Example of a Hybrid Life-Cycle Assessment

of Construction Processes
Melissa Bilec, S.M.ASCE1; Robert Ries2; H. Scott Matthews, A.M.ASCE 3; and
Aurora L. Sharrard, S.M.ASCE4

Abstract: Design and construction industries, along with owners, have an increasing interest in and responsibility for the environmental
impacts of buildings. Since the environmental impacts of a building’s life cycle are considerable, quantification of all phases is important,
especially the construction phase, which is often glossed over. This research focuses on the construction phase of a building project.
Generally, the life-cycle inventory created in performing a life-cycle assessment 共LCA兲 is developed using either a process or input–output
approach; however, both techniques have distinct advantages and disadvantages. A hybrid approach combining both methods’ advantages
has been proposed and will be demonstrated. Existing hybrid models are reviewed, along with a recommendation of a hybrid model for
construction. A preliminary case study of a precast concrete parking garage’s construction using hybrid LCA methodology is presented.
Preliminary investigations indicate transportation, equipment activity, and support functions have the largest effects on the environment.
DOI: 10.1061/共ASCE兲1076-0342共2006兲12:4共207兲
CE Database subject headings: Construction management; Life cycles; Environmental issues; Buildings.

Background and Introduction
The environmental impacts of a building’s life cycle—design, raw
material extraction, processing, construction, use, and end of
life—are considerable. Buildings annually use over 70% of total
United States electricity 共EIA 2001a兲, require over 39% of United
States primary energy 共EIA 2001b兲, emit 39% of the United
States’ greenhouse gas emissions 共EIA 2003兲, contribute 136 t
million of construction and demolition waste 共EPA 1998兲, con-
sume 11% of the United States’ potable water 共USGS 2000兲, and
use 40% of raw materially globally 共USGBC 2004兲. Increasing
interest in leadership in energy and environmental design
共LEED兲, the green building rating system created and facilitated
by United States Green Building Council 共USGBC兲, demonstrates
that the owners and design and construction industries have begun
to recognize these important environmental impacts of buildings.
Previous environmental analysis of the built environment has
generally focused on areas such as building materials; energy,
1 exist because the construction industry does not report emissions
Doctoral Student, Dept. of Civil and Environmental Engineering,
Univ. of Pittsburgh, 949 Benedum Hall, 3700 O’Hara St., Pittsburgh, PA
15261 共corresponding author兲. E-mail: mmp23@pitt.edu
2
Assistant Professor, Dept. of Civil and Environmental Engineering,
Univ. of Pittsburgh, 949 Benedum Hall, 3700 O’Hara St., Pittsburgh, PA
15261. E-mail: robries@pitt.edu
3
Associate Professor, Depts. of Civil and Environmental Engineering/
Engineering and Public Policy, Carnegie Mellon Univ., 119 Porter Hall,
Pittsburgh, PA 15213. E-mail: hsm@cmu.edu
4
Doctoral Student, Dept. of Civil and Environmental Engineering,
Carnegie Mellon Univ., 119 Porter Hall, Pittsburgh, PA 15213. E-mail:
aluscher@cmu.edu
Note. Discussion open until May 1, 2007. Separate discussions must
be submitted for individual papers. To extend the closing date by one
month, a written request must be filed with the ASCE Managing Editor.
The manuscript for this paper was submitted for review and possible
publication on July 26, 2005; approved on September 22, 2005. This
paper is part of the Journal of Infrastructure Systems, Vol. 12, No. 4,
December 1, 2006. ©ASCE, ISSN 1076-0342/2006/4-207–215/$25.00.
water, and material use in operation and maintenance; construc-
tion and demolition waste management; and off-road vehicle
emissions; this research focuses on construction processes. An
environmental management tool that can be used to assess the life
cycle of buildings is life-cycle assessment 共LCA兲, a framework
that holistically evaluates the environmental effects of a product
or process by analyzing the entire life cycle of that particular
product or activity. Many LCAs utilize either the process-based
共Keoleian et al. 2001, Zapata and Gambatese 2005兲 or input–
output techniques 共Ochoa et al. 2002兲. This research examines
and develops hybrid LCA modeling for construction, which at-
tempts to combine the strengths of both techniques.
Some existing research has assumed that the impacts of the
construction phase are negligible 共Junnila and Horvath 2003兲,
while others have indicated that the environmental impacts asso-
ciated with construction are underestimated 共Hendrickson and
Horvath 2000兲. Quantification can be difficult because the con-
struction data can be inaccurate: widespread reliable data does not
to the Environmental Protection Agency 共EPA兲.
Because the environmental effects of the construction process
have not been consistently quantified, the assumption that these
effects are negligible in comparison with other building phases
has not been sufficiently explained. Even if the environmental
impacts from construction are small compared to other phases,
these impacts may be large when looked at in different temporal
and spatial frames. If one is to understand the entire life cycle of
a building, then an LCA should be performed for the construction
phase. Another possible reason this phase of the building life
cycle has not been previously advanced is due to the lack and
inconsistency of data supplied by the construction industry.
This research attempts to assess the construction phase of a
building project with a hybrid LCA model that utilizes both pro-
cess and input–output life-cycle inventory methods. This paper
provides general information on LCA, discusses existing hybrid

JOURNAL OF INFRASTRUCTURE SYSTEMS © ASCE / DECEMBER 2006 / 207

Downloaded 13 Jun 2012 to 128.237.190.83. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org
Table 1. Process LCA and EIO-LCA Model Comparisons
Process model
Advantages Detailed analysis of specific processes
Product comparisons
Identify process improvements
Disadvantages Subjective boundary selection
Lack of comprehensive data in many cases
Time and cost intensive
Proprietary data
Uncertainty
LCA models, proposes a hybrid LCA model for construction, and
presents a case study of a preliminary hybrid model for construc-
tion processes.
Methodology: LCA Modeling
Life-cycle assessment is a systematic approach to analyze and
assess the environmental impacts of a product or process over its
entire life cycle. A typical LCA includes the major stages of raw
material extraction, manufacturing, use, and end-of-life scenarios
for a product or process. Guidelines for performing an LCA are
delineated by the American National Standards Institute 共ANSI兲
and International Organization of Standardization’s 共ISO兲 14040
series 共ANSI/ISO 1997兲. LCA is an iterative four-stage process
including goal and scope definitions, life-cycle inventory 共LCI兲
analysis, life-cycle impact assessment 共LCIA兲, and interpretation.
The LCI step is often where the LCA terminates due to frame-
work, development, and subjectivity inconsistencies in the LCIA
stage. There are two approaches for generating an LCI. The pro-
cess method systematically computes the known environmental
inputs and outputs by utilizing a process flow diagram. The scope
of the process model continues to the point where the flow be-
tween process and emissions are negligible. The process approach
was further developed with the framework established in the ISO
14040 series. This approach requires data collection from pubic
sources, company or product specific information, and published
research.
Another LCI method is input-output 共I-O兲 analysis, which was
developed in terms of economics by Wassily Leontief in the
1930s 共Leontief 1936兲. Leontief developed an interdependency
model that quantifies proportional interrelationships among eco-
nomic sectors in an entire economy. I-O LCA combines national
sector-by-sector economic interaction data, which quantifies the
dependencies between sectors, with sector level environmental
effects on resource use data. Using matrix operations, a change in
economic demand from a sector can be quantified into environ-
mental or resource terms. Carnegie Mellon University has devel-
oped an I-O based LCA tool, economic input–output life-cycle
assessment 共EIO-LCA兲 共Lave et al. 1995; Hendrickson et al.
1998; Green Design Institute, Carnegie Mellon University 2004兲.
The United States economy is represented by about 500 sectors,
and many sectors represent a wide range of product types. EIO-
LCA considers both direct and indirect impacts. For example, for
the purchase of a car, direct impact would include steel, alumi-
208 / JOURNAL OF INFRASTRUCTURE SYSTEMS © ASCE / DECEMBER 2006
Downloaded 13 Jun 2012 to 128.237.190.83. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org
EIO-LCA
Boundary is defined as the entire economy
Economy-wide, system LCA
Publicly available data
Reproducible results
Aggregated level of data
Identification of process improvements are difficult
Imports treated as United States products
Uncertainty
Limited non-United States data
Product use and end-of-life options not included
num, and plastic; and also, the indirect impacts from the produc-
tion of steel as well as the entire supply chain of the automobile
through the economy.
Because there are advantages and disadvantages associated
with process and I-O methods, researchers have proposed a hy-
brid model that combines the strengths of both methods. A sum-
mary of advantages and disadvantages are listed in Table 1. One
of the major limitations of the process model is the subjective
determination of the boundary location; conversely, I-O effec-
tively eliminates the boundary issue by considering an entire
economy. As an example of the boundary issue, again consider
the production of a car, whose direct components include materi-
als like steel, aluminum, tires, plastic, and paint. For each of those
direct components, the LCA practitioner is required to trace the
inputs to environmental impacts, a process that continues both
upstream and downstream from the car until the set boundary
limits are reached. The boundary for process LCAs is typically
drawn around the direct impacts and to a point acceptable to the
LCA practitioner; whereas, the boundary using I-O is able to
capture all of the components.
While the process LCA model can be limited by boundary
definitions, I-O limitations center on the level of aggregation in
the economic sector model. With I-O LCA, evaluation of a spe-
cific product is not possible due to the aggregation of products
into sectors. For example, a comparative LCA of a Ford Focus
and a Honda Civic cannot be done with I-O LCA because there is
only one “passenger car” sector in which make and model are not
distinguished. Consequently, because specific products cannot be
modeled, it is also difficult to model process improvements; I-O
cannot easily identify the processes within car manufacturing that
could be improved. However, I-O analysis can be used for system
analysis. For example, if the United States government offered
substantial incentives for residential housing, it is possible to
model the expansion of the direct and indirect sectors that would
result in growth of the residential sector. Another major disadvan-
tage of the process LCA model is that it is often difficult for an
LCA practitioner to obtain the necessary data to complete a thor-
ough LCA either because data are not available or a company is
unwilling to contribute the proprietary information. EIO-LCA
data are based on United States publicly available sources. The
United States economic data system has a large number of sec-
tors: the number of sectors and amount of environmental data
available in other countries is not as extensive 共Suh et al. 2004兲.
Hybrid Models
The goal of a hybrid LCA is to combine the advantages of both
approaches. There are several types of hybrid models including
tiered, I-O based hybrid, integrated, and “augmented process-
based” 共Bullard et al. 1978; Joshi 2000; Suh 2004; and Guggemos
2003兲. While these models are presented distinctly, the models are
similar in that they combine process and I-O methods in some
way and differ primarily in the proportions of input–output and
process data. The models can be thought of as a continuum of
hybrid model development. Tiered hybrid analysis uses mainly
input–output data and augmented process-based uses the largest
proportion of process information.
Tiered hybrid analysis was developed by Bullard et al.
共1978兲.The model uses input–output analysis iteratively each step,
increasing the detail of the model in a process-like framework in
order to determine the energy burden of a product system. Input–
output analysis is used to determine the energy flows crossing the
process system boundary. The first approximation is at the whole-
economy level: the cost of a product is multiplied by the energy
intensity per unit gross domestic product 共GDP兲. Increasing levels
or tiers of detail can be added by associating parts of the product
system to individual I-O sectors, increasing the specificity of the
analysis. The disaggregated parts of the product system are cat-
egorized as either typical or atypical products of existing I-O
sectors. The energy requirements of typical products can be de-
termined directly from the I-O sector and energy use factors. The
atypical products require further disaggregations and an iterative
input–output approach. 共Note that Bullard uses input–output for
both typical and atypical products.兲 An error term determines the
appropriate tier based on the uncertainty goals of the analysis. All
investigation levels include adjustments for margins of error, bud-
get uncertainty, and energy intensity uncertainty. In summary, this
hybrid methodology determines the energy demand for an atypi-
cal product by disaggregating the product with budget or material
data into representative sectors with associated energy intensities.
In terms of the hybrid model continuum, the tiered hybrid analy-
sis model uses mainly input–output data. Other examples of the
tiered approach are Suh et al. 共2004兲, Moriguchi et al. 共1993兲,
Hondo et al. 共1996兲, and Munksgaard et al. 共2000兲.
The second hybrid analysis method, input–output based hybrid
analysis, focuses on disaggregating sectors according to detailed
economic information. Joshi 共2000兲 describes three models in this
category, namely Models II, III, and IV. Model II can be used
when an existing product is not well represented in an existing
I-O commodity sector or a completely new product is introduced
into the economy. If the production inputs and the environmental
burdens are known, then a new sector can be inserted into the
EIO-LCA model to determine the economywide economic and
environmental effects. However, the key assumption, and key dis-
advantage of this approach is that the original technical coeffi-
cient matrix remains unchanged. Model II is similar to the tiered
hybrid model, except environmental process data is included, in-
creasing the amount of process data included in the life-cycle
inventory. Model III creates a framework for interaction with the
technical coefficient matrix through disaggregation of a sector.
The advantages of this approach are that detailed process infor-
mation can be included without double counting and the frame-
work is presented in a consistent manner. Model IV expands the
technical and environmental matrices through an iterative process
of disaggregating existing sectors and adding new sectors for
which process LCA information is available. Model IV combines
the advantages of both process LCA and I-O, namely detailed
JOURNAL OF INFRASTRUCTURE SYSTEMS © ASCE / DECEMBER 2006 / 209
Downloaded 13 Jun 2012 to 128.237.190.83. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org
process information and a whole economy boundary, respectively.
The third method, integrated hybrid analysis, was developed
by Suh 共2004兲. Integrated hybrid analysis integrates process-
based LCA with I-O in a mathematical framework. The process
data is described in a technology matrix with physical units per
operation time for each process, while the units of the I-O model
are monetary. The process-based and I-O data for an integrated
hybrid model is linked through a make and use framework that is
connected through the flows at the boundary of each system. The
advantages of the integrated hybrid are the consistent mathemati-
cal framework for the entire life cycle, avoidance of double
counting, and easy application in analytical tools; the disadvan-
tages are that it is data and time intensive 共Suh et al. 2004兲.
An augmented process-based hybrid model was utilized by
Guggemos 共2003兲 for modeling the life cycle of a commercial
building. This hybrid approach begins with a process description
of a system and uses both EIO-LCA and process data in the
analysis. The life cycle is represented as material manufacturing,
construction, operation, maintenance, and end of life. The manu-
facturing stage is modeled by estimating the materials used and
their costs, and then using the best-matching EIO-LCA sectors.
EIO-LCA is also used to model the life-cycle inventory for
temporary materials in the construction phase, fossil fuel and
electricity in the operations phase, and lifetime maintenance
materials. Emission factors are used for on-site equipment and
transportation in the construction, maintenance, and end-of-life
stages. The result of this approach is similar to Joshi’s Model IV
because it: 共1兲 incorporates a detailed process framework; 共2兲
maintains the whole economy as the boundary of analysis; 共3兲 and
includes missing processes with process models.
The tiered, I-O based hybrid, integrated, and augmented
process-based analyses are four examples of combining process
and I-O models. Tiered and I-O based are similar in that they both
begin with I-O models and model a system primarily with I-O
data. The augmented process approach begins with a process dia-
gram, uses I-O sectors to model individual components, and uses
additional sources for processes that cannot be modeled with
I-O methods. The tiered and Model II I-O-based models use the
materials cost breakdown and result in a similar approach. Both
the tiered and I-O hybrid models rely heavily on budget informa-
tion within an I-O framework and/or the applicable emission and
energy use. The final outcome of both of these approaches may
still be a higher level of aggregation than ultimately desired since
the model relies heavily on the I-O framework. Conversely, the
augmented process-based model relies heavily on a process
framework along with relevant process and I-O data for unit pro-
cesses that cannot be modeled efficiently with process data. While
the integrated hybrid model may provide a comprehensive hybrid
framework, the time and data constraints of this model may make
this hybrid model time and cost prohibitive. After considering the
applicability of all four hybrid LCA models to construction, the
augmented process-based LCA approach will be used in the fol-
lowing construction case study. This approach was chosen be-
cause of the lack of life-cycle assessment data for construction
and the nature of the construction industry, which is primarily
driven by schedules, budgets, and quality standards. Economy-
wide public construction emissions data are lacking because the
construction industry does not consistently report emissions to the
EPA and these emissions are not typically tracked by the con-
struction industry or others. When available, construction process
data was derived from typical project records and practices; this
data will be used in conjunction with LCI processes and equip-
ment emission factors.
 Additionally, the construction industry has existing financial
construction data in estimating software and scheduling tools.
This monetary data can be applied in I-O tools such as EIO-LCA
without the need to collect additional data. Two main goals in
hybrid LCA models are improving the time and cost associated
with process-based LCAs and developing an inclusive boundary.
Using the augmented process-based hybrid LCA achieves both of
those goals, hence its selection for application to this case study.
Case Study
The goal of this case study was to create an order of magnitude
framework for the construction phase of a parking garage and
begin to develop a hybrid LCA for construction processes. The
scope of the case study is the construction phase of a precast
parking garage located in Pittsburgh. This project was selected
because the materials in a parking garage are limited. The limited
number of materials will allow for analysis of the future hybrid
LCA construction processes model. Results and methodology are
preliminary. A comprehensive description of the parking garage
construction process was developed through daily construction
reports, drawings, specifications, and utility bills. In terms of en-
vironmental impacts, the life-cycle inventory focused on emis-
sions of CO2, CH4, N2O and CFC; and SO2, CO, NO2, VOC,
lead, and PM10.
Overview of System: Precast Concrete Parking
Structure
The functional unit for this LCA is the construction phase of a
precast parking structure. The facility is a five-level parking struc-
ture located on a recently remediated brownfield site. This project
was publicly funded with four prime contractors who were
awarded the project based on the lowest responsible bid. In addi-
tion to the four primes, an agency construction manager was hired
by the owner to oversee daily construction activity. Construction
began at the beginning of 2003 and was completed in October
2003. The parking facility has 377 spaces and is supported by a
deep foundation with augercast piles, pile caps, and grade beams.
The main structural system is precast concrete with the main
components of pre-topped double tees, columns, lite walls, span-
drels, panels, stairs, and hollowcore planks.
The construction process is a complex system, and the com-
plete inclusion of every construction activity, material purchase,
and delivery is not possible. This analysis is a high-level study
that begins the framework to help understand the most critical
elements of the construction process. The initial assumptions for
the case study are as follows:
1. Once the precast member was delivered to the site, the mem-
ber was moved off the trailer bed by a crane and erected. It is
assumed that the precast member was not stored in a staging
area on-site and double-handling of material did not occur.
The transportation of the construction equipment to the con-
struction site was not considered due to lack of available
data.
2. The dust created during construction activities was not in-
cluded because the specifications required dust control mea-
sures and the monitored data on dust generation was not
available. Storm water runoff quantity and quality were not
included in this analysis since a management plan was imple-
mented. The storm water management plan was approved by
the Allegheny County Conservation District and included el-
210 / JOURNAL OF INFRASTRUCTURE SYSTEMS © ASCE / DECEMBER 2006
Downloaded 13 Jun 2012 to 128.237.190.83. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org
Fig. 1. Building life-cycle and LCA boundary
ements such as inlet basins covers, silt fencing, and rock
construction entrances.
3. The use of small tools was considered only in terms of elec-
tricity usage. The electric switch gear and transformer were
not included because this equipment was shared with adja-
cent buildings.
Hybrid Construction Model
This section describes the development of the hybrid construction
model along with more specific information on the considered
impacts and modeling.
1. Process flow diagram. A general construction process flow
diagram was developed. For this case study, the process flow
diagram was at a relatively high level.
2. Determination of the boundary. The boundary for this LCA is
only the construction and design phases represented as sup-
port services 共see Table 3兲, while the entire building life
cycle includes design, raw material extraction and process-
ing, construction and design, operations and maintenance,
and end of life. Fig. 1 illustrates the building life cycle and
boundary. Included in the boundary is the transportation of
the final building material from the suppliers to the construc-
tion site and support services. It is important to note that this
transport is for the building material from the final supplier
to the construction site. For example, the transportation of
the precast concrete does not include either the transportation
of the raw materials or the mixing of the concrete at the
plant, only the transportation of the final finished precast
member from the plant to the job site. The boundary was
selected to focus on construction processes in more detail.
3. Process inventory data assessment. The process inventory
data assessment was completed by gathering all available
process information. Data collection included information
from the architect, engineer, prime contractors, and owner.
The main data sources include construction drawings, tech-
nical specifications, approved submittals, bid tabulation, and
schedule. Daily construction reports 共DCRs兲 from four
primes and the on-site construction manager were used for
determining the daily activity and daily job progress, such as
the number of workers and the on-site equipment use. The
main data extracted from these sources related concrete
quantity, precast weight and configuration, amount of excess
soil removed, equipment durations and types, and electricity
and water usage. The DCRs were used to determine con-
Table 2. Construction Process and Model Used
Construction process
Transportation
Construction equipment in terms of diesel usage and emissions
from combustion
Construction service sectors
Upstream production and maintenance effects of construction
equipment
On-site electricity usage
On-site water consumption
struction equipment quantities and duration of use. The con-
crete delivery tickets for the poured concrete and the supplier
information were obtained from the general contractor. A da-
tabase including the quantity, weight, date, delivery time in-
formation, and building location 共when available兲 was
created. The concrete data for the augercast piles was deter-
mined from the construction drawings and submittals. For
the precast concrete, the concrete piece drawings were ob-
tained from the precast manufacturers. Each piece drawing
indicates the weight and quantity. This information was used
to determine the ton-kilometers for precast transport. Be-
cause the site had excess soil and was a brownfield site, all of
the excess soil was transported to a state approved residual
waste landfill. The manifests indicated the amount of mate-
rial transported and the weight; this information was used to
determine the ton-kilometers for excess soil transportation.
On-site electricity and water bills were also obtained for
quantifying on-site utility use.
4. EIO-LCA inventory assessment. After analyzing the process
inventory data, EIO-LCA data were obtained to augment the
process data and to identify and include major impact areas
not quantified in the process data, namely, service sectors and
upstream effects, and maintenance of construction equip-
ment. The 1992 model of EIO-LCA was selected over the
1997 model because of uncertainties in the construction
sector from the Standard Industry Classification 共SIC兲 and
North American Industry Classification System 共NAICS兲
conversion.
5. Identification of impact processes. Six main processes
emerged from this data: 共1兲 transportation; 共2兲 construction
equipment in terms of diesel usage and combustion emis-
sions; 共3兲 construction service sectors; 共4兲 upstream and
maintenance of construction equipment; 共5兲 on-site electric-
ity usage; and 共6兲 on-site water consumption. The six catego-
ries and their respective methodologies are shown in Table 2
along with respective software tool information. Additional
information on each impact follows.
6. Modeling. The hybrid construction model was implemented
in an LCA process modeling software tool, SimaPro 5.0. Si-
maPro is a widely used process based LCA software program
JOURNAL OF INFRASTRUCTURE SYSTEMS © ASCE / DECEMBER 2006 / 211
Downloaded 13 Jun 2012 to 128.237.190.83. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org
Methodology and tool
Process LCA
40 t truck and 28 t truck
SimaPro
Process LCA
Diesel equipment
SimaPro
Input–output LCA
Selected supply chain sectors from new construction office, industrial, and
commercial buildings
EIO-LCA
Input–output LCA
Selected supply chain sectors from new construction office, industrial, and
commercial buildings
EIO-LCA
Process LCA
U.S. average electricity
SimaPro
Quantity determined
with extensive databases and different impact assessment
methods. With SimaPro, the LCA practitioner develops an
LCA by using the software’s existing processes or creating
new processes 共SimaPro 2001兲. The life-cycle inventory data
for transportation, construction equipment’s diesel usage and
emissions, and on-site electricity usage are derived from pro-
cesses in the software tool. New processes were created from
EIO-LCA data for the construction service sectors and equip-
ment manufacturing and maintenance. Water use was quan-
tified in terms of usage; however, treatment and distribution
of the water is not included in the assessment. The hybrid
construction model inventory was then generated with the
LCA software tool. The results in this paper only show the
life-cycle inventory, not the results from the life-cycle impact
assessment 共LCIA兲.
Transportation
The transportation processes include the delivery of concrete, pre-
cast concrete, and excess material to and from the site. The data
for concrete delivery was obtained from concrete tickets. The
critical information for this analysis includes the distance from
the plant to the site, the weight of the loaded truck, the return
distance, and the weight of the unloaded truck. The assumed
weight of the empty concrete truck was 11,793 kg 共26,000 lb兲;
total ton-kilometers were determined from this data.
This analysis does not include the idle time of the trucks, since
that data was not available. Because the concrete tickets for the
augercast piles were not available, the transportation of the con-
crete for the augercast piles was derived from the other concrete
data. The same distances were used and an average loaded truck
weight was determined. A quantity takeoff was done to determine
the amount of concrete in the augercast piles and the total ton-
kilometers were calculated. All concrete trucks were assumed to
return empty.
The transportation 共in ton-kilometers兲 for precast pieces was
determined by adding the precast piece weight plus the average
weight of the tractor trailer and multiplying the total tons by the
distance traveled. Since three different precast manufacturers
were used on this project, three different distances were utilized
Table 3. Construction Sectors
Construction service sectors
共A兲
Engineering, architectural, and surveying services
Wholesale trade
Retail trade, except eating and drinking
Real estate agents, managers, operators, and lessors
Management and public relations services
Banking
Other business services
Advertising
Accounting, auditing and bookkeeping, and miscellaneous
services, n.e.c.
Computer and data processing services
Legal services
Insurance carriers
U.S. Postal Service
Security and commodity brokers
Insurance agents, brokers, and services
Motion picture services and theaters
Credit agencies other than banks
Signs and advertising specialties
depending on the piece. The precast pieces were evaluated for
configuration and sizing to determine the number of pieces per
shipment. It was assumed that all precast trucks returned to their
respective manufacturing facilities empty.
Total ton-kilometers for the excess soil transport was deter-
mined by multiplying the total loaded truck weight by its weight
and the distance from the job site to the disposal facility. For the
return trip, the total unloaded truck weight was multiplied by the
distance from the disposal site to the job site. It was assumed that
all trucks returned empty from the disposal facility and did not
return to the truck’s original destination, the trucking facility.
The LCI for these three transportation activities used existing
process descriptions in the LCA software tool. Concrete and pre-
cast deliveries were modeled with a 40-t truck and the excess soil
transport was modeled with a 28-t truck. The processes are based
on average fuel consumption and emissions factors. Efficiencies
were adjusted to reflect actual activity in this case study. The
process includes construction of the roadway infrastructure,
manufacturing the trucks, direct energy, combustion emissions,
materials used during the use phase, and end-of-life scenarios.
The emission from the combustion of fuel was considered 共PRe
Consultants and ESU-Services 2001兲.
Construction Equipment: Diesel Usage
and Combustion Emissions
The DCRs contain information related to the weather, daily ac-
tivities, problems, safety incidents, number and type of workers,
and equipment used with hours of operation. The list of construc-
tion equipment included in this assessment is shown in Fig. 1.
The critical data for this analysis were the equipment used and the
respective hours of operation. In some instances when the model
was not recorded on the DCRs, assumptions were made based on
project knowledge and photographs. After models, makes, and
engines were determined and in some cases assumed for each
piece of equipment, the manufacturer’s websites were consulted
for the associated horsepower. The amount of diesel fuel used in
gallons was determined by multiplying DCR hours by horse-
212 / JOURNAL OF INFRASTRUCTURE SYSTEMS © ASCE / DECEMBER 2006
Downloaded 13 Jun 2012 to 128.237.190.83. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org
Construction equipment and maintenance sectors
共B兲
Miscellaneous equipment rental and leasing
General industrial machinery and equipment, n.e.c.
Industrial and commercial machinery and equipment, n.e.c.
Hoists, cranes, and monorails
Motors and generators
Electrical equipment for internal combustion engines
Electrical machinery, equipment, and supplies, n.e.c.
Fluid power equipment
Costruction machinery and equipment
Hand and edge tools, except machine tools and handsaws
Power-driven handtools
Transportation equipment, n.e.c.
power by 0.04 共Peurifoy and Schexnayder 2002兲. This formula
was developed for diesel engines operating under standard condi-
tions. The equipment processes were then modeled in the LCA
software tool using the existing unit process, Diesel Equipment,
which accounts for extraction, production, and combustion 共PRe
Consultants 2001兲.
Construction Service Sectors
The construction service sectors portion of the construction pro-
cess were modeled using EIO-LCA. The total cost of the con-
struction project was $5,440,000, but because EIO-LCA requires
producer’s price in lieu of the purchaser’s price along with an
assumption regarding profit, a dollar amount of $5,000,000 was
input into the EIO-LCA sector: “office, industrial, and commer-
cial buildings.” The data extracted included economic, conven-
tional pollutants, global warming potential, and energy usage,
which was provided for all 485 sectors, using the U.S. Dept of
Commerce 1992 EIO-LCA database. Results were then subjec-
tively analyzed to determine the relevant construction service sec-
tors 共e.g., architects, engineers, and surveyors兲. Column 共A兲 of
Table 3 summarizes the construction service sectors selected from
the EIO-LCA model.
Upstream Production and Maintenance
of Construction Equipment
“Upstream and maintenance effects of construction equipment”
refers to allocating the environmental impacts from the manufac-
turing of the construction equipment as well as other life-cycle
stages related to the equipment. The same EIO-LCA approach
used for the construction service sectors was used for this process
as well. The results from the $5,000,000 producer’s price were
subjectively analyzed to determine the sectors related to construc-
tion equipment, and the data from the construction equipment
categories were totaled and conversions were performed for the
LCA software tool. The EIO-LCA sectors selected are listed in
Table 3, Column 共B兲. It should be noted that this equipment pro-
Table 4. Transportation Table 6. Transportation Emissions Comparison between Two Precast
Bidders
Transportation t/km
Emission Lowest bid Second lowest bid
Precast delivery 3,783,100
Excess soil delivery 270,515 CO2E 共kg兲 397,495 124,790
Concrete delivery 21,105 SO2 共kg兲 975 340
CO 共kg兲 1,015 350
NO2 共kg兲 3,135 1,065
VOC 共kg兲 1,420 490
cess does not include the truck or concrete sectors because this
PM10 共kg兲 210 70
information is included in the transportation process.

On-Site Electricity Usage and Water Consumption

Utility bills were obtained from the general contractor. The on-
site electricity bill was provided for both the trailers and the con-
struction site, while the water bills were only for the construction
site itself, as the trailers did not have publicly supplied water. This
data were used to determine the electricity and water consumption
for the project. Electricity usage was mainly for small equipment,
temporary lighting, and trailers, while water usage was used gen-
erally during pile cap installation, concrete pours, and clean up.
The total metered electricity and water usage for the project was
92,700 kW· h and 227,124 L, respectively.
Life-Cycle Inventory Analysis
The life-cycle inventory analysis is shown in Tables 4 and 5.
Table 4 summarizes the total ton-kilometers for each construction
process included in the analysis. Precast delivery is the most sig-
nificant transportation process with 3,783,100 t-km. The data in
Table 4, along with other data, were used to determine construc-
tion emissions 共Table 5兲. Global warming potential 共GWP兲 values
have been converted to CO2 equivalents 共CO2E兲 using the Adri-
aanse Report 共Adriaanse 1993兲. In terms of GWP, transportation
has the highest impact, followed by equipment 共diesel兲, and con-
struction services. Transportation has the highest impact in terms
of SO2, NO2, VOC, lead, and PM10. The construction service
sectors produce the highest amount of CO. The environmental
impacts of on-site electricity usage were negligible.
Discussion
Of the construction activities included in this analysis, transpor-
tation contributed the highest environmental impacts in all cat-
egories except two. The transportation of the precast concrete was
the most significant. The lowest bidder’s plants and subcontrac-
tor’s plants were located in three different locations with the
farthest in Toronto. Conversely, the batch concrete plant’s loca-
tion was less than 8 km away, potentially minimizing the impacts
of transportation. Not included in this analysis was idling time,
which could be significant due to the long wait times at the
United States–Canada border.
Since this was a publicly bid project, the owner accepted the
lowest responsible bidder; however, the second lowest bidder’s
architectural precast plant was closer, 200 km, versus the plant in
Toronto at 530 km. A comparison in terms of impact indicators
between the lowest bid and the second lowest bid is shown in
Table 6. The second lowest bidder was an average of 33% lower
in most emission categories. If social externalities for emissions
are added to bid prices, then the lowest bidder’s bid price would
have increased by about $14,400 and the second lowest bidder’s
price would have increased by about $5,000 共Matthews and Lave
2000兲.
If one would calculate the social costs of the additional emis-
sions, the environmental impacts associated with transportation
illustrate the importance of promoting the purchase of local ma-
terials, especially high mass materials such as precast concrete
and steel. Local materials are addressed in Leadership in Envi-
ronmental and Economic Design (LEED)—New Construction,
Version 2.1 green building rating system 共USGBC 2002兲. Credit
can be obtained if 20% of the building’s material is purchased
“locally”, where “locally” is defined as within a 800 km radius
from the site. In this case, the precast concrete from Canada
would have been considered a local product. Since this analysis
indicates that transportation is dominant in most impact catego-
ries, in the construction phase, additional incentives for shorter
distances should be considered.
The NOx emissions from the equipment fuel combustion were
the second highest. In May 2004, the EPA finalized a comprehen-
sive rule to reduce emissions from nonroad diesel engines 共i.e.,
construction equipment兲. The overall goal was to reduce PM and
NOx by 90%. As a part of the implementation of the program, the
EPA developed a voluntary program in conjunction with the As-
sociated General Contractors of America 共AGC兲 entitled “Clean
construction USA” 共EPA 2005兲. This program encourages the use
of emission control technologies by promoting retrofit incentives
and technical assistance. The diesel equipment process selected in

Table 5. Emissions
Total
SO2 CO NO2 VOC Lead PM10 CO2E
Construction process 共kg兲 共kg兲 共kg兲 共kg兲 共kg兲 共kg兲 共kg兲
Construction services 25 1,880 1,115 410 0 40 125,680
Upstream and maint. construction equipment 5 95 5 40 0 0 11,860
Transportation 975 1,015 3,135 1,420 0 210 397,495
Equipment 共diesel兲 285 540 2,385 445 0 165 128,115
On-site electricity 140 10 65 15 0 5 18,570

JOURNAL OF INFRASTRUCTURE SYSTEMS © ASCE / DECEMBER 2006 / 213

Downloaded 13 Jun 2012 to 128.237.190.83. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org
the LCA software tool to determine the emissions for this case
study did not include the new regulations.
Many process LCAs do not include the associated service sec-
tors; excluding these sectors can lead to an assessment that may
greatly underestimate the environmental emissions. For this as
sessment, the construction services produced the most CO and
were the third highest process with respect to GWP total, SO2,
NO2, VOC, and PM10 created. If construction service sectors
were excluded from this LCA, GWP would be approximately
20% lower than the total GWP for the entire construction process.
Finally, SimaPro software was used with applicable processes.
Several commercial LCA software packages are available, and
results can vary depending on the processes selected and software
used.
Conclusion and Future Research
Life-cycle assessment methodology can be used to holistically
evaluate construction processes. After reviewing several types of
hybrid models: tiered, input–output, integrated, and augmented
process-based, an augmented process-based hybrid model was
used for the case study of the construction of a precast parking
garage. The dominant process was transportation, demonstrating
the importance of setting standards related to procurement of
local materials. Including construction service sectors was shown
to be significant 共20% of total GWP兲; excluding the service sec-
tors can lead to an incomplete analysis that may not represent all
of the environmental impacts. Future research will further de-
velop the hybrid model. The case study assisted in creating a
conceptual model and uncovering potential development areas.
One example is the use of EPA’s Nonroad program, which will
allow for additional modeling of construction equipment in terms
of type, horsepower, and emission standards. Additionally, this
case study utilized the producer’s construction contract value in
EIO-LCA and then subjectively selected the service and construc-
tion equipment sectors considered relevant. An alternative ap-
proach would be to determine the actual contract values of each
support sector 共i.e., architect, engineer, and construction manager兲
and determine the relevant sectors for these activities. The results
would then be direct impacts, instead of indirect impacts. Simi-
larly, it is possible to determine the dollar values of the equip-
ment, allocate these prices according to expected engine hours
used on the project divided by the expected lifetime engine hours,
and utilize these as direct inputs instead of indirect results. The
model will also be expanded to include estimates of productivity
and contract durations and will include several more construction
processes. Additional research will be explored between the envi-
ronmental externalities, bidding processes, and project delivery
methods; specifically, developing the relationship between includ-
ing the environmental costs with submitted bid prices.
The case study also revealed other areas necessitating further
investigation including surveys of on-site energy expenditures,
construction equipment with associated emissions, and an in-
depth investigation of construction sectors as represented in the
current NAICS. On-site energy sources are usually provided by
diesel, electricity, and natural gas. Determining the energy expen-
ditures can be roughly estimated using Census reports for the
construction sectors 共US Department of Commerce 2000兲; how-
ever, more site specific information is needed according to con-
struction project, cost, and duration to further understand energy
usage per construction project. This information is currently being
collected as a part of this research 共Matthews et al. 2005兲. There
214 / JOURNAL OF INFRASTRUCTURE SYSTEMS © ASCE / DECEMBER 2006
Downloaded 13 Jun 2012 to 128.237.190.83. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org
is a need to more fully understand the sector mapping for con-
struction between U.S. Bureau of Economic Analysis’s I-O tables,
NAICS, and EIO-LCA tool. For most other sectors in the
economy, there is an existing mapping between I-O sectors and
NAICS industries. Attempts are being made to produce a clear
mapping between I-O sectors, NAICS industries, and EIO-LCA
construction sectors.
Acknowledgments
Financial support for this research was provided by National
Science Foundation Grant No. CMS-0327878.
References
Adriaanse, A. 共1993兲. “Environmental policy performance indicators-A
study on the development of indicators for environmental policy in
The Netherlands.” SDU Publishers, Hague, The Netherlands.
American National Standards Institute/International Organization for
Standardization 共ANI/ISO兲. 共1997兲. “Environmental management—
Life cycle assessment—Principals and framework,” Geneva,
Switzerland.
Bullard, C. W., Penner, P. S., and Pilati, D. A. 共1978兲. “Net energy analy-
sis, handbook for combining process and input-output analysis.” Re-
sources and Energy, 1共3兲, 267–313.
Green Design Institute, Carnegie Mellon University 共2004兲. “Economic
input–output analysis—Life cycle assessment.” 具http://www.eiolca.
net典 共November 2004兲.
Guggemos, A. 共2003兲. “Environmental impacts of on-site construction:
Focus on structural frames.” Doctoral dissertation, Univ. of Califor-
nia, Berkley, Calif.
Hendrickson, C., and Horvath, A. 共2000兲. “Resource use and environ-
mental emissions of U. S. construction sectors.” J. Constr. Eng. Man-
age., 126共1兲, 38–44.
Hendrickson, C., Horvath, A., Joshi, S., and Lave, L. 共1998兲. “Economic
input–output models for environmental life-cycle assessment.” Envi-
ron. Sci. Technol., 32共7兲, 184A–191A.
Hondo, H., Nishimura, K., and Uchiyama, Y. 共1996兲. “Energy require-
ments and CO2 emissions in the production of good and services:
Applications of an input-output table to life cycle analysis.” CRIEPI
Rep. No. Y95013, Central Research Institute of Electric Power Indus-
try, Tokyo.
Joshi, S. 共2000兲. “Product environmental life-cycle assessment using
input-output techniques.” J. Ind. Ecol., 3共2&3兲, 95–120.
Junnila, S., and Horvath, A. 共2003兲. “Life-cycle environmental effects of
an office building.” J. Infrastruct. Syst., 9共4兲, 157–166.
Keoleian, G., et al. 共2001兲. “Life-cycle energy, costs, and strategies for
improving a single-family house.” J. Ind. Ecol., 4共2兲, 135–156.
Lave, L., Cobas-Flores, E., Hendrickson, C., and McMichael, F. 共1995兲.
“Using input-output analysis to estimate economy wide discharges.”
Environ. Sci. Technol., 29共9兲, 420A–426A.
Leontief, W. W. 共1936兲. “Quantitative input and output relations in the
economic systems of the Unites States.” Rev. Econ. Stat., 18共3兲,
105–125.
Matthews, H. S., and Lave, L. 共2000兲. “Applications of environmental
valuation for determining externality costs.” Environ. Sci. Technol.,
34共8兲, 1390–1395.
Matthews, H. S., Roth, M., Sharrard, A., and Bilec, M. 共2005兲. “Eco-
nomic and environmental implications of construction energy use
under new EPA emission standards.” Proc., Construction Research
Congress, San Diego.
Moriguchi, Y., Kondo, Y., and Shimizu, H. 共1993兲. “Analyzing the life
cycle impacts of Cars: The case of CO2” Industry and Environment,
16共1–2兲, 42–45.
Munksgaard, J., Pedersen, K. A., and Wier, M. 共2000兲. “Impacts of
 household consumption on CO2 emissions.” Energy Economics,
22共4兲, 423–440.
Ochoa, L., Hendrickson, C., and Matthews, H. S. 共2002兲. “Economic
input-output life-cycle assessment of U. S. residential buildings.” J.
Infrastruct. Syst., 8共4兲, 132–138.
Peurifoy, R. L., and Schexnayder, C. J. 共2002兲. Construction planning,
equipment, and methods, 6th ed., McGraw Hill, New York.
PRe Consultants. 共2001兲. SimaPro 5.0, database manual, the Franklin US
LCI Library, Written by Sylvatica, Greg Norris, The Netherlands.
PRe Consultants and ESU Services. 共2001兲. SimaPro 5.0, database
manual, The ETH-ESU 96 Library. Written by ESU-Services,
R. Frischknecht and N. Jungbluth, The Netherlands.
SimaPro 5.0 共2001兲. “SimaPro 5.0 software.” PRe Consultants, The
Netherlands.
Suh, S. 共2004兲. “Functions, commodities and environmental impacts in an
ecological-economic model.” Ecological Economics, 48, 451–467.
Suh, S., et al. 共2004兲. “System boundary selection in life-cycle invento-
ries.” Environ. Sci. Technol., 38共3兲, 657–664.
U.S. Department of Commerce. 共2000兲. “1997 Economic census, Industry
summary: Construction, subject series.” EC97C23S-IS, Washington,
D.C.
U.S. Energy Information Administration 共EIA兲. 共2001a兲. “Table F11:
Electricity consumption, price, and expenditure estimates by
sector, 2001.” 具http://www.eia.doe.gov/emeu/states/sep_fuel/html/
fuel_es.html典 共July, 13 2005兲.
JOURNAL OF INFRASTRUCTURE SYSTEMS © ASCE / DECEMBER 2006 / 215
Downloaded 13 Jun 2012 to 128.237.190.83. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org
U.S. Energy Information Administration 共EIA兲. 共2001b兲. “Table F18:
Total energy consumption, price, and expenditure by sector,
2001.” 具http://www.eia.doe.gov/emeu/states/sep_fuel/html/fuel_te.
html典 共July, 13 2005兲.
U.S. Energy Information Administration 共EIA兲. 共2003兲. “Figure 110. Car-
bon dioxide emissions by sector and fuel, 2003 and 2005.” 具http://
www.eia.doe.gov. oiaf/aeo/excel/figure110_data.xls典 共July, 25 2005兲.
U.S. Environmental Protection Agency 共EPA兲. 共1998兲. “Characterization
of building related construction and demolition debris in the United
States,” Rep. No. EPA 530-R-98-010, Prepared by Franklin
Associates.
U.S. Geological Survey 共USGS兲. 共2000兲. “Estimated use of water in the
United State in 2000.” 具http://water.usgs.gov/pubs/circ/2004/circ1268/
htdocs/text-total.html典 共July 17, 2005兲.
U.S. Green Building Council 共USGBC兲. 共2002兲. Green building rating
system for new construction and major renovations (LEED-NC),
version 2.1. 共November 2002兲.
U.S. Green Building Council 共USGBC兲. 共2004兲. “Global use of LEED.”
具http://www.usgbc.org/Docs/About/usgbc_intro.ppt典 共April 20, 2004兲.
Zapata, P., and Gambatese, J. A. 共2005兲. “Energy consumption of asphalt
and reinforced concrete pavement materials and construction.” J. In-
frastruct. Syst., 11共1兲, 9–20.
U.S. Environmental Protection Agency 共EPA兲. 共2005兲. “Clean diesel.”
具http://www.epa.gov/cleandiesel/典 共May 12, 2005兲.