POWERING DOWN THE EL -- Reducing Electricity Consumption From Train Idling at the Chicago Transit Authority

POWERING DOWN THE EL

REDUCING ELECTRICITY CONSUMPTION

FROM TRAIN IDLING AT THE

CHICAGO TRANSIT AUTHORITY

Submitted to the Department of Public Policy

in partial fulfillment of the requirements

of the degree of Bachelor of Arts

Presented to: Raymond Lodato (adviser) and Dustin Gourdin (preceptor)

Sonya Dekhtyar

ABSTRACT

The Chicago Transit Authority (CTA), which operates trains on an electrical system utilizing

third rails, has no policies or methodology in place to control the amount of idling that occurs

when a train stands in a yard – that is, when it is not on a run but is using electricity for

temperature control or to minimize engine wear. Consequently, until this research, there has been

no concrete data on the amount of electricity used by these idling trains across the system, even

though the problem of energy waste has been fully acknowledged by the agency.

This paper will bring together data and information from interviews to calculate the total

amount of electricity consumed by idling trains between 2011 and 2013 and the cost of that

idling. It will also conduct a cost-benefit analysis in relation to the private costs, environmental

costs, and public benefits derived from idling in an effort to propose a variety of localized

policies dedicated to minimizing the amount of idling occurring at the CTA every year.

Special thanks to Paul Zielinski, Mark Kokodynsky, Peter Ballard, James Harper, Paras Bhayani,

and Karl Peet at the CTA for all their incredible help, support, and, above all, patience.

TABLE OF CONTENTS

Part I: Introduction ............................................................................................................... 1

Part II: Background ........................................................................................................................ 3

Electricity Sources and Measurement ...................................................................................... 3

Electricity Consumption and Costs .......................................................................................... 4

Electric Trains ......................................................................................................................... 5

Third Rail ................................................................................................................................. 6

Train Idling ............................................................................................................................ 10

Part III: Literature Review ................................................................................................. 12

National Cooperative Transit Research Program .................................................................. 12

Electrification 101 ................................................................................................................. 13

Steering Group Research Programme ................................................................................... 15

Chicago Locomotive Idle Reduction Project .......................................................................... 15

Part IV: Methodology and Data ......................................................................................... 17

Part V: Data Analysis......................................................................................................... 19

Total Traction Power Consumption ....................................................................................... 19

Railcar Availability ................................................................................................................ 22

Revenue Mileage .................................................................................................................... 24

Cost-Benefit Analysis ............................................................................................................. 25

Private Cost ............................................................................................................................... 25

Public Cost ................................................................................................................................ 28

Public Benefit ............................................................................................................................ 31

Part VI: Policy Perspectives............................................................................................... 34

Overview ................................................................................................................................ 34

Policy History ........................................................................................................................ 34

Level of Policy ....................................................................................................................... 35

Recommendations .................................................................................................................. 38

Electricity Caps or Cap-and-Trade ........................................................................................... 38

Increasing Electricity Prices...................................................................................................... 40

Technological Mandates ........................................................................................................... 40

Subsidizing alternative energy resources .................................................................................. 42

Conclusion ......................................................................................................................... 43

References .......................................................................................................................... 44

TABLE OF CONTENTS, CNTD.

Appendices

I: Annual Idling Calculations, by Line ................................................................................... 47

II: Annual Idling Cost Breakdown ......................................................................................... 53

Figures

1: ComEd sources of electricity in Chicago ............................................................................. 3

2: 2011 electricity flow by energy source ................................................................................. 4

3: CTA train shoe ..................................................................................................................... 6

4: Top contact on third rail ...................................................................................................... 7

5: AC v. DC current ................................................................................................................. 7

6: Electrification from substation to third rail ......................................................................... 8

7: CTA substations ................................................................................................................... 9

8: Total annual electrical consumption, by month ................................................................. 19

9: Total annual traction costs, by month ................................................................................ 20

10: Annual cost per kWh, by month ....................................................................................... 21

11: Average monthly railcar availability, by year .................................................................. 22

12: Average monthly railcar availability by line .................................................................... 23

13: Actual monthly revenue mileage by line .......................................................................... 24

14: Sample idling and cost calculation (Blue Line, 2011) ...................................................... 25

15: Per-car idling cost, by year ............................................................................................. 27

1

PART I: INTRODUCTION

The Chicago Transit Authority (CTA), which provides transit to all of Chicago and several

nearby townships, sustains a large fleet of trains which run on electricity supplied by third rails.

More technical information is provided below, but the most concerning discussion related to the

CTA’s rail electrification system has to do with the concept of train idling, which is the excess

usage of electricity in operating a motionless train while it stands in a yard warming up, cooling,

or preparing for a subsequent run. Not only does idling increase electricity expenses to the CTA

itself, but it also, by extension of the process of electricity generation, leads to significant costs to

the environment.

From 2011 to 2013, the years for which the research in this paper was conducted, the

greenhouse gas emissions associated with the amount of electricity consumed through train

idling every year were equivalent to the amount of emissions released by upwards of 6,000

average cars on the road,1 and the cost of idling to the agency each of those years far surpassed

$2 million – almost 8% of the CTA’s entire yearly electric bill. These private and public costs,

when added up over time, indicate a significant problem with the current system of train idling.

The largest attempt to reduce the amount of idling throughout the system came on the heels

of a September 2013 accident between two Blue Line trains, one of which had mysteriously

pulled out of its yard. This spontaneous movement was partially blamed on the fact that the

runaway train had been keyed up for 48 hours prior to the accident. After this accident, the CTA

mandated that all trains be powered off when not in service, but electricity has continued being

used in order to maintain interior air comfort and external equipment temperatures. According to

estimates by Rail Operations managers at the CTA, 90% of all trains in yards were fully keyed

1 Environmental Protection Agency, “Clean Energy: Calculations and References”

2

on 80% of the time prior to the accident, but 80% of yarded trains have been using electricity

90% of time after. Unfortunately, this change did nothing to decrease the total electricity usage

and electrical cost to the agency, but I will use this “80-90 rule” to calculate total consumption

below.

Part II of this paper introduces electricity in relation to the third-rail system employed at the

CTA. Part III reviews some literature related to train idling at large and to electric train idling

specifically. Part IV briefly discusses the methodology employed in collecting data for analysis,

and Part V presents this analysis in detail, with a cost-benefit comparison of public and private

costs and benefits involved in train idling. Part VI recommends some preliminary policies which

could be undertaken in order to curb the extent of the practice.

3

PART II: BACKGROUND

Before delving into a deeper statistical analysis of electricity usage by idling trains at the

CTA, it may be useful to understand some background information about electricity, electric

trains, and the CTA’s rail system specifically.

Electricity Sources and Measurement

Because the CTA’s rail system is entirely dependent on electricity, it is important to first

understand how electrical power is generated. Electric power is usually produced by generators

or batteries, which, in turn, derive energy from coal, nuclear, natural-gas, solar, wind, or biofuel

sources. The vast majority of electricity in Chicago is generated and delivered by ComEd, which

supplies electricity to 3.8 million consumers. According to ComEd’s 2013 Environmental

Disclosure Statement, 44% of Chicago’s electricity used in the year prior to June 30, 2013 was

sourced from coal, 35% from nuclear energy, and 17% from natural gas (Figure 1). All of the

electricity used by CTA trains is sourced from coal power plants.

Sources of Chicago electricity

supplied for 12 months

% of

Total

Coal-fired 44%

Nuclear 35%

Natural Gas-fired 17%

Wind 2%

Hydro 1%

Other 1%

Biomass 0%

Solar 0%

Oil-fired 0%

TOTAL 100%

Figure 1: Sources of electricity supplied for the year ending

June 30, 2013, by the percentage of total each source makes up

(ComEd, “Environmental Disclosure Statement”).

% of Total

4

In Chicago, as in many other cities, the passage of electricity from a power source to a

specific destination consists of a very simple process, the network of which is termed as “the

grid.” ComEd’s system begins at a power plant; transmission lines from these plants extend to

substations, which distribute power to distribution circuits; these circuit lines run along highways

and major roads, serving businesses and smaller communities directly. From there, spur or

service lines carry energy to neighborhoods or individual customers.2

Electrical power is measured in numerous units, but the most standard is the kilowatt hour

(kWh), which is equivalent to one kilowatt of power expended for one hour, and this is the unit

that will be used throughout this paper.

Electricity Consumption and Costs

Of the 4,054 billion kilowatt hours used in the United States in 2012, less than 3% were

utilized by the transportation sector, and even less than that were used specifically by electric

trains. Figure 2 below shows the movement of electricity from different sources to various

industry sectors in 2011.

2 California Public Utilities Commission. “Electric Transmission”

Figure 2: 2011 electricity flow by energy source. Transportation received only 0.23% of the total power

production in the U.S. (EIA, “Annual Energy Review”).

5

According to the U.S. Energy Information Administration, Illinois used 17.89 billion kWh of

energy in 2012, much of it centralized in the Chicagoland region.3 Assuming the ratio of the

area’s electricity usage by the transportation sector is proportional to the national ratio

demonstrated in Figure 2, it can be said that only about 2% of the total regional energy is

utilized by electric trains (and considering that other regional transit systems such as Metra use

predominantly diesel locomotives, this ratio may be even lower).

In 2011, the average retail price of electricity for the transportation sector was 10.46¢ per

kWh nationwide and about 6.5¢ in Chicago, versus an average retail price of diesel of $3.68

nationwide and $3.90 in Chicago.4 There exist numerous arguments about the higher energy

efficiency of electricity (that is, its ability to reduce the amount of energy required to provide a

product or service): by some estimates, the electric power grid has an average efficiency of 92%,

while diesel comes in at roughly half of that.5

Electric Trains

The first electric locomotive was built by chemist Robert Davidson in 1837 in Aberdeen,

Scotland. Nearly 42 years later, inventor Werner von Siemens, by creating the first electric

passenger train in Berlin, showcased the first recorded use of a third rail6 – a rail running parallel

to train tracks with which a train’s “shoes,” small metal blocks or pedals, have constant contact

in order to consistently maintain functional electric circuits.7 Today, both the third-rail system

and electric trains have become a widespread phenomenon in railway systems across the world.

3 California Public Utilities Commission. “Electric Transmission”

4 EIA, “Factors Affecting Electricity Prices”

5 Light Rail Now!, “Electrification 101”

6 Gordon, “The Underground Electric,” vol. 2, p. 156

7 Middleton, “Railroad Standardization,” vol. 27.4, p. 10-11

6

One of the largest such systems encompasses the ‘L’ and subway lines of the CTA, the third

largest system in the United States by mileage and ridership.8

The CTA was formed as an independent governmental agency in 1945, after acquiring

citywide properties formerly owned by the Chicago Rapid Transit Company (CRTC) and the

Chicago Surface Lines.9 The CRTC, founded in 1892, built the city’s first rail lines; four were

operating by the beginning of the 20th

century, and steam-powered locomotives were quickly

being replaced with electric ones. By 1947, the CTA had begun running electric train service

across the city, placing cars on tracks equipped with a third rail.10

Third Rail

As mentioned above, the third-rail electrification system involved in train operations consists

of two major parts: an extra rail parallel to train tracks, called the conductor rail, and a metal

contact block, called a “shoe,” which makes contact with the conductor rail, as Figure 3 below

shows. The contact of the “shoe” with the rail creates a closed circuit with the train tracks,

generating electricity to provide power to the train. In most transit systems using a third rail,

trains automatically switch to battery or diesel power at locations where the third rail is

interrupted (at level crossings, for instance, or crossovers).11

8 www.transitchicago.com, “Facts at a Glance”

9 Ibid

10 Chicago Transit Authority, “President’s 2014 Budget Recommendations,” p. 83

11 Middleton, p. 10-11

Figure 3: A CTA train’s shoe

making contact with a

third rail (Harrison,

www.chicago-l.org).

7

Figure 4 shows a technical view of top contact – the type of third-

rail contact utilized at the CTA – which involves shoes hanging from a

beam at the bottom of a train car making contact with the top surface

of the rail. Although top-contact systems are the most vulnerable to

extreme temperature changes, they are also the cheapest to maintain

and replace, which is the driving reason behind their frequent use.12

A vast majority of third-rail systems supply direct-current (DC)

electric power to trains. Overhead-wire train systems generally employ AC power. DC current

moves unidrectionally from the positive end of a circuit to its negative end (in this case, from the

third rail to the train); AC, oppositely, can move

between ends in either direction and its movement

most often resembles a sine wave in order to

achieve the most efficient transmission of energy

while carrying higher voltages (Figure 5).13

The costs and benefits of third-rail systems are numerous and debatable. Some critics argue

that third rails, especially uncovered ones in top-contact systems, are too exposed, making them

prone to dangerous accidents. (In an extreme case, the Supreme Court of Illinois, in 1992, ruled

the CTA guilty of landowner negligence for failing to prevent a drunken man from urinating on

the third rail, an act which fatally electrocuted him. The agency was fined $1.5 million.14

). Third-

rail systems often require that trains move at slower speeds than possible with alternating current

(AC) or overhead wires in order to avoid blowing rails’ electrical circuits. On the other hand,

besides being cheaper to install than overhead wires, third rails are less visually polluting than

12

Steering Group Research Programme, “Investigating the economics of the 3rd rail DC system,” p. 24-26 13

Sims, The Boy Electrician, p. 40 14

U.S. Supreme Court, Lee v. Chicago Transit Authority, 152 Ill.2d 432, 605 N.E.2d 493

Figure 4:

CTA trains use the top contact style of third rail

(www.railway-

technical.com/etracp.shtml).

Figure 5: The movement of AC vs. DC current

(www.railway-technical.com/etracp.shtml).

8

wires and eliminate the impact of electromagnetic interference (a disturbance affecting an

electric circuit); there are also arguments that wires are just as prone to severe weather conditions

as top-contact conductor rails.15

The CTA’s conductor rails derive all their power from 60 of ComEd’s 1,300 total

substations. These substations convert electric power from the grid and supply it to nearby third

rails in order to electrify trains, as Figure 6 below demonstrates. From a technical standpoint, all

third-rail systems use DC because they can carry roughly 41% more power than an AC system

operating at the same peak voltage.16

Most of the agency’s substations are separate buildings located close to major stops or

roughly two miles from each other. The map on the following page (Figure 7) shows the

location of all the CTA’s substations relative to rail lines.

15

Middleton, p. 12-13 16

Sims, p. 43

Figure 6: Electrification, from power grid to substations to third rail (www.railway-technical.com/etracp.shtml).

9

All substations include two conversion systems; one is essentially a backup for the other in

case of failure. Incoming power is routed through a transformer called the switchgear, which

reduces and converts the power’s electric voltage from a 12,600-volt AC current to the 600-volt

Figure 7: Substations of

the CTA. Map

created with

location data provided by

the agency.

10

DC current supplied to the system’s third rails.17

At the CTA, there is one exception to the usage

of DC power: the new 5000 Series cars convert third-rail DC power to alternating current (AC)

traction in order to decrease the amount of electricity required for movement and to create

smoother rides.18

The CTA is just one of over 90 electric rail systems across the United States, and only one of

27 systems nationwide and one of 142 systems globally which uses a third-rail system,19

but its

extensive history, innovative railcar design, and changing technology makes it a fascinating

organization to study specifically in regards to train idling.

Train Idling

Idling in diesel locomotives has long been known to cause environmental and economic

controversy, especially with regard to fuel usage and air and noise pollution. In a highly

controversial and influential case, the Massachusetts Bay Transportation Authority (MBTA) and

the Massachusetts Bay Commuter Railroad Company were required to pay nearly $3 million in

litigated fees and fines for failing to comply with the Clean Air Act by running diesel engines for

longer than the permitted half-hour.20

According to the settlement report, cutting down diesel

idling time by one hour a day per locomotive at the MBTA, given up-to-date engines and low-

sulphur fuel, could reduce yearly emissions of carbon dioxide by roughly 800 tons, of nitrogen

oxides by 170 tons, of carbon monoxide – by 80, of particulate matter – by about 23, and of

sulphur dioxide by a few tons.21

To put this in financial terms: the annual social cost of one ton

of carbon dioxide is $39,22

of one ton of nitrogen oxide – $16,23

and of one ton of sulphur

17

Barry, “CTA to begin construction on new Red Line substation near Morse” 18

Bombardier, “CTA 5000-Series Rapid Transit Car” 19

Chicago-L, “The CTA Takes Over: Resurrection by Modernization” 20

Environmental Protection Agency, “Clean Air Act Settlement for Commuter Train Idling Violations,” p. 1 21

Environmental Protection Agency, “Clean Air Act Settlement for Commuter Train Idling Violations,” p. 2 22

Environmental Protection Agency, “Social Cost of Carbon”

11

dioxide – $65.24

That is, cutting diesel idling time by one hour a day could save society at least

$32,000 a year in environmental costs per year.

Indeed, the effects of reducing diesel idling are significant and well-documented, not only

because of the politicized aspect of the problem, but also because data is much more easily

tracked technologically and is more readily available to the public. Electric train idling, on the

other hand, is an extremely under-researched topic; very little analysis exists on its trends and

policy implications. Much of the existing research is based in Europe, especially England, but it

largely focuses on the costs and benefits of different systems of electrification for trains rather

than on idling itself, and no analysis has been conducted into the matter at the CTA, either.

23

U.S. Energy Information Administration, “Emissions allowance prices for SO2 and NOX remained low in 2011” 24

Burtraw et al. “Cost-Effective Reduction of NOX Emissions from Electricity Generation,” p.10

12

PART III: LITERATURE REVIEW

As noted, very little research, international or local, has been conducted on electric train

idling. Limited studies have observed issues of electrification of transit systems overall, without

specific focus on idling, and others have discussed the various third-rail systems and their

respective pros and cons. This section will discuss four main texts of interest about railway

electrification broadly and about electricity consumption in transit systems specifically

(excepting one article discussing diesel train idling). The first is a Department of Transportation

report on reducing trains’ electrical demand during rush hour; the second is a series of five

commentaries on the potential of rail electrification; the third is the British Rail Safety and

Standard Board’s (RSSB) engineering research program study on the economics of third-rail

systems in comparison to other electrification systems; and the fourth is a case study on diesel

locomotives in Chicago conducted by the Environmental Protection Agency (EPA). Existing

research on environmental consequences of electric idling will be discussed in Part IV.

National Cooperative Transit Research Program (NCTRP)

The NCTRP, sponsored by the Urban Mass Transportation Administration of the Department

of Transportation, conducts regular studies on the administration, financing, energetics, and

operations of transit systems in America and across the world. In 1983, the group published a

report titled “Reduction of Peak-Power Demand for Electric Rail Transit Systems,” which

analyzed, econometrically, financially, and environmentally, the causes and effects of high

electricity usage by trains during morning and evening rush hours.

The most interesting correlation the report provided was between ambient temperatures and

electricity usage. As mentioned earlier in the introductory section, trains idle largely to sustain

interior air levels and appropriate equipment temperatures for proper functioning and

13

comfortable riding experiences. Conducting regression analyses on the relation between energy

consumption and temperature, the NCTRP determined that the peak average power levels

occurred in the heating and cooling regions, which lie beyond the ambient temperature at which

heating and cooling equipment is fully on, respectively.

Although the cost-benefit analysis provided in the report is slightly outdated in terms of

dollar value, it did provide did provide some insight into load management – “the monitoring,

prediction, and control of power demand.”25

The NCTRP recommends several energy

conservation techniques for peak load demands which can likewise be applied to idling trains.

Two of these are most closely related to the issue at hand: the first is the concept of batch

processing of metering information or regular analysis of electric bills, which, at the moment,

does not exist at the CTA but would be incredibly helpful in monitoring levels of electricity

consumption in rail yards; the second is the regeneration of braking energy – precisely what the

new 5000 Series cars are capable of. To think that this technology was formally proposed as

early as 1983, but not initiated at the CTA (or at any major transit agency in the country, for that

matter) until years later, is indicative of the enormous progress and technological development

yet to be made in energy and electricity conservation in the future. If the CTA were to invest

more in similar railcar capabilities that would build up stores of recycled electricity, the amount

of new electricity required to power trains – and, consequently, the cost of paying for energy

consumption – would decrease over time.

Electrification 101

The Light Rail Now! project, designed to research and improve light rail systems across the

world, conducted a series of five analyses grouped under the name “Electrification 101.” The

25

p. 7

14

first and third commentaries deal with the relationship between transit electrification and global

oil prices; the second discusses the cost-effectiveness of electric rail; the fourth proposes

potential urban-rail solutions to lower national dependence on oil; and the fifth makes

observations about the measurement of rail electricity and its cost.

The relationship of electricity and oil is a well-investigated and long-problematic one. Oil

prices have skyrocketed over the past few years, and the world – and especially the U.S. – does

not appear to be significantly reducing the amount of oil it consumes today. However, some

analyses estimate that America’s oil use will decrease by 10% in the next decade or so, while the

consumption of electricity, though having slowed in 2011 and 2011, is expected to increase by

28% by 2040, with most electrical energy coming from coal plants.26

As oil becomes more

expensive and limited in supply as electricity becomes more widely used and cheaper to produce,

Light Rail Now! argues, a switch to electrified light urban rail is inevitable.27

Moreover, there is quantified research proving that electric rail is faster and cheaper than

diesel power – this latter point not just in the sense of per-gallon versus per-kWh cost, but rather

in the overall cost-benefit sense. Light Rail Now! presents a sample case of the Southeastern

Pennsylvania Transportation Authority, where passenger-rail operating costs are $300 per car-

hour for electric trains and $400 per car-hour for diesel trains. Further, in some transit systems,

electric trains can complete one-hour routes for half the operating cost of a diesel locomotive

running the same route for an hour. In fact, diesel locomotives are so much slower than electric

ones that running diesel trains becomes unprofitable sooner; it has been shown in various diesel-

engine case studies that transit systems with diesel trains can save money by providing fewer

26

U.S. Energy Information Administration, “Growth in electricity use…still increases by 28% from 2011 to 2040.” 27

Light Rail Now!, “Electrification of Transportation as a Response to Peaking of World Oil Production”

15

trains in service, but at the expense of decreasing customer revenue and satisfaction and

becoming weaker in competition with electric trains.28

Steering Group Research Programme

The RSSB research program study is one of the most comprehensive analyses of third-rail

electrification available. The overarching idea behind the study is that DC power in general is not

nearly as efficient as AC power, since trains cannot travel as fast on DC and instances of current

failure are much higher with lower DC voltage. In the Programme’s opinion, all DC third-rail

power should be converted to AC overhead wires.

Whether or not this is the most efficient way of providing better electric service is still under

debate, both in terms of financial means and environmental consciousness. AC current requires a

higher voltage to operate than DC, making it less suitable for idling reduction. However, the

study does appropriately observe that energy efficiency and benefits to electrification system

renewals generally increase dramatically as initial capital spending rises marginally. Finally, the

report concedes that, regardless of whether or not a switch to AC power is made, there is

incredible potential for cost savings in the long run with a stronger awareness of the possible

methods of reducing electricity usage.

Chicago Locomotive Idle Reduction Project

In March 2004, the Environmental Protection Agency (EPA) published a case study

investigating the effects of diesel idling reduction on fuel savings, reduced pollution, and

extended service life. The study presented four major conclusions about diesel idling: diesel

engines are left to idle primarily as a temperature-setting process; most idling occurs in rail

28

Light Rail Now!, “The Cost-Effectiveness of Electric Rail”

16

yards; idle-reduction technology does exist, albeit to a limited extent, in different varieties and

price ranges; but, most importantly, it is difficult to initiate the use of these technologies because

of uncertainty of reduction payback, large initial capital costs, lack of knowledge about these

idle-reduction systems, or, largely, ingrained habits and procedural customs. All four of these

conclusions can be related directly to the electric idling at the CTA, as noted in the policy

recommendations below.

The project proposed a $28,000 “diesel-driven heating system,” a 72-volt alternator which is

capable of powering the electric immersion heater for water in the main engine, charging train

batteries, and powering heaters when the train is shut off. This technology reduced idling time by

nearly a third, polluting emissions by a fifth, and noise by 10 decibels. At the same time, the

study noted several key points regarding this technology’s use: crew compliance in shutting

down idling trains in order to automatically power on the diesel-driven heating system was

variable, reliant largely on past training and focus; several pieces of technology functioning

together are most effective when the shutdown decision is taken away from the train operator;

the most efficient technology is that which allows for easy train restart in the coldest

temperatures of the region; and, finally, the new system could be technologically set up to

provide accurate data collection of idle time, shutdown time, and idling location using the

internet or satellites.

This study is important in regards to electric train idling because similar technology is just as

attainable on electric locomotives. Just as idling diesel engines consume more fuel and produce

more pollution, electric idling consumes significant amounts of electricity which instead could

be regenerated, recycled, or obtained from renewable sources. Studying the various approaches

taken in reducing diesel idling may help better grasp the possibility available to electric trains.

17

PART IV: METHODOLOGY AND DATA

I was granted full access to any and all data I would require by the CTA. Unfortunately, by

general rule and safety procedures, both within the agency and among national transit systems,

trains are never supposed to be left idling between runs or overnight, and no metering system

exists in yards. Therefore, the data I collected was either slightly broad (for instance, traction

electricity usage encompassing both idling and moving trains) or estimated (no hard records of

idling schedules exist, and only approximations could be made based on train schedules and car

availability). However, this data, when combined together with interviews, observations, and the

“80-90 rule,” was accurate enough to give a good picture of the excess electricity usage by idling

trains, to estimate a dollar value to lost electricity, and to propose policy changes to the practice.

Besides amassing a variety of contractual documents and agreements which provided insight

to yearly trends and cost variations, I collected geographical information on the location of

substations, the maps of which are included above, and yards. Also provided were spreadsheets

documenting total traction power, electrical consumption, and costs by line for each month

between January 2011 and December 2013. I only went back to 2011 because the CTA changed

its electrical supplier policy in 2013, which allowed the agency to purchase electricity from

certified Retail Electric Suppliers (RES) other than ComEd29

– that is, for years before this

change, the electrical costs per kWh were approximately the same and three years of detailed

data sufficed. No complete data had been collected for 2014 to date, so I did not include this year

in my analysis. Further, I received datasets of railcar availability by line for each day of the year

in the same time frame, which I compiled into one larger dataset to reflect trends and averages by

month and line, as well as monthly records of actual revenue mileage (that is, the total mileage

29

Chicago Transit Authority, “Ordinance No. 013-48”

18

run during service on a particular line) and train schedules, which I used to calculate the average

amount of monthly time a car spends in a yard. I was likewise able to conduct interviews with

various representatives of departments related to train idling – Rail Operations, Rail

Maintenance, Energy, Engineering, Performance Management, Accounting, and Scheduling.

Finally, I gathered data on the social costs of electricity from an environmental standpoint.

Because electricity generation requires the use of coal and thus creates carbon dioxide, sulphur

dioxide, nitrogen oxide, and other toxic emissions as byproducts, I attempted to quantify, as

accurately as possible, the effects of coal-generated electricity use on the environment in order to

determine what the true cost of the CTA’s electricity consumption due to train idling is.

19

PART V: DATA ANALYSIS

Total Traction Power Consumption

The first trend analysis I conducted concerned the total kWh usage and traction costs across

the entire rail system. It is important to recognize that total electricity costs within any transit

agency consist of the sum of traction and non-traction components: the traction costs, by

definition, refer to all electricity usage related to rail electrification; non-traction costs include

fixed costs such as facility charges and all other electrical bills. Although the traction dataset I

received contains both idling and running trains, it provides a very clear overview of overall

monthly, seasonal, and year-over-year trends of electricity consumption.

Figure 8 shows very clear monthly (and, by extension, seasonal) patterns. January, July,

August, and December have the highest electrical consumption rates (as demarcated in red) –

most likely an effect of the need to sustain comfortable train-car temperatures in the winter (both

Figure 8: Year-over-year total electrical consumption, in kWh, by month.

20

interior, to keep cars warm, and exterior, to keep equipment from freezing over) and summer

(largely interior, to keep cars cool and fans functional) – with 2013 recording three-year highs of

electricity usage. April and October have the lowest kWh totals each year (represented in green),

most likely because the need for excess heating or cooling is lost during the spring and fall, and

because train service is generally cut back slightly during those seasons, due to the decreased

necessity to keep riders out of extreme heat or cold. 2011 has the overall lowest rates of kWh

usage of the three years, but 2013 saw a greater-than-average dip – and a three-year low – in

electricity consumption in October. That month was the one immediately following the runaway

train accident, after which all train idling was curbed, and thus its electricity consumption levels

were at a record low.

Surprisingly, even though 2011 had the lowest rates of kWh totals, it had the highest total

traction costs of the three years under consideration, as Figure 9 shows.

Figure 9: Year-over-year total traction costs, in dollars, by month.

21

The same months each year had the highest and lowest costs as did those with the highest

and lowest electricity consumption, but the reason for the varying costs year-over-year lies

largely in contractual agreements between the CTA and Exelon. Each year, the CTA draws up a

new contract with the company, which modifies the price levels for electricity for that year. The

price that is settled on reflects predicted electrical generation and delivery costs by Exelon and

expected electricity consumption by the CTA. In 2011, the contractual price per kWh was $0.07,

but in the two subsequent years dropped to $0.06, due to new agreements with RES electrical

suppliers. A clearer way to see the fluctuations in total costs as a function of electricity

consumption is to graph the price of each consumed kWh per month (Figure 10). Similar

monthly and seasonal patterns present themselves in this metric as before.

Figure 10: Year-over-year cost per kWh, by month.

22

Railcar Availability

I was given daily car availability data for each weekday between January 1, 2011 and

December 31, 2013 (unfortunately, no data is recorded during weekends) and compiled the

values in calendar form across the years in order to find any patterns in the average numbers of

cars available by month or train line. In order to eventually be able to estimate the overall

electricity consumption by idling trains, and because that consumption relies directly on the

number of cars in service at any given time, it was important for me to distinguish trends in

availability in order to properly estimate the amount of time an average car would spend idling.

Figure 11 shows the average car availability by month across the three years observed. There

are obvious differences between the three years – but these are all explained by one simple

explanation: the retirement of the old 2200 Series cars and the influx of new 5000 Series cars.

The 2200 Series, put into service in 1970, predominantly on the Blue Line, was fully retired on

Figure 11: Average availability of railcars from 2011 to 2013, by month.

23

July 31, 2013, but the process of removing these cars from the system began in late 2010.30

In

November 2011, the 5000 Series entered service on the Pink Line; hence the visible spike of

availability at the end of that year. The series was temporarily pulled out of service in the first

third of 2012 due to manufacturing issues, but by May 2012, these cars were back in the system

and the availability increased once again. By 2013, the Green and Pink Lines had been fully

outfitted with the new cars, and the Red Line was nearly equally comprised of new and older

cars. By the end of 2013, the CTA had received and placed in service 380 5000-Series cars.31

It is not enough, however, to only observe the monthly and yearly patterns of overall car

availability. There are not equal numbers of railcars assigned to each line on any given day;

therefore, it would not be enough to take an overall system average as the baseline for idling

calculations. The figure below, Figure 12, shows the average number of railcars available each

month in the observed time frame by line (the Yellow Line is not represented here because it

only runs a maximum of 6 cars at any given point).

30

Chicago-L, “2200 Series Cars” 31

Chicago-L. “5000-Series Cars”

Figure 12: Average railcar availability by line for each month between 2011 and 2013.

.

24

It is clear that the Red and Blue Lines have the highest car availability, which is expected,

considering that these two lines are the longest and busiest. The Pink-Line fluctuations in 2012

are due to the influx of 5000-Series cars during their testing phase; the same is true about the

Green Line in late 2012 and early 2013. The Red line peak in early 2013 is likewise due to the

addition of the new cars, while the Blue-Line drop in mid-2012 is a result of the retirement of the

2200 Series.

Revenue Mileage

Because there is no technology to track how long trains stand in yards, nor whether or not

they are idling, I had to employ a sort of backward induction methodology in my final

calculations. The formulation used hinges largely on the revenue miles run by each line every

month. The CTA records the scheduled and actual mileage each month in order to track by how

much trains are being overused in service. I used only the actual revenue mileage in an effort to

determine how much time the train was not in service. Figure 13, below, shows the trends in

average revenue mileage, by line and month, across the three years. (Again, the Yellow Line is

excluded here, because its revenue mileage is miniscule compared to the other lines.)

Figure 13: Actual revenue mileage by line for each month between 2011 and 2013. .

25

Cost-Benefit Analysis

Private Cost

In order to reconcile gathered data with a lack of more detailed idling information, I created a

formula to calculate the potential idling time of an average car, depending on what line it served,

based on the maximum amount of time it could stand in a yard. It is important to remember that

these calculations reflect only weekdays – no data is recorded for weekends; therefore, my final

cost and consumption calculations are, in reality, lower bounds of the total possible amount each

month. Figure 14, below, shows a sample of the complete formulation of idling time, energy

consumption, and cost – in this case, for the Blue Line in 2011. (Appendix I reproduces the full

set of calculations, by year and line.)

The first four rows contain purely empirical data, with no calculations. Row 1 lists the

average number of railcars available each month (as demonstrated in Figure 12 above). Row 2

notes the total track mileage of each line – in both directions, as each car is expected to make a

round-trip from and to the same end terminal – as determined by information provided by the

CTA. These values remain largely consistent per line across the three years, with the exception

of the Green and Red Line in 2013, when the Red Line South project shifted the two lines’

service areas between May and October due to construction. Row 3 records the total two-way

Figure 14: Idling time, electricity consumption, and cost calculation for the Blue Line in 2011. .

26

trip length, in minutes, as determined by CTA train schedules. Row 4 contains the actual revenue

mileage per line and month (from Figure 13).

The following rows contain the steps necessary to calculate final consumption and costs.

Row 5 finds the average actual trips per car by dividing the actual revenue mileage from the

preceding row by the track mileage, and then dividing that value by the monthly average number

of cars. Row 6 multiplies the value from the previous row by trip length in order to determine the

average amount of time each car spends in service. Row 7 reflects the total number of possible

minutes allotted to each car’s service each month; this value depends on each line’s run schedule

and the number of weekdays in each month, and is therefore different for each line and month.

Row 8 finds the difference between the previous two rows in order to determine the time each

car is not running, under the simplistic assumption that if a car is not in service, it is parked in a

yard (in reality, this is not always true, as cars are sometimes run between lines or terminals,

depending on service needs; however, for the purposes of my model, I will assume that any car

not running in scheduled service is not running at all).

The following two rows use the “80-90 rule.” Row 9 calculates the minutes the average car

spends idling by taking into account the total non-run time from the previous row and the

percentage of time trains are estimated to be idling; thus, prior to October 2013, the multiple for

this calculation was 0.9, and after – 0.8. Row 10 shows the average total idling time possible per

car, given the assumed estimate of the percentage of cars idling at any given moment – that is,

essentially, the probability that the average yarded car is idling; once again, pre-October 2013,

the estimate translates to a 0.8 multiple, and after – to 0.9.

Finally, Row 11 and 12 calculate the total consumption and cost, on a per-car basis, of idling.

The Rail Operations and Engineering departments at the CTA calculate that the requirements for

27

railcar power are 53 kWh with air conditioning, 60 kWh with heat, and an additional 20 kWh for

full power (i.e. when the train is fully keyed up). Heat is used, roughly, between October and

March; air conditioning is used the remainder of the year. Row 11 takes these values into

account; after the runaway-train accident, the full power component drops and only the heat

electricity usage remains (for the last three months of 2013). Row 12, lastly, multiplies the total

kWh consumption by the retail price of electricity for the year – $0.07 in 2011 and $0.06 the

following two years.

Figure 15 shows the per-car cost of idling over the three years analyzed, by line:

The following table provides the total yearly electrical consumption and cost of idling,

calculated by multiplying the per-cost consumption and cost by the total number of cars available

at the end of each year (a full cost breakdown can be found in Appendix II):

YEAR KWH CONSUMPTION TOTAL COST

2011 38,487,712.7 $2,694,139.89

2012 38,069,296.7 $2,284,157.80

2013 42,251,369.7 $2,535,082.18

Figure 15: Per-car cost of idling, summed up by year. .

28

The total consumption follows a clear pattern: after a small dip in 2012, the number of kWh

consumed jumped by almost four million in 2013. The primary reason for the rapid spike of

electricity consumption due to idling in 2013 is the harsh winter that struck in the last quarter of

the year. Even though trains were no longer allowed to be keyed up following the runaway train

accident in September of that year, Rail Operations managers at the CTA confirm that the vast

majority of the trains were idling consistently in order to sustain warm temperature inside trains

and to maintain exterior equipment temperatures to avoid freezing and malfunction. (Because

there was no accurate modified estimate of idling trains for the winter of 2013, I used the “80-90

rule” in my formulation through the end of the year, but the total consumption was likely even

higher.) The only reason the total 2013 electrical cost is slightly lower than the 2011 price is

because each kWh cost only six cents in 2013, compared to seven cents in 2011.

These costs represent only the private costs to the agency and comprise between 7% and 8%

of the CTA’s total electrical bill each year – a substantial amount. And yet it would be inexact to

only calculate the private costs to the agency of idling: it is also important to consider the public

environmental costs of the practice. As noted in the first section of this paper, the CTA derives

its entire electricity supply from coal-powered plants, which implies, by extension, that the

tremendous excessive consumption of energy by idling trains also requires the excessive

generation of electricity by ComEd. The generation of electricity from coal has been a

contentious and well-analyzed pollution issue in recent years, so the following section will

investigate the environmental costs associated with this electrical waste in more detail.

Public Cost

The generation of electricity from coal affects the environment in three major ways: it

releases greenhouse gases and harmful particulate matter from power plants into the air; it uses

29

up excess water for steam and cooling processes; and, less importantly to the discussion at hand,

excess demand for coal-powered electricity may lead to the construction of additional plants,

which take up land that could be used for other purposes. I will not focus on negative individual

health consequences of pollution, because that is an entirely different topic of research; however,

it is important to remember that health expenses due to pollution-induced lung or heart disease,

for example, are also incredibly expensive, and thus my costing below could be considered only

a lower bound of the total environmental impact of excess electricity usage. I will likewise not

discuss the broader effects of coal’s impact on global warming, due to the enormous breadth of

the subject; nonetheless, as with health impacts, it is crucial to remember that, if my costing were

to include the overall social cost of global warming, the final price would be much higher.

Numerous studies, by government agencies and private organizations alike, have created a

range of average estimates of how much water is consumed and greenhouse gas is emitted per

kWh of electricity used. The most detailed report by far, “Full Cost Accounting for the Lifecycle

of Coal,” was published in 2011 by a group of scientists from some of the nation’s most elite

medical, health, and environmental schools, led by Harvard Medical School affiliate Paul

Epstein. I will use data provided in this report, as well as some external information when

necessary, and my above findings on the electricity consumption of the CTA’s idling trains to

monetize the environmental impact of that excess energy use.

The first step in computing the total environmental cost of the CTA’s idling consumption is

to attach prices to each byproduct of coal-generated electricity. I used the City of Chicago’s

sewer rates for the three years at hand in order to determine the public cost of the water

consumed in the process of electricity production at a coal plant. For the cost of emissions, I

turned to the Epstein report, focusing on the three major gases – carbon dioxide (CO2), sulphur

30

dioxide (SO2), and nitrogen oxide (NOx). Because the report was published in 2011, I also used

statistics provided by the Environmental Protection Agency to supplement cost estimates for

2012 and 2013.32

The chart below shows the social costing of water and each of the three

gaseous emissions:

YEAR H2O (PER 1,000 GAL.) CO2 (PER TON) SO2 (PER TON) NOX (PER TON)

2011 $2.00 $13.85 $2.12 $15.89

2012 $2.51 $24.97 $23.84 $16.34

2013 $2.88 $39.01 $64.97 $18.63

The next step is to find out how much each of these four byproducts is used relative to a

certain level of electricity consumption. Here again the Epstein report proves useful. The

following table records the consumption or emission levels for each product, based on 2011 data:

H2O (GAL/KWH) CO2 (TON/KWH) SO2 (TON/KWH) NOx (TON/KWH)

0.461029 0.000994 0.000005 0.000002

The final step is to calculate the total social cost of the CTA’s electricity consumption due to

idling by combining the data in the previous two charts and the year-by-year agency

consumption, as calculated in the prior section. Thus the total environmental cost of train idling

at the CTA, given this model, is:

YEAR H2O CO2 SO2 NOx TOTAL

2011 $26,279.84 $392,375.95 $313.62 $973.21 $419,942.61

2012 $32,604.44 $699,328.87 $3,486.45 $989.33 $736,409.09

2013 $45,054.36 $1,315,770.47 $11,442.78 $1,358.45 $1,373,626.05

TOTAL $103,938.64 $2,407,475.28 $15,242.85 $3,320.99 $2,529,977.76

Again, it is important to remember that this cost is not all-encompassing; for instance, it does

not account for health costs related to air pollution, fatalities, loss of tourism from coal plant

construction, and the grander social cost of global warming. According to the Epstein report,

32

Environmental Protection Agency, “Clean Air Act Settlement for Commuter Train Idling Violations,” p. 1-2 & “Social Cost of Carbon”

31

which includes all these additional costs and more, the average cost of one coal-generated kWh

is 17.8ȼ. If I were to apply this base value to the CTA’s idling energy consumption, the total cost

would rise drastically:

YEAR KWH CONSUMPTION TOTAL COST

2011 38,487,712.7 $6,850,812.86

2012 38,069,296.7 $6,776,334.81

2013 42,251,369.7 $7,520,743.81

Nonetheless, even $2.6 million over the course of three years is a considerable environmental

cost. When added to the three-year total electrical cost of idling, the total public and private costs

sum to more than $10 million – an enormous expense.

Public Benefit

Considering the large electrical cost of idling to the CTA, there exist practically no private

benefits to the agency for continuing the practice. Likewise, relative to the combined private and

public costs of idling, as calculated above, the public benefit from the practice is almost entirely

negligible. Nonetheless, it is worth a closer analysis.

Many studies have been dedicated to various benefits the public derives from transit. While

some have focused on the elasticity of transit demand relative to service, others have focused on

the effect of weather on ridership. The common consensus is that fare increases and weather

affect ridership levels, but that it is extraordinarily difficult to pinpoint consistent relationships

between ridership and various variables, not least because consumers of public transit in large

metropolitan areas like Chicago are largely split between transit-dependent riders and choice

riders; the first category has a very low elasticity for service change (that is, these customers will

use public transit no matter what minimal changes in fares, weather, or service quality there may

be), but the second category will have a much higher elasticity.

32

In the case of idling trains, the only benefit the public would derive would be from the

interior air temperature sustained by the train upon service. Personally, any rider would much

rather prefer an air-conditioned car in summer over a stiflingly hot one and a heated car in winter

instead of a freezing one, but the vast majority of CTA customers are transit-dependent riders

who are reliant on transit regardless of railcars’ interior temperatures. Besides, any train keyed

up for service automatically operates a fan or a heater (depending on the season), so trains in

service would eventually acquire an optimal temperature throughout their runs. Thus, even

though an individual may prefer to ride in a car at optimal temperature, he will likely use the

train regardless, and the interior temperature will eventually reach optimality without having

been sustained for long hours prior to a run.

The second part of the public benefit discussion consists of the use of idling to maintain

equipment temperatures in extreme weather. From my discussions with CTA personnel, this

seemed like a secondary objective relative to the sustenance of air comfort levels, but it is

important to note that, in the extreme case of weather affecting the ability of trains to run due to

equipment malfunction, the public will be negatively affected by scheduling delays and

overcrowded trains. Thus, in this instance, it can be said the public also derives benefit from

having properly-functioning trains, and if idling helps ensure that equipment is well-maintained,

the benefit truly is important. But the same CTA personnel who observe the need for functioning

equipment regulated by proper temperatures also note that it is, realistically, completely

unnecessary to keep trains idling for as long as they do now in order to ensure proper equipment

functioning. Thus, the benefit to the public from keeping trains idling for this particular purpose

is largely nonexistent.

33

Lastly, it is important to consider that whatever public benefit may be derived from train

idling, it is limited exclusively to those who actually use trains. Choice riders or people who

rarely, if ever ride public transit, have no preference about interior or exterior temperatures, and

therefore would not derive any public cost of benefit from the absence or presence of idling

either way.

Put together, the potential benefit of train idling to the riding public is very minimal. Both

from a personal and a technical standpoint, idling does little to appease any sort of urgency for

service quality or quantity, and therefore one can assume the public benefit from idling to be

nonexistent.

It is thus evident that the costs of train idling far exceed the benefits; in economic terms,

then, this process is very inefficient and should not be continued.

34

PART VI: POLICY PERSPECTIVES

Overview

As the data and analysis above have shown, the costs of train idling in yards, both to the CTA

and to the environment at large, are significant. Undoubtedly, these costs are not unique to the

CTA – train idling is common practice in all transit agencies, and electric idling is equally costly

to all systems using third rail. The above cost findings offer a direct response to current policy

debates over energy efficiency standards for transportation. Transit policies have, historically,

entirely omitted electric train idling from efficiency calculations, and there may be some debate

over the importance of such policies and the difficulty of their implementation. Thus, although

the issue at hand concerns transit systems across the nation, it may be difficult to propose a

policy which will appropriately balance the needs of the agency, the environment, and the

consumer in order to attain an optimal quantity of idling.

Policy History

To date, there has been no policy on electric train idling. Much more governmental and

public intervention has been conducted on the matter of diesel-engine train idling, both as federal

and local policy and as grassroots, community-led petitions and legislative appeals. The reason

behind the general absence of electric train idling policy is, largely, a lack of previous data

analysis and understanding of the subject. Unfortunately, as shown above, the public benefits of

electric train idling are hard to quantify, and the idea of “keeping the public happy” – that is,

keeping the riding experience comfortable – often prevails over the consideration of the costs of

idling in environmental and financial terms. On the other hand, diesel train idling effects on

35

noise and air pollution are very quantifiable, and thus its policy background is much more

developed.

Still, it is important to remember that the costs of electric train idling are high compared to

what they could potentially be if various technological and financial measures were undertaken.

Such possibilities include, but are not limited to, mandates for updated digital tracking of time

spent idling or financial caps on the amount of electrical cost due to idling allowable for a certain

time period. Before considering these policies in detail, the question of which level of

government or authority they should arise from should be addressed.

Level of Policy

Approaching the question of potential policies concerning controlling, limiting, or changing

electricity usage for third-rail systems requires a careful consideration of the various levels at

which such regulations can be enforced. In this particular case, there exist three possible levels of

policy initiatives: federal, local, and agency-specific. National policies would imply one set of

regulations for all third-rail systems; local policies would provide more flexibility to regional

lawmakers in order to account for variability in service area and electrical costs, for example;

and firm-specific policies would be the least rigid, allowing each individual transit agency to

control its electricity-supply system use and cost completely independently of various policies or

other firms’ regulations.

Historically, national environmental policy has never been too successful. Efforts to regulate

a whole country’s environmental sustainability have in the past almost always been crushed due

to a separation of ideas and expectations. Beginning in the 1970s, civic environmentalism, a

movement led by people and localities rather the federal government, became much more

prevalent, and statewide policies accomplished much more on the environmental front, due to

36

more centralized (and subsidized) commitments.33

This approach is clearly evident in the

copious lawsuits and petitions brought by residents of communities where diesel trains continue

to present pollution problems. Although there may be some benefit in sustaining a federal

environmental policy – for instance, it can set regulations on issues that affect all citizens

equally, such as air pollution – it is evident that there are often too many variables to properly

control it. Because the prices of electricity supply per kWh and the sizes of transit agencies

employing third rail vary greatly across the nation,34

it will be extremely difficult to legislate any

sort of national policy controlling the amount of third-rail electricity usage across all transit

agencies nationwide. Interestingly, no national policies of the same sort exist in foreign

countries, likely for similar reasons (it is hard to imagine the numerous local and regional transit

systems in the United Kingdom or France, for instance, submitting to electricity-consumption

control by the countries’ respective federal governments).

On the other hand, localized policies have the potential of being more effective in this

particular case. Of course, even state- or city-wide policies have limits to their functionality: it is

often hard to attain perfect cooperation among all levels of authority passing the regulations, or

specific goals may be prioritized differently between various groups (in this case, between the

CTA and ComEd, for example). In this regard, finalizing and enforcing local environmental

policies has always been slightly problematic.35

However, when considering the specific issue of

electricity usage in third-rail systems, a localized policy is much more viable than a nationwide

one, largely due to the location-specific effects of transit agencies. Because the high costs of

train idling are undeniable, potential policies requiring all third-rail systems to limit or control

the amount of electricity lost to idling should not be expected to face much conflict.

33

DeWitt, “New Directions for Environmental Policy and Politics,” p. 7-8 34

Marketplace , “The Price of Electricity” 35

DeWitt, “An Overview of State Environmental Policy,” p. 52

37

The amount of social benefit decreases if the policies in questions become too agency-

specific: if individual transit systems using third rail are given too much flexibility to regulate

their own policies concerning electricity, there will likely be a lack of attempt to decrease

electricity waste in an effort to preserve riding comfort and bypass expenses on additional

technology or limits controlling the amount of idling. Nonetheless, if policies related to train

idling are to be regulated on an agency-by-agency basis, there must be some overarching system

of controls in place in order to maintain certain standards for electricity consumption by idling

trains, depending on location-specific ambient temperatures, ridership, and train-car and -service

differences. If each transit agency were to have complete leeway over train idling schedules and

electricity consumption without any regulation, it would become too difficult to track the

benefits and costs of related policies.

In short, the most plausible policy recommendations would occur on a localized level – either

within a certain region (or regions similar to each other in terms of, for example, yearly

temperature fluctuations) or among agencies similar to each other in size and service – and

would largely concern small-scale technological and regulatory improvements mandated by

maximum limits to electrical consumption or cost due to idling. Any possible policy solutions

should be minimally costly, should require low levels of investment from as few related parties

as possible, and should be technologically simple.

Unfortunately, even given this straightforward approach to potential policy, it is very difficult

to predict the response from involved transit agencies. First of all, most agencies, like the CTA,

have a very long-established, tight culture among workers and with the system as a whole, and

any change would be very hard to promote instantaneously. Departments often try to appear self-

sustaining without requesting help, and there have historically been some tensions between

38

vehicular and financial or data-driven divisions. Thus, recommended policies will not only have

to control electricity usage by idling trains, but will also have to be put forth in a way that

encourages the agency culture to mold itself to change and to communicate better internally.

Second of all, a vast majority of transit agencies, especially large metropolitan ones, are

operated by employees who have worked their way up from low-level jobs within the agency

(such as bus drivers) – that is, workers who are no longer young and as amenable to

technological change as younger tech-savvy generations. This is certainly the case at the CTA,

where all but two departmental and terminal managers are over the age of 45. The digital

systems in place for tracking vehicle location, mileage, and schedules are often complicated to

the point that managers prefer not to use them, resorting instead to handwritten notes. In

proposing various policies mandating updated technologies for the tracking of idling and

electricity consumption, then, it will be important to ensure that the agency can move past the

technological gap but work in sync to solve the collective action problem of reducing electrical

consumption due to idling.

Recommendations

Electricity Caps or Cap-and-Trade

The consumption of electricity by idling trains, as noted previously, is not only expensive to

the host agency, but is also harmful to the environment; thus, it may be feasible to institute a

series of caps on electricity consumption by idling trains. The extreme of this policy would be to

completely ban all idling (that is, to impose a maximum cap of zero kWh consumed by idling

trains), but since there is still a minimal public benefit to train idling, a total ban would not be

economically optimal or efficient. Because this type of policy would only be focusing on third-

rail systems, and because these systems differ significantly in location, service, and ridership, it

39

would be impossible to sustain a policy impacting all related agencies equally – not to mention a

federal policy capping electricity consumption for all transit agencies in general.

Instead, a policy of this type could be implemented in one of two ways. The first would be as

a cap and trade, whereby third-rail systems could be granted a certain yearly amount of kWh

consumption, beyond which they would be fined, but would retain the option of trading units of

this allotted electricity with each other, depending on each system’s electrical demands. The

second would be a more agency-specific – or, at minimum, a region-specific – policy which

would cap electricity usage by an agency utilizing a third-rail system to a certain number,

determined by average traction usage by non-idling trains, of kWh per rail car or per hour, day,

or month. This would allow agencies to distribute their allotted electricity according to the

number of railcars available, system schedules, ambient temperatures while maintaining

adherence to imposed electricity limits.

A maximum cap or cap-and-trade policy would be efficient, considering that similar systems

already exist across the nation for controlling greenhouse gas emissions. Illinois, for one, does

not cap such emissions, although non-binding targets do exist. The state also has no cap-and-

trade system for gas emissions – for the time being, at least, although, as a member of the

Midwestern Regional Greenhouse Gas Reduction Accord between six regional governors and a

Canadian premier, a trade scheme is eventually expected to be enacted.36

Some arguments for the insufficiency of cap-and-trade for toxic gas emissions observe that

the relative prices of these emissions are currently too low, from a market perspective, to initiate

a major reduction in levels. Instead, some experts note that a carbon-emission portfolio standard,

which would gradually reduce the amount of gases like carbon dioxide emitted per kWh of

36

Institute for Energy Research, “Illinois Energy Facts,” p. 2

40

electricity, could be sold by utility companies to end users directly.37

Nonetheless, many

economists and policymakers agree that cap-and-trade serves as a functional first step in

environmental policy implementation, aiding in the regulation of excess consumption and

emissions of resources or harmful chemicals.

Increasing electricity prices

An alternative to capping electricity use could be to simply increase electricity prices in order

to encourage implementing some method of decreasing energy consumption due to idling. This

would be a localized policy, largely drawn out between each transit agency and its electrical

supplier, but it would likely only work in the short term – if a supplier keeps pushing prices

upwards, the agency may eventually seek out a new supplier. Taking the CTA as a case study

and raising ComEd prices from six or seven cents per kWh to ten or twelve cents, it can be

predicted that the agency will choose to cut back on electricity consumption where possible,

including decreasing idling. However, to properly identify and curb sources of excess electricity

consumption, an agency would require some sort of digitalized method of tracking electricity in

real time, and the next proposed policy discusses this issue.

Technological Mandates

As mentioned previously, technological mandates will be hard to implement, not least

because the culture and average age of those most closely involved with train idling precludes

rapid transitions to advanced digitalization. In this regard, any policy pertaining to advanced

technological systems of tracking idling will also have to include some element of necessary

training for those employees who will in any way be utilizing the new technology.

37

Samaras et al., “Cap and Trade is Not Enough,” p. 1, 3-4

41

Because most third-rail transit systems across the county have no standardized digital method

of recording how long trains stand in yards and whether or not they are idling, this policy could

be attained on a federal level: directed exclusively at transit agencies operating third-rail systems,

it could mandate the use of a single technological platform at each concerned agency in order to

normalize any gathered data across the country. This platform could be two-sided: on the one

hand, it could digitally track and record how long a particular railcar stands in a yard and how

many kWh are used by each car during idling, as well as include tracking for outside variables,

such as temperature, in order to aid in the correlation of data; on the other hand, it could require

all railcars to include an on-board digital tracker which would record how long prior to service

they are turned on to idle and could be capable of automatically shutting off idling trains after a

certain amount of time or, oppositely, powering them up remotely a specified amount of time

prior to service.

This digital system could be mandated for all related agencies, and a penalty for not adhering

to the technology could be imposed. This penalty would have to be high enough to incentivize

transit agencies to make the transition to the new technology. In addition, mandatory training

sessions could be required of all employees involved in any way with tracking electricity

consumption by idling trains in order to ensure that no technological gap exists from the outset.

Disciplinary action could be prescribed to all those who do not partake in training.

Arguably, the setup costs to a national platform such as this one would be sizeable. But

considering the cost of electricity consumption from idling at the CTA alone – approaching $3

million annually to the agency (not to mention the additional environmental costs) – the

technological startup costs would quickly pay themselves out. The most large-scale technological

database platform currently in place at the CTA costs, by various estimates, between $6 and $9

42

million to implement. The system proposed for tracking idling could cost about as much or less

for an agency the size of the CTA (and considerably less for smaller agencies) – which means

that the startup costs could be paid out, in terms of saved electrical costs, within just a few years.

Subsidizing alternative energy resources

In today’s global environmental move toward alternative energy resources, even the

seemingly-small costs to the environment of train idling could benefit from more efficient

energy-generation techniques. A federal subsidy to agencies using third-rail systems, based on

the system’s size and service capacity, could be allocated to the specific purpose of adding more

energy-efficient sources of electricity supply, such as solar panels, wind turbines, or cars with

regenerative braking (such as the 5000 Series at the CTA). The cost of solar panels has

plummeted over the past few years; their proper placement throughout a transit agency’s service

area could help lower electrical costs overall by replacing electricity consumed through

substations and third rails with electricity supplied through solar power. Industrial wind turbines

are still incredibly expensive, but subsidizing small-scale turbines in one or two key locations

within a transit system could also help reduce electrical costs over time. Lastly, subsidizing the

production of more cars using regenerative braking, like the current 5000 Series cars, would help

transit systems store and recycle more energy over time, thus driving down electrical demand

and costs due to idling.

43

CONCLUSION

In this paper, I have tried to calculate, as accurately as possible, the significant private and

public costs of electric train idling at the Chicago Transit Authority in an effort to propose

potential policies which could serve similar agencies at large in decreasing the amount of

electricity consumed through idling. Although the number of transit agencies nationwide using

third rail is not large, this study of costs to the CTA itself and to the environment as a whole

demonstrates that a reduction in energy consumption due to idling can be – and should be –

easily reduced.

44

REFERENCES

Barry, Patrick. “CTA to begin construction on new Red Line substation near Morse.” 16 Jan

2013. CTA Station Watch. 30 Oct 2013. <ctastationwatch.com/main-line/636>.

Burtraw, Dallas et al. “Cost-Effective Reduction of NOX Emissions from Electricity

Generation.” Washington, D.C.: Resources for the Future, July 2001.

California Public Utilities Commission. “Electric Transmission.” San Francisco, CA. 2009.

Chicago-L. “The CTA Takes Over: Resurrection by Modernization.” Chicago-L.org. 24 Oct

2013. <www.chicago-l.org/history/CTA2.html>.

Chicago-L. “2200 Series Cars.” Chicago-L.org. 5 Apr 2014. <www.chicago-

l.org/trains/roster/2200.html>.

Chicago-L. “5000 Series Cars.” Chicago-L.org. 5 Apr 2014. <www.chicago-

l.org/trains/roster/5000.html>.

Chicago Transit Authority. “Facts at a Glance.” 2013. Chicago Transit Authority. 23 Oct 2013.

<www.transitchicago.com/about/facts.aspx>.

Chicago Transit Authority. Modernizing Transit for the Future: President’s 2014 Budget

Recommendations. Chicago, IL. October 2013.

Chicago Transit Authority “Ordinance No. 013-48.” Document Library. 19 Apr 2013.

Ordinances. 25 Mar 2014. <www.transitchicago.com/assets/1/ordinances/013-

048Electricity_Contract.pdf>. p. 1.

City of Chicago. “Water Sewer Rates.” Service. 2010. Water Management. 16 Apr 2014.

<www.cityofchicago.org/city/en/depts/water/provdrs/cust_serv/svcs/know_my_water_sewerr

ates.html>.

Commonwealth Edison. “Environmental Disclosure Statement.” Chicago, IL. July 2013.

Commonwealth Edison. “Schedule of Rates for Electric Service.” 2 Oct 2013. Electricity. ILL.

C. C. No. 10. 23 Mar 2014. <www.comed.com/Documents/customer-service/rates-

pricing/rates-information/current/Ratebook.pdf>. p. 540.

“CTA 5000-Series Rapid Transit Car.” 2013. Bombardier. 30 Oct 2013.

<www.bombardier.com/en/transportation/projects/project.metro-chicago-usa.html>.

DeWitt, John. “New Directions for Environmental Policy and Politics.” Civic Environmentalism:

Alternatives to regulation in States and Communities, pp. 1-20. Washington, D.C.:

Congressional Quarterly, Inc., 1994

45

DeWitt, John. “An Overview of State Environmental Policy.” Civic Environmentalism:

Alternatives to regulation in States and Communities, pp. 51-83. Washington, D.C.:

Congressional Quarterly, Inc., 1994

“Electric Traction Power Supplies.” 2013. Railway Technical. 29 Oct 2013. <www.railway-

technical.com/etracp.shtml>.

Environmental Protection Agency. “Case Study: Chicago Locomotive Idle Reduction Project.”

Mar 2004. Air and Radiation.

<www.epa.gov/smartway/documents/publications/420r04003.pdf>.

Environmental Protection Agency. “Clean Air Act Settlement With the Massachusetts Bay

Transportation Authority (MBTA) and Massachusetts Bay Commuter Railroad Company

(MBCR) for Commuter Train Idling Violations.” 4 August 2010. Environmental Protection

Agency. 1 Nov 2013. <www.epa.gov/region1/enforcement/air/pdfs/CAA-MBTA-MBCR-

Fact-Sheet.pdf>.

Epstein, Paul et al. “Full Cost Accounting for the Lifecycle of Coal.” Annals of the New York

Academy of Sciences. Ecological Economics Review, vol. 10.1111. p. 73-98. 2011.

Gordon, William. “The Underground Electric.” Our Home Railways, vol. 2. 1910.

Institute for Energy Research. “Illinois Energy Facts.” Illinois. 2014. State Energy Regulations.

19 Apr 2014. <www.instituteforenergyresearch.org/state-regs/pdf/Illinois.pdf>.

Light Rail Now!. “Electrification of Transportation as a Response to Peaking of World Oil

Production.” 20 Dec 2005. Electrification 101. 31 Oct 2013.

<www.lightrailnow.org/features/f_lrt_2005-02.htm>.

Light Rail Now!. “Ready-to-Go Urban Rail Projects as a Medium-Term Response to America’s

Oil Problems.” 8 May 2007. Electrification 101. 31 Oct 2013.

<www.lightrailnow.org/features/f_lrt_2007-04a.htm>

Light Rail Now!. “The Cost-Effectiveness of Electric Rail: Examples from Real-World

Experience.” 7 Mar 2006. Electrification 101. 31 Oct 2013.

<www.lightrailnow.org/features/f_lrt_2006-03a.htm>.

Light Rail Now!. “Transport Energy Debate: How Many BTUs on the Head of a Pin?

...er...Power Line?” 10 Dec 2007. Electrification 101. 31 Oct 2013.

<www.lightrailnow.org/facts/fa_lrt_2007-08a.htm>.

Middleton, William D. “Railroad Standardization: Notes on Third Rail Electrification.” Railway

& Locomotive Historical Society Newsletter, vol. 27.4. 9 Sept 2002.

N/A. “The Price of Electricity.” Marketplace. 20 May 2011. Maps – Electricity. 25 Jan 2014.

<www.marketplace.org/topics/sustainability/maps-electricity>.

46

Samaras, Constantine et al. “Cap and Trade is Not Enough.” Improving U.S. Climate Policy. Mar

2009. Department of Engineering and Public Policy at Carnegie Mellon University. 17 Apr

2014. <www.cmu.edu/gdi/docs/cap-and-trade.pdf>.

Sims, John Walter. The Boy Electrician. London: George G Harrap & Co Limited, 1960.

Steering Group Research Programme. “Investigating the economics of the 3rd rail DC system

compared to other electrification systems.” Rail Safety and Standards Board. Aug 2011.

Supreme Court of the United States. Lee v. Chicago Transit Authority, 152 Ill.2d 432, 605

N.E.2d 493. 22 Oct 1992. Illinois Courts. 24 Oct 2013.

<www.state.il.us/court/Opinions/archive.asp>.

Uher, Richard and Neena Sathi. Reduction of Peak-Power Demand for Electric Rail Transit

Systems. Washington, D.C. : Transportation Research Board, National Research

Council, 1983.

U.S. Energy Information Administration. “Annual Energy Review” 2012. Total Energy.

<www.eia.gov/totalenergy/data/annual/pdf/sec8_3.pdf>.

U.S. Energy Information Administration. “Emissions allowance prices for SO2 and

NOX remained low in 2011.” 2 Feb 2012. Today in Energy. 5 Apr 2014.

<www.eia.gov/todayinenergy/detail.cfm?id=4830>.

U.S. Energy Information Administration. “Factors Affecting Electricity Prices.” 13 Aug 2012.

EIA. 2 Nov 2013.

<www.eia.gov/energyexplained/index.cfm?page=electricity_factors_affecting_prices>.

U.S. Energy Information Administration. “Growth in electricity use slows but still increases by

28% from 2011 to 2040.” 2011. Market Trends – Electricity. 2 Nov 2013

<www.eia.gov/forecasts/aeo/MT_electric.cfm>.

U.S. Energy Information Administration. “Use of Electricity.” 26 Apr 2012. EIA. 2 Nov 2013.

<www.eia.gov/energyexplained/index.cfm?page=electricity_use>.

U.S. Environmental Protection Agency. “Calculations and References.” 16 Apr 2014. Clean

Energy. 18 Apr 2014. <www.epa.gov/cleanenergy/energy-resources/refs.html>.

U.S. Environmental Protection Agency. “The Social Cost of Carbon.” 26 Nov 2013. Climate

Change. 5 Apr 2014. <www.epa.gov/climatechange/EPAactivities/economics/scc.html>.

47

APPENDIX I: ANNUAL IDLING CALCULATIONS, BY LINE

2011

48

49

2012

50

51

2013

52

53

APPENDIX II: ANNUAL IDLING COST BREAKDOWN

2011

2012

54

2013