Embed Size (px)
Citation preview
LETTER • OPEN ACCESS
Impacts of rising air temperatures on electrictransmission ampacity and peak electricity load inthe United StatesTo cite this article: Matthew Bartos et al 2016 Environ. Res. Lett. 11 114008
View the article online for updates and enhancements.
You may also likeStudy on the ampacity of single-coresubmarine power cable with returnconductorFAN Youbing, LIU Songhua, LI Te et al.
-
Calculation Method of Cable AmpacityBased on Field-circuit Combined ModeDan Pang, Haipeng Wu, Bin Dai et al.
-
Study of Temperature Field and Ampacityof 110kV AC Submarine Cables underDifferent Laying ConditionsZhenxin Chen, Yanjie Le, Heng Liu et al.
-
Recent citationsQuantification of Climate-InducedInterannual Variability in Residential U.S.Electricity DemandHadi Eshraghi et al
-
Future flood riverine risk analysisconsidering the heterogeneous impactsfrom tropical cyclone and non-tropicalcyclone rainfalls: Application to daily flowsin the Nam River Basin, South KoreaAngelika L. Alcantara and Kuk-Hyun Ahn
-
Electricity load forecast considering searchengine indicesXinyu Wu et al
-
This content was downloaded from IP address 65.21.228.167 on 17/11/2021 at 03:46
Environ. Res. Lett. 11 (2016) 114008 doi:10.1088/1748-9326/11/11/114008
LETTER
Impacts of rising air temperatures on electric transmission ampacityand peak electricity load in the United States
MatthewBartos1,4,Mikhail Chester1, Nathan Johnson2, BrandonGorman1, Daniel Eisenberg1,3,Igor Linkov3 andMatthewBates3
1 Department of Civil, Environmental and Sustainable Engineering, Arizona StateUniversity, USA2 The Polytechnic School, Ira A. Fulton Schools of Engineering, Arizona StateUniversity, USA3 USArmyEngineer Research andDevelopment Center, Vicksburg,MS,USA4 Author towhomany correspondence should be addressed.
E-mail:[email protected],[email protected], [email protected], [email protected], [email protected], [email protected] [email protected]
Keywords: climate change, infrastructure, resilience, electrical grid
Supplementarymaterial for this article is available online
AbstractClimate changemay constrain future electricity supply adequacy by reducing electric transmissioncapacity and increasing electricity demand. The carrying capacity of electric power cables decreases asambient air temperatures rise; similarly, during the summer peak period, electricity loads typicallyincrease with hotter air temperatures due to increased air conditioning usage. As atmospheric carbonconcentrations increase, higher ambient air temperaturesmay strain power infrastructure bysimultaneously reducing transmission capacity and increasing peak electricity load.We estimate theimpacts of rising ambient air temperatures on electric transmission ampacity and peak per-capitaelectricity load for 121 planning areas in theUnited States using downscaled global climatemodelprojections. Together, these planning areas account for roughly 80%of current peak summertimeload.We estimate climate-attributable capacity reductions to transmission lines by constructingthermalmodels of representative conductors, then forcing thesemodels with future temperatureprojections to determine the percent change in rated ampacity. Next, we assess the impact ofclimate change on electricity load by using historical relationships between ambient temperature andutility-scale summertime peak load to estimate the extent towhich climate changewill incuradditional peak load increases.We find that bymid-century (2040–2060), increases in ambient airtemperaturemay reduce average summertime transmission capacity by 1.9%–5.8% relative to the1990–2010 reference period. At the same time, peak per-capita summertime loadsmay rise by4.2%–15%on average due to increases in ambient air temperature. In the absence of energy efficiencygains, demand-sidemanagement programs and transmission infrastructure upgrades, theseload increases have the potential to upset current assumptions about future electricity supplyadequacy.
Glossary
qc Convective heat loss fromconductor to air (Wm–1)
qr Radiative heat loss fromconductor to surround-ings (Wm–1)
qs Radiative heat transferfrom sun to conductor(Wm–1)
qj Resistive heating of theconductor (Wm–1)
I Conductor current (A)
OPEN ACCESS
RECEIVED
31December 2015
REVISED
26 September 2016
ACCEPTED FOR PUBLICATION
14October 2016
PUBLISHED
2November 2016
Original content from thisworkmay be used underthe terms of the CreativeCommonsAttribution 3.0licence.
Any further distribution ofthis workmustmaintainattribution to theauthor(s) and the title ofthework, journal citationandDOI.
© 2016 IOPPublishing Ltd
Tcond Average conductor temp-erature (K)
R AC resistance of conduc-tor (Ωm–1)
h̄ Average heat transfercoefficient (Wm–2 K–1)
D Conductor diameter (m)
Tamb Ambient air temper-ature (K)
Tf Film temperature (K)
e Emissivity of conductorsurface (dimensionless)
s Stefan–Boltzmann con-stant (5.670e–8Wm−2 K−4)
d Incident solar radiation(Wm–2)
as Absorptivity of conductorsurface (dimensionless)
d Per-capita electricity load(Wper capita)
1. Introduction
Climate change may adversely affect electricity supplyadequacy by reducing generation and transmissioncapacity while simultaneously increasing electricitydemand. Extreme heat and drought can impair powergeneration capacity [1–3], limit the current-carryingcapacity (ampacity) of transmission lines [3], andincrease peak electricity loads [3, 4]. As atmosphericcarbon concentrations increase, extreme heat anddrought events are expected to occur with greaterfrequency, meaning that power infrastructure may beplaced under greater strain for longer periods of time [5].Future drought conditions have the potential to limitpower generation at ‘base-load’ power plants—whichrequire a consistent supply of water for cooling [1, 2, 6–9]. Similarly, extreme heatmay reduce the power outputof peaking generation sources like gas turbines, whichbecome less efficient as the density of air decreases [1, 3],and photovoltaic solar cells, which lose efficiency at highair temperatures due to increased carrier recombinationrates [10]. By mid-century, changes in climate mayreduce vulnerable generation capacity in theWesternUSby as much as 1.1%–3.0% in an average year, and up to7.2%–8.8% under a ten-year drought scenario [1].Although these capacity reductions are significant intheir own right, impacts to power generation are likely tooccur alongside impacts to electricity transmission anddemand. Elevated air temperatures can reduce the ratedcapacity of electric transmission lines,meaning that theirability to transmit power will be diminished during peakhours. Higher temperatures may also increase electricityloads for air conditioning during the summer period,
and although only one-half of United States utilitiesexperience peak summer load in the summer, theaggregation of these utilities create a nationwide peakload that is higher in the summer than in other parts ofthe year, thereby indicating summer as the period ofinterest when considering nationwide transmissioninfrastructure vulnerability to rising ambient air tem-peratures [11, 12]. The opposite effect may be observedin other periods of the year with different energy usepatterns, such as colder regions with electric resistanceheating or electric heat pumps in buildings that create anannual peak load in the winter rather than the summer[13], and this peak may be reduced as ambienttemperatures increase in winter [12]. Nevertheless, forthe summer period, rises in ambient temperature createcoincident impacts to electricity generation, transmis-sion and peak load that may result in a more vulnerableelectric grid in certain regions of the US. Traditionally,risk analysis for electric power systems has taken theform of probabilistic methods based on historicalconditions [14]. In a future characterized by changingclimatic conditions, increased grid complexity, andaugmented operational ‘interconnectedness’, thisapproach may no longer be sufficient to protect againstoutages and other grid disruptions [14, 15]. Failure toaccount for the effects of climate change on the electricalgridmay leave regional planning authorities unpreparedfor future electricity shortages.
The effects of climate change on overall electricpower reliability have yet to be fully explored. It is,however, possible that rising electrical loads combinedwith a reduction in electrical generation and/or linecarrying capacity will narrow the band of operationalstability and safety. Broadly, electricity supply ade-quacy can be described as a function of three factors:(1) achievable generation capacity, (2) transmissionand distribution capacity, and (3) expected peak load.While previous research has assessed climate impactsto each of these components in isolation, it has so farproven difficult to assess how the combined impacts togeneration, transmission and peak load may affectoverall electricity supply adequacy. Evaluation offuture electricity supply adequacy is complicated by ahost of factors, including climate non-stationarity,system redundancy, power system regulations andindustry norms. However, a major obstacle to evalua-tion of future electricity supply adequacy is the factthat impacts to generation, transmission and demandhave been evaluated at varying local and regionalscales, but have not been evaluated at the scale neces-sary for grid-level assessment. Several studies haveassessed the potential impacts of climate change onpower generation at national, interconnection andsub-interconnection scales [1–3]. However, transmis-sion ampacity and peak electricity load also play animportant role in determining a region’s overall elec-tricity supply adequacy. Although the effects of ambi-ent temperature on electric transmission capacity havelong been recognized, power providers typically rate
2
Environ. Res. Lett. 11 (2016) 114008
system ampacity using historical temperature profiles[16, 17]. Sathaye et al [3] use global climate model(GCM) output to characterize the effects of climatechange on transmission infrastructure in California,and find that transmission capacity may be reduced by7%–8% by the end of the twenty-first century [3].However, to our knowledge there is no study that esti-mates the impacts of rising temperatures on transmis-sion capacity nationwide.
Several recent studies have assessed the potentialimpacts of climate change on energy consumption.However, these studies generally fall short of forecast-ing potential effects to electricity supply adequacy for anumber of reasons: (1) they concentrate on increasesto annual building energy consumption (i.e. MWh), asopposed to instantaneous peak load (i.e.MW), (2) theyuse a spatial extent or resolution that is not scalable toan interconnection-level assessment, or (3) they useassumed temperature increases, as opposed to spa-tially explicit GCM projections. Almost all previousstudies focus on climate impacts to annual buildingenergy consumption instead of peak electricity load[4]. Several studies assess changes to annual buildingenergy demand at the national scale [18–23]; othersfocus on a particular region [24], using finer-scaletemperature and infrastructure data [24]. While thesestudies help to predict the increased primary energyburden resulting from climate change, they do notindicate how climate change will affect the capacityneeded during peak times. Franco and Sanstad [25]and Sathaye et al [3] estimate climate impacts to peakload in California. However, in terms of electricalinfrastructure, California is not an isolated system,and relies on interstate electricity transfers to meetdemand [26]. Thus, impacts to California’s peak per-iod infrastructure do not necessarily reflect overallelectricity supply adequacy at the interconnectionscale. Reference [27] investigate spatially explicit chan-ges in electricity load at the neighborhood scale usinghigh-resolution downscaled climate model output inthe Southeastern United States. Dirks et al [4] use ahighly detailed building energymodel to assess climateeffects on peak electricity load in the Eastern Inter-connection using one climate model scenario [4].However, to our knowledge there has been no com-prehensive, national-scale assessment of peak loadimpacts under a range of possible climate change sce-narios. Because of the relative lack of research on peakdemand and transmission impacts at the interconnec-tion scale, there remain significant unknowns as to theeffects of climate change on electricity reliability. Toaddress these knowledge gaps, we estimate impacts toelectricity transmission capacity and summertimepeak per-capita load over the continental UnitedStates.
2.Methodology
We estimate the impacts of climate change on (a) therated ampacity of transmission lines and (b) expectedsummertime peak per-capita electricity load for 121electric service planning areas across the US. Together,these planning areas supply roughly 650 GW of power[28]—about 80% of current peak summertime load intheUS [29]. Our analysis takes place in two parts. In thefirst phase, we estimate climate-attributable capacityreductions to aerial transmission lines by constructingthermal models of representative conductors andforcing these models with temperature projections(from 2010–2100) to determine the relative change insafe operating ampacity. In the second phase, wecharacterize climate impacts on electricity load by usinghistorical relationships between ambient temperatureand summertime per-capita peak load to estimate theextent to which climate change may incur additionalload increases. To account for variability in GCMprojections, we use output from 11 different GCMmodels with 2–3 representative concentration pathway(RCP) scenarios per model. After estimating climate-attributable impacts to both electricity transmissionand load, we discuss how atmospheric carbon reduc-tions, heat-resistant conductors and smart-grid tech-nologies can mitigate the effects of rising airtemperatures on electricity peak-period infrastructure.
2.1. Estimating climate impacts to electrictransmission capacityWe estimate the impacts of climate change on electrictransmission capacity by developing thermal modelsof representative electric transmission cables, thenforcing these models with downscaled temperatureprojections over the period 2010–2100. Electric powercables suffer decreased transmission capacity as thetemperature of the conductor increases [30]. A portionof these capacity losses result from increased electricalresistance at higher conductor temperatures [30].However, the current-carrying capacity of a transmis-sion line is primarily limited by a conductor’s max-imum allowed operating temperature [30]. Maximumoperating temperatures are prescribed for differenttypes of conductors to (a) ensure compliance withclearance regulations, and (b) prevent damage to theconductor and other line hardware [16]. For a typicalaluminum conductor steel-reinforced cable, allowabletemperatures may range from 50 °C to 180 °Cdepending on the duration of heat exposure, andengineering practice and judgment [30]. Continuedoperation beyond a conductor’s maximum operatingtemperature can result in excessive sag or damage [30].To avoid surpassing a transmission line’s maximumoperating temperature, operators typically curtail thecurrent in an at-risk conductor such that thermal
3
Environ. Res. Lett. 11 (2016) 114008
limits are satisfied [3, 30]. Thus, electric power cablesare generally given a ‘rated ampacity’, which representsthe maximum current for which conductor temper-ature limits are met under standard ambient temper-ature and wind conditions [3, 30]. Hotter airtemperatures due to anthropogenic global warmingmay reduce the effective ampacity of transmissionlines by interfering with their ability to dissipate heat[30]. To gauge potential ampacity reductions underchanging climatic conditions, we develop a thermalbalance model to estimate the rated ampacity oftransmission lines based on cable properties andmeteorological forcings. We then force this thermalbalance model with future temperature projections todetermine the percent change in rated ampacity fortransmission lines in theUS.
We simulate the rated ampacity of power linesusing an energy balance approach
D = + - + - ( )E q q q q q . 1j rs c d
Assuming steady-state conditions (D = )E 0 andno conduction =( )q 0 ,d the energy balance of a con-ductor can be resolved into four heat transfer compo-nents: (1) heat gain from the electrical current flowingthrough the conductor, q ,j (2) heat gain from solarradiation striking the top half of the surface of the con-ductor, q ,s (3) heat loss due to convection, q̇ ,c and (4)heat loss due to radiation, qr [30]. To satisfy equili-brium conditions, the total heat transfer into the con-ductor must equal the total heat transfer out of theconductor
+ = + ( )q q q q . 2jc r s
Heat gain due to electrical loading (known as Jouleheating) is a function of the current transferredthrough the conductor (I) and the resistance of theconductor at a given conductor temperature, ( )R Tcond
[30]
= ⋅ ( ) ( )q I R T . 3j2
cond
Rearranging the heat balance yields the maximumallowable current (the rated ampacity) as the depen-dent variable
=+ -
( )( )I
q q q
R T. 4c r s
cond
In expanded form, the rated ampacity of an over-head conductor can be expressed in terms of ambientweather conditions (temperature and wind speed),solar insolation, and cable properties (diameter, sur-face area and material properties). For a full deriva-tion, see section 1.1.1 in the supplementaryinformation (SI) document
where h̄ is the average heat transfer coefficient, D isthe cable diameter, Tamb is the ambient air temper-ature, e is the emissivity of the conductor surface, s isthe Stefan–Boltzmann constant, d is the incident solarradiation, and as is the absorptivity of conductorsurface.
We apply the thermal balance model to existingand proposed transmission lines in the United Statesto predict the relative decrease in rated ampacity underfuture climatic conditions. We use the HomelandSecurity Infrastructure Program (HSIP) database todetermine the locations, geometries and voltage clas-ses for both existing and proposed transmission lines[31]. This dataset includes 50 822 existing and 1184proposed high-voltage transmission lines throughoutthe continental US. Representative model cables aregenerated for each standard voltage class (69, 138, 230,345 and 525 kV) based on reported conductor specifi-cations (see SI section 1.1.3). Manufacturer data areused to determine relevant design specifications formodel cables, such as conductor diameter and ACresistance [32–34].We use amaximum allowable con-ductor temperature of 75 °C, a representative windspeed of 0.61 m s−1 perpendicular to the conductoraxis, and a solar insolation intensity of 1000Wm−2 torepresent full sun conditions, per standard ampacitycalculation protocols used by manufacturers to ratethe ampacity of aluminum-conductor steel-reinforced(ACSR) conductors (the most common conductortype currently in use) [32–35]. Although average solarinsolation varies by geographical location, we use afixed solar insolation value of 1000Wm−2 to (i) evalu-ate ampacity under ‘worst-case’ (full sun) conditions,and (ii) tomaintain consistency with existing ampacityrating protocols. We do not model long-term changesinwind speed, given that wind speed in the continentalUS is not expected to change significantly as a result ofclimate change (see SI section 2.5), though it should benoted that the effect of climate change on continentalwind speed is still a debated subject. For our base-casescenario, we assume that the conductor technologiesused for proposed cables will be similar to those usedfor existing cables. However, adoption of more heat-resistant conductor technologies—such as aluminumconductor steel supported (ACSS) conductors—canhelp to mitigate impacts posed by rising air tempera-tures. The effect of using more heat-resistant con-ductors is explored in SI section 2.3. Other assumedparameters used in the thermal model (includingemissivity and absorptivity) can be found in SItable S3.
The thermal balance model is executed at a dailytime step (using daily maximum temperature) to
p p e s d=
⋅ ⋅ ⋅ - + ⋅ ⋅ ⋅ ⋅ - - ⋅ ⋅¯ ( ) ( )( )
( )Ih D T T D T T D a
R T5cond amb cond
4amb
4s
cond
4
Environ. Res. Lett. 11 (2016) 114008
determine the reduction in rated ampacity betweenhistorical and future scenarios. For the historical per-iod, we force the thermalmodel with gridded observeddaily maximum temperature data from 1990–2010[36]. The temperature forcings included in this datasetare derived from National Oceanic and AtmosphericAdministration (NOAA) Cooperative Observer sta-tions, which are gridded to a spatial resolution of 1/8degree (roughly 12 km) using the synergraphic map-ping system (SYMAP) algorithm [37, 38]. For thefuture period, we use downscaled daily maximumtemperature data from the Coupled Model Inter-comparison Project Phase 5 (CMIP5) multi-modelensemble of GCMs [39]. CMIP5 is the latest iterationof the CMIP project, which aims to cultivate a standar-dized protocol for comparing outputs of coupledatmosphere-ocean general circulation models. Pro-jected temperature data from GCMs are gridded at aspatial resolution of 1/8 degree using the bias-cor-rected constructed analogue downscaling method.Temperature data are joined to each transmissioncable in the HSIP dataset based on spatial location. Tocapture a range of possible futures, we use the RCP 2.6,4.5 and 8.5 RCPs proposed by the IntergovernmentalPanel on Climate Change. These scenarios prescribeanticipated ranges of anthropogenic warming basedon divergent trends in atmospheric carbon concentra-tions, and provide a basis for measuring the impacts ofclimate change under different social, technical andpolicy scenarios. The specific GCMs and RCP scenar-ios used in this study are shown in table 1.
2.2. Estimating peak per-capita load increases due toelevated air temperaturesClimate change may strain power infrastructure byincreasing loads during the summer peak period.During the summer months, electricity load typicallyincreases with hotter air temperatures due to increasedair-conditioning usage [3, 40]. Across the continental
US, average daily maximum summertime tempera-tures are expected to increase by as much as 1 °C–5 °Cbymid-century [39]. In the absence of building energyefficiency gains and loadmanagement programs, thesehigher temperatures are likely to result in higher per-capita electricity loads for summer-peaking regions,where air conditioning use is closely tied to ambientair temperature. Wemodel potential increases to peakper-capita summertime electricity load for 121 uniqueplanning areas (comprising 1044 individual retailelectric service providers) in the United States basedon historical relationships between air temperatureand load. These planning areas account for about 650GW of summertime load (or about 80% of peaksummertime load in the United States). To estimatepotential per-capita load increases under future cli-mate scenarios, we first develop regression relation-ships between peak daily per-capita load andmaximum daily temperature using historical popula-tion, load and temperature data. Next, we force theseregression models with projected temperature datafrom theCMIP5multi-model GCMensemble.
Because average annual electricity load scales withpopulation [3], it is necessary to remove the effect ofpopulation growth before a relationship betweentemperature and peak load can be observed. To thisend, we estimate peak per-capita electricity load forplanning areas using reported hourly loads along withpopulation estimates from the US Census. Hourlyloads from 1993–2010 are determined for major elec-trical planning areas using Federal Energy RegulatoryCommission (FERC) Form 714 [41]. Next, populationestimates are generated for each planning area usingcensus tract-level population estimates for the years1990, 2000 and 2010 [42–44]. Census tracts are spa-tially joined to the planning areas that intersect them.We use census tract boundary files from the US Cen-sus Bureau’s Topologically Integrated GeographicEncoding and Referencing (TIGER/Line) database
Table 1.Global climatemodels selected from theCoupledModel Intercomparison Project Phase 5 (CMIP5)multi-model ensemble for usein this study. Themodeling group responsible for eachGCMmodel is shown on the leftmost column,while the representative concentrationpathway (RCP) scenarios included are shown in the rightmost column. The r1i1p1 ensemblemember is used for all GCMmodels.
Modeling group GCMmodel
RCP scenarios
included
CanadianCentre for ClimateModelling andAnalysis CanESM2 2.6, 4.5, 8.5
National Center for Atmospheric Research CCSM4 2.6, 4.5, 8.5
Community Earth SystemModel Contributors CESM1-BGC 4.5, 8.5
CentreNational de RecherchesMétéorologiques /Centre Européen deRecherche et
FormationAvancée enCalcul Scientifique
CNRM-CM5 4.5, 8.5
Commonwealth Scientific and Industrial ResearchOrganization in collaborationwith
QueenslandClimate Change Centre of Excellence
CSIRO-Mk3.6.0 2.6, 4.5, 8.5
NOAAGeophysical FluidDynamics Laboratory GFDL-ESM2G 2.6, 4.5, 8.5
NOAAGeophysical FluidDynamics Laboratory GFDL-ESM2M 2.6, 4.5, 8.5
Institut Pierre-Simon Laplace IPSL-CM5A-MR 2.6, 4.5, 8.5
Atmosphere andOceanResearch Institute (TheUniversity of Tokyo), National Institute forEnvironmental Studies, and JapanAgency forMarine-Earth Science andTechnology
MIROC5 2.6, 4.5, 8.5
Max Planck Institute forMeteorology MPI-ESM-LR 2.6, 4.5, 8.5
Max Planck Institute forMeteorology MPI-ESM-MR 2.6, 4.5, 8.5
5
Environ. Res. Lett. 11 (2016) 114008
[45–47], and retail electric service area boundariesfrom the HSIP GIS database [28], which are aggre-gated into their respective planning areas. The popula-tion in each planning area is taken as the sum of thecensus tract populations that intersect that planningarea. Per capita electricity load is then computed foreach planning area for the period 1993–2010 by com-bining population and hourly load data developed inthe previous steps. It should be noted that socio-eco-nomic factors—such as gross domestic product,household income, and electricity prices—can alsoinfluence electricity load [48–50]. When predictingtotal electricity load under future scenarios (particu-larly over a short time horizon), it is sometimes neces-sary to take these factors into account. In this study, wedo not account for changes in socio-economic indica-tors for two primary reasons: (1) these socio-economicindicators cannot be meaningfully projected over the35–85 year time horizon used in this paper, and (2)weare interested primarily in isolating the marginalincrease in per-capita electricity load due to long-termchanges in air temperature, rather than absolute futureelectricity load. In this formulation, socio-economicfactors appear as part of the variance in the temper-ature-load response function. Similarly, technologicalchanges—such as improved building energy efficiencyor market saturation of air conditioning units—canalso affect the temperature-load response function.Rising temperatures may, for instance, promote theadoption of air conditioning units in regions wheretheir usage was historically more limited [51]. Weacknowledge that changes beyond what we estimate inour model are possible, and discuss these sources ofuncertainty in greater detail in section 4.
Having estimated per-capita load for major plan-ning areas in the US, historical temperature data areused to construct temperature-load regressionequations, which are then forced with projected temp-erature data to determine the expected changes in peakper-capita summertime load under future climaticconditions. Gridded historical daily maximum temp-erature data are first assigned to each planning areausing a spatial join [36]. For planning areas with a sin-gle urban area, temperature data nearest to the cen-troid of the urban area are used. For planning areasthat contain multiple urban areas, the process of ren-dering a representative temperature cell takes place intwo steps: (1) gridded temperature data nearest to thecentroid of each urban area are collected, and (2) aweighted average temperature for the entire planningarea is generated, with the weights being determinedby the populations of each urban area. After determin-ing representative temperatures for each planning areafor the period 1993–2010, regression relationshipsbetween daily maximum temperature and daily peakload are developed for the summer peak period (June–August). Regression relationships take the form of aquadratic equation, with ambient daily maximumtemperature as the independent variable, and peak
per-capita electricity load as the dependent variable
a b g= ⋅ + ⋅ + ( )d T T , 6amb2
amb
where a, b and g are empirically determinedcoefficients. A quadratic equation is selected for theregression model because it provides a good fit formost planning areas (with an average R2 coefficient of0.6 over the 121 planning areas considered), andbecause higher-order models can result in very largeloads when extrapolated beyond the historical temper-ature range (SI section 2.4). Planning areas with an R2
coefficient below 0.5 are not included in the results. Itshould be noted that although this study develops aregression relationship between ambient temperatureand total per-capita load, not all of electricity load issensitive to changes in air temperature. The regressionrelationships developed for each planning area areforced with downscaled CMIP5 GCM projections ofdaily maximum temperature to determine theincrease in summertime peak per-capita load over the21st century [39].
3. Results
By mid-century (2040–2060), rising air temperaturesmay reduce summertime transmission capacity by1.9%–5.8% on average, relative to the 1990–2010reference period (with the range of impacts beingdependent on GCM model and RCP scenario selec-tion). These reductions in rated ampacity may posechallenges for transmission planning authorities, whotypically design peak period infrastructure based onhistorical conditions. Figure 1 showsmaps of expectedampacity reductions by decade under the ‘medium’
atmospheric carbon concentration scenario (RCP4.5). Impacts to transmission capacity vary with bothgeographic location and conductor technology.Ampacity reductions are typically largest in the FERCMidcontinent (MISO) electricity market, where aver-age summertime daily maximum temperatures arelikely to rise by as much as 2 °C–5 °C. By contrast,transmission lines experience smaller ampacity reduc-tions along the coastal regions, where temperatureincreases are more moderate. Although impacts varyby geographic region, almost all transmission lines inthe US are expected to experience ampacity reduc-tions. In addition to geographic location, conductorproperties also play a role in determining expectedclimatic impacts to transmission capacity. In general,high-voltage lines suffer greater capacity reductionsthan lower-voltage lines. This is because high-voltagelines are generally thicker, making heat dissipationmore difficult. While geographic and technologicalfactors play a role in determining the vulnerability oftransmission lines, reductions in rated ampacity aremost sensitive to increases in ambient air temperature.Figure 2 shows histograms of percent reduction inrated ampacity by decade for three carbon
6
Environ. Res. Lett. 11 (2016) 114008
Figure 1.Climate attributable reductions to transmission capacity by decade (from top left to bottom right: 2020, 2040, 2060, 2080).Colors indicate the percent reduction in transmission capacity under the average RCP 4.5 scenario, relative to the 1990–2010 referenceperiod. For additional information on transmission line visualizations, see SI section 4.6.
Figure 2.Histograms indicating percent reduction to transmission capacity by decade (from top left to bottom right: 2020, 2040,2060, 2080). RCP scenarios (2.6, 4.5 and 8.5) are indicated by their respective colors (blue, yellow and red). The horizontal axisrepresents bins of percent ampacity reductions, while the vertical axis represents the number of transmission lines falling into eachbin.
7
Environ. Res. Lett. 11 (2016) 114008
concentration scenarios—RCP 2.6, 4.5 and 8.5. Bymid-century, average transmission capacity reduc-tions range from 1.9%–3.9% under the lowest carbonconcentration scenario (RCP2.6) to 2.2%–4.3%underthemedium carbon concentration scenario (RCP 4.5),to 3.6%–5.8%under the highest carbon concentrationscenario (RCP 8.5). This outcome suggests thatimpacts to transmission ampacity can be mitigatedby policy efforts to reduce atmospheric carbonconcentrations.
Increases to peak per-capita electricity load pre-sent perhaps the greatest climate impact-related chal-lenge to electricity supply adequacy. By mid-century,peak per-capita summertime loads may rise by 4.2%–
15% on average due to increases in daily maximum airtemperature (with GCM and RCP selection account-ing for the range of variability). In the absence ofenergy efficiency gains and demand-side managementprograms, these load increases could narrow the bandof safe operating conditions, indicating greater gridvulnerability and reduced reliability. Figure 3 presentsmaps of summertime peak per-capita load by decadeunder the RCP 4.5 scenario, while figure 4 displaystime series data of summertime peak per-capita loadfor all RCP scenarios over the period 1990–2100.Impacts to peak load are strongly correlated with thedegree of air temperature rise. For the lowest carbonconcentration trajectory (RCP 2.6), increases to percapita load level out at roughly 4.2%–9.2% by mid-century, while under the highest carbon concentrationscenario (RCP 8.5), per capita loadsmay be 7.9%–15%
higher by mid-century, and may rise to as much as30% by 2100. As with impacts to transmission capa-city, climate-attributable load increases vary by geo-graphical location. However, unlike impacts totransmission capacity, peak load increases are notnecessarily correlated with the degree of temperaturerise. In general, urban areas show greater climate-attri-butable load increases than rural areas. Populousstates with large metropolitan areas—such as Cali-fornia, Illinois, Texas and New York—show some ofthe largest load increases, while sparsely populatedstates—such as Wyoming, Utah and Nevada—showrelatively small load increases, despite being located inregions that are expected to receive above-averagewarming. This effect could be attributed to greaterpenetration and density of air conditioning units inurban areas, where building proximity prevents pas-sive cooling. While increases to peak load vary byatmospheric carbon concentration and geographicregion, overall per capita peak load is expected toincrease for all scenarios considered. These climate-attributable increases to peak load are currently notaccounted for by many planning authorities, meaningthat planned peak period infrastructure may not opti-mallymeet future electricity needs.
4.Uncertainty assessment
Climate impacts to electricity supply adequacy may beinfluenced by several sources of uncertainty—includ-ing climate model variability, unknown conductor
Figure 3.Climate-attributable increases to peak per-capita summertime load for electric retail service providers in theUnited Statesby decade (from top left to bottom right: 2020, 2040, 2060, 2080) under the average RCP 4.5 scenario. Colors indicate the percentincrease in peak summertime load relative to the 1993–2010 reference period. For additional information on utility service areavisualizations, see SI section 4.6.
8
Environ. Res. Lett. 11 (2016) 114008
specifications, and future technological changes.Although these sources of uncertainty may affect thedegree to which climate change impairs electricityreliability, it is unlikely that a scenario will occur inwhich transmission capacity is not reduced, or inwhich per capita peak electricity load does notincrease. Variability in GCMmodel output representsthe largest source of uncertainty in this study. BetweenGCM models, median impacts to transmission capa-city range from about 1.7% (GFDL-ESM2M, RCP 2.6)to 5.6% (CANESM2, RCP 8.5) by mid-century, while
increases to electricity load range from about 4.3%(GFDL-ESM2M, RCP 2.6) to almost 15% (CSIRO-Mk3-6-0, RCP 8.5). Note that although both transmis-sion and load losses depend on the amount oftemperature rise, upper-bound impacts occur underdifferent GCM models. This result is explained byspatial variability in temperature rise between models:impacts to electricity load are dependent on temper-ature changes over major population centers (non-uniformly distributed), while high voltage transmis-sion lines are more evenly distributed, meaning that
Figure 4.Climate-attributable increases to peak summertime load byNERC region (from top left to bottom right:MAPP, SERC,ECAR, SPP,WECC, and the entire continental United States). The horizontal axis indicates the year (1990–2100), while the verticalaxis indicates the percent increase in peak per-capita summertime load relative to the 1993–2010 reference period. For reference, amap ofNERC regions is shown in SI section 4.7. Individualmodel runs are shown in SI section 4.8, alongwith time series plots of allNERC regions.
9
Environ. Res. Lett. 11 (2016) 114008
impacts to transmission are more reflective of overallcontinental temperature rise. Although impacts varywidely between GCMmodels, this variability does notaffect the primary conclusions of the study: all modelspredict overall reductions to transmission capacityand overall increases to peak electricity load. To ensurethat the predicted impacts are not an artifact of GCMmodel variability, we test for the significance of ourresults using a Wilcoxon rank-sum test (see SI section2.1). Transmission capacity and per capita electricityload are both found to be significantly affected bychanges in climate (p<0.001) for all GCM modelsandRCP scenarios considered.
In addition to GCM model uncertainty, we alsoassess several secondary sources of uncertainty,including (1) choice of representative conductorcables, (2) choice of maximum allowable conductortemperature, (3) choice of temperature-load regres-sion model, and (4) potential for future changes inwind speed. Conductor model choice has a relativelyminor effect on climate-attributable ampacity reduc-tions (<1%), and is unlikely to affect the primaryresults of this study (see SI section 2.2). Maximumallowable conductor temperature, on the other hand,can significantly affect temperature-related ampacityreductions (see SI section 2.3). Under the most con-servative maximum conductor temperature (75 °C),ampacity reductions range from 1.9%–5.8% by mid-century, while for a more permissive maximum con-ductor temperature (100 °C), climate-attributableampacity reductions are on the order of 1.1%–3.4%.Ultimately, we select a maximum conductor temper-ature of 75 °C for ACSR cable, because it is the indus-try standard for continuous operation under normalconditions [35]. However, it should be noted thatsome climate impacts to transmission capacity may beavoided by allowing transmission lines to run at highertemperatures, or by installing more heat-resistantcables (such as ACSS). Because ACSS conductors aredesigned to withstand temperatures of 250 °Cwithoutloss of strength [52], the effect of ambient temperaturerise on these cables is comparatively small. In additionto assessing uncertain conductor parameters, we alsoassess the sensitivity of results to different temper-ature-load regression models (see SI section 2.4). Weinvestigate three regression models (linear, quadraticand cubic) and find that the choice of regressionmodel can significantly influence climate-attributableincreases to peak electricity load. Impacts are fairlysimilar under the linear and quadratic regressionmodels: by mid-century, summertime peak load isexpected to increase by about 4.2%–13% under thelinear regression model, and about 4.2%–15% underthe quadratic regression model. However, the cubicregression model predicts much higher impacts thanthe other regression models, with expected peak loadincreases of roughly 11%–40%. This outcome followsfrom the fact that high-order polynomial models tendto predict large loads when extrapolated beyond the
historical temperature range. All models offer similargoodness of fit, with average R2 values of 0.58, 0.6 and0.6 for the linear, quadratic and cubic models, respec-tively. However, the linear model shows bias at theupper and lower ends of the temperature spectrum,where the temperature-load response function tendsto show nonlinear behavior. Ultimately, we select thequadratic regressionmodel because (i) it offers a bettergoodness of fit than the linear model, (ii) it captureshistorical temperature-load profiles over the entirerange of expected temperatures and (iii) it is less sus-ceptible to extrapolation errors than the cubic model.Conductor ampacity is strongly dependent on windspeed, suggesting that future changes in wind speedcould potentially influence anticipated ampacityreductions. However, based on an assessment of pro-jected wind speeds in the CMIP3multi-model ensem-ble, we find that wind speeds are not expected tochange significantly as a result of climate change (SIsection 2.5). It should be noted, however, that theeffects of climate change on wind speed are a subject ofdebate, with various studies predicting global trends ofincreasing wind speed [53], global trends of decreasingwind speed [54], 1.4%–4.5% reductions in wind speedthroughout the US over the next hundred years [55],decreased wind speed throughout the Western USHigh Plains Region [56], no detectable changes in USwind speed over the next half-century [57], and smallmagnitude changes (±0.1 m s−1) of uncertain signthroughout theUS [58].
Changes in social, economic and technologicalfactors may also affect the degree to which climatechange will impact overall electricity supply adequacy.Electricity load, for instance, is influenced by a host ofunderlying drivers, including socio-economic factors,residential behaviors, and electricity use for commer-cial and industrial activities. If these underlying driversof electricity consumption change, then it is possiblethat peak loadwill change as well. Future technologicalchanges—like conductor upgrades and buildingenergy efficiency improvements—represent anothermajor source of uncertainty (SI section 2.6). However,unlike GCM models or conductor cable parameters,technological changes are subject to economic, socialand political considerations, and are therefore difficultto quantify. Upgrading existing ACSR cable to heat-resistant ACSS cable offers the potential to offsetnearly all climate-attributable ampacity losses. How-ever, it is unclear whether upgrading existing ACSRcables is economically or politically viable. Similarly, areview of the literature reveals that upgrades to build-ing energy efficiency and demand-side managementprograms have the potential to offset virtually all cli-mate-attributable increases to electricity load (SIsection 2.6)—however, it is unclear to what extentthese gains will actually be realized. While social andtechnological changes could have a significant impacton the results of this study, it is difficult to assess whatlevel of change is likely to occur. Thus, rather than
10
Environ. Res. Lett. 11 (2016) 114008
attempting to characterize the effects of social andtechnological changes as a source of uncertainty, werecommend viewing these changes as a potential miti-gation strategy—one that future research shouldaddress in detail.
5.Discussion
While climate change may present challenges to peak-period electric power infrastructure, impacts can beoffset through carbon mitigation initiatives, re-con-ductoring of congested transmission corridors, build-ing energy efficiency improvements, and new ‘smart-grid’ technologies. Perhaps the most striking result ofthis study is the degree to which infrastructure impactsare correlated with atmospheric carbon concentra-tion. Impacts to mid-century transmission capacityunder the high emissions scenario are about 74%higher than those under the low emissions scenario,while impacts to peak electricity load are about 71%higher. The wide range of impacts between carbonconcentration trajectories suggests that policies limit-ing carbon emissions can go a long way to preventingclimate impacts to electricity reliability. Technologicalimprovements also offer the potential to mitigateclimate impacts. Upgrades to overhead conductors(such as replacing ACSR cables with more temper-ature-resistant ACSS cables) have the potential tolessen climate-attributable reductions to transmissioncapacity. On the demand side, investments in buildingenergy efficiency and demand response programscould offset load increases due to higher air tempera-tures. Although these changes are difficult to predict inany quantitative capacity, a review of the literatureindicates that expansions to demand-side manage-ment programs have the potential to reduce futureelectricity load by 4.0%–27% relative to a ‘business-as-usual’ scenario (SI section 2.6.1). These reductions arecomparable to climate-attributable increases in elec-tricity load (4.2%–15%), meaning that investments inbuilding energy efficiency and demand responseprograms could effectively negate the adverse effects ofclimate change on the grid.
In addition to conventional best practices, pro-posed ‘smart grid’ technology may enable the elec-tricity grid to respond more flexibly to climate-constrained conditions (SI section 2.6.2). Smart griddevelopment couples transmission and distributioninfrastructure with information and communicationtechnologies to enable advanced power system ser-vices [59], including wide-areamonitoring systems forlarge-scale reliability and security assessment,dynamic electricity pricing schemes based on demandresponse, and the distributed deployment of energystorage, electric vehicle, and renewable power genera-tion resources. Power line temperature and sag mon-itoring systems can be used with advanced actuatorand electricity routing infrastructure to direct
electricity down lines that are less stressed by changingtemperature in peak conditions [60]. Moreover,through dynamic ampacity systems, it is possible forthermal limits to better reflect operating conditions byusing real-time electrical and environmental data todetermine the maximum allowable current for a givenpower line [60]. This practice enables greater flexibilityin how power lines are used to deliver electricity and inplanning infrastructure maintenance and upgrades.Smart grid technologies may also help to soften theimpact of climate change on electricity load. In futuresmart grids, distribution systems may monitor andcommunicate with smart appliances and HVAC sys-tems to control their operation in temperature con-strained conditions [61]. For example, to prevent loadshedding and infrastructure failures during peak loadhours, electric service providers may change the set-tings on smart A/Cunits to reduce load within a singlehome or acrossmultiple buildings within the same cir-cuit. While rising air temperatures may lead todecreased transmission capacity and increased peakloads, impacts to overall electricity reliability can beavoided with the proper mitigation measures in place.Investments in carbon mitigation strategies, demand-sidemanagement programs, and smart-grid technolo-gies may lessen or even negate anticipated climateimpacts.
The results of this study show that increasingambient air temperatures may increase electricitydemand and decrease transmission ampacity. Addi-tionally, a complementary study on electrical powergeneration predicts reductions in instantaneous gen-eration capacity in hot summer months for the Wes-tern Electricity Coordinating Council (WECC) region[1], which comprises 14 states in the Western US [62].Bymid-century, existing vulnerable summertime gen-erating capacity in the WECC region may be reducedby 1.1%–3.0% in an average year, with proposed capa-city (mostly from combustion turbine and solar pho-tovoltaic sources) suffering similar losses [1]. Inaddition to these losses, increases in ambient air temp-erature may increase peak per-capita load in theWECC region by as much as 1.7%–14%, whileWECCtransmission capacity may be reduced by as much as1.4%–5.6%. This suggests that the joint effect ofdecreases in generation capacity, decreases in trans-mission ampacity, and increases in peak load mayresult in amore constrained electricity grid as ambienttemperatures rise. This effect can differ betweenregions with a summer peak and regions with a winterpeak [1, 13]. Regions with a summer peak may havereduced ampacity and increased load (e.g., due toincreases in loads from central air conditioners)whereas regions with a winter peak may haveincreased ampacity and decreased load (e.g., due todecreases in loads from radiative heating). This high-lights the need to better understand the impacts of cli-mate change and subsequent rises in ambienttemperature on overall electricity system reliability,
11
Environ. Res. Lett. 11 (2016) 114008
both at a general level and for site-specific locations.Currently, some regional electricity supply adequacyassessments do not explicitly account for the effects ofclimate change [63], meaning that long-term invest-ments in electricity infrastructure may not optimallymeet future electricity needs. Site-specific models areneeded to explore the combined effects of temperaturerise on electricity generation, transmission, distribu-tion, and demand. Further, more work is needed toassess the asynchronous impacts of network conges-tion and market forces on interconnection-scale elec-tricity supply adequacy. While electricity supplyadequacy is adversely affected under all the climatescenarios considered, the results offer several optionsfor alleviating potential impacts, including carbonmitigation initiatives, heat-resistant conductors,building energy efficiency upgrades, and new smart-grid technologies. These measures offer the potentialto lessen or even completely offset potential climateimpacts, depending on how aggressively they are pur-sued. While the effects of climate change on electricpower infrastructure could be considerable, this studyshows that these impacts can be overcome with exist-ing technologies and management techniques, giventhat anticipatory governance practices can beinstituted.
Acknowledgments
This material is based on work supported by theNational Science Foundation (grant numbers WSC1360509, IMEE 1335556, and RIPS 1441352). Thiswork was supported in parts by the USACE MilitaryPrograms. Permission was granted by the authority oftheUSACEChief of Engineers to publish thismaterial.The views and opinions expressed in this article arethose of the individual authors and not those of theU.S. Army or other sponsor organizations.
References
[1] BartosMDandChesterMV2015 Impacts of climate changeon electric power supply in theWesternUnited StatesNat.Clim. Change 5 748–52
[2] VanVlietMTH, Yearsley J R, Ludwig F, Vögele S,LettenmaierDP andKabat P 2012Vulnerability ofUS andEuropean electricity supply to climate changeNat. Clim.Change 2 676–81
[3] Sathaye J,Dale L, LarsenP, FittsG,KoyK, Lewis S andLucenaA2011Estimating risk toCaliforniaEnergy Infrastructure fromProjectedClimateChange,CECPublicationCEC-500-2012-057(www.energy.ca.gov/2012publications/CEC-500-2012-057/CEC-500-2012-057.pdf)
[4] Dirks J A et al 2015 Impacts of climate change on energyconsumption and peak demand in buildings: a detailedregional approachEnergy 79 20–32
[5] ParryML (ed) 2007Climate Change 2007—Impacts,Adaptation andVulnerability:WorkingGroup II Contributionto the Fourth Assessment Report of the IPCC vol 4 (Cambridge:CambridgeUniversity Press) (www.ipcc.ch/pdf/assessment-report/ar4/wg2/ar4_wg2_full_report.pdf)
[6] HartoCB et al 2012Analysis of drought impacts on electricityproduction in theWestern and Texas interconnections of theUnited StatesANL/EVS/R-11/14
[7] Kimmell TA andVeil J A 2009 Impact of drought onUS steamelectric power plant coolingwater intakes and related waterresourcemanagement issuesDOE/NETL-2009/1364
[8] KochH andVögele S 2009Dynamicmodelling of waterdemand, water availability and adaptation strategies for powerplants to global changeEcolog. Econ. 68 2031–9
[9] BartosM andChesterM2016Methodology for estimatinggenerating capacity losses in thermoelectric power facilitieswith recirculating cooling systems under climate changeASUDigital RepositoryASU-SSEBE-CESEM-2016-RPR-003(https://repository.asu.edu/items/40430)
[10] Dubey S, Sarvaiya J and Seshadri B 2013Temperaturedependent photovoltaic (PV) efficiency and its effect on PVproduction in theworld—a review Energy Proced. 33 311–21
[11] Janko S, ArnoldMand JohnsonN2016 Implications of high-penetration renewables for ratepayers and utilities in theresidential solar photovoltaic (PV)marketAppl. Energy 18037–51
[12] Energy Information Administration 2016EIA Form 861:Electric power sales, revenue, and energy efficiency (https://eia.gov/electricity/data/eia861/) (Accessed: September 2016)
[13] Parkinson S andDjilali N 2015Robust response to hydro-climatic change in electricity generation planningClim.Change 130 475–89
[14] Linkov I et al 2014Changing the resilience paradigmNat. Clim.Change 4 407–9
[15] Roege P E,Collier ZA,Mancillas J,McDonagh J A and Linkov I2014Metrics for energy resilience Energy Policy 72 249–56
[16] PJM Interconnection 2010Guide for Determination of BareOverhead Transmission Conductors (http://pjm.com/
~/media/planning/design-engineering/maac-standards/bare-overhead-transmission-conductor-ratings.ashx)(Accessed:May 2015)
[17] PJM Interconnection 2004Outdoor Substation ConductorRatings (http://pjm.com/~/media/planning/design-engineering/maac-standards/outdoor-substation-conductor-ratings.ashx) (Accessed:May 2015)
[18] RosenthalDH,GruenspechtHK andMoran EA 1995 Effectsof global warming on energy use for space heating and coolingin theUnited States Energy J. 16 77–96
[19] ZhouY, Eom J andClarke L 2013The effect of global climatechange, population distribution, and climatemitigation onbuilding energy use in theUS andChinaClim. Change 119979–92
[20] ZhouY et al 2014Modeling the effect of climate change onUSstate-level buildings energy demands in an integratedassessment frameworkAppl. Energy 113 1077–88
[21] Hadley SW, EricksonD J,Hernandez J L, BroniakCT andBlasing T J 2006Responses of energy use to climate change: aclimatemodeling studyGeophys. Res. Lett. 33 L17703
[22] Mansur ET,MendelsohnROandMorrisonW2005Adiscrete-continuous choicemodel of climate change impacts onenergy, SSRNYale SOMWorking PaperNo. ES-43 (abstractnumber 738544),submitted to J. Environ.Manag.
[23] McFarland J et al 2015 Impacts of rising air temperatures andemissionsmitigation on electricity demand and supply in theUnited States: amulti-model comparisonClim. Change 131111–25
[24] AmatoAD, RuthM,Kirshen P andHorwitz J 2005Regionalenergy demand responses to climate change:methodology andapplication to the commonwealth ofmassachusettsClim.Change 71 175–201
[25] FrancoG and SanstadAH2007Climate change and electricitydemand inCaliforniaClim. Change 87 139–51
[26] USEnergy InformationAdministration 2011 Fig. ofWesternU.S. net regional power flows fromFERC’s Form714Aquarterof California’s electricity comes from outside the state (http://eia.gov/todayinenergy/detail.cfm?id=4370) (Accessed:July 2015)
12
Environ. Res. Lett. 11 (2016) 114008
[27] AllenM, Fernandez S, Fu J andOlamaM2016 Impacts ofclimate change on sub-regional electricity demand anddistribution in the southernUnited StatesNat. Energy 1 16103
[28] Homeland Security Infrastructure Program 2014RetailElectric Service Territories Shapefile., United StatesDepartment ofHomeland Security) (CDROM)
[29] NorthAmerican Electric Reliability Corporation 2012 2012Long-TermReliability Assessment (http://nerc.com/news/Headlines%20DL/2012%20Long%20Term%20Reliability%20Assessment.pdf) (Accessed: June 2015)
[30] IEEE Standard forCalculating theCurrent-TemperatureRelationship of BareOverheadConductors 2013 IEEE Std 738-2012 (Revision of IEEE Std 738-2006—Incorporates IEEE Std738-2012 Cor 1-2013) (doi:10.1109/ieeestd.2013.6692858)
[31] Homeland Security Infrastructure Program2014 ElectricTransmission Lines Shapefile, United StatesDepartment ofHomeland Security (CDROM)
[32] Southwire Company 2012 Product Specification—AluminumConductor, Steel Reinforced (http://southwire.com/
ProductCatalog/XTEInterfaceServlet?contentKey=prodcatsheet16) (Accessed: April 2015)
[33] OmanCables Industry 2011 Overhead Line Conductors(http://omancables.com/brochure/Overhead_Line_Conductor.pdf) (Accessed: April 2015)
[34] NexansOlexNewZealand 2012 PowerCable CatalogueConductor characteristics and product sheets (http://nexans.co.nz/NewZealand/2013/Power%20Cable%20Catalogue%20Full%20version%202012.pdf) (Accessed: April 2015)
[35] HouseHE andTuttle PD 1958Current-carrying capacity ofACSRTrans. Am. Inst. Elect. Eng. III 77 1169–73
[36] Maurer E P,WoodAW,Adam JC, LettenmaierDP andNijssen B 2002A long-termhydrologically based dataset ofland surfacefluxes and states for the conterminousUnitedStates J. Clim. 15 3237–51
[37] ShepardD 1984Computermapping: the SYMAPinterpolation algorithm Spatial Stat.Mod. 133–45
[38] WidmannMandBrethertonC 2000Validation ofmesoscaleprecipitation in theNCEP reanalysis using a new gridcelldataset for the northwesternUnited States J. Clim. 13 1936–50
[39] Reclamation 2013 DownscaledCMIP3 andCMPI5ClimateProjections: Release ofDownscaledCMIP5ClimateProjections, Comparisonswith Preceding Information, andSummary ofUserNeeds. Denver, CO:USDepartment of theInterior, Bureau of Reclamation, Technical Service Center(http://gdo-dcp.ucllnl.org/) (Accessed:May 2015)
[40] California Energy Commission 1999High Temperatures andElectricity Demand: AnAssessment of Supply Adequacy inCalifornia (Sacramento, CA: theCalifornia EnergyCommission staff) (www.energy.ca.gov/reports/1999-07-23_HEAT_RPT.PDF)
[41] Federal Energy Regulatory Commission 2015 FormNo. 714—Annual Electric Balancing Authority Area and Planning AreaReport, 1993-2013 (Data set). RetrievedMay 2015 from(http://ferc.gov/docs-filing/forms/form-714/data.asp)
[42] USCensus Bureau 2015 Population by census tract, 1990 (Dataset)Prepared by Social Explorer (https://socialexplorer.com/
explore/tables) (Accessed:May 2015)[43] USCensus Bureau 2015 Population by census tract, 2000 (Data
set)Prepared by Social Explorer (https://socialexplorer.com/
explore/tables) (Accessed:May 2015)
[44] USCensus Bureau 2015 Population by census tract, 2010 (Dataset)Prepared by Social Explorer (https://socialexplorer.com/explore/tables) (Accessed:May 2015)
[45] USCensus Bureau 2015 2014 TIGER/Line Shapefiles—CensusTract Geometries (Data set) (ftp://ftp2.census.gov/geo/tiger/TIGER2014/) (Accessed:May 2015)
[46] USCensus Bureau 2015 2000 TIGER/Line Shapefiles—CensusTract Geometries (Data set) (ftp://ftp2.census.gov/geo/tiger/tiger2k/) (Accessed:May 2015)
[47] USCensus Bureau 2015 1992 TIGER/Line Shapefiles—CensusTract Geometries (Data set) (ftp://ftp2.census.gov/geo/tiger/TIGER1992/) (Accessed:May 2015)
[48] Cayla J-M,MaiziN andMarchandC 2011The role of incomein energy consumption behaviour: evidence fromFrenchhouseholds data Energy Policy 39 7874–83
[49] Wyatt P 2013Adwelling-level investigation into the physicaland socio-economic drivers of domestic energy consumptionin England Energy Policy 60 540–9
[50] LiseW andVanMontfort K 2007 Energy consumption andGDP inTurkey: is there a co-integration relationship?EnergyEcon. 29 1166–78
[51] SailorD J and Pavlova AA 2003Air conditioningmarketsaturation and long-term response of residential coolingenergy demand to climate changeEnergy 28 941–51
[52] Southwire Company 2012 Product Specification—AluminumConductor, Steel Supported. Bare (http://southwire.com/
ProductCatalog/XTEInterfaceServlet?contentKey=prodcatsheet28) (Accessed:May 2015)
[53] Young I R, Zieger S andBabanin AV 2011Global trends inwind speed andwave height Science 332 451–5
[54] McVicar TR et al 2012Global review and synthesis of trends inobserved terrestrial near-surface wind speeds: implications forevaporation J. Hydrol. 416–417 182–205
[55] BreslowPB and Sailor D J 2002Vulnerability of wind powerresources to climate change in the continental United StatesRenew. Energy 27 585–98
[56] Greene J S, ChatelainM,MorrisseyMand Stadler S 2012Projected future wind speed andwind power density trendsover thewesternUShigh plainsAtmos. Clim. Sci. 02 32–40
[57] Pryor SC andBarthelmie R J 2011Assessing climate changeimpacts on the near-term stability of thewind energy resourceover theUnited StatesProc. Natl Acad. Sci. 108 8167–71
[58] Eichelberger SC J, Nijssen B andWoodA 2008Climate changeeffects onwind speedN.Am.Windpower 7 68–72
[59] FangX,Misra S, XueG andYangD 2012 Smart grid—the newand improved power grid: a survey IEEECommun. Surv.Tutorials 14 944–80
[60] DebAK2000Powerline Ampacity System: Theory,ModelingandApplications (Boca Raton, FL: CRC)
[61] NiyatoD, Xiao L andWang P 2011Machine-to-machinecommunications for home energymanagement system insmart grid IEEECommun.Mag. 49 53–9
[62] USDepartment of Energy 2015 Learnmore aboutinterconnections (http://energy.gov/oe/services/electricity-policy-coordination-and-implementation/transmission-planning/recovery-act-0) (Accessed: July 2015)
[63] NorthAmerican Electric Reliability Corporation 2015 2015Long-TermReliability Assessment (http://nerc.com/pa/RAPA/ra/Reliability%20Assessments%20DL/2015LTRA%20-%20Final%20Report.pdf) (Accessed: September 2016)
13
Environ. Res. Lett. 11 (2016) 114008
