transcript
by
RU. AYRES* and
L.W. AYRES** 93/78/EPS
This working paper was published in the context of INSEAD's Centre
for the Management of Environmental Resources, an R&D
partnership sponsored by Ciba-Geigy, Danfoss, Otto Group, Royal
Dutch/Shell and Sandoz AG.
* Professor of Environmental Economics, Sandoz Chair in Management
and the Environment, at INSEAD, Boulevard de Constance,
Fontainebleau 77305 Cedex, France.
** Research Associate, at INSEAD, Boulevard de Constance,
Fontainebleau 77305 Cedex, France.
A working paper in the INSEAD Working Paper Series is intended as a
means whereby a faculty researcher's thoughts and findings may be
communicated to interested readers. The paper should be considered
preliminary in nature and may require revision.
Printed at INSEAD, Fontainebleau, France
CHEMICAL INDUSTRY WASTES; A MATERIALS BALANCE ANALYSIS
Robert U. Ayres and Leslie W. Ayres INSEAD November 1993
Abstract
The paper is a systematic derivation of aggregate production wastes
for the U.S. chemicals industries, SIC 28 (1975 - 1988). To
facilitate this derivation we have classified chemicals within this
sector by key element(s) from which they are derived. In principle,
each is thought of as initiating a transformation process,
converting material (and energy) inputs to outputs. Both inputs and
outputs are published in well-established government statistics,
with rare exceptions. This makes it possible to use the
materials-balance methodology by comparing aggregate inputs and
outputs. Unlike other sectors of the economy, knowledge of the
transformation processes themselves is needed in evaluating the
chemical industry.
We compare the results of this "bottom up" approach with other
estimates of waste residuals. In several cases, significant
discrepancies have been identified. However, the major value of
this approach is to clearly distinguish between dry and wet wastes.
Our approach is probably superior to the conventional one in this
regard.
The data used is sufficiently standard so that it should be
possible for a government agency to compile and present these data
on a routine basis. Where there are major differences with other
sources (including direct measures) the underlying data probably
need revision.
1. Introduction
The chemical industries (roughly, SIC 28) are unquestionably major
polluters, and major sources of hazardous wastes. According to
USEPA [USEPA 1988, 1991] the inorganic chemical sectors (SIC 281)
generate 920 million tons of non-hazardous industrial waste each
year; the plastics and resins sector (SIC 282) generates 181
million tons, and the fertiliz- er/agrichemical sector (SIC
2873-79) generates 166 million tons. Unfortunately, it is
completely unclear how these numbers were generated and how much
(if any) of the weight of wastes is actually water.
An earlier (1985) study by Science Applications Inc., sponsored by
EPA [SAI 1985] estimated the "dry" weights of waste from these
sectors as follows: inorganic chemicals (1979), 28.8 million tons;
plastics and resins manufacturing (1982), 49.6 million tons;
fertilizer and agrichemicals (1983), 65.0 million tons. In all
three cases, however, there is some doubt whether the numbers
really refer to dry weight; in cases of disagreement, EPA has
recommended its own numbers in preference to those of SAI [see
Allen & Behmanesh 92, Table III, footnote a].
R.U. Ayres & L.W. Ayres November 12, 1993
Based on a comprehensive survey conducted in 1987 by EPA (which,
however, did not include the mining industry) the chemical industry
is, apparently the largest source of hazardous wastes. In millions
of tons, the key subsectors reported hazardous emissions in 1986 as
follows:
SIC 2800 49 SIC 2812 5 SIC 2816 4 SIC 2819 17 SIC 2821, 2873 9 SIC
2843, 2869, 2899 5 SIC 2865, 2869 6 SIC 2865 11 SIC 2892 24
The sum total of hazardous wastes reported by the U.S. chemical
manufacturing sector in 1986 was 130 million tons, or about 115
MMT. It is certain that much of this weight is water, although the
data provide no basis for estimating the water fraction.
Apart from the problem of water, there are other problems with the
TRI system, notably its limited coverage and the fact that there is
very little verification. The materials balance approach has not
been much used in this context, but one study was carried out by
Industrial Economics Inc for EPA. It covered 21 chemicals, or
groups, among the most toxic or hazardous known. Most were produced
and used in fairly large quantities, and reasonably good statistics
could be obtained on imports, exports, usage as such and conversion
to other chemicals. In several cases there were fairly large
inadvertent fluxes resulting from incidental activities, such as
transportation or petroleum refining. However, in each case IEI
identified all known uses, including consumption for purposes of
conversion to other materials. Direct conversion losses were
estimated from published process yield estimates.
The results of this study are summarized in Table 1. below. In each
case, the TRI estimates are compared with the missing or
"unaccounted for" mass of the chemical. It is not claimed that all
of the missing mass should immediately be classified as emissions
to the environment. Apart from statistical errors and limitations
on coverage, the IEI methodology is a way of estimating
pretreatment losses, not actual releases. Some of the difference is
probably due to the presence of end-of-pipe treatment facilities,
such as incinerators, that can convert a toxic or hazardous
chemical such as benzene to a non-hazardous chemical form (e.g.
CO2). On the other hand, the most treatment methods merely shift a
waste stream from one environmental medium to another, and the
waste should still be reported as a waste. In the case of metals,
conversion to another non-hazardous form is impossible by
definition. Besides, discrepancies between unaccounted for missing
mass and TRI are very large indeed, in many cases.
How then can a hazardous chemical "disappear" from the statistics?
Deliberate under-reporting is one possibility, of course. But the
most likely explanations, in our view, are the following: (i) the
missing mass is lost as a minor contaminant in some very large
waste flux 1 such as sludge or incinerator ash or (ii) it becomes
incorporated in some product designed for
One is reminded of the adage from the 1970's "the solution to
pollution is dilution".
2
Chemical Industry Waste: A Materials Balance Analysis
TABLE 1: Toxic Chemicals; Materials Balance Estimates Compared to
TRI (1000 metric tonnes)
Chemical
version & Incidental Emission Use Losses Emissions Losses
TRI Losses un- accounted for by TRI
TRI metal com-
Xylenes, mixed 3419.2 295.2 544.2 839.4 86.5 752.9
m-Xylene 34.6 6.9 0.2 7.1 1.6 5.5 o-Xylene 509.6 101.6 2.6 104.2
1.3 102.9
p-Xylene 2510.9 179.0 11.6 190.6 3.2 187.4
Carbon tetrachloride 400.2 9.9 1.2 11.2 2.3 8.9
Chloroform 224.8 26.5 19.9 46.4 12.2 34.2 Methylene chloride 183.1
161.7 1.2 162.9 70.3 92.6
Perchloroethylene 252.9 182.3 0.7 183.0 17.1 165.9
Trichloroethylene 68.9 65.2 0.3 65.5 26.2 39.3
1, 1, 1-Trichloroethane 303.4 281.5 1.6 283.1 88.1 195.1 Methyl
ethyl ketone 239.2 237.7 4.3 242.0 73.1 168.9 Methyl isobutyl
ketone 91.2 81.6 0.5 82.1 20.2 61.9 Cadmium 3.6 3.6 1.0 4.6 0.9 3.7
0.16 0.72
Chromium 536.9 418.1 24.9 443.0 31.2 411.8 9.74 21.46 Mercury 1.6
1.3 0.3 1.6 0.1 1.5 0.13 0.01
Nickel 159.2 118.1 23.7 141.8 8.7 133.1 4.02 4.73 Cyanides 629.7
151.4 31.1 182.5 5.3 177.2
Hydrogen cyanide 543.0 64.6 31.1 95.7 1.4 94.2 Cyanides, other 86.8
86.8 0.0 86.9 3.9 82.9
TOTAL 20000.8 2890.5 1836.7 4727.2 621.6 4105.6 Basic data source:
[1E1 1991]
3
R.U. Ayres & L.W. Ayres November 12, 1993
dissipative use, either as a contaminant, additive, or even as a
major ingredient. Dubious products, such as solvents and cleaning
agents, are doubly hard to track if they are widely distributed and
consumed by small businesses (not covered by the TRI legislation)
or private individuals. Many chemical products are used
dissipatively. This applies to pesticides, solvents, fuel
additives, cleaning agents, catalysts, and a number of other
categories. These points are not in dispute. However, they raise
serious doubts about TRI. It is these doubts, in part, that have
led us to attempt a materials balance analysis of the chemical
industry as a whole.
The complexity of the industry, and the large number of its
products make this task difficult. However, some simplification is
possible by focussing attention very strictly on inputs to, and
"final" products of the industry, disregarding intermediates for
the most part, except with regard to production losses. Because
many chemical products are used largely, or entirely, to make other
chemical products, "final" output tonnage is considerably smaller
than total production. MI material inputs to the industry must
ultimately become outputs, either as products or as wastes. It is
not necessary to know much about the details of chemical processes
to assert that any inputs not embodied in products must end up in
some waste stream.
The conventional SIC classification of chemicals is potentially
confusing, inasmuch as it lumps together some chemical commodities
with no functional relation to each other and, in some cases,
includes processes much more naturally associated with other
sectors. In particular, one of the biggest items in SIC 2819
(inorganic chemicals not elsewhere classified) is aluminum oxide
(alumina). This material is used almost entirely for the production
of primary aluminum, and would more logically be included in the
primary metals sector (SIC 33). However, it would be even more
confusing to try to reinvent the SIC. Hence we simply note the
difficulty.
Except for hydrocarbons, it is convenient to classify chemicals by
key element(s). The major categories are; aluminum-based,
nitrogen-based, sulfur-based, phosphorus-based, chlorine- based,
sodium-based, and hydrocarbon based 2. Our approach is to begin by
cataloging the major chemical inputs to the chemical industry, both
inorganic and organic, by element and (for hydrocarbons) by
starting point. We then try to identify the final products of the
chemical industry — treating aluminum oxide as a special case, for
reasons already mentioned — considering fertilizer chemicals (SIC
287)3, inorganic pigments (SIC 2816) and other metal- based
chemicals (except sodium chemicals), and finally non-fertilizer
chemicals. The latter group is the remainder of SIC 28; hereafter
we designate this remainder as R-28. We attempt to account for each
of the input categories by end use. We proceed by identifying
those
2 Many chemicals include two (or even more) key elements. In case
one of them is a hydrocarbon base, we include it with the other
element. (Example: methyl chloride). Hydrocarbon-based chemicals
apart, there are relatively few chemicals of importance that appear
in more than one of these categories (e.g. ammonium sulfate,
ammonium chloride, phosphorus trichloride, diammonium phosphate,
sodium chlorate).
3 N.B. Nitrogenous and phosphatic fertilizers account for most of
the nitrogen and sulfur based chemicals, and almost all of the
phosphorus-based chemicals.
4
Chemical Industry Waste: A Materials Balance Analysis
elements (other than oxygen, which is discussed later) used in
significant quantities for industrial chemicals.
In the following we use the convention metric ton = MT and million
metric tons = MMT.
2. Aluminum-Based Chemicals (Included in SIC 2819)
All aluminum-based chemicals are produced from aluminum oxide, or
alumina (Al203). Alumina is manufactured by from bauxite, mostly
imported, by the so-called Bayer process. In 1988 U.S. bauxite
consumption for alumina production was 8.97 MMT, yielding 4.995 MMT
gross weight, and 4.575 MMT calcined equivalent. The apparent
weight loss (dry) was 4.239 MMT, consisting mostly of "red mud" and
particulates. For each ton of alumina produced, 37 kg of limestone
(CaCO 3) and 37.5 kg of caustic soda (NaOH) are required, or about
0.170 MMT of limestone and 0.172 MMT of caustic soda. These also
become part of the Bayer process waste stream.
Alumina is the primary feedstock for the electrolytic reduction of
aluminum metal. In 1988 U.S. primary aluminum production was 3.944
MMT. This would have required 7.730 MMT of alumina feedstock, based
on a calculated 1.96 kg of alumina per kg of primary aluminum
including 0.03 kg of alumina converted to aluminum fluoride and/or
synthetic cryolite. This would correspond to roughly 0.160 MMT of
cryolite [(3NaF)A1F3] and 0.120 MMT of aluminum fluoride (A1F3)
produced and consumed by the aluminum industry. The Bureau of Mines
estimated that 97% of U.S. alumina production was used by domestic
aluminum smelters. The U.S. imported 4.634 MMT of calcined alumina,
and exported 1.036 MMT, for a domestic supply of 8.173 MMT. So
apparent domestic consumption of alumina for other (chemical)
purposes was 0.443 MMT in 1988, but only 0.15 MMT of this was
manufactured domestically.
The Bureau of Mines reported 1988 U.S. aluminum sulfate production
of 1.243 MMT (17% Al203 basis; see also Sulfur), plus 0.587 MMT
aluminum hydroxide trihydrate (Al203.3H20), and 0.06 MMT of
aluminates. Aluminum sulfate is made directly from bauxite. The
latter two accounted for virtually all of the alumina not consumed
by smelters.
3. Nitrogen-based Chemicals'
As noted in footnote 2 above, most ammonia compounds and urea
belong to the fertilizer roup SIC 2873. Fixed nitrogen (in the form
of ammonia and its derivatives) is an interesting case, since the
nitrogen is from the atmosphere, combined with hydrogen from
natural gas. A detailed account of nitrogen flows for the U.S.
economy is reserved for another paper [Ayres et al 1993a]. However,
the main results are summarized in Figure 1. below.
4 We acknowledge the kind assistance of Ray Cantrell of the U.S.
Bureau of Mines. He is not responsible for
any remaining errors.
ADN ..1
Hexa- methylene diamine
November 12, 1993
Figure 1: Materials Balance of Ammonia-based Chemical Production in
the U.S. 1988 (1000 MT N-content)
6
Chemical Industry Waste: A Materials Balance Analysis
Domestic production of ammonia in 1988 was 12.544 MMT (N-content).
Net imports (imports less exports) plus stock changes increased
apparent domestic consumption of ammonia to 14.745 MMT (N). In
addition, there were significant imports and exports of
N-chemicals. The major (net) import items were urea (0.483 MMT N)
and ammonium nitrate (0.091 MMT N), while major (net) export items
were ammonium phosphates (1.150 MMT N) and ammonium sulfate (0.155
MMT N). Figure 1. accounts for 89 % of the total fluxes. On the
input side, ammonia not accounted for was used outside the chemical
industry (e.g. pulp and paper), or was used in manufacturing
chemicals for which there is no published production data.
On the output side, N-content of monomers embodied in plastics and
resins in 1988 (SIC 282) added up to 0.669 MMT. Nitrate and
nitro-explosives (excluding amines) account for about 0.777 MMT
(N). Dyes (aniline, etc) account for at least 0.007 MMT (N). Other
uses of nitric acid account for 0.135 MMT N). Including the
fertilizers and animal feeds, plus known production losses of 0.230
MMT (N), and allowing for major import-export flows cited above, we
can account for 89% of the nitrogen flows, leaving about 1.75 MMT
(N) — 11% — unaccounted for. Part of the missing ammonia goes into
other "final" chemicals, including military explosives, pesticides,
plasticizers, rubber chemicals and so on. Part of it must be
ammonia used for non-chemical purposes, as noted above. Process
losses probably are considerably larger than the 0.23 MMT (N) we
have identified. This is only about 1.6% of the total nitrogen
fixation. We reckon the true figure is perhaps twice as large, or
roughly 3%. (Nitrogen wastes are smaller than they might otherwise
be, however, because ammonia- bearing waste streams are easily
neutralized by sulfuric acid to produce a useful fertilizer,
ammonium sulfate). We estimate that process losses will consume
0.25 MMT of the "missing" N.
In terms of pollution of the environment, the 3% loss rate
suggested above would be insignificant in comparison with
dissipative uses of nitrogenous chemicals. Apart from fertilizers
and animal feeds, these include industrial explosives, pesticides
and herbicides, dyes, surfactants, flotation agents, rubber
accelerators, plasticizers, gas conditioning agents, and so on. In
fact, except for the plastics and resins (and plasticizers), it is
safe to assume that virtually all nitrogenous chemicals are
dissipated in normal use, but mainly by other sectors, or final
consumers. Over a slightly longer time span —5 to 10 years — the
same is also true of plastics, synthetic rubber, fibers and
military explosives.
The ultimate forms of nitrogenous wastes, and even the
environmental media into which they are dispersed, are not
particularly well-known. One soluble waste generated by many
processes is ammonium bisulfate (NH 4)HSO4. This can be converted
to ammonium sulfate, in principle, but the conversion is not
economic in most cases. Many processes probably dissipate some N as
ammonia, some as cyanides (e.g. gold mining) and some as NO.. The
latter contributes to environmental acidification. Adipic acid
production, and possibly nitrogenous explosives, seem to produce
emissions of N20. In particular, we have calculated that in
explosive decomposition of ammonia nitrate under dry conditions,
about 9% of the nitrogen is converted to N20 [Axtell 1993]. This
implies a total output of 0.053*44/28= 0.083 MMT of N20 (mainly
from the mining and quarrying sectors, SIC 10,14. Nitrous oxide, a
potent "greenhouse gas", is a major source of concern for the
global environment.
7
4. Sulfur-based Chemicals
Sulfuric acid (37.7 MMT in 1988) is derived from elemental sulfur
(11.584 MMT in 1988). It is the product of the chemical industry,
subsector SIC 28193. It is the starting point for most sulfur based
chemicals. Sulfuric acid is also recovered from petroleum
refineries and from copper, zinc and lead smelters (1.125 MMT
S-content, 1988). What is recovered in petroleum refineries is
mostly also used there (0.786 MMT S-content), but over half of the
spent acid from refineries is reclaimed: 0.43 MMT (S) . While
recycling is relatively efficient, the annual consumption
represents very substantial makeup requirements to compensate for
losses that cannot be economically recovered. These waste streams
are not well-characterized, but some go to the air as SO 2, some
probably are in the form of insoluble sulfates or organic
complexes. These end up mostly in landfills. However, some refinery
wastes go to waterways.
Similarly, acid recovered from copper smelters is mostly used for
the so-called hydro- metallurgical process of copper and other
non-ferrous ore leaching (0.543 MMT S). In the case of copper
mines, some copper sulfate is recovered from the leach piles, and
this is recycled, but much of the leaching acid remains in the ore
heaps where it presumably reacts with other minerals and remains as
insoluble sulfates.
By far the largest use of sulfuric acid is for phosphate rock
processing (8.404 MMT S). This is the phosphate fertilizer
subsector, SIC 28734. The primary product is wet-process phosphoric
acid, which is the source of phosphate fertilizers. The sulfur in
this case ends up as calcium sulfate ("phospho-gypsum") and is
discarded as waste. This is recoverable in principle and could be
used for building materials (e.g. plaster board) but is not
recovered for use in the U.S. due to the presence of significant
radioactivity. Another important use of sulfuric acid is in the
sulfate (Kraft) pulping process (0.288 MMT S). In the case of
paper/pulp mills there are increasing efforts to recover sulfuric
acid and/or sodium sulfate from the waste stream ("black liquor")
but the annual makeup requirements represent that which is
discharged, either into the air as SO 2 or H2S, or into streams and
rivers as sulfates and complex sulfonated organic materials. A
detailed sulfur balance accounting for most sulfur uses is
summarized in Figure 2. below [see also Ayres & Norberg-Bohm
19931
The "pickling" process for cleaning rolled steel prior to
galvanizing or tin-plating uses sulfuric acid. This process
accounted for 0.074 MMT S, In the case of steel pickling, the waste
is mostly ferrous sulfate, which is recoverable but has few
markets. Sulfuric acid containing an additional 0.024 MMT S was
used by other metallurgical processes (such as plating). Automotive
batteries account for a further 0.051 MMT S. The above items, plus
exports, add up to 11.162 MMT of embodied sulfur, or >90% of the
total. The remainder, 1.172 MT of embodied sulfur is either used
elsewhere in the chemical industry or for unidentified non-
chemical purposes.
Excluding phosphates and sulfuric acid itself, about 2.169 MMT (S)
was used directly by other parts of the chemical industry, R-28. Of
this, 0.684 MMT was used as elemental sulfur and the rest as
sulfuric acid. Of the sulfur used in chemicals, 0.566 MMT (S) was
used to produce ammonium sulfate fertilizer, either directly or
indirectly as a by-product of other chemical processes using
sulfuric acid (e.g.caprolactam — a nylon monomer — and hydrogen
cyanide). Apart from ammonium sulfate fertilizer, aluminum sulfate
and some detergents (and
8
L 1376 • (378)
Petroleum products
1028 (508)
Non-ferrous metals 49 (16)
Non-ferrous mining 'SIC 10
Uranium, vanadium 104 (34)
Copper ores
1490 (487)
(nn) = thousand metric tons sulfur content
* except for sulfuric acid manufacturing
Other Data source: [USBuMines 1989]
ores 67 (22)
Figure 2: Materials Balance of Sulfur-based Chemicals Production in
the U.S. 1988 (1000 MT)
9
R.U. Ayres & LW. Ayres November 12, 1993
sulfuric acid itself) very few final products actually contain
sulfur. This means most of the sulfur used within the chemical
industry is also dissipated within the chemical industry, becoming
a process waste of the industry.
Summarizing, 93% of elemental sulfur (excluding a small number of
"unidentified" uses) is consumed in the manufacture of sulfuric
acid. Of the latter, 89.5% (again, excluding unidentified uses) is
used within SIC 28, of which by far the largest single use — 68% of
all sulfuric acid — is in the treating of phosphate rock to
manufacture phosphoric acid. Most other uses are dissipative, and
the final form of waste is normally either insoluble calcium
sulfite/sulfate, soluble (dilute) sodium sulfate, or SO 2 emitted
to the air. Only a small amount of sulfuric acid is recycled (about
0.864 MMT S, or 7.3%), also mostly from petroleum refineries and
synthetic materials.
It is probably appropriate to note again that 7.33 MMT of sulfuric
acid is consumed by the non-fertilizer part of the chemical
industry (R-28), of which 2.17 MMT was embodied sulfur. About 0.566
MMT (S) can be accounted for in ammonium sulfate fertilizer (SIC
28731). Of the rest, 0.17 MMT was embodied in aluminum sulfate (SIC
28196), 0.185 MMT was in the form of by-product sodium sulfate (SIC
28197), and about 0.1 MMT in the form of alkyl aryl sulfonate and
related chemicals for detergents (SIC 2843). A small amount of
sulfur is embodied in pesticides and other final products. We can
account for a total of 1.016 MMT of embodied sulfur, and there may
be another 0.1 to 0.2 MMT in other chemicals. However it seems
certain that, allowing for these flows, losses to the environment
as H2, SO2, calcium sulfite or calcium sulfate, probably account
for around 1 MMT (S) from sulfuric acid uses. We do not know the
fate of the surprising large amount of elemental sulfur (0.46 MMT)
used for "other" agricultural chemicals. However, it is possible
that much of this ends in products used on farms; we do not count
it as pollution.
5. Phosphorus-based Chemicals'
Phosphate rock is the only source of phosphorus chemicals,
including fertilizers. Phosphoric acid (P205) is the end product of
phosphate rock processing. U.S. production of crude phosphate rock
in 1988 was 224.1 MMT, which was reduced to 45.1 MMT fertilizer
grade superphosphate (13.833 MMT phosphorus pentoxide or P
2O5-content).6 The dry weight of material loss in the concentration
process was 179 MMT. A schematic diagram of the process is shown in
Figure 3. below. It is worth noting that approximately 1.8 kg of
water is used in the process for each kg of phosphate rock
processed, of which 1.4 kg is discharged (the remainder being
combined with calcium sulfate. Phosphate rock processing in 1988
consumed about 25.5 MMT of 100% sulfuric acid or 8.4 MMT of sulfur.
This was more than 68% of total U.S. sulfur consumption in 1988
(see "Sulfur").. The sulfur content of the acid is disposed of as
calcium sulfate ("phospho-gypsum") waste. Concentration wastes from
phosphate rock processing in 1988 were approximately 204.5 MMT,
dry, plus about 400
5 We thank Tom Llewellyn of the U.S. Bureau of Mines for his
assistance. He is not responsible for any
remaining errors.
6 Phosphorus pentoxide dissolved in water is phosphoric acid, the
active ingredient in most phosphate fertilizers (e.g.
"superphosphates"), not used, generally, in pure form.
10
Gypsun
3000
1120 1418 CaSO4 2H20 1359 Si02 75 P205 34 H2SiF6 5 Other 109
Phosphoric acid H 3PO455} H3PO4 409 to
A
H2SiF6 1 H2SO4 1 v,.
Other 3
R.U. Ayres & LW. Ayres November 12, 1993
MMT of water, for a total of roughly 600 MMT. Evidently this
process also accounts for virtually all of the waste from the
entire fertilizer sector in the U.S. and most of the waste from the
whole chemical industry.7
Among the major by-products of this processing was 46,565 MT (0.047
MMT) of fluosilicic acid, with a fluorine content equivalent to
90,000 MT of fluorspar.
Fertilizers (SIC 2873) now account for close to 95% of all
phosphorus used. Of the 1988 production (13.833 MMT P205
equivalent), exports — mostly as ammonium phosphates — accounted
for 2.608 MMT, leaving 10.549 MMT for domestic consumption. Most of
this, 9.329 (88.4%) was converted into "wet process" phosphoric
acid (H 3PO4). Elemental phosphorus production in the U.S. in 1988
was 0.32 MMT (0.79 MMT P 205 equivalent), of which 85% is
reconverted to "furnace grade" phosphoric acid for chemical
manufacturing. Some of this goes back into fertilizer (to make
triple superphosphate), but about half of it was used to
manufacture sodium tripolyphosphate (Na 5P30/0), a detergent
builder. Production of this chemical in 1988 was 0.497 MMT (P).
Some phosphoric acid is used in the food and beverage industry. A
minor but growing use of phosphorus is in the manufacture of
lubricating oil additives such as zinc dithiophosphate, which
starts from phosphorus pentasulfide (made by direct reaction of
phosphorus metal and elemental sulfur). This use accounted for
0.015 MMT of phosphorus metal in 1974; it is probably greater
today.
The starting point for organic phosphate synthesis is phosphorus
trichloride (PC1 3). Production figures are not published, but on
the basis of absorbing 1% of chlorine output (see chlorine below),
we can conclude that about 0.03 MMT of phosphorus metal would have
been used. The trichloride is later converted to phosphorus
oxychloride (POCI 3) by direct reaction with chlorine and P205. The
oxychloride, in turn, is the basis of organic phosphate esters that
now have many uses. The most important of them is the plasticizer
tricresyl phosphate (TCPP). Phosphate esters are also used as flame
retardants and fire resistant hydraulic fluids. Such phosphate
esters totalled .043 MMT (P205) in 1990 [USITC 1991]; detailed data
for each chemical are not published, but the phosphorus content is
rather small (8.5% in the case of TCPP). TCPP is also used as a
gasoline additive. All uses are, of course, dissipative. Uses of
phosphorus are summarized in Figure 4. In most cases the
dissipation does not occur within the chemical industry itself (SIC
28), however.
6. Chlorine-based Chemicals
Chlorine chemicals are mostly produced from elemental chlorine.
They are the major product (with sodium carbonate) of the
chlor-alkali industry, SIC 2812. The exceptions are sodium chloride
(salt or halite) used as such for a variety of purposes including
snow removal, for cattle feeding and food processing, calcium
chloride from brine and used for snow removal, potassium chloride
(sylvite), used as potash fertilizer. One other chemical, sodium
chlorate
It is important to emphasize this fact; to associate an
undifferentiated "waste coefficient" with the "fertilizer sector"
as a whole, in a country without a phosphate rock processing
industry, would be grossly misleading. By the same token, it must
be recognized that most countries (such as Morocco and Algeria)
with significant phosphate rock processing sectors do not bother to
control fluoride emissions, still less recover the fluorine for
beneficial use. In most of the world this industry is extremely
hazardous to workers and nearby residents.
12
superphosphate 334 3.6%
Other 155 1.7%
Normal superphosphate 325 2.7%
1988
R.U. Ayres & LW. Ayres November 12, 1993
(for bleaching pulp) are manufactured from salt.
Elemental chlorine together with sodium hydroxide (NaOH) — caustic
soda — are coproduced by electrolysis of sodium chloride (salt),
mainly in the form of brine. In 1988 U.S. salt production was 35
MMT. U.S. chlorine output in 1988 was 10.21 MMT, plus 9.77 MMT of
sodium hydroxide. The wastes from this process (mostly spent
brines) amount to about 15% of the weight of the products, or about
3 MMT. Major uses of chlorine include chemicals manufacturing
(chlorinated solvents, plastics and other chemicals), water and
sewage treatment (5%), pulp & paper bleaching (15%), titanium
dioxide manufacturing (3%), and miscellaneous (2%). Virtually all
uses of chlorine are dissipative, with the major exception of PVC,
which is used for structural purposes (e.g.water and sewer pipes,
siding, window frames, calendered products and bottles). PVC
accounts for about 20% of elemental chlorine output.
The chlorine used for bleaching in the pulp and paper industry has
become a very contentious subject in recent years, due to the
discovery of dioxin in bleached paper products. As a consequence —
whether justified or not — this bleaching process is rapidly being
phased out in Europe and North America.
Roughly 75% of elemental chlorine output (7.65 MMT) is absorbed by
the chemical sector (Figure 5.). By far the biggest use is ethylene
dichloride or EDC (40.5%), which is the starting point for PVC (EDC
-> vinyl chloride -> PVC). Other major uses are chlorinated
methanes (8%), chlorinated ethanes (5%), hydrochloric acid by
direct chlorination of hydrogen (3%), phosgene (2%), and
chlorinated benzenes (1.5%). Some elemental chlorine is used to
manufacture inorganic chemicals, mostly bleaches — e.g. calcium and
sodium hypochlorite- (7.8% of C1) and phosphorus trichloride (1%).
Most Hcl (about 91%) is recovered as a by- product of one of the
chlorination processes. However, we suspect that as much as a third
of the total flux of Hcl (including the recovered portion) is
eventually dissipated within the chemical industry, in the sense
that the chlorine is not exported from the sector embodied in a
product. This would account for about 10% of all chlorine, or 1
MMT, more or less.
Except for PVC, most chlorinated chemicals are dissipated in use.
Most of the uses (e.g. of solvents) are not in the chemical
industry itself. However, chlorination processes are relatively
inefficient. Hence, a fairly large proportion (5%-10%) of the input
chlorine eventually becomes a production waste in manufacturing
other chemicals. However, we cannot be more precise without more
detailed analysis.
7. Sodium-based Chemicals'
In 1988, as noted above, 9.77 MMT of sodium hydroxide was produced
as a byproduct of chlorine production in SIC 2812. Major uses of
sodium hydroxide are in the chemical industry (46%), in the pulp
and paper industry (16%), the preparation of alumina by the Bayer
process and to make synthetic cryolite for aluminum manufacturing
(6%), petroleum refining (6%),
8 We are grateful for assistance by an anonymous reviewer from the
U.S. Bureau of Mines. He is not responsible for any remaining
errors.
14
TEL
sue% percent of CI production Name (no%) = percent of precursor yy
= CI-cooteat (MMT) Data sources: XS= 1992, 1E1 1991]
Vinyl chloride
CH2-CHCI 2.331
Solvents DDT
Other (222%) Solvents (33.6%) Chemicals (10.4%) Aerosols (16.8%)
Export (16.9%)
Other (1.9%) CFCs (92.7%) Export (5.3%)
Chloro- form
CCI3H 0.211
CFCs (99%)
rbon tetra-
melting
0.240
Other (24.0%) Solvents (49.9%) Chemicals (7.4%) Aerosols (11.1%)
Export (7.6%)Natural-0 25sources
Sodium chlorate NaCIO3 0.082
Sodium chloride Nag
ELE- MENTAL IlLORIN
10.212 0.4 Export
aniline chloro- benzene C2H5CI 0.030
Poly- chloroprene (neoprene)
!dor 'notion
411. H2
5% -410.. Water treatment TiO2
Fer- rous
Chemical Industry Waste: A Materials Balance Analysis
Figure 5: Materials Balance of Chlorine-based Chemical Production
in the US 1988 (MMT Cl-content)
15
R.U. Ayres & LW. Ayres November 12, 1993
dyeing of textiles (4%), rayon manufacturing (3%), soap and
detergents (3%), cellulose acetate (2%) and miscellaneous purposes
(14%), including exports [Lowenheim & Moran 1975, p.
742].
Taking into account alumina, rayon and soap/detergent manufacturing
60% (6 MMT) is probably absorbed by the chemical manufacturing
sector. Except for the rayon industry (which produces sodium
sulfate as a by-product) and the detergents, virtually all uses of
sodium hydroxide are dissipative within the chemical sector.
Virtually none is actually embodied in products. Thus, we estimate
that dissipative losses of caustic soda within the industry account
for about 5.6 MMT.
To cite a few examples, the manufacture of ammonia requires about 4
kg per metric ton of product; the extraction of benzene from light
oils requires 14.5 to 28.5 kg per metric ton; glycerine requires
100 kg per metric ton; caprolactam, the monomer for nylon 6,
requires 125 kg of NaOH per MT of product; 63 kg of NaOH is
required per MT of adipic acid from cyclohexane. In none of these
cases is the sodium embodied in the product. In virtually all
cases, NaOH is used to regulate the acidity or alkalinity (pH) of
the reaction.
Soda ash is another alkali sodium chemical (SIC 2812) that was once
manufactured synthetically (by the Solvay process). However sodium
carbonate is now extracted from brines and evaporite deposits,
called trona. According to one source [Lowenheim & Moran 1975
p. 709] 1.5 T of ore are required to produce 1 T of product. This
implies 0.5 unit of waste for each unit of product. However, the
process is really more complicated. Process waters are later used
for solution mining and recycled for additional soda ash. Purge
liquors are also sold and not considered "waste", although both
these streams must eventually return to the environment in some
form. However, it appears that insolubles (returned to the mine)
only account for 10% of the ore mass. Hence the 8.7 MMT of sodium
carbonate produced in 1988 resulted in 0.9 MMT of solid waste, plus
an unknown amount of process water.
Out of the U.S. production in 1988, 2.117 MMT was exported (net).
Domestic uses of sodium carbonate accounted for 7.6 MMT. It is used
in glass manufacturing (50%), alkaline cleaners (12%), pulp &
paper (2%), flue gas scrubbing (2%) and water treatment (1%). The
glass industry uses about 0.28 metric tons of soda ash per metric
ton of glass. However, around 22% (1.9 MMT) is used in other
chemical manufacturing. Important sodium chemicals based on sodium
carbonate as a feedstock include sodium cyanide, sodium silicates
(0.737 MMT, Si02 equivalent) sodium dichromate, etc. Most of these
sodium containing chemicals are dissipated in use, but not within
the chemical sector SIC 28.
8. Pigments and Other Metal-based Chemicals
Compounds of iron, chromium, copper, lead, titanium and zinc also
have important chemical uses, especially for pigments. Titanium
dioxide has already been mentioned. It is the most important
metallic pigment — being used for most exterior paints, as well as
in paper. U.S. production in 1988 was 0.926 MMT, mostly from
ilmenite. For each ton of TiO 2 produced, 1.2 tons of waste is
generated, implying 1.11 MMT of waste from this source in
1988.
16
Partial oxidation
Methanol CII3011 Acetaldehyde
MTBE xx% = % 1988 U.S. production (MMT)
Data source: [Gaines & Shen 1980]
Silicones
Solvents
Figure 7: Product Flows from Methanol (U.S.A. 1978)
18
•••..
1.4%
3%
15% thanolamines
19
10% 15%
16% Acrylic
fibers 6.7%
Methacrylates 2.8%
Data source: [Gaines & Shen 19801
0%
Ammonia
Figure 9: Product Flows from Propylene (U.S.A. 1978)
20
Imports
Imports
yy% = percentage of precursor Data source for Benzene &
Toluene: [IN 19911 Data source for other percentages: [Mitteihauser
Corp. 19801
Chemical Industry Waste: A Materials Balance Analysis
Figure 10: Product Flows from Toluene & Benzene (U.S.A. 1988,
1000 MT)
21
R.U. Ayres & LW. Ayres November 12, 1993
Net imports of product Stock changes, losses, other outputs not
allocated y Process losses of input
x
Figure 11: Product Flows from Xylene (U.S.A. 1988, 1000 M.7)
22
Chemical Industry Waste: A Materials Balance Analysis
Iron oxide is a red pigment. Ferrous chloride is used as a soil
conditioner. Ferrous sulfate is used to make iron oxide, to
manufacture ferrites, as a catalyst, in sewage treatment, etc.
However it is really a waste product of the steel industry (2-4
MMT/yr) and a disposal problem. Copper sulfate (0.0342 MMT) is the
basis of most copper chemicals (fungicides, algicides, pesticides,
catalysts, flotation reagents, etc). Chromic acid, sodium chromate
and sodium dichromate (0.145 MMT, Cr203 equivalent) are the basis
of pigments, algicides, leather tanning agents, and chrome plating
chemicals. Lead and zinc sulfates and oxides are primarily
pigments, but also the basis for other lead and zinc chemicals.
They are still produced in fairly large, though decreasing,
quantities. Tetraethyl and tetramethyl lead were once produced in
very large quantities as a gasoline additive, although production
and use have declined sharply since 1970 as a result of
environmental regulation. All uses are dissipative. However, we do
not discuss chemical uses of metals further in this paper.
9. Hydrocarbon-based Chemicals
Most organic industrial chemicals are based on petrochemical
(hydrocarbon) feedstocks. There are three categories: paraffins
(methane, ethane, propane, butane), olefins (ethylene, propylene,
butylene, butadiene), and aromatics (benzene, toluene, xylenes and
naphthalene). These feedstocks, totalling 32.44 MMT in 1988 were
derived from natural gas liquids (22.46 MMT), refinery off-gas
(1.12 MMT) and naphtha (8.864 MMT) [OECD 1991]. All downstream
products, with minor exceptions, are derived, in turn, from these
starting points (or the inorganic intermediates discussed above).
The structure of the petrochemical industry is outlined in Figures
6.-12.
However, for purposes of the present analysis it is conve- nient to
redefine the list of "feedstocks" to include metha-
SBR rubber
from natural gas, or methane), Butadiene
plus chlorine. It is convenient loon
to exclude the alkanes: meth- 20% 20%ane, ethane, propane and bu-
Other
(including(intane, since the latter — from hexa-
natural gas or petroleum refin- methylene diamine),
eries — are almost entirely nylon
used as fuels or converted immediately to olefins. It is the latter
which we consider to be Figure 12: Product Flows from Butadiene
(U.S.A. 1978) feedstocks. To illustrate, hydro- carbon feedstocks
like ethane (C2H6), propane (C3H8), and butane (C4H8) are first
dehydrogenated to ethylene (C 2H4), pro- pylene (c314), butadiene
(C4H6), and butylene (C4H8) respectively. Large amounts of
hydrogen-rich off-gases are produced at this stage, but these are
mostly consumed for fuel. In some cases the hydrogen is captured
and used for downstream hydrogenation steps.
-----",7*-Poly- butadiene
23
R.U. Ayres & LW. Ayres November 12, 1993
The vast majority of downstream chemical processes involve further
dehydrogenation, hydrogenation, recombination (e.g. alkylation,
reforming), ammoniation, chlorination or (partial) oxidation or
dehydration. Most hydrogen produced in the organic chemical
industry itself is combined with oxygen and lost as water vapor.
Carbon losses (as HC, CO or CO2) are relatively minor, perhaps on
the order of 10% overall. On the other hand, chlorine, nitrogen
(from ammonia) and oxygen (from the air) are added. A great deal of
sulfuric acid and other sulfur chemicals are used in the organic
chemicals sector, but very little is embodied in final products of
the sector, mainly in pesticides and detergents.
10. Materials Balance for R-28
The rest of the chemical industry (R-28) consists of two
components. One is the manufacture of hydrocarbon feedstocks,
listed above, insofar as they are produced within SIC 28 and are
not products of SIC 29 (petroleum refining). As noted in the
previous section, the first step in conversion from alkanes to
olefins is dehydrogenation followed by separation (e.g. by
distillation). Wastes from these activities are quantitatively
small to insignificant, as a fraction of the inputs. This can be
verified quite easily from published process data, though not from
production/consumption data. The processes in question are highly
evolved and quite efficient.
The second component of R-28 consists of all other processes
leading to "finished" organic chemical products, not including
agrichemicals. Most of these products are organic. To estimate the
losses in this component of R-28 it is convenient to sum up the
"feedstocks" as defined above: total hydrocarbon inputs in 1988
were 38.73 MMT, plus 3.693 MMT of methanol (about half imported),
and 7.5 MMT of chlorine.
As regards chemicals, the situation is very confusing due to
imports, exports and byproducts at various stages. As regards
ammonia, over 80% goes to fertilizer and feed. We can account for
1.9 MMT (N-content) in terms of major N-based chemicals, and 1.5
MMT (N-content) is unaccounted for (but some of this is outside the
chemical sector). We estimate that c. 3 MMT (N-content), or 3.6 MMT
ammonia, was consumed in R-28 in 1988.
We can account for 1.85 MMT of oxygen embodied in the methanol,
plus about 3.5-4 MMT of oxygen used in downstream oxidation
processes. Much of this comes from the acids HNO3 and H2SO4). We
can account for about 1.6 MMT (S) entering R-28 in 1988, which
would correspond to 4.9 MMT H2SO4, of which 3.2 MMT was oxygen. We
must also allow for 5.5 MMT of NaOH used in processes. Details are
too complex for a complete explanation here, since most chemicals
are used in the production of other chemicals, and oxygen can be
added or lost at any point. However major oxygenation stages
include production of ethylene and propanol oxides; production of
ethanol, isopropanol and butanol, production of phthalic anhydride;
production of terephthalic acid (TPA). Also, oxy-chlorination of
ethylene to EDC. Summing up, we can account for total inputs — as
we have defined them — adding to close to 64 MMT. Little or no
sodium or sulfur was embodied in organic chemical products.
The major tonnage outputs (sales) of the organic chemicals industry
— indeed, for all practical purposes, the entire chemical industry
— are listed in Tables 2. - 6., for 1990 [USITC 1991]. (We were
unable to find 1988 data in as much detail, but we note that
the
24
Table 2: Plastics & Resins U.S. Production & Sales
1990
1000 metric tons
Alkyd resins 349.0 264.3 phthalic anhydride 290.5 227.8
Epoxy resins, unmodified 315.9 230.0 Melamine-formaldehyde resins
109.2 89.6 Phenolic & other tar acid resins 943.9 621.9
Polyester resins, unsaturated 537.2 512.7 Polyether & polyester
polyols for urethanes 760.6 561.4 Polymethane elastomers &
plastics 114.2 98.2 Urea formaldehyde resins 1103.9 742.7 All other
thermosetting resins 75.6 56.2
Thermoplastic resins, total 25201.3 22093.9
Acrylic resins 683.8 610.8 PMMA 304.7 189.2
Engineering plastics 526.1 470.5 Polyamide resins, total 288.6
279.9
nylon type 261.9 253.9 Polyester resins, saturated, total 1598.4
1381.6
PET 1347.6 1147.7 Polyethylene resins, total 9070.9 8125.2
Polypropylene resins, total 3465.5 3146.7 Styrene plastics, total
4624.1 3648.9 Vinyl resins, total 4943.9 4430.3
PVC 4247.0 1379.1
Nylon 6, 6/6 1090.3 PAN 220.7 Other 1047.5 1379.1
Polymers, water soluble 309.7 264.7
Data source: [USITC 1992]
1000 metric tons
GRAND TOTAL 2233.1 1555.1
Ethylene-propylene type 2323 204.1 Polybutadiene type 349.2 175.6
Silicone type 94.7 68.4 SBR type 900.2 604.3 Thermoplastic
elastomers 256.8 167.0 All other 399.9 335.7
Source: [USITC 19921
1000 metric tons
Source: [USITC 19921
1000 metric tons
Cationic, total 844.6 744.3 Non-ionic, total
Source: [USITC 19921
Table 6: Miscellaneous Chemicals U.S. Production & Sales
1990
1000 metric tons
Chelating agents 137.2 101.5 Fuel additives 4224.7 1935.6
MTBE 4029.7 1800.1 Lube oil & grease additives 387.2 343.6
Textile chemicals (ex. surfactants) 22.4 19.8 Urea 5455.7
5270.5
for feed 573.7 551.2 for liquid fertilizer 1141.0 1037.2 for solid
fertilizer 3741.0 3682.1
N.E.C. 2096.6 1421.9
grand totals for the two years were nearly identical). Major items
include thermosetting resins (3.18 MMT), thermoplastic resins
(22.094 MMT), synthetic elastomers (1.555 MMT), polymers for
synthetic fibers (1.379 MMT), rubber processing chemicals (0.136
MMT), plasticizers (0.827 MMT), surfactants (2.718 MMT), gasoline
additives - mostly MTBE - (1.936 MMT), Tube oil and grease
additives (0.344 MMT), antifreeze (0.9 MMT) and CFC's (0.3 MMT).
Pesticides, fungicides, etc. accounted for 0.442 MMT. Including a
collection of miscellaneous small items, such as dyes, medicinals,
pigments, chelating agents, water soluble polymers, etc. the above
adds up to about 37.65 MMT.
In addition, one must allow for solvents and organic explosives not
elsewhere classified, for which use data is not specifically broken
out, Table 7. These include acetone (16% to solvent use, or 0.16
MMT), isopropyl alcohol (40% to solvent use, or 0.27 MMT),
methylene chloride (paint remover), methyl chloroform (dry
cleaning), trichloroethylene (degreasing) and perchloroethylene
(dry cleaning). The above-listed chlorinated solvents, in toto,
added up to around 0.8 MMT in 1990. Altogether, including solvents,
final outputs of the organic chemicals industry tentatively
accounted for very close to 39 MMT in 1988.
Comparing product output weights - 39 MMT - with inputs accounted
for (-64 MMT) implies material losses of -25 MMT give or take a
little. See Figure 13. This translates to 39 MMT of aggregate
material inputs in R-28. (Obviously process water is not included).
This loss is not as difficult to explain, as it might at first
seem, given the fact that most products require a sequence of
several processes, each of which might have a yield efficiency
(based on inputs) of 80-95%. A sequence of three to four such
processes would easily account for a total loss of the order of
39%. On the other hand, our estimate for the entire R-28 is
considerably less than the 49.6 million tons estimated by SAI for
the plastics and resins sector alone. (Again, one is forced to
conclude that SAT's numbers include a good deal of process
water).
27
R.U. Ayres & L.W. Ayres November 12, 1993
Table 7: End Use Organic Chemical Products US. Production &
Sales 1989, 1990
1000 metric tons
TOTAL 50338.8 39107.8
Dyes 174 117.0 146 104.0 Organic pigments 50 53.0 43 45.0
Medicinals 130 144.0 204 107.0 Flavor & perfume materials 64
60.0 38 37.0 Rubber processing chemicals 176 179.0 129 136.0
Pesticides 572 557.0 461 442.0 Thermosetting resins 4309.5 3177.0
Thermoplastic resins 25201.3 22093.9 Polymers for fibers 2358.5
1379.1 Polymers, water soluble 309.7 264.7 Elastomers 2233.1 1555.1
Plasticizers 890.7 826.5 Surfactants 5848.7 2718.1 Antifreeze 920
900.9 900.0 CFC's 417 308.3 300.0 Solvents 1200.0 Chelating agents
137.2 101.5 Fuel additives 4224.7 1935.6 Lube oil & grease
additives 387.2 343.6 Textile chemicals (ex. surfactants) 22.4 19.8
Miscellaneous chemicals N.E.C. 2096.6 1421.9
Source. [USITC 19921
A more detailed analysis might enable us to estimate roughly the
allocation of emissions between gases vented to the atmosphere and
solids, mostly disposed of on land. As a guess, is not unreasonable
to assume that 3-5% of HC inputs end up in waterborne wastes, along
with inorganic residues such as chlorides and sulfates. On this
basis we estimate that organic wastes to watercourses range from
1-1.5 MMT, although the numbers could be somewhat larger. On the
other hand, 22 MMT must be accounted for. If EPA's estimate of 3.9
million tons (3.6 MMT) emitted to air [USEPA 1986, 1988] is
correct, then about 15 MMT of wastes are being disposed of on land
or in wells. (Most of these wastes are probably hazardous.)
11. Conclusions
In summary, the major source of wastes within the chemical industry
consists of ore concentration wastes: from phosphates (204 MMT dry
+ 400 MMT water); from alumina (4.6
28
Chemical Industry Waste: A Materials Balance Analysis
MMT dry); from trona ore (1 MMT); from titanium dioxide processing
(1.1 MMT). The total of these is about 210 MMT, dry weight. Others
for which we have no quantitative estimates include potash pro-
cessing wastes and other evap- orate chemical wastes. There is no
way to reduce these quanti- ties as long as primary raw materials
are the source.
Oxygen from air 3.5 - 4.0
Hydrocarbon feedstocks 38.7
Methanol 3.7 Chlorine 7.5 Sulfuric acid 4.9 Caustic soda 5.5
Products 39
Wastes (mputed) 25
The second major category of chemical industry wastes con- Figure
13: Materials Balance for R-28 (U.S.A. 1988, MMT) sists of
materials used dissipa- tively within the sector, and not
incorporated into any product of the sector. The major examples are
sulfuric acid, of which about 3 MMT (1 MMT S) is lost, and caustic
soda (Na0H), of which as much as 5.5 MMT seems to be dissipated and
lost within the sector. A fair amount of hydrochloric acid is also
lost; we estimated 1 MMT. Total of dissipative losses of reagents:
9.5 MMT, mainly in R-28.
The third category of losses within the chemical industry consists
of process losses of unreacted feedstocks (not including the acids
or alkalis) or combustion wastes. We estimated above that perhaps
0.5 MMT (3%) of fixed nitrogen might be lost in chemical conversion
processes. The 22 MMT difference between feedstock weight and
product weight is a combination of process waste and dissipative
uses. Most of these losses probably occur in sector 2815 (cyclic
intermediates), 2818 (organic chemicals) and 2821 (plastics and
resins). In tonnage terms, the major emissions are probably water
vapor (from combustion) and other combustion products, notably CO2
and CO, plus unreacted volatiles of various sorts. A significant
fraction of "off-gases" is used for fuel, as noted above, but some
is simply flared, vented, or lost as "fugitive emissions". The
remaining residuals are due to incomplete reactions. In terms of
hydrocarbon inputs, we estimate an average loss of 5% per process
stage, although most of this is oxidized to CO2 and H20.
Non-oxidized residuals are typically either volatile hydrocarbons
in the vapor phase or heavier hydrocarbons or chloro-carbons in the
aqueous phase. The former is consistent with EPA's estimated 2 MMT
of VOC emissions from the organic chemical industry.
It seems evident that the rather large estimates of "dry" chemical
wastes by EPA [USEPA 1986, 1988] cannot actually represent dry
weight. For example, EPA assigns 28.9 million tons of dry waste to
the inorganic chemical sector (SIC 2812, 2819) and 49.6 million
tons to the plastics and resins sector (SIC 2821). These must
consist largely of contaminated process water, or wet sludge. This
is a very misleading measure of waste emissions, since it is
relatively easy, in many cases, to reduce water-borne emissions by
increased internal recycling of process water.
29
R.U. Ayres & L.W. Ayres November 12, 1993
Opportunities for waste reduction in the chemical industry are not
particularly great, as the industry is now structured. By far the
greatest quantity of wastes arises from processes of concentrating
ores (notably phosphate rock). Only substitute sources or reduced
demand for chemical products could reduce these waste flows. (In
fact, they will increase in future as ore grades decline). Such
opportunities as do exist are of two kinds. In the first place,
process wastes are reduced if yields are improved or (even better)
if new reactions are developed that skip a step in the process
chain. Second, wastes are reduced by definition if uses can be
found for them. For instance, it would be helpful if new uses could
be discovered for calcium chloride, ferrous sulfate or calcium
sulfate, all of which are in excess supply.
References
[Allen & Behmanesh 1992] Allen, David T. & Nasrin
Behmanesh, "Non-Hazardous Waste Generation", Hazardous Waste &
Hazardous Materials 9(1), Winter 1992 :91-96.
[Axtell 1993]
[Ayres & Norberg-Bohm 1993] Ayres, Robert U. & Vicki
Norberg-Bohm, Industrial Metabolism of Sulfur, Working Paper,
INSEAD, Fontainebleau, France, 1993.
[Ayres et al 1993] Ayres, Robert U., Leslie W. Ayres & Vicki
Norberg-Bohm, Industrial Metabolism of Nitrogen, Working Paper,
INSEAD, Fontainebleau, France, 1993.
[Gaines & Shen 1980] Gaines, L.L. & S.Y. Shen, Energy &
Materials Flows in the Production of Olefins & Their
Derivatives, (ANL/CNSV-9), Argonne National Laboratory, Argonne,
IL, August 1980. [for U.S. Department of Energy, Washington,
DC].
[IEI 1991] Industrial Economics, Incorporated, Materials Balance
Profiles for 33150 Chemicals, Draft Report (EPA Contract
68-W1-0009), Industrial Economics, Incorporated, Cambridge MA,
September 1991.
[Lowenheim & Moran 1975] Lowenheim, Frederick A. &
Marguerite K. Moran, Faith, Keyes, & Clark's "Industrial
Chemicals", Wiley-Interscience, New York, 1975. 4th edition.
[Mittelhauser 1979] H.M. Mittelhauser Corporation, Energy/Material
Flows Associated with Cyclic Petrochemicals, (ANL/CNSV-TM-56), H.M.
Mittelhauser Corporation, Downers Grove, IL, October 1979. [for
Argonne National Laboratories].
[OECD 1991] Organization for Economic Cooperation &
Development, The State of the Environment, OECD, Paris, 1991.
[OECD-a 1991] Organization for Economic Cooperation &
Development, Environmental Indicators - A Preliminary Set, OECD,
Paris, 1991.
[SAI 1985] Science Applications International Corporation, Summary
of Data on Industrial Nonhazardous Waste Disposal Practices, EPA
Contract (68-01-7050), Science Applications International
Corporation, Washington DC, 1985.
[USBuMines 1975] United States Bureau of Mines, Mineral Facts &
Problems, United States Government Printing Office, Washington DC,
1975.
[USBuMines 1977] United States Bureau of Mines, Minerals Yearbook,
United States Government Printing Office, Washington DC,
1977.
30
Chemical Industry Waste: A Materials Balance Analysis
[USBuMines 1988] United States Bureau of Mines, Minerals Yearbook,
United States Government Printing Office, Washington DC,
1988.
[USBuMines 1989] United States Bureau of Mines, Minerals Yearbook,
United States Government Printing Office, Washington DC,
1989.
[USEPA 1986] United States Environmental Protection Agency Office
of Solid Waste, Waste Minimization Issues & Options, NTIS
PB-87-114369 (EPA-530-SW-86-041), United States Environmental
Protection Agency Office of Solid Waste, Washington DC, October
1986.
[USEPA 1988] United States Environmental Protection Agency, Solid
Waste Disposal in the United States, Report to Congress
(EPA-530-SW-88-011), United States Environmental Protection Agency,
Washington DC, 1988.
[USEPA 1991] United States Environmental Protection Agency Office
of Solid Waste, 1987 National Biennial Report of Hazardous Waste
Treatmen4 Storage & Disposal Facilities Regulated Under RCRA,
NTIS PB-87-114369 (EPA-530-SW-91-061), United States Environmental
Protection Agency Office of Solid Waste, Washington DC, 1991.
[USITC 1991] United States Trade Commission, Synthetic Organic
Chemicals 1991, United States Government Printing Office,
Washington DC, 1991.
[USITC 1992] United States Trade Commission, Synthetic Organic
Chemicals 1992, United States Government Printing Office,
Washington DC, 1992.
31