Full Text CEPIS/OPS/OMS

Automatic Sorting for Mixed Plastics

Introduction
Available sorting systems
Marker Systems
Where will systems be used?


Introduction

NCREASINCLY, designs for large scale materials recovery facilities specify that the majority of sorting is done automatically. Magnets, trommels and air jets, for example, separate tin from glass and cans from plastics. But further separation within categories (paper by type, glass by color, plastics by resin) continues to be done by hand. Experts are predicting that this too will soom become a thing of the past. We are on the verge of a technical revolution that will extend automatic sorting to individual materials - specially for plastics.

In the rnid-l980s, The Society of the Plastics Industry, Inc. developed a resin identification system compatible with manual sorting which was designed to help handlers and reclaimers identify the plastic used in a particular brand or product line. Although it helped meet an immediate need, everyone acknowledged that automatic sorting technology would soon be required.

In recycling programs where HDPE milk jugs and PET soft drink bottles are the only plastics collected, separation is fairly simple. Reasonable sorting rates of 500 to 600 pph (pounds per hour) are achieved. But as residents set out more types of plastics, sorting problems are significantly compounded. Now a variety of shapes, sizes and colors are added, and decreases in both rate and quality are common. In addition, because turnover is high among sorting line workers consistency is difficult to achieve.

The automatic sorting systems being developed by entrepreneurs, corporations and universities are designed to improve efficiency and quality, and reduce the cost of plastics recycling. Automatic plastics separation can be accomplished at three different levels: macro, micro, or molecular. This report focuses on macro separation of whole containers.

Macro separation of plastics is the only type of sorting currently practiced commercially. Manual sorting is macro separation and is also the first method of sorting that is being automated. The pioneer in this technology is Prof. Henry Frankel, at the Rutgers University Center for Plastics Recycling Research. In 1989, with funding from the Council for Solid Waste Solutions and the Plastics Recycling Foundation, Dr. Frankel developed a system to separate PET, unpigmented HDPE and pigmented containers using the light transmission properties of the resins. During the past three years, he has brought this concept to a single station, multidetector system with computer software to track the containers and eject them by resin and color into separate downtream containers.

Because PET and PVC are contaminants te each other when mixed, industry pressure for improved quality of the recycled streams led him the look for a detector that could identify the PVC that occasionally appeared in the PET stream. He enenuraged Asoma Instruments, Inc. of Austin, Texas to modify one of its standard laboratory instruments so that it could respond to this need. Asoma was able to provide an instrument with an ultra fast analysis capability (200 analyses per second), so that it could check each bottle many times as it went by. This made possible accurate identification of bottles traveling past the detector at high speeds, despite interference's from labels, dirt, etc. The systems subject the containers to low-level radiation and the chlorine atoms in the PVC molecules respond with a unique peak in the x-ray spectrum which is detected by the instrument.

Significant improvements have been made since the first unit was sold in 1989. Each of Asoma's Model VS-2's can identify singulated PVC bottles traveling anywhere in a one foot wide channel at a speed of 10 feet per second, with a miss rate of less than one bottle per 100,000. Several units can be placed side by side to cover wider channels. These units sell for $35,000 to $45,O00. The limiting factor on accuracy is determined more by the mechanical details of the material handling system than by the detectors as is the case for most macro sorting systems.


Available Sorting Systems

National Recovery Technologies Inc. (NRT) of Nashville, Tenn., has developed a system called Vinyl Cycle based on a principle similar to that of the Asoma PVC detector. The system works on whole or crushed bottles, but does not require that the containers be delivered single file. Vinyl Cycle consists of a row of detectors beneath the width of a conveyor belt. When a PVC container passes over a detector, its position on the belt is noted. The container is computer tracked and ejected by an air jet further down stream .Other containers in the proximity of the PVC container are ejected along with the PVC, so the result is a PVC-rich stream while the original feed stream is essentially PVC-free. Initial experience with commercially installed units has shown a better than 99 percent recovery rate of vinyl containers.

NRT manufactures machines that process a mixed plastic container stream at a rate of 2,000 or 5,000 pounds per hour. The 2,000 pph unit sells for under $100,000, and the 5,000 pph unit sells for under $200,000. The actual price depends on the specific design parameters. The smaller machine is a new offering, but eight 5,000 pph VinylCycle systems were sold during the past two years.

These pioneering efforts led to the recent introduction of an integrated high-speed sorting system for the separation of the four most commonly collected packaging plastics: PET, PVC, HDPE, and PP. Commercialized with the assistance of the Council for Solid Waste Solutions, the first such automatic plastic sorting system was sold to Eaglebrook Plastics in Chicago and started up in late December 1991. It was engineered by Magnetic Separation Systems (MSS) of Nashville, Tenn., to scrl 5,000 pounds per hour of mixed plastic bottles, and consists of four parallel lines fed by a bale breaker and singulator system. The process works with either whole or baled bottles.

The MSS «Bottlesort» system employs a primary detector to separate the bottles into three streams: PET and PVC; unpigmented HDPE and PP; and mixed color opaque. The process can be designed to stop at this point or separate the three prirnary streams into subcategories by the addition of other sensor and eject modules. Four other modules are available: PVC Module - separates PVC from PET bottles, PET Module - separates green and amber PET from clear PET; PP Module - separates unpigmented PP from unpigmented HDPE and; Color Module - separates mixed color opaque into seven individual colors or combinations of colors. The system has the ability to add as few or as many modules as required to handle the incoming material. It could be designed as a hand-fed, single line handling 1,250 pph using only the primary detector, or a multiline system integrated with a bale breaker and singulator employing all of the modules and diverting the separated streams directly into individual grinders. Costs for these systems would range from $65,000 for a basic hand-fed detector to $750,000 for a 5,000 pph system with all the options.

A second macro system for sorting baled bottles funded by the CSWS is being developed by Autornation Industrial Control (AIC) of Baltimore, Maryland. This system uses a pair of detectors at a single station - one for determining resin type and one for identifying color - to control the separation. The system is capable of detecting and separating all of the common packaging resins (PET, HDPE, LDPE, PVC, PP, PS) and polycarbonate (PC).

A demonstration line currently is under construction and is designed to process three bottles per second (approximately 1,500 pph). The bottles will be carried to the detector by a singulator (equipment which positions the bottles in a single file prior to the detector) developed by Chamberlain MRC of Hunt Valley, Maryland, a subcontractor to AIC. The prototype line, expected to be completed in June 1992, will be designed to separate bottles into specific categories of resin/color combinations with one pass. The demonstration line will provide for separation into seven categories such as clear PET, natural PP, red HDPE, blue PVC, etc. The categories may be changed by modifying the computer programs. The cost of the AIC system will not be determined until after the completion of the prototype demonstration, but it is anticipated that a basic 1,500 pph system (singulator, detector and diversion system) will be in the range of $350,000 to $400,000, and 3,000 pph system from $425,000 to $475,000.


Marker Systems

The concept of marking either the container or the resin with some type of "flag" that would be readily detectable is one that has attracted the attention of a number of major corporations. Continental Container Corporation developed the concept of marking all containers with an easily detectable invisible ink which would allow separation not only by resin type, but also by color, resin additive package,. and package contents, if desired. The major drawback of this type of system is the overall cost - even if the price of each individual system were low, the need to install a marking unit on every packaging production line in the country would be significant. For the system to be effective, this concept requires all of the involved organizations, both producers and recycles, to participate.

Eastman Chemical Cornpany announced in December the development of a molecular marker that could be used to identify each type of resin. Eastman has indicated that the cost of adding the markers would be relatively low and that it would be able to distinguish between different grades of the same resin - a task that could not be accomplished with any of the other systems currently under development.1 This system also would require the participation of all resin producers, but this is considerably easier than providing markers on containers since it is within a common industry.


Where Will Systems be Used?

The relatively high capital cost of macro sorting systems (compared to the cost of a conveyor, for example) will limit the number of locations that purchase them. But the deciding factor for most buyers likely will be the volume of material they are processing on a monthly or annual basis. This would lead one to predict that regional handlers and PRFs (plastic recovery facilities) - where the economies of scale would be conducive to investing in systems to improve speed and quality - will install the majority of automatic systems. There are a large number of MRFs -.handlers and recycles who process sufficient volumes of material to warrant this level of investment.

________________________
Peter Dinger is director of technology for the Council for Solid Waste Solutions, the plastics industry's task force, based in Washington. D.C. Developments in automatic plastics separation at the micro and molecular levels (Part II) will be covered in an upcoming issue.
________________________