Plastic Technical
Paper
An Overview of Recycling Plastics from
Durable Goods: Challenges and Opportunities
Dr. Michael B. Biddle
President, MBA Polymers, Inc.
Peter Dinger
Director of Technology, American Plastics Council
Dr. Michael M. Fisher,
Director of Technology, American Plastics Council
Introduction
The recycling of plastics from packaging, particularly bottles,
has grown significantly during the last ten years for most
industrial countries. While the recycling industry has
experienced significant market challenges due to price
fluctuations, the recovery of polyethylene terepthalate (PET) and
high-density polyethylene (HDPE) is still being carried out in
numerous large scale operations throughout the world. The growth
of bottle recycling has been facilitated by the development of
processing technologies that increase product purities and reduce
operational costs. The recovery of plastics from other streams,
such as durable goods is growing in interest. This stream
includes items like: automobiles, appliances, computer and
business equipment, electrical goods, and even sporting
equipment. This interest is being driven by a number of factors
such as actual and proposed take-back and producer responsibility
legislation on end-of-life products, as well as green marketing
initiatives by a number of durable goods manufacturers. Metals
and even reusable components are frequently recovered from these
streams of end-of-life equipment, but the plastics are not being
recovered at similar levels. This is due to the fact that the
plastics used in these goods are newer engineering materials and
represent greater recovery challenges—both technically and
economically. Many end-users and plastic recyclers recognize that
the plastics used in durable goods are often more valuable than
those found in packaging. The recovery of these plastics,
however, is complicated by a number of unique challenges, such
as: a much wider range of different and incompatible plastics; a
less developed collection infrastructure; more varied end
products; lower overall volumes of these 2 materials,
particularly on an individual grade basis; and a much wider range
of attached foreign materials such as metal, rubber, foams,
fabrics, etc. This paper prepared for IdentiPlast II* will first
review the unique challenges associated to this part of the
recycling business, then discuss approaches to address these
challenges. It will provide an overview of the general approach
used to recover plastics from durable good streams. Much of the
work on recycling plastics from durable goods has been sponsored
by the American Plastics Council (APC). Specifically, the paper
will briefly review the main steps associated to the recovery of
plastics from end-of-life durable goods shown as a recycling loop
in Figure 1:
1)
Identification and Sorting,
2) Size
Reduction and Liberation,
3) Separation
of Nonplastic Materials,
4) Separation
of Mixed Plastics, and
5) Upgrading.
It will touch
on some of the major advances in these areas, and set the stage
for the more detailed discussions that will follow from other
presenters. Discussion As with other types of materials such as
metals and glasses, different types of postuse plastics must be
separated from foreign materials and from one another to achieve
good performance and acceptable market values. Most plastics are
not highly compatible with one another, and while some commingled
applications have been demonstrated, particularly with
compatibilization technology, they typically capture much lower
values than virgin plastic. In summary, the primary reasons for
segregating plastics by type are:
• Most foreign
material contaminants decrease the performance of the host
material, and most plastics are not uniformly compatible.
• The
properties will be consistent and understood. Even if compatible
blends are produced, it would be difficult to ensure consistent
composition of a blend made from a mixed recycle stream. Markets
for other than generic resins are difficult to define.
• The maximum
formulation and upgrading flexibility is available to materials
in a pure state. Some method to sort plastics by type is required
because there are many different types of plastics used in
plastic products (particularly in the durables area) and the
pre-labeling of plastics to assist with identification does not
facilitate the sorting in a factory environment. * IdentiPlast
II, April 26-28, 1999, Brussels, Belgium, sponsored by APME, APC,
PWMI, and EuPC is the second international conference on plastics
recycling technology. IdentiPlast I, held October 1997, focused
on plastics identification and sorting technology. IdentiPlast II
focuses on plastics separation technology. Furthermore, part
labeling will not impact most end-of-life durables streams for
many years, reflecting the long life-cycle of many of these
products. . One approach is to first identify the plastic and
then effect some type of mechanically assisted sorting based on
that identification. Another common approach is to depend on some
type of intrinsic material property, such as density, to effect
the sorting during the recycling operation. Both approaches will
be discussed in the following sections. Identification and
Sorting Technology In general, this approach refers to a fast and
accurate identification of the primary plastic contained in a
particular item followed by some type of manual or automated
sorting of that item based on the identification. This area was
the subject of the first IdentiPlast Conference, held in Brussels
in October 1997, therefore only a general overview will be
provided. The reader is referred to the proceedings from that
conference and to an APME (Association of Plastics Manufacturers
in Europe) summary paper for more detailed information (reference
1). In the area of rigid plastics packaging, high speed automated
plastic bottle sorting technology has come a long way since it
was first envisioned many years ago as a way to meet demanding
quality requirements. Bottles can now be identified and sorted at
rates of over one ton per hour in some cases. In the most typical
approaches, the system scans each bottle multiple times, and
sometimes with multiple types of radiation (visible light,
infrared light, X-ray) as it passes on a rapidly moving conveyor
belt. The multiple scans ensure that the plastic has been
measured independently from the labels or other non-plastic
items, while the multiple types of radiation are used to pinpoint
the plastic’s chemical make-up. This type of approach requires an
identification speed of hundreds of measurements per second. This
performance is a significant improvement from the many minutes of
time required for the quickest laboratory methods just a few
years ago, and would not have been possible without significant
research effort. There are over 250 automated bottle-sorting
lines in commercial operation today. Advanced systems can
identify all of the commercially used packaging resins and can
sort by color. Other systems are in place which sort flakes by
color and resin type (PVC and PET only today) at rates of over
two tons per hour. Reviews of some of these bottle-sorting
technologies can be found in references 2 through 7. While the
recycling of plastics from bottles is widely practiced
commercially, the recycling of plastics from durable goods such
as automobiles, computer and electronic equipment, appliances,
building and construction and even sporting goods, is a more
recent interest as the recovery of these products becomes more
commonplace (8-11). As with plastics from packaging, the plastics
used in durable goods must be sorted according to type.
Unfortunately, the automated bottle sorting technologies that
were mentioned earlier are not applicable to most of the plastics
found in end-of-life durable products for a variety of reasons:
• Durable
parts come in a much wider variety of shapes and sizes compared
to bottles.
• The average wall thickness of items is much greater in durable
goods.
• The parts from durable products are often opaque and often
contain carbon black.
• Coatings are used much more frequently for both decorative and
functional reasons on plastics from durable goods.
• There are a much greater variety of plastic types found in
durable products to meet the correspondingly wide variety of
performance requirements. The wide variety of shapes and sizes
implies that the parts from durable goods will be difficult to
“singulate” on a traditional conveying system. It also makes
probing with a remote sensing device more difficult because the
orientation of the surface and the distance to the surface with
respect to the probe may change significantly with each part. The
thicker walls and part opaqueness, make energy transmission
through the part much more difficult. Carbon black, in
particular, absorbs much of the radiation from traditional
spectroscopic identification techniques, making it difficult to
obtain information from the underlying host polymer. Some bottle
sorting techniques also rely on the simple fact that the degree
of light transmission through a sample provides information
regarding polymer type. Light transmission is not possible
through most plastic parts from durable goods. Coatings,
particular metallic ones, also interfere with most types of
analysis techniques. Finally, a large variety of plastic types,
fillers, reinforcements and additive combinations are found in
durable goods to meet the wide variety of aesthetic and
performance needs in modern durable products. This means that the
identification and sorting system must be capable of
accommodating a much wider variety of materials than those
developed for bottles, which typically focus on just three to
five materials. Beyond the differences in technical challenges,
there are several factors associated with durable goods that make
a slower and possibly more manual approach feasible, at least as
an interim technique:
• Durable parts are more likely to be handled manually at some
point in the recovery process (disassembly).
• In many cases, the value of the plastic used in a durable
product is greater than that used in packaging.
• The average plastic part from a durable product weighs
considerably more than an individual plastic bottle.
Bottles
are collected at curbside and are found in high piece volumes of
similar parts, whereas most durable parts are collected by
numerous different recycling infrastructures and occur in a much
lower number of individual pieces, particularly of similar parts.
All of these factors suggest that a slower identification
technique could be economically feasible for plastics coming from
durable products. The parts are already frequently being handled
manually in existing dismantling infrastructures, so a manual
device could be easily incorporated into the process.
Furthermore, each identification and sorting event, on the
average, will result in greater value for durable parts than
bottles because the amount of plastic being sorted is larger,
and, in many cases, the value of the plastic itself is greater.
Finally, it may prove difficult to collect large numbers of
individual plastic parts from durable goods in one location,
making a justification for potentially expensive and complicated
automated technologies more difficult. The identification and
sorting considerations just discussed and their implications are
summarized in Table 1. As described above, the implications
suggest that a manual approach might be economically and
technically feasible for some of the plastics coming from
specific durable products. It was also clear that an automated
approach would take much longer to develop. For these reasons,
both manual and automated approaches have been pursued
independently and concurrently by different companies and
research organizations. Three types of manual equipment are being
developed worldwide: hand-held, portable and bench. As might be
expected, truly hand-held devices might not be able to identify
the full spectrum of polymers, but could be very useful in niche
applications. Bench top units are typically modified laboratory
instruments that have been adapted to facilitate rapid plastics
identification. Portable units attempt to bridge the gap between
truly handheld and bench-top units, much like laptop computers do
between the personal digital assistants (PDAs) and desktop
computers. The most desired device characteristics are summarized
below: •
Accurate identification (less than 1%
error rate) •
The ability to identify a wide variety of
plastics in any color •
Fast response times (less than five
seconds) •
Portable and rugged enough to be used in a
recycling environment •
Economical enough for widespread use by
recyclers •
Easy to use by non-professionals Before
1994, no commercial devices existed that met the desired
instrument characteristics. In fact, many minutes were required
for an experienced spectroscopist to identify the type of plastic
from which a sample was made, and the sample preparation was
usually somewhat tedious and destructive.
6 The
biggest single challenge in both developing and evaluating
technologies to meet these goals is the wide variety of plastics
used in durable products. Even within a specific plastic family,
such as ABS, the number of formulation varieties available may be
well over 100. This high degree of material tailorability has
been one of the primary reasons for the dramatic increase in the
use of plastics in durable products. Numerous additives can be
used to achieve certain properties and the pigments used in
polymers can also vary significantly, particularly between
different application areas. Finally, many of the plastics from
end-of-life durable products were formulated ten or more years
ago and the materials have continuously changed and improved. The
American Plastics Council and MBA Polymers, Inc., in particular,
developed an extensive calibration and testing library containing
some of the most commonly used plastics based on market data and
feedback from endusers. This library has been used to help
develop, evaluate and demonstrate many different technologies
(11). Numerous techniques have been proposed and explored by
various organizations throughout the world. Some of the most
notable technologies are listed below:
• Mid infrared
spectroscopy (MIR) or Fourier Transform Infrared spectroscopy (FTIR)
• Near
infrared spectroscopy (NIR) •
Shortwave NIR (SWNIR)
• Raman
spectroscopy •
Pyrolysis mass spectroscopy (Py-MS)
• Pyrolysis IR
spectroscopy (Py-IR) •
Laser-induced emission spectral analysis (LIESA)
•
Infrared thermography •
X-ray methods •
Triboelectric property measurements
Each technology has its own set of advantages and disadvantages,
many of which are briefly summarized in Table 2. A much more
detailed discussion on these techniques is contained in reference
1. Additional discussions of plastics identification technology
can be found in references 12 through 20. No technology has yet
been identified which addresses all of the needs of
identification and sorting technology for durable goods, however
new ones are being developed all of the time.
More Challenges
Even if all of the plastic parts
from a durable goods stream can be identified and sorted into
different categories, many challenges remain before most parts
can be transformed into reusable material. They may contain
various paint and coatings, 7
the must be size reduced, and most
will be made of many mixed materials that are attached to one
another. These challenges are reviewed below.
Paints and Coatings
Paints, coatings and coverings (such as
fabrics, sheets and films) are encountered among plastics from
durable goods rather often, and represent challenges to both the
identification and recycling of plastics. In the appliance,
electronic, computer and automotive industries, paints and
coatings can be used for both decorative and functional reasons.
Paint and coatings, if not removed, can cause property reductions
in some recycled plastics from stress concentrations created by
the coating particles. Degradation of the coating can also lead
to chemical degradation of the plastic during reprocessing. The
level of potential property reduction depends on the combination
of the type of plastic substrate, coating type and coating
thickness. Appearance properties and surface characteristics can
also be effected by residual paint and coatings. The paints and
coatings must usually be removed or rendered compatible with the
plastic substrate to achieve the highest possible mechanical
properties of the recycled material, although there have been
reports of good property retention with certain coating/substrate
combinations without any special attention to the coatings (21).
The approach taken to remove or compatibilize the coating depends
on the nature of the coating and its interaction with the
substrate material. As noted in the Identification and Sorting
Section, there are many different types of plastics used in
durable products. When these are multiplied by the numerous
varieties of coatings, the number of different coating/substrate
combinations are staggering. It is unlikely that any single
technique would be optimal for all combinations. The chrome from
plated plastics has been recovered for years with simple
grinding, sometimes assisted with cryogenic methods (to enhance
the liberation process and prevent the plating from being
embedded in the plastic granules). This has been widely practiced
due to the value of the chrome and its ease of separation from
the plastic using strong magnets (some of the material in the
"chrome" coating is slightly magnetic). Fine grinding of most
painted plastics may result in a fair amount of liberation, but
the separation of the paint and plastic particles becomes
difficult, if not impossible. The aerospace industry has
developed numerous abrasive paint removal techniques in response
to environmental concerns with solvent stripping methods. These
techniques, however, are more applicable to large whole parts and
a manual approach. Several continuous and automated abrasive
techniques were 8
investigated as part of an APC project
using large flakes of coated plastics in an effort to identify a
dry coating removal technique, but none proved completely
satisfactory. High temperature aqueous-based approaches
demonstrated through APC projects and by other organizations,
however, have shown promise on many coatings and substrates, and
continue to be investigated further (22-25). The high temperature
aqueous environment can hydrolyze many coatings, but the plastic
substrates might also be susceptible to degradation, so the
processing conditions must be very carefully controlled. In the
case of olefin-based car bumpers, Toyota has demonstrated that
the coatings can be changed sufficiently using a high temperature
water process to compatibilize them with the plastic, and removal
of the paint is not deemed necessary (26). The olefinic plastic
is apparently not degraded under these conditions, however this
particular plastic is less susceptible to hydrolytic degradation
than some of the other engineering thermoplastics (particularly
condensation polymers).
Size Reduction
Most of the plastic parts returning from
EOL durable goods have other materials attached to them, such as
ferrous and nonferrous metal inserts, screws, bolts, clips,
brackets, etc.; metal, paper and plastic labels; foam insulation;
wiring and mixed plastics. It is usually not economically
feasible to remove most of these items manually, so they must be
liberated and separated in an automated fashion if the plastic is
to be recycled. Before any automated separation can be performed,
the parts must be size reduced and the foreign material
contaminants liberated from one another. The size reduction must
be sufficiently extensive and vigorous to cause even molded-in
items to be liberated or to create particles in which the
contaminant represents the majority of the particle volume so
that its characteristics can be used to effect a separation.
Furthermore, many of the downstream separation techniques require
that the particles be fairly uniform in size and shape for
efficient performance.
In summary, the size reduction step has three primary purposes:
1) generation of particles that can be more easily handled than
bulky parts,
2) generation of uniformly sized and shaped particles that can
be separated effectively in downstream processes and
3) liberation of dissimilar materials from one another. Plastic
bottle recycling size reduction challenges have been mostly met
by shredders and standard granulators.
Plastics from durable goods, however, have several unique
characteristics that make this step even more challenging.
Plastic parts from durable goods
come in a wider variety of shapes and are usually much larger.
9
Many of the plastic parts contain
significant amounts of metal that can damage traditional plastics
size reduction equipment, such as granulators.
The parts often contain other materials intimately bonded to the
plastic substrate, requiring aggressive liberation.
The parts are thicker and the
materials are stronger, making size reduction more difficult.
Traditional plastics size reduction involves high-speed
granulators with fixed screens or grates to control particle
size. The knives of the granulators can be quickly dulled or even
damaged by hard materials like most metals. In some cases, the
high speed rotating knives can catastrophically fail upon hitting
metal pieces and cause significant damage to the equipment. Some
granulator manufacturers have developed more robust knives that
can accommodate limited metal contamination, but not the full
range of metal found in end-of-life durable products. Metal
detectors can be placed in line before the granulation step, but
this is not an effective metal removal technique for streams with
significant metal content because the devices reject some of the
target plastic with the metal each time metal is detected. In
some respects, shredders and hammer mills are at the other
extreme of the recycling size reduction spectrum. They are used
by metal recyclers to size reduce items such as entire
automobiles and large appliances, and are generally designed to
perform coarse size reduction and liberation with high
throughputs. Shear shredders operate at low speeds and rely on
stacked opposing circular cutters with "hooks" or fingers on two
counter-rotating shafts to grab and shear the materials in a
single pass, while hammer mills operate at much higher speeds and
beat the material until it is small enough to fit through the
openings in screens or grates typically fixed below the units.
None of these techniques proved viable as solitary size reduction
operations for plastics from durable goods. Studies showed that
traditional granulators could not accommodate significant amounts
of metal, standard shredders were not effective at producing well
controlled particle sizes or adequate liberation and hammer mills
were noisy, were not effective at close particle size control,
generated a large amount of fines and actually imbedded
contaminants in the plastics in a few instances. However, a
combination of these techniques used with metal removal equipment
(see Separations Section) could be used to provide the necessary
size control and liberation. As a result of considerable
searching, developmental efforts and trials undertaken as part of
an APC project, several types of stand-alone equipment were
eventually uncovered which provided the necessary size reduction
and material liberation necessary for large plastic items with
significant metal content: four-shaft shear shredders, modified
two-shaft shear shredders and rotary grinders.
10 The
four-shaft shear shredders are most commonly found in Europe and
often with a removable fixed screen for size control. Like a
standard shredder, they can accommodate high metal content
material, but the four shaft design provides added cutting
opportunities and an efficient recirculation function that causes
the material to pass through the cutters multiple times until it
can pass through the holes of the screen, thus producing a
particle with the desired size characteristics and significant
liberation achieved with granulators, but with the capability of
accommodating higher amounts of hard materials. Some standard
two-shaft shredder manufacturers have also realized the need to
generate smaller and more controlled particles for some
applications, and have developed narrower cutters and fixed
screens that can be placed below the shredder to cause the
material to experience repeated cuts until it can fit throughout
the screen openings. The placement of screens on shredders,
however, significantly reduces the achievable throughput rates.
Rotary grinders were initially developed to size reduce wood. The
typical design employs numerous one to two-inch square teeth
mounted in various patterns on a horizontally rotating shaft.
These teeth take small "bites" from material that is pushed into
the rotating shaft via a large sliding ram. The ram is typically
equipped with a load sensor to maintain a fairly uniform pressure
at the cutting interface and to indicate when the ram should
retract to allow additional material to fill the grinding area.
These units, which operate at speeds between that of granulators
and shear shredders, can be fitted with a wide size range of
screens sizes for particle size control. This type of equipment
can accommodate rather large parts and moderate amounts of metal,
but typically not as great in either case as a shredder. Once the
majority of the metal is removed from these streams, more
traditional size reduction techniques can be used to cause
further liberation, such as granulation and milling. These
additional size reduction steps are usually only necessary for
recycle streams containing very well-adhered foreign material
requiring aggressive liberation. Alternatively, or additionally,
wet and cryogenic techniques can be used to enhance liberation
(see Separations Section). A summary of some of the major size
reduction approaches is presented in Table 3. APC and MBA
Polymers put together a report that summarizes many different
types of size reduction technologies that were evaluated over
several years. That report should be available through APC soon.
Materials Separations Technology
As noted above, even if the
plastic part has been identified to determine the primary plastic
from which it was made and sorted into groups of parts made of
this plastic before size reduction, in most cases there will be
many different types of materials mixed together because very few
plastic parts used in durable goods have nothing attached to
them. If these parts have been size reduced into sufficiently
small pieces to liberate most or all the different materials from
the target plastic(s), the result is a mixture of commingled
flakes of various materials that may include:
The target
plastic(s)–usually the majority component(s)
Ferrous metals
Nonferrous metals
Paper, plastic film and other label
materials
Foam
Fabric
Cables and wiring
Glass
Wood
Other plastics
Other materials
These foreign materials must be separated
from the target materials for recovery. Furthermore, these
foreign materials themselves should be separated into pure steams
to the greatest extent possible because it enhances their
recyclability as well. Technologies to facilitate these
separations have been developed by entrepreneurs, equipment
manufacturers and the plastics industry (8, 9 10, 27, 28). In
many cases, the technologies have been borrowed from other
industries, such as food processing, agriculture, mining, waste
management, and plastics processing. APC and MBA Polymers
concurrently with the investigations into processes for
identification technology and size reduction equipment also
carried out investigations into material separation technologies.
One APC Project, designated M-234, specifically investigated
methods to liberate and separate automotive parts with coverings,
such as seats, instrument panels and interior door panels. This
project led to the development of a mechanical process to recover
polyurethane foam from car seats. A report was written by APC and
MBA on this project and will be available for publication soon.
The work on separation processes was much broader than that
carried out under APC Project M-234. In fact, two other APC
projects were undertaken: one to investigate existing separation
technology (M-131), and one to investigate new and advanced
mechanical recycling technology (M-132). APC worked with MBA and
wTe Corporation on Project M-131 and with MBA on Project M-132.
Some of the major techniques that were investigated under these
projects are discussed briefly below. These were also summarized
at a Workshop put together by APC and MBA Polymers for the
Society of Plastics Engineers’ Annual Recycling Conference in
Chicago in 1998 (ARC98).
Plastic People
Knowledge Base
Knowledge Base
