistical result would occur as a result of being able to retire older, high maintenance items. Also, train crew manpower needs declined as a result of declining tonnage.

2) You have a conclusion about diesel fuel costs in your text, but I don't see any comparison with costs of coal on a comparable basis during the time frame in question.

The whole point of Dieselization, of course, was to purchase a more efficient motive power fleet. What was missing from most of the public relations promotionals was the key part of the process: "purchase," -- and because the economic service life of the machine being purchased was considerably shorter than equivalent steam, those studies that took the time to look at the financing implications of the changeover found that, for road power, the additional financing charges exceeded the cost savings obtained from efficiency gains in operation. This showed up very clearly at Milwaukee Road:

[Click to enlarge]

An interesting paper on a similar project from a modern perspective:

 http://www.trainweb.org/tusp/Coal%20Locomotive%20Final%20Paper.pdf

Just a thought on your project.  Include a section that is as complete as you can make it showing the complexity of your endouver - such as Michael wrote up about the maintenence cost accounting.  That should eat up a lot of your "space" requirement --- and it will positively impress your grader very much.

An added part of the so-called efficiency of the diesel over steam was the manpower needed to care for and feed the beasts - and I do not mean the roundhouse although that is a significant part of the cost structure.  I am refering to all those folks that serviced the locomotives and the hostlers and hostler-helpers.  These folks had to literally baby-sit steam engines doe to steam leaks in the throttle that would cause the engine to move off all on its own.  Into the turntable pit.  Into other locomotives.  Off the end of the track.  Manpower savings in the support structure were the primary savings followed by utilization.

Utilization increased because, among other things, the steam locomotive usually needed to be cut off and taken to the "house" at each crew change and thus replaced by another locomotive, whereas the diesel could - and soon, often did - operate through crew changes for the 500 mile ICC mechanical check of the trains thus reducing time loss, duplicated requirements for locomotives, duplicated service facilities, duplicated repair facilities, and eliminated delays due to having to stop for water enroute. 

This would show up instantly in any comparison of steam and diesel showing about a 50%+ fewer HP to move the same ton-miles.  The postulation would be that running through one locomotive change is a reduction of 1/2, running through 2 changes would require 1/3, 3 changes would be 1/4, where the diesels required would be the numerator and steam the denominator.  It should be somewhat plain that the real savings for the railroad was availability, actual HP that was needed to be owned to move ther tonnage, and the elimination of as much as 80% of the servicing and maintenence sites and personel.

It is absolutely amazing the cost of even one minutes delay to a train when propagated over the chain of delays that ensue and the equipment and crew costs when calculated out. 

And, as the Paper I cited to above can show, an awful lot of work can go into these projects, for just a school paper. A modern consideration, which I consider somewhat of a puzzler, is that while early diesel-electrics were advertised at 20 years, 1,000,000 miles, Iowa Curve analysis quickly showed that the depreciation period, upon which railroads based their purchase decisions, was wildly optimistic; and was quickly reduced in 1957 from 20 years to 14. Anytime there is a 30% negative error in assumptions, you can pretty well bet the original studies were worthless, but, by 1957, where was the industry to go?

But a further puzzler is this. While Class I experience with Northerns suggested annual mileage of 200,000 miles, and a service life well beyond 2,000,000 miles, modern BN high horsepower units (of less hp than a Northern) are averaging only about 90,000 miles per year, and overhauls are occuring at about 8 years; about 750,000 miles. I don't have good published data on that, but its my impression from a variety of anectdotal sources. This really skews the use of availability measures and fleet size costs compared to what was utilized in the original decisions to dieselize.

MichaelSol wrote:

A modern consideration, which I consider somewhat of a puzzler, is that while early diesel-electrics were advertised at 20 years, 1,000,000 miles, Iowa Curve analysis quickly showed that the depreciation period, upon which railroads based their purchase decisions, was wildly optimistic; and was quickly reduced in 1957 from 20 years to 14. Anytime there is a 30% negative error in assumptions, you can pretty well bet the original studies were worthless, but, by 1957, where was the industry to go?

Question about accounting and large companies...

The last time I operated a small business, what you could depreciate and the rate at which you could depreciate it was fixed in the tax code.  A notable case was PC's which had to be depreciated over something like 5 years, but most early PC's were being replaced in 3 years. 

Are large companies faced with the same rigid tax codes, and could that be the reason the period went from 20yr depreciation to 14 year depreciation?  It is plausable that rates were initially set for depreciating Diesels based on experience with Steam, but that when the RR's were able to show cause, they were allowed to set a more reasonable depreciation schedule.

JSGreen wrote:

Are large companies faced with the same rigid tax codes, and could that be the reason the period went from 20yr depreciation to 14 year depreciation?  It is plausable that rates were initially set for depreciating Diesels based on experience with Steam, but that when the RR's were able to show cause, they were allowed to set a more reasonable depreciation schedule.

No, the original 20 years was set based on studies that EMC/EMD said represented the "long service" life of road diesels. It was heavily promoted in the advertising of the time and which advertising is readily available to review for specific proof of the nature of the claim and the strong emphasis that EMD put on that particular claim.

It simply turned out not to be true, but by the time this became statistically obvious, dieselization was essentially completed on US Railways.

MichaelSol wrote:

No, the original 20 years was set based on studies that EMC/EMD said represented the "long service" life of road diesels.  It simply turned out not to be true,

Michael, I would guess if you took every road diesel built by EMD/GMD before 1980, their service lives would average more than 20 years. E units would probably average quite lower, but this had more to do with the reduction in passenger trains(?). Other builder's diesels built before 1980 probably averaged less than 20 years.

Are you thinking more in terms of only the early EMD's, say pre 1948?

nanaimo73 wrote:

Michael, I would guess if you took every road diesel built by EMD/GMD before 1980, their service lives would average more than 20 years. E units would probably average quite lower, but this had more to do with the reduction in passenger trains(?). Other builder's diesels built before 1980 probably averaged less than 20 years. Are you thinking more in terms of only the early EMD's, say pre 1948?

There is a difference between "service life" -- which can be indefinite if you want to spend the money -- and "economic service life" -- which has a specific econometric measure.

The "economic service life" is generally the useful life of an asset that results in the minimum combination of capital cost plus operating cost.

Once that first engine overhaul is necessary, the decision is made to either overhaul to engine or buy a new locomotive -- and because of the cost of the overhaul, it was considered a new capital investment either way. And that is what H.F. Brown and others pointed out in measuring the "economic" part of the service life: if you are going to consider the economic service life as extended by the overhaul, then the costs of the overhaul have to be included the operating costs. And that is where the EMD numbers for estimated maintenance were misleading but at the same time apparently plausible: the ongoing maintenance costs for the diesel-electric were cheaper but did not utilize proper accounting principles for comparable economic service lives with Steam. And when those principles were properly applied, the numbers reversed.

This is an extremely interesting thread. And I'm truly in awe of your knowledge on the topic, MichaelSol. You deserve a serious pat on the back!

I'm involved in a couple of geofiction games called Aurora (which has a 1945-55 technology level) and Scandia (which is late 20th centrury technology). From time to time, I try to include railroad related stuff. In one Aurora case, my fictional nation has a company that builds locomotives (steam) for market and I was trying to gen up some cost and time to build estimates. It's not easy data to find, trust me. I can say, unequivocably, that I've gotten better and more significant information in this thread here than hours of internet research got me previously.

nanaimo73 wrote:

Are you thinking more in terms of only the early EMD's, say pre 1948?

Here's  a real world example [click to enlarge]:

Milwaukee's mechanical department, which was strenuously trying to get rid of the Electrics, fudged the numbers for the GP40 after 1973, showing that the rate of cost increase in repairs slowed down from their recorded blistering pace of actual increases, doubling, for instance, every two years during the first four years of their service life -- which is pretty much what H.F. Brown's numbers show.

But, this where the capitalization comes in. The GP40 was incurring a substantial capital cost, which the EF-4's were not. Historically, the capitalization costs on any measured basis exceeded after 1960 the maintenance cost. So here, even if the electric power costs were the same as diesel fuel costs, the maintenance costs for a 6 year old Diesel-electric had already equalled the maintenance costs of a 25 year old straight Electric. If you were in charge of actually choosing the machine with the best "economic service life" alternative -- that six year point for the Diesel-electric becomes problematic.

Historically, Steam maintenance costs were somewhere in between Diesel and Electric if the capitalized cost of the overhaul were pro-rated. But, here, you can see that the much longer economic service life has a substantial effect on the impact of the capitalization cost of motive power since it adds to the cost of the Diesel-electric fleet, but is paid off at some much earlier date for the Electric fleet. The Steam fleet had characteristics more similar to the Electric fleet in that regard than to the Diesel-electric fleet.

GE didn't buy the figures on this chart at all; estimating that the EF-4 maintenance costs were, in 1972, approximately 47% of the GP40 costs.