EMBODIED ENERGY ANALYSIS AND EMERGY ANALYSIS: A Comparative View
M.T. Brown' and R.A. Herendeenb
"Dept of Environmental Engineering Sciences, Univ. of Florida, Gainesville, FL 32611
Illinois Natural History Survey, Champaign, IL 61820
Similarities and differences between energy analysis and EMERGY analysis are discussed and
highlighted using the two approaches to analyze the same systems. With particular emphasis on
accounting schemes, parallel quantitative analyses of several simple model systems and two
proposed hydroelectric dams in Thailand are performed. For the first time in the open literature
EMERGY accounting procedures are given in detail. The discussion is presented in alternating
sections since the authors still disagree on several fundamental issues.
1.1 Statement of the Problem
This is a dialogue between a practitioner of EMERGY analysis (EMA) and one of energy
analysis (EA) to assess questions of usefulness, comprehensiveness, self-consistency, and
consistency with accepted science. The dialogue began in earnest after the May 1990 meeting of
the International Society for Ecological Economics in Washington, D.C., where H. T. Odum
called for parallel analysis using the two approaches. It has continued through the American
Association for the Advancement of Science meeting in Chicago in January 1992 and the Denver
meeting of the International Society for the Systems Sciences in July 1992.
Our hope is that this paper will exhibit our approaches and viewpoints for scrutiny by
others. We do not agree on many points, but by analyzing the same system we have been able to
crystallize our areas of agreement and disagreement and to present them consistently. The energy
analyst (Herendeen) enters into this discussion because he sees the need for, and admires, the
broad purview of EMA, in which there is a comprehensive scheme to quantify all environmental
services that sustain humans. He acknowledges that the scope of EA is much narrower, but is
concerned that the conceptual details of EMA, which ultimately determine if EMA will be applied
in real decisions, are currently inadequate. The EMERGY analyst (Brown) has agreed to this
discussion as a means of presenting differences and similarities between the two methods with the
hopes that a better understanding of the conceptual basis and quantitative details of the approach
can be realized.
In Section 2 we present definitions and principles of the two methods. Our discussion is
concerned to a major degree with accounting procedures, which we present in Section 3. In
Section 4 we compare the results of the two approaches for specific, simple flow diagrams. In
Section 5 we compare results of applying the approaches to a real example, two hydroelectric
dams in Thailand. In Section 6 we review agreements and disagreements, in Section 7 we
summarize, and in Section 8 we present a dialogue table as a summary of the differences and
similarities. Sections representing only one author's viewpoint are identified as such. We suggest
that readers already familiar with the debate first look at the dialogue table before reading the text.
2. DEFINITION OF TERMS AND PRINCIPLES
2.1 Energy Analysis (EA) and Embodied Energy (Herendeen)
Energy analysis is the process of determining the energy required directly and indirectly to
allow a system (usually an economic system) to produce a specified good or service (IFIAS,
1974). The bookkeeping used in this determination can be used for embodied anything-for
example, copper, SO, (Leontief 1970, 1973), nitrogen (Herendeen, 1990), or labor (Bezdek and
Hannon, 1974). The basic motivation for energy analysis is to quantify the connection between
human activities and the demand for this important resource. The implication is that energy is
more important than conventional economic reckoning indicates. Of course the question of how
much more is arguable. In the 1970s the oil embargo brought energy to economic center stage,
and many environmental analysts were satisfied to treat energy use as a first-order indicator of
overall environmental impact. In the 1980s the world oil price dropped and environmental
analysts wanted a more detailed accounting of environmental impacts. In the 1990s the
greenhouse implications of fossil fuel burning have again promoted energy's use as an
Energy analysis explicitly and rigorously calculates indirect effects. For example, about
86% of the energy required to produce an automobile is burned in industries outside the auto
assembly plant (Bullard, Hannon, and Herendeen, 1975). The bookkeeping to account for
indirect flows has strong similarities to that in Input-Output (1-0) economics (Bullard and
Herendeen, 1975), and some of the machinery of that economic technique has often been used in
energy analysis for ecological systems as well as economic ones. (Hannon, 1973; Finn, 1976,
1980; Patten, Higashi, and Burns, 1990). Indirect effects are especially important in the question
of net energy: how the energy produced by an energy technology compares with the energy
required to produce its inputs (Chapman, 1975; Chambers, et al, 1979; Herendeen, Kary, and
Rebitzer, 1979: Herendeen, 1988).
Energy analysis can include renewable energy sources; attentive bookkeeping is required
to keep them separate from non-renewable sources. While energy analysis is based on the notion
that energy is more important than most people think, it typically is not used to support an
"energy theory of value." The more moderate view is that energy analysis is one information
input, like economics, to the process of making a decision (Herendeen, 1988).
There are two things EA does not do:
1. EA does not have an optimizing principle.
2. While direct and indirect pollution releases can be calculated using the EA framework,
EA does not quantify the environment's role in absorbing and processing pollution.
2.2. EMERGY. Transformity. and Maximum EMERGY (Brown)
EMERGY analysis (EMA) is a technique of quantitative analysis which determines the
values of non-monied and monied resources, services and commodities in common units of the
solar energy it took to make them (called Solar EMERGY). The technique is based on the
principles of energetic (Lotka, 1922, 1925, 1945), system theory (von Bertalanffy, 1968) and
systems ecology (Odum, 1975, 1983, 1988, 1991). One of its fundamental organizing principles
is the maximum EMERGY principle. Stated as simply as possible the maximum EMERGY
principle is as follows:
Maximum EMERGY Principle: Systems that will prevail in competition with
others, develop the most useful work with inflowing EMERGY sources by
reinforcing productive processes and overcoming limitations through
It is important that the term "useful" is used here. Useful work is related to using
inflowing EMERGY in reinforcement actions that ensure and, if possible, increase the inflowing
EMERGY. Energy dissipation without useful contribution to increasing inflowing EMERGY is
not reinforcing, and thus cannot compete with systems that use inflowing EMERGY in self-
reinforcing ways. For example, drilling oil wells and then burning off the oil may use oil faster (in
the short run) than refining and using it to run machines, but it will not compete, in the long run,
with a system that uses oil to develop and run machines that increase drilling capacity and
ultimately the rate at which oil can be supplied.
The maximum EMERGY principle suggests a system of value that is donor based rather
than receiver based. By this we mean that the value of something is derived from how much goes
into it rather than how much one is willing to pay for it. The line of reasoning is that since sys-
tems are organized to maximize power, (using their inflowing energies in ways that reinforce pro-
ductive processes) any expenditure of energy has to return useful work equivalent to at least what
was expended. We believe that this holds for all systems over all time and spatial scales. Yet, we
recognize that at any one moment in time, one might observe an expenditure of energy that does
not, in any way, appear to maximize power. And while such a circumstance may seem a violation,
it is apparent to us, that because something at a moment in time appears to violate a rule, it does
not mean the rule is not operating, only that the rule hasn't caught up with the violation.
The EMERGY of renewable energies, nonrenewable resources, goods, services and even
information are determined by the energy required to make them. When values are expressed in
these terms, we call the new measure EMERGY and define it as the amount of one type of energy
that it takes to make another. When expressed as the amount of solar energy that was used, the
units of EMERGY are solar EMERGY, the units of which are solar a.mjoules (abbreviated sej).
To derive solar EMERGY of a resource or commodity, it is necessary to trace back
through all the resources and energy that are used to produce it and express them in the amount
of solar energy that went into their production. This has been done for a wide variety of resources
and commodities and the renewable energies driving the biogeochemical process of the earth.
When expressed as a ratio of the total EMERGY used to the energy produced, a transformity
results (dimensions are sej/J). As its name implies, the transformity can be used to "transform" a
given energy into EMERGY, by multiplying the energy by the transformity. For convenience, in
order not to have to calculate the EMERGY in resources and commodities every time a process is
evaluated, we use transformities that have been previously calculated.
This use of transformities calculated at other times and for processes that may be in
another part of the globe is bothersome to some. In fact, we know full well that there is no single
transformity for most commodities, but a range of transformities. There is probably a lower limit,
below which a commodity cannot be made, and there is some upper limit, although in theory, one
could make corn, for instance, with an infinite amount of wasted fuel and thus have an infinitely
high transformity. For some commodities we have calculated several transformities in different
parts of the globe and for different ways of making them. Shrimp is an example. In the Gulf of
California, shrimp caught by the mechanized Mexican shrimping fleet have a transformity of 13.0
E6 sej/J, while those caught with artisianal fishing techniques have a transformity of 4.0 E6 sej/J
(Brown et al 1988). Shrimp grown in shrimp ponds in Ecuador have still another transformity
equal to 18.9 E6 sej/J (Odum and Arding 1991). Average transformities are used whenever the
exact origin of a resource or commodity is not known.
Uncertainties surrounding transformities are related to the resource in question. For
instance, transformities for renewable energies like wind, rain, tides and so forth were calculated
using global inputs from sunlight, deep heat, and tidal momentum. The best estimates (from the
literature) of the total energy in global winds, total amount of global rainfall, and global tidal
energy, were divided by the total EMERGY in sunlight, deep heat and tidal momentum to obtain
their transformities. These renewable energy transformities are the basis for most other
transformities, since all "higher order" processes and commodities include some proportion of
renewable energy. The uncertainty then becomes relative, since if the transformities for renewable
energies are too high or too low, then the EMERGY in higher order products are off as well, but
by equal amounts. (End Brown)
3. ACCOUNTING RULES AND PROCEDURES
3.1. Energy Analysis Accounting Procedures (Herendeen)
The accounting framework for EA has been in the literature for 20+ years (IFIAS, 1974;
Bullard and Herendeen, 1975; Bullard, Penner, and Pilati, 1978). This framework is subject to
inherent, inevitable difficulties which must be dealt with explicitly (Herendeen, 1988). Examples
are questions of system boundary, how to merge several kinds of energy, and energy credit for
byproducts. It should be stressed that these problems do not result from confusion or lack of
study. On the contrary, they are fundamental issues which research has shown can nly be
resolved byjudgmental decision.
To determine the energy to produce a product, the most direct approach would be to
perform a detailed "vertical analysis" covering the manufacturer, its suppliers, their suppliers, and
so on. At each stage one tallies the energy inputs per unit of output, and then the inputs of
everything else. One crucial assumption is that the measured quantities (say tons, liters, or even
dollars) are adequate carriers, or numeraires, for embodied energy. This process spreads out
dendritically (see Figure 1) and can even turn back to the beginning (steel is an input to cars, and
cars are an input to steel), thus leading to an infinite series, but the calculation converges
mathematically. Often the process is truncated after just a few steps with negligible loss in
accuracy. A classic example of a vertical analysis is Berry and Fels' (1973) study of automobile
Vertical analyses are expensive. To save money, analysts were drawn to the Input-Output
economic technique, which organizes large amounts of economic flow data between economic
sectors to allow calculation of monetary indirect effects. Such data bases (e.g., US Department of
Commerce, 1984) cover all economic sectors and are checked and adjusted for self-consistency.
The attraction of a complete flow table for ca. 350 sectors covering the US economy is strong,
and much energy analysis has been based on using these dollar flows to calculate energy
intensities (Bullard and Herendeen, 1975). The approach has also been used for foreign
economies (Denton, 1975; Herendeen, 1978; Peet, 1986). Use of the consumer expenditures
portion of the data base in conjunction with the energy intensities has yielded the energy impacts
of specific consumer market baskets (Herendeen, 1978; Herendeen, Ford, and Hannon, 1981).
Drawbacks of this data base are:
1. it is typically 5-7 years old at the least,
2. dollars may not be appropriate for conveying the linkages that embodied energy implies
(instead of tons or liters),
3. even with 350 sectors, a sector may not be disaggregated enough for the purpose at
hand. For example, the (aggregated) sector "automobiles and parts" is potentially
not detailed enough to compare Fords and Cadillacs.
The machinery of EA and I-O analysis solves n simultaneous linear equations. One way to
do this is to invert a matrix of coefficients, and that matrix inverse can be written as a converging
infinite series of matrix products, each progressively representing a more indirect process. This
infinite series corresponds exactly to the implied infinite step process in a vertical analysis; the two
methods are equivalent for identical technologies.
The easiest way to see how EA calculates energy intensities is by use of a diagram (Figure
2). We start with compartment (or sector) "j", with inputs of goods and services Y, output X,
and actual energy input Ej. All are flows, measured per unit time. We now assume that all flows
carry an embodied energy given by ei, which defines the energy intensity of sector i as e,. ei is
measured in units of E (say kcal d-') divided by units of X (say $ d-') = (therefore) kcal S-'. The
fundamental assumption of EA is that sector j is in embodied energy balance. Figure 2 thus
represents a balance equation for the conservation of embodied energy:
I E rl'yEj = Ej (
If there are n sectors, there are n simultaneous balance equations for the n energy
intensities, which can be expressed concisely in matrix form:
where E is a vector of energy inputs, j. a vector of energy intensities, X a matrix of the
flows Xij, and X a diagonal matrix of the outputs Xji.
To reemphasize, the balance diagram in Figure 2 and the resulting equations are applicable
not only to energy, but any input-materials, labor, etc. In that case Ej is replaced with the flow of
the desired input, the X, are chosen to be acceptable carriers for the embodied input, and the e4
are replaced with the intensities for that input.
The balance of embodied energy is an assumption. I argue that this assumption captures
the intent of calculating indirect effects: to allocate something normally not accounted for, which
may (e.g., energy) or may not (e.g., copper) be dissipated, to the products ultimately produced.
This balance has nothing fundamental to do with any thermodynamic law, as it is assumed when
calculating non-energy intensities such as labor (Bezdek and Hannon, 1974) and nutrient
(Herendeen, 1990). I use embodied energy/labor/nutrient analysis as the core of an accounting
scheme designed to keep track of and account for something otherwise lost from scrutiny.
The sector-by-sector conservation of embodied energy leads to overall conservation for
the entire system, which is desirable from an aggregation criterion. See Figure 3.
Mathematically, this overall balance is expressed as
Z Ei = E EiYi (3)
i= 1 i= 1
X = X.. + Yi (4)
Yi being the compartment's flow to exports/final consumption. Equation 3 results from
substituting Equation 4 into Equation 1 (Bullard and Herendeen, 1975).
An unavoidable consequence of this embodied energy balance is that internal embodied
energy flows can exceed the actual real energy inputs to the system when there is feedback. At
first this seems impossible. A potential disclaimer is that embodied energy is not actual free
energy, so that this flow which seems too large should not be of concern. However, this is not
the fundamental reason. In fact this internal flow "problem" is a logical consequence of feedback
flows, and applies whether or not one uses energy as numeraire. For example, one could verify it
by measurement for the case of nutrient intensity in a system that "leaks" no nutrient, as discussed
Consider Figure 4a. In this system, biomass energy is the numeraire, there is feedback,
and nutrient enters compartment A. Nutrient does not leak or dissipate. Therefore embodied
nutrient flows = actual nutrient flows, and all entering nutrient is embodied in the output. The
resulting nutrient intensities are rlT = 10 g kcal"' and TI, = 100 g kcal'". As shown in Figure 4b,
the internal embodied nutrient flows A->B and B-->A do exceed the system input of 100 g d'V.
The apparently too-large flow is physically possible because feedback occasions and requires that
molecules of nutrient passing through A come from both downstream and upstream. The nutrient
intensity is increased by the feedback flow of high nutrient-intensity from B->A. If one measures
the nutrient flux (g d"') one will find that much nutrient actually does flow. Because this flux is
measurable for a non-dissipated input, I argue that it is permissible, in fact desirable, to use the
same accounting scheme for the embodied flows of a dissipated input.
Following from I-O analysis, there are a number of conventions and manipulations that
can be applied in EA to more complicated flows than those in Figs 3 and 4. Ifa sector has
multiple outputs but there are still n total products for n sectors, then there are at least two
manipulations which reduce the problem to the form in Figure 3 and 4. If there are more
commodities than sectors, there is a manipulation ("market shares" assumption) which again
reduces the problem to the form in Figs 3 and 4. If there are byproducts, EA, following I-O,
manipulates to assume them away. There is no way to maintain conservation if the product and
byproducts are independently assigned the total input to produce product and byproduct together.
One must choose one convention or the other. EA chooses to maintain conservation of embodied
energy, believing that to be more useful than preserving the byproduct option.
Thus in Figure 5 EA can manipulate outputs in at least three ways:
1. EA could sum the output into an aggregated commodity "2 lumber + 1 sawdust" (of
which there are 10 units) with an energy intensity = 2100/10 = 210 J/aggregated unit (Fig,
2. EA could sum that the output as if sawdust and lumber are identical (of which there are
30 units) with an energy intensity of 2100/30 = 70 J/disaggregated unit (Figure 5c), or
3. EA could assume the products are made in two parallel processes and assign input
energy based on any scheme that assures that the total embodied energy out is 2100 J.
3.2. EMERGY Accounting Procedures (EMERGY Algebra) (Brown)
EMERGY is the amount of a source energy it takes to make another form of energy.
There are definite rules that are followed to assign EMERGY to flows of energy. We have
termed the sum total of these rules, EMERGY Algebra (Scienceman, 1987).
The first rule is:
ALL SOURCE ENERGY TO A PROCESS IS ASSIGNED TO THE
PROCESSES' OUTPUT (S).
Figure 6 shows a process that has several sources of EMERGY. In Figure 6a, the
pathways are evaluated in heat energy, and all energy is accounted for as either output, or
dissipated energy. In Figure 6b, EMERGY is assigned based on rule 1. The outflow pathway has
a total EMERGY of 1000 sej. Figure 6c shows that the transformity of the output is 100 sej/J.
The second rule concerns processes with more than one output (byproducts)
BYPRODUCTS FROM PROCESS HA VE THE TOTAL EMERGY
ASSIGNED TO EACH PATHWAY.
The third rule relates to an output from a process that is divided into two separate flows (a
split of EMERGY):
WHEN PATHWAYSPLITS, THE EMERGYIS ASSIGNED TO EACH
"LEG" OF THE SPLIT BASED ON ITS PERCENT OF THE TOTAL
ENERGY FLOW ON THE PATHWA Y.
Figure 7 shows the assignment of EMERGY to byproducts (A and B) from the same
process and assignment of EMERGY to a split (Al and A2). The pathways are evaluated in heat
energy in Figure 7a, then total EMERGY is assigned to the outputs equally in Figure 7b. Notice
that the same EMERGY is assigned to both outputs, A and B, and then A is split based on the
percentage of total energy that is on each of the two split pathways. Transformities are calculated
in Figure 7c by dividing the EMERGY on each pathway by the energy.
The difference between byproducts and splits of the same output is important. Many
processes produce more than one output (for instance, agriculture that produces ears of corn and
corn stalks, a saw mill that produces lumber and sawdust, or a manufacturing process that
produces a good and one or more "waste" byproducts). Since a process that produces two or
more outputs cannot produce one without producing the other, the total EMERGY input is
assigned to each output. Each is required, and each requires the total input of EMERGY for its
The fourth rule describes how EMERGY is assigned within systems of interconnected
EMERGY CANNOT BE COUNTED TWICE WITHIN SYSTEM
a) EMERGYin feedbacks cannot be double counted
b) byproducts, when reunited, cannot be added to equal a sum greater
than the source EMERGYfrom which they were derived.
Figure 8 is a simple system of two components having two energy sources and a
"feedback" from component B to component A. In practice it is often easiest to assign EMERGY
by writing the EMERGY assigned to each pathway as the sum of inputs from different sources.
Beginning on the left the output from A is the sum of 400 sej from source S and 3/5ths of source
F, or 60sej, for a total EMERGY of 460 sej. Notice that only that portion of the feedback from B
that did not come from source S, through A, is counted in the output of A. The 400 sej coming
originally from A cannot be counted a second time.
Figure 9 illustrates a second consequence of the fourth rule. The system illustrated has
two parallel pathways that are "reunited" at component D. To simplify the illustration, it is drawn
with only byproducts (no splits), and only one feedback from D to C. Beginning on the left, the
EMERGY assigned to both outputs of A is the total of source S (400 sej). The output from B is
500 sej, or the sum of the EMERGY coming from A and source F. The output from C is the sum
ofEMERGY coming from A and the EMERGY of source F coming through components B and
then D. Since EMERGY cannot be counted twice within the same system, the input to
component D can only be a total of 500 sej. The EMERGY from source S that comes through
components B and C cannot be counted twice when these two pathways are reunited at D.
4. COMPARISON OF EMERGY AND EMBODIED ENERGY ACCOUNTING
4.1 Brown comments
Figure 10 is a systems diagram having 5 components (A E) and two energy sources (S
and F). The flows are also given in Table 1. Components of the system are organized and
interconnected with pathways of energy flow, where the numbers on each pathway represent
hypothetical yearly energy fluxes. The energy sources are of different "forms", the one to the left
(S) is a renewable flow-limited energy source such as the sun. Its inflow to the system is "split"
between components A and B, 3/10ths to A and 7/10ths to B. Components of the system are
arranged hierarchically and according to energy flow from left to right. Using a food chain
analogy, lower, more abundant members of the chain (green plants) are to the left, while higher
and higher trophic levels are to the right. The energy source from the top (F) is a nonrenewable
source of higher "quality" than S.
Figure 10 illustrates the fundamental reason that recognition of the differences in form
energy is necessary. From a thermodynamic perspective the system is correct, having all energy
accounted for because the inflows equal the outflows. However, when evaluated in heat energy
(as the diagram is) energy flows to the right are so small as to be insignificant when compared
with the flows farther to the left. Yet it is apparent, from a systems perspective that the processes
and flows to the right cannot exist without the inputs, nor since there are feedbacks, can the
processes to the left exist without those to the right. In other words since the system is
interconnected all components and flows are necessary, yet when evaluated in their heat value,
many flows (especially those at the top of energy hierarchies) seem insignificant and of little
importance. EMERGY evaluations make the assumptions that they are essential for the entire
system, their value is the total EMERGY that contributes to them.
The flows from each component can be expressed in the amount of form energy of type S
it takes to support it, deriving an equivalent basis for relative comparison. The only thing that is
needed is the expression of energy source F in the equivalent form energy of S.
Figure 1 la shows the amount of solar EMERGY assigned to each pathway using the four
conventions (rules) described in the previous section. Figure 1 lb shows the embodied energy
assigned to each pathway using the conventions of embodied energy accounting and matrix
inversion in Equation 2.
Differences in the magnitudes of EMERGY and embodied energy assigned to each
pathway result from the different techniques and rules. Quantitatively, there are some significant
differences between the diagrams evaluated in EMERGY and embodied energy. One of the first
differences is that the outputs (Y and Z) total 37500 sej for the EMERGY diagram (Figure 10)
and 30000 Joules, for the embodied energy diagram (Fig 1 lb). When the evaluated pathways (Y
and Z) are added together, their sum in the EMERGY diagram (37500 sej) is greater than the sum
of the inputs, an apparent problem and the basis for much confusion and misguided criticism since
it appears EMERGY is not conserved1. In the embodied energy diagram the sum of the outputs
equals the sum of the inputs.
Using the rules of EMERGY algebra, it is impossible for any single pathway to have more
EMERGY assigned to it than is inflowing to the system. Using the matrix inversion techniques of
embodied energy, however, pathways often have more embodied energy assigned than is
inflowing (pathways from C to A, and from A to D), again providing fodder for much confusion
and misguided criticism. We will discuss these apparent problems in a later section.
The most important difference that results when the two accounting procedures are used,
is that on pathways from lower order components the embodied energy is about 1.8 times as
much as the EMERGY, but on the highest order pathways, the EMERGY is from 1.1 to 2.0
times as great as the embodied energy. The significance is that the differences between lower and
higher order pathways are amplified between the two accounting systems. Embodied energy
accounting gives more weight (more relative importance) to lower order pathways over higher
order ones, while EMERGY accounting gives more relative importance to higher order pathways.
Figure 12 shows the solar transformities and embodied energy intensities that result when
the energy on pathways (Figure 10) is divided into EMERGY (Figure 11a), and embodied energy
The definition of conservation being used in those criticisms is the sum of inputs equals the outputs. The law of
conservation of energy can be stated as follows: "The total energy is neither increased or decreased in any process.
Energy can be transformed from one form to another and transferred from one body to another, but the total amount
remains constant" (General Physics. D.C. Giancoli, 1984). It should be kept in mind that EMERGY isnt energy, but
an accounting of the energy used previously (energy memory) and thus does not behave like energy...its algebra is
different, reflecting this fact
(Figure 1 lb). The units of solar transformities are sej/J. The units of intensities are Joules of
embodied energy per Joule. The relative importance of lower and higher order pathways,
discussed above, is more easily seen when transformities and intensities are compared. This can
be seen in Figures 12a and 12b, where the relative values oftransformities and intensities for
lower order outputs (component A and B) and higher order outputs (components D and E) are
significantly larger in the EMERGY evaluated diagram (Figure 1 la) than in the embodied energy
accounting (Figure 1 b). Transformities for higher order pathways are from 13 to 16 times as
great as lower order pathways, while intensities for higher order pathways are from 3 to 10 times
as great as lower order pathways.
The significance of these differences is in interpretation of what transformities and
intensities mean. If as theory suggests, transformities are a relative measure of flexibility,
substitution, or value (the energy required), and if embodied energy intensities are analogous to
transformities, then relative importance can be significant. In EMERGY terms, the output of
component E has a transformity that is 13 times as great as the output from component A (Figure
12a), while the embodied energy intensity is only 3.5 times as great for the same outputs (Figure
12b). In like manner the transformity for the output from component D is 16 times that of
component B, while the intensity factor is only 10 times as great.
Differences in relative importance of outputs is also revealed from the fact that each
system assigns highest intensity and transformity to different outputs. The largest intensity is
associated with the output from C (5000 embodied energy joules per Joule), while the largest
transformity is the output from E (7500 sej/J). In other words, embodied energy accounting
assigns highest relative importance (highest intensity) to an intermediate component and its
output, while EMERGY analysis assigns greatest importance to the highest order component (top
of the energy hierarchy).
The Conservation Controversy
As has already been discussed, when a process results in the output of two different
products (i.e. byproducts) the entire input EMERGY is assigned to both outputs, since each
cannot be made without the other and all EMERGY is required to make each. This fact creates
much confusion since at first glance it appears that more EMERGY is output from the process
than is input, and thus a violation of the Conservation Law of Thermodynamics. However, under
no circumstances should the EMERGY outputs from a process be added together. It is a
violation of rule four and results in double counting the EMERGY. Additionally, at times, all
EMERGY inflows to a process may not all be assigned to its outflows (see for instance Figure 9b
where the inflows to component C total 900 sej [if they were added together], while the output is
500 scj). Since EMERGY cannot be counted twice in the same system, if through feedback
actions EMERGY returns to a component, it cannot be added into its output again.
EMERGY has been referred to as energy memory (Scienceman 1989). This is often a
convenient way of visualizing EMERGY. When a system is evaluated in solar EMERGY the
quantities represented are the "memory" of the solar energy used to make it. As a result, the
quantities are not energy and do not behave like energy. (End Brown)
4.2. Herendeen comments
Equation 2 is applied to the flows in Figure 10 (also given in Table 1) to give the energy
intensities in Figure 12b and the embodied energy flows in Figure 1 lb. The energy intensities in
Figure 12 range from 258 to 5000 sej/j, a range of 19.4. Feedback is the reason that the spread in
values is not greater, feedback has the effect of mixing what otherwise would be "high" and "low"
compartments and destroying the otherwise more hierarchical structure. This is not a defect of
EA; it is a strong point. EA explicitly accounts for feedback loops. (Other indications of the
strength of feedback can be had by calculating trophic position and path length (Ulanowicz, 1986;
Wulff and Ulanowicz, 1989; Herendeen, 1990). Similar mixing of "high" and "low" levels occurs,
but I do not illustrate here). Figure lb shows that each compartment is in embodied energy
balance, as is the overall system. (End Herendeen)
5. PARALLEL ANALYSES OF TWO HYDROELECTRIC DAMS IN THAILAND
5.1. EMERGY Analysis of Hydroelectric Dams in Thailand (Brown)
The Mekong River forms the northern boundary between Thailand and Laos until it flows
eastward through Cambodia. As part of the UN-sponsored initiative to develop the hydroelectric
potential throughout the Mekong Basin, proposals for dams along the main reaches in the upper
basin have been made. Among the first of these proposals involved two sites in northern Thailand
known as the Upper Chiang Khan and Lower Pa Mong dams. Numerous studies evaluating both
sites have been conducted over the past decade as the governments involved have tried to
reconcile costs and benefits of the two locations. An EMERGY analysis of both dams was
conducted to lend additional insight and to provide a practical demonstration of the EMERGY
analysis technique. To complete the analysis of the proposed dams, it was necessary to evaluate
the EMERGY in concrete and the EMERGY value of rice. These analyses are not given here but
can be found in the original report (Brown and McClanahan, 1992).
An overview diagram of the proposed dam on the Mekong (Figure 13) shows that the
main loss from the proposed dam is the loss of area for agricultural production and the
displacement of rural households. The primary benefits are irrigation and generated electricity
and for use in urban and rural households and manufacturing.
EMERGY analyses of both dams (Tables 2 and 3) includes the potential dam benefits
(electricity, aquatic productivity, and irrigation supporting farm production) and losses, (direct
costs of dam and irrigation system construction, losses of agricultural productivity, and losses
associated with human population resettlement). The dam is assumed to have a 50-year life span,
thus construction costs were divided by 50 to present data on a yearly basis. The analysis
indicates that electricity production, by far, is the major EMERGY benefit of dam construction.
Irrigation and, to a much lesser extent, aquatic productivity are relatively unimportant. The
analysis assumes that irrigation has the effect of doubling the annual yield of crops through dry
season irrigation. While irrigation has a very high return on investment its inclusion in the
development project appears to be a relatively unimportant effect to the net EMERGY of the
project, since the EMERGY value of electricity produced is more than an order of magnitude
greater than the expected agricultural production. The most significant costs associated with the
dam construction are resettlement followed by direct use of goods and services. The lost
agricultural production is significant but is more than made up for with increased production
resulting from irrigation of other lands. Overall, both dams have positive EMERGY yield ratios.
The Upper Chiang Khan option has a better ratio (20.3/1) than the Low Pa Mong option (12.3/1).
While electrical production is relatively similar between the two sites, costs at the Upper Chiang
Khan site are proportionately lower.
Two EMERGY yield ratios are given in Tables 2 and 3. The first ratio in each table does
not include the loss of sediments. The second ratio includes sediments as a cost which
substantially reduces the overall benefit. While it can be argued that sediments trapped behind the
dams are not lost, their contribution to downstream productivity is temporarily disrupted. The
loss of wetland, agricultural, and fishery productivity (to name but a few) is indirectly accounted
for through the value of sediments, since the EMERGY of sediments is one of the main inputs to
these subsystems. The inclusion of sediments in the EMERGY yield ratios reduces the Low Pa
Mong ratio to 1.39/1 and that of the Upper Chiang Khan to 1.34/1.
The EMERGY analysis of these dam proposals illustrates the ability to evaluate not only
direct energy used (termed cultural energy by Herendeen), but environmental losses as well. The
inclusion of sediments in the evaluations significantly alters the net effect of the dam proposals.
Yield ratios not including sediments suggest that both dams would have a positive influence on
the Thailand economy. However, with the inclusion of sediments, the ratios decrease to levels
that should influence policy makers to give second thoughts to constructing the dams. The net
effect may be to stimulate the economy through higher electric production, but seriously
undermine its environmental support base. The energies of Thailand's economy could be better
invested in alternative proposals that yield higher net EMERGY. (End Brown)
5.2. Embodied Energy Analysis of Hydroelectric Dams in Thailand (Herendeen)
This analysis is very approximate and is done for illustration only. EA is considerably less
ambitious than EMA. It usually, but not always, confines itself to cultural energy and typically
excludes consideration of many of the environmental impacts that EMA includes. The inputs
which EA evaluates are converted to energy using energy intensities from the US economy
(Casler, 1994) because I do not have them for the specific economies that produced the inputs to
these projects, say Japan, Russia, etc. EA does not evaluate the energy impacts of river
geopotential or sedimentation. Lost agricultural production is energy-costed by assuming that the
production is accomplished elsewhere with American technology. The difference between the
(cultural) energy for the alternative production and the (assumed zero) cultural energy cost of the
original is counted as an energy input to the project. This procedure likely overestimates this
energy input, but there is no point to seeking the "correct" way to handle this here; it depends on
the boundaries to the question, which are arrived at by conscious judgmental decision. For the
case of inputs measured in physical instead of monetary terms, energy intensities in BTU/$ are
converted to BTU/unit using the price of the product.
Because of many approximations the energy inputs are very uncertain, easily varying
+100% or -50%. The energy intensities are the sum of fossil + nuclear + (the fossil equivalent of)
hydroelectric energy, as detailed in Bullard and Herendeen (1975). The intensities and the
expenditures are for different years; corrections are required for overall energy/GNP changes for
the producing economy and for monetary inflation in the specific sector involved. A technique to
make these corrections is given in Bullard and Herendeen (1975). All quantities in Tables 4 and 5
are on a per-year basis and hence the energy obtained can be used directly to calculate the 50-year
lifetime energy ratio.
The energy output is converted using 1 kwh = 3413 Btu = 3.60 x 106 Joules, which is the
direct conversion of electricity to heat and does not account for the energy requirements of a
fossil plant to produce that electricity. Thus the incremental energy ratio (IER), which is defined
as (energy out)/(energy in), actually contains different approaches between numerator and
denominator. (This is an another example of the need for judgmental decisions to "solve" such
problems.) If desired, one could multiply the numerator by a factor of approximately 3-4 to
correct for this. Tables 4 and 5 show that, even without that correction, the two dams have
energy ratios of 19 (Low Pa Mong) and 40 (Upper Chiang Khan). Both dams are thus net energy
producers, and the uncertainties are unlikely to be large enough to change that conclusion. For
each dam, more than half the energy input is from the assumed energy cost of resettlement and
alternative agricultural production; if these were not counted the IER would be over twice as
large. The IERs obtained from EA exceed the EMERGY benefit/cost ratio from EMA, even
when the lost-sediment energy cost is not considered. In general one expects EMA to count more
costs than EA, which would tend to decrease an energy ratio.
EA would stress using these results in context of other energy analyses and other types of
analysis. For example, while the dams seem to be net energy producers, other energy
technologies may have higher or lower energy ratios. Pilati (1977) found energy ratios of 2-7 for
a number of coal, natural gas, and nuclear electricity generators. This was for mine to busbar and
neglected the fuel (otherwise energy ratio would be <1), and is hence directly comparable with the
dam results here. We see that these hydroelectric facilities have higher IERs than the fossil plants.
At least one proposed electricity facility has shown an IER close to breakeven; this is the solar
power satellite (Herendeen, Kary, and Rebitzer, 1979) (End Herendeen)
6. AGREEMENTS AND DIFFERENCES
6.1 Assigning Embodied Energy Based on Dollars. Materials. or Energy (Brown) In
practice, embodied energy can be assigned to products based on money flows, material flows, or
energy flows. Odum (1994) has shown that very different results are obtained in a system when
each of these are used to assign embodied energy to the same pathways. The practice of using
commodities and money to assign embodied energy is a questionable one, since there is no logical
reason that embodied energy is related to money, carbon, labor, or anything else.
6.2.1 Accounting for Labor and Renewable Energy in EMERGY Analysis
Most embodied energy evaluations researched in the literature do not include labor, or if it
is included, only a portion of the energy of a human is considered as an input to the process
(Costanza and Herendeen, 1984; Hall, Cleveland, and Kaufmann, 1986). In general, there is
much debate over whether or not labor should be included, and if included, how to account for it.
In EMERGY analysis, all labor (service) is accounted for using the dollar costs of
products, since dollars spent in the economy always purchase services. A ratio of EMERGY to
dollars is calculated from year to year for an economy and used to determine how much
EMERGY is "behind" any expenditure for services. All dollars spent for a commodity, service, or
fuel, when transformed to EMERGY, are necessary inputs to the process. They are accounted for
as being used to produce it, even if the EMERGY may be partially used by labor for activities not
directly related to production of the commodity purchased. Accounting for labor is a question of
whether the outputs of human production are considered byproducts or splits. Figure 14 shows
the outputs both ways, first as a split where each output is a percentage of the total output (Fig
14a) and as byproducts, where each of the outputs are considered necessary and equal in
importance, thus all EMERGY is assigned to each output (Figure 14b). In EMERGY accounting
the outputs of human production are considered byproducts and thus the EMERGY value of
labor is the sum of all EMERGIES used in support of humans.
Renewable energy inputs to processes are evaluated in EMERGY units (solar emjoules)
by using previously calculated transformities based on the distribution of solar energy in the
biosphere, the output energy of the various processes (rainfall, total wind energy, total wave
energy, etc) and the assumption that these processes are byproducts of the biosphere.
In practice, both renewable and nonrenewable energies are included in EMERGY
evaluations (as illustrated in the previous example of the Mekong dam proposals). This is
extremely important, since an economy does not run on fossil fuels alone, but requires an
environmental support base. Energy analysis on the other hand, routinely leaves out the
renewable energy contributions to economies (as was done in the previous example) including
only, what Herendeen terms, "cultural energy" sometimes accounting for differences in quality
(Cleveland, 1992). Such omissions can leave out significant portions of economies (in developing
nations over half their resource base often comes from nonrenewable sources) leading to inflated
Incremental Energy Ratios and underestimating environmental impacts and net effects of
development proposals. (End Brown)
6.2.2 Accounting for labor and renewable energy in Energy Analysis (Herendeen)
EA is a conserving accounting scheme and the framework can apply to any input,
including renewable resources. Indeed it is routinely applied to solar energy intensities of
different compartments in ecological food chains and webs (Hannon, 1973; Hannon and Joiris,
1989; Herendeen 1990). The input can be labor as well (Bezdek and Hannon, 1974; Hannon, et
al, 1975; Herendeen and Sebald, 1975). There is a side issue about labor: whether consumers are
free to alter their purchasing patterns, i.e., whether the economic system is closed or open with
respect to personal consumption expenditures (See Costanza and Herendeen, 1984), but that is
peripheral to this discussion. (End Herendeen)
6.3.1 Disaggregation. Byproducts and Splits (Herendeen)
There is an aggregation problem which EA handles as follows. Suppose we have an
aggregated system with one energy input and one observed commodity output. The energy
intensity of the output is (energy flow in)/(output flow). If we later learn that this system actually
has the potential to produce two outputs, then revealing the internal flows should tell us the
energy intensity of the second product, but not change the intensity of the first. This is different
from the situation in Figure 5, where we know already that the system can produce more than one
output, and for which we now require internal details to decide how to allocate inputs to those
EA's position is illustrated by Figure 15. In Figure 15a, 1000 kcal d"' enter, and I kcal d'1
of product B exit. Therefore e6 = 1000 kcal/kcal. In Figure 15b and 15c, additional structure is
revealed and we see that the system can produce another product A. In Figure 15b there is no
feedback, and EA, using Equation 2, gives EA = 10 kcal/kcal and 6e = 1000 kcal/kcal. In Figure
15c there is feedback and EA gives eA = 100 kcal/kcal and EB = 1000 kcal/kcal. This shows that
the intensity of the revealed new product (A) depends on the details of the revealed internal
structure, but that the intensity of the sole product known in the aggregated case (B) is unaffected
by the revealed internal details. Disaggregation gives us new results for things we learn about
only by disaggregating, but does not require us to change what is known about the aggregated
system. (End Herendeen)
6.3.2 Disaggregation. Byproducts and Splits (Brown)
In EMERGY analysis as in all investigations into the nature of things it is essential that the
system be diagrammed and understood as completely as possible prior to evaluation. For
comparative purposes Figure 16 shows the same simple aggregated system and the two
disaggregations used in the previous discussion (Figure 15). In the top diagram, all that is
"known" is that the process has a single output. The total inflowing EMERGY is assigned to the
output and its transformity is 1000 sej/J. However, when the system is studied in more detail, it is
revealed that the system is composed of two processes (A and B). The revealed internal structure
does not change the transformity of the output, however, we now have a new "intermediate"
product (A) whose transformity is 10 sej/J. Further study reveals there is a feedback from B to A.
The new information does not change the previously calculated transformities, but does allow us
to calculate a third, whose value is 111 sej/J.
Two aspects of EMERGY analysis are important in this example: first, because of the
feedback from B, EMERGY analysis assumes that the process cannot be "decomposed" into two
separate processes, one that makes product A and one that makes product B And second,
neither of the outputs from B can be produced without the other. In this way all input EMERGY
is assigned to both outputs, since it takes the entire input to produce both outputs. In all, the new
information does not change previously calculated transformities but does allow for calculation of
additional transformities for the newly revealed internal structure.
Energy analysis, on the other hand, has no set rule for manipulating byproducts. Since the
matrix process by which embodied energy is calculated cannot handle two different outputs from
the same process, it must divide the process into two different processes or assume the products
are the same. Operationally, for processes like that in Figure 16c, energy analysis must either
split them into two different processes, or assume that the outputs are the same product.
Obviously, the first is to be preferred. For instance, if the process B is a sawmill producing
sawdust (B) and lumber (B1), to assume that sawdust and lumber are the same is a fatal error,
especially if one were purchasing a material with which to build a house. However, there is a
flaw in splitting any process into different processes for the sake of the convenience of the tool by
which accounting is preformed. The flaw is simply one of logic...they are not two different
process and the byproduct cannot be produced separately. Assuming the saw mill process can be
separated into a process of making lumber and one of making sawdust and that the two processes
can exist independently is inconsistent with reality. The first is not possible in a thermodynamic
world and the second may be possible, but not competitive in a world that prizes order over
The EA "byproducts flaw" leads to another problem that has to do with reproduceability.
When a process that has byproducts is split into two different processes, some decision has to be
made concerning the energy inputs and how much of the inputs will be assigned to each of the
new, separate processes." If all the energy inputs are assigned to both processes, the assignment
violates the basic EA rule that all systems conserve, since assigning the entire amount to both
processes will double count the input. The only other alternative is to split the inputs between the
two new processes, with an infinite number of possibilities for their allocation. Depending on
how input energy is assigned, embodied energy and intensities of outputs will be quite different.
There is no rule that suggests how the input energies should be allocated and thus, depending on
who does the analysis, the results might vary. (End Brown)
This joint paper was undertaken as a means of illustrating the similarities and differences
between energy analysis and EMERGY analysis. One important similarity is apparent: intensities
and transformities are analogous concepts, although, in practice they differ markedly because of
the energies included and methods employed in their calculation. Differences between the two
methods stem from different conceptual underpinnings and accounting procedures.
Probably the most significant difference is related to the "form" of EMERGY and
embodied energy. EMERGY is defined as the energy of one type (usually solar energy) that is
required to produce something. Energies of different types (i.e. solar, tidal, chemical potential
energy in rain, fuels, or electricity) are expressed in the equivalent solar energy required to make
them. Embodied energy analysis, as practiced, uses strictly the heat energy of fuels and does not
include environmental energies. The embodied energy in goods and services, for instance, does
not include the environmental support that is derived from solar, geophysical, and tidal energies
that drive all economies. It is suggested that these energies could be included, but there is no
formal way of including them, no intensities have been calculated for them, and the fact that the
biosphere is a single process with multiple outputs precludes the use of the matrix inversion
process for their calculation.
EMERGY analysis includes the service input of humans in all evaluations, and does so by
considering that the work output of humans is one of several multiple outputs and therefore the
total EMERGY support to humans is assigned to their work. Embodied energy analysis routinely
does not include human service inputs to processes, but it can. When it does, considers their work
output to be only some fraction of the total fuel energy used in their support.
Another significant difference is related to the system properties of EMERGY and
embodied energy. EMERGY is a system property and is defined as the energy required to
produce something. Processes (or systems) having more than one output cannot be decomposed
into separate processes each having only one output because a given process (or system) is an
interdependent whole. As a result, processes having more than one output require all the
inflowing EMERGY for the production of each output. Energy analysis, on the other hand.
chooses to assign embodied energy in a manner that cannot account for multiple products from
one process without additional assumptions. Thus processes with multiple outputs are either
decomposed into separate processes, each having a single output, or outputs are treated as if they
were the same product, or multiple products are ignored in favor of a single output.
It is quite apparent that EMERGY and embodied energy are two very distinct concepts.
EMERGY analysis recognizes and, in fact, has pioneered the concept of energy "form,"
developing the conceptual and empirical basis that all energies are not of the same quality.
Embodied energy analysis, and the units of embodied energy, do not recognize the qualities of
energy across the energy spectrum of the biosphere, but instead account for only what has been
termed "cultural" energies. In so doing about half of the total energy driving the economies of the
biosphere is ignored.
We both acknowledge that EMA attempts a bolder and more comprehensive synthesis of
interdependencies driving ecological systems and the economic systems that depend on them than
does EA. The question raised by EA is whether the assumptions of EMA, especially regarding
bookkeeping of flows and calculations of interdependency, are useful and defensible.
8. DIALOG TABLE (Brown and Herendeen)
1. WHY DO THIS ANALYSIS?
In order to quantitatively evaluate
public policy options,
environmental impacts and to
lend quantitative insight into
questions of resource
management EMERGY analysis
expresses all energies and
resources in the same type of
energy...most often in solar
Energy is important, more so than
current $ price implies. For
example, energy is a good Ist-
order indicator of environmental
impact Indirect effects are
important, and EA explicitly
addresses them. Results are to be
used in conjunction with other
indicators, likely including
economic and ecological ones.
The approach is easily extended
to indirect effects in materials,
pollution, and employment
Useful in "net energy" questions.
If all energies are expressed in
heat calories, large errors can
result because energy of different
types is not accounted for in an
Central concept Indirect effects
are quantified. The EMERGY
required to produce a good or
service divided by the energy
of the good or service equals its
While EA quantifies "indirect"
energy it does not account for
energy of different types. EA does
not explicitly require that all
energy be expressed as energy of
Central concept Indirect effects
are quantified. Embodied energy
= the direct and indirect energy
required to produce a good,
service, or entity. Examples: the
sunlight to produce an eagle, the
coal to produce a car or an opera
production, the energy to produce
ethanol from grain.
EA has used different types of
energy (coal, crude oil, natural
gas, nuclear and hydro) in the
same analysis (Bullard, Hannon,
and Herendeen, 1975). In
ecological applications, EA can
account for trophic levels and
Maximum EMERGY Principle:
Systems that will prevail in
competition with others develop
the most useful work with
inflowing EMERGY sources by
reinforcing productive processes
and overcoming limitations
through system organization.
Useful work is defined as
reinforcing production processes
and overcoming limitations.
All energy required to make a
product is accounted for in energy
of one type. This includes
renewable, non-renewable and
energy in services.
There are four rules:
o Source EMERGY to a process
is assigned to the process's
o Byproducts from a process have
total EMERGY assigned to each
o Split EMERGY flows are
proportional to energy flows.
o EMERGY cannot be counted
twice within a system:
o ENERGY in
feedbacks should not be
o byproducts, when
reunited, cannot be
Every compartment is in
embodied energy balance, as
expressed in Figure 2. This
balance is not a direct
consequence of the 1 st Law of
Thermodynamics. Rather, it
follows from the bookkeeping
requirement to incorporate losses
into final products. The same
'conserving" accounting is used
for applications of the method to
materials, pollution, and
employment Resulting equations
yield energy intensities, measured
in units of energy/(flow variable).
Byproducts are manipulated to
maintain this balance. A
consequence is that the entire
system is also in embodied energy
balance. Embodied energy flows
= (energy intensity)*(flow).
Ratio of EMERGY flows so
reckoned, to original energy
flows, is transformity.
EA has no explicit methodology
for dealing with byproducts, but
must assume them away to satisfy
EA performs a necessary, but
transparent, manipulation to
remove byproducts to prevent
what EA considers double
EA has no optimizing principle.
There is often an underlying
assumption that minimizing non-
renewable energy inputs is
The Maximum EMERGY
Principle is incomplete, basically
replacing the term to be defined
(maximum EMERGY) with
another, undefined term (useful
a conservation principle.
5. CHOICE OF FLOW
6. HOW MANY
THERE FOR AN
The flow variable is usually
energy, although mass is
sometimes used where
transformities have been
calculated on a mass basis (i.e.
EA's allowance for anything as a
carrier produces quite different
results in the same system if
analyzed using different carriers.
This is because energy, dollar,
and nutrient flows are not
EMA accounting rules are
straightforward and easily
mastered. All feedbacks within a
system are explicitly dealt with.
There are at least n, sometimes
more. EMERGY is a systems
property and therefore if a
compartment has a byproduct
output, these two flows may have
Can be energy (in ecological
applications, for example), but
usually is not in economic
applications: dollars of product,
tons of steel, liters of herbicide,
etc. Different units can be used
for different compartments.
Comparing results with different
flow variables is often insightful,
e.g., biomass energy v. nutrient in
ecosystems (the latter have more
feedback) (Herendeen, 1990).
EMA'S preoccupation with
energy as carr has reduced
attention to feedback and
produced accounting rules that
are confounded and cumbersome
when feedback occurs.
EMA's rules on double counting
prematurely truncate feedback
Exactly n, unless extra
assumptions are made.
7. RELATIONSHIP OF
INTERIOR FLOWS TO
Accounting rules above forbid
any interior EMERGY flow to
exceed summed EMERGY inputs
to system. Figure I la shows this.
Interior embodied energy flows
can exceed summed embodied
energy inputs to system. Figure
1 lb shows this. All
compartments are still in balance,
as is entire system: system output
= system input This is no
contradiction. It quantitatively
reflects the existence of feedback.
The bookkeeping that produces
this apparent contradiction for
embodied energy also predicts
that nutrient flows should show
the same effect But nutrient
flows are measurable. Therefore
the effect is real and the
bookkeeping is valid. Storage
change is irrelevant, as this
argument assumes steady state.
8. RELATIONSHIP OF
SYSTEM OUTPUT TO
By definition the EMERGY value
of an output of a system is equal
to the sum of the inputs. When
there are byproducts all
EMERGY of inputs are assigned
to all outputs.
Embodied energy output must
equal embodied energy input
The fact that interior flows of
embodied energy can exceed
input flows makes it difficult to
balance units. Intensities have
units of energy/unit carrier. A
paradox develops then. Because
of dimensions of inflows and
pathways are carrier/time, and
intensities are embodied
energy/unit carrier. Then
embodied energy assigned to a
pathway should have the
dimensions of embodied
energy/time (since embodied
energy = intensity flow). This
would require that no more
embodied energy can be assigned
to a pathway in a period of time
than is inflowing to the system. It
is apparent that some inflowing
energy is double counted in the
matrix manipulations. It is
therefore difficult to understand
how practitioners of EA can
suggest the entire system is in
balance. Thus, embodied energy
on pathways that are larger than
inflows are not rigorously
defined. To have higher
embodied energy (or nutrients)
flowing on interior pathways can
only be accomplished if a system
has accumulated storage over
time that were made possible
because inputs exceeded outputs.
EMA's lack of overall
conservation of EMERGY is a
fatal flaw: how can one make
valid quantitative comparisons
with a non-conserved quantity?
What is the point of rigorous
adherence to Rule 4 when it leads
to non-conservation such as this?
The argument for conservation
has little to do with the First Law
of Thermodynamics, because
similar arguments apply to labor
intensity, nutrient intensity, etc.
9. RELATIVE WEIGHT GIVEN
TO HIGHER V. LOWER
TROPHICIC) LEVELS, AS
EMA attaches more relative
importance to higher order
pathways than does EA. This
importance is indicated by the
ratio of the largest to smallest
transformity (for EMA) or
of the largest to smallest energy
intensity (for EA).
Adherents of EA assume that
because feedback flows often.
have more embodied energy
assigned to them than the sum
total of inputs, they somehow
account for feedback better than
EMA. Yet the very fact that they
do not reflect the hierarchical
structure of systems (i.e. "they
tend to have less disparity...')
shows that the black-box
approach to calculating
intensities does not account
adequately for the value of higher
order flows and compartments.
The problem is that with blending
of intensities, EA double counts
It can go either way depending on
the details of the original energy
flows. Feedback tends to make
the energy intensities converge.
What is "high" and 'low" depends
on the degree of feedback.
For a linear chain without
byproducts, EMA's transformities
= EA's energy intensities. In EA,
a feedback-rich system will tend
to have less disparity in energy
intensities of different
compartments; feedback is
blending them together. EMA's
rule against "double counting
feedback energy" diminishes
feedback's role, which tends to
support EMA's claim. However,
the EMA rules on byproducts can
reverse this, as can be shown
explicitly. But the real question
is "what's the problem?" If
herbivores also eat detritus, that
Adherents to EA have developed
a straw dummy and called it a
fatal flaw, and in this way cloud
the issue of EA's inability to deal
with byproduct flows. When a
process produces an output and a
byproduct, one cannot be made
without the other and therefore
both are assigned the total
EMERGY input To a EA
practitioner this appears to
violate their definition of
conservation. Valid quantitative
comparisons can be made
regardless of adherence to EA's
definition of conservation as long
as they are made within the same
algebraic system and those things
being compared are quantitatively
determined within that system.
embodied energy and gives
higher weight to lower order
compartments and flows leading
to erroneous comparisons... for
instance with the so called
"blending" of intensities, an
energy analysis of an ecosystem
might lead one to the conclusion
that it takes just as much sunlight
to support a herbivore as it does
to support a carnivore. As with
EA, EMA deliberately includes
the effects of feedback, and does
not truncate these effects as
alluded to. In item number 8,
above, EMA was criticized
because the EMERGY of an
output from a system cannot be
adequately determined unless the
interior details are known.
Knowing the interior details (and
the existence of feedback flows)
affects the output In other words,
the existence of feedback has very
real and consequential affects on
system output and is always
accounted for using EMA.
will affect how much sunlight is
required to produce a carnivore
that eats the herbivore. EA
deliberately includes the effects of
this feedback, as it should. EMA
truncates these effects.
10. IS HUMAN LABOR
INCLUDED IN THE
It routinely is, because to not
count labor leaves out a
significant portion of the energy
required to make economic
It can be, though EA usually does
not include it A common
application of energy intensities
has been to investigate the energy
requirements of different
consumer spending patterns
("market baskets). Including
labor as an aggregated sector
with an immutable market basket
negates this idea. EA can include
labor if that is desired.
While proponents of EA suggest
that labor can be included if it is
desired, in practice it is not
included, and in so doing EA
suggests that human services are
not necessary inputs to economic
11. IS RENEWABLE ENERGY
It routinely is included because to
not include renewable energy
would leave out a significant
portion of the energy required to
make most products. At the
present time renewable energy
accounts for roughly 1/2 the total
energy driving the combined
system of humanity and nature.
Proponents of EA suggest
renewable energy can easily be
included in the analysis; however
routinely it is not included and in
those rare occasions where it is
included, it is quantified as heat
energy of sunlight, with no
accounting of rain, tides, waves,
etc. and their differing qualities.
EA was originally developed for
fossil and hydro energy, but
renewable energy can easily be
EA does not attempt to quantify,
comprehensively, in energy or
other terms, the environmental
services that support human
activity. This laudable objective
is pursued by EMA.
12. HOW ARE BY-
PRODUCTS DEALT WITH?
Byproducts are necessary
outflows from most processes and
as such must be accounted for. In
some case they are useful, for
instance, sawdust may be burned
to generate steam. In other cases
they are not immediately useful
such as lignin wastes from a pulp
mill. Never-the-less, they are
outputs and their EMERGY
content is equal to the sum of the
EMERGY inputs to the process.
Byproducts cannot be dealt with
using the EA methodology.
Therefore they are either assumed
not important and neglected, or a
process that generates more than
one output is split into a number
of process that equal the number
of outputs. This requires that the
input energies be split in some
manner among the new
processes. The logical extension
of this splitting is that outputs and
byproducts can be produced
EA performs a necessary but
transparent manipulation to
remove byproducts to prevent
what EA considers double
It is true that EA's manipulation is
equivalent to assuming that
byproducts can be produced
separately. This is unfortunate,
but is the price paid for
conservation of embodied energy.
It is impossible to conserve
embodied energy while tracking
byproducts as EMA does. One
must choose between the two
Berry, R. S., and M. Fels 1973. Energy cost of automobiles. Bulletin of the Atomic Scientists
29:11-17, 58-60. December.
Bezdek, R., and B. Hannon. 1974. Energy, Manpower, and the Highway Trust Fund. Science
Boskin, M.J., Bradford, D.F., Wonnacott, P. 1992. Economic Report of the President. U.S.
Government Printing Office, Washington, D.C.
Brown, M. T., and T. R. McClanahan. 1992. EMergy Analysis Perspectives of Thailand and
Mekong River Dam Proposals. Final Report to The Cousteau Society. Gainesville, FL: Center
for Wetlands and Water Resources, Univ. of FL. 56 pp. (CFW-92-03)
Brown, M. T., S. Tennenbaum, and H. T. Odum. 1988. Emergy Analysis andPolicy
Perspectives for the Sea of Cortez, Mexico. Report to the Cousteau Society. Gainesville, FL:
Center for Wetlands, Univ. of FL. 58 pp. (CFW-88-04)
Bullard, C. and R. Herendeen. 1975. The Energy Costs of Goods and Services. Energy Policy
Bullard, C. W., B. Hannon, and R. Herendeen. 1975. Energyflow through the United States
economy. Wall chart, University of Illinois Press, Urbana.
Bullard, C., P. Penner, and D. Pilati. 1978. Net Energy Analysis: Handbook for Combining
Process and Input-Output Analysis. Resources and Energy 1:267-313.
Bureau of the Census. 1993. StatisticalAbstract of the United States. U.S. Dept. of Commerce.
U.S. Printing Office, Washington, D.C.
Casler, S. 1994. (Allegheny College, Meadville, PA). Energy Intensities for 1985. Personal
Chambers, R. S., R. A. Herendeen, J. J. Joyce, and P. S. Penner. 1979. Gasohol: Does it or
doesn't it...produce positive net energy? Science 206:789-795.
Chapman, P. 1975. Energy analysis of nuclear power stations. Energy Policy 3:285-298.
Cleveland, C. 1992. Energy Quality and Energy Surplus in the Extraction of Fossil Fuels in the
U.S. Ecological Economics. 6:139-162.
Costanza, R., and R. Herendeen. 1984. Embodied energy and economic value in the United
States economy: 1963, 1967, and 1972. Resources and Energy 6:129-164.
Denton, R. V. 1975. The energy costs of goods and services in the Federal Republic of
Germany. Energy Policy 3:279-284.
Energy Information Administration. 1992. Annual Energy Review. U.S. Dept. of Energy. U.S.
Printing Office, Washington, D.C.
Finn, J. T. 1980. Flow Analysis of Models of the Hubbard Brook Ecosystem. Ecology
Finn, J. T. 1976. Measures of ecosystem structure and function derived from analysis of flows.
Journal of Theoretical Biology 56: 115-124.
Giancoli, D.C. 1984. Physics, principles with applications. Prentice-Hall, Englewood Cliffs, N.J.
Hall, C., C. Cleveland, and R. Kaufmann. 1986. Energy and Resource Quality: The Ecology of
the Economic Process. John Wiley and Sons, New York.
Hannon, B., and C. Joiris 1989. A seasonal analysis of the southern North Sea ecosystem.
Hannon, B., R. Herendeen, F. Puleo, and A. Sebald. 1975. The energy, employment, and dollar
impacts of alternative transportation options, pp. 105-130. In R. Williams, ed., The Energy
Conservation Papers. Ballinger.
Hannon, B. 1973. The Structure of Ecosystems. Journalof TheoreticalBiology 41:535-546.
Herendeen, R., C. Ford, and B. Hannon. 1981. Energy cost of living, 1972-73. Energy
Herendeen, R. 1978. Total energy cost of household consumption in Norway, 1973. Energy
Herendeen, R., and A. Sebald. 1975. Energy, employment, and dollar impacts of certain
consumer options, pp. 131-170. In R. Williams, ed., The Energy Conservation Papers.
Herendeen, R., T. Kary, and J. Rebitzer. 1979. Energy analysis of the solar power satellite.
Herendeen, R. 1990. System-level indicators in dynamic ecosystems: comparison based on
energy and nutrient flows. Journal of Theoretical Biology 143:523-553.
Herendeen, R. 1988. Net energy considerations, pp. 255-273. In R. West and F. Kreith, eds.,
Economic Analysis of Solar Thermal Energy Systems. MIT Press.
IFIAS. 1974. Energy Analysis Workshop on Methodology and Conventions. International
Federation of Institutes for Advanced Study, Nobel House, Sturegatan 14, Box 5344, S-102
Stockholm, Sweden. 89 pp.
Kirkpatrick, K. 1974. Independent verification of I/O energy results. Center for Advanced
Computation, Technical Memorandum No. 26, University of Illinois at Urbana-Champaign.
Leontief, W. 1973. Input-Output Economics. New York, Oxford University Press.
Lotka, AJ. 1922. Natural selection as a physical principle. Proceedings of the National
Academy of Sciences 8:147-155
Leontief, W. 1970. The dynamic inverse. In Carter, A. and A. Brody, eds., Contributions to
Input-Output Analysis, New York, American Elsevier.
Lotka, A.J. 1922. Natural selection as a physical principle. Proceedings of the NationalAcademy
of Sciences. 8:147-155.
Lotka, A.J. 1945. The Law of Evolution as a Maximum Principle. Human Biology 17:167
Lotka, A.J. 1925. PhysicalBiology. Williams and Wilkins, Baltimore
Odum, H. T. 1975e. "Energy Quality and Carrying Capacity of the Earth" (Response at Prize
Ceremony, Institute La Vie, Paris). TropicalEcology 16(1). 14 pp. (CFW-75-33)
Odum, H. T. 1983b. Systems Ecology: An Introduction. New York, NY: John Wiley & Sons.
644 pp. (CFW-83-02)
Odum, H. T. 1988a. "Self-Organization, Transformity, and Information." Science, 24 November
1988, Vol 242:1132-39. (CFW-88-12)
Odum, H. T. 1991. "Emergy and Biogeochemical Cycles." In C. Rossi and E. Tiezzi (Eds.)
Ecological Physical Chemistry, Proceedings of an International Workshop, 8-12 November
1990. Amsterdam: Elsevier Science Publishers B.V. 16 pp. (CFW-91-15)
Odum, H.T. 1994. Environmental Accounting: EMERGY and Decision Making. Draft
Manuscript for John Wiley, N.Y. Gainesville, FL: Environmental Engineering Sciences, Univ. of
Fla., 416 pp.
Odum, H. T. and J. E. Arding. 1991. "Emergy Analysis of Shrimp Mariculture in Ecuador."
Working paper prepared for Coastal Resources Center, University of Rhode Island, Narragansett.
124 pp. (CFW-91-03)
Patten, B., M. Higashi, and T. Bums 1990. Trophic dynamics in ecosystem networks:
significance of cycles and storage. Ecological Modelling 51:1-28.
Peet, N.J. 1986. Energy requirements of the output of the New Zealand economy. 1976-1977.
Pilati, D. 1977. Energy analysis of electricity supply and energy conservation options. Energy
Pimentel, D. 1980. Handbook of Energy Utilization in Agriculture. CRC Press, Inc., Boca
Scienceman, D. M. 1989. "The Emergence of Emonomics." In Proceedings of The
International Society for Social Systems Science, 33rd Meeting, 1989, Edinburgh, Scotland,
Volume III, pp. 62-68. (CFW-89-02)
Scienceman, D. M. 1987. "Energy and Emergy." In G. Pillet and T. Murota (eds.),
Environmental Economics: The Analysis of a Major Interface. Geneva: R. Leimgruber. pp.
U. S. Department of Commerce. 1984. Input-Output Structure of the United States Economy,
1977. Survey of Current Business 64:42-84.
Ulanowicz, R. 1986. Growth and Development: Ecosystems Phenomenology. New York,
von Bertalanffy, L. 1968. General System Theory. George Braziller Co,. New York
Wulff F., and R. Ulanowicz. 1989. A comparative anatomy of the Baltic Sea and Chesapeake
Bay Ecosystems. In F. Wulff J. Field, and K. Mann, eds., Flow Analysis in Marine Ecosystems,
New York, Springer.
Table 1. Energy flows in Figure 10. Units = J per time unit. The sum of actual energy inputs,
10200 exactly balances the sum of dissipation and exports by the First Law of Thermodynamics.
The sum of total outputs exceeds this figure because, for example, part of Compartment A's
output is counted in Compartment D's.
FROM\TO A B C D E Dissip Export Total
I__ I Output
A 0 0 0 50 0 2960 0 3010
B 0 0 0 100 0 6905 0 7005
C 10 0 0 0 0 201 0 211
D 0 0 10 0 10 120 10 150
E 0 5 1 0 0 3 1 10
Column sum __ 10189 11
Energy input 3000 7000 200
Embodied energy 3000 7000 20000
Table 2. EMERGY Evaluation of Low Pa Mong Dam and Irrigation.
2.56E + 04 p/yr
5.91E + 16 J
9.70E + 08
1. River Geopotential
3. Aquatic Product
5. Operation and Maintenance
10. Ag. Prod. (Rice)
11. Ag. Prod. (Maize)
13. Irrigation (Services)
5.68E + 16.5
3.62E + 16J
9.30E + 14J
1.15E + 11g
2.04E + 07 $
1.35E + 11 g
2.31E + 08 g
6.66E + 08 g
3.78E + 07 S
1.23E + 11 g
7.42E +14 J
4.00E + 06 $
4.80E +06 $
ENERGY YIELD RATIO NOT INCLUDING
ENERGY YIELD RATIO INCLUDING SEDIMENTS:
3.46E + 12
7.00E + 07
1.80E + 09
6.70E + 09
3.46E + 12
9.70E + 08
4.75E + 04
3.46E + 12
3.46E + 12
2.98E + 15
6.30E + 04
Table 3. EMERGY Evaluation of Upper Chiang Khan Dam.
Transformity Solar EMERGY
Note* Item Raw Units (sej/unit) (E18 sej/yr)
1. River Geopotential
3. Aquatic Product
5. Operation and Maintenance
10. Ag. Prod. (Rice)
11. Ag. Prod. (Maize)
13. Irrigation (Services)
14. Social Disruption
5.19E + 16 J
3.20E + 16J
6.10 E+ 14J
3.68E + 10g
7.40E + 06 $
1.1E + 11 g
1.19E + 08 g
4.16E + 08 g
2.770E + 07 $
2.30E + 09 g
1.14E +15 J
2.63E + 06 $
1.54E +06 S
1.68E + 04 p/yr
5.65E + 16 J
9.7E + 08
3.46E + 12
7.00E + 07
1.80E + 09
6.70E + 09
3.46E + 12
9.70E + 08
4.75E + 04
3.46E + 12
3.46E + 12
2.98E + 15
6.30E + 04
EMERGY YIELD RATIO NOT INCLUDING
ENERGY YIELD RATIO INCLUDING SEDIMENTS:
Table 4. Energy Analysis of Low Pa Mong Dam and Irrigation (1992)
Item Amount Units Energy BTU/Unlt YR ofTechnology/YR of Technology Multiplier(b) Inflation Corrected Energy Energy
(*X1992) Intenstly Dollar (YR/YR) Multiplier(c) Intenslty(d) (BTU/$) (BTUIYR)
1 River Uoopoterial
3 Aquatic Prod.
5 OP and Maint
10 Rice Production
11 Maize Production
14 Social Disruption
Incremental Energy Ratio (IER)
N- -Not i.dered
3.62E+16 J 9.48E-04 BTU/J
2.07E07 S 1.71E.04 BTU/S (1985/1985)
1.35E+11 g 1.23E+00 BTU/g (1977)
2.31E+08 g 5.63E+01 BTU/g (1963)
6.66E+08 g 9.84E+01 BTU/O (1967)
3.78E+07 S 5.51E+03 BTU/S (1985-1985)
1.23E+11 g 1.17E+01 BTU/g (1975)
7.42E+14 I 2.08E+04 BTU/J (1975)
4.00E+06 S 1.38E+04 BTU/S (1992)
a. From Tables 2 and 3
b. Energy/ONP from Energy Information Administralion (1992), Bureau of the Census (1993)
c Deflators frno Bkin Bradford, and Wonnacot (1992)
d, Current energy intensity (Energy intenity)(Technical Multiplier)*(Inflation Multiplier)
2. 1KWII-3413 BTU, IBTU-1055J
5. Casier, 1994: Section 12
6. Ilercndee. t al, 1979
7. lerendeen et al. 1979
8. Kirkpatrick, 1974
9. Casler, 1994: Sector 73
10. Pimentel, 1980
11. Pimentel., 1980
12. National average(b)
Table 5. Energy Analysis of Upper Chiang Khan Dam (1992)
Item Amount Units Energy BTU/Unt YR of Technology/YR of Technology Multlpler(b) Inflation Corrected Energy Energy
(aX1992) Intensity Dollar (YR/YR) Multipler(c) Inienslty(d) (BTU/S) (BTU/YR)
I river ueopotentia
3 Aquatic Prod.
5 OP and Maint
10 Rice Production
II Maize Production
14 Social Disruption
Incremental Energy Ratio (IER)
3.20E+16 J 9.48E-04 BTU/J
7.40E+06 S 1.71E+04 BTU/S (1985/1985)
1.10E+11 g 1.23E+00 BTU/g (1977)
1.19E+08 g 5.63E+01 BTU/g (1963)
4.16E+08 g 9.84E+01 BTU/G (1967)
2.77E+07 S 5.51E+03 BTU/S (1985/1985)
2.30E+10 g 1.17E+01 BTU/g (1975)
1.14E+15 1 2.08E+04 BTU/J (1975)
2.63E406 S 1.38E+04 BTU/$ (1992)
a. From Tables 2 and 3
b. Energy/GNP from Energy Information Aminis tratin (1992) Bureau of the Census (1993)
c. Deflators from Boakin, Bradford, and Wonacott (1992)
d. Current energy intensity (Energy int sity)(Technical Multiplir)(Inflation Multiplier)
2. IKWH-3413 BTU, IBTU-1055J
5. Caler, 1994: Section 12
6. llerendeen, et al, 1979
7. Herendeen,t al, 1979
8. Kirkpatrick, 1974
9. Casler, 1994: Sector 73
10. Pimentel, 1980
11. Pimentel., 1980
12. National average(b)
Figure 1. Schematic diagram of dendritic structure symbolizing vertical analysis.
Figure 2. The fundamental sector embodied energy balance equation of Energy Analysis.
- - -- -I
E k X
E k Xk k k
1 J Xj -^----EjYj
Figure 3. Embodied energy flows (from EA) in a 2-compartment system. E's represent actual
energy per unit time; e's represent energy intensities (energy per unit of X). EA. is the embodied
energy flowing out of compartment j. Ej+E, is the energy entering the system per unit time; ejYj +
EkYk is the embodied energy leaving the system. In EA, these two flows are equal.
i B I 1
Biomass (Kcal d 1)
Nutrient (g d-1
Nutrient (g.d )
Figure 4. Effect of feedback on embodied energy and nutrient flows, a) Biomass flows and
nutrient input. b) Embodied nutrient flows (g d'1). Because no nutrient is dissipated, embodied
nutrient is nutrient. These flows are actual, measurable flows.
100 20 Lu mner
I Sawmill Sawdust
100 Sawmill 10 (2 lumber + 1 sawdust)
-- Sawmill -- 30 (lum-ber/sawdust)
2000 2000- Sawmill
10 x Dept. -
Figure 5. Example flow diagram to illustrate possible ways to account for byproducts in EA. All
units=J/time. a) Example: sawmill produces lumber and sawdust. Embodied energy in each output
is undefined without more information. b) Output defined in units of(2 lumber + 1 sawdust), of
which output = 10 units. c) Output defined as lumber or sawdust, of which output = 30 units. d)
Inputs and dissipation assigned arbitrarily to two parallel processes, one producing lumber and
one producing sawdust. x, y and z can have any values consistent with energy conservation.
Embodied energy out of sawmill = energy input.
EMERGY of Form "S"
Figure 6. Diagram to illustrate EMERGY flows and transformity in a single-output system. a)
energy flows, b) EMERGY flows, and c) transformities.
(given: F=t S)
S 400 Energy
B 20 <
B20 A2 =3
EMERGY of form "S"
B 25 A 50
0v A2 =50
Figure 7. EMERGY flows and transformities in a dual output system, illustrating both byproducts
and splits. a) Energy flows, b) EMERGY flows, and c) transformities.
-I '-J 200
(400 100) (160*40)
EMERGY of Form "S"
Figure 8. EMERGY flows and transformities in a 2-compartment system with feedback. a)
Energy flows, b) EMERGY flows, c) transformities.
Figure 9. EMERGY flows and transformities in a 4-compartment system, demonstrating the
fourth rule of EMERGY algebra.
A 50 5 split
S7000 B D
6905 2950 '201 J120 3
Figure 10. Energy flows in 5-compartment system in units ofjoules/time.
Figure 11. EMERGY (a) and embodied energy (b) flows of the 5-compartment system in Figure
Figure 12. Transformties (a) and energy intensities (b) of the 5-compartment system in Figure 10.
on the Mekong River
M.T. Brown 2/0O
Figure 13. Energy diagram of relationships between urban and rural populations and the proposed
hydroelectric dars on the Mekong River.
33- o rk
-- -. Reproduction
Figure 14. EMERGY assigned to human activities assuming (a) human outputs are splits and (b)
human outputs are byproducts. In EMERGY Analysis, human outputs are considered byproducts.
Transformity of the input is 1 sej/J.
1000 > B 1
1000 A 100 B1-- 1
1000 A 100 B B 1
Figure 15. Two-compartment system to illustrate EA treatment of aggregation. All flows in kcal
d-1. a) Initial stage: one compartment with one output. EB = 1000 kcal/kcal. b) Internal details are
revealed; system can also produce a second product. e^= l0kcal/kcal; EB = 1000 kcal/kcal. c)
Internal details include feedback. E^ = 100 kcal/kcal; E = 1000 kcal/kcal.
1 J Transformity of Output
- = J100L .J = 1000 sej/J
Transfonnity of A
= 100cQXse = 100 sej/J
Transformity of B
= 1000se = 1000 sej/J
Transformity of A
= l100 sel = 10 sej/J
Transfonnity of B1
= 100Qsel = 1000 sej/J
Transfanity of B2
= 1000sel = 111 sej/J
Figure 16. A 2-compartment system (the same as Figure 16) to illustrate aggregation questions in
-- AI100 i -I