Frameworks for Assessing the
Vulnerability of U.S. Rail Systems to
Flooding and Extreme Heat
30 April 2015

Mikhail Chester, Ph.D.
Assistant Professor
mchester@asu.edu
Andrew Fraser
Graduate Student
andrew.fraser@asu.edu
Matthew Bartos
Research Scientist
matthew.d.bartos@gmail.com
Civil, Environmental, and Sustainable Engineering
Arizona State University

ASU Report No. ASU-SSEBE-CESEM-2015-RPR-001
http://repository.asu.edu/items/29203
Prepared for the
National Transportation Center – University of Maryland

Project No. NTC2014-SU-R-12

Page 1 of 19

Table of Contents
1

Introduction ............................................................................................................................................... 3
1.1

Climate Change, Flooding, and Extreme Heat ............................................................................. 3

1.2

Climate Change and Passenger Rail................................................................................................ 3

2

Life Cycle Vulnerability Assessment ....................................................................................................... 4

3

Methodology for Assessing the Vulnerability of Rail Systems to Flooding ..................................... 5

4

5

6

3.1

Consequences to Infrastructure and Service ................................................................................. 5

3.2

Vulnerability Assessment Methodology......................................................................................... 6

3.3

Inland and Coastal Flood Potential ................................................................................................ 7

3.4

Conclusion.......................................................................................................................................... 8

Methodology for Assessing Reliable Electricity Supply to Rail with Extreme Heat ....................... 9
4.1

Introduction ....................................................................................................................................... 9

4.2

Determining the temperature-load response function for utility service areas ........................ 9

4.3

Estimating electricity load at electrical substations ....................................................................12

4.4

Determining the thermal limits of electrical transmission lines ...............................................14

4.5

Estimating transmission bottlenecks ............................................................................................15

4.6

Conclusion........................................................................................................................................15

Discussion .................................................................................................................................................16
5.1

Positioning LCA for Climate Change Vulnerability Assessment .............................................16

5.2

Ensuring Rail Service with Climate Change ................................................................................16

References.................................................................................................................................................18

Frameworks for Assessing the Vulnerability of U.S. Passenger Rail to Flooding and Extreme Heat
M Chester, A Fraser, and M Bartos
Page 2 of 19

1 Introduction
Recent climatic trends show more flooding and extreme heat events and in the future transportation
infrastructure may be susceptible to more frequent and intense environmental perturbations. Our
transportation systems have largely been designed to withstand historical weather events, for
example, floods that occur at an intensity that is experience once every 100 years, and there is
evidence that these events are expected become more frequent. There are increasing efforts to better
understand the impacts of climate change on transportation infrastructure (NRC, 2008). An
abundance of new research is emerging to study various aspects of climate change on transportation
systems. Much of this research is focused on roadway networks and reliable automobile travel. We
explore how flooding and extreme heat might impact passenger rail systems in the Northeast and
Southwest U.S..
1.1

Climate Change, Flooding, and Extreme Heat

In the Northeast U.S. average annual temperatures have increases by 2ºF and average winter
temperatures by 4ºF. Heavy precipitation events have increased in both magnitude and frequency
and the majority of precipitation now falls as rain, not snow (USGCRP, 2009, EPA, 2015). This
combination of temperature rise and increasing precipitation has increased the risks from flooding.
Furthermore, the rate of sea level rise in the Northeast exceeds the global average of approximately
8 inches (USGCRP, 2009).
Average temperatures in the Southwest U.S. have increased 1.6ºF (+/- 0.5ºF) from 1901-2010, with
isolated areas experiencing increases up to 3.6ºF (SWA, 2012). Summer heat waves are also expected
to become more frequent and intense around the globe (Meehl and Tebaldi, 2004), including in the
Southwest (Miller et al., 2008). Throughout the United States, the frequency of extreme heat (or
temperatures closely correlated with sharp increases in human mortality) has increased by about 20%
over the last fifty years (Gaffen and Ross, 1998).
1.2

Climate Change and Passenger Rail

There are a variety of ways in which climate change can impact passenger rail systems. In this report
we focus on how flooding might impact tracks in the Northeast and how extreme heat events are
likely to create greater demand for electricity thereby pushing electrical distribution systems closer to
their capacities, compromising reliable electricity delivery to rail systems. We start by describing how
the life cycle assessment (LCA) framework can be used to assess climate change vulnerability in
passenger transportation systems. Then, using case studies in the Northeast and Southwest, we
present methodologies for assessing how flooding and extreme heat can impact long-distance and
local passenger rail systems. The findings are relevant to current long-distance (i.e., Amtrak), future
long-distance (high-speed rail) and urban rail systems.

Frameworks for Assessing the Vulnerability of U.S. Passenger Rail to Flooding and Extreme Heat
M Chester, A Fraser, and M Bartos
Page 3 of 19

2 Life Cycle Vulnerability Assessment
The LCA framework has primarily been used to assess the environmental impacts of products,
processes, services, activities, and the complex systems in which they reside, and can be positioned
to inform vulnerability mitigation strategies for passenger transportation systems. We use LCA as a
guiding framework to assess rail life cycle processes that may be affected by flooding and extreme
heat, and develop case studies for the Northeast and Southwest U.S. to explore vulnerabilities. While
we focus on only a few life cycle processes (primarily infrastructure design and operation, and energy
production and deliver) in the provision of passenger rail services, the vulnerability LCA framework
(further referred to as vLCA) could be used to assess how vehicle (manufacturing and maintenance),
infrastructure (construction, operation, maintenance, and rehabilitation), and energy production
processes are affected by climate change, in addition to vehicle propulsion. We view vLCA as a form
of anticipatory LCA in that it can aid policy and decision makers in understanding how components
of transportation infrastructure are vulnerable to climate change and how to proactively govern and
invest resources towards reducing vulnerabilities. Life cycle processes of rail systems are shown in
Table 1 and are based on the assessments of Chester and Horvath (2009) and Chester et al. (2012).
Table 1: Passenger Rail Life Cycle Processes
Passenger Rail Life Cycle Processes
VEHICLE
Manufacturing
Operation/Propulsion
Maintenance
INFRASTRUCTURE
Construction







Train Manufacturing
Transport to Point of Use
Propulsion (Electricity Generation)
Idling (Electricity Generation)
Train Maintenance

 Tracks
 Terminals
Operation
 Track Lighting
 Herbicide Use
 Train Control
 Equipment
Maintenance/Rehabilitation
 Track Maintenance
ENERGY PRODUCTION & DELIVERY
Extraction & Processing
 Primary Fuels Extraction and Processing
 Transmission & Distribution
We assess how flooding (urban and coastal) and extreme heat can impact (long-distance and local)
passenger rail systems by focusing on a subset of the life cycle processes shown in Table 1. A
methodology for assessing how flooding can impact Northeast U.S. rail systems is developed
focusing on infrastructural processes. A methodology for assessing how the design of electricity
Frameworks for Assessing the Vulnerability of U.S. Passenger Rail to Flooding and Extreme Heat
M Chester, A Fraser, and M Bartos
Page 4 of 19

distribution infrastructure and the ability of the electricity network to provide demand for rail
electricity is developed for Southeast U.S. cities (Phoenix, Arizona and Los Angeles, California).
These studies begin to develop insight into how electric passenger rail systems in the U.S. will be
impacted by climate change and how LCA can be positioned to aid policy and decision makers in
proactively mitigating these vulnerabilities.

3 Methodology for Assessing the Vulnerability of Rail Systems to
Flooding
As a region, the Northeast is one of the densest in the country which makes it well-suited for
passenger rail service. The proximity of major U.S cities including Washington D.C., Baltimore,
Philadelphia and Boston have facilitated the development of an already well-utilized passenger rail
system and population growth projects have increased interest for future investments to expand and
improve existing services (e.g. upgrading Amtrak’s Acela line to high-speed). At the same time, the
Northeast is expected to experience impacts from climate changes that may threaten existing
infrastructure and should be accounted for when developing plans for new infrastructure and
services. Notably, climate projections forecast rising sea levels and more frequent and severe
precipitation resulting in coastal storm surges and inland flooding. These extreme weather events
could threaten day-to-day passenger rail operations as well as the long-term stability of the rail
infrastructure and the health of the people they provide service to.
3.1

Consequences to Infrastructure and Service

More frequent and intense precipitation can negatively impact rail infrastructure in several different
ways and while some may result in immediate consequences others may contribute to a slow
deterioration of components. Table 1 details the potential climate effects of more precipitation and
their infrastructure consequences. We propose a methodology to assess the vulnerability of existing
passenger rail infrastructure to flooding resulting from the increasing frequency and severity of
extreme storms. The methods are generalizable to other climate effects and infrastructure systems.

Frameworks for Assessing the Vulnerability of U.S. Passenger Rail to Flooding and Extreme Heat
M Chester, A Fraser, and M Bartos
Page 5 of 19

Table 2: Climate Impacts on Rail Infrastructure (Oslakovic et al., 2013)
Heavy Precipitation

Snowfall

High Winds
Low Temperatures

Impact
 Flooding












Flooding
Freezing
Damage to cables
Loss of electricity
Track obstructions
Coastal storm surge
(flooding)
Same as snowfall
Damage to cables
Loss of electricity
Freezing and frost

Consequences
 Rail embankment and slope
damage
 Scour of bridge supports
 Flooding of underground
structures
 Damage to rail track
 Material damage to other
equipment and infrastructure
components
 Same as heavy precipitation








3.2

Supply cable sag, tensional
failure
Same as heavy precipitation
Material damage to equipment
and infrastructure
Frost cracking, freezing of
equipment and structures on
track
Supply cable sag, tensional
failure
Damage to rail track

Vulnerability Assessment Methodology

A three-step method is utilized to assess the vulnerabilities of Northeast passenger rail systems based
on a methodology proposed by Oslakovic et al. (2013). First, a review of existing literature on
infrastructure failures resulting from weather effects, including scholarly, industry-focused and
design principles will be conducted to understand the causal link between extreme weather events
and infrastructure impacts. A synthesis of this literature forms the basis of a model which would
predict the probability of infrastructure impacts and consequences to extreme weather scenarios.
These probabilities are based upon threshold values for flooding determined from the literature.
Secondly, the current Northeast passenger rail system is studied in conjunction with climate
predictions to identify the geographic locations where infrastructure is more likely to be impacted by
future weather events. In this step we also identify critical sections of the infrastructure such as
bridges and heavily utilized rail links. The final step synthesizes the first two to create a probabilistic

Frameworks for Assessing the Vulnerability of U.S. Passenger Rail to Flooding and Extreme Heat
M Chester, A Fraser, and M Bartos
Page 6 of 19

assessment of risk to individual components and then prioritizes individual components for
preventative maintenance and rehabilitation.
3.3

Inland and Coastal Flood Potential

The increasing severity and frequency of storms in the Northeast will lead to increasing runoff and
flooding which has the potential of inundating low-lying rail lines. Additionally, increased storm
surge will likely impact facilities at or below sea level along the Eastern seaboard. Similarly, sea-level
rise may inundate and damage low-lying coastal rail infrastructure components. In addition to
immediate consequences, reoccurring rain runoff and flooding can slowly undermine critical
structural elements including bridges and railway beds and also contribute to the deterioration of
secondary infrastructure (dams, electricity distribution) which may also impact passenger rail service.
The likelihood of floods impacting passenger rail systems is based upon the location of Federal
Emergency Management Agency’s 100 Year Flood Zones and the location of passenger rail
infrastructure (FEMA, 2015). The 100 year flood plan predicts where there is a 1% chance of
flooding every year. However, climate models predict the increasing frequency and severity of
extreme storms so the probability of what is currently described as a 100 year event is likely to
increase. Alternative flooding scenarios are developed from climate forecasts. In addition to area
flooding we propose that the magnitude of stream flow (which is correlated with the magnitude of
flooding impacts) is also considered. Normal and flood stream flows available from the USGS
should be utilized (USGS, 2015). The assessment should include rail infrastructure associated with
both commuter rail systems and Amtrak along the 450 mile Northeastern Corridor stretching from
Washington D.C. to Boston, Massachusetts (Figure 1) (FRA, 2014, FRA, 2015). This includes 17
tunnels and approximately 1,200 rail bridges.

Frameworks for Assessing the Vulnerability of U.S. Passenger Rail to Flooding and Extreme Heat
M Chester, A Fraser, and M Bartos
Page 7 of 19

Figure 1: Northeastern Corridor Passenger Rail – Infrastructure Components and FEMA
100 Year Flood Planes
3.4

Conclusion

The methodology presented here provides a basic framework for assessing the risk and vulnerability
of northeastern passenger rail to flooding impacts resulting from climate change. These estimates
should be largely based on a model utilizing historical infrastructure failures and hypothetical
impacts based on historical engineering design standards. Due to the uncertainty associated with the
increasing frequency and severity of extreme weather events these methods likely lead to a
conservative estimate of risk. As rail infrastructure ages and extreme weather events increase,
resulting failures should be continually added to future iterations of the risk assessment model.

Frameworks for Assessing the Vulnerability of U.S. Passenger Rail to Flooding and Extreme Heat
M Chester, A Fraser, and M Bartos
Page 8 of 19

4 Methodology for Assessing Reliable Electricity Supply to Rail with
Extreme Heat
4.1

Introduction

Electric rail reliability may be jeopardized by power shortages resulting from future extreme heat
events. Extreme heat events pose a threat to electric power reliability because they result in increased
electricity demand while simultaneously causing decreased generation and transmission capacity.
Electricity demand is typically largest during periods of high ambient temperature, due to air
conditioning loads. At the same time, extreme heat and drought may limit the available generation
capacity at thermoelectric, hydroelectric and photovoltaic power facilities. Transmission lines and
substations are also most likely to run up against thermal limits during periods of high ambient
temperature. During extreme heat events, these effects interact in unpredictable ways, and may result
in electricity shortages. Over the next fifty years, extreme heat events are expected to occur with
greater frequency, intensity and duration; however, the combined effects of climate change on
electricity generation, transmission and demand are currently not well understood. In this study, we
assess potential power distribution bottlenecks arising from extreme heat events, and how they may
affect future electric rail reliability. To evaluate potential points of failure in the system, we use a
bottom-up approach:
1) Temperature-load response functions are created for relevant utilities to determine the
increase in electricity load for a given increase in temperature.
2) Loads at each substation are estimated by using a Voronoi tessellation to partition the utility
service area into areas served by individual substations.
3) Thermal limits of power lines are determined based on meteorological forcings and
characteristic power line properties.
4) A network model is forced with expected loads and meteorological forcings to determine
where failures are most likely to occur.
The results of this analysis are used to determine where electric power shortages are most likely to
occur in the transmission network, and to estimate the probability that electric rail reliability may be
affected by future extreme heat events.
4.2

Determining the temperature-load response function for utility service areas

Extreme heat events may strain power transmission infrastructure by incurring large coincident peak
loads. Electricity shortages were reported in 1998 during record temperatures in the Western United
States (CEC, 1999). Stage II alerts were issued on several occasions during this period by the
California Independent System Operator, signaling that operating reserves had fallen below 5
percent (CEC, 1999). During this time, interruptible-load customers were asked to curtail electricity
usage such that a 5 percent reserve margin could be maintained (CEC, 1999). Following this event,
the California Energy Commission developed a study in 1999 to examine the relationship between
temperature and electricity demand for 67 utilities in the Western Systems Coordinating Council
Frameworks for Assessing the Vulnerability of U.S. Passenger Rail to Flooding and Extreme Heat
M Chester, A Fraser, and M Bartos
Page 9 of 19

region. The study found that “small changes in average temperature … had a large impact on peak
demand” (CEC, 1999). As the Southwestern U.S. becomes hotter and drier, electricity shortages may
become more commonplace. To determine potential increases in electricity load due to climate
change, we develop temperature-load response functions for each utility service area in the
Southwest based on historical temperature and electricity demand data. For each utility service area,
the four following steps are applied to determine the temperature-load response function: (i) hourly
electricity load data are collected for the period 1993-2013, (ii) census tract data are used to construct
population estimates for the period 1990-2010, (iii) the effect of long-term population growth is
removed from hourly electricity load, and (iv) detrended hourly electricity load data are combined
with hourly temperature data in a regression model to determine the expected increase in electricity
load for a given increase in temperature.
To isolate the effect of temperature on electricity demand, the effect of population growth must first
be removed. Time series of electricity demand data can typically be separated into a long memory
component and a short memory component. The long memory component reflects year-to-year
changes in population, electricity prices, and demographic characteristics. The short memory
component reflects periodic trends in electricity demand and is mainly attributable to changes in
temperature. To isolate the short memory component of electricity demand, hourly loads from
1993-2013 are first determined for major electrical utilities in the WECC region using data from
FERC Form 714 (FERC, 2015). Next, the population of each utility service area is estimated over
the same time period. This process takes place in three steps: (i) population estimates are collected at
the census tract-level for the years 1990, 2000 and 2010 (Census, 1990a, Census, 2000a, Census,
2010), (ii), the census tracts contained within each utility service area are determined using a spatial
join (Census, 2014a, Census, 2000b, Census, 1990b, DHS, 2014a), and (iii) populations of all
“contained” census tracts are summed to determine the total population within a utility service area.
Having determined the population of each utility service area from 1993-2010, the long-memory
component of electricity demand is isolated by constructing a regression between mean annual
electricity load and mean annual population. This long-memory component is then subtracted from
observed hourly load data to determine the short-memory component of electricity demand.
Removing the effect of population growth yields the “detrended” load (normalized around zero),
which represents the portion of load that can be attributed to fluctuations in temperature. This
“detrended” load (referred to as the “load anomaly”) is indicated by the green curve in Figure 2
(right panel).

Frameworks for Assessing the Vulnerability of U.S. Passenger Rail to Flooding and Extreme Heat
M Chester, A Fraser, and M Bartos
Page 10 of 19

Figure 2. (Left): average annual load for the Salt River Project (SRP) utility service territory vs.
population growth in Maricopa County (which contains the utility service territory). (Right): Hourly
demand for the SRP service territory (1993-2013) is shown in blue. Removing the effect of population
growth from hourly demand yields the “detrended” load, which is indicated by the green curve. The
green curve shows the “short-memory” component of hourly electricity demand (i.e. the portion of
electricity demand that can be attributed to temperature effects) (Bartos and Chester, 2015).
After removing the effect of population growth, hourly electricity demand is combined with
observed temperature data to determine the relationship between hourly temperature and hourly
load. Because not all utility service areas contain representative temperature gauging stations,
historical gridded temperature data at a spatial resolution of 1/8 degree are used (Maurer et al.,
2002). In Figure 3 (left), the load anomaly (i.e. the detrended load) is plotted against the temperature
anomaly (which represents the hourly temperature’s deviation from the annual mean temperature as
a decimal fraction). This plot shows a strong positive correlation between air temperature and
electricity load. However, the relationship is not linear. Rather, the plot of load anomaly vs.
temperature anomaly exhibits a hysteresis loop, which likely reflects the effect of thermal storage in
air conditioned buildings: buildings heat up during the morning hours, following the lower bound of
the loop. After reaching a peak (when daily temperature is greatest), the load begins to decrease over
the evening and nighttime hours, following the upper bound of the loop. The highest hourly
temperatures are associated with the highest hourly electricity loads. For each utility service area, a
logistic regression is used to relate a given load anomaly to a given temperature anomaly. Linear
regressions are also used to relate daily peak loads with daily maximum temperatures (see Figure 3,
right).

Frameworks for Assessing the Vulnerability of U.S. Passenger Rail to Flooding and Extreme Heat
M Chester, A Fraser, and M Bartos
Page 11 of 19

Figure 3. (Left) Temperature-load relationship for the Salt River Project (SRP) service territory

during summer months (June, July and August). The horizontal axis represents the temperature
anomaly (the degree to which temperature exceeds the mean temperature as a decimal fraction). T he
vertical axis represents the load anomaly (hourly electricity load with the effect of population growth
removed, normalized around zero). Colors represent the kernel density of observed values (with red and
yellow areas corresponding to a greater number of observations). Note that the relationship between
temperature and load exhibits a hysteresis loop, most likely resulting from the effects of building
thermal storage: buildings heat up during the day, following the lower bound of the loop, then cool of f
at night, following the upper bound of the loop. (Bartos and Chester, 2015) (Right) The relationship
between maximum daily temperature and peak load is nearly linear; however, the slope of the
relationship varies from year to year, and is influenced by socioeconomic factors.

Some utility service areas (such as the Arizona Public Service Co.) span multiple climate zones and
population centers. For these utilities, a single representative temperature cannot be used to
determine the relationship between temperature and electricity demand. Instead, multiple regression
is used to relate temperatures at major population centers to total electricity load. First, major
population centers in each utility service area are determined using the U.S. Census Bureau’s Urban
Areas dataset (Census, 2014b). Gridded observed temperatures at each urban area are then used in a
multiple linear regression relating daily hourly temperature to electricity demand. The populations of
the representative urban areas are used as the initial weightings in this multiple regression.
4.3

Estimating electricity load at electrical substations

Power failures typically do not occur everywhere at once within a transmission network, but rather
occur at the “weakest link” in the system. To determine the probability of intra-urban electricity
shortages, it is necessary to estimate electrical loads at individual transmission lines and substations
such that the most vulnerable regions of the network can be identified. In this section, we outline a
method for estimating component-scale loads based on hourly utility-scale demand and housing
densities.
Electricity load at the sub-utility scale is generally not accessible to the public. To estimate loads at
individual substations, utility-level demand data are downscaled to component substations by (a)
determining the approximate region served by each substation, and (b) determining the number of
housing units located in each of these regions. First, to determine the approximate area served by
Frameworks for Assessing the Vulnerability of U.S. Passenger Rail to Flooding and Extreme Heat
M Chester, A Fraser, and M Bartos
Page 12 of 19

each substation, a Voronoi tessellation is generated, with substation locations representing the
“seed” points of each Voronoi cell (DHS, 2014b). Voronoi tessellation is a method of partitioning a
plane into unique regions based on a set of “seed points”. For each seed point, Voronoi tessellation
yields a corresponding region consisting of all points closer to that seed point than any other. In this
case, the cells generated by the Voronoi tessellation represent the area that is closest to each
substation to the exclusion of all other substations. After determining the approximate region served
by each substation, the relative load at each substation is estimated based on the number of houses
in each Voronoi polygon. To this end, we use gridded housing density data from the EPA’s ICLUS
dataset for the year 2010 (EPA, 2010). The EPA ICLUS dataset is used because it offers housing
density data at a fine scale, contains projected populations up to the year 2100, and is integrated with
SRES climate scenarios. Figure 4 and Figure 5 show Voronoi tessellations for Los Angeles and
Maricopa counties, respectively. The color of each Voronoi cell indicates the approximate load at
the contained substation.

Figure 4. Voronoi tessellation of Los Angeles electrical substations. Substations are
indicated by black dots, while the colored regions indicate the Voronoi cells. The col or
of the Voronoi cell indicates the load, in MW, with blue cells representing small loads,
and red cells representing large loads.
Frameworks for Assessing the Vulnerability of U.S. Passenger Rail to Flooding and Extreme Heat
M Chester, A Fraser, and M Bartos
Page 13 of 19

Figure 5. Voronoi tessellation of Maricopa County electrical substations. Substations
are indicated by black dots, while the colored regions indicate the Voronoi cells. The
color of the Voronoi cell indicates the load, in MW, with blue cells representing small
loads, and red cells representing large loads.
4.4

Determining the thermal limits of electrical transmission lines

Extreme heat events may contribute to electricity shortages by causing transmission lines to surpass
their thermal ratings. Thermal ratings represent the maximum capacity at which an overhead
conductor can operate for given weather conditions. Continued operation beyond a conductor’s
thermal rating can result in excessive sag or damage to the conductor (IEEE, 2006). To avoid
damage, utilities typically either curtail the current in an at-risk conductor, or shut off power to than
conductor entirely. Additionally, transmission lines suffer incremental power losses at elevated
conductor temperatures, meaning that elevated temperatures could result in reduced transmission
capacity even when thermal limits are not breached (IEEE, 2006).
Thermal ratings (i.e. design ampacities) are determined based on a maximum allowable conductor
temperature, meteorological conditions, and conductor characteristics (such as the geometry of the
conductor and its material properties) (IEEE, 2006). The thermal limits of overhead conductors are
estimated using a steady-state heat balance (IEEE, 2006):
Frameworks for Assessing the Vulnerability of U.S. Passenger Rail to Flooding and Extreme Heat
M Chester, A Fraser, and M Bartos
Page 14 of 19

𝑞𝑐 + 𝑞𝑟 = 𝑞𝑠 + 𝐼 2 𝑅(𝑇𝑐 )
Where 𝑞𝑐 represents the convected heat loss rate per unit length (W/m), 𝑞𝑟 represents the radiated
heat loss rate per unit length (W/m), 𝑞𝑠 represents the heat gain rate from the sun (W/m), 𝐼
represents the conductor current (A), 𝑅(𝑇𝑐 ) represents the AC resistance of the conductor at
temperature 𝑇𝑐 (Ω/m). This heat balance can be rearranged to yield the maximum current for a
design conductor temperature:
𝐼= √

𝑞𝑐 + 𝑞𝑟 − 𝑞𝑠
𝑅(𝑇𝑐 )

Design ampacities are calculated for each transmission line in the area of interest based on
characteristic transmission line designs and meteorological forcings (i.e. temperature and wind
speed) at design conditions (IEEE, 2006). These design ampacities are then used as inputs to a loadbalancing simulation, which aims to determine where power shortages are most likely to occur
within a transmission network.
4.5

Estimating transmission bottlenecks

After determining the effect of climate change on both electricity demand and electricity
transmission capacity, a load-balancing model is forced with projected meteorological data to
determine the extent to which elevated temperatures may induce power shortages. In this model,
power shortages occur when peak demand cannot be satisfied—either because the temperatureinduced demand exceeds the carrying capacity of the connecting lines, or because exceedance of
thermal ratings results in a momentary power outage. The load balancing model is forced with
downscaled outputs from three GCM models and three carbon emissions scenarios for the period
2010-2050. For projections of temperature and wind speed, gridded forcings from the CMIP3 multimodel are used (DOI, 2013). The load-balancing model is used to deliver a first order estimate of
which transmission lines are most at risk of power shortages. Impacts to rail lines are expected to
occur if the substations serving these rail lines are susceptible to frequent power outages under
future extreme heat events.
4.6

Conclusion

This study uses a bottom-up approach to determine how electric rail reliability may be affected by
power shortages resulting from future extreme heat events. This process takes place in four steps.
First, temperature-load response functions are developed for each utility in the Southwest, such that
peak loads resulting from future extreme heat events may be predicted. Second, utility-level
electricity demand is allocated to substations within each utility service area using a Voronoi
tessellation approach. Next, thermal limits are determined for each transmission line in the location
of interest based on transmission line characteristics along with wind speed and temperature
Frameworks for Assessing the Vulnerability of U.S. Passenger Rail to Flooding and Extreme Heat
M Chester, A Fraser, and M Bartos
Page 15 of 19

forcings. Finally, using the substation-level loads and thermal ratings developed in the previous
steps, a load-balancing simulation is used to determine the probability of power shortages under
future extreme heat events, and to determine the location within the network at which failures are
most likely to occur. The results of the load-balancing simulation are used to determine whether
substations serving electric rail lines are likely to experience power shortages. In this way, the
reliability of electric rail under future extreme heat events can be assessed.

5 Discussion
5.1

Positioning LCA for Climate Change Vulnerability Assessment

Currently there exists no systems-oriented framework for proactively managing the potential
vulnerabilities associated with climate change to passenger transportation systems and the vLCA
framework offers significant opportunity for guiding decision makers towards protecting their
systems, including vehicle, infrastructure, and energy production components, in addition to
propulsion. LCA has traditionally focused on quantifying environmental effects. However, the
framework is well-suited for structured decomposition of complex passenger transportation systems
to understand, from material and primary fuel extraction, transformation into finished products, use,
maintenance, rehabilitation, and operation of the vehicles, where vulnerabilities will exist due to
environmental perturbations, and how those vulnerabilities may cascade through the system. While
we do not offer a case study in this report that comprehensively uses vLCA, we instead intend to
setup the framework for future studies of passenger transportation systems.
5.2

Ensuring Rail Service with Climate Change

The flooding and extreme heat case studies impact infrastructure use and vehicle operations, which
have the potential to create major service disruptions, and service providers can protect existing and
future systems with proactive management strategies. Engineers should recognize that flooding
events that occurred with low frequencies (e.g., once in 100 years) are forecast to occur with greater
frequency (USGCRP, 2009). As such, design standards which likely follow historical weather events,
should be updated to account for these greater frequencies. We should also recognize that more
frequent and more intense precipitation events are likely to increase our transportation system’s
exposure to water. Proactive management strategies for flooding should consider:





Electrical components should be protected with more robust insulation or raised above or
moved away from potential water intrusion.
The design of track beds should be reassessed to ensure that they maintain their structural
integrity.
Station designs should be reconsidered to protect passengers (in both safety and comfort)
from the increases in precipitation.
Protections should be afforded for ingress to and egress from stations.

Frameworks for Assessing the Vulnerability of U.S. Passenger Rail to Flooding and Extreme Heat
M Chester, A Fraser, and M Bartos
Page 16 of 19

Similarly, physical components of transportation systems and passengers are likely to experience
more frequent and intense exposure to extreme heat, particularly in the U.S. Southwest (Bartos and
Chester, 2014). Extreme heat is likely to impact passenger transportation systems through a number
of pathways. As such, proactive management strategies should consider:





Reassessing the design specifications of physical components of vehicles and infrastructure
to ensure that they will reliably perform during hotter temperatures over longer durations.
Protecting people through increased use of shading, fans, and misters (structural and natural)
at stations and during ingress and egress.
Implementing renewable energy generation systems for electricity delivery for propulsion,
station operations, and track and signal control.
Reassessing transmission and distribution network capacity to ensure that rail systems will
receive reliable and sufficient supply during times of increased demand.

It’s important to recognize that large-scale system failures can stem from component failures that
cascade through the system (Chester, 2013). Strategies to protect transportation systems against
climate change must take an entire systems view.

Frameworks for Assessing the Vulnerability of U.S. Passenger Rail to Flooding and Extreme Heat
M Chester, A Fraser, and M Bartos
Page 17 of 19

6 References
BARTOS, M. & CHESTER, M. 2014. Assessing Future Extreme Heat Events at Intra-urban Scales:
A Comparative Study of Phoenix and Los Angeles. Tempe, AZ: Arizona State University
Report No. ASU-CESEM-2013-WPS-001.
BARTOS, M. & CHESTER, M. 2015. Impacts of climate change on electric power supply in the
Western United States. Nature Climate Change, Volume and Issue Forthcoming.
CEC 1999. High Temperatures & Electricity Demand: An Assessment of Supply Adequacy in
California, Trends & Outlook. Sacramento, CA: California Energy Commission.
CENSUS 1990a. Population by Census Tract. Washington, DC: U.S. Census Bureau.
CENSUS 1990b. TIGER/Line Shapefiles - Census Tract Geometries. Washington, DC: U.S.
Census Bureau.
CENSUS 2000a. Population by Census Tract. Washington, DC: U.S. Census Bureau.
CENSUS 2000b. TIGER/Line Shapefiles - Census Tract Geometries. Washington, DC: U.S.
Census Bureau.
CENSUS 2010. Population by Census Tract. Washington, DC: U.S. Census Bureau.
CENSUS 2014a. TIGER/Line Shapefiles - Census Tract Geometries. Washington, DC: U.S. Census
Bureau.
CENSUS 2014b. TIGER/Line Shapefiles - Urban Area Geometries. Washington, DC: U.S. Census
Bureau.
CHESTER, M. 2013. Human and Organizational Factors that Contributed to the US-Canadian
August 2003 Electricity Grid Blackout. Tempe, AZ: Arizona State University Report No.
ASU-SSEBE-CESEM-2013-RPR-003.
CHESTER, M. & HORVATH, A. 2009. Environmental Assessment of Passenger Transportation
Should Include Infrastructure and Supply Chains. Environmental Research Letters, 4, 024008.
CHESTER, M., PINCETL, S. & ALLENBY, B. 2012. Avoiding unintended tradeoffs by integrating
life-cycle impact assessment with urban metabolism. Current Opinion in Environmental
Sustainability, 4, 451-457.
DHS 2014a. Electric Retail Service Territories. Washington, DC: U.S. Department of Homeland
Security's Infrastructure Program.
DHS 2014b. Electric Substation Shapefiles. Washington, DC: U.S. Department of Homeland
Security's Infrastructure Program.
DOI 2013. Downscaled CMIP3 and CMIP5 Climate and Hydrology Projections: Release of
Downscaled CMIP5 Climate Projections, Comparison with preceding Information, and
Summary of User Needs. Denver, CO: U.S. Department of the Interior, Bureau of
Reclamation.
EPA 2010. ICLUS v1.3 Housing Density Projections. Washington, DC: Global Change Research
Program, National Center for Environmental Assessment, Report No. EPA/600/R09/143F.
EPA. 2015. Climate Impacts in the Northeast [Online]. Washington, DC: U.S. Environmental Protection
Agency. Available: http://www.epa.gov/climatechange/impacts-adaptation/northeast.html
[Accessed April 29, 2015.
FEMA 2015. 100 Year Flood Zones. In: U.S. DEPARTMENT OF HOMELAND SECURITY
(ed.). Washington DC: Federal Emergency Management Agency.
FERC 2015. Form No. 714 – Annual Electric Balancing Authority Area and Planning Area Report.
Washington, DC: Federal Energy Regulatory Commission.
Frameworks for Assessing the Vulnerability of U.S. Passenger Rail to Flooding and Extreme Heat
M Chester, A Fraser, and M Bartos
Page 18 of 19

FRA 2014. Railroad Bridges (GIS). In: U.S. DEPARTMENT OF TRANSPORTATION (ed.).
Washington, DC: Federal Railroad Administration.
FRA 2015. National Rail Network. In: U.S. DEPARTMENT OF TRANSPORTATION (ed.).
Washington DC: Federal Railroad Administration.
GAFFEN, D. J. & ROSS, R. J. 1998. Increased summertime heat stress in the US. Nature, 396, 529530.
IEEE 2006. IEEE Standard for Calculating the Current-Temperature of Bare Overhead
Conductors. New York, NY: IEEE Power Engineering Society.
MAURER, E., WOOD, A., ADAM, J., LETTENMAIER, D. & NIJSSEN, B. 2002. A Long-Term
Hydrologically Based Dataset of Land Surface Fluxes and States for the Conterminous
United States. American Meteorological Society, 15, 3237-3251.
MEEHL, G. & TEBALDI, C. 2004. More intense, more frequent, and longer lasting heat waves in
the 21st century. Science, 305, 994-997.
MILLER, N., HAYHOE, K., JIN, J. & AUFFHAMMER, M. 2008. Climate, Extreme Heat, and
Electricity Demand in California. Journal of Applied Meteorology and Climatology, 47, 1834-1844.
NRC 2008. Potential Impacts of Climate Change on U.S. Transportation. Washington, DC: National
Research Council Committee on Climate Change and U.S. Transportation.
OSLAKOVIC, I. S., TER MAAT, H., HARTMANN, A. & DEWULF, G. Risk Assessment Of
Climate Change Impacts On Railway Infrastructure. proceedings of the Engineering Project
Organization Conference Devil’s Thumb Ranch, Colorado, July, 2013. 9-11.
SWA 2012. Assessment of Climate Change in the Southwest United States: A Technical Report
Prepared for the U.S. National Climate Assessment. Tucson, AZ: Southwest Climate
Alliance.
USGCRP 2009. Global Climate Change Impacts in the United States. Cambridge, MA: U.S. Global
Change Research Program.
USGS 2015. National Hydrography Dataset. In: INTERIOR, U. S. D. O. T. (ed.). Washington, DC:
U.S. Geological Survey.

Frameworks for Assessing the Vulnerability of U.S. Passenger Rail to Flooding and Extreme Heat
M Chester, A Fraser, and M Bartos
Page 19 of 19