Why are Railroad Replacement Costs Higher than Book Value?

Implications: One reason a railroad's replacement cost is higher than its book value is that the company has been harvesting the infrastructure by deferring normal replacement cycles such that the average quality and life remaining of the installed base is much less than what would be expected under a steady-state, nomalized maintenance program. The railroads generally say that they are properly maintaining their track and investing capital at a rate that ensures the plant is in a state-of-good repair; but, they also say that there are enormous "replacement cost" obligations that are not reflected in the book value of their assets.  Bridges are a clear example where many structures are over 100 years old.  But, the condition, quality and life of the track assests (rail, ties, and ballast) is a very complex issue and very difficult to assess.  In this article, I explain this complexity and show how we have developed methods for discerning this important aspect of railroad evaluation.

Analysis: Overview of the Business Problem

The railroads’ ability to handle traffic volumes depends on the quantity and the quality of their physical plant.  Railroad tracks must provide relatively direct movement between markets.  The tracks must enable the railroads to operate trains at economical speeds and frequencies, requiring a balance be struck between too little capacity and too much money spent on track structure and condition.  As agricultural and industrial production changes location over time, the railroads have to relocate their infrastructure to follow the markets.  As traffic grows between existing markets, railroads initially increase capacity through operational means (e.g. running more trains), but eventually have to expand capacity on existing lines through capital improvements (e.g. adding sidings and double tracks or upgrading the signaling system). 

Even if traffic volumes remain constant, time and use will dictate some level of track replacement on an ongoing basis.  We would expect railroads with relatively older physical plant and/or shifting markets to spend proportionately more in the medium term (next 10 years) than railroads with newer track and/or stable markets.  This suggests some railroads will have better cash-generating potential than others.  

Total railroad capital spending has three major components: Signals & Communications, Track & Structures, Rolling Stock.  In this analysis we will focus on Track, which comprises about 2/3 of the total dollars. 

Of the track components, Rail and Ties by far account for the largest dollar amounts and therefore we will further concentrate our focus on them.  Capital investment and operating expense of ballast and OTM is largely connected to work activity in rail and ties.  When rail or ties are replaced, the railroad also must renew a certain portion of the surrounding ballast and must replace a proportional amount of spikes, plates, rail bars, and bolts (OTM).  As such, we can simply ensure that our unit costs for ties and ballast reflect the “all in” cost of installation in order to capture the total cost of track inclusive of ballast, OTM, and switches.   We have determined the expected remaining life of each railroad’s rail and ties, and forecast how much railroads need to spend in the future to keep tracks in a state of good repair, given expected changes in traffic, materials quality, and maintenance practices.

What is the underlying age and condition of existing track?

Over the last decade, the Class I railroads have generally shed lower-density lines, selling them to short-line railroads that can operate them more economically.  Railroads have also torn up underutilized yards, eliminated second main line tracks on light density corridors, and abandoned lines that were made redundant through mergers.  This downsizing has yielded a Class I network with increasing traffic density.  All other things being equal, one would expect higher density track to yield economies of scale, lowering track cost per ton-mile of traffic.   

However, as railroads operate heavier axle load rail cars, track component wear rates will get exponentially worse, all other things being equal.  Similarly, higher train speeds (e.g. due to increased intermodal traffic) exponentially amplify dynamic loads, leading to increasing capital track costs per unit of traffic ($/GTM).   Railroads have responded to the increased demands (weight and speed) on their track structure by deploying much higher grades of track materials and inventing modern track inspection and maintenance techniques. 

As a result, tie and rail life can be quite different by railroad or even within a railroad, due to the wide ranges of traffic density, maintenance practices, the quality of materials used, and track geometry (hills & curves).  This analysis is enabled by two key assumptions: 1.     Over the long run, railroads keep the average condition of the track at an acceptable state of good repair.  The close inspection, monitoring, and regulation by the Federal Railroad Administration serves as a lower bound and the capital rationing of railroad investors serves as an upper bound.  The rate of track-caused accidents is a good indicator of track condition.  Accident statistics from the Federal Railroad Administration show a relatively flat to declining rate of track- and signal-caused train accidents over the last 10 years    2.     Maintenance and installation techniques and materials improve over time, so the life of track components becomes longer over time.  Railroads are increasingly replacing wood ties with non-wood (mostly concrete) ties, but wood ties still comprise 92% of all newly installed ties.  This is because most track in the U.S. was originally built with wood ties, and thus wood ties are appropriate for use in replacement.  However, on new lines, wood ties are only about half of the total.  As wood ties have a shorter lifespan than non-wood, we expect to see tie life increasing over time.

What causes track components to wear out?

Ties: Wood ties largely deteriorate because of time (wood rot).  Heavier traffic, wet environmental conditions, severe track geometry, and poor maintenance practices accelerate the natural deterioration process.  Typical wood ties will last between 10 and 40 years depending on the conditions at that location. Non-wood ties (most are concrete, other types are metal and plastic) last 40 to 60 years.  Their lives are much less affected by traffic or maintenance, as long as they are in reasonably well-maintained track.  Derailments can have a catastrophic effect on a patch of concrete ties as they tend to shatter versus wood ties which are more resilient and can absorb the cushion of derailed wheels being drug along the track surface.  

Rail:  Rail wears out because of the accumulated tonnage it supports, not because of rust or any other time-based deterioration.  Consequently, rail life is traditionally expressed in millions of gross tons (MGTs) accumulated.  Typically, today’s new rail lasts around one thousand MGTs, with a range of 0.8-1.3 billion GTs.  Besides tonnage carried, other factors include: 

Rail weight: the heavier the rail cross section, the more tonnage it can carry before wearing out.       

Location: Rail on straight (“tangent”) track wears more evenly and has a longer functional lifespan than rail laid on curves.  A significant component of track maintenance and capital spending is “curve patch” – the replacement of rail on curves. 

The increasing hardness of rail steel reduces rail wear rates.  The increasing cleanliness of rail steel (very clean steel has very few inclusions or air bubbles that initiate cracks or fractures) reduces rail defects.

How do railroads maximize the life of track components?

Ties:  Railroads do not generally re-use wooden ties.  However, if ties are being removed from, say, a yard that is being closed, and the ties have significant life left, the railroad may re-use them in a light–traffic application.  Non-wood ties can be reused, but few of them are ever pulled up from their initial location.  

Rail: In contrast to ties, railroads “cascade” rail to maximize its life in terms of years.  They lay new rail on high-traffic (heavy GT) lines.  After several years, they pick it up and re-lay it on lighter traffic (lower GT) lines. For example, new rail is initially put in track getting 50 MGTs/year.  At the end of 19 years, it has accumulated 50 * 19 MGTs.  After 19 years, it is picked up and re-laid in track getting only 1 MGTs/year.  At that rate, it lasts an additional 50 years, so the total life span of the rail is 69 years.

How do maintenance and operating practices affect the amount of capital spending needed?

Railroads make a trade-off between spending operating expense money to maintain track and capital money to upgrade or install new track.  The life of track components can be extended with appropriate maintenance and operating decisions. Keeping the track bed well-surfaced, straight, and level ensures the track is smooth and firm, reducing dynamic loads from “bumpy rides”. Keeping the track well-drained extends the life of wood ties by reducing rot. Keeping rail well-lubricated reduces the amount of wear per GT carried. Keeping track in gauge reduces the probability of train accidents that in turn damage the track components. Running slower trains reduces the wear-and-tear on the track.

What are the costs of installing ties and rail?

Ties:  Wood ties are installed 18” apart, so 3,580 are required per mile.  Non-wood (primarily concrete) ties can be installed from 24”-30” apart.  At 24”, only 2,640 are required per mile, 26% fewer than the number of wood ties.  However, wood ties are cheaper per unit, so the total initial installation cost of non-wood ties can be 3 times the cost for wood ones.  Wood ties, though, do not last as long and require more maintenance than non-wood.  Under heavy traffic conditions, track with wood ties may need to be re-gauged frequently – perhaps twice a year – whereas track with concrete ties rarely go out of gauge. 

Therefore the economics of wood versus non-wood will depend on traffic density, environmental conditions and maintenance costs over the life of the track, not just the initial cost per unit.  

Rail: There are 2 categories of rail, New and Relay.

New:  Rail purchased new from the factory. The costs to install new rail are Material + OTM + Labor + ballast and surfacing. New rail will last longer than existing rail because inventory of rail in track has lower average weight of rail than new rail. Average hardness of new rail is more than the average hardness of old rail. Maintenance practices are better now than in past.  So existing rail has a history of poorer maintenance and will wear out faster.

Relay:  Rail picked up from existing track and re-installed at a different location. The costs to install new rail are OTM + Labor + ballast and surfacing.

One goal is to isolate and quantify the major drivers of track capital spending.  Our hypothesis is that improvements in technology and maintenance practices, coupled with the economies of scale of increased GTM per track mile more than offset the need to increase capital spending due to traffic growth.  Rather, what railroads and investors perceive as “growth capital” may simply be the symptom of a railroad’s failure to actively manage their physical plant in a way the ensures capital investment  and disinvestment closely maps to changes in traffic volumes; changes that occur from year to year and from corridor to corridor.