| || |
| || || |
Industrial Ecology: Concepts And Approaches
L. W. JELINSKI*, T. E. GRAEDEL, R. A. LAUDISE, D. W. MCCALL,
AND C. K. N. PATEL AT&T Bell Laboratories. Murray Hill,
NJ 07974 .
This paper serves as an introduction to the following papers, which
were presented at a colloquium entitled "Industrial Ecology," organized
by C. Kumar, N. Patel, held May 20-21, 1991, at the National Academy
of Sciences, Washington, DC.
ABSTRACT: Industrial ecology is a new approach to the industrial
design of products and processes and the implementation of sustainable
manufacturing strategies. It is a concept in which an industrial
system is viewed not in isolation from its surrounding systems
but in concert with them. Industrial ecology seeks to optimize
the total materials cycle from virgin material to finished material,
to component, to product, to waste product, and to ultimate disposal.
To better characterize the topic, the National Academy of Sciences
convened a colloquium from which were derived a number of salient
contributions. This paper sets the stage for the contributions
that follow and discusses how each fits into the framework of
It is not difficult to find evidence that human activities are
beginning to overrun the resources of the planet. The exhaustion
of the oil that has fueled the industrial revolution appears to
be less than a century away, waste disposal sites are in increasingly
short supply, the atmosphere over Antarctica is being radically
changed as a consequence of industrially synthesized halogenated
gases, and so forth. In many of these cases, the industrial processes
that have benefited society are also among the sources of the
problems. It is clear that "business as usual" is not an option
that industry can maintain for long. Nonetheless, the complexities
involved in new approaches to industrial design and operation
are many and individuals who are adequately equipped to make decisions
with the full spectrum of those complexities in mind are few.
Recognizing that new approaches are required for the industrial
design of products and processes and the implementation of sustainable
manufacturing strategies, the National Academy of Sciences, in
cooperation with the AT&T Foundation, sponsored a colloquium
in Washington, DC, on May 20-21, 1991. In this communication,
we wish to frame the concept on which that discussion was based
and to describe how the papers that follow fit into that overall
Industrial Ecology: A Systems Description
Traditional biological ecology is defined as the scientific study
of the interactions that determine the distribution and abundance
of organisms. The relationship between this concept and that of
industrial activities has been discussed by Frosch and Gallopoulos:
In a biological ecosystem. some of the organisms use sunlight,
water, and minerals to grow, while others consume the first. alive
or dead, along with minerals and gases, and produce wastes of
their own. These wastes are in turn food for other organisms,
some of which may convert the wastes into the minerals used by
the primary producers, and some of which consume each other in
a complex network of processes in which everything produced is
used by some organism for its own metabolism. Similarly, in the
industrial ecosystem, each process and network of processes must
be viewed as a dependent and interrelated part of a larger whole.
The analogy between the industrial ecosystem concept and the biological
ecosystem is not perfect, but much could be gained if the industrial
system were to mimic the best features of the biological analogue.
It is instructive to think of the materials cycles involved with
the earliest of earth’s life forms. At that time, the potentially
usable resources were so large and the amount of life so small
that the existence of life forms had essentially no impact on
available resources. This individual component process might be
described as linear, that is, as one in which the flow of material
from one stage to the next is independent of all other flows.
We term this pattern type I ecology; schematically, it takes the
form of Fig. 1.
A contrasting picture is an ecosystem in which proximal resources
are limited. In such a system, the resulting life forms become
strongly interlinked and form the complex networks we know today
as biological communities. In this system, the flows of material
within the proximal domain may be quite large, but the flows into
and out of that domain (i.e., from resources and to waste) are
quite small. Schematically, such a type II system might be expressed
as in Fig. 2.
The type 11 system is much more efficient than the previous one,
but it clearly is not sustainable over the long term because the
flows are all in one direction, that is, the system is "running
down." To be ultimately sustainable, biological ecosystems have
evolved over the long term to be almost completely cyclical in
nature, with "resources" and "waste" being undefined, since waste
to one component of the system represents resources to another.
[The exception to the cyclicity of the overall system is that
energy (in the form of solar radiation) is available as an external
resource.] This type III system may be pictured as in Fig. 3.
It is important to recognize that the cycles within the system
tend to function on widely differing temporal and spatial scales,
a behavior that greatly complicates analysis and understanding
of the system.
The ideal anthropogenic use of the materials and resources available
for industrial processes (broadly defined to include agriculture,
the urban infrastructure, etc.) would be one that is similar to
the ensemble biological model. Many uses of materials have been
and continue to be essentially dissipa-tive, however. That is,
the materials are degraded, dispersed, and lost in the course
of a single normal use, mimicking the type I unconstrained resource
diagram. This trend can be associated with the maturation of the
Industrial Revolution of the 18th century, which, together with
exponential increases in human population and agricultural production,
took place essentially in a context of global plenty in which
copious quantities of energy and raw materials were available.
Many present day industrial processes and products still remain
largely dissipative. Examples include lubricants, paints, pesticides,
and automobile tires.
Not all present day technological processes are completely dissipative,
however. Where specific materials are sufficiently precious, some
demonstrate ecological type 11 behav-ior, at least in part. An
example is precious metals recovery.
On the broadest of product and time scales, there are many demonstrations
today that the flows in the ensemble of industrial ecosystems
are so large that resource limitations are setting in: the rapid
changes in atmospheric ozone, increases in atmospheric carbon
dioxide, and the filling of available waste disposal sites being
examples. Accordingly, industrial systems (and other anthropogenic
systems) are and will increasingly be under selective pressure
to evolve so as to move from linear (type 1) to semicyclic (type
11) modes of operation.
The central domain of industrial ecology is conveniently pictured
with four central nodes the materials extractor or grower, the
materials processor or manufacturer, the con-sumer, and the waste
processor. To the extent that they perform operations within the
nodes in a cyclic manner or organize to encourage cyclic flow
of materials within the entire industrial ecosystem, they evolve
into modes of oper-ation that are more efficient, have less disruptive
impact on external support systems, and are more like type 111
ecological behavior. Examples range from the recycling of iron
scrap by the heavy metals industry to the popularity of garage
sales for the reuse of products within the consumer domain. The
schematic model of such an industrial ecosystem is shown in Fig.
4. Note the flows within the nodes and within the industrial ecological
system as a whole are much larger than the external resource and
Characterization of Industrial Ecology
The first of the contributed papers in this issue is that of Patel
(1). His presentation emphasizes the basic characteristics of industrial
ecology and points out the many perceptions that inhibit such a
system from functioning properly. He calls for enhanced communication
among specialists, a restraint on the use of jargon, and a willingness
to develop new concepts to deal with new approaches to understanding
the functioning of industrial processes.
For many of the papers from the colloquium, it is helpful to
use the schematic model of Fig.4 to place them into the framework
of the industrial ecosystem. From this perspective, the paper
by Frosch (2) treats the ensemble of constituents within the broken
line. Perhaps the most striking of Frosch’s many points is that
natural ecosystems, evolving over millions or billions of years,
have produced every entity needed for a type 111 ecosystem. The
industrial ecosystem, in contrast, is a system moving from type
I to type 11 behavior. If we wish it to approach a type 111 system,
we may need to artificially create some of the missing entities.
Frosch regards a healthy recycling and remanufacturing entity
as an example of an entity that is missing, or nearly so, and
may need to be created by fiat if we wish industrial ecology to
proceed at a rapid pace.
Smart’s essay (3) asks why an individual entity in the industrial
ecology system is motivated to act in ways that benefit the system.
He finds the metabolism of this entity to be responsive to economic
forcing and shows how price signals to that entity for such factors
as the true value of natural resources can alter the metabolism
of a corporation. His approach can be taken to refer to either
the "Raw Materials Extractor" or "Materials Processor" boxes in
Fig. 4 and to the factors that control the flows to and from those
Examples of Industrial Ecology in Manufacturing
Industrial ecology may be approached in either of two ways. The
first is material specific; that is, it selects a particular material
or group of materials and analyzes the ways in which it flows through
the industrial ecosystem. Such an analysis in manufacturing operations
is generally made while products are in their manufacturing cycle,
and any modifications to materials or processes tend to be costly
The second type of industrial ecology analysis is one which is
product-specific; that is, it selects a particular product and
analyzes the ways in which its different component materials flows
may be modified or redirected in order to optimize product-environment
interaction. Such an analysis is particularly appropriate at the
initial design stage of a product, when decisions on alternative
materials or processes can often be made at a stage preceding
the investment of large amounts of capital for equipment or process
development, an action that often locks in a particular material
or process for the long term.
Several of the papers that follow present examples of situations
in which one or more aspects of industrial ecology were used in
actual industrial situations. These are activities that fall largely
within the "Materials Processor or Manufacturer" box in Fig. 4
or in the flows to and from that box. An example of a material
specific analysis is that presented by Boyhan (4), in which he
describes the ways in which AT&T has revised its approach
to cleaning electronic circuits while eliminating the use of chlorofluorocarbons
(CFCs). An important message from this paper is that modern manufacturing
processes often use materials and processes at the forefront of
technology and that it may be difficult in the short run to sustain
that technology while implementing major changes in the processes
on which it is based.
A broader material specific analysis is that of Ayres (5), in
which the materials flows of a number of toxic heavy metals are
presented. Among the results of that study is the real-ization
that very large amounts of material must be processed to extract
the metals from their ores and that many uses of the metals are
dissipative. The sustainability of the biosphere over time is
probably incompatible with these current practices.
A product specific analysis is that of McFarland (6), who discusses
DuPont’s analyses of possible replacements for CFCs by more environmentally
benign substances. An important result of this paper is that it
is often difficult to determine which of several possible choices
is most desirable from an environmental standpoint. In other words,
the environmental forcing function can sometimes be rather ambiguous.
Mitchell (7) discusses a technique for eliminating the hazards
involved in the transportation and storage of potentially harmful
chemicals by synthesizing them as a part of the industrial process.
Here the forcing function. both from an industrial hygiene standpoint
and a regulatory standpoint is very clear. Success was possible
only by bringing new and insightful scientific and technological
approaches to bear on a problem with a long history.
Influencing Industrial Ecology Flows
The Fig. 4 diagram includes not only inputs of energy and materials
(from the left) and outputs (to the right), but also the quasi-cyclic
flows within and between individual entities. Two of the papers
that follow deal specifically with energy use. Ross (8) points out
that much of the energy used in industry is concentrated in very
few sectors: aluminum smelting and cement production, for example.
Major reductions in energy will follow from the trend to declining
quantities of materials use, since modern technologies are much
less energy-intensive than the processes they replace. Hoffman (9)
presents a simple and more universal goal: to reduce the energy
used in lighting. Many of the technologies needed for this process
are already in hand; the challenge appears to lie rather in achieving
satisfactory economic incentives.
Stein (10) discusses the level of recycling of polymers today
and the prospects for the future. He envisions today’s 30% recycling
rate increasing to about 60% but sees difficulty in progressing
much further so long as plastics are designed into products in
such a way that their extraction after use is unfeasible or causes
excessive degradation of the basic polymeric material. Thoughtful
initial design of products, utilizing industrial ecology principles,
will be needed. Ways around this problem may come from innovative
work like that of Luzier (11), whose use of biotechnology and
materials science as part of the manufacturing cycle has resulted
in a family of plastic materials that are produced from renewable
resources while being completely biodegradable.
Constraints and Incentives for Industrial Ecology
Industrial ecology cannot be studied and optimized in isolation
from the human institutions of various kinds that promote or constrain
the materials or energy flows indicated in Fig. 4. Consider the
Several of the issues mentioned above are discussed in more depth
in papers by Nordhaus (12). Duchin (13), Henrichs ( 14). and Pariza
( 15). Nordhaus ( 12) discusses industrial ecology in the context
of today’s intricate web of economic activity. He points out that
one could force the economy into the economic equivalent of a closed
ecosystem by ensuring that all environmental costs and benefits
are priced at their full social value.
- (i) Engineering excellence can often promote cyclic behavior
within the manufacturing node by designing processes to promote
- (ii) The desire to avoid toxic wastes may promote process
changes to reduce the quantity of wastes or (better) to substitute
materials or components that result in less toxic or nontoxic
- (iii) The economic system may make it difficult to raise
capital to alter a process and render it more efficient, that
is, to improve its cyclic nature.
- (iv) Taxation may promote raw materials flows or import-export
flows that are contrary to cyclization of the industrial ecosystem.
- (v) Government regulations may make reuse of materials so
difficult that enhanced waste flow is de facto encouraged.
- (vi) The price system, by failing to include relevant externalities
in prices and costs may preclude adoption of industrial ecology
by manufacturers and producers.
- (vii) The standard of living of the consumer may encourage
long product use or, alternatively, may promote early product
- (viii) The rapid rate of technological evolution and obsolescence
contributes to an enhanced waste stream.
Duchin’s paper (13) concentrates on the ways in which economic
factors may be assessed in the industrial ecology perspective.
Appropriate methodologies are vital, since environmental tradeoffs
are certain to arise, just as do economic and technological tradeoffs.
Duchin describes how economic input-output analysis is particularly
suitable to industrial ecology.
Legal constraints to industrial ecology are discussed by Henrichs
(14). She points out that regulations that are based on incentives
rather than on sanctions are those that have proven to be the
most effective. In addition, such legal approaches are far easier
to change as knowledge changes than are those that dictate every
detail of the industrial process they seek to regulate.
Pariza (15) looks at regulatory matters from the point of view
of risk assessment and public perceptions. This perspective is
important for industrial ecology in that processes and products
can be designed to avoid the use of materials with known ecological
consequences but cannot be designed to avoid the use of materials
whose use might be constrained by irrational legislation. In this
connection, a good case can be made for close cooperation among
industry, government, and the public if real environmental progress
is to be achieved.
Educating Industrial Ecologists
No matter how clearly industrial ecology is defined, it cannot be
implemented without a sufficiently numerous cadre of trained personnel.
For the most part, such individuals do not now exist, and the process
of creating them from the potential reservoirs of industrial hygienists,
process engineers, and students is a challenging one. Troxell (16)
argues that in the short term most of the people will come from
industry itself. Universities and professional societies will initially
train these people through short courses, industry visits, and the
like. Industry itself will then need to expand its expertise by
extensive in-house training.
For the longer term, Hutchinson and Lynch (17) present a possible
syllabus for a master’s degree program in industrial ecology.
They model this syllabus after that of a two-year professional
school, and structure it around three primary academic core topics:
environmental science and engineering, management science, and
public policy. Starr (18) directs his comments chiefly to the
latter topic and argues that the crucial item in educating industrial
ecologists is bridging the traditional separation between the
study of technology and of society.
Looking to the Future
To be broadly implemented. industrial ecology will require the setting
of broad goals and of effective ways of reaching those goals. Speth
speaks to that topic in his paper (19), in which he points out that
the solutions to today’s environ-mental problems lie mostly outside
the established "envi-ronmental sector." He calls on the business
and technological communities to move beyond compliance to leadership
and cites the colloquium as a promising step in that direction.
A similar perspective but a somewhat more directed focus is taken
by Piasecki (20), who makes the point that industrial ecology
requires not only fresh approaches to technology but also institutional
innovations in management. He then discusses several ‘’popular
fallacies’’ to demonstrate the many natural alliances that have
been discovered and that will occur in the future among economic
considerations, govern-ment, industry. and the environment.
Finally, Brown (21) contributes a plan of action from the perspective
of one whose career is within the political system. He calls for
a roundtable organization to bring together appropriate personnel
from the industrial, scientific, engineering, and economic sectors
in order to promote activities in industrial ecology and to minimize
constraints to it, intended or accidental.
The colloquium whose discussions are reported in this issue of the
PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES USA demonstrated
that industrial ecology is a concept that is relevant to the modern
industrial community and that is already being implemented in various
ways, at least on a partial and preliminary basis. In his conference
summary, Ausubel (22) reflects that one important accomplishment
is the obvious awareness that many in industry, academia, and government
are now bringing to the field. He cites the potential advantages
to be gained through industrial ecology and notes the great disadvantages
to all of us if the tenets of industrial ecology are not followed.
A simple and useful synopsis of the colloquium is to cite several
characteristics of industrial ecology that, in one way or another,
permeate the papers that follow:
Two hundred years ago, industry changed from a small, labor-intensive,
inobtrusive activity to one that has become large, obtrusive, and
potentially destructive to the resources that support it. Industry
now has the opportunity to take a step as great as was taken in
the industrial revolution of the 18th century: to move from unconstrained
use and disposal of materials to manufacturing approaches that take
products and impacts into account in the same design and with the
same degree of foresight. The twentytwo papers that follow show
us all how to get started.
- (i) Industrial ecology is proactive not reactive. That is,
it is initiated and promoted by industrial concerns because
it is in their own interest and in the interest of those surrounding
systems with which they interact, not because it is imposed
by one or more external factors.
- (ii) Industrial ecology is designed in not added on. This
characteristic recognizes that many aspects of materials flows
are defined by decisions taken very early in the design process
and that optimization of industrial ecology requires every product
and process designer and every manufacturing engineer to view
industrial ecology with the same intensity that is brought to
bear on product quality or manufacturability.
- (iii) Industrial ecology is flexible not rigid. Many aspects
of the process may need to change as new manufacturing processes
become possible, new limitations arise from scientific and ecological
studies, new opportunities arise as markets evolve, and so on.
- (iv) Industrial ecology is encompassing not insular. In the
modern international industrial world, it calls for approaches
that not only cross industrial sectors but cross national and
cultural boundaries as well.
- Patel. C. K. N. (1992) Proc . Natl. Acad. Sci. USA 89, 798-799.
- Frosch. R. A. (1992) Proc. Natl. Acad. Sci. USA 89, 800-X03.
- Smart, B. (1992) Proc. Natl. Acad. Sci. USA 89, 804-806.
- Boyhan. W. S. (1992) Proc. Natl. Acad. Sci. USA 89, 807-811.
- Ayres, R. U. (1992) Proc. Natl. Acad. Sci. USA 89, 812-814.
- McFarland. M. ( 1992) Proc. Natl. Acad. Sci. USA 89, 815-820.
- Mitchell. J. W. (1992) Proc. Natl. Acad. Sci. USA 89, 821-826.
- Ross. M. H. /1997) Proc. Natl. Acad. Sci. USA 89, 827-831.
- Hoffman. J. S. (1997) Proc. Natl. Acad. Sci. USA 89, 832-834.
- Stein. R. S. (1992) Proc. Natl. Acad. Sci. USA 89, 835-838.
Back to Top