A P P L I C A T I O N S A N D I M P L E M E N T A T I O Nh.docx

A P P L I C A T I O N S A N D I M P L E M E N T A T I O N
http://mi tpress.mit.edu/JIE Journal of Industrial Ecolog y 107
© Copyright 2002 by the
Massachusett s Institute of Technology
and Yale University
Volume 6, Number 1
The Greening of a Pulp and
Paper Mill
International Paper’s Androscoggin Mill,
Jay, Maine
Marquita Hill, Thomas Saviello, and Stephen Groves
Keywords
corporate social responsibility (CSR)
kraft pulping
pollution prevention (P2)
public par ticipation
pulp and paper industry
toxics use reduction (TUR)
Address correspondence to:
Marquita Hill
Department of Chemical Engineering
University of Maine
5737 Jenness Hall
Orono, ME 04469-5737 , USA

[email protected]
Summary
International Paper (IP), the world’s largest forest products
company, owns the Androscoggin Mill, a large pulp and paper
mill in Jay, Maine, in the nor theastern United States. This case
study describes the transformation of the Androscoggin Mill
from an object of public opprobrium and con�ict to a show-
case for environmental management. In the late 1980s, an 18-
month strike had embittered workers and townspeople and
left the mill’s reputation in tatters. In response to mill environ-
mental violations, some of which were considered crimes by
state regulators, the town of Jay passed its own environmental
ordinance to control mill emissions. Early in the 1990s, new
management, including two former corporate-level employ-
ees, sought to change the mill’s business approach and turn
the Androscoggin Mill into IP’s best environmental performer.
The initial emphasis on establishing and maintaining compli-
ance was expanded to include aggressive pollution prevention
effor ts that led to cooperative projects with the Maine De-
par tment of Environmental Protection, the U.S. Environmental
Protection Agency, and stakeholder groups. The mill’s ap-
proach in the 1990s evolved further to essentially follow prin-
ciples of industrial ecology. New approaches focused on “clos-
ing the loop” by �nding bene�cial uses for previously
land�lled
wastes, replacements for most hazardous chemicals, and re-
ductions in solid and hazardous waste generation. The mill also
pursued the establishment of symbiotic relationships with a
facility that began using a mill by-product on-site and an on-
site natural gas burning facility that provided par t of the mill’s
steam demand. IP also established a public advisory committee
in 1992 to advise management on operational and “big-
picture” issues, which later included the application of sustain-
ability criteria to the mill. IP has since formed community ad-
visory committees at each of their integrated pulp and paper

mills.
108 Journal of Industrial Ecolog y
Introduction
From 1965, when it was built, until 1986, In-
ternational Paper’s (IP’s) Androscoggin Mill, lo-
cated on the Androscoggin River in Jay, Maine,
was a typical large pulp and paper operation. It
was the economic mainstay for Jay, a community
of 5,000 in western Maine, and was considered
no better or worse than other mills. A strike that
began in 1987 and lasted into 1988 greatly
changed the company’s reputation, embittering
mill workers and the Jay community. Environ-
mental problems that had previously received lit-
tle notice brought a quick and negative reaction.
State and local authorities cited the mill for vi-
olating land�ll and air emissions requirements. In
July 1991 , �ve criminal indictments were
brought against the mill, alleging misrepresen-
tations in its wastewater license application and
the burning of unlicensed waste. The mill’s abil-
ity to continue operating was also at risk, as its
land�ll was close to capacity at a time when it
looked unlikely that it could obtain a permit for
expansion.
Against this backdrop, the next decade saw a
profound change in how the mill was managed;
its relationship to its workers, the Jay community,

regulators, and other stakeholders; and its envi-
ronmental performance. In effect, IP’s worst per-
former became its best. This article examines the
drivers of change at the mill, speci�cally new
management, the Jay environmental ordinance,
the relationship of the mill with its regulatory
agencies, and the activities of its public advisory
committee (PAC). It shows the evolution of
thinking on environmental issues as the mill be-
gan with end-of-pipe control measures, initiated
pollution prevention projects, began treating
mill by-products as useful materials, and worked
with colocated facilities in an industrial symbi-
otic relationship (Ehrenfeld and Gertler 1997) to
reduce or eliminate pollution. We explore the
reasons for the establishment of a formal advisory
process, the activities of the PAC, and its pro-
gression from nuts and bolts issues to a focus on
sustainability. The conclusions focus on how this
transformation occurred and the nature of the
drivers for change in both the early stages and as
the company’s efforts matured in the 1990s on
into the 2000s.
Background to Change
The Androscoggin Mill
At the end of 2001, the Androscoggin Mill
employed about 1,200 people, 150 of whom were
in salaried positions. The mill uses the kraft pulp-
ing process to produce about 1,600 tons of mostly
coated paper per day, plus some specialty-grade
papers and dried pulp. Pulping is the process of
taking wood �ber and turning it into the raw ma-
terial for making paper, paperboard, and card-

board. Appendix 1 describes the chemical pulp-
ing process. The Androscoggin Mill has two
Kamyr digesters, two recovery boilers, one waste-
fuel incinerator, two limekilns, two bleaching
lines, �ve paper machines, a ground-wood mill,
and a �ash dryer. Pulp and paper mills are
resource-intensive operations that can have a
signi�cant impact on the environment (Servos
et al. 1996; Springer 2000a).
IP, the mill’s owner, is based in Connecticut
and is the world’s largest forest products com-
pany, with about 117,000 employees worldwide.
It is vertically integrated, owning the raw mate-
rials (forests and wood lots), producing inter-
mediates (pulp and chemicals), and manufactur-
ing products for business and consumer markets.
Its businesses include manufacturing printing pa-
per, packaging, building materials, and chemi-
cals.
Corporate Level
In 1990, President George H. Bush estab-
lished the President’s Commission on Environ-
mental Quality (PCEQ) to seek advice from the
private sector on environmental issues (PCEQ
1990). IP and other major corporations were in-
vited to become members. David Critch�eld, IP’s
corporate director of regulatory affairs and recy-
cling, typically attended meetings on behalf of
John Georges, then chief executive of�cer. The
PCEQ, staffed by Michael Deland, chair of the
White House Council for Environmental Qual-
ity, urged companies to integrate pollution pre-
vention principles into corporate environmental

programs, test new strategies, and share results.
Several commission recommendations became
relevant to the Androscoggin Mill:
Hill, Saviello, and Groves, The Greening of a Pulp and Pa per
Mill 109
� Pursue pollution prevention projects
� Form public participation groups in com-
munities where they operated and become
more open to community involvement and
input
� Take one facility and develop it into an
environmental model, from which other
facilities could learn
Most of the companies that adopted this last rec-
ommendation focused on improvin g well-
performing facilities. In contrast, IP chose to take
its worst facility, the Androscoggin Mill, and
make a conscious effort to turn it into its best.
New Mill Management
In 1990, IP asked Larry Stowell, then a cor-
porate manufacturing manager, to become man-
ager of the Androscoggin Mill. David Critch-
�eld, in a September 2001 interview in Portland,
Maine, said Stowell was basically asked to “turn
the mill around and keep it out of the headlines.”

He said that IP corporate policy as applied to all
mills meant that “we will comply with regula-
tions, period, and we will strive to minimize our
environmental impacts.” In 1991 Thomas Sav-
iello joined the mill as environmental superin-
tendent and, in 1992, Stowell persuaded David
Critch�eld to leave his corporate-level position
and join the mill as environmental manager, su-
pervising Saviello. The three immediately fo-
cused on the need to set very high standards for
the mill while, at the same time, working to en-
gender the trust of regulators and the local com-
munity. In�uenced by the PCEQ recommenda-
tions, they took on the challenge of making the
Androscoggin facility, still reeling from viola-
tions and criminal indictments, into IP’s model
of environmental excellence. Also, as suggested
by the PCEQ, they formed a PAC to provide an
outside source of perspective that might help the
mill mitigate the negative perceptions of the me-
dia and others. The PAC, however, chose to stay
out of the public domain and evolved in its own
way (see below).
David Critch�eld described his approach to
the Androscoggin Mill as systematic, regularly
consulting a text on systems analysis (Church-
man 1968): “To approach something as compli-
cated as a paper mill, you have to be systematic.
Otherwise you would �nd yourself trying to move
in all directions at once. Too many problems
needed a solution.” Of necessity, his �rst priori-
ties were improving a terrible safety record and
bringing the mill into compliance with environ-
mental regulations. Saviello had already reorga-

nized the mill’s environmental department, hir-
ing specialists in air, water, hazardous waste, and
solid waste to mirror the regulatory structure of
the Maine Department of Environmental Pro-
tection (DEP).
The Early 1990s
The Jay Ordinance
As a result of citizens’ poor regard of mill en-
vironmental performance, the town of Jay insti-
tuted its own environmental ordinance, subject-
ing the mill to more restrictive local regulations
along with those already established by the state
and federal governments. This local ordinance
became a driver of change when it required more
monitoring wells for the mill land�ll than re-
quired by the state and mandated the adoption
of technologies that were not required at other
kraft mills. For example, the mill installed a re-
generative thermal oxidizer in 1994 at the town’s
behest to capture and destroy odorous chemicals.
Subsequently, the U.S. Environmental Protec-
tion Agency (U.S. EPA) has adopted a regula-
tion that requires such emission controls by
2006.
Gaining Employee Cooperation
In a large pulp and paper mill, management
of complex environmental programs requires the
support and participation of workers and man-
agers. This was dif�cult in an environment in
which workplace communications were still poor
after the strike and mill environmental infrac-

tions remained common. The recovery boilers
were averaging 56 opacity1 incidents a year. The
incidents resulted from a lack of consistency in
controlling boiler operations and recognizing
which conditions would lead to an infraction.
The number of opacity incidents began to de-
cline when department managers, who had to re-
port the infractions, began backing up repeated
environmental staff requests for action and when
each member of the boiler staff was given control
over a speci�c section of equipment. As each
came to know a section well, they developed the
skill to maintain proper conditions. Later, exter-
nal recognition, such as the governor’s Pollution
Prevention Award and IP’s Corporate Award for
Environmental Excellence, maintained em-
ployee motivation. Further improving perfor-
mance and communications, staff also recognized
that a failure to control one section contributes
to control failures elsewhere. Recovery boiler
opacity infractions dropped to zero.
Changing Relations with the Maine DEP
Although the new management team began
to get the mill under control, exceedences of li-
cense limits of various types still occurred in
1992, and the mill continued to pay �nes to set-
tle earlier violations. But examples surfaced that

demonstrated changing attitudes. One was to re-
move the lawyers previously standing between
the mill and the Maine DEP: Saviello had ini-
tially taken lawyers with him to meetings at the
Maine DEP, and the lawyers took charge of the
meetings, a practice to which DEP personnel re-
acted with animosity. In an attempt to better re-
lations, Stowell told Saviello in 1991 to stop in-
viting lawyers. This approach initially felt very
different to Saviello, but he became increasingly
comfortable with direct discussions with DEP
personnel. And, as seen below, mill relations
with DEP did improve.
Evolving Approaches
to Managing Environmental
Performance
Beginning at the End of the Pipe
To make wood pulp suitable for paper, mills
must remove the lignin that binds cellulose in
the �bers (Appendix 1). Lignin and other or-
ganic materials that end up in wastewater con-
tribute to biochemical oxygen demand (BOD), a
major pollutant produced by pulp and paper mills
(Springer 2000b) . Because of water-quality con-
siderations, the Maine DEP Water Bureau issued
a wastewater license in 1991 to the mill limiting
BOD releases to the river to 10,500 lb/day, a level
more than 2 times as restrictive as federal re-
quirements. Although mill personnel did not be-
lieve that this limit could be achieved, they —
again bypassing lawyers’ advice —accepted the
invitation of Stephen Groves, then director of

the DEP’s Water Bureau, to work collaboratively
to solve the problem. The mill implemented a
number of Maine DEP suggestions, including the
installation of dozens of aerators in the waste-
water treatment lagoon to achieve higher dis-
solved oxygen levels and promote greater micro-
bial degradation of the BOD. The collaboration
proved successful, as the mill quickly reduced
BOD discharges to meet the permit and achieved
a consistent level of approximately 4,000 lb/day
or lower through 2001 (�gure 1).
Moving up the Pipe
Mill and Maine DEP engineers formed an
“environmental quality team” in 1992 to identify
pollution prevention opportunitie s (Springer
2000d). One concern of the team was the impact
of pulp-bleaching processes on emissions. Mill
staff introduced two process changes in response
to team recommendations that bleaching be
minimized by reducing lignin in pulp (Springer
2000b). The mill combined extended deligni�-
cation, which involved cooking wood chips for
a longer period at lower temperatures, while
maintaining pulp yield and quality (McDonough
2000), and oxygen deligni�cation, treating pulp
with high-pressure oxygen (Sjostrom 1981), to
reduce lignin content and required bleaching.
The result was lower quantities of adsorbable or-
ganic halide (AOX) (chlorine-containing by-
products formed during bleaching) compounds,
dioxin, and furan in the mill’s ef�uent and low-
ered chloroform emissions to air and water. Di-
oxin, furan, and chloroform emissions were ul-
timately eliminated follow ing the mill’s

conversion to 100% chlorine dioxide bleaching2
in 1995 (see Table 1).
Additional pollution prevention projects re-
sulting in process changes continued in 2002 in
conjunction with the mill’s two U.S. EPA XL
projects.3 Although several pollution prevention
efforts, such as extended deligni�cation/oxygen
deligni�cation, were not unique to this mill, the
Mill 111
Figure 1 Biochemical oxygen demand (BOD) loss to mill
ef�uent (pounds per day).
XL projects resulted in the development of new
technological approaches to pollution preven-
tion. Indeed, one criterion that U.S. EPA uses to
judge XL projects is the likelihood of developing
technology that is transferable to similar facili-
ties. The use of a computer model of the mill’s
waste-fuel incinerator developed under the di-
rection of Thomas Saviello prompted process
changes that to date have reduced particulate
emissions by 50%. In the second XL project, the
mill began modifying production processes to re-
duce spent liquor in wastewater, which can affect
ef�uent color and chemical oxygen demand.4
Both XL projects involved a stakeholder team as

well as a technical team, the latter including paid
consultants.
Addressing Pollution through Supply
Chain Management
Tests in April 1998 by the mill revealed that
the concentration of mercury in upstream river
water was 6.5 parts per trillion (ppt), whereas the
ef�uent concentration was 19.2 ppt. Investiga-
tion revealed two contaminated feedstocks. One
was alkali purchased from a chloralkali plant us-
ing a mercury process. The mill found a supplier
that did not use the mercury cell process. The
second supply chain issue was the mill’s purchase
of sulfuric acid from a Canadian lead smelter that
converted its captured sulfur dioxide emissions to
sulfuric acid, with mercury as an unintended con-
taminant. The mill switched suppliers to pur-
chase uncontaminated sulfuric acid from a nickel
smelter. By August 2000, without end-of-pipe
control, ef�uent mercury was reduced to 3.4 ppt,
equivalent to the river background of 3.9 ppt
(�gure 2).
Toxics Use Reduction
Although it requires only reporting of toxics
use and hazardous waste generation, the Maine
Toxics Use Reduction Act was the initial driver
in identifying opportunitie s to eliminate or at
least reduce the use of hazardous substances in a
variety of operations. The use of elemental chlo-
rine, transported by tank car into the mill over
most of its operating life, posed an ongoing dan-

ger to workers and the community if a leak de-
veloped. By 1995, the mill completed a switch to
100% chlorine dioxide bleaching. Chlorine di-
oxide also poses a risk, but because it is generated
on-site as needed, the risk associated with trans-
Table 1 Mill pollution prevention example
Year AOX and dioxin discharges to ef�uent
Dioxin (2,3,7,8-TCDD) 1988 88 pg/L
1996 Nondetecta and remains sob
Furan (2,3,7,8-TCDF) 1988 420 pg/L
1997 Nondetect and remains sob
AOXc 1994 1.44 lb/ton bleached pulp/day
2000 0.52 lb/ton bleached pulp/day
a Nondetect is less than 10 pg/L.
b Neither dioxin (2,3,7,8-TCDD) nor furan (2,3,7,8-TCDF) are
detected in the bleach-plant ef�uent where, if
present at all, they would occur at the highest concentration.
c Adsorbabl e organic halide, chlorine-containing by-product s
formed during bleaching.
Figure 2 Mercury in mill ef�uent (nanograms per liter).
portation and the use of large quantities is sub-
stantially lessened. Another product, Nalco

Chemical Co. TRI-ACT 1804, used to inhibit
corrosion in recovery boiler tubes, contained the
hazardous chemical cyclohexylamine. This was
replaced with a proprietary Nalco product (TRI-
ACT 1826) not containing any hazardous con-
stituents.
The mill also replaced two hazardous products
that had provided nitrogen and phosphorus nu-
trients to promote the growth of microorganisms
that degrade organic materials in mill wastewater
(Springer and Maxham 2000). Anhydrous am-
monia as a source of nitrogen was eliminated be-
cause of Jay of�cials’ concerns about shipping
safety and U.S. EPA’s requirement for risk man-
agement planning (U.S. EPA 2001). It was re-
placed with the much less hazardous nitrogen-
containing chemical, urea. Use of phosphoric
acid as a source of phosphorus declined 57% be-
tween 1993 and 2000 as a mill team progressively
reduced the amount added to the lowest level
that would still suppor t optimal microbial
growth. Additional reductions occurred when
the mill switched to a single product that blends
urea with ammonium polyphosphate, eliminat-
ing the phosphoric acid use in this application.
Mill efforts to reduce hazardous waste gener-
ation included greater attention to preventing
caustic liquor spills, the substitution of barium

Mill 113
Figure 3 Hazardous waste disposal (thousands of pounds per
year).
chloride with a nonhazardous chemical for use in
a titration analysis (eliminating 7,000 lb of haz-
ardous waste each year), screening of paint prod-
ucts before purchase to avoid hazardous constit-
uents, and purchasing low-mercury �uorescent
bulbs to minimize hazardous waste generation
when relamping. These and other actions re-
duced hazardous waste generation from a high of
60,000 lb in 1990 to a low of 3,260 lb in 2000
(see �gure 3).
Changing Approaches to Wastes
Pulp and paper mills generate a large variety
and quantity of wastes (Springer 2000e), much
of which may be land�lled, including tree bark,
�ume grit (dirt and contaminants carried with
logs into the mill), sludge from wastewater treat-
ment, green-liquor dregs and lime mud,5 wood
knots and screenings,6 mill garbage, and some
waste metal and paper. In 1988, the mill operated
an on-site land�ll that averaged 1,643 cubic
yards (yd3) of new waste a day and was close to
capacity. Intensive efforts at recycling, pollution
prevention, incineration, and bene�cial reuse re-
sulted in average daily land�ll rates in 2001 of
150 yd3, a 91% reduction (�gure 4). Mill pro-
grams included the following:
� Recycling wood, metals, and paper

� Compacting nonrecyclable paper into
burnable pellets
� Improving limekiln operations to allow �r-
ing of all lime mud produced
� Selling �ume grit to a contractor that pro-
cessed it into landscape material (similar to
peat or perlite used for potting media and
erosion control)
� Burning bark and sludge and incorporating
the ash into AshCrete, a product devel-
oped at the mill (see below)
� Incorporating green-liquor dregs into
AshCrete
Sludge and Ash
The Androscoggin Mill produces about 10%
of the 1 million tons of sludge produced each year
from wastewater treatment plants at Maine’s
pulp and paper mills. Sludge use on IP timber-
lands as a soil supplement was discontinued be-
cause of the high cost of transporting sludge that
was only 40% solids by weight (Springer 2000e).
Adjusting the waste-fuel incinerator allowed all
sludge except for that used as land�ll cover
(where it replaced virgin clay) to be burned. A
portion of the mill ash is shipped to a contractor

Figure 4 Solid waste land�lled (cubic yards per day).
in Unity, Maine, which uses it in composting
municipal sewage sludge for farm application.7
Most ash from sludge and bark incineration,
however, is incorporated into a product called
AshCrete, developed in 199 8 by Stephen
Groves.8 A contractor makes AshCrete on-site
from ash, green-liquor dregs, and other proprie-
tary by-products. For the next 15 years, all of the
AshCrete is expected to be used to reduce the
size of the wastewater lagoon, which had been
designed and built when the mill produced sub-
stantially more BOD. Approved by the Maine
DEP, AshCrete use negates the need to purchase
gravel to �ll the lagoon. The DEP also approved
the use of AshCrete in closing the mill land�ll
and as a subbase for a concrete pad. Two southern
IP facilities also now produce AshCrete with a
somewhat different formula. They use it as a
berm and dike material and, similar to the An-
droscoggin Mill, to recon�gure or close waste-
water treatment lagoons.
Facilities Colocated with the
Androscoggin Mill
Much like Kalundborg, Denmark (Ehrenfeld
and Gertler 1997), a small “industrial ecosystem”
has evolved slowly around the mill, with several
companies locating facilities at the site to take

advantage of by-products and market opportu-
nities.
� Specialty Minerals, Inc. produces precipi-
tated calcium carbonate (PCC) by reacting
carbon dioxide with calcium oxide in a
proprietary process. Specialty Minerals
needed a source of carbon dioxide and an
outlet for PCC and set up operations at the
mill in 1997 using carbon dioxide emis-
sions from a limekiln.9 In return, the mill
buys PCC at an attractive price and elim-
inates transportation costs.
� A contractor operates an on-site facility
owned by the mill to process the ash pro-
duced from burning mill sludge and bark
into AshCrete.
� Androscoggin Energy is a natural-gas burn-
ing facility, generating electricity with
high-temperature steam (sold off-site) and
selling low-temperature steam to meet a
portion of the mill’s needs. The mill boil-
ers, which had burned number-six fuel oil
containing 1.8% sulfur, went into standby
mode, resulting in lower sulfur dioxide, ni-
trogen oxides, particulate, and carbon di-
oxide emissions. (Mill recovery boilers,
Mill 115

which burn spent pulping liquor as part of
the chemical and energy recovery opera-
tions, along with the waste-fuel incinera-
tor, furnish the rest of the mill’s steam
needs, as well as providing electricity.)
The Public Advisory
Committee
Members, Mission, and Early Challenges
In developing a PAC in 1992, IP followed
through on a recommendation of the PCEQ. The
PAC originally de�ned its mission as to “ . . . help
identify environmental issues the Androscoggin
Mill must address, and proactively assist in
choosing the options. This will be accomplished
by developing trust and respect for each other.”
By 2000, members expanded that mission to
“. . . act as a public board to identify and respond
to the environmental, social, economic, and
community issues that the Androscoggin Mill
must address, and proactively assist in choosing
sustainable options.” According to David Critch-
�eld, the mill in 1992 sought members who
“would not pull punches, but who also had a
strong constructive side.” Members included en-
vironmentalists, forestry and business experts, a
mill customer, and a member of the mill’s hourly
staff (table 2). Initial PAC meetings included the
mill environmental manager, its environmental
superintendent, and frequently the mill manager.
(The current mill manager, Michael Craft, has
attended all recent meetings.) Subsequently, the
mill engineer responsible for developing energy

conservation and ef�ciency measures became a
regular participant. The PAC also hears from
other mill personnel as necessary.
A major challenge to the PAC in early years
was to understand mill operations, environmen-
tal matters, and issues associated with local, state,
and federal agencies. To assist in the challenge
of dealing with a substantial amount of infor-
mation, the PAC developed a report card with
data on ef�uent quality, solid waste generation,
energy and water use, and process- and energy-
related air emissions (including carbon dioxide).
The report card also provided comparisons with
the previous month, license limits, and annual
goals and provided a focus and data to help moni-
tor performance. The mill used the report card
as a way to monitor itself by putting the goals
and progress in constant view. The mill also used
it on occasion with customers to demonstrate the
mill’s environmental commitment.
An ongoing discussion at early meetings cen-
tered on the town of Jay and its environmental
regulations. PAC members strongly advised the
mill to accommodate Jay’s concerns, to make a
strong effort to establish a working relationship
with Jay, to be candid in all dealings with the
town, and to not see it as an adversary. Over the
years, the relationship with Jay became positive.
Jay citizens served on the PAC and Project XL
stakeholder teams. They also served on the col-
laborative stakeholder team formed for relicen-
sing the mill’s dams, and the team developing the
mill’s federally mandated risk management plan.

The Jay ordinance was a driver of early change,
but, over time, the impetus for change increas-
ingly came from the mill. The ordinance re-
mained in existence, but the PAC did not deal
with compliance issues after the �rst few years.
Moving beyond Compliance
Over time, PAC members began to think be-
yond issues of compliance to sustainability issues.
In particular, because a pulp and paper mill is not
viable without a sustainable wood supply, PAC
members began considering mill wood supply.
Where did it come from, and what was the con-
dition of the forests supplying it? Did loggers
practice responsible forestry? Because IP forests
are administered separately from the mill, the
PAC invited forestry personnel to its meetings,
and members also visited some of IP’s forests. IP
follows the Sustainable Forestry Initiative (SFI)
standards of the American Forest and Paper As-
sociation10 (AFPA 1994) and was the �rst Amer-
ican forest products company to earn ISO 14001
certi�cation of its forest management system.11
IP forestland supplies barely 20% of the An-
droscoggin Mill’s �ber requirements, however,
because most wood goes to higher-value uses, es-
pecially lumber. The other 80% of the wood
comes from independent loggers, who must com-
plete SFI training and agree to abide by SFI stan-
dards. Some loggers have chosen not to provide
wood to the mill because of these requirements.
In other cases, the mill has refused to buy wood

Table 2 Public advisory committee members, 2002
Member Af�liation and location
Deborah Burd* Western Mountains Alliance, Farmington
(regional sustainable community
development)
Harold Burnett Two Trees Forestry, Winthrop (forestry
consulting)
Richard Cormier Franklin Savings Bank, Jay
Carla Dickstein Coastal Enterprises, Inc., Wiscasset
(community development)
William Harlow Androscoggin Mill (hourly employee) and Jay
Planning Board, Jay
Marquita Hill* Department of Chemical Engineering, University
of Maine, Orono
Donald Hopkins* Hearst Corporation, New York City (mill
customer)
David Kraske† Retired University of Maine professor and
paper-company executive,
Canton, Maine
Patrick Flood IP, regional forestry operations
Daniel Sosland* Environment Northeast and environmental
lawyer, Rockport
* Members who served since the inception of the PAC in 1992.
Sosland served as chair since shortly after the
inception of the PAC.

† Kraske did not work for IP.
from loggers they believe are in violation of SFI
standards. Although PAC members were basi-
cally satis�ed that IP was genuine in its efforts to
maintain sustainable forests, they continued to
follow the topic closely. Among the PAC mem-
bers was an independent forester; and to further
assist the PAC in its efforts, the manager of IP’s
regional forestry operations became a PAC mem-
ber in 2001.
Valuing the PAC
IP’s policy is to have a PAC at each of its
integrated pulp and paper mills. Except for the
Androscoggin PAC, other committees are com-
munity based. Committees provide a means for
IP to promote an understanding of how the com-
pany operates, in the belief that knowledgeable
communities will be more supportive. Environ-
mental staff at the Androscoggin Mill viewed the
PAC as helpful in the internal effort to under-
stand the mill’s environmental performance and
to identify issues that might otherwise be missed.
PAC members brought a different set of perspec-
tives to these issues, avoiding “group think” that
can blind internal staff to critical problems. The
PAC was also a means of bringing external ac-
countability to the mill, which can reestablish a
“franchise to operate” when it is threatened, as
it was in Jay.
PAC members found it rewarding to have
contributed to the positive changes at the An-

droscoggin Mill. Other values important to
members were opportunities to delve into mill
issues and obtain a greater understanding of the
complexities often involved, seeing how the mill
operates, how management works, and sharing
perspectives and information among themselves
and between the PAC and mill administration.
At another level, some organizations, including
the University of Maine, see participation in an
activity such as a PAC as part of its public service
mission.
At the Androscoggin Mill, the PAC’s role be-
gan to change in the late 1990s. Providing feed-
back and assistance on compliance and pollution
prevention was no longer central. The mill was
well run and well regarded. PAC members began
to turn to the longer-term issues and quickly
came up with a dif�cult set of questions: How do
PAC members de�ne sustainability and effec-
tively discuss it with the mill? How does the PAC
tell the mill that it should be setting higher goals,
requiring a greater stretch to reach them? What
is an appropriate level of natural resources for the
mill to use and how can it get there? And how
can the mill get corporate authorization for more
environmental capital investments?
In April 2002, the PAC was disbanded. Mill
management believed that it had ful�lled its mis-
sion. Moreover, it felt a need for a local com-
munity advisory committee whose major purpose

Mill 117
would be to promote good communication with
the Jay community. It expects to start a com-
munity advisory committee in the summer of
2002. To minimize bias in member selection, the
mill will hire a contractor to recommend mem-
bers. PAC members supported management’s de-
cision while also voicing regret. Members be-
lieved they served as that “extra pressure”
pushing for positive change. But they also be-
lieved that the mill has come so far that, espe-
cially with its image of environmental leader, it
was unlikely to backslide. Additionally, there is
a continuity of environmental staff, and Michael
Craft (mill manager since 1999) comes from an
environmental management background. More-
over, XL projects are still ongoing, and two PAC
members are participants in those committees.
The mill is also part of the U.S. EPA Star Track
Program (U.S. EPA Star Track Program 1997).
Discussion and Conclusions
Although crisis stimulated change at the An-
droscoggin Mill, the outcome could have been
less positive. Critical to the successful outcome
was the IP corporate decision to �nd capable in-
dividuals and charge them with turning the mill
around. The early leadership in this effort fo-
cused on making the Androscoggin Mill the best
facility at IP. They were willing to risk giving the
Maine DEP greater access to the mill, to form a
PAC, and then to be responsive to both. External

demands on the mill, especially in the early
1990s, played a critical role as well. Jay’s envi-
ronmental ordinance, highly unusual for a small
community, was a signi�cant driver for change,
along with the town’s unremitting pressure on
the mill.
The mill, after its success in reaching out to
the DEP, to the PAC, and increasingly to Jay,
continued its outreach by forming collaborative
projects, as when it relicensed its dams on the An-
droscoggin River in the mid-1990s and the U.S.
EPA XL projects, which continue into the pres-
ent. Each had membership from inside and out-
side the mill, from Jay, and from stakeholders
outside the immediate community. Although
stakeholders had different perspectives, all held
a common interest in the success of the project
with which they were involved. Thus, many peo-
ple came to care about mill success: employees,
PAC members, Jay citizens, the Maine DEP, U.S.
EPA personnel, and members of participating
nonpro�t organizations. Rather than simply �nd-
ing fault with the mill, these collaborative efforts
brought a sense of collective investment to �nd-
ing solutions.
The PAC’s contribution to the mill’s environ-
mental successes was substantial if not public. Its
access to current environmental data provided a
means to constantly challenge the mill to im-
prove. Relatively little turnover in the PAC
meant that members were in a position to press
mill management if change on a particular issue
seemed too slow. Conversely, the turnover in mill

managers (four in the PAC’s �rst decade) created
some confusion and delays. The tenure of envi-
ronmental staff, however, helped smooth transi-
tions and provide some constancy in the inter-
face between the PAC and the mill.
The environmental improvements at the An-
droscoggin Mill did not emerge from a formal
system, but important elements of a systems ap-
proach were in place. The mill bene�ted from a
combination of talented individuals with vision,
support, and pressure from within and outside the
mill and hard consistent effort. This case dem-
onstrates the value of management developing
long-term goals and a framework for change,
built on high standards and a desire to gain the
trust of regulators and the community that lead
to environmental excellence.
Acknowledgments
We dedicate this article to the memory of Pe-
ter Bernard, public advisory committee member
1993 – 2000, Androscoggin Mill employee, and
Jay Planning Board member. We are grateful to
Mill Manager Michael Craft and to public ad-
visory committee members for their patience
with the development of the chronicle from
which this article is derived and the slow evo-
lution of this article. We thank Androscoggin
engineers John Cronin and Vickie Gammon for
information and assistance in plotting data and
David Critch�eld (now CEO of EMSource, Port-
land, Maine, USA) for assistance in reconstruct-
ing events. Marquita Hill thanks the University
of Maine, Orono, for travel assistance and John

Hassler and Adriaan Van Heiningen for valuable
advice. She thanks her fellow authors for their
decade of un�agging effort: for persistence, cour-
age, and responsiveness and for zest, creativity,
and surety that any problem can be solved.
Notes
1. Opacity is an optical measure of particulat e emis-
sions.
2. The mill has not calculate d chloroform emissions
since 1995, when it completed the switch to
100% chlorine dioxide bleaching. The National
Council for Air and Stream Improvement of the
pulp and paper industry indicates that chloroform
is not formed in 100% chlorine dioxide bleach-
ing, a conclusion accepte d by both the U.S. EPA
and the Maine DEP. The procedure for calculat -
ing chloroform emissions is given in the NCASI
Handbook of Chemical Speci�c Informatio n for
SARA Section 313 Form R Reporting Chemical-
Speci�c Information for Chloroform: Section 3.1
Manufacture the Toxic Chemical. This reference
is updated yearly.
3. U.S. EPA XL projects are designed to enhance
environmental protectio n while introducing reg-
ulatory �exibility and innovative environmental

approaches at exemplary facilitie s (U.S. EPA XL
Projects 2000).
4. Chemical oxygen demand is the measure of ox-
ygen required to oxidize organic and inorganic
compounds in ef�uent. It can adversel y affect or-
ganisms in receiving waters.
5. Green-liquo r dregs and lime mud result from
chemical recover y operations : After burning
black liquor in the recover y boiler, a smelt that
contains sodium carbonate results. When dis-
solved, the smelt forms green liquor, leaving be-
hind green-liquo r dregs, which are removed.
Green liquor is reacted with lime (calcium oxide)
to regenerate white liquor and precipitat e lime
mud (calcium carbonate) . The mud is �red in the
limekilns to regenerate lime. After adjustin g its
chemical composition , the white liquor is again
used in cooking operation s (Smook 1992, 149 –
153).
6. Wood knots are overthick chips or other irregula r
wood pieces that are dif�cult to digest during
cooking. Screenings are particles of wood that
may contaminate pulp and paper if not screened
out before cooking operations .
7. The Maine DEP sets standards and requires test-
ing for heavy metals and other hazardous contam-
inants in ash and sludge as a condition of ap-
p r o v a l f o r b e n e �c i a l r e u s e , s u c h a s l a n d
application . It also sets limits on the amount of
heavy metals and other contaminants that can be
applied to a parcel of land. Code of Maine Reg-

ulations, Chapter 419, “Agronomic Utilization of
Residuals” (effectiv e December 19, 1999).
8. Groves left the Maine DEP to work at the mill
in 1994, becoming environmental manager in
1995.
9. Limekilns recover calcium oxide (lime) from cal-
cium carbonate by driving off carbon dioxide.
Used as a coater and �ller, PCC use is not unique
to the Androscoggin Mill. Some coated papers
contain 30% PCC by weight. SMI sells about
60% of its output to the mill and the rest to other
paper mills.
10. There is disagreemen t over competing standards
in forest management. The primary alternativ e to
the American Forestry and Paper Association’s
SFI is the Forest Stewardship Council’s standard.
See the compariso n prepare d by the Meridian In-
stitute (2001) for further details.
11. The Androscoggi n Mill also has an environmen-
tal management system that is audited and ap-
proved by the Maine DEP and the U.S. EPA.
References
AFPA (American Forest and Paper Association) .
1994. Sustainable forestry initiative program.
www.afandpa.org /iinfo/iinfo.html. Accessed May
2002.
Churchman, C. W. 1968. The systems approach. New
York: Dell.

Ehrenfeld, J. R. and N. Gertler. 1997. Industrial ecol-
ogy in practice: The evolution of interdepen -
dence at Kalundborg. Journal of Industrial Ecology
1(1): 67 – 79.
McDonough, T. 2000. Pulping and bleaching technol-
ogies for improved environmental performance .
In Industria l environmenta l control, pulp and paper
industry. Third edition. Edited by A. Springer.
Atlanta: TAPPI Press.
Meridian Institute. 2001. Comparative analysis of the
Forest Stewardshi p Council© and Sustainabl e For-
estry Initiative®certi�cation programs. Washington,
DC: Meridian Institute. http://madison.merid
.org/meridian/home.nsf/projectarea . Accessed Au-
gust 2002.
President’s Commission on Environmental Quality
(PCEQ). 1990. Executive Order 12737, Decem-
ber 12, 1990, 55 F.R. 51681.
Servos, M. R., K. R. Munkittrick, J. H. Carey, and G. J.
Van Derkraak, eds. 1996. Environmental fate and
Mill 119
effects of pulp and paper mill ef�uents. Delray
Beach, FL: St. Lucie Press.
Sjostrom, E. 1981. Wood chemistry. New York: Aca-

demic Press.
Smook, G. A. 1992. Handbook for pulp and paper tech-
nologists. Second edition. Bellingham, WA: An-
gus Wilde Publications.
Springer, A. 2000a. Industria l environmenta l control,
pulp and paper industry. Third edition. Edited by
A. Springer. Atlanta: TAPPI Press.
Springer, A. 2000b. Impact of water pollutants on the
pulp and paper industry. In Industria l environmen -
tal control, pulp and paper industry. Third edition.
Edited by A. Springer. Atlanta: TAPPI Press.
Springer, A. 2000c. Air pollution : A problem without
boundaries . In Industrial environmenta l control,
pulp and paper industry. Third edition. Edited by
A. Springer. Atlanta: TAPPI Press.
Springer, A. 2000d. Pollution prevention. In Industrial
environmenta l control, pulp and paper industry.
Third edition. Edited by A. Springer. Atlanta:
TAPPI Press.
Springer, A. 2000e. Management of residuals. In In-
dustrial environmenta l control, pulp and paper in-
dustry. Third edition. Edited by A. Springer. At-
lanta: TAPPI Press.
Springer, A. and J. Maxham. 2000. Secondary aerobic
biologica l treatment. In Industrial environmenta l
control, pulp and paper industry. Third edition. Ed-
ited by A. Springer. Atlanta: TAPPI Press.
U.S. Environmental Protection Agency (EPA) Chem-

ical Accident Prevention and Risk Management
Programs. 2001. www.epa.gov/swercepp/acc-pre
.html. Accessed August 2002.
U.S. EPA Star Track Program. 1997. StarTrack Envi-
ronmental Performance Report. www.epa.gov/
region01/steward/strack/eprip.pdf . Accessed May
2002.
U.S. EPA XL Project. 2002. What is Project XL?
www.epa.gov/projectxl /�le2htm. Accessed Au-
gust 2002.
Appendix 1: Kraft Pulping and
Pollution Associated with It
Background
Wood, the source of the �ber used to make
most paper, contains two major components.
One is carbohydrate (cellulose and hemicellu-
lose). The other is lignin, which binds tightly to
the carbohydrate and imparts strength to the
wood. In chemical pulping, �ber (cellulose and
hemicellulose) is separated from lignin by cook-
ing wood chips with inorganic chemicals at high
temperature and pressure.
The paper indust ry typically used two
chemical-pulping technologies in the 1950s and
1960s, sul�te and kraft. Sul�te mills of that time
discharged spent cooking liquor into rivers to
avoid the cost of recovering cooking chemicals.
In contrast, the kraft process incorporates an eco-
nomic recovery of 97% to 98% of cooking chem-

icals (Smook 1992). The result was the closure
of a growing number of sul�te mills in the 1950s
as kraft mills came on line. Kraft mills, although
more economical to operate, produce a dark pulp
(the color of a typical grocery bag), which re-
quires stronger bleaching. This led to the use of
elemental chlorine as a bleaching agent.
Chemical and Energy Recovery
The spent pulping liquor that results from
kraft cooking contains about half the wood’s or-
ganic substance, especially lignin-containing
chemicals, as well as the degradation products of
the inorganic cooking chemicals. In the kraft
process, the inorganic chemicals are recovered in
a closed-loop process that relies on energy gen-
erated by the burning of lignin-rich liquor: The
spent liquor initially contains about 15% solids.
This is concentrated in evaporators to about 75%
solids and then burned in recovery boilers to gen-
erate steam. The steam is used to evaporate and
concentrate additional spent liquor, continuing
the cycle and also providing electricity and steam
to the paper mill. The smelt left after burning
the liquor contains the inorganic chemicals.
These are recovered and converted back to cook-
ing chemicals that can be used again.
Making Paper
The cooked chips (separated from the bulk of
the liquor) are disintegrated into their compo-
nent �bers, the pulp. The pulp is washed and
then often undergoes further treatment such as
oxygen deligni�cation. It is bleached and washed

again. Finally, the pulp is diluted and fed along
with additives into the machines producing pa-
per products.
Pollutants Produced by Kraft
Pulping
Wastewater treatment plants generally can-
not be 100% effective, and some pollution still
reaches rivers. Pollutants from kraft mills include
BOD, pigments (primarily from the dark color of
lignin), and total suspended solids. The use of
elemental chlorine to bleach pulp results in
AOXs, including trace amounts of dioxins, in
mill ef�uents. Chlorine dioxide bleaching signi�-
cantly lowers AOX generation and virtually
eliminates dioxi n formation (McDonough
2000). Kraft mills produce air pollutants that are
typical of many industrial facilities. Those most
annoying to nearby communities are malodorous
sulfur-containing chemicals, such as methyl sul-
�de, that result from the cooking process. Strin-
gent emissions controls are required to avoid re-
leases (Springer 2000c).
About the Authors
Marquita Hill is cooperatin g professo r in the De-
partment of Chemical Engineering at the University

of Maine in Orono, Maine, USA, and was a member
of Androscoggi n Mill’s public advisor y committee
from 1992 until 2002. Thomas Saviello was the su-
perintenden t of environmental services , 1991 – 2001,
and has been manager of environmental health and
safety since April 2001 in the environmental depart-
ment at International Paper’s Androscoggi n Mill, in
Jay, Maine, USA. Stephen Groves was the Andro-
scoggin Mill’s manager of environmental health and
safety, 1995 –2001, and has been International Paper’s
corporat e senior program manager for environmental
initiatives since April 2001.
R E S E A R C H A N D A N A LY S I S
Greenhouse Gas Emissions Reduction
Opportunities for Concrete Pavements
Nicholas Santero, Alexander Loijos, and John Ochsendorf
Summary
Concrete pavements are a vital part of the transportation
infrastructure, comprising nearly
25% of the interstate network in the United States. With
transportation authorities and
industry organizations increasingly seeking out methods to
reduce their carbon footprint,
there is a need to identify and quantitatively evaluate the
greenhouse gas (GHG) emis-
sion reduction opportunities that exist in the concrete pavement
life cycle. A select few

of these opportunities are explored in this article in order to
represent possible reduc-
tion approaches and their associated cost-effectiveness:
reducing embodied emissions by
increasing fly ash content and by avoiding overdesign;
increasing albedo by using white
aggregates; increasing carbonation by temporarily stockpiling
recycled concrete aggregates;
and reducing vehicle fuel consumption by adding an extra
rehabilitation. These reduction
strategies are evaluated for interstate, arterial, collector, and
local road designs under urban
and rural scenarios. The results indicate that significant GHG
emission reductions are pos-
sible, with over half of the scenarios resulting in 10%
reductions, compared to unimproved
baseline designs. Given the right conditions, each scenario has
the potential to reduce GHG
emissions at costs comparable to the current price of carbon.
Keywords:
carbon
cost effectiveness
global warming potential (GWP)
industrial ecology
life cycle assessment (LCA)
life cycle costing (LCC)
Supporting information is available
on the JIE Web site
Introduction
The construction, operation, and maintenance of the U.S.
roadway system are responsible for substantial energy and re-

source consumption. Although the cumulative environmental
impact of the road network is unknown, there is reason to
believe that significant greenhouse gases (GHGs) are released
during the construction and operation of pavements. Accord-
ing the United States Geological Survey (USGS), 460 million
metric tons of crushed aggregate alone go into the construc-
tion, rehabilitation, and maintenance of the U.S. pavement
network (USGS 2011) in order to provide service for over
5 trillion vehicle-kilometers per year (USDOT 2011).1 Pas-
senger and freight movement on roadways accounts for 83%
of carbon dioxide (CO2) emissions from the transportation
sector and 27% of total CO2 emissions in the United States
(EPA 2009). Although the bulk of the these road transport
Address correspondence to: Nicholas Santero, PE International,
71 Stevenson Street, Suite 400, San Francisco, CA 94105, USA.
Email: [email protected]
© 2013 by Yale University
DOI: 10.1111/jiec.12053 Editor managing review: Ester van der
Voet
Volume 17, Number 6
emissions should not be attributed to the pavements them-
selves, the materials and serviceability levels required of this
infrastructure system give rise to a notable GHG emission
source.
Reducing the GHG emissions of pavements requires a com-
plete understanding of how it impacts the natural environment.
Like any other product or service, pavements generate GHG
emissions throughout their service life, beginning with raw ma-
terials extraction and manufacturing, continuing through con-
struction, operation, and maintenance, and, finally, ending with
waste management and recycling. Life cycle assessment (LCA)
is designed to capture each of these phases in order to create

a portrayal of the sources and magnitude of emissions over the
life cycle. This approach not only quantifies the current foot-
print, but also is useful in identifying and quantifying potential
opportunities to reduce those impacts.
www.wileyonlinelibrary.com/journal/jie Journal of Industrial
Ecology 859
mailto:[email protected]
This article focuses on GHG reduction opportunities for
concrete pavements, as measured by their global warming po-
tential (GWP). Emissions are quantified for a select number
of strategies, then evaluated for their cost-effectiveness using
life cycle cost analysis (LCCA) principles. The strategies are
developed and applied to representative designs for each Fed-
eral Highway Administration (FHWA) roadway classification
in the United States, spanning from rural local roads to urban
interstates. The results demonstrate a set of opportunities and
economic impacts that Departments of Transportation (DOTs)
and other stakeholders can use to decrease the GHG emissions
of their roadway networks.
Opportunities for Greenhouse Gas Emission
Reductions
Concrete pavements offer an abundance of opportunities
for GHG reductions. Four broad approaches are explored in
this article: reducing embodied emissions; increasing albedo;
increasing carbonation; and reducing vehicle fuel consumption.
Most strategies can be grouped into one of these approaches,
making this a convenient organizational method in which to
characterize potential improvement opportunities.

Embodied emissions are those released during the manufac-
turing and construction of paving materials. Essentially, these
are the emissions embodied in the pavement when it begins
its service life, as well as those materials that are added during
maintenance operations. These emissions can be reduced by us-
ing fewer natural resources, substituting less emission-intensive
materials, or increasing production efficiency.
Albedo measures the fraction of incoming solar radiation
that is reflected by the pavement surface. Increasing albedo
reduces the climate impacts from both the urban heat island
effect and direct radiative forcing. Albedo also correlates with
lighting demand, thus affecting the electricity needed to illu-
minate a roadway. Concrete naturally enjoys a relatively high
albedo, but improvements can be made to the concrete mix
that increase the albedo even further, such as the use of white
aggregates, white cement, and slag.
Carbonation is a chemical process by which CO2 is natu-
rally sequestered in the concrete. Carbonation for an in-situ
concrete pavement is usually minimal, penetrating only a few
centimeters into the pavement over its service life and thus se-
questering only a fraction of the CO2 release during calcination.
Following Fick’s law of diffusion, carbonation is expedited with
increases in the surface-area/volume ratio—something that oc-
curs when concrete pavements are crushed at the end of their
service life. Crushed concrete is typically recycled as base, fill,
or concrete aggregates, which all present an opportunity for
carbon sequestration.
Vehicle fuel consumption is affected by the choice in pave-
ment design, maintenance, and materials. Most vehicle fuel
consumption is unaffected by pavement-related issues, but
GHG emissions from increased vehicle fuel consumption result-
ing from pavement-vehicle interaction (e.g., increased rough-

ness or reduced stiffness) and traffic delay (caused by pavement
construction activities) can be significant and should be allo-
cated to the pavement life cycle (Santero and Horvath 2009).
With upward of 100,000 vehicles per day traveling over certain
structures, pavement characteristics that offer even slight fuel
economy improvements can significantly decrease the GHG
emissions associated with the pavement life cycle.
The Role of Economics
Economics provide the critical link that helps implement
environmental impact reduction strategies into DOT decision-
making frameworks. Although most DOTs and other stakehold-
ers are interested in reducing GHG emissions, the primary goal
remains to provide maximum pavement performance within
budgetary constraints. Reaching environmental targets neces-
sarily becomes a secondary priority. In order to effectively in-
tegrate GHG reduction strategies into DOT decision making,
it is essential to appreciate that reductions must be achieved at
minimal costs.
LCCA offers a method of analyzing the economic impacts
of pavements and is often used at DOTs for deciding between
design alternatives. LCCA can also be used to determine the
cost-effectiveness of environmental improvement strategies—
the application used in this research. Coupling LCCA with
LCA provides a holistic view of both the economic and envi-
ronmental impacts of a given strategy, thus providing decision
makers with a more complete set of information.
Methodology
The baseline designs and reduction scenarios are evaluated
using a pavement LCA model developed at the Massachusetts
Institute of Technology (MIT) Concrete Sustainability Hub.

The model captures impacts from each phase of the pavement
life cycle: materials; construction; use; maintenance; and end
of life (EOL). The model is built primarily in the GaBi LCA
software package, with external data (e.g., albedo impacts and
carbonation rates) and models (e.g., traffic delay models and
pavements design models) supplemented as necessary. More
information about the MIT model and its workings are available
in work by Santero and colleagues (2011) and Loijos (2011).
Baseline Designs and Emissions
Baseline designs are created and evaluated for 12 functional
units, which collectively characterize each roadway classifi-
cation in the United States. The functional units are based
on centerline-kilometers (cl-km), rather than lane-kilometers
(lkm), in order to capture the impacts of a typical structure as a
whole, including both the mainline and shoulders. Estimates
for the more traditional lkm metric can be back-calculated
using the given data. Geometric and traffic data are taken
from Highway Statistics 2008 (FHWA 2008); accompanying
pavement structures are designed using the American Associ-
ation of State Highway Officials (AASHTO) method for rigid
860 Journal of Industrial Ecology
Table 1 Baseline pavement designs and global warming
potential
Roadway Traffic Total Paved Concrete Base Estimated GWP
classification (AADT/AADTT) lanes width (m)a thickness (mm)
thickness (mm) (Mg CO2-eq/cl-km)c

Rural Interstate 22,000/4,400 4 23 292 152 3,800
Principal arterial 6,400/710 2 12 203 152 1,300
Minor arterial 3,100/310 2 12 191 152 1,200
Major collector 1,200/85 2 10 152 152 770
Minor collector 570/40 2 10 127b 0 540
Local 180/12 2 8 102b 0 340
Urban Interstate 79,000/6,300 6 34 305 152 6,700
Freeway 54,000/2,200 4 23 279 152 2,400
Principal arterial 20,000/790 4 20 216 152 2,100
Minor arterial 9,700/3,980 2 12 178 152 1,400
Collector 4,200/170 2 12 165 0 960
Local 980/39 2 10 127b 0 610
aIncludes mainline and shoulders.
bThese pavements may be thinner than some states allow.
However, the 1993 AASHTO design procedure was still
followed to remain consistent.
cResults from Santero and colleagues (2011).
Note: AADT = annual daily traffic; AADTT = annual daily
truck traffic; GWP = global warming potential; cl-km =
centerline-kilometer; one meter
(m, SI) ≈ 3.28 feet (ft); one millimeter (mm) = 10−3 meters (m,
SI) ≈ 0.039 inches; carbon dioxide equivalent (CO2-eq) is a
measure for describing the
climate-forcing strength of a quantity of greenhouse gases using
the functionally equivalent amount of carbon dioxide as the
reference. One megagram
(Mg) = 1 metric ton (t) = 103 kilograms (kg, SI) ≈ 1.102 short
tons.
pavements (AASHTO 1993, 2004). The concrete mix has a
flexural strength of 4.5 megapascals and uses 335 kilograms per
cubic meter of cementitious material (90% portland cement
and 10% coal fly ash).2 It is important to note that the 10% fly
ash is a gross average for use in a concrete pavement (ACAA

2009; USGS 2009), but is not necessarily a typical replacement
rate for concrete mixes because of potentially poor resistance to
alkali silica reaction.
A 40-year analysis period is used for the baseline designs,
which includes rehabilitation activities at years 20 and 30 con-
sisting of slab replacement (4%) and diamond grinding. Note
that the 40-year analysis period is an assumption, and that con-
crete pavement service lives will, in practice, vary widely. The
analysis period and rehabilitation schedules and activities are
based on surveys of state DOTs with respect to their LCCA
procedures (Rangaraju et al. 2008; Minnesota Department of
Transportation 2007).
Table 1 shows the relevant designs inputs and estimated
life cycle GWPs as determined by the MIT pavement LCA
model. Table S1 in the supporting information available on
the Journal’s Web site contains mass and other relevant data.
More complete descriptions of the baseline designs and the
calculation of the life cycle GWP values are found in work by
Santero and colleagues (2011).
Greenhouse Gas Emission Reduction Strategies
Five GHG reduction strategy strategies are explored, with at
least one strategy from each of the categories presented in the
in-
troduction: (1) reducing embodied emissions through increased
fly ash replacement of cement; (2) increasing albedo using
white
aggregates; (3) increasing carbonation through EOL waste con-
crete management; (4) reducing fuel consumption by adding an
extra rehabilitation activity; and (5) reducing embodied emis-
sions by avoiding overdesign through the use of advanced
design

models. A summary of strategies and the notable differences be-
tween the baseline scenarios are given in table 2. The relevant
inventory emission data are given in table 3.
Of note is that the chosen strategies are not meant to be
an exhaustive set of options for reducing GHG emissions, but
rather an exploratory set of opportunities. Also of note is that
these reductions are based on average roadway dimensions and
structures, thus lacking the project-specific inputs that are nec-
essary to obtain context-specific results. The intent is to provide
estimates for a select number of generalized strategies in order
to gain insight into the magnitude of possible GHG reductions.
1. Fly ash is already widely used in the concrete industry as
a supplementary cementitious material (SCM). An in-
crease from 10% (the average fly ash used in concrete
pavement mixes) to 30% fly ash replacement is mod-
eled here to exemplify the possible reduction from one
embodied emissions reduction strategy. The 30% replace-
ment of cement with fly ash is based on a survey of DOT
practices (ACPA 2011), but is admittedly a conserva-
tive ceiling. An added benefit of higher fly ash contents
is expedited carbonation: Mixes with 10% and 30% re-
placement have been shown experimentally to increase
the carbonation coefficient (i.e., the rate of carbonation)
by approximately 5% and 10%, respectively (Lagerblad
2006).
2. White aggregates (both fine and coarse) are used in pave-
ment design to increase the pavement albedo. Increased
Santero et al., GHG Reduction Oppor tunities for Concrete
Pavements 861

Table 2 Summary of key differences between baseline and GHG
reduction scenarios
Strategy Description Baseline scenario GHG Reduction scenario
1. Increasing fly ash Increased usage of fly ash to replace
portland cement
10% fly ash replacement 30% fly ash replacement
2. White aggregate Switch to high-albedo fine and coarse
aggregates
αconcrete = 0.33 αconcrete = 0.41
3. EOL stockpiling Crush and expose recycled concrete to
expedite carbonation
0% EOL carbonation 28% EOL carbonation
4. Extra rehabilitation Grind at year 10 to reduce pavement
roughness
See table S3 on the Web See table S3 on the Web
5. Avoiding overdesign Reduce material demand by using a
mechanistic-empirical design approach
See table S3 on the Web See table S3 on the Web
Note: GHG = greenhouse gas; EOL = end of life.
Table 3 Inventory data for significant materials and processes
relevant to the reduction scenarios

GWP emissions factor Source
Cement 0.93 kg CO2-eq/kg Marceau and colleagues (2006)
Fly ash 0.01 kg CO2-eq/kg PE International (2011)
Water 0.005 kg CO2-eq/kg PE International (2011)
Aggregate 0.0032 kg CO2-eq/kg Zapata and Gambatese (2005)a
Dieselb 3.2 kg CO2-eq/L PE International (2011)
Gasolineb 2.6 kg CO2-eq/L PE International (2011)
Truck transport 0.089 kg CO2-eq/Mg-km PE International
(2011)
Pavement roughness Cars: 0.01 L/km per 1 m/km increase in
IRI Zaabar and Chatti (2010)a
Trucks: 0.04 L/km per 1 m/km increase in IRI
Diamond grinding 1,600 L diesel/lkm IGGA (2009)
Radiative forcing 2.6 kg CO2-eq/m2 per 0.01 decrease Akbari
and colleagues (2009)
Urban heat island 4.9 g CO2-eq/m2 per 0.01 decrease in albedo
Rosenfeld and colleagues (1998)
Lighting 0.040 kWh/lumen/yr AASHTO (2005)a
Electricity (input) 0.79 kg CO2-eq/kWh PE International (2011)
In-situ carbonation 1.58 mm/y0.5 (1.65 mm/y0.5 for 30% fly
ash) Lagerblad (2006)
aCO2-eq emission factor value was derived based on data
reported in the given source.
bincludes upstream and combustion emissions.
Note: IRI = international roughness index; kg CO2-eq/kg =
kilograms carbon dioxide equivalent per kilogram; CO2-eq/L =
carbon dioxide equivalent
per liter; CO2-eq/Mg-km = carbon dioxide equivalent per
megagrams per kilometer; m/km = meters per kilometer; lkm =
lane-kilometers; mm/y0.5 =

millimeters per square root of years. One liter (L) = 0.001 cubic
meters (m3, SI) ≈ 0.264 gallons (gal); one square meter (m2, SI)
≈ 10.76 square feet
(ft2); one kilowatt-hour (kWh) ≈ 3.6 × 106 joules (J, SI) ≈
3.412 × 103 British Thermal Units (BTU).
albedo increases the reflectivity of the pavement surface,
allowing for reduced lighting demand, decreased urban
heat island effect, and increased radiative forcing. The
average albedo of the baseline concrete pavement is taken
to be 0.33; the white aggregate pavement has an albedo
of 0.41 (Levinson and Akbari 2002). Tables S1 and S2
in the supporting information on the Web contain infor-
mation on the estimated lighting demands for the various
roadway classifications.
3. EOL stockpiling consists of crushing and stockpiling the
concrete for 1 year, during which time it was assumed to
sequester 28% of the initial CO2 released from carbona-
tion, or 155 grams of CO2 per kilogram of cement in the
mix (Dodoo et al. 2009). It should be noted that actual
carbonation is difficult to pinpoint and that the empirical
data used for this estimate should be refined as more pre-
cise models become available. There are also practicality
issues to consider, such as the willingness of DOTs and/or
industry to stockpile recycled concrete for months at a
time. This strategy represents only one option available
at the EOL, although other options are likely similar in
terms of the magnitude of emission reductions.
4. Adding an extra rehabilitation at year 10 reduces vehicle
fuel consumption by creating a smoother ride. Zaabar and
Chaati (2010) estimate that a decrease in roughness of 4
meters per kilometer (m/km) reduces fuel consumption
by 4.2% for cars and 2.8% for trucks. The extra rehabil-

itation itself consumes additional energy from diamond
grinding and requires that the structure is 1 centimeter
thicker at the initial construction in order to account
for the material that will be removed during the grinding.
The additional activity benefits the life cycle in two ways:
First, the pavement roughness is brought back down to an
initial international roughness index (IRI) of 1.0 m/km;
second, completely uncarbonated concrete is exposed to
the environment and carbonation resumes again at its
faster, initial rate. The average IRI values for years 10
through 20 are given in table S3 in the supporting infor-
mation on the Web for both the baseline and reduction
scenarios.
5. Avoiding overdesign decreases embodied emissions by
optimizing the materials necessary to construct the pave-
ment structure. Long-term pavement performance data
collected by the FHWA suggest that concrete pave-
ments routinely supported up to ten times the traf-
fic that they were designed to carry (CEMEX 2010).
In order to evaluate the GWP of the more accurate
designs, the Mechanistic-Empirical Pavement Design
Guide (MEPDG) models was used to create alternative
designs using equivalent traffic and service life inputs,
assuming moderate climate conditions. The structure de-
signs for six of the twelve roadway classifications are listed
in Table S3 in the supporting information on the Web, as
compared to their 1993 AASHTO equivalents. MEPDG
is primarily a high-traffic volume design tool and does

not provide outputs of less than 178 millimeters for the
concrete slab thickness, so the low-volume classifications
are not analyzed.
Cost-Effectiveness Analysis
Cost-effectiveness analysis (CEA) is most commonly asso-
ciated with the health and medicine fields, where it is used to
evaluate the cost of different interventions with respect to their
ability to increase quality of life (Gold 1996). Applying the
con-
cept to pavements and GHG emissions, reduction strategies can
be evaluated not only on their reduction potential, but also on
the relative cost of that reduction. Thus, cost-effectiveness in
this study speaks to the cost to reduce GHG emissions,
measured
in U.S. dollars per megagrams of CO2 equivalent ($/Mg CO2-
eq).3 Equation 1 provides the basic relationship between costs,
emissions, and cost-effectiveness (CE). The “alt” and “base”
subscripts refer to the reduction alternative and baseline case,
respectively.
CEalt =
costalt − costbase
emmissionsalt − e mi s s i onsbase
(1)
= �costalt−base
�emissionsalt−base
The outputs of the GWP reduction analysis determine the
values for the denominator of equation (1); the numerator is
determined through economic analysis. Following established
LCCA protocols, the absolute cost of each strategy is not neces-
sary to compute if the difference between the base and

reduction
strategy cases is known. Because many cost inputs will be iden-
tical between alternatives (e.g., construction processes, mobi-
lization, and unit costs), the demand for data is significantly
reduced. Practitioners can focus on the differences between
designs rather than calculating comprehensive, but largely ir-
relevant, absolute costs. Sensitivity analyses are performed for
selected parameters in order to estimate a range of expected
costs. Table 3 summarizes the cost and other data used in the
CEA.
This analysis uses a transportation agency perspective on
cost abatement, thus adopting the LCCA approach that DOTs
currently use in their decision-making process. In general, the
FHWA (Walls and Smith 1998) recommends using the dis-
count rate published in the most current version of the White
House Office of Management and Budget (OMB) Circular A-
94; accordingly, this analysis discounts future costs at a rate of
2.3% (OMB 2010). It should be noted that many abatements
analyses, such as McKinsey & Company (Creyts et al. 2007),
use levelized costs, particularly in the field of energy improve-
ments where the concept was first established (Meier 1984).
This approach annualizes the economic impact over the life of
the reduction strategy. In order to equitably compare the results
in this CEA with other abatement curves, it may be necessary
to convert the results to levelized costs using the data already
provided.
Results
Greenhouse Gas Emission Reductions and Costs
The reductions in GWP for each scenario are shown in
figure 1. The absolute values show quantity of GWP reduced.
Higher-volume roadways, such as urban interstates, have larger

absolute reduction potentials because of the larger structures
and from roughness-related vehicle fuel consumption. Accord-
ingly, reducing embodied emissions (through increased fly ash
or avoided overdesign) and reducing smoothness (through an
extra rehabilitation) have the largest reductions for interstates.
Although lower-volume roadways have smaller absolute re-
duction potentials, the reductions relative to their baseline sce-
narios are significant. Local roads contribute roughly ten times
less life cycle GWP than their interstate counterparts, so even
small reductions can have a large influence on the overall foot-
print. In particular, increasing albedo (through white aggre-
gates) results in high relative reductions for local roads and
collectors, with the strategy reducing GWP by 20% for these
scenarios.
The cost-effectiveness of the GHG reduction strategies are
shown in figure 2. The solid bars represent the results using the
best-estimate data shown in table 4; the error bars represent
the sensitivity to the low- and high-estimate data. Note that
for clarity purposes, the y-axis stops at $250/Mg CO2-eq saved,
even though some points are above that threshold. Strategies
at that cost magnitude are significantly higher than estimated
carbon prices and are thus considered to be above reasonable
cost-effectiveness limits.
Each scenario for both of the embodied emissions strate-
gies has a negative cost-effectiveness value, meaning that the
strategies reduce both costs and emissions. Avoiding overde-
sign essentially reduces the thicknesses of the concrete and/or
base layers, thus mitigating the costs and emissions associated
with extraction, production, and handling of natural resources.
Pavements 863

0%
10%
20%
30%
40%
50%
R
el
at
iv
e
to
B
as
el
in
es
Increased fly ash White aggregate EOL stockpiling Extra
rehabilitation MEPDG case study

0
200
400
600
800
1,000
In
te
rs
ta
te
Pr
in
ci
pa
l a
rt
er
ia
l
M
in

or
a
rt
er
ia
l
M
aj
or
c
ol
le
ct
or
M
in
or
c
ol
le
ct
or
L
oc

al
ro
ad
In
te
rs
ta
te
Fr
ee
w
ay
Pr
in
ci
pa
l a
rt
er
ia
l
M
in

or
a
rt
er
ia
l
C
ol
le
ct
or
L
oc
al
ro
ad
A
bs
ol
ut
e
(M
g

C
O
2-
eq
/c
l-k
m
)G
W
P
R
ed
uc
tio
n
Rural Roadways Urban Roadways
Figure 1 Life cycle GWP reductions shown in terms of absolute
emission reductions (bottom) and relative reductions compared
to
the baselines (top). EOL = end of life; MEPDG = Mechanistic-
Empirical Pavement Design Guide; GWP = global warming
potential; Mg
CO2-eq/cl-km = megagrams carbon dioxide-equivalent per
centerline-kilometer.
Table 4 Costs and other data used to conduct the CEA for the
GHG emission reduction strategies

Parameter Best estimate Low estimate High estimate Source
Cement ($/Mg) $102 — — USGS (2009)
Fly ash ($/Mg) $50 $25 $65 Tikalsky and colleagues (2011)
Truck transport ($/Mg-km) $0.10 — — Assumed
Extra aggregate haul (km) 50 0 200 Assumed
Recycled concrete value ($/Mg) $7.43 — — USGS (2008)
Annual carrying cost (%/Mg/yr) 25% 20% 40% Hendrickson
(2008)
Grinding cost ($/m2) $4.31 $4.00 $5.00 Caltrans (2011)
Concrete pavement ($/m3) $212 $151 $273 Caltrans (2011)
Aggregate base ($/m3) $83 $51 $114 Caltrans (2011)
Note: $/Mg = U.S. dollars per megagram; $/Mg-km = dollar per
megagram per kilometer; km = kilometer; %/Mg/yr = percent
per megagram per year;
$/m2 = dollars per square meter; $/m3 = dollars per cubic
meter.
Increasing the fly ash content has negative cost-effectiveness
values for the same reason, although the magnitude is consider-
ably lower because the cost reductions are limited to the
binding
agent, rather than the structure as a whole.
The strategies to increase albedo, increase EOL carbona-
tion, and reduce vehicle fuel consumption result in positive
cost-effectiveness values. Following the trend from the emis-
sion reductions results, the use of white aggregates is more
cost-effective on low-volume pavements, whereas the extra re-
habilitation is more cost-effective on high-volume roadways.
The cost-effectiveness of increasing carbonation through EOL
stockpiling is consistent across all the classifications.

Figure 3 combines the absolute GWP reduction and the
associated cost-effectiveness for the urban interstate and rural
local road scenarios. The plot exemplifies the differences that
exist between different roadway classifications. For instance,
the
-$1,500
-$1,250
-$1,000
-$750
-$500
-$250
$0
$250
In
te
rs
ta
te

Pr
in
ci
pa
l a
rt
er
ia
l
M
in
or
a
rt
er
ia
l
M
aj
or
c
ol
le
ct

or
M
in
or
c
ol
le
ct
or
L
oc
al
ro
ad
In
te
rs
ta
te
Fr
ee
w
ay

Pr
in
ci
pa
l a
rt
er
ia
l
M
in
or
a
rt
er
ia
l
C
ol
le
ct
or
L
oc

al
ro
ad
C
os
t E
ff
ec
tiv
en
es
s
($
/M
g
C
O
2-
eq
re
du
ce
d)

Increased fly ash White aggregate EOL stockpiling Extra
rehabilitation MEPDG case study
Rural Roadways Urban Roadways
Figure 2 Cost-effectiveness of five greenhouse gas reduction
strategies (solid bars represent best-estimate data, and error
bars represent
low and high data).
-$800
-$600
-$400
-$200
$0
$200
$400
C
os
t E
ff
ec
tiv
en
es

s
($
/M
g
C
O
2-
eq
s
av
ed
)
GWP Reduction
(vertical gridlines are in 100 Mg CO2-eq increments)
MEPDG case study
Increased fly ash
EOL stockpiling
Extra rehabilitation
White aggregate
Urban interstate Rural local road
Figure 3 Cost of GWP abatement comparison of urban
interstates versus rural local roads. The width of the bars
represents

the total reduced GWP, with the vertical gridlines representing
100 Mg CO2-eq increments. GWP = global warming potential;
Mg
CO2-eq/cl-km = megagrams carbon dioxide-equivalent.
white aggregate strategy is not practical for urban interstates:
The high cost and relatively small reduction potential is a poor
combination. Conversely, for rural local roads, the white aggre-
gate strategy offers a significant GWP reduction at costs that
are comparable to the price of carbon.
Discussion
The fly ash scenario is a good example of reducing embodied
emissions by adjusting the amount of cement in the mix. Ce-
ment has been shown to be the largest GWP contributor over
the concrete life cycle (Santero et al. 2011), so it is reasonable
to assume that decreasing the cement content through the use
of SCMs or optimized mix designs is a reasonable reduction
approach. The results coincide with this assertion, showing a
10% to 20% reduction across all roadway classifications. The
replacement of cement with a by-product of the coal com-
bustion process also reduces costs: A metric ton of CO2-eq
corresponds to a $40 savings in material costs. Blast furnace
ground-granulated slag and silica fume are examples of other
SCMs that may provide similar results. It should be noted that
the quality and regional availability of the fly ash (and other
SCMs) will affect the efficacy of this reduction option; this
study
assumes high-quality class F fly ash that is practically
available.
Moreover, any health hazard concerns associated with fly ash
are considered outside the scope of this study.

Pavements 865
The driving forces behind the cost-effectiveness of using
white aggregates to increase albedo are the depth of the con-
crete and the local availability of the aggregates. If an extra
haul
distance of 50 km is necessary to acquire white aggregates, this
is a relatively cost-effective GHG reduction strategy for low-
volume classifications (e.g., $41/Mg CO2-eq reduced for rural
local roads). As the extra haul distance increases, both the cost
and total emissions increase, causing the cost-effectiveness to
quickly rise to levels well above the price of carbon. Consid-
ering that white aggregates may not be locally available for
many projects, this strategy is not universally applicable. This
strategy favors low-volume classifications because of the thin-
ner concrete layer needed for the structure. Because albedo is a
surface property, pavements with high surface-area/concrete-
thickness ratios will have better cost-effectiveness: Only
the fine and coarse white aggregates at the top of the structure
will contribute to the albedo reduction. Alternatively, concrete
overlays and two-lift concrete structures could take advantage
of this concept by utilizing the albedo benefits of white ag-
gregate while minimizing the “wasted” white aggregates in the
structure.
Facilitating the natural carbonation process of recycled con-
crete aggregates presents an opportunity to sequester a
consider-
able amount of the CO2 released during cement manufacturing.
Stockpiling and exposing recycled aggregate for 1 year is a rel-
atively cost-effective approach ($31/Mg CO2-eq reduced), but
standard practices of DOTs and other stakeholders may make

this an impractical method. In particular, recycled concrete ag-
gregates tend to be used quickly after they are processed, some-
times even immediately in the case where a mobile crushing
unit is available at the construction site. Moreover, if this strat-
egy were continuously applied for many pavements, the result
would be an effective removal of some tonnage of aggregate
sup-
ply from the available stock. The effect would be an induced
de-
mand for virgin aggregate to replace the lost stock—something
that a more thorough LCA might consider within its bound-
aries. Although EOL carbonation arguably is most effective
when the recycled concrete aggregate is directly exposed to the
environment, research shows that even buried crushed concrete
sequesters a significant amount of CO2 (Collins 2010). If EOL
stockpiling is considered impractical, then engineers should at
least consider alternative methods of promoting carbonation at
the end of the concrete pavement life cycle.
Adding an extra rehabilitation is potentially a cost-effective
method for reducing emissions of high-volume roadways, al-
though the results presented here seem to suggest otherwise.
The representative structures and inputs are for average con-
ditions across the 12 roadway classifications and thus do not
capture many outlying scenarios—such as those with high traf-
fic volumes and/or high IRI values—that could benefit from this
strategy. Adding an extra rehabilitation for urban interstates has
a cost-effectiveness of $140/Mg CO2-eq reduced (significantly
higher than the price of carbon), but are only modeled for the
average traffic of 79,000 vehicles per day. With volumes
ranging
up to 130,000 and higher on some urban interstates, an extra re-
habilitation could provide significantly better cost-effectiveness
for pavements under different conditions. Moreover, the road-
ways modeled here are in relatively good condition at year

10, which is when the extra rehabilitation is assumed to oc-
cur: The IRI at year 10 is 1.2 m/km, with grinding assumed to
reduce the roughness to 1.0 m/km. Roadways with higher pre-
rehabilitation IRI values will benefit more from grinding, lead-
ing to larger emission reductions and better cost-effectiveness.
Considering that the average urban interstate has an IRI of
1.5 m/km (FHWA 2008), there should be ample opportunities
to reduce emissions through diamond grinding.
Avoiding overdesign shows significant potential as a cost-
effective method of reducing GHG emissions. Using MEPDG
(rather than AASHTO 1993) to design the pavements essen-
tially reduces the thicknesses of the concrete and/or base layers,
thus mitigating the costs and emissions from the associated ma-
terials and processes. MEPDG is climate specific, so data from
a
particular location (Oxnard, CA, USA) were used. A moderate
climate was specifically chosen to show the potential GWP and
cost benefits of MEPDG-derived designs, but it should be noted
that the results may differ in other climates. The perceived
advantage of using MEPDG (or other advanced design proce-
dures) to avoid overdesign will differ from project to project. In
some cases, the baseline design may already be relatively accu-
rate (or even underdesigned), thus leaving little to no oppor-
tunities to reduce emissions using this technique. Additionally,
any benefits associated with avoiding overdesign are correlated
to the analysis period itself: Overdesigned pavements will out-
last their intended service life, thus providing service beyond
the analysis period. This study operates under the assumption
that pavements are strategically designed for a particular
service
life and thus an efficient design is one that meets that service
life and minimizes the risk of functional obsolescence (Santero
et al. 2010).
Conclusions

There are multiple approaches to reduce GHG emissions of
concrete pavements. Reducing embodied emissions (the quan-
tity and emission intensity of the materials and designs) can
be complemented by increasing the pavement albedo, increas-
ing carbonation at the EOL, and decreasing the fuel consump-
tion of vehicles during the use phase. Each of these approaches
are explored using representative strategies: reducing embodied
emissions by increasing fly ash content and avoiding overde-
sign; increasing albedo by using white aggregate; increasing
car-
bonation through EOL stockpiling; and reducing vehicle fuel
consumption by adding an extra rehabilitation.
The analyzed designs and input parameters are meant to
represent average concrete structures and conditions for each
of the FHWA roadway classifications. In reality, there is a sig-
nificant variation within each roadway classification, making
it difficult to adopt a single representative structure. Concrete
pavement designs will vary significantly from one pavement
to the next, changing based on regional climate, local design
practices, budget, service life, material availability, and other
factors. For instance, urban interstates routinely support be-
tween 30,000 and 130,000 vehicles per day (FHWA 2008), but
the weighted average (79,000) is used in this analysis. This not
only affects operating emissions (e.g., roughness-related vehicle
fuel consumption), but also the materials and geometry of the
structure. This approach is useful in generally characterizing a
large breadth of pavement functions, but may also fail to ade-

quately capture the impacts caused by atypical structures within
each classification. Project-specific analyses are better suited to
accurately quantify the impacts associated with a particular,
well-defined pavement. Even with the generalized approach
adopted in this research, several overarching conclusions can
be drawn:
� Significant GHG emission reductions are possible. Over
half of the scenarios result in emissions reductions greater
than 150 Mg CO2-eq per cl-km, with high-volume road-
ways generally offering higher absolute reduction poten-
tials as a result of their more massive designs and higher
traffic volume. Relative to the unimproved baseline de-
signs, over half of the scenarios reduce emissions by over
10%. Relative emission reductions tend to be greater
for low-volume roadways because of the combination
of small baseline emissions, disproportionality of albedo
with structure depth, and larger dependence on materials-
based emissions.
� There are cost-effective methods to reduce GHG emis-
sions for concrete pavements. Both embodied emis-
sions strategies produce negative cost-effectiveness val-
ues, meaning that costs and emissions are saved simulta-
neously. The strategies for increasing albedo, increasing
carbonation, and reducing vehicle fuel consumption have
positive cost-effectiveness values, but there are scenarios
where each is comparable to the price of carbon. Eval-
uating economic impacts alongside emission reduction
potentials is essential in order to identify the feasibility of
implementing a given reduction strategy.
� The emission reduction potential and cost-effectiveness
of a GHG emission reduction strategy changes based on
the classification of the roadway. Scarcity concerns aside,
increasing albedo by using white aggregate stands out as

an effective method of reducing GWP for low-volume
roadways. For high-volume roadways, the inefficient use
of the specialty material (only a small fraction of the ag-
gregates contribute to the increased albedo) limits the
reduction potential and disproportionately increases the
costs, resulting in a poor effectiveness for these roadways.
Conversely, adding an extra rehabilitation in order to re-
duce vehicle fuel consumption has the potential to be ef-
fective on high-volume roadways, but is not effective (and
potentially counterproductive) for low-volume roadways.
Acknowledgments
This work was performed at the Concrete Sustainability Hub
at the Massachusetts Institute of Technology. The program is
funded by the Portland Cement Association and the National
Ready-Mix Concrete Education and Research Foundation.
Notes
1. One vehicle-kilometer (km, SI) ≈ 0.621 vehicle-miles (mi).
2. One kilogram (kg, SI) ≈ 2.204 pounds (lb). One cubic meter
(m3,
SI) ≈ 35.3 cubic feet (ft3).
3. Throughout this article, $ indicates U.S. dollars.
References
AASHTO (Association of State Highway and Transportation
Offi-
cials). 1993. AASHTO guide for the design of pavement
structures.
Washington, DC: American Association of State Highway and
Transportation Officials.

Offi-
cials). 2004. A policy on geometric design of highways and
streets, 5th
ed. Washington, DC: American Association of State Highway
and Transportation Officials.
Offi-
cials). 2005. Environmental stewardship practices, procedures,
and
policies for highway construction and maintenance.
Washington,
DC: American Association of State Highway and Transportation
Officials.
ACAA (American Coal Ash Association). 2009. 2008 coal
combustion
product (CCP) production & use survey report. Aurora, CO,
USA:
American Coal Ash Association.
ACPA (American Concrete Pavement Association). 2011. ACPA
Application Library: Agency Practices Explorer. http://1734298.
sites.myregisteredsite.com/apps/. Accessed April 2011.
Akbari, H., S. Menon, and A. Rosenfeld. 2009. Global cooling:
Increas-
ing world-wide albedos to offset CO2. Climatic Change 94(3–
4):
275–286.
Caltrans. 2011. Contract Cost Database (2010 prices). Cal-
ifornia Department of Transportation. http://sv08data.dot.
ca.gov/contractcost. Accessed July 2011.

CEMEX. 2010. Mechanistic Empirical Pavement Design Guide
(MEPDG): Why it is important and why the industry should be
pushing its adoption. Paper presented at the ACPA 47th Annual
Meeting, 1–3 December, Bonita Springs, FL, USA.
Collins, F. 2010. Inclusion of carbonation during the life cycle
of built
and recycled concrete: Influence on their carbon footprint.
Inter-
national Journal of Life Cycle Assessment 15(6): 549–546.
Creyts, J., A. Derkach, S. Nyquist, K. Ostrowski, and J.
Stephenson.
2007. Reducing U.S. greenhouse gas emissions: How much at
what
cost? Washington, DC: McKinsey & Company.
Dodoo, A., L. Gustavsson, and R. Sathre. 2009. Carbon
implications
of end-of-life management of building materials. Resources,
Con-
servation and Recycling 53(5): 276–286.
EPA (Environmental Protection Agency). 2009. 2009 U.S.
greenhouse gas inventory report. www.epa.gov/climatechange/
emissions/downloads09/GHG2007entire_report-508.pdf. Acce-
ssed February 2010.
FHWA (Federal Highway Administration). 2008. Highway
statistics
2008. Federal Highway Administration, United States Depart-
ment of Transportation. www.fhwa.dot.gov/policyinformation/
statistics/2008. Accessed July 2011.

Pavements 867
http://1734298.sites.myregisteredsite.com/apps/
http://1734298.sites.myregisteredsite.com/apps/
http://sv08data.dot.ca.gov/contractcost
http://sv08data.dot.ca.gov/contractcost
http://www.epa.gov/climatechange/emissions/downloads09/GH
G2007entire_report-508.pdf
http://www.epa.gov/climatechange/emissions/downloads09/GH
G2007entire_report-508.pdf
http://www.fhwa.dot.gov/policyinformation/statistics/2008
http://www.fhwa.dot.gov/policyinformation/statistics/2008
Gold, M. 1996. Cost-effectiveness in health and medicine. New
York:
Oxford University Press.
Hendrickson, C. 2008. Project management for construction:
Fundamental
concepts for owners, engineers, architects, and builders.
Pittsburgh,
PA: Carnegie Mellon University. http://pmbook.ce.cmu.edu. Ac-
cessed August 2011.
IGGA (International Grooving and Grinding Association). 2009.
Conserving fuel when rehabilitating concrete roads. West
Coxsackie,
NY, USA: International Grooving & Grinding Association.
Lagerblad, B. 2006. Carbon dioxide uptake during concrete life
cycle—State
of the art. Stockholm: Swedish Cement and Concrete Research
Institute, CBI.

Levinson, R. and H. Akbari. 2002. Effects of composition and
exposure
on the solar reflectance of portland cement concrete. Cement
and
Concrete Research 32(11): 1679–1698.
Loijos, A. N. 2011. Life cycle assessment of concrete
pavements:
Impacts and opportunities. Master’s thesis, Massachusetts In-
stitute of Technology, Cambridge, MA, USA. http://hdl.
handle.net/1721.1/65431. Accessed August 2011.
Marceau, M., M. Nisbet, and M. VanGeem. 2006. Life cycle
inventory of
portland cement manufacture. Skokie, IL, USA: Portland
Cement
Association.
Meier, A. 1984. The cost of conserved energy as an investment
statistic.
Proceedings from the Sixth Annual Industrial Energy
Conference
Volume II, 15–18 April, Houston, TX, USA.
OMB (White House Office of Management and Budget). 2010.
Circular A-94 appendix C: Discount rates for cost effec-
tiveness, lease purchase, and related analysis. White House
Office of Management and Budget. www.whitehouse.gov/
omb/circulars_a094/a94_appx-c. Accessed July 2011.
PE International. 2011. GaBi software. www.gabi-software.com.
Ac-
cessed May 2011.
Rangaraju, P., S. Amirkhanian, and Z. Guven. 2008. Life cycle

cost
analysis for pavement type selection. Report no. FHWA-SC-08-
01.
Columbia, SC, USA: South Carolina Department of Transporta-
tion.
Rosenfeld, A., H. Akbari, J. Romm, and M. Pomerantz. 1998.
Cool
communities: Strategies for heat island mitigation and smog re-
duction. Energy and Buildings 28(1): 51–62.
Santero, N. and A. Horvath. 2009. Global warming potential of
pave-
ments. Environmental Research Letters 4(3), DOI:
10.1088/1748-
9326/4/3/034011.
Santero, N., J. Harvey, and A. Horvath. 2010. Environmental
policy
for long-life pavements. Transportation Research Part D 16(2):
129–136.
Santero, N., A. Loijos, M. Akbarian, and J. Ochsendorf. 2011.
Meth-
ods, impacts, and opportunities in the concrete pavement life
cycle.
Cambridge, MA, USA: Massachusetts Institute of Technology.
Tikalsky, P., P. Taylor, S. Hanson, and P. Ghosh. 2011.
Development of
performance properties of ternary mixtures: Laboratory study on
con-
crete. National Concrete Pavement Technology Center. Report
no. DTFH61-06-H-00011. Washington, DC: Federal Highway
Administration.

USDOT (United States Department of Transportation). 2011.
Trans-
portation statistics annual report 2010. Washington, DC: United
States Department of Transportation.
USGS (United States Geological Survey). 2008. 2008 minerals
yearbook: Sand and gravel, construction. http://minerals.usgs.
gov/minerals/pubs/commodity/sand_&_gravel_construction.
Accessed July 2011.
USGS (United States Geological Survey). 2009. 2009 miner-
als yearbook: cement. http://minerals.usgs.gov/minerals/pubs/
commodity/cement. Accessed July 2011.
USGS (United States Geological Survey). 2011. 2011 minerals
yearbook: Stone, crushed.
http://minerals.usgs.gov/minerals/pubs/
commodity/stone_crushed. Accessed July 2013.
Walls, J. and M. Smith. 1998. Life-cycle cost analysis in
pavement design:
Interim technical bulletin. Report no. FHWA-SA-98-079. Wash-
ington, DC: Federal Highway Administration.
Zaabar, I. and K. Chatti. 2010. Calibration of HDM-4 models
for esti-
mating the effect of pavement roughness on fuel consumption
for U.S.
conditions. Transportation Research Record (2155): 105–116.
Zapata, P. and J. Gambatese. 2005. Energy consumption of
asphalt
and reinforced concrete. Journal of Infrastructure Systems
11(1):
9–20.

About the Authors
Nicholas Santero was a research scientist at the Mas-
sachusetts Institute of Technology in Cambridge, MA, USA,
when this research was performed. He is currently a senior
consultant at PE International in San Francisco, CA, USA.
Alexander Loijos was a graduate student researcher at the Mas-
sachusetts Institute of Technology when this research was per-
formed. He is currently cofounder and director of LinkCycle,
LLC, in Cambridge, MA, USA. John Ochsendorf is an asso-
ciate professor at the Massachusetts Institute of Technology.
Supporting Information
Additional Supporting Information may be found in the online
version of this article at the publisher’s web site:
Supporting Information S1: This supporting information
provides several tables with input data for baseline scenarios,
the
roughness index (IRI) and design thickness for various
strategies, and estimates of global warming reduction potentials
and
cost effectiveness for different scenarios.
http://pmbook.ce.cmu.edu
http://hdl.handle.net/1721.1/65431
http://hdl.handle.net/1721.1/65431
http://www.whitehouse.gov/omb/circulars_a094/a94_appx-c
http://www.whitehouse.gov/omb/circulars_a094/a94_appx-c
http://www.gabi-software.com
http://minerals.usgs.gov/minerals/pubs/commodity/sand_&_grav
el_construction
http://minerals.usgs.gov/minerals/pubs/commodity/sand_&_grav

el_construction
http://minerals.usgs.gov/minerals/pubs/commodity/cement
http://minerals.usgs.gov/minerals/pubs/commodity/cement
http://minerals.usgs.gov/minerals/pubs/commodity/stone_crushe
d
http://minerals.usgs.gov/minerals/pubs/commodity/stone_crushe
d
Copyright of Journal of Industrial Ecology is the property of
Wiley-Blackwell and its content
may not be copied or emailed to multiple sites or posted to a
listserv without the copyright
holder's express written permission. However, users may print,
download, or email articles for
individual use.

A P P L I C A T I O N S A N D I M P L E M E N T A T I O Nh.docx

Recommended

Recommended

More Related Content

Similar to A P P L I C A T I O N S A N D I M P L E M E N T A T I O Nh.docx

Similar to A P P L I C A T I O N S A N D I M P L E M E N T A T I O Nh.docx (20)

More from fredharris32

More from fredharris32 (20)

Recently uploaded

Recently uploaded (20)

A P P L I C A T I O N S A N D I M P L E M E N T A T I O Nh.docx