Transportation Research Part D 17 (2012) 380–388
Contents lists available at SciVerse ScienceDirect
Transportation Research Part D
journal homepage: www.elsevier .com/ locate / t rd
Life cycle assessment of pavement: Methodology and case study
Bin Yu ⇑, Qing LuCivil and Environmental Engineering, University of South Florida, 4202 E. Fowler Avenue, Tampa, FL 33620, United States
a r t i c l e i n f o
Keywords:Pavement designInfrastructure life cycle assessmentPavement overlay systems
1361-9209/$ - see front matter � 2012 Elsevier Ltdhttp://dx.doi.org/10.1016/j.trd.2012.03.004
⇑ Corresponding author.E-mail address: [email protected] (B. Yu).
a b s t r a c t
A life cycle assessment model is built to estimate the environmental implications ofpavements using material, distribution, construction, congestion, usage, and end of lifemodules. A case study of three overlay systems, Portland cement concrete overlay, hotmixture asphalt overlay, and crack, seat, and overlay, is presented. The case leads to thefollowing conclusions. It is reasonable to expect less environmental burdens from thePortland cement concrete and crack, seat, and overlay options as opposed to hot mixtureasphalt while although the results have a high degree uncertainties. The material,congestion, and particularly usage modules contribute most to energy consumption andair pollutant. Traffic related energy consumption and greenhouse gases are sensitive totraffic growth and fuel economy improvement. Uncertainties exist in the usage module,especially for the pavement structure effect.
� 2012 Elsevier Ltd. All rights reserved.
1. Introduction
Life-cycle analysis (LCA) in pavement assessment is still at an immature stage. Typically, a LCA model of pavement con-sists of the following components: material, construction, use, maintenance and rehabilitation (M&R), and end of life (EOL).However, most of the work does not incorporate all the components (Chan, 2007). Two most important elements in LCAmodels, usage and traffic congestion resulted from construction and M&R activities are typically ignored due to great com-plexity. Huang et al. (2009) suggested that additional fuel consumptions and pollutant emissions due to traffic delay duringroadwork periods are significant. Even for studies that incorporate the use and congestion phases, there is still room toimprove (Keoleian et al., 2005; Zhang et al., 2010): first, data in some studies are outdated; second, the usage phase isnot complete; third, the EOL phase is simply taken as landfill while practically, most hot mixture asphalt (HMA) is recycledand old Portland cement concrete (PCC) is crushed to substitute base course aggregates.
2. Methodology
We begin by defining a functional unit needed to build the LCA model framework. A functional unit quantifies a standardamount to be compared between alternatives that serve this function. Equivalent functionality shall be maintained for allcandidates of a LCA model. For pavement, this means that they should serve the same traffic over the same analysis spanwith the same performance. To assess the environmental impacts of pavement, a system of LCA model is developed, asshown in Fig. 1. The LCA functionality is fulfilled by six components, including material module, distribution module, con-struction module, congestion module, usage module, and EOL module, with various supplementary models attached to thecorresponding modules.
. All rights reserved.
Material Module
Distribution Module
Construction Module
Congestion Module
Usage Module
End of life Module
Life Cycle Assessment Model
Reference Model
GREET Model
NONROAD Model
QuickZone Model
Miscellaneous Models
Mobile Model
Model Parameters and User Inputs
Energy Consumption and Environmental Burden
Fig. 1. Relationship among various components.
B. Yu, Q. Lu / Transportation Research Part D 17 (2012) 380–388 381
Material consumption is modeled, with data from various sources including the Portland Cement Association (Marceau etal., 2007), the Swedish Environmental Research Institute (Stripple, 2001), and the Athena Institute (2006) and covers, energyconsumption and discharged environmental pollutants, including carbon dioxide (CO2), carbon monoxide (CO), methane(CH4), nitrogen oxide (NOx), sulfur oxide (SOx), volatile organic compound (VOC), particulate matter (<10 lm) (PM10), etc.
The distribution module is closely linked to the material module and the EOL module. All materials, equipment, andwastes are transported by a combination of road, rail, and waterway. Greenhouse Gases, Regulated Emissions, and EnergyUse in Transportation (2010) is used to model greenhouse gas emissions and energy embracing a data for fuel and electricityproduction, truck transportation, tie and dowel bar production, and natural gas burned that may be used during the pave-ment life time. Emission data for all non-road construction and vehicular equipment are obtained from the US Environmen-tal Protection Agency’s (EPA) NONROAD 2008 model for construction and maintenance activities. For each piece ofconstruction equipment, an estimate of the engine horsepower is made on the basis of one or two typical machines.NONROAD2008 model provides emission factors for various ranges of horsepower.
Most prior work has not included a EOL module because the pavement structure is assume to have an indefinite life. It isdesired to investigate the role of EOL module on the LCA model. Environmental burdens to dismantle and transport the oldpavement, the environmental savings due to the reuse of old pavement materials, and the potential additional energy con-sumption to process the old pavement materials before they can be used, need to be identified. They can be modeled in asimilar fashion as the way of material, distribution, and construction modules.
Traffic delay induced by construction and rehabilitation activities has significant influences on energy consumption andpollutant emissions compared with those under normal vehicular operations. The changes in traffic flow, traffic delay, andqueue length are estimated using the QuickZone model. Outputs include detour rate, queue length and speed reductionwithin work zones. Once vehicle delays due to construction and maintenance events are determined, they are coupled withfuel consumption and vehicle emissions to measure their environmental impacts. CO2 is calculated by the fuel consumptions(Emission Facts, 2005). Other vehicle emissions are calculated using US EPA’s MOBILE 6.2 model, which supplies the tailpipeemissions and evaporative emissions at varying traffic speeds on a per year basis through 2050.
Fuel consumptions and environmental burdens are calculated as the differences between those of construction and reha-bilitation periods and those of normal operations:
Ytotal ¼ VMTqueue � Yqueue þ VMTworkzone � Yworkzone þ VMTdet our � Ydet our � VMTnormal � Ynormal ð1Þ
where Yi is the value of different environmental indicators, such as fuel usage (L/km) or emission values (g/km), VMTi is themiles traveled by vehicles (km or mile), i is scenario index, representing the total, waiting in queue, passing through workzone, taking detour, or operating under normal conditions.The usage module focuses on the fuel consumption and pollutant emissions due to vehicle operations within the analysisperiod; roughness effect, pavement structure effect, albedo, and carbonation, are investigated.
Three major factors pose great influences on the LCA inventory, including: traffic volume, fuel economy, and pavementroughness. The fuel economy is derived directly from Vision model, which provides the fuel economies for passenger carsand trucks until 2100. In the LCA model, a fleet on-road average fuel economy is used.
Increasing pavement roughness causes more vibrations and reduces driving speed, and thus increases fuel consumptionand pollutant emissions of vehicles. An international roughness index (IRI, m/km) is generally used to describe the level ofroughness, which is a ratio of a vehicle’s accumulated vertical movement and the vehicle distance traveled during the mea-surement. Increase in IRI reduces fuel economy, a relationship found by Amos (2006) in Missouri where fuel economy rosefrom 21.30 mpg to 21.47 mpg for gasoline powered cars, and from 5.91 mpg to 6.11 mpg for diesel trucks as the IRI was ame-liorated from 2.03 m/km to 0.95 m/km. Thus, a fuel consumption factor (FCF) is used to describe real fuel consumptions ofvehicles driving on pavements with different IRIs:
FCF ¼ 7:377� 10�3IRI þ 0:993 for passenger cars
FCF ¼ 2:163� 10�2IRI þ 0:953 for trucksð2Þ
382 B. Yu, Q. Lu / Transportation Research Part D 17 (2012) 380–388
Besides the influence on fuel economy, increase of IRI reduces the driving speed and thus leads to a reduction of highwaycapacity. The speed-reduced fleet may witness significant fuel consumption increase and pollutant emission changes.Moreover, additional roughness causes increased friction and vertical acceleration of the vehicle body, and thus leads tomore vehicle fuel consumption and pollutant emissions. How these factors contribute to the life cycle inventory will beaddressed by the case study in detail.
Pavement structures have significant influence on the fuel consumptions of vehicles, especially for asphalt compared withPCC and composite pavements (Taylor et al., 2000). Taylor and Patten’s (2006) study suggested that PCC and composite pave-ments have significant fuel economy advantages over HMA pavement.
Albedo directly contributes to global cooling by adjusting the radiative forcing of the earth’s surface. As a surface covering,pavements can reflect a portion of the incoming solar radiation back into space, thus adjusting the global energy balance.Akbari et al. (2008) estimated that for every square meter, 2.55 kg of emitted CO2 is offset for every 0.01 increase in albedodue to increased radiative forcing. Eq. (3) gives the means to calculate the benefit:
DmCO2 ¼ 100� C � A� Da ð3Þ
where DmCO2 is the mass equivalents of CO2 mitigated (kg), C is the CO2 offset constant (kg CO2/m2), A is the area of pavement(m2) and Da is the change is albedo.
Over time, much of the CO2 that was originally liberated from limestone during cement kiln processes will rebind itself tothe cement in the pavement through the carbonation process. The carbonation of concrete can be modeled using a simpli-fication of Fick’s second law of diffusion (Lagerblad, 2006):
dc ¼ kffiffi
tp
ð4Þ
where dc is the depth of carbonation (mm), k is the rate factor (mm/y1/2) and t is time (year).Not all of the calcium in the concrete, however, is expected to bind CO2 molecules; the binding efficiency is suggested to
be roughly 75% (Stolaroff et al., 2005). The mass of CO2 sequestered is given by:
mCO2 ¼ dc � A� qconcrete �mcement=concrete �MCO2
MCaO� e ð5Þ
where mCO2 is the mass of CO2 sequestered through carbonation (kg), dc is the depth of carbonation (m), A is the surface areaof pavement (m2), qconcrete is the density of concrete (kg/m3), mcement/concrete is the mass ratio of cement in concrete, MCO2 is themolar mass of CO2, MCaO is the molar mass of CaO and e is the binding efficiency of CO2 to CaO.
3. Case study of three pavement overlay systems
3.1. The study
For the case study we consider an old PCC pavement that is at the end of its service life and requires rehabilitation torestore the serviceability. The existing base course is assumed to perform well and can function without intensive mainte-nances. This pavement has a 225 mm PCC surface with 250 mm crushed aggregate as base course, and subgrade. In eachdirection, the widths of the inner paved shoulder, main lanes, and outsider paved shoulder are 1.2 m, 3.6 � 2 m, and2.7 m. There is an annual average daily traffic flow (AADT) of 70,000, with 8% being truck that is growing at growth of 4%a year. Three replacement options frequently adopted in Florida are considered:
� Remove and replace the 225 mm PCC pavement with 250 mm new PCC (the PCC option). Diamond grinding is frequentlyused to restore surface smoothness and reported to be viable for 16–17 years (Stubstand et al., 2005) and thus is per-formed every 16 years as a periodic rehabilitation strategy.� Remove and replace the existing pavement with 225 mm HMA (the HMA option). Use a mill-and-fill (remove 45 mm
HMA surface and replace the same depth of new HMA) plan every 16 years as a periodic rehabilitation strategy (Weilandand Muench, 2010).� Crack, seat, and overlay (the CSOL option). Crack and seat the existing PCC pavement and then overlay with 125 mm HMA.
Use the same mill-and-fill plan as the periodic rehabilitation strategy every 16 years (Weiland and Muench, 2010).
The pavement overlay designs follow the Florida Department of Transportation (FDOT) pavement design manual as ver-ified by the Mechanistic-empirical Pavement Design Guide (MEPDG) software using Florida local weather data. Thus thefunctional unit is a one kilometer overlay system over an existing PCC pavement with four lanes in two directions that wouldprovide satisfactory performance over a 40-year period.
For material module, cement concrete production uses data from the Portland Cement Association (Marceau et al., 2007)while HMA production uses data from the Swedish Environmental Research Institute (Stripple, 2001). Distribution and con-struction modules are estimated based on the quantities of construction and maintenance activities.
For EOL module, reclaimed concrete material (RCM) is frequently used to substitute aggregates in base course. Reclaimedasphalt pavement (RAP) is now routinely accepted in asphalt paving mixtures with substitution rates ranging from 10% to
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
IRI (
m/k
m)
Age (Year)
PCC option HMA option CSOL option
Fig. 2. Development trends of IRI as predicted from MEPDG.
Table 1Fuel economy comparisons of two pavement structures to PCC.
Season Winter Spring Fall Winter
Pavement type HMA CSOL HMA CSOL HMA CSOL HMA CSOL
For passenger carsComparison (%) 3.1 �2.07 �0.42 1.2 �0.42 1.2 �0.42 1.2
For trucksComparison (%) 0.86 2.0 1.58 0.9 1.6 �1.34 1.6 �1.34
B. Yu, Q. Lu / Transportation Research Part D 17 (2012) 380–388 383
50% or more, depending on state specifications. In EOL module, two scenarios are tested. One is to crush PCC pavement anduse 10% and 20% of the crushed materials in base course layer. The other is to recycle the milled asphalt mixture into theasphalt drum plant with a portion of 10% and 20%. RAP are treated as free of any inherent or feedstock energy, whereas,in fact, it retains its feedstock value indefinitely. We include the energy used to extract RAP (roadway milling) and the trans-portation to the asphalt plant where it is remixed, but excluded other possible energy demands because milled RAP are veryconsistent and can be used in new mixes without further screening or crushing (Federal Highway Administration, 2011). Theresults suggest that the effect of recycling is very limited compared with the other modules and will not be discussed in de-tail here. One can find more information in the report by Yu and Lu (2011).
For construction activities, it is assumed that the two lanes in each direction are both closed so that all traffic takes detour,with a speed reduction from 65 mph to 40 mph, and a longer travel distance of 2.4 km (1.5 mi) (Zhang et al., 2010). For therehabilitation periods, it is assumed that only one lane will be temporarily closed. These assumptions are fed into QuickZonemodel to estimate the parameters like detour rate, queue length, speed reduction within work zones. Traffic delays are thencoupled with fuel consumptions and vehicle emissions to measure their environmental impacts. The fuel consumption andenvironmental burdens are calculated using Eq. (1).
For the usage module roughens effect, pavement structure effect, albedo, and carbonation, are considered separately. Forroughness effects, the IRI developing trend for the three options are estimated by MEPDG model (Fig. 2).
It is assumed that IRI is restored to its initial values when rehabilitation activity of every 16 years is performed. The LCAinventory is calculated as the differences between driving on real pavement and on an ideally smooth pavement. Accordingto Chandra’s research (2004), highway capacity is reduced by approximately 150 vehicles per hour per lane when IRI is in-creased by 1 m/km. Under the IRI development scenarios from Fig. 2, the potential highway capacity reductions are esti-mated accordingly, which are then reflected into the QuickZone model to estimate the possible delay and amounts ofdetours. A typical torque curve for an engine (Tunnell and Brewster, 2005) is used to estimate the emission due to additionalfriction and vertical acceleration of the vehicle body, which defines a not-to-exceed zone. Constrained by this zone, a con-stant emission rate is assumed for a typical operation speed (90–105 km/h). Any additional emissions produced from engineload increase can be estimated as proportional to the fuel consumption increase calculated by Eq. (2).
Pavement structure effects are based on Taylor et al. (2000) and applied using Florida’s temperature range, Table 1 (theCSOL option is treated as composite pavement). The concrete pavement is set as the baseline. The transformation is carriedout discarding winter data, applying spring data to the winter, summer data to summer and fall seasons, fall data to winter.
384 B. Yu, Q. Lu / Transportation Research Part D 17 (2012) 380–388
The additional fuel consumptions are expressed as the differences between the HMA, CSOL pavements and the PCC pave-ments. The associated air emissions are calculated following the convention of estimating the air emissions in the ‘‘IRI Ef-fect’’. For the albedo effect, in this LCA, the three options are compared with the old PCC pavement with an albedo of0.25. For the PCC option, the albedo is set to 0.35; for the HMA and CSOL options, the albedo is set to 0.15 (Pomerantzand Akbari, 1997). A study by Portland Cement Association found carbonation rate factors of 8.5, 6.7, and 4.9 for concretewith compressive strengths of 21, 28, and 35 MPa (Gajda, 2001). The k value used is 6.3 via linear interpolation. The Albedoeffect can thus be estimated by Eq. (3). Carbonation of PCC option can be calculated using Eq. (5) based on the area of PCCpavement surface layer.
4. Results
The energy consumption of each module and associated components are plotted in Fig. 3. It is found that the energy con-sumed for 1 km of the PCC, HMA, and CSOL overlays are 61 � 103 GJ, 129 � 103 GJ, and 101 � 103 GJ. The energy consump-tions for three scenarios are all dominated by material, congestion, and usage modules. If usage module is not considered, asmany previous studies did, the energy consumptions for PCC, HMA, and CSOL options witness reductions of 40%, 50%, and44%. Feedstock energy occupies a significant portion of the consumed energy, and will significantly reduce the energy con-sumption for the HMA and CSOL options if not counted. Details of the life cycle inventory (LCI) are listed in Table 2.
It seems from Table 2 that the PCC option is most environmental friendly. However, uncertainties exist in the usage mod-ule because of the assumptions used.
The global warming impact is expressed as GHG emissions in tonnes of CO2 equivalent. Fig. 4 shows the global warmingimpact of each overlay system. GHG is dominated by material, congestion, and usage modules for all the three pavementrehabilitation options. And the GHG emissions from the usage phase are dominating for HMA and CSOL. Moreover, carbon-ation gives credit to the PCC overlay but very limited; albedo brings benefits to the PCC option as compared with the HMAand CSOL options; pavement structure effect brings great amount of GHG burdens for the HMA and CSOL options. For CO2,CH4, and N2O, CO2 dominates, with more than 90% for the three options. As for the usage module, it suggests all components,IRI development, pavement structure, albedo, and carbonation, give credits to the PCC option. However, similar to ‘‘EnergyConsumption’’, plenitude of assumption and uncertainty are introduced and shall be captured in the sensitivity analysis.
In Table 2, the emissions of NOx and CO for the HMA and CSOL options show negative values in comparison with the PCCoption in congestion module. The reason for this phenomenon is the emission rates of NOx and CO are higher at low speedsthan those at high speeds while the fleet speed decreases significantly during construction periods (Zhang et al., 2010).
The traffic growth rate in the baseline scenario is set to zero, while in reality traffic growth may significantly affect thefuel consumptions and air pollutant emissions. Several traffic scenarios with various annual growth rates are selected toinvestigate their impacts on fuel consumptions. The results are shown in Fig. 5. Only traffic related fuel consumption inthe congestion and the usage modules are counted.
At various annual traffic growth rates, traffic related energy consumption increases significantly, almost following a linearmode. The slopes of the HMA and CSOL options are much steeper than that of the PCC option, which suggest that the formertwo are more sensitive to traffic growth. The CSOL option consumes less energy during congestion and usage phases due to asmaller intercept compared with HMA option.
Fuel economy is also a critical factor influencing the traffic related energy consumption. The baseline scenario is set to azero fuel economy improvement. Three alternatives, with 1% annual fuel economy improvement, hybrid technology
Feedstock
Primary
Distribution
Construction
Congestion
Usage-IRI Effect
Usage-Pavement Structure
EOL
PCC HMA CSOL0
20
40
60
80
100
120
140
Tot
al E
nerg
y C
onsu
mpt
ion,
103
GJ
Fig. 3. Energy consumption by life-cycle phase.
Table 2Inventories associated with the alternatives.
Input–output Energy (GJ) CO2 (tonne) CH4 (kg) N2O (kg) VOC (kg) NOx (kg) CO (kg) PM10 (kg) SOx (kg)
Primary Feedstock
PCC Material 12,709 NA 1219 659 4 111 2194 14,118 3168 1158Distribution 185 NA 14 16 0.3 5 17 8 2 4Construction 70 NA 4 0a negligible 23 291 133 14 8Congestion 11,274 NA 759 0a 0a 877 �2908 �27,414 116 1Usage 37,083 NA 1863 0a 0a 3057 3376 73,470 55 59EOL 100 NA 13 8 0.2 5 44 17 4 3
HMA Material 13,958 39,034 930 2247 1 205 1994 199 64 879Distribution 245 NA 19 21 0.4 7 22 11 3 5Construction 97 NA 54 0a negligible 30 390 172 30 11Congestion 10,792 NA 726 0a 0a 1103 �1625 �15,291 67 3Usage 64,688 NA 4964 0a 0a 4814 5343 115,670 85 92EOL 143 NA 37 7 0.14 22 297 168 22 8
CSOL Material 9539 26,668 636 1535 1 140 1362 136 44 60Distribution 118 NA 9 10 3 11 5 1 2Construction 74 NA 41 0a negligible 23 312 143 24 9Congestion 8190 NA 551 0a 0a 1104 �1625 �15,291 67 3Usage 56,419 NA 4340 0a 0a 4767 5227 115,215 86 92EOL 79 NA 21 4 0.1 12 165 93 12 5
Note: 0a; the item is not within outputs of the models and a zero value is assigned although this does not influence the results significantly because CO2
emissions are three orders bigger than other GHGs, CO2, CH4, and N2O.
Usage-IRI Effect Usage-Carbonation Usage-Albedo Usage-Pavement Structure Material EOL Congestion Construction Distribution
PCC HMA CSOL
-500
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
5500
6000
6500
7000
CO
2 eu
qiva
lent
, ton
ne
Fig. 4. Greenhouse gas emission by life-cycle phase.
B. Yu, Q. Lu / Transportation Research Part D 17 (2012) 380–388 385
(Heywood et al., 2004), and 2% annual fuel economy improvement, are studied to measure the uncertainty of fuel economyparameter. The results are plotted in Fig. 6. For traffic related energy consumption, fuel economy improvement of 2% annu-ally brings a fuel reduction of 26%, 27%, and 28% for the PCC, HMA, and CSOL options.
The Usage module dominates the LCIs but carries uncertainty that may alter the results. Specifically, IRI increase rate is afactor that would influence the traffic related energy consumption. At a 2% higher IRI development rate (compared with theoriginal MEPDG IRI predictions), the additional fuel consumption is 2%, 1.4%, 1.6% for the PCC, HMA, and CSOL options. At a4% higher IRI development rate, the additional fuel consumption is 4%, 2.7%, 3.2% for the PCC, HMA, and CSOL options.
Influenced by many factors, such as chemical composition of the concrete, pavement structural dimensions, and theambient environment, the carbonation process can take from years to millennia to complete (Damtoft et al., 2008). Forthe extreme low scenario, no carbonation is expected for the PCC option; for the extreme high scenario, an 8.5 mm/y½ ofcarbonation rate is used (Gajda, 2001).
Albedo brings advantage to the PCC option but disadvantage to the HMA and CSOL options at the selected albedo values.Actually, albedo tends to have a broad range for typical concrete pavement and asphalt pavement, being 0.27–0.58 and 0.12–0.46 (Pomerantz et al., 1997). Extreme cases are then calculated accordingly.
0.5% 1% 1.5% 2% 2.5%
40
60
80
100
120
140
160
Ene
rgy
Con
sum
ptio
n, 1
03 G
J
Annual Traffic Growth Rate
PCC
HMA
CSOL
Fig. 5. Sensitivity of traffic related energy consumptions due to traffic growth.
PCC HMA CSOL0
10
20
30
40
50
60
70
80
90
Ene
rgy
cons
umpt
ion,
103
GJ
baseline
1%
hybrid
2%
Fig. 6. Energy consumption based on different fuel economy improvement scenario.
Table 3Fuel economy comparisons between asphalt, composite pavements and concrete pavement. Souces: Zaniewski (1989), Taylor (2002), Taylor and Patten (2006),and Beuving et al. (2004).
Item Asphalt versus concrete (%) Composite versus concrete (%)
Passenger car Heavy trucks Passenger car Heavy trucks
1 0 20 NA NA2 NA 0.2–4.9 NA �1.1 to 3.23 �0.3 to 2.9 0.8–1.8 �2.3 to 1.5 �1.5 to 3.14 0.05–0.88 0.05–0.88 NA NA
386 B. Yu, Q. Lu / Transportation Research Part D 17 (2012) 380–388
Pavement structure influences fuel consumption and air emissions of vehicles considerably, and prior comparisons ofpavement structures have produced a variety of results. The possible ranges of additional fuel consumption for asphaltand composite pavements versus concrete pavement are listed in Table 3; extreme values are used to calculate the range.The three ‘‘what-if’’ calculations are plotted in Fig. 7.
Fig. 7 suggests that the effect due to the albedo and carbonation are completely overwhelmed by the pavement structureeffect, that there are large uncertainties associated with the pavement structure effect, and that the low and high bounds for
Carbonation
Albedo
Albeo
HMA vs PCC
Albedo
CSOL vs PCC
-4 .0 x1 0 5 -2 .0 x1 0 5 0 .0 2 .0 x1 0 5 4 .0 x1 0 5 6 .0 x1 0 5
CSO
LPC
C
CO2 equivalent, tonne
HM
A
High bound Low bound
Fig. 7. Uncertainty of carbonation, albedo and pavement structure effects.
B. Yu, Q. Lu / Transportation Research Part D 17 (2012) 380–388 387
the ‘‘CSOL vs. PCC’’ bar is approximating while the high bound is orders larger than the low bound for the ‘‘HMA vs. PCC’’ bar.In this sense, one expects a smaller environmental burdens from the PCC as compared with the HMA option while less cer-tain about the conclusion if compared with the CSOL option.
5. Conclusions
A LCA model of pavement embracing six modules – material module, distribution module, construction module, conges-tion module, usage module, and EOL module – is developed and used to explore three overlay options, PCC, HMA, and CSOLcompared with the old PCC pavement. It is found that the overall model is a useful and a relatively complete tool to estimatethe environmental impacts of pavement. Further, it is reasonable to expect a smaller environmental burden from the PCCand CSOL options as opposed to HMA although comparisons between the former are indeterminate because of uncertaintiesin the usage stage, and especially pavement structure effects. Materials, congestion, and usage are the three major sources ofenergy consumptions and air pollutant emissions in the usage module. Traffic related fuel consumption is emerges as verysensitive to traffic growth and fuel economy improvements. Fuel consumption basically increases linearly with the trafficgrowth rate.
References
Akbari, H., Menon, S., Rosenfeld, A., 2008. Global cooling: increasing world-wide urban albedos to offset CO2. Climatic Change 95, 275–286.Amos, D., 2006. Pavement Smoothness and Fuel Efficiency: An Analysis of the Economic Dimensions of the Missouri Smooth Road Initiative. Final Report:
OR07-005, Jefferson City.Beuving, E., De Jonghe, T., Goos, D., Lindahl, T., Stawiarski, A., 2004. Environmental Impacts and Fuel Efficiency of Road Pavements. Europe Asphalt Pavement
Association and Eurobitume, Brussels.Chan, A.W.C., 2007. Economic and Environmental Evaluations of Life Cycle Cost Analysis Practice: A Case Study of Michigan DOT Pavement Projects, Master
of Science Thesis in Natural Resource and Environment, University of Michigan.Chandra, S., 2004. Effect of road roughness on capacity of two-lane roads. Journal of Transportation Engineering 130, 360–364.Damtoft, J.S., Lukasik, J., Herfort, D., Sorrentino, D., Gartner, E.M., 2008. Sustainable development and climate change initiatives. Cement and Concrete
Research 38, 115–127.Emission Facts, 2005. Average Carbon Dioxide Emissions Resulting from Gasoline and Diesel Fuel. EPA420-F-05-001. Office of Transportation and Air
Quality, Washington, DC.Federal Highway Administration, 2011. Reclaimed Asphalt Pavement in Asphalt Mixtures: State of the Practice. Publication No. FHWA-HRT-11-021,
Washington, DC.Gajda, J., 2001. Absorption of Atmospheric Carbon Dioxide by Portland Cement Concrete (revised in 2006). Portland Cement Association. PCA R&D Serial No.
2255a, Skokie.Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation, 2010. Model. Energy Systems Division, Argonne National Laboratory, Argonne,
and University of Chicago, Center for Transportation Research.Heywood, J.B., Weiss, M.A., Schafer, A., 2004. The Performance of Future ICE and Fuel Cell Powered Vehicle and Their Potential Fleet Impact. Laboratory for
Energy and the Environment, Massachusetts Institute of Technology, Cambridge, Mass.Huang, Y., Bird, R., Bell, M., 2009. A comparative study of the emissions by road maintenance works and the disrupted traffic using life cycle assessment and
micro-simulation. Transportation Research D 14, 197–204.Athena Institute, 2006. A Life Cycle Perspective on Concrete and Asphalt Roadways: Embodied Primary Energy and Global Warming Potential, Prepared for
the Cement Association of Canada, Ottawa.Keoleian, G.A., Kendall, A., Dettling, J., Smith, V.M., Chandler, R.F., Lepech, M.D., Li, V.C., 2005. Life cycle modeling of concrete bridge design: comparison of
engineered cementitious composite link slabs and conventional steel expansion joints. Journal of Infrastructure Systems 16, 51–60.Lagerblad, B., 2006. Carbon Dioxide Uptake During Concrete Life Cycle – State of the Art. Swedish Cement and Concrete Research Institute, CBI. Nordic
Innovation Centre Project Number 03018, Stockholm.Marceau, M.L., Nisbet, M.A., Van Geem, M.G., 2007. Life Cycle Inventory of Portland Cement Concrete. Portland Cement Association, Skokie.Pomerantz, M., Akbari, H., Chen, A., Taha, H., Rosenfeld, A.H., 1997. Paving Materials for Heat Island Mitigation. Lawrence Berkeley National Laboratory.
LBNL-38074.
388 B. Yu, Q. Lu / Transportation Research Part D 17 (2012) 380–388
Stolaroff, J.K., Lowry, G.V., Keith, D.W., 2005. Using CaO- and MgO-rich industrial waste streams for carbon sequestration. Energy Conversion andManagement 46, 687–699.
Stripple, H., 2001. Life Cycle Assessment of Road: A Pilot Study for Inventory Analysis, second revised ed. Swedish National Road Administration. IVL B 1210E., Borlänge.
Stubstand, R., Darter, M., Rao, C., Pyle, T., Walid, T., 2005. Report on the Effectiveness of Diamond Grinding Concrete Pavements in California. CaliforniaDepartment of Transportation, Sacramento.
Taylor, G.W., 2002. Additional Analysis of the Effect of Pavement Structure on Truck Fuel Consumption. Prepared for Government of Canada Action Plan2000 on Climate Change. Concrete Roads Advisory Committee, Ottawa.
Taylor, G.W., Patten, J.D., 2006. Effects of Pavement Structure on Vehicle Fuel Consumption – Phase III. National Research Council of Canada. Project 54-HV775, Technical Report CSTT-HVC-TR-068, Ottawa.
Taylor, G.W., Marsh, P., Oxelgren, E., 2000. Effect of Pavement Surface Type on Fuel Consumption – Phase II: Seasonal Tests. Portland Cement Association.CSTT-HWV-CTR-041, Ontario.
Tunnell, M.A., Brewster, R.M., 2005. Energy and emissions impacts of operating higher-productivity vehicles. Transportation Research Record 1941, 107–114.
Weiland, C., Muench, S.T., 2010. Life-cycle assessment of reconstruction options for interstate highway pavement in Seattle, Washington. TransportationResearch Record 2170, 18–27.
Yu, B., Lu, Q., 2011. Life Cycle Assessment of Pavement Overlay Systems with Focus on Usage Stage. Department of Civil and Environmental Engineering,University of South Florida. <http://qlu.myweb.usf.edu/>.
Zaniewski, J.P., 1989. Effect of Pavement Surface Type on Fuel Consumption. Portland Cement Association. PCA R&D SR289, Skokie.Zhang, H., Lepech, M., Keoleian, G., Qian, S.Z., Li, V., 2010. Dynamic life-cycle modeling of pavement overlay systems: capturing the impacts of users,
construction, and roadway deterioration. Journal of Infrastructure Systems 16, 299–309.
