Miljøprojekt, 1054 – Affaldssortering baseret på online detektion af grundstoffer

Affaldssortering baseret på online detektion af grundstoffer

Summary and conclusions

This project is the continuation of work by the Danish EPA and FORCE Technology on the development of sensors dedicated to waste sorting. This project is based on the sensors developed by the project partners in the project, "Sensor for detection of salt impregnated waste wood" - based on the principle of, Prompt Gamma Neutron Activation Analysis, PGNAA.

The purpose of this project was:

1. to optimize PGNAA-sensors,

2. to examine the possibilities for using PGNAA-sensors in to the benefit of the environment,

3. to demonstrate and document the sorting of a single waste stream in a pilot scale test.

The main activities were to:

build the theoretical and technological basis for design, optimization, and construction of PGNAA-sensors for a given task,
optimize the sensor through a theoretical design study, also clarifying the technological potentials and requirements of different applications,
investigate the demand with respect to problematic waste streams with high treatment costs,
plan, design and build a pilot sorting plant demonstrating sorting of a single selected waste stream,
conduct sorting tests, improve the sensor and characterize or sort other waste streams/items,
disseminate the results of the project at one international conference.

Need analysis (paragraph 3)

Based on detailed investigations and calculation of the impact on the environment, as well as of the economical potential, the steering committee of the project prioritized the focus of the project on the following waste fractions:

1. PVC

2. Industrial waste

3. Circuit boards: precious metals, PCB-capacitors, etc.

4. Quality control/training at municipal recycling stations

5. Fuel pellets

6. Slag from waste incineration plants

7. Waste wood

Because the same companies normally treat circuit boards with precious metals and PCB-capacitors, these waste fractions are discussed together.

Outside this project, FORCE Technology has examined the economical potential of sorting used batteries based on the detection of elements by PGNAA. The assessment is not described in detail in this report, but it clearly confirms that PGNAA can detect NiCd-batteries in a mixed battery fraction with high certainty and speed. The speed of sorting matches commercial conditions without any problems.

We estimate that the 15 old EU Member States will have a demand for 40-50 battery sorting plants with a capacity of 3,000-7,000 tonnes/year/plant in the near future. Where the plants will be located and how big they will be depend on the coming EU Directive and, of course, on how the market players will organise.

Advantages of sorting out PVC

The impact on the environment from sorting out PVC from a mixed waste plastic fraction is very positive, since less PVC will end up in waste incineration plants decreasing the need of deposit capacity for flue gas cleaning residues and the tendency of dioxin formation. The sorting plant will make it possible to exploit plastic waste containing low concentrations of PVC and thereby increasing the trade potential of PVC-treatment plants.

The economy of a commercial sorting plant is very good – we estimate the treatment costs to be in the area of 26 DKK/tonnes (3.5 EUR/tonnes) sorted PVC for a plant treating 48,000 tonnes PVC-waste/year.

Commercial sorting actuators available on the market can do the sorting job, which combined with the high sensitivity of PGNAA gives a high probability of success.

Technical design study

In the theoretical design study knowledge about possible ways of constructing PGNAA-sensors was collected and systematized, just as several aspects about the connection between design variables and the performance of the sensors were considered.

Moreover, we developed methods and software for neutron transport simulation in relevant materials and interactions between neutrons and these materials. In this context, we developed a scalable simulation model for a class of PGNAA-sensors, including the sensor used in the pilot plant.

Finally, we set up the theoretical tool needed for analyzing the data from measurements, calibrating the sensor and planning and evaluating the quality of the fractionation for a certain sorting problem.

Pilot plant

The pilot plant was constructed according to the configuration: "common moderator, two neutron sources, and two detectors," as described in the theoretical design study. The plant was constructed with possibility of characterizing and/or sorting other types of waste than PVC-containing plastic waste.

The detection field was designed "as narrow as possible" (giving the highest sensitivity), still allowing reasonable test samples to pass through. This was done primarily in preparation for unknown sample sizes of PVC. The size of the detection field was designed so the plant has capacity for industrial use.

Tests

Two types of tests were performed:

Stationary tests: Several samples were measured for a longer period, while located in the detection field.
Continuous tests: Samples were transported through the sensor by a conveying belt, while time series of the sensor signal were recorded.

Some of the results from the stationary tests were used for calibration.

The conclusions from the tests are:

PVC:

Sorting of a mixed fraction of plastic waste in PVC / non-PVC is possible and easy. The data analyses show that the characterization is sufficiently robust for the sorting job.

The speed of characterization is sufficient for industrial use, sorting thousands of tonnes/year.

PCB-capacitors

We can conclude that it is possible to sort out capacitors in PCB-containing and non-PCB-containing fractions.

The speed of characterization is sufficient for industrial use, sorting thousands of tonnes/year.

NiCd-batteries

Sorting in NiCd / non-NiCd-containing fractions is possible and easy. The data analyses show that the characterization is sufficiently robust for the sorting job.

The speed of characterization is sufficient for industrial use, sorting thousands of tonnes/year.

Future tasks

The possibilities of PGNAAs in PVC treatment processes

Several tests on the demonstration plant, sorting out both soft and hard PVC from a mixed plastic waste fraction, have proved that it is technically possible to sort out PVC pieces down to some 20g, with the certainty and speed required of an industrial plant. In traditional PVC-containing waste treatment processes – i.e. RGS90's two technologies – the PVC waste is shredded to a particle size of 2-3 mm (corresponding to a mass below 1g) before chemical/thermal treatment. We have proved, both theoretically and practically, that the time necessary to detect such small PVC particles would be unrealistically long (~ seconds), and that it would be economically realistic to implement a PGNAA sorting machine at a pre-sorting plant.

If shredding of the waste plastic fraction could be divided into two stages, the first one shredding the waste down to a particle size of 20g, a PGNAA-based sorting system would have a great and positive impact on the environment and be very economical.

FORCE Technology will inform the suppliers of PVC-treatment plants about this advantage.

Cadmium in PVC

Our tests have shown that PGNAA can detect Cadmium (Cd) in small quantities, e.g. as stabilizer in PVC, with great certainty and speed. The PGNAA-sensor can detect chlorine in PVC and Cd simultaneously, giving an extra opportunity of positive impacts on the environment and economical advantages.

Multi-sensor sorting of batteries

The PGNAA-sensor is efficient when detecting Cd, but the speed is more moderate when the task is to detect Manganese (Mn) and Nickel (Ni), which are very common elements in batteries. FORCE Technology has therefore invented a new concept based on the combination of an X-ray vision system and PGNAA.

FORCE Technology will be looking for funding for a pilot plant.

Determination of traces of Mercury in batteries

According to theory, PGNAA-sensors should be able to detect Mercury in relatively small quantities, but this requires a sensor equipped with special and cooled detectors. Such a sensor has not been installed at the existing pilot plant.

FORCE Technology will be looking for the funding for this purpose, when the demand is recognized.

Characterization/sorting of fuel pellets

The consumption of wood pellets for energy production has increased dramatically during the last few years and is expected to increase in the coming years. There is noticeable demand for fast analyses for both laboratory check and online control of fuel pellets quality. PGNAA could probably detect unwanted heavy metals and Nitrogen, which are important elements in the glue in chipboard and laminated wood, etc.

Systematic optimization of PGNAA-sensors

Through the theoretical design study FORCE Technology laid the basis of the implementation of a systematic design optimization tool. If the demand for PGNAA-based sorting machines increases as expected, e.g. for sorting batteries and capacitors, it would be a natural task to optimize the design phase with an automatic design tool.