Chemical Engineering Science 54 (1999) 1959}1973

Environmentally friendly processes

Carmo J. Pereira*
DuPont Engineering, Wilmington, DE 19898, USA
Abstract
Environmental concerns have raised public awareness of environmental issues and are driving forces for regulation. The impact of
regulation on the cost of production is expected to become important in determining the international competitiveness of the US
chemical industry. In response to cost pressures, industry has launched a number of initiatives aimed at improving e$ciency and
reducing environmental impact. Some of these environmental success stories are receiving increased national attention due to
programs such as the Presidential Green Chemistry Challenge Awards Program. In addition to traditional metrics for evaluating
process performance, such as productivity, environmental considerations increasingly are important in process development.
Chemical processes evolve through life cycle phases, beginning with research, and then moving to process engineering, plant
operation, and eventually, decommissioning. The number of technology options available for reducing environmental impact are
highest early on in the life cycle and then decrease drastically. In contrast, costs associated with resolving an environmental problem
typically increase exponentially as the process matures and the scale of equipment gets larger. There is, therefore, a considerable
incentive to address and resolve environmental issues early in the life cycle.
Chemical reactions responsible for producing high value-added products are, in most cases, also responsible for generating
by-products and pollutants. New chemical and biochemical approaches are providing new reaction concepts. As in the development
of traditional chemical and petrochemical processes, reaction engineering, broadly de"ned as the "eld that quanti"es the engineering
aspects of chemically reactive systems, is providing enabling tools that accelerate the development of environmentally friendly
processes. Core reaction engineering methods are being utilized for kinetic modeling, reactor selection, scale-up, and design.
Meanwhile, the research frontiers are providing new reaction engineering tools, from computational chemistry to probe the nature of
catalytic active sites to computational #uid dynamics modeling for designing the internals of reaction-separation systems.
The long-term goal is to develop processes having 100% raw materials utilization, or zero waste. The near-term strategy for
controlling emissions is to institute pollution prevention programs and install cost-e!ective end-of-pipe technologies. These
technologies typically control generic classes of air pollutant emissions such as carbon monoxide, volatile organic compounds
(VOCs), nitrogen oxides (NO
V
) and sulfur oxides (SO
V
). Technologies for treating wastewater are also available. Over time, a shift in
focus is expected, from mere compliance to a point where environmentalism, like safety, is fully integrated into the corporate culture.
The present paper discusses the role of reaction engineering in the development of environmentally-friendly processes. Select examples
of processes, from the author's experience, that bene"t from reaction engineering will be presented. 1999 Elsevier Science Ltd. All
rights reserved.
Keywords: Reaction engineering; Environment; Chemical processes; Life cycle analysis; Pollution control; Research trends
1. Introduction
Industrial growth on a global scale has spawned envir-
onmental concerns on a global level. These concerns
transcend local problems, such as point-source pollutant
emissions and site remediation, to truly global problems,
such as ozone depletion and global warming. They have
*Tel.: 001 302 774 2337; fax: 001 302 774 2457; e-mail: carmo.j.
[email protected].
raised public awareness of environmental issues and are
driving forces for regulation. The impact of regulation on
the cost of production is expected to become more im-
portant in determining the chemical industry's com-
petitiveness than in earlier years when the US industry
and producers in a few countries were dominant in world
markets (O$ce of Technology Policy, 1996).
The US chemical industry is important to the nation's
economic health. In 1994, the industry directly supported
nearly a million jobs and shipped more than 70,000
products valued at $341.3 billion. On a value-added
0009-2509/99/$} see front matter 1999 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 0 9 - 2 5 0 9 ( 9 8 ) 0 0 5 0 4 - 1
Fig. 1. Schematic of the industrial ecological approach over time
(Tibbs, 1993).
basis, the industry provided about 1.9% of the gross
domestic product (GDP). Chemical industry exports
accounted for more than $1 out of every $10 of US goods
exported in recent years. The industry trade surplus, of
$18.4 billion in 1994, has increased to $20.5 billion in
1997 (O$ce of Technology Policy, 1996; Storck, 1998). In
contrast, the 1994 US environmental industry supported
1.27 million jobs and reported revenues of $172.5 billion.
Total 1993 pollution abatement and control expendi-
tures were reported at about $110 billion: $31.9 billion
for air, $39.0 billion for water, and $38.8 billion for solid
waste (US Department of Commerce, 1997). It is gener-
ally believed that environmental issues cost the industry
approximately 3% of sales revenue, or about $10 billion
each year (Carberry, 1998).
Regulatory and cost pressures are driving forces for
developing environmentally-friendly and energy-e$cient
products and processes. A long-term framework for de-
veloping an environmental agenda has been proposed. In
the concept of &&industrial ecology'', industrial plants are
designed as if they were a series of interlocking man-
made ecosystems interfacing with the natural global envi-
ronment. The need for such systems is driven by the fact
that the scale of industrial production is now so great
that even formerly-tolerable emissions, such as carbon
dioxide, are becoming a threat. For example, industrial
#ows of nitrogen and sulfur are reported to be equivalent
or greater than natural #ows, while #ows of metals (such
as lead, cadmium, nickel, vanadium, and the like) are
twice natural #ows (Tibbs, 1993). Since the long-term
survival of the industrial systemdepends on the function-
ing of a natural global ecosystem, the environmental
agenda must allowthe two systems to coexist. The indus-
trial ecology approach, over time, will require a shift in
focus from mere compliance to a point where environ-
mentalism, like safety, is fully integrated into the corpo-
rate culture (see Fig. 1).
Industry recognizes the need to proactively reduce
pollutant emissions rather than to react to increasingly
stringent regulations. A key approach for reducing emis-
sions is pollution prevention (Allen and Rosselot, 1997).
Companies are developing programs to improve e$ci-
ency in the use of materials, energy and water, and to
develop inherently-safer processes, distribution, and
products. For example, DuPont has committed to a 15%
reduction in energy required per pound of product by the
year 2000, from a 1991 base. In addition, the company
has documented a 20 billion gallons per year reduction in
in#uent water use since 1993, on a global basis (Chemical
Engineering, 1997). Along with traditional objectives
such as productivity and safety, environmental consider-
ations have become important.
In the US, green chemistry is receiving national atten-
tion. The Presidential Green Chemistry Challenge
Awards Program was established in 1995 to &&recognize
and support fundamental and innovative chemical
methodologies that are useful to industry and that ac-
complish pollution prevention through source reduc-
tion'' (US EPA, 1996}97). Awards during the past two
years are listed in Table 1. Environmentally-oriented and
&&lower-emissions'' research proposals are being sup-
ported by organizations, such as the American Institute
of Chemical Engineers, the Council for Chemical Re-
search, the Department of Commerce, the Department of
Energy, the Environmental Protection Agency, and the
National Science Foundation (see e.g., Beaver, 1996).
These initiatives will likely accelerate the development of
new tools and new chemical and process concepts.
ife cycle assessment (LCA), an environmental aud-
iting tool that quanti"es the environmental burdens of an
activity, can help a company reduce environmental risk
and thereby acquire a competitive advantage in the mar-
ketplace. A schematic of a life cycle framework is shown
in Fig. 2. LCA provides a methodology for the cradle-to-
grave analysis of materials and energy that includes raw
material extraction, the manufacturing process, product
use, and product disposal. It involves an inventory (of
wastes, emissions, raw materials and energy usage asso-
ciated with the entire life cycle of the product) and an
assessment of the environmental impact at each life cycle
stage. The methodology can be used to compare the
environmental impact of each of several possible options.
The data for such an analysis is often di$cult to obtain.
The cost of acquiring the data, however, often is o!set by
savings that result from identifying options that avoid
future environmental costs. LCA has been used typically
for products and not for processes.
The atom utilization (or by-products per pound prod-
uct) is an indicator of the raw materials e$ciency of
a chemical process. Atom utilization varies widely across
the chemical process industry. The basic chemicals indus-
try can be viewed as quite e$cient in its use of raw
materials with overall yields of over 90%. On the other
hand, when viewed on a wet basis, the amount of waste
generated in chemical manufacturing is high, pointing to
the need for lower overall water usage (Allen, 1996).
1960 C.J. Pereira/Chemical Engineering Science 54 (1999) 1959}1973
Table 1
US Presidential Green Chemistry Challenge Awards in 1996 and 1997 (US EPA, 1996}1997)
Type Year Title
Alternative synthetic 1996 The catalytic dehydrogenation of
Pathways Diethanolamine
1997 BHC Company Ibuprofen Process
Alternative solvents/ 1996 The development and Commercial
reaction conditions Implementation of 100 Percent Carbon Dioxide as an Environmentally
Friendly Blowing Agent for the Polystyrene Foam Sheet Packaging Market
1997 DryView'` Imaging Systems
Designing safer 1996 Designing an Environmentally Safe
chemicals Marine Antifoulant
1997 THPS Biocides: A new Class of Antimicrobial Chemistry
Small business 1996 Production and Use of Thermal Polyaspartic Acid
1997 Coldstrip'`, A Revolutionary Organic Removal and Wet Cleaning Technology
Academic 1996 Conversion of Waste Biomass to Animal Feed, Chemicals, and Fuels
1997 Design and Application of Surfactants for Carbon Dioxide
Fig. 2. Life cycle framework for clean manufacturing technology and
green design (O$ce of Technology Assessment, 1992).
Table 2
Some environmental process objectives
Higher product selectivity
High conversion
Improved energy e$ciency
Use of benign solvents
Raw materials substitution
Conversion of hazardous products to less hazardous products prior to
shipment
On-site production of hazardous materials
Conversion of by-products to products
Lower secondary emissions
Low aqueous waste
By-products per pound of product are much higher in the
pharmaceutical industry. Product manufacture, in this
segment, may require that a number of stoichiometric
organic synthesis steps be conducted in series. As a result,
the overall yield can be low even if the yield for each
individual synthesis step is high. Given the high-value
nature of pharmaceutical products, the economic driving
force for increasing selectivity has been relatively low.
New environmentally-oriented chemistry and processes
in this industry have been recently reviewed (Mills and
Chaudhari, 1997).
From an environmental standpoint, a process may
possess a number of desirable attributes (Table 2).
Clearly, as discussed above, higher atom utilization or
product selectivity is extremely desirable. Beyond the
obvious impact on production, conversion impacts costs
associated with product re"ning and recycle. The ability
to reduce or eliminate the use of hazardous rawmaterials
or solvents is an advantage. Higher energy e$ciency by
improving heat integration and introducing novel pro-
cess concepts is desirable. There is a high interest in
"nding new uses for byproducts and in developing pro-
cesses that convert by-products to useful products. There
is a trend away from the large-scale manufacture, trans-
portation, storage and handling of hazardous chemicals.
In some cases, the control of one pollutant can lead to
a secondary emission problem (e.g., the formation of
toxic chlorinated products during the catalytic oxidation
of halogenated hydrocarbons). Such secondary emissions
must be reduced or eliminated.
An optimization framework that includes the environ-
mental costs associated with a process is needed. This
involves "rst de"ning a process boundary and ac-
counting for raw materials, energy and wastes #owing
across the boundary. For example, the boundary for
a vinyl chloride monomer manufacturing process may
include processes for making ethylene and chlorine feed
(Stefanis et al., 1995). The next step is to de"ne an objec-
tive function, such as total process cost:
Minimize
,

G
C
G
or
Minimize

I
C
''' I
#
*

H
C
"`''" H

, (1)
C.J. Pereira/Chemical Engineering Science 54 (1999) 1959}1973 1961
Fig. 3. Schematic of life cycle phases of a typical chemical process
(adapted from Bollinger et al., 1996).
where C
''' I
are process-related costs and C
"`''" H
are
environmental costs. Such an approach has recently been
reported for a nitric oxide plant (Kniel et al., 1996).
C
"`''" H
can include levelized end-of-pipe equipment and
operating costs, waste disposal costs, and the hidden
costs associated with environmental monitoring and re-
porting. The cost of emissions from the standpoint of
human health and the environment can also be included.
Though several methods for doing this have been pro-
posed (e.g. Stefanis et al., 1995), a good general method is
not yet available. Regardless of whether or not such costs
are included, the optimum design of a process likely will
change as environmental costs become signi"cant.
As in the development of traditional chemical and
petrochemical processes, reaction engineering, broadly
de"ned as the "eld that quanti"es the engineering aspects
of chemically reactive systems, provides enabling tools
that accelerate process development. Core reaction
engineering methods are useful for kinetic modeling,
reactor selection, scale-up, and design. Reactor models,
that link independent process variables (e.g., reactor type,
operating conditions) to performance variables (e.g., con-
version and selectivity), together with the appropriate
explicit and implicit constraints, are essential for process
optimization. In addition, the research frontiers are pro-
viding tools for new process development, from com-
putational chemistry to probe the nature of catalytic
active sites to Computational Fluid Dynamics modeling
for designing the internals of reaction-separation sys-
tems. The increased use of reaction engineering in devel-
oping biochemical processes is anticipated. Biochemical
processes for producing chemicals such as acrylamide,
cyclodextrin, hydroquinone, and ethanol are commercial
in Japan (Tanaka et al., 1993).
A comprehensive review of the environmental "eld is
a considerable challenge. The status of a number of
environmental technologies has been discussed in the
Proceedings of the 5th World Congress of Chemical
Engineering (AIChE, 1996). A summary of environ-
mental developments in Europe is also available
(DECHEMA, 1990). The present paper, while recogniz-
ing the considerable advances in the development of both
products and processes, will primarily address process
development. The life cycle phases for a typical process
will "rst be presented. Some examples from the authors'
experience, where reaction engineering is playing a key
role in improving environmental performance, will then
be discussed. The discussion is biased toward catalysis
and catalytic reactors.
2. Process life cycle phases
Chemical processes typically evolve through life cycle
phases. Life cycle phases have recently been discussed in
a monograph on inherently safer chemical processes
(Bollinger et al., 1996). They include research, process
engineering, plant operation and, eventually, decommis-
sioning (Fig. 3). The research phase typically explores the
feasibility of a chemical process concept. Reactor and
process design is accomplished during the extensive pro-
cess engineering phase. This phase includes development
of kinetics and mathematical modeling, "rst at the labor-
atory scale and later at the pilot scale. Durability issues
are typically investigated at this stage. If questions re-
garding feasibility arise, the project may return to the
research phase, or be discontinued. Commercial model-
ing software may be used to design unit operations and
to develop a preliminary #ow sheet. The #ow sheet is
used for economic evaluation. If economic criteria are
met, the plant is designed and constructed. During the life
of the process, materials used in manufacture, such as
spent catalysts and by-product wastes, may have to be
disposed from time to time.
The number of options available for reducing environ-
mental impact decreases drastically as the project pro-
gresses from one life cycle phase to the next. A larger
number of alternatives are available at the research
phase. The constraints on what can be reasonably con-
sidered increase as the project continues. Extensive reac-
tor and process modi"cations can be only implemented
prior to commissioning of a new plant. During plant
operation, the scope of changes is impacted by the econ-
omics of modifying downstream processing units and
installed utilities. If environmental compliance issues still
remain after the plant is commissioned and optimized,
end-of-pipe technologies provide the only remaining
option.
The costs associated in implementing an option dra-
matically increase as the size of the equipment gets larger.
Within the context of process safety, it is believed that,
relatively speaking, "xing a safety-related problem costs
$1 at the research stage, $10 at the process #ow sheet
stage, $100 at the "nal design stage, $1000 at the pro-
duction stage, and $10,000 at the post-incident stage
(Bollinger et al., 1996). Cost parallels are likely to be
similar in the case of environmental problems as well.
There is, therefore, a considerable incentive to address
and resolve environmental issues early in the life cycle.
1962 C.J. Pereira/Chemical Engineering Science 54 (1999) 1959}1973
Fig. 4. Active steps in the microkinetic model for methane partial
oxidation over (a) MoO
`
/SiO
`
, and (b) V
`
O
`
/SiO
`
(Amiridis et al.,
1991).
3. Research
In many cases, the catalytic active site, responsible for
making product, also generates undesired by-products
and pollutants. Beyond reducing rawmaterials e$ciency,
these undesired materials have to be separated and
disposed. Discovery research aimed at identifying a
100%-selective active site is ongoing both in industry and
in academia. As an example, a catalyst that decreases
by-product pollutant CCl
"
selectivity during the manu-
facture of phosgene has been recently discovered. The
conventional coconut carbon-based catalyst produced
about 500 ppm of CCl
"
in the exhaust, which had to be
reduced to below 100 ppm to meet the tighter regula-
tions. The new catalyst has reduced CCl
"
levels to below
50 ppm and produces more phosgene (Cicha and
Manzer, 1998). The identi"cation of such catalysts would
be especially welcome in partial oxidation catalysis.
Computational chemistry is being increasingly used to
probe the active site. For example, a new, more-selective
transport riser-regenerator process for oxidation of
n-butane to maleic acid has been discussed (Lerou and
Ng, 1997). The process concept was based on the "nding
that the reaction proceeds via a redox mechanism that
involves oxygen on the surface of the vanadium pyro-
phosphate catalyst (Haggin, 1995). Computational chem-
istry is providing insights on how the oxygen interacts on
the catalyst surface. Optimization calculations of the
monovanadium species at the density functional theory
(DFT) level are providing clues on how oxygen moves
from one vanadium to another (Dixon et al., 1997) and
may provide leads for further improving selectivity.
A detailed understanding of the reaction rate mecha-
nism is useful for identifying operating conditions that
minimize by-products and developing new catalysts. In
the partial oxidation of methane to formaldehyde, macro-
kinetic analysis has revealed that formaldehyde and car-
bon dioxide are primary products over a MoO
`
/SiO
`
catalyst. In contrast, formaldehyde is the primary prod-
uct over V
`
O
`
/SiO
`
catalyst. A microkinetic approach,
using a single set of physically realistic input assump-
tions, has identi"ed the elementary steps needed to "t the
data for both of these catalysts (Amiridis et al., 1991).
Reaction pathways for each catalyst are shown in Fig. 4.
The approach can also be used as a design tool to identify
catalyst properties that may provide improved selectiv-
ity. A monograph discussing the microkinetic approach
is now available (Dumesic et al., 1993). Such analyses are
particularly important in the design of catalytic devices
for the removal of pollutants in the ppm range where
customary approximations in catalytic reaction engin-
eering are unreliable over wide ranges of process vari-
ables (Boudart, 1994).
In addition to tools that allow a mechanistic under-
standing of the surface chemistry, there is a need for tools
that represent and track molecular species in complex
feeds. This is especially true in petroleum re"ning where
regulations are placing restrictions on the composition of
the fuel. For example, reformulated gasoline regulations
mandate a minimumlevel of oxygenated compounds and
that the aromatic (especially benzene) content, sulfur
content, and vapor pressure be controlled. Models that
contain an increased level of molecular detail, far beyond
the traditional lumped approaches, are needed. The
structure oriented lumping (SOL) approach provides
a methodology for representing molecular structure, de-
scribing process chemistry, and estimating molecular
properties (Quann and Ja!e, 1996). Such models can help
track molecules as they are processed in the re"nery and
provide insights on howto optimize re"nery operation to
meet regulatory requirements or to generate a higher-
value product slate.
Molecular modeling is being used to estimate transport
coe.cients in zeolites (Theodorou et al., 1996). There is
a continuing need to extend the work to more complex
feed and catalyst systems. An example, from the pharma-
ceutical industry, is the hydroxylation of phenol with
hydrogen peroxide over ZSM-5 and TS-1 to yield hydro-
quinone. The use of hydrogen peroxide, although expen-
sive when compared to molecular oxygen, is preferred
because it is simpler and safer to use. The reaction has
99% selectivity to dihydroxybenzenes and produces very
little by-product salt. In contrast, the conventional route
using Fenton's reagent produces 10 kg of salt per kg of
product. A second example is the acylation of alkyl
benzene with carboxylic acids over Ce-Y zeolite to yield
aromatic ketones. The conventional route uses
stoichiometric quantities of AlCl
`
along with acid chlor-
ides, which creates environmental problems (Mills and
Chaudhari, 1997).
Traditional reaction engineering methods for ana-
lyzing and optimizing intra- and inter-particle resistances
have been used extensively in the industry. As new
chemistries emerge, there may be a need to extend our
C.J. Pereira/Chemical Engineering Science 54 (1999) 1959}1973 1963
Fig. 5. Biochemical routes to chemical intermediates (C&E News,
1998).
Fig. 6. Dodecane conversion over ZSM-5 crystals of various sizes (in microns). A nitrogen-containing poison is injected at 25 h (Kennedy et al., 1991).
understanding to these new systems. An example is the
hydrolysis of adiponitrile (Fig. 5), using a supported
enzyme catalyst, for the production of 5-cyanovaler-
amide, an agricultural intermediate (Di Cosimo, 1998).
The conventional manganese dioxide catalyst is not reus-
able, di$cult to recover and reactivate and, for selectivity
reasons, operates at lower than 20% conversion. The
bioconversion process, in contrast, has high conversion
and high selectivity.
An interesting example in which di!usion, reaction,
and the mechanical properties of the catalyst support are
all important is in ethylene polymerization. During early
polymerization, the catalyst pores "ll up with polymer
product. Ethylene continues to di!use through the prod-
uct even after the pores are "lled and, eventually, the
support fragments like popcorn (Webb et al., 1991). As
polymerization continues, the active sites separate from
each other as the polymer grains grow in size. Reaction
engineering models of polymer grain growth are largely
due to Ray and his coworkers (e.g., Floyd et al., 1986).
The discovery of homogeneous &&single-site'' metallocene
catalysts, that are at least an order of magnitude more
active than conventional catalysts, has created a need to
&&heterogenize'' the homogeneous catalyst, engineer sup-
port pore structure, and optimize process operation.
Metallocene catalysts are making it possible to create
plastics with physical properties "nely tuned to particu-
lar uses. The development of single polymers (versus
polymer blends) having specialized properties is expected
to reduce polymer usage and the amount of waste and
simplify polymer recycle (Ewen, 1997).
An improved understanding of the deactivation mecha-
nism can accelerate process development. An example is
the shape-selective catalytic cracking of normal para$ns
over ZSM-5 to improve the low temperature #uidity of
fuels. A reaction engineering analysis of the cracking and
deactivation behavior of ZSM-5 crystals using a model
n-dodecane feed suggested that cracking is intra-crystal
di!usion limited, coking occurs uniformly throughout
the crystal and that poisoning, due to nitrogen species
such as 5,6 benzoquinoline, occurs via a shell progressive
mechanism. This explains why the smallest crystals,
though the most active initially, are more easily poisoned
(Fig. 6) and provides guidance on the desired crystal size
(Kennedy et al., 1991). The process, originally developed
to improve operating e$ciency and reduce energy costs,
has largely replaced prior solvent extraction technology
which produced fugitive emissions of solvent. The ZSM-5
process eliminates these fugitive emissions and cracks the
normal para$ns to useful, gasoline-range products
(McWilliams, 1993).
Environmentally benign alternative solvents, such as
supercritical carbon dioxide, are being increasingly inves-
tigated. For example, methyl methacrylate has been poly-
merized in a supercritical carbon dioxide medium by
using a specially engineered free radical initiator and
a polymeric stabilizer. The carbon dioxide is easily
separated from the reaction mixture, and the approach
o!ers the possibility of avoiding the generation of
hazardous waste streams (Illman, 1994). Supercritical
1964 C.J. Pereira/Chemical Engineering Science 54 (1999) 1959}1973
#uids are being considered in a number of industrial
applications, such as extraction of food and pharma-
ceutical compounds and fragrances, dyeing and clean-
ing textile "bers, and synthesis of nanopowders and
polymers. The role of supercritical #uids in the synthesis
of organic chemicals has been recently reviewed (Dinjus
et al., 1997). From a reaction engineering standpoint,
a continuing need to develop physical property
databases and kinetics for these new chemistries is in-
dicated.
As novel technologies are developed, new reaction en-
gineering tools may be needed.
E The environmental opportunities for electrochemical
technology have been recently discussed (Fenton,
1998). As an example, a process for the direct gas phase
electrolysis of anhydrous hydrochloric acid to chlorine
in a polymer electrolyte membrane fuel cell (PEMFC)
is being developed (Trainhamet al., 1995). The process
converts by-product hydrochloric acid to chlorine for
reuse.
E The electrical engineering aspects of inductively heat-
ing a granular bed of conductive particles has recently
been examined (Duquenne et al., 1994). The suitability
of using induction heating for endothermic reactions
has been discussed in the literature (Gardner, 1991).
E Hydrogen cyanide has been produced in small quantit-
ies from methane and ammonia over a Pt/Al
`
O
`
cata-
lyst housed in a microwave cavity. Selectivity is found
to be higher than that of thermal processes (Wan and
Koch, 1994).
E Reactor design aspects of a loop reactor using a sono-
chemical power ultrasound module have been dis-
cussed for the hydrolysis of methyl benzoate (Martin
and Ward, 1992).
E Photocatalytic reactors are being developed for the
oxidation of organic compounds in air and water
systems. Reactor design issues include conversion and
selectivity to partially-oxygenated products (Braun
et al., 1993).
4. Process engineering
The engineering phase provides an opportunity for
assessing the design aspects of a new process. Typical
technical goals include a detailed quanti"cation of reac-
tion products, reactor modeling and the development of
scale-up criteria, durability evaluation and process syn-
thesis. An obvious "rst step is to explore reactor design
and operating conditions that reduce byproduct selectiv-
ity. Theoretical analyses of conditions that minimize
waste for series and parallel reaction networks are avail-
able (Allen and Rosselot, 1997; Hurme, 1995). For
example, for parallel reactions, the ratio of the rate of
the undesirable to the desirable reaction for power law
kinetics is given by
R"
A
S
A
B
e'#S#B'02 CK`K

. (2)
The values of the reaction rate constants in Eq. (2) pro-
vide guidance on reactor selection and on operating
conditions favoring the desired reaction. Such analysis
has been used to study allyl chloride synthesis, acrylo-
nitrile synthesis, a reactive distillation process, and a
petrochemical process (Hopper et al., 1994). The authors
report that emissions from two of these processes have
been reduced by implementing changes proposed by
modeling.
Mixing and -ow distribution hardware can have a large
in#uence on product distribution. Computational #uid
dynamics (CFD) code allows the numerical solution of
the equations of motion (mass, momentum, and energy)
in a #ow geometry of interest. Species material balance
equations may also be solved to include chemical reac-
tions. Many challenges in the implementation of CFD
code remain. Some examples include models for multi-
phase #ow, the e!ects of turbulent micro-mixing of fast
reactions with characteristic times comparable to the
time scale of the smallest (Kolmogorov) eddies, and the
computational e!ort required to solve problems with
wide ranges of geometrical scales and time-dependent
behavior. CFD can provide guidance on the design of
reactor internals, such as feed nozzles, ba%es, distributor
plates, and impellers. When reactions are included, CFD
can provide semi-quantitative information on the e!ect
of reactor geometry on product and by-product selectiv-
ity (Preisigke, 1995). Neglecting micro and meso-mixing
can sometimes lead to large errors in predicted selectivity
(Baldyga et al., 1997), even when velocity and turbulence
"elds agree quite well with experiment. A reviewof litera-
ture together with an identi"cation of future needs in the
chemical processing industry has been recently published
(Harris et al., 1996).
The design of reactors for high temperature, short con-
tact time operation poses a challenge. Syngas has been
produced over noble metal surfaces directly from meth-
ane at temperatures of &1,0003C and contact times of
&1 m second (Schmidt and Goralski, Jr., 1997). Reactor
design considerations include high mass and heat trans-
fer rates, narrowresidence time distributions, elimination
of homogeneous processes which lead to multiple prod-
ucts, #ames and explosions, and materials of construc-
tion issues (Schmidt et al., 1996). For rational reactor
design, models must contain detailed kinetic descriptions
of both heterogeneous and homogeneous reactions. De-
tailed gas-phase and surface kinetics and transport pro-
cesses have been included in a model for the catalytic
ignition of H
`
/air mixtures over platinum catalysts (Bui
et al., 1997). These approaches will have to be extended to
include methane and higher alkanes.
C.J. Pereira/Chemical Engineering Science 54 (1999) 1959}1973 1965
ransient analysis of the di!usion, reaction and ad-
sorption processes are important in the design and op-
eration of solid-supported liquid-catalyst reactors. Such
reactors have been proposed to replace conventional
re"nery alkylation technology that uses liquid hydro-
#ouric and sulfuric acids. In this concept, the reactor is
packed with a high surface area support material that
strongly adsorbs a #ourinated sulfonic acid catalyst.
Reactants, together with a small amount of supple-
mental acid, are fed to the reactor to produce alkylate
(Hommeltoft, 1996). Alternately, the reverse #ow reactor
concept has also been discussed. In this case, reactant
#ows over the bed driving the acid front toward the
reactor exit. At periodic intervals, the #ow of reactant is
reversed (Hommeltoft, 1991). This reactor concept is the
mass transfer analog of the heat storage concept prac-
ticed in reverse-#ow oxidation reactors. The approach is
the next best thing to the development of a commercially
viable solid acid alkylation catalyst (e.g., see Rao and
Vatcha, 1996; Corma and Martinez, 1993).
There is a trend toward reactors that combine reaction
and separation processes. An example is the synthesis of
methyl-tert-butyl-ether (MTBE), a reformulated gasoline
additive, from methanol and isobutylene. In contrast to
the conventional approach of using "xed bed catalytic
reactors followed by distillation columns, a distillation
column containing catalyst is used (e.g., see Smith, 1984).
The overall material balance for reactive distillation is
(Doherty and Buzad, 1992)
x
$G
!
D/B
D/B#1
x
"G
!
1
D/B#1
x
G
#Da
,2

H

x`
H`
!
x
H
x
H`
K
, i"1, 2, (3)
where x is the mole fraction of component i in the liquid
phase, D and B denote the #ow of distillate and bottoms,
Da is the Damkohler number, and the last term in
parenthesis represents the reaction rate for a reversible
reaction. Potential advantages to this approach include
reduced thermodynamic limitations to conversion (as the
MTBE moves down the column and away from the
reaction zone), a simpler separation system, fewer pieces
of process equipment (e.g., valves, #anges, heat exchange-
rs, etc.) reducing the possibility of fugitive emissions, and
lower capital and energy costs (Allen, 1996). The ap-
proach does reduce the operating degrees of freedom,
since the catalyst temperature is limited to a range where
separation is possible as well. Another example, reported
in the trade literature, is the process for producing methyl
acetate in which about seven process steps are con-
solidated into a single reactive distillation vessel. Com-
pared to the traditional process, reduced investment,
improved operability, and an energy savings of over 70%
are claimed (Chem. Eng. Prog., 1990). Other reaction-
separation systems, based on porous inorganic and other
membranes, are also being investigated (Allen and
Rosselot, 1997).
Separate reactor and separation systems can be opti-
mally combined to increase product yield. Such is the
case for the methane oxidative coupling reaction to con-
vert abundant natural gas to C
`
products. Progress on
developing catalysts having high conversion and selectiv-
ity has been slow in recent years. The fundamental prop-
erties of these catalysts, such as p-semiconductivity and
oxygen anion mobility, have been investigated in more
detail and a kinetic model has been proposed (Zantho!
et al., 1992). The desired products of the reaction, ethy-
lene and ethane, are far more reactive with oxygen than
methane and are therefore easily oxidized to carbon
oxides. Process concepts that couple catalytic reactors
with adsorbers have been investigated for increasing the
overall yield at high conversion. The simulated counter-
current moving-bed chromatographic reactor (SCMBCR)
periodically cycles the methane}oxygen feed mixture
between four reactor units and four separating columns
to achieve a yield of 50% (Tonkovich et al., 1993).
Modeling work on the SCMBCR reactor has been re-
ported (Tonkovich et al., 1994). C
`
hydrocarbon yield of
up to 88% has been claimed using a gas recycle loop
reactor combined with an appropriate molecular sieve
trap in the recycle loop (Jiang et al., 1994).
Process technology for the on-site manufacture and
immediate use of hazardous chemicals eliminates the
need to transport, store and handle hazardous chemicals.
Methylisocyanate has been conventionally produced by
the phosgenation of methylamine and then stored and
shipped as needed. A two-step process for the on-site
generation of methylisocyanate from methylamine
has been developed. Methylamine is "rst reacted in
a bubble-column reactor to form methylformamide. The
oxidative dehydrogenation of methylformamide using
a silver catalyst yields methylisocyanate (Rao and
Heinsohn, 1985). The approach of developing small
modular reactors for making hazardous chemicals, such
as HCN and phosgene, is also being explored. The ulti-
mate in modularization and miniaturization is the devel-
opment of microreactors (Lerou et al., 1995).
An alternate approach to on-site generation is to de-
velop technology that renders materials non-hazardous
prior to shipping. Hydrocyanation technology is being
used to prepare nitriles, which can then be shipped to the
customer for further functionalization. A di!erent but
related approach, used in the delivery of HCN to mine
sites for gold extraction, is to reduce the risk of exposure
by converting HCN to solid NaCN prior to shipment
(Bell et al., 1995).
There is an active interest in "nding new uses for
by-products or converting by-products to useful products.
A vintage example is re"nery hydrodesulfurization,
which selectively removes organic sulfur in gasoline
1966 C.J. Pereira/Chemical Engineering Science 54 (1999) 1959}1973
and distillate fuel and converts it to hydrogen sul"de.
Hydrogen sul"de can be routed to a Claus plant and
converted to salable elemental sulfur. Parenthetically,
the reduced sulfur content in gasoline bene"cially a!ects
the oxidation performance of the automobile catalytic
converter (since sulfur oxides inhibit noble metal
catalysts). In the chemical industry, by-products from
the process for making hexamethylenediamene, a nylon
6,6 intermediate, are being converted to branched dia-
mines, used in comonomers and curatives (Bell et al.,
1995).
Reaction engineers are called upon to design reactors
as new chemistries are discovered. A much cited example
of a &&green'' chemistry is the new process for manufactur-
ing the drug ibuprofen. The conventional process utilizes
a six-step stoichiometric process having a molar e$ci-
ency of about 40%. The "rst step involves the reaction of
isobutylene and an acetylating agent using an environ-
mentally undesirable AlCl
`
catalyst. The new process
uses a reactive-extractor system with liquid HF as cata-
lyst (Lindley et al., 1991). The product is hydrogenated
over a Raney

catalyst to form the alcohol. The alcohol

is carbonylated by carbon monoxide using a Pd catalyst
to make product with an overall molar e$ciency of
about 80% (Chemical Engineering, 1997). Another
example is the enzyme-based synthesis of dopamine.
A reactor model has revealed that reactor volume can
be minimized by using separate sequential (rather that
simultaneous) enzymatic reactions (Anderson et al.,
1992). More user-friendly software for screening alternate
reactor concepts will clearly be useful.
5. Plant operation
At this stage, the nature and emissions of pollutants
are known. Regulatory compliance and cost issues often
drive technology selection and the timing for implemen-
tation. Pollution prevention approaches can be e!ec-
tively used to minimize emissions. Obvious approaches
include good housekeeping, simple process modi"ca-
tions, product reformulation, raw material substitution,
and recycling. Each unit operation in the process can be
examined from the pollution reduction standpoint (Allen
and Rosselot, 1997). A rational methodology to pollution
prevention, that involves selecting the waste stream,
generating ideas for emissions reduction, and implemen-
ting the most cost-e!ective changes, has recently been
discussed (Dyer and Mulholland, 1998a).
In some special cases, a detailed understanding of
process chemistry can result in the development of in-
process additives for emissions reduction. An example of
this is for the #uid catalytic cracking (FCC) process, that
converts high-boiling feeds to transportation fuels. Pro-
cess emissions of particulates, carbon monoxide (CO),
nitrogen oxides (NO
V
), and sulfur oxides (SO
V
) have to be
reduced because of regulations. In response to these
pressures, manufacturing methods have been modi"ed to
produce more attrition resistant catalysts. In-process ad-
ditives, having the same size distribution as FCC cata-
lysts, have been developed to control emissions. Noble
metal-promoted additives have reduced regenerator CO
and NO
V
emissions. SO
V
transfer additives have been
developed to reduce SO
V
emissions. The SO
V
transfer
mechanism is shown schematically in Fig. 7. SO
V
is
adsorbed by the additive in the regenerator and released
as H
`
S together with product gases from the FCC riser
(Cheng et al., 1998).
5.1. End-of-pipe technologies
These technologies control generic classes of air pollu-
tant emissions (such as CO, volatile organic compounds
(VOCs), NO
V
and SO
V
). Sometimes, technology for deal-
ing with atypical pollutants (e.g., methyl bromide from
polyester plants, nitrous oxide from adipic adic plants,
etc.) is needed. The speci"c choice of control technology,
in most cases, is driven by the need to meet regulatory
compliance at the lowest cost. The most cost-e!ective
strategy for a given application, depends on a number of
factors, including the nature of the pollutant, its concen-
tration, the #ow rate of the exhaust, and regulations.
Reaction engineering contributions to emission con-
trol go back to the early days of the 3-way automobile
catalytic converter (Hegedus and Gumbleton, 1980). Bead
catalysts with non-uniform intraparticle impregnation
patterns were developed and later analyzed (e.g., see
Gavriilidis et al., 1993, and references therein). Reaction
engineering tools were used to predict poisoning behav-
ior of catalysts and to optimize pore structure for max-
imum poison resistance (e.g., see Pereira et al., 1988, and
references therein). More recently, work has been re-
ported on modeling the Electrically Heated Converter
(EHC), which is an option for meeting the tough regula-
tory standards in California (see Oh et al., 1993). The
EHC is a catalyst-coated metal pre-converter device that
is resistively heated by an electric current just prior to
engine ignition. Electric heating reduces the CO and
hydrocarbon emissions during the "rst 2 min following
the cold start of the engine, which account for as much as
80% of the emissions during a typical driving cycle.
5.1.1. Volatile organic compound (VOC) control
The two primary options for destroying VOCs are by
thermal or catalytic oxidation. hermal oxidizers are
robust combustion devices that achieve high destruction
e$ciency if properly operated at the 3Ts: temperature,
turbulence, and residence time. Fuel costs and emissions
of carbon monoxide and nitrogen oxides (formed in
the #ame tube of the oxidizer) can be issues. Catalytic
oxidizers may be bead beds or monoliths. The active
C.J. Pereira/Chemical Engineering Science 54 (1999) 1959}1973 1967
Fig. 7. Mechanism for in-process SO
`
transfer additive for SO
V
control (adapted from Cheng et al., 1998).
catalytic ingredients (usually noble metals) are typically
supported on a high surface area alumina (Spivey, 1987).
The activity and durability of these catalysts have been
addressed for select applications (e.g., see Libanati et al.,
1998, and references therein). Sometimes, as in the case of
the destruction of N
`
O from adipic acid plants, new
catalysts are needed (Anseth and Koch, 1994). Catalytic
systems can typically operate at a much lower excess
oxygen concentration than thermal oxidizers.
The destruction of halogenated hydrocarbons, often
used as solvents, cleaners and degreasers, is a signi"cant
challenge. In thermal systems, complete combustion has
to be ensured by proper design and operation of the
#ame, to avoid emissions of toxic chlorinated partial
oxidation products (Chang et al., 1986). Noble metal
catalysts deactivate in the presence of halogens and more
robust catalyst formulations are needed. The perfor-
mance of supported-molten salt catalysts (Saraf et al.,
1989), zeolite-supported chromium (Chatterjee and
Greene, 1991), and combined sorbent-catalyst systems
(Greene et al., 1996) has been discussed.
Under normal operation, VOC oxidation catalysts op-
erate at or near external mass transfer-limited conver-
sion. Since this conversion provides an upper bound on
control e$ciency, there have been considerable e!orts
aimed at increasing the mass transfer-limited perfor-
mance of monolithic structures at a "xed pressure drop.
Under mass transfer-limited control (k
K
is the external
mass transfer coe$cient)
k
K
"!S< ln(1!)/ (4)
where SV is the space velocity, is the geometric surface
area per volume and is the conversion. A variety of cell
geometries and metal structures have been evaluated.
Mass transfer correlations have been developed for sev-
eral of these structures. For example, as shown in Fig. 8,
Camet metal monoliths were found not to follow the
mass transfer-pressure drop analogy and have a higher
mass transfer limited conversion at a given pressure drop
compared to commercial square-channeled ceramic
monoliths (Gulian et al., 1991). Theoretical calculations
aimed at increasing the destruction e$ciency of metal
structures are providing some design guidance (Anders-
son and Schoon, 1993). The development of a thinner-
walled, steel monolith substrate having a heat capacity
that is 40% that of commercial monoliths has been
reported. A catalyst based on this substrate is claimed to
have a 30% higher &&overall purifying e$ciency'' at
a third of the pressure drop (Chemical Engineering,
1997).
New process concepts for dealing with halogenated
hydrocarbons have recently been proposed. A new
"xed-bed reactor technology combines oxidation with
dry scrubbing. The reactor contains a base metal oxide
1968 C.J. Pereira/Chemical Engineering Science 54 (1999) 1959}1973
Fig. 8. Variation of f and j
"
for square channel ceramic and Camet
catalysts (Gulian et al., 1991).
catalyst (such as Mn or Mg) supported on an alkali or
alkaline earth (such as Na, K, or Ca) metal carbonate.
The halogenated hydrocarbon is oxidized to carbon di-
oxide, water and acid gas. The acid gas reacts with the
metal carbonate support as soon as it is formed. Event-
ually, the support is consumed and has to be regenerated
or replaced (Berty, 1997). Another proposed approach is
the catalytic hydrogenation of chlorinated solvents (Kim
and Allen, 1997). In this approach, there is a need to
carefully analyze the exhaust for the presence of toxic
partially halogenated products.
Scrubbing of VOCs is an attractive option for treating
exhausts having low concentration but high exhaust
#ow. Froma materials standpoint, the search has been to
identify materials that have high capacity and regenera-
bility. Pressure swing and temperature swing "xed bed
adsorbers have been used for a long time. =heel concen-
trators are being more recently deployed. The VOCs are
desorbed by passing hot regenerative gas through a small
sector of the wheel (Schwartz and Pereira, 1994). Typi-
cally, the VOC concentration in the regenerative gas is
higher than the concentration of the exhaust by a factor
of 6}10. VOCs in the regenerative gas may be oxidized or
recovered. Membranes also present a viable option for
concentrating VOCs.
Most oxidizers have built-in recuperative or regen-
erative heat recovery systems. Recuperative systems are
shell and tube heat exchangers. The regenerative or &&hot
rocks'' systems transfer heat by direct heating and cooling
of solids. Though the original regenerative or &&reverse-
#ow'' concept patent was issued a long time ago (Cottrell,
1938), regenerative catalytic and thermal systems for
VOC control have been commercialized in the US only
in the past decade. Reaction engineering work on
reverse-#ow reactors is largely due to Matros and
Eigenberger and their coworkers (e.g., see Matros and
Bunimovich, 1996; Matros and Bunimovich, 1995;
Eigenberger and Nieken, 1994).
5.1.2. Nitrogen oxides (NO
V
) control
NO
V
is de"ned as the sum of the nitric oxide (NO) and
nitrogen dioxide (NO
`
) concentration in the exhaust. The
easiest way to reduce nitrogen oxides is to prevent them
from forming in the "rst place. Nitrogen oxides are prim-
arily formed by the oxidation of fuel-bound nitrogen and
by reaction of air-derived nitrogen with oxygen at high
temperature. The latter pathway for nitrogen oxide
formation can be minimized by reducing high temper-
ature hot spots, increasing turbulence, and increasing the
residence time. An example is a low-NO
V
burner in which
gradual mixing with the centrally injected fuel is achieved
by radial strati"cation of the #ame. This strati"cation is
achieved by a combination of swirling air #owand strong
positive radial density gradients in the #ame (Beer, 1994).
Another approach is the development of an inward-"red
premixed radiant burner that operates at high levels of
excess air (Bartz et al., 1996). A third approach is via
catalytic combustion.
In catalytic combustion, the catalyst lights o! combus-
tion reactions which are sustained in the gas phase after
the catalyst. Peak temperatures obtained are signi"cantly
lower than those achieved in admixed or premixed
#ames. Along with materials developments, proposed
reactor designs include two or more reactors in series,
e.g., the "rst stage for light o!, the second for controlled
conversion, the third stage for high temperature stability
(Zwinkels et al., 1993; Dalla Beta et al., 1993a). As shown
in Fig. 9, hybrid heat exchange-catalyst reactors having
only a fraction of the metal monolith structure coated
with catalyst are being developed to control conversion
and thereby limit the maximum temperature of the cata-
lyst (Retallic and Alcorn, 1993; Dalla Betta et al., 1993b).
Yet another concept for preventing nitrogen oxide
formation is that of &&hybrid catalytic combustion'', which
combines a partial oxidation reactor with air staging.
Natural gas and a small fraction (&7.5%) of the required
air are fed to the heat exchange-catalyst reactor which is
externally cooled by about half the required air. The
reaction products, together with the remaining air are
combusted in the primary zone of the combustor. The
heated air is injected into the secondary zone and sent to
a gas turbine (Colket et al., 1993). Reaction engineering
can play a useful role in the design and optimization of
these heat exchange-reactor concepts.
The commercial catalytic method for controlling NO
V
requires the injection of ammonia which, in the presence
of a V
`
O
`
/TiO
`
catalyst, selectively reacts with NO
V
to
form nitrogen (SCR). Exhaust gases containing NO
V
also
often contain SO
V
. The reaction engineering aspects of
the SCR process have been examined both for NO
V
reduction and SO
`
oxidation (Beeckman and Hegedus,
1991; Tronconi et al., 1994). NO
V
conversion is intrapar-
ticle di!usion limited while SO
`
oxidation is kinetically
limited. Optimization of the pore structure has increased
NO
V
activity and durability while minimizing SO
`
C.J. Pereira/Chemical Engineering Science 54 (1999) 1959}1973 1969
Fig. 9. Catalyst structure employing integral heat transfer (Dalla Betta
et al., 1993).
oxidation selectivity. An issue with this technology is the
cost of catalyst and monitoring systems and the emission
of unconverted (or slip) ammonia.
In some cases, control technologies can be integrated
into plant operations. An example of such integration is
the Ljungstrom regenerative air preheater used in coal
"red power plants. The air preheater is a large wheel (up
to 20 to 30 in diameter), packed with metal sheets, that
rotates between the feed and exhaust stream of a coal-
"red power plant. Hot exhaust gases, at a temperature of
about 3503C, directly heat the packed metal sheets. As
the hot packing rotates to the cool inlet side, it exchanges
heat with the inlet air to the plant. An approach under-
going evaluation is to coat the metal sheets in the top
section of the wheel with SCR catalyst (or pack the top
layer with ceramic SCR catalyst) and then to inject am-
monia for increased control of NO
V
emissions (Michel-
felder and Michalak, 1986). The objective function, for
purposes of optimization, is to maximize NO
V
conver-
sion subject to a lower bound on regenerative heat trans-
fer and an upper bound on pressure drop.
There is some progress on replacing ammonia in SCR
with a hydrocarbon, such as methane. The most desirable
way of controlling NO
V
is to catalytically decompose it
into nitrogen and oxygen. Steady-state NO
V
decomposi-
tion in the presence of excess oxygen has only been
demonstrated relatively recently; however, the catalyst is
inhibited by oxygen and water and poisoned by sulfur
oxides (see Pereira and Amiridis, 1995, and references
therein).
5.1.3. Combined technologies
In special cases, technologies may be combined to
provide a cost-e!ective control option. For example,
noble metals are good oxidation catalysts. In addition,
when ammonia is added, noble metals are known to
convert NO
V
at lower temperatures than V
`
O
`
/TiO
`
catalysts. These observations present the possibility of
operating the reactor in a narrow temperature window
with ammonia injection to simultaneously control both
carbon monoxide oxidation and NO
V
. A downside of this
technology is that noble metals are sensitive to the pres-
ence of low levels of sulfur in the exhaust and that the
product of the NO
V
reaction is N
`
O. The technology has
been commercialized primarily for natural gas-"red sys-
tems (Pereira et al., 1990).
There may also be opportunities to trade emissions of
one pollutant against another. For example, in the case of
natural gas-"red cogeneration plants, nitrogen oxide
emissions may be controlled by injecting steam or water
into the gas turbine combustion chamber. A negative
consequence, however, is that emissions of CO and un-
burned hydrocarbons can increase. These emissions may
be controlled by installing an oxidation catalyst. If only
modest ((50%) reduction in NO
V
emissions is needed,
this approach can obviate the need for a more expensive
SCR system (Pereira and Thomas, 1993).
Processes for sulfur oxides reduction involve either wet
or dry scrubbing and have been around for several dec-
ades (Cooper and Alley, 1994). More recent work aimed
at technologies for combined sulfur oxide and nitrogen
oxide control is based on the recognition that coal com-
bustors, that generate a large fraction of the sulfur oxides,
are also large emitters of nitrogen oxides. Developmental
technologies include solid adsorption-regeneration pro-
cesses, electron beam irradiation processes (oxidation
followed by scrubbing), denox}desox (SCR followed by
sulfur dioxide oxidation and scrubbing), combined
denox}desox wet scrubbing, dry injection processes, and
electrochemical reduction processes (Chicanowicz et al.,
1991).
5.1.4. Wastewater treatment
Wastewater treatment is a signi"cant issue for the
chemical processing industry. Bio,ltration and biotreat-
ment are being increasingly used for treating air and
water streams, respectively. Natural microorganisms are
used together with natural materials (such as mulch, soil,
etc.) or synthetic substrates (shaped pellets, monoliths,
etc.). Considerations in bio"lter design have been re-
viewed recently (Govind et al., 1993). Batch activated
sludge data has been used to elucidate the performance of
activated sludge systems and to design continuous #ow
reactors (Argaman, 1991). Reaction engineering work on
the design and operation of anaerobic treatment pro-
cesses has been discussed. Stoichiometry and kinetics,
mass transfer and reactor type have been found to be
important in such processes (Lema et al., 1991). Research
horizons include the development of reliable bio-moni-
tors and methods for minimizing and disposing the bio-
mass produced.
The practicing engineer is often called upon to trouble-
shoot problems associated metals emissions in waste-
water. Commercial simulation packages, such as the OLI
Software System, which has a database containing the
thermodynamic and physical properties for over 3000
inorganic and organic species, are now available for
1970 C.J. Pereira/Chemical Engineering Science 54 (1999) 1959}1973
estimating the solubility of metal hydroxides/oxides in
water as a function of pH (Dyer et al., 1998b).
The need to better understand the thermochemical and
kinetic aspects of aqueous treatment systems is expected to
become increasingly important. Experimental and
modeling work on the base catalyzed hydrolysis of
acrylonitrile suggests that concentration reduction has
a double exponential dependence on pH. The model
predicts that dramatic reductions in concentration may
be realized at higher pH, suggesting the use of basic
conditions (LaMarca et al., 1996). Reaction engineering
tools can quantify the time required to achieve the re-
quired pollutant destruction and thereby provide
a sound technical basis for the water treatment and site
remediation technologies.
6. Decommissioning and disposal
The safe handling and disposal of plant equipment has
been discussed by Bollinger et al. (1996). There are also
materials that have to be disposed, from time to time,
during plant operation. For example, the disposal of
spent catalyst is an important issue for chemical and
power plants. Noble metal spent catalysts are routinely
reclaimed for their metals value. Spent hydrotreating
catalysts contain about 5 wt% NiO or CoO and 15 wt%
MoO
`
. Demetallation catalysts can pick up an addi-
tional 10 w% Ni, and as much as 100% V, from heavy
feeds. These catalysts may be reclaimed to recover cobalt,
molybdenum, nickel, vanadium, and alumina. (C&E
News, 1992a). Re"ning typically entails dissolution in
strong acid or base followed by wet chemical recovery.
Reclamation economics can vary widely depending on
the sometimes-#uctuating price of metals. It is estimated
that 1/3 of the market for vanadium is met by reprocess-
ing spent demetallation catalysts (C&E News, 1992b).
A number of proprietary technologies are being de-
veloped for treating waste polymers. An attractive option
is to depolymerize the waste for reuse. However, this is
di$cult to do in cases where the waste is a blend of
individual polymers. As option receiving some research
attention is to blend the polymer together with heavy oil
feed and to crack it in a re"nery FCC unit (Lin et al.,
1998). The addition of polymers may a!ect FCC perfor-
mance, especially if the polymers contain heteroatoms,
such as chlorine and nitrogen.
7. Conclusions
Regulatory pressure and their impact on the cost of
production are driving forces for the development and
deployment of inherently-less polluting processes. From
the standpoint of the number of options available and
cost, it is desirable to deal with environmental issues
early in the process life cycle rather than during plant
operation. While the near-term approach is regulatory
compliance, in the longer term, environmental consider-
ations, like safety, will become an integral part of process
design. Continued development and application of reac-
tion engineering tools will accelerate the development of
both environmentally-friendly processes and end-of-pipe
technologies.
Acknowledgements
The author would like to thank the many past collab-
orators at W.R. Grace and Co. and present colleagues
at DuPont, especially J.B. Carberry, L.L. Hegedus,
T.R. Keane, C. LaMarca, J.J. Lerou, K.A. Saturday,
Stefanova-Petkova (from the DuPont Brandywine
Library), R.W. Sylvester and J.N. Tilton. The author also
thanks Professors D.T. Allen and B. Subramanian for
providing some reference materials.
References
Allen, D.T. (1996). In R. Gottlieb (Ed.), Reducing toxics: A new approach
to policy and industrial decisionmaking. Washington, DC: Island
Press.
Allen, D.T., & Rosselot, K.S. (1997). Pollution prevention for chemical
processes, New York: Wiley.
American Institute of Chemical Engineers (1996). Proceedings of the 5th
=orld Congress of Chemical Engineering: echnologies Critical to
a Changing =orld (Vol. 3, p. 111).
Amiridis, M.D., Rekoske, J.E., Dumesic, J.A., Rudd, D.F., Spencer,
N.D., & Pereira, C.J. (1991). A.I.Ch.E. J., 37(1), 87.
Anderson, W.A., Moo-Young, M., & Legge, R.L. (1992). Biotechnol.
Bioengng., 40(3), 388.
Andersson, S.L., & Schoon, N.-H. (1993). Ind. Engng. Chem. Res., 32,
1081.
Anseth, K., & Koch, T.A. (1994). U.S. Patent 5,314,673.
Argaman, Y. (1991). =ater Res., 12, 1583.
Baldyga, J., Bourne, J.R., & Hearn, S.J. (1997). Chem. Engng Sci., 52(4),
457.
Bartz, D.F., Marshall, B.N., Sullivan, J.D., Bruce, K., & Lombardo, A.
(1996). Proceedings of the International Conference on Incineration
and hermal reatment echnologies, (p. 823).
Beaver, E.R. (1996). Chemtech., 9.
Beeckman, J.W., & Hegedus, L.L. (1991). Ind. Engng Chem. Res., 29,
969.
Beer, J.M., (1994). Chem. Engng Sci., 49(24A), 4067.
Bell, A.T., Manzer, L.E., Chen, N.Y., Weekman, V.W., Hegedus, L.L.
& Pereira, C.J. (1995). Chem. Eng. Prog., 26.
Berty, J.M. (1997). Ind. Engng Chem. Res., 36, 513.
Braun, A.M., Jakob, L., Oliveros, E., & Oller do Nascimento, C.A.
(1993). In D. Volman, & G.S. Hammond (Eds.), Advances in Photo-
chemistry, (Vol. 18, p. 235).
Bollinger, R.E., Clark, D.G., Dowell III, A.M., Ewbank, R.M., Hender-
shot, D.C., Lutz, W.K., Meszaros, S.I., Park, D.E., & Wixom, E.D.,
(1996). In: D.A. Crowl (Ed.), Inherently safer processes: A life cycle
approach, Center for Chemical Process Safety, American Institute of
Chemical Engineers.
Boudart, M. (1994). Ind. Engng Chem. Res., 33, 2900.
C.J. Pereira/Chemical Engineering Science 54 (1999) 1959}1973 1971
Bui, P.-A., Vlachos, D.G., & Westmoreland, P.R. (1997). Ind. Engng
Chem. Res., 36, 2558.
Carberry, J.B. (1998). DuPont Company. Personal communication.
Chang, W-D., Karra, S.B., &Senkan, S.M. (1986). Environ. Sci. echnol.,
20(12), 1243.
Chatterjee, S., & Greene, H.L. (1991). J. Catal., 130, 76.
Chem. & Engng News (1992a). October 26, 20.
Chem. & Engng News (1992b). November 23, 2.
Chem. & Engng News (1997). August 4, 35.
Chem. Engng (1997). November 25, EC-25.
Chem. Engng Prog. (1990). February, 40.
Cheng, W.-C., Kim, G., Peters, A.W., Zhao X., Rajagopalan, K.,
Ziebarth, M.S., & Pereira C.J. (1998). Cat. Rev. - Sci. Engng,
40(1&2), 39.
Cicha, W.V., & Manzer, L.E. (1998). WO Patent 9800364.
Cichanowicz, J.E., Dene, C.E., DePriest, W., Gaikwad, R., & Jarvis, J.
(1991). EPRI/EPA/DOE SO
`
Controls Symposium. Washington, DC.
Colket, M.B., Kesten, A.S., Sangiovanni, J.J., Zabielski, M.F., Pandy,
D.R. & Seery, D.J. (1993). U.S. Patent 5,235,804.
Cooper, D.C. & Alley, F.C. (1994). Air pollution control: A design
approach, (p. 449). Waveland Press.
Corma, A., & Martinez, A. (1993). Catal. Rev. - Sci. Engng 35(4), 483.
Cottrell, F.G. (1938). U.S. Patent 2,121,733.
Dalla Beta, R.A., Ezawa, N., Tsurumi, K., Schlatter, J.C., & Nickolas,
S.G. (1993a). U.S. Patent 5,183,401.
Dalla Beta, R.A., Ribeiro, F.W., Tsurumi, K., Ezawa, N., & Nickolas,
S.G. (1993b). U.S. Patent 5,250,489.
DECHEMA (1990). Produktionsintegrierter ;mweltschutz in der
chemischen Industrie. ISBN 3-926959-21-5.
Di Cosimo, R. (1998). 2nd Annual Green Chemistry and Engineering
Conference. Washington, DC, June.
Dinjus, E., Fornika, R., & Scholz, M. (1997). In R. van Eldik & C.D.
Hubbard (Eds.), Chemistry under extreme or non-classical conditions.
New York: Wiley.
Dixon, D.A., Nichols, J.A., Gutowski, M., Jones, D.R., Nicholas, J.B.,
& Neurock, M. (1997). Preprints of Symposium on Applications and
Advances of Computational Chemical Modeling to Heterogeneous
Catalysis, ACS Meeting (p. 132).
Doherty, M.F., & Buzad, G. (1992). rans Ind. Chem. Engng, 70A, 448.
Dumesic, J.A., Rudd, D.F., Aparicio, L.M., Rekoske, J.E., & Trevino,
A.A. (1993). he microkinetics of heterogeneous catalysis. Washing-
ton, DC: American Chemical Society.
Duquenne, P., Deltour, A., & Lacoste, G. (1994). Canad. J. Chem.
Engng, 72, 975.
Dyer, J.A., & Mulholland, K.L. (1998a). Chem. Engng Prog., 34.
Dyer, J.A., Scrivner, N.C., & Dentel, S.K. (1998b). Environ. Prog.,
17(1), 1.
Eigenberger, G., & Nieken, U. (1994). Int. Chem. Engng, 34(1), 4.
Ewen, J.A. (1997). Sci. Am., 86.
Fenton, J.M. (1998). Electrochem. Soc. Interface, 30.
Floyd, S.K., Choi, K.-Y., Taylor, T.W., & Ray, W.H. (1986). J. Appl.
Polym. Sci., 32, 2935.
Gardner, D. (1991). Chem. Bri., 27(12), 1120.
Gavriilidis, A., Varma, A., & Morbidelli, M. (1993). In E.R. Becker,
& C.J. Pereira (Eds.), Computer-aided design of catalysts (p. 137).
New York: Marcel Dekker.
Govind, R., Utgikar, V., Zhao, W., Shan, Y., Parvatiyar, M., & Bishop,
D.F. (1993). Proceedings of the IG Symposium on Gas, Oil and
Environmental Biotechnology. Colorado Springs.
Greene, H.L., Prakask, D.S., & Athota, K.V. (1996). Appl. Cata. B:
Environ., 7, 213.
Gulian, F.J., Rieck, J.S., & Pereira, C.J. (1991). Ind. Engng Chem. Res.,
30, 122.
Haggin, J. (1995). Chem. Engng News, 20.
Harris, C.K., Roekaerts, D., Rosendal, F.J.J., Buitendijk, F.G.J., Das-
kopoulos, Ph., Vreenegoor, A.J.N., &Wang, H. (1996). Chem. Engng
Sci., 51(10), 1569.
Hegedus, L.L., & Gumbleton, J.J. (1980). Chemtech., 10(10), 630.
Hommeltoft, S.L. (1991). European Patent EP0433954A1.
Hommeltoft, S.L. (1996). U.S. Patent 5,498,820.
Hopper, J.R., Yaws, C.L., Vichailak, M., & Ho, T.C. (1994). =aste
Management, 3}4, 187.
Hurme, M. (1995). Chemical reactor design for waste minimization. NTIS
PB96-173901.
Illman, D.L. (1994). Chem. Engng News, 22.
Jiang, Y., Yentekakis, I.V., & Vayenas, C.G. (1994). Science, 264, 1563.
Kennedy, C.R., LaPierre, R.B., Mikovsky, R.J., & Pereira, C.J. (1991).
Ind. Engng Chem. Res., 30, 12.
Kim, D.I., & Allen, D.T. (1997). Ind. Engng Chem. Res., 36(8), 3019.
Kniel, G.E., Delmarco, L., & Petrie, J.G. (1996). Environ. Prog., 15(4),
221.
LaMarca, C., Scrivner, N.C. Gron, L.U., Izzo, B.J., Chu, A., & Klein,
M.T. (1996). In J.A. Apps, &C.-F. Tsang (Eds), Deep injection disposal
of hazardous and industrial waste (p. 601). New York: Academic Press.
Lindley, D.D., Curtis, T.A., Ryan, T.R., de la Graza, E.M., Hilton, C.B.,
& Kenesson, T.M. (1991). U.S. Patent 5,068,448.
Lerou, J.J., Harold, M.P. Ryley, J. Ashmead, J. O'Brien, T.C. Johnson,
M., Perrotto, J., Blaisdell, C.T., Rensi, T.A., & Nyquist, J. (1995).
DECHEMA Monographs, 132, 51.
Lerou, J.J., & Ng, K.M. (1996). Chem. Engng Sci., 51(10), 1595.
Lema, J.M., Mendez, R., Iza, J., Garcia, P., & Fernandez-Polanco, F.
(1991). =ater Sci. echnol., 24(8), 79.
Libanati, C., Ullenius, D.A., & Pereira, C.J. (1998). Appl. Catal. B:
Environ., 15, 21.
Lin, Y.-H., Sharratt, P.N., Garforth, A.A., & Dwyer, J. (1998). Energy
& Fuels, 12, 767.
Martin, P.D., & Ward, L.D. (1992). rans. Ind. Chem. Engng, 70A, 296.
Matros, Y.Sh., & Bunimovich, G.A. (1995). Ind. Engng, Chem. Res., 34,
1630.
Matros, Y.Sh., & Bunimovich, G.A. (1996). Catal. Rev. - Sci. Engng,
38(1), 1.
McWilliams, J.P. (1993). Spring Seminar, sponsored by the Central
Jersey Section of A.I.Ch.E., Princeton University.
Michelfelder, S., & Michalak, S. (1986). U.S. Patent 4,602,673.
Mills, P.A., & Chaudhari, R.V. (1997). Catal. oday, 37, 367.
O$ce of Technology Assessment (1992). Green products by design:
Choices for a cleaner environment, OTA-E-541, U.S. Government
Printing O$ce.
O$ce of Technology Policy (1996). Chemtech, 46.
Oh, S.H., Bissett, E.J., & Battiston, P.A. (1993). Ind. Engng Chem. Res.,
32, 1560.
Pereira, C.J., Kubsh, J.E., & Hegedus, L.L. (1988). Chem. Engng Sci., 43,
2087.
Pereira, C.J., Gulian, F.J., Czarnecki, L.J., & Rieck, J.S. (1990). Proceed-
ings. of Air & =aste Management Association, Paper 90-105.6.
Pereira, C.J., & Thomas, A.H. (1993). Proceedings of the Conference on
Integrated Energy and Environmental Management. New Orleans.
Pereira, C.J., & Amiridis, M.D. (1995). ACS Symp. Ser., 587, 1.
Preisigke, H.-J. (1995). In: J.B. Rawlings (Ed.), Dynamics and control of
chemical reactors, distillation columns, and batch processes, Oxford:
Pergamon Press.
Quann, R.J., & Ja!e, S.B. (1996). Chem. Engng Sci., 51(10), 1615.
Rao, P., & Vatcha, S.R. (1996). Oil Gas J., 56.
Rao, V.N.M., & Heinsohn, G.E. (1985). U.S. Patent 4,537,726.
Retallick, W.B., & Alcorn, W.R. (1993). U.S. Patent 5,202,303.
Saraf, A., Greene, H.L., Kosusko, M.S., & Narayan, S. (1989). J. Catal.,
120, 478.
Schwartz, R.J., & Pereira, C.J. (1994). Proceedings of Air and =aste
Management Association, Paper 94-FA148.04.
Schmidt, L.D., Hu!, M., & Bharadwaj, S.S. (1994). Chem Engng Sci., 49
(24A), 3981.
Schmidt, L.D., & Goralski, Jr., C.T. (1997). In R.K. Grasselli, S.T.
Oyama, A.M. Ga!ney, & J.E. Lyons (Eds.), 3rd =orld Congress on
Oxidation Catalysis, Elsevier, Amsterdam (p. 491).
1972 C.J. Pereira/Chemical Engineering Science 54 (1999) 1959}1973
Smith, L.A. (1984). U.S. Patent 4,443,559.
Spivey, J.J. (1987). Ind. Engng Chem. Res., 26, 2165.
Stefanis, S.K., Livingston, A.G., & Pistikopoulos, E.N. (1995). Com-
puters Chem. Engng, 19, S39.
Storck, W. (1998). Chem. Engng News, 18.
Tanaka, A., Tosa, T., & Kobayashi, T. (1993). Industrial applications of
immobilized biocatalysts (p. 91). New York: Marcel Dekker.
Theodorou, D.N., Snurr, R.Q., & Bell, A.T. (1996). Compr. Supramol.
Chem., G. Alberti & T. Bein (Eds.), 7, 508.
Tibbs, H. (1993). Industrial ecology: An environmental agenda for indus-
try, Global Business Network.
Trainham, J.A., Law, C.G., Newman, J.S., Keating, K.B., & Eames, D.J.
(1995). U.S. Patent 5,411,641.
Tronconi, E., Beretta, A., Elmi, A.S., Forzatti, P., Malloggi, S., & Bal-
dacci, A. (1994). Chem. Engng Sci., 24A, 4277.
Tonkovich, A.L., Carr, R.W., & Aris, R. (1993). Science, 262, 221.
Tonkovich, A.L., & Carr, R.W. (1994). Chem. Engng Sci., 49 (24A), 4657.
U.S. Department of Commerce (1997). Statistical Abstract of the ;nited
States 1997 (117th Ed.).
U.S. Environmental Protection Agency (1996). he Presidential Green
Chemistry Challenge Awards Program. EPA744-K-96-001.
U.S. Environmental Protection Agency (1997). he Presidential Green
Chemistry Challenge Awards Program. EPA744-S-97-001.
Wan, J.S.K., & Koch, T.A. (1994). Res. Chem. Intermed., 20, 29.
Webb, S.W., Weist, E.L., Chiovetta, M.G., Laurence, R.L., & Conner,
W.C., (1991). Canad. J. Chem. Engng, 69, 665.
Zantho!, H., Zhang, Z., Grzybek, T., Lehmann, L., &Bearns, M. (1992).
Catalysis oday, 13, 469.
Zwinkels, M.F.M., Jaras, S.G., Menon, P.G., & Gri$n, T.A. (1993)
Catal. Rev. Sci. Engng, 35, 319.
C.J. Pereira/Chemical Engineering Science 54 (1999) 1959}1973 1973