See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/245558261

Comparison of Environmental Implications of

Asphalt and Steel-Reinforced Concrete
Pavements

Article in Transportation Research Record Journal of the Transportation Research Board · September 1998
DOI: 10.3141/1626-13

CITATIONS READS

87 130

2 authors, including:

Chris Hendrickson
Carnegie Mellon University
257 PUBLICATIONS 6,468 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

https://www.nap.edu/catalog/14670/underground-engineering-for-sustainable-urban-development
View project

Environmental Life Cycle Assessment View project

All content following this page was uploaded by Chris Hendrickson on 17 April 2015.

The user has requested enhancement of the downloaded file.

TRANSPORTATION RESEARCH RECORD 1626
Comparison of Environmental Implications of
Asphalt and Steel-Reinforced
Concrete Pavements
ARPAD HORVATH AND CHRIS HENDRICKSON
The public, industry, and governments have become increasingly inter-
ested in green design and sustainable development. Construction activi-
ties affect the environment significantly, so environmental issues should
be considered seriously. Thousands of miles of roads are paved every year
with asphalt and steel-reinforced concrete. What are the environmental
effects of the two materials? If asphalt has been used overwhelmingly
over concrete, is it a better choice for sustainable development? We
present results of a life cycle inventory analysis of the two materials
based on publicly available data. We find that for the initial construc-
tion of equivalent pavement designs, asphalt appears to have higher
energy input, lower ore and fertilizer input requirements, and lower toxic
emissions, but it has higher associated hazardous waste generation and
management than steel-reinforced concrete. When accounting for the
uncertainty in the data and when annualizing environmental effects based
on assumed average service lives of the two pavement types, the resource
input requirements and the environmental outputs are roughly compa-
rable for the two materials. However, asphalt pavements have been re-
cycled in larger quantities than concrete pavements, with consequent
resource savings and avoided pollution, which suggests that asphalt may
be a better choice from a sustainable development viewpoint. Of course,
special functional requirements or economics may dictate the use of one
material over the other in particular applications regardless of the overall
environmental effects.
Road construction is a major construction activity. The two most
common materials used for pavement construction are asphalt and
cement concrete (often reinforced with steel). Of the 1 440 366 km
(895,000 mi) of rural and urban Interstate and other highway, arte-
rial, and collector roads and streets in the United States, approx-
imately 1 189 307 km (739,000 mi) (or nearly 83 percent) of
pavement are of flexible type (asphalt wearing surface), 94 951 km
(59,000 mi) (or nearly 7 percent) are of rigid type (portland cement
concrete roadway with or without a bituminous wearing surface),
and 154 497 km (96,000 mi) (or less than 11 percent) are of com-
posite type (bituminous surface on a cement concrete base) (1). The
statistics on the surface type of Interstate highways and other free-
ways and expressways show a more balanced use of asphalt and
concrete: as much as 57 percent of the mileage may have cement
concrete either as a wearing surface or as a base. Other roads and
streets largely have asphalt surface.
What are the environmental implications of using one pavement
material type versus the other? If asphalt is used overwhelmingly
over concrete, is it a better material in terms of environmental sus-
tainability? These questions can be answered by performing a life
cycle assessment (LCA).
Department of Civil and Environmental Engineering, Carnegie Mellon
University, Pittsburgh, PA 15213.
Paper No. 98-0661 105
LCA
LCA is a method that looks systematically at the environmental
effects of various stages of the entire life cycle of a product or a
process: the materials extraction stage, the manufacturing/production
stage, the use phase, and the ultimate disposal phase (or end-of-life).
It involves quantification of the environmental burdens (inventory
analysis), estimation of the impacts of these burdens on humans and
nature (impact analysis), and identification of areas where improve-
ments are possible (improvement analysis). There are many efforts
in the world to produce LCA studies that are comprehensive and
useful (2,3). Several computer-based LCA tools provide templates
and libraries of data to users. Existing studies differ in the number
of environmental effects quantified and in the scope of the analy-
sis. Scope is determined by where the boundary of the analysis is
drawn. Currently, there are two approaches to boundary setting: a
model developed most intensively by the Society of Environmental
Toxicology and Chemistry (SETAC) and the U.S. Environmental
Protection Agency (EPA), and a model called economic input–
output-based LCA (EIO-LCA), developed by researchers in Carnegie
Mellon University’s Green Design Initiative (4,5). The SETAC-EPA
approach divides each product into individual process flows and tries
to quantify their environmental effects. For example, in the manu-
facturing stage of products, it attempts to go as far back (upstream)
in the flow as possible. This assessment is typically limited by data
availability, time, and cost; it includes the first tier (direct) of suppli-
ers but seldom the complete hierarchy of suppliers—that is, all the
suppliers of suppliers (indirect). In contrast, the EIO-LCA model
uses the 498 × 498 economic input–output matrix (commodity-by-
commodity) of the U.S. economy to identify the entire chain of
suppliers (both direct and indirect) to a commodity, thus setting the
boundary of the materials extraction and the manufacturing stage
assessment at the level of the national economy. The 498 × 498
matrix is based on commodities such as cement, motor vehicle bod-
ies, sugar, and so forth. To obtain the direct plus indirect (total) eco-
nomic demand, final purchase amounts (final demand) are input into
the model. The resulting total demand is then multiplied by matrices
of average environmental burden factors calculated on economic sec-
tor level: toxic emissions per dollar of economic output, fuel use per
dollar of output, and so forth.
A life cycle inventory analysis of the two most common pave-
ment material types (asphalt and steel-reinforced concrete) is
presented here. The EIO-LCA method was used to assess the envi-
ronmental effects of the materials extraction and the manufactur-
ing stages; data from the literature were used for the other life
cycle phases. This article is based on an unpublished doctoral
dissertation (6 ).
106 Paper No. 98-0661
DESIGN AND MATERIAL COMPOSITION
OF PAVEMENTS
New 1-km-long asphalt and cement concrete pavement sections were
used that have roughly the same functionality, representing equiv-
alent designs, based on AASHTO guidelines (7). The pavement
cross sections are intended to represent a typical two-lane highway
with high volumes of traffic in the United States. The sections are
designed for 10 million 80-kN (18-kip) equivalent single-axle loads,
which is an estimate of 10 or more years of Interstate highway traf-
fic. The serviceability index at the end of the design life of the pave-
ment sections is selected to be 2.5 (on a scale of 0 to 5, with 5 being
a perfect pavement). Both pavement sections are 720 cm (24 ft) wide
and are assumed to sit on 15 cm (6 in.) of high-quality cement-treated
soil subbase [E = 6.9 GPa (1,000,000 lbf/in.2)]. Because the subbase
is the same for both pavement types, only the environmental burdens
of the surface course materials are compared.
The concrete pavement section is selected to be continuously
reinforced (CRCP), with these design parameters:
• Continuity factor, 2.2;
• Flexural stress for concrete, 4.1 kPa (600 lbf/in.2);
• Working stress for concrete, 4.1 kPa × 0.75 = 3.1 kPa
(450 lbf/in.2);
• Modulus of elasticity for concrete, 27.6 GPa (4,000,000 lbf/in.2);
• Composite modulus of subgrade reaction (kc), 330 lbf/in.3;
• Amount of longitudinal steel required, 0.6 percent;
• Subbase friction (F ), 1.9;
• Allowable working stress in steel, 345 MPa (50,000 lbf/in.2);
• Amount of transverse steel required, 0.046 percent [based on
Figure D.4-4 (7 )]; and
• Reinforcing bar used, No. 4 [cross-sectional area, 1.29 cm2
(0.2 in.2)].
Figure D.4-2 (7 ) was used to design the concrete pavement slab
thickness: 22 cm (8.5 in). To determine the required longitudinal
steel reinforcement, Figure D.4-7 (7 ) was used: No. 4 bars at 10 cm
(4 in.) center to center [73 bars, for a total of 73 000 m (239,440 ft)].
For the required transverse steel reinforcement, we used Figure
D.4-7 (7 ): No. 4 bars at 130 cm (51 in.) center to center [769 bars,
for a total of 5537 m (18,161 ft)]. The extra reinforcing for the con-
struction joints is relatively low and is neglected in this assess-
ment. Therefore, a total of 78 537 m (257,601 ft) of No. 4 bars are
required. At 0.994 kg/m (0.668 lb/ft) for No. 4 bars, the steel con-
tent of the CRCP section is 78 066 kg (172,077 lb). For an aver-
age producer price of $0.30/kg ($0.135/lb) (8), the steel demand
for the CRCP section is estimated to be $23,420. (Because of
rounding of numbers, the steel demand based on pounds would be
$23,230.)
For design of the asphalt pavement section, the following para-
meters were used:
• Soil support value, 3.0;
• Regional factor, 1.5;
• Structural number, 6.0;
• Layer coefficient for the subbase, 0.15; and
• Layer coefficient for the asphalt concrete surface course, 0.44.
Based on Section C.5 (7), the thickness of the asphalt pavement was
determined to be 30 cm (12 in.).
TRANSPORTATION RESEARCH RECORD 1626
The density of both materials is assumed to be 2323 kg/m 3
(145 lb/ft3) (8). A 1-km-long asphalt pavement section, therefore,
requires 5,018 metric tons of asphalt, and a 1-km CRCP section
requires 3680 metric tons of concrete. The national average prices
of asphalt and cement concrete [assuming transportation within
a 32-km (20-mi) radius and 20.67 MPa (3,000 lbf/in.2) or the con-
crete] were taken to be $29/metric ton for both materials (8). There-
fore, the material cost for the asphalt pavement section was $145,500
and the cost for the CRCP was $106,700 for the concrete and $23,400
for the steel (total of $130,100).
PERFORMING A LIFE CYCLE INVENTORY
After the designs and costs of the two pavement sections made of dif-
ferent materials were established, the EIO-LCA model was used to
calculate the environmental burdens of both materials in the manu-
facturing stages. The final demands are the costs of the pavement
materials: $145,500 for the asphalt pavement, and $106,700 for the
concrete and $23,400 for the steel content in the CRCP. Both pave-
ment types were estimated in this analysis by commodities in the
498 × 498 input–output matrix. For the asphalt pavement, the sector
“asphalt paving mixtures and blocks” [standard industrial classifica-
tion (SIC) code 2951] was used. For the concrete content of the pave-
ment, the sector “ready-mixed concrete” (SIC code 3273) was used,
and for the steel reinforcement, the sector “blast furnaces and steel
mills” (SIC code 3312) was used. The total economic transactions
in the supply chain for the two pavements were similar: $349,000
for asphalt, and $289,000 for steel-reinforced concrete. This total
demand indicates that as a result of the final demand in the asphalt
and the concrete and steel industries, there was additional demand for
products and services in the entire national economy for $203,500 in
the case of asphalt, and for $158,900 in the case of CRCP.
Resource inputs and environmental burdens for the two pavement
materials were quantified in this LCA study. Resource inputs include
consumption of electricity, fuels, ores, and fertilizers (9). The eco-
nomic input–output methodology allows for identification of the
inputs not directly related to pavement material manufacturing, such
as fertilizer demand. In this case, fertilizer consumption occurs with
some of the agricultural suppliers of some of the direct suppliers to
the asphalt and concrete sectors. Environmental effects include toxic
chemical discharges to air, water, land, underground injection wells,
and transfers to off-site treatment plants (10); ozone depletion poten-
tial of chemical releases (10,11); hazardous waste generation and
management (12); and conventional pollutant emissions to air (9,13).
Similar to the resource input requirements, the environmental effects
of not only the direct suppliers (such as the cement industry for con-
crete) but also the indirect suppliers (such as the agricultural sector
for concrete) are included in this assessment.
UNCERTAINTY IN THE DATA
All the data used in this study are uncertain. For example, the toxic
chemical release data are obtained from the EPA’s toxics release
inventory (TRI) collected from manufacturing plants. Facilities have
to report emission estimates to the TRI, which leads to inaccuracies
(14). Similar uncertainties exist regarding the Resource Conserva-
tion and Recovery Act (RCRA) Subtitle C hazardous waste data.
Conventional pollutant emissions are based on the fuel consumption
of facilities, and the AP-42 emission factors (AP-42) from the EPA.
Horvath and Hendrickson
The AP-42s are estimates of air emissions from processes. Resource
input calculations are based on data gathered by the U.S. Department
of Commerce. Companies’ accounts of these costs are imperfect; in
fact, many facilities fail to report them.
RESOURCE INPUT REQUIREMENTS
Table 1 presents a summary of the resource input requirements
associated with paving a 1-km section of a typical highway with
asphalt and concrete. Ore and fertilizer consumption appeared to be
higher for the CRCP. The electricity consumption is estimated to be
roughly the same for the two materials. There appear to be differ-
ences in consumption of some fuel types. Bituminous coal con-
sumption appears to be larger for the steel-reinforced concrete
pavement, whereas the use of natural gas, liquefied natural gas, and
liquefied petroleum gas appears to be larger for the asphalt pave-
ment. Overall, converted to a common unit, asphalt pavement pro-
duction for a 1-km section requires about 7 million MJ of energy,
and the equivalent CRCP design requires roughly 5 million MJ.
Embodied energy figures for concrete (without the steel reinforc-
ing) are available from the literature. Figure 1 compares the result
of the EIO-LCA with a few values from other studies. The EIO-LCA
yields a value of 900 MJ/metric ton for the ready-mixed concrete
TABLE 1 Environmental Effects of Asphalt and Steel-Reinforced Concrete
(Concrete + Rebars) Production for 1 km of Pavement: Summary of Resource Inputs
Paper No. 98-0661 107
(without electricity) and is roughly comparable to the value reported
by Fog and Nadkarni (16 ) and to the higher value of Boustead and
Hancock (18). However, it appears to be lower than the values
reported by the American Institute of Architects (15) and by Ashby
(17 ) and significantly higher than the lower figure of Boustead and
Hancock (18). Unfortunately, these literature sources do not define
what exactly their embodied energy figures contain, so comparison
with EIO-LCA results may be inaccurate.
ENVIRONMENTAL OUTPUTS
Table 2 presents a summary of the environmental burdens asso-
ciated with the material extraction and manufacturing stages of
asphalt and concrete materials (concrete and steel reinforcement) for
pavements. Three major groups of environmental effects are quan-
tified in this assessment: TRI chemical emissions, hazardous waste
generation and management, and conventional pollutant emissions
to air. For TRI emissions, we assessed both the unweighted amounts
of TRI emissions (in kilograms, as reported by facilities to the EPA),
and amounts of TRI emissions weighted by relative toxicity of the
discharges, normalized to sulfuric acid (H2SO4) (called CMU-ET,
and expressed in kilograms of H2SO4 equivalent) (19). Except for
the RCRA Subtitle C hazardous waste amounts generated and man-
FIGURE 1 Comparison of embodied energy figures for concrete (without the
steel reinforcing).

TABLE 2 Environmental Effects of Asphalt and Steel-Reinforced Concrete (Concrete +

Rebars) Production for 1 km of Pavement: Summary of Environmental Outputs
Horvath and Hendrickson
aged, and the emissions of total suspended particulates, emissions
appeared to be higher for the steel-reinforced concrete design.
CONSTRUCTION STAGE EFFECTS
During construction, asphalt pavements emit bitumen fumes. Out-
gassing occurs during material mixing, pavement laying, and for some
time after the pavement is finished depending on the characteristics
of the asphalt and the local climate.
Bitumens have variable chemical compositions and can be a mix-
ture of residues from many crude oil types. Because of the variabil-
ity of the chemical composition, environmental risks associated with
asphalt may vary a great deal. At least 55 chemical compounds are
commonly found in asphalt (20). The ones considered of environ-
mental concern when emitted include polycyclic aromatic hydro-
carbons (PAHs) (for example, naphthalene, fluorene, anthracene,
fluoranthene, and benzopyrenes) and volatile organic compounds
(VOCs) (for example, benzene, benzaldehyde, alkylated benzenes,
naphthalene, and alkylated naphthalenes) (21). Many PAHs have
been thought to contribute, at least in part, to the carcinogenicity of
fossil fuels and their products. They have been reported to cause skin
cancer in laboratory animals and possibly skin and lung cancers in
exposed workers. The EPA has reviewed the literature on 15 PAHs
(20). Seven PAHs are accepted as probable human carcinogens and
confirmed animal carcinogens, and eight have not been classifiable as
human or animal carcinogens because of inadequate evidence. Some
VOCs are known to have toxic and potentially cancerous effects on
exposed mammals, and they contribute to formation of smog. For
example, benzene has been classified as a human carcinogen by
almost all relevant agencies (22).
A report (20) lists several epidemiological studies that have
researched the relationship between exposure to asphalt and effects
on human health. Some studies involved exposure of workers to
paving asphalt, and some involved exposure to roofing asphalt. The
interpretation of these studies is complicated because there have
not been adjustments to other exposures (such as coal-tar pitch or
tobacco smoke), or they were based only on employment on the date
of the census.
Despite some of its components being suspected or confirmed
carcinogens, asphalt fumes (Chemical Abstract Service number
8052-42-4) are considered carcinogenic only by the National Insti-
tute for Occupational Safety and Health (NIOSH). Other agencies
did not have sufficient evidence or have not yet classified asphalt.
With regard to asphalt workers’ occupational exposure hazards,
the American Conference of Governmental Industrial Hygienists’
(ACGIH) threshold limit value time-weighted average (TLV-TWA)
for asphalt fumes is 5 mg/m3 (22), which indicates the concentration
of asphalt fumes needed to develop adverse health effects in work-
ers exposed to it over time. In contrast, sulfuric acid has a TLV-TWA
of 1 mg/m3. Heavy metals have a TLV-TWA that is two to three
orders of magnitude smaller (indicating that smaller concentrations
are hazardous). Australia, Belgium, Denmark, and the United King-
dom have recommended workplace exposure values for asphalt
fumes that are the same as TLV values (23). In contrast, NIOSH has
established a recommended exposure limit ceiling of 5 mg/m3 per
15 min, a considerably stricter limit than the TLV-TWA. Modern
paving machines, however, have the capability to channel fumes
away from asphalt workers.
A report (20) summarizes studies on the exposure of workers in
a mixing plant and in four paving crews to PAHs in asphalt fumes.
Paper No. 98-0661 109
For the 19 substances measured in the studies, many concentrations
were below the detection limit or differed substantially from study
to study. Occupational Safety and Health Administration permissi-
ble exposure limits, which have the same definition as the ACGIH
TLVs, were found for three of the substances. For these three chem-
icals, the measured concentrations were under the threshold limit
that could cause adverse health effects in workers.
The transportation required for the manufacturing and the con-
struction stages of the two materials, as well as the energy use during
construction, are assumed to be comparable.
END-OF-LIFE OPTIONS FOR
PAVEMENT MATERIALS
Asphalt
A survey of 29 state highway agencies in the United States (21) indi-
cated that around 80 percent of the removed asphalt concrete surface
get recycled into highway applications. This amounts to more than
70 million metric tons of asphalt pavement debris kept out of land-
fills annually. A key to this rate is the demonstrated recyclability of
asphalt pavements. Most state highway agencies’ specifications per-
mit the contractor to retain ownership of reclaimed asphalt pavement.
This practice may help the contractor in the bidding process as well
as in managing the equipment capacities and material inventories.
Depending on the type and place of recycling asphalt pavements,
the most frequently used methods are plant (hot-mix asphalt) re-
cycling, cold in-place recycling, and hot in-place recycling (21).
Although hot-mix plant recycling has been widely used, both cold
and hot in-place recycling are recommended only for low-traffic
roads. Nevertheless, asphalt can be recycled numerous times. Report-
edly, about 25 percent of hot-mix asphalt (produced in off-site plants
in 1985 and 1986) used recycled asphalt pavement content, around
9300 lane-km were resurfaced using the cold in-place recycling tech-
nique, and about 3900 lane-km used hot in-place recycling. Re-
cycling of pavements also preserves the vertical roadway clearances
(under bridges, at the curb, and manholes), which is an important
benefit.
However, there are no reliable, long-term statistical data on the
strength and durability of recycled asphalt pavements. Advantages
as well as disadvantages of reclaimed pavements have been quoted
in the literature (21,24). A hot mix with recycled content reportedly
reaches high strength quickly (24). Although the moisture contents
of reclaimed asphalt pavements are generally found to be higher than
desired, the Florida Department of Transportation (DOT) reported
that the reclaimed asphalt pavement was better than the locally
available, highly absorptive virgin aggregate because the precoated
aggregate had formed a “more uniform film of asphalt in the final
mix” and that the pavements with recycled content appeared to age
slower. Short-term experiences with reclaimed hot-mix asphalt have
been good, but cold in-place recycled pavements have had stability
problems due to the moisture content of the recycled aggregate. The
Arizona DOT has had problems with uniformity of the reclaimed
asphalt that contained seal coats, friction courses, and other materi-
als. Specifications have been lax with some older pavements that
used reclaimed material: inadequate gradation may have affected
durability. Initial performance was good, but long-term behavior has
not yet been documented.
Even though recycling of old asphalt pavements is an established
technology and evidence from individual projects documented in
110 Paper No. 98-0661
the literature suggests that it might be economically feasible, com-
prehensive, national data on the life cycle costs of recycled material
asphalt pavements are unavailable. Individual projects have experi-
enced initial construction cost savings from 20 to 40 percent over
virgin aggregate mixes (24). In initial construction costs, recycling
of asphalt pavements saves
• Aggregates (especially important in areas where good quality
aggregate is hard to find or is expensive);
• Asphalt binder [mixtures with virgin aggregate typically require
around 6 percent asphalt cement, but reclaimed aggregate mixes
need only 1 to 3percent additional asphalt, saving approximately
37.9 L (10 gal) of asphalt per ton of mix (24)];
• Energy (energy savings in mining, processing, and transport-
ing of the aggregate and the bitumen).
Many millions of tons of old asphalt pavement still end up in land-
fills. The National Asphalt Pavement Association (NAPA) reports
that many states have tested landfilled asphalt pavement material
and determined that it can be regarded as clean fill (25). The EPA’s
toxicity characteristic leaching procedure test did not indicate any
appreciable quantities of toxic materials from discarded asphalt
paving material.
Concrete
The recyclability of concrete has been demonstrated for a variety of
applications—for example, road base, embankment, backfill, non-
load-bearing (decorative) applications—as well as in new concrete
and asphalt pavements, although it has been suggested (21) that the
latter two applications would benefit from more research on design
and performance. The economic feasibility of recycling concrete
pavements has not yet been documented on a large scale. It was
reported (21) that only about 3 million metric tons of concrete pave-
ment are reclaimed annually. However, it is not known what frac-
tion of concrete pavements gets recycled. One reason for the low
rates of recycling might be that concrete pavements are often left in
place after they become structurally or functionally obsolete; they
get an asphalt overlay and become adequate road bases.
Cement concrete recycling projects are scarcely reported in liter-
ature. One report (26) appeared in Public Works about a small-scale
concrete pavement recycling project. A 25-year-old concrete pave-
ment from a two-lane highway in Kansas was reclaimed, crushed,
and used as aggregate for portland cement-treated base (PCTB) for
1.6 km (1 mi) of portland cement concrete pavement (PCCP) and for
0.8 km (0.5 mi) of road shoulders for a new, four-lane highway sec-
tion for the Kansas DOT. To meet the strength requirements for the
PCTB, 20 percent limestone had to be added to the recycled aggre-
gate. Eventually, the compression strength of the new PCTB proved
to be about 40 percent higher than for conventional PCTB, and the
base appeared to set up quicker with the recycled aggregate, even
though higher moisture content and lower density were observed. It
was reported that the unit price using recycled aggregate was 10 per-
cent lower than the unit price for virgin aggregate used by the same
contractor on a similar project. The conclusion from this report is sur-
prising given that limestone had to be added to the recycled aggre-
gate; that the steel reinforcement from the old pavement had to be
removed with a backhoe and manual labor, often with difficulty; and
that crushing the reclaimed pavement was reported to have been
“more expensive than crushing virgin aggregate due to problems
TRANSPORTATION RESEARCH RECORD 1626
with mesh and rebar plugging the crusher and magnet.” However, the
DOT agreed to take over the excess reclaimed pavement, so the con-
tractor did not have to store it or dispose of it. For the PCCP, 50 per-
cent natural sand had to be added to offset the water and cement
demand of the reclaimed aggregate. Strength figures were compara-
ble to conventional PCCP. About 50 percent recycled aggregate was
used for the road shoulders, but too much large fraction aggregate
reportedly made placing of the mix difficult, and internal fracturing
of aggregate particles contributed to lower than conventional bear-
ing capacity. It has not been reported how the recycled aggregate
affected the durability of the PCTB, the PCCP, or the shoulders.
The recycling of steel reinforcing bars in concrete pavements is
currently not feasible on a large scale because of high labor costs
associated with the activity.
In summary, although not all recycling is economically and envi-
ronmentally beneficial, recycling of both asphalt and concrete pave-
ments (without the reinforcement) appear to be so. The reclaimed
materials are typically reused on site or stored at a nearby mixing
plant, with relatively little transportation involved. It is not eco-
nomical to transport reclaimed materials to distant project locations,
but local markets generally can absorb them if there are efficient
channels of communication between construction professionals. For
example, an Internet-based exchange of information may facilitate
buying and selling of local reclaimed pavement material (27).
Because the EIO-LCA model allows for the economy-wide
assessment of environmental effects, we have to be careful to not
double count the advantages (saved resources and avoided pollu-
tion) of using recycled materials as feedstock of new material pro-
duction. Secondary materials enter the feedstock streams of new
material production and thus are already figured in the economic
input–output matrices. Therefore, recycling materials in the end-of-
life stage should not get credit for the second time. However,
reclaimed asphalt and concrete pavement materials are different:
their recycling is relatively new, and these secondary materials may
not yet have entered the feedstock streams in large enough volumes
and thus show up in the economic statistics. (We were using 1992
economic data in this assessment.) Thus, recycling the entire asphalt
pavement would avoid 5000 metric tons of landfill disposal and
would likely save energy and other resources. In contrast, not re-
cycling the CRCP would imply 3700 metric tons of landfill disposal
at end-of-life.
DISCUSSION OF RESULTS
Some environmental implications associated with the material extrac-
tion, the manufacturing, the construction, and the end-of-life stages
of a 1-km section of asphalt and cement concrete pavements were
presented. Most of the resource inputs and environmental outputs
appeared to be higher for the CRCP in this assessment. Therefore,
building our roads with reinforced concrete instead of asphalt will
have different implications for sustainable development. Estab-
lishing the impacts of resource use and waste outputs on sustain-
able development is a difficult task, but a comparison with annual
national totals (6 ) can be interesting and insightful. If all the con-
crete pavements in the United States [at least 95 000 km (59,000 mi)
(1)] were assumed to be continuously reinforced, and the same
resource input and environmental output values as found in this study
applied, for what percentage of the annual national totals would these
values be the primary factor? Furthermore, if all the 1 440 000 km
(895,000 mi) of rural and urban Interstate and other highway, arterial,
Horvath and Hendrickson
and collector roads and streets in the United States had CRCP, what
percentage of the annual national totals would the inputs and out-
puts represent then? Our findings are presented in Table 3. (Only the
inputs and outputs that were at least twice as high for CRCP as for
asphalt were analyzed.) Of course, the 95 000 km of concrete roads
have been built over decades, and replacing 1.4 million km of pave-
ments would take many decades. Also, roads with less traffic would
not have the same CRCP design assumed for a major highway in this
study. But given that concrete pavements need to be replaced after
20 to 25 years at the most, it is evident that significant resources are
“embedded” in our pavements and that the corresponding environ-
mental effects are significant. One would hardly expect that concrete
pavements (95 000 km) alone are “responsible” for 6 percent of the
iron ore consumption in the United States based on 1992 numbers
and several percentage points of the 1995 TRI emissions. If 1.4 mil-
lion km of roads had CRCP, iron ore demand would almost double
the U.S. consumption based on 1992 figures, and the concrete and
the steel for roads alone would be responsible for 15 percent of the
total TRI emissions of the U.S. manufacturing industries in 1995 and
as much as one-quarter of the TRI emissions weighted by toxicity
(CMU-ET).
The two types of pavements were compared above based on sum-
mary environmental effects. It would be a more realistic to compare
the annualized environmental burdens, taking into account the
expected service life or durability of the two pavement types.
TABLE 3 Resource Inputs and Environmental Outputs for Steel-Reinforced Concrete
Pavements, As a Percentage of Annual National Totals in the United States, for Different
Road-Length Scenarios
Paper No. 98-0661 111
Determining the average service life of pavements is difficult. A
paper (28) reported that the Interstate highways “are designed to last
20 years before major repairs are necessary, but many do not.” A
1949 national survey estimated the average age of road surfacing at
12 years, with 13 percent of the roads being more than 20 years old
and “nearing the end of their useful lives.” As reported by the Port-
land Cement Association (29), the most ambitious study of pavement
types was conducted from 1929 to 1968 by the FHWA and its pre-
decessor, the Bureau of Public Roads. Nineteen states monitored sev-
eral pavement types and traffic conditions. A 1971 study based on
the collected data reported that the average service life for concrete
pavements was 25 years; and for asphalt pavements it was 15 years.
However, the design life of pavements depends on many local fac-
tors, such as climate, traffic types and volumes, and quality of work,
materials, subbase, and so forth. Therefore, using national averages
for service life estimation is problematic.
State or regional data on expected design life of pavements would
be more helpful. For example, the Portland Cement Association
reports (29) that in 1979 the state of Louisiana evaluated pavement
sections built between the mid-1950s and the early 1960s. It found
that the average service life of the 110 test sections (49 were concrete
pavements) was 18 years, with the “majority” of concrete pavements
meeting or exceeding their 20-year design life “in good to excellent
condition.” In contrast, asphalt pavements had to be resurfaced after
an average of 13 to 15 years (with 82 percent requiring a structural
112 Paper No. 98-0661
overlay after 13 years). Of course, both pavement technologies have
improved since these data were collected.
If we assume that the service life of asphalt pavements is 13 to
15 years and the service life of concrete pavements is 20 to 25 years,
annualized environmental effects of the materials extraction and the
manufacturing stages (assuming that the time value of pollution is
constant and that the data have uncertainties) are roughly comparable
for the two pavement materials.
It is also important to consider maintenance requirements. Until the
end of their service life, pavements are typically repaired many times,
whenever a pavement distress occurs. The overall environmental
effects of the two pavement types depend also on how much repair is
needed on average for one pavement type versus the other. It is diffi-
cult to obtain reliable data. Both the Portland Cement Association
(29) and NAPA (25) quote several studies in which one material has
capital and operating cost advantages over the other. Unfortunately,
the numbers from these reports are not useful because it is unclear
how the costs were established, whether equivalent designs were
compared, and whether the time value of money had been considered.
For credible analyses, detailed and comprehensive statistical data
are needed, not merely anecdotal evidence. The most promising
source of newer, presumably better, data is FHWA’s Long-Term
Pavement Performance program (30), the largest effort in pavement
research ever undertaken in the world. However, data are not yet
available from this program.
SUMMARY
When accounting for uncertainty in the data and annualizing the
environmental effects (based on assumed average service lives), the
resource input requirements and the environmental outputs of asphalt
and steel reinforced concrete pavements appear to be roughly com-
parable. However, some other environmental burdens have not been
quantified in this assessment. For example, dust emissions, water
usage in manufacturing, nonhazardous solid waste generation and
disposal, generation and disposal of hazardous waste by type, envi-
ronmental effects of landfilling, noise and vibration, visual impacts,
and other burdens are currently difficult to quantify because of a lack
of data, and—as in the case of visual impacts—lack of an acceptable
metric. If these other environmental effects had been included, our
assessment might have yielded different conclusions.
During construction, asphalt fumes might pose occupational
exposure hazards to workers in the form of respiratory problems and
eye and throat irritation. (There is no conclusive evidence that they
are carcinogenic.) However, modern paving machines have the
capability to channel fumes away from asphalt workers. In the dis-
posal phase of the two pavement types, recyclability is demonstrated
for both, but in the United States asphalt pavements have been re-
cycled in larger quantities than concrete pavements. NAPA reported
in 1997 that asphalt recycling saves over $300 million a year (25).
However, this figure is minuscule compared with the annual expen-
diture in the nation for pavement [roughly $39 billion in 1996 (31)].
Based on the LCA performed in this study, asphalt pavements
appear to be an environmentally better choice if they are recycled
effectively. However, uncertainty in the data is large and may have
skewed the results. Also, many other important environmental fac-
tors have not been included in this analysis. Regardless of the envi-
ronmental effects, special functional requirements or economics
may dictate the use of one material over the other in particular
applications.
TRANSPORTATION RESEARCH RECORD 1626
A great majority of the roads and streets in the United States have
an asphalt wearing course today. If the conclusions of this analysis
are correct, lower overall environmental effects of asphalt pavements
(due to recycling) coincide with their widespread use, providing an
example for sustainable development.
ACKNOWLEDGMENTS
The authors wish to thank the U.S. Department of Energy’s Office of
Health and Environmental Research and the National Science Foun-
dation’s Structures, Geomechanics and Building Systems program
for financial support for this work.
REFERENCES
1. Highway Statistics 1994. Report FHWA-PL-95-042. U.S. Department of
Transportation, FHWA, Office of Highway Information Management,
Washington, D.C., 1995.
2. Curran, M. A. Life Cycle Analysis. Island Press, New York, 1996.
3. Vigon, B. W., D. A. Tolle, B. W. Cornaby, H. C. Latham, C. L. Harri-
son, T. L. Boguski, R. G. Hunt, and J. D. Sellers. Life Cycle Assessment:
Inventory Guidelines and Principles. Report EPA 600/R-92/245. EPA,
Washington, D.C., Feb. 1993.
4. Hendrickson, C., A. Horvath, S. Joshi, and L. Lave. Use of Economic
Input-Output Models for Environmental Life Cycle Assessment. Envi-
ronmental Science & Technology, Vol. 3, No. 4, April 1997.
5. Lave, L. B., E. Cobas, C. Hendrickson, and F. C. McMichael. Using
Input/Output Analysis to Estimate Economy-Wide Discharges. Environ-
mental Science & Technology, Vol. 29, No. 9, Sept. 1995, pp. 420A–426A.
6. Horvath, A. Estimation of the Environmental Implications of Construc-
tion Materials and Designs Using Life Cycle Assessment Techniques.
Ph.D. thesis. Department of Civil and Environmental Engineering,
Carnegie Mellon University, Pittsburgh, Pa., June 1997.
7. AASHTO Interim Guide for Design of Pavement Structures 1972.
AASHTO, Washington, D.C., 1974.
8. Means Building Construction Cost Data 1992. R. S. Means Co.,
Kingston, Mass., 1991.
9. Input-Output Accounts of the U.S. Economy, 1992 Benchmark. Computer
Diskettes, U.S. Department of Commerce, Interindustry Economics
Division, Washington, D.C., 1997.
10. 1987–1995 Toxics Release Inventory. CD-ROM, EPA 749-C-97-003.
EPA, Office of Pollution Prevention and Toxics, Washington, D.C.,
Aug. 1997.
11. <http://www.epa.gov/ozone/title6/sec602.html>.
12. National Biennial RCRA Hazardous Waste Report 1993. Computer Out-
put, EPA, Solid Waste and Emergency Response, Washington, D.C.,
1993.
13. Air CHIEF. CD-ROM, EPA, Washington, D.C., July 1995.
14. U.S. General Accounting Office. Toxic Chemicals—EPA’s Toxic
Release Inventory Is Useful but Can Be Improved. Report GAO/RCED-
91-121. Washington, D.C., June 1991.
15. American Institute of Architects. Environmental Resource Guide. John
Wiley & Sons, New York, 1997.
16. Fog, M. H., and K. L. Nadkarni. Energy Efficiency and Fuel Substitu-
tion in the Cement Industry with Emphasis on Developing Countries.
World Bank Technical Paper 17. The World Bank, Washington D.C.,
1983.
17. Ashby, M. F. Materials and the Environment. Technical Report. Cam-
bridge University Engineering Department, Cambridge, U.K., Jan. 1992.
18. Boustead, I., and G. F. Hancock. Handbook of Industrial Energy Analy-
sis. Ellis Horwood, Chichester, U.K., 1979.
19. Horvath, A., C. Hendrickson, L. B. Lave, F. C. McMichael, and T. Wu.
Toxic Emissions Indices for Green Design and Inventory. Environmen-
tal Science & Technology, Vol. 29, No. 3, March 1995, pp. 86A–90A.
20. Bloomquist, D., G. Diamond, M. Oden, B. Ruth, and M. Tia. Engineer-
ing and Environmental Aspects of Recycled Materials for Highway Con-
struction—Final Report. Report FHWA-RD-93-088. U.S. Department
of Transportation, FHWA, and EPA, Washington, D.C., July 1993.
 Horvath and Hendrickson
21. U.S. Department of Transportation. A Study of the Use of Recycled
Paving Material: Report to Congress. Report FHWA-RD-93-147,
EPA/600/R-93/095. FHWA, Washington, D.C., June 1993.
22. 1996 Threshold Limit Values and Biological Exposure Indices. Ameri-
can Conference of Governmental Industrial Hygienists, Cincinnati,
Ohio, 1996.
23. Occupational Exposure Limits for Airborne Toxic Substances, 3rd ed.
International Labour Office, Geneva, Switzerland, 1991.
24. Goldstein, G. Pavement Recycling: Stopgap or Remedy? Civil Engi-
neering, Vol. 54, Sept. 1984, pp. 54–57.
25. http://www.hotmix.org/geninfo/faq.htm.
26. Love, R. M. State DOT Experiments with Concrete Pavement Re-
cycling. Public Works, Vol. 118, No. 6, June 1987, pp. 83–85.
27. Hendrickson, C., A. Horvath, L. Lave, and F. McMichael. Recycled
Materials Markets for the Transportation Infrastructure. TR News,
May–June 1996.
View publication stats
Paper No. 98-0661 113
28. Marland, G., and A. M. Weinberg. Longevity of Infrastructure. In Cities
and Their Vital Systems (J. H. Ausubel, and R. Herman, eds.), National
Academy Press, Washington, D.C., 1988, pp. 312–332.
29. Portland Cement Association and National Asphalt Pavement Associa-
tion. Asphalt vs. Concrete. American City & County, Vol. 101, No. 7,
July 1986, pp. 30–38.
30. Long Term Pavement Performance Information Management System—
Data Users Guide. Report FHWA-RD-93-094. FHWA, Washington,
D.C., July 1993.
31. Construction Put in Place. Engineering News-Record. Feb. 24, 1997,
p. 17.
Publication of this paper sponsored by Committee on Waste Management in
Transportation.