Skip to main content

Automated Sorting and Recycling of Waste Lamps

Final Report Summary - ILLUMINATE (Automated Sorting and Recycling of Waste Lamps)

Executive Summary:
The Challenge

Currently there are 1500 million lamps and bulbs sold in Europe. The number of end-of-life fluorescent lamps in Europe will steadily grow from today’s levels of 150 million annually to a peak in 2020 of 800 million. This will put an increased demand on collection, recycling and material recovery, including the capture of critical rare earth materials. Current collection and transportation methods for waste lamps often give rise to extensive lamp breakage, materials contamination, and mercury emissions which may cause severe environmental impact, processing difficulties and lower materials value in the recycling step. Recycling operations in many European countries usually handle mixed fractions of bulb types which is then subjected to materials separation and recycling. In a mixed stream it is difficult to accurately detect/identify bulbs that contain mercury due to the vastly varying nature of the waste stream. ILLUMINATE aims to remedy these issues.
Project Objectives

The concept of the ILLUMINATE project is to develop methods and processes for two main areas of the supply chain: collection of the waste streams and sorting of the waste, then develop automated systems that are able to effectively sort bulbs into different classes and remove foreign objects. This is essential for an economically viable process. The unit will be a multisensor system, combining machine-based recognition technique with sensors measuring e.g. weight, colour and/or shape, and being able to rapidly recognize the main types of lamps (brands) in the waste stream. Once the identification and separation has been achieved the materials from both mercury containing and non mercury containing waste streams can then be handled by the appropriate processing steps in order to cost effectively recycle the waste bulbs.

The key enabling process is an automated sorting step which can rapidly and accurately identify and sort mixed lamp types into pure streams for dedicated pre-processing treatments. By adjusting the collection methods and hardware to the downstream processes, a seamless integration of collection, transportation, sorting, and pre-processing steps will be obtained. This will facilitate lamp waste treatment, maximize the recovery rates for glass, metals, plastics and critical rare earth materials and improve the working environment. Connected to the unit will be subsequent treatment lines adjusted to the specific and sorted fractions. RELIGHT and MERCURY have facilities across the EU and will host demonstration equipment that will allow the performance of sensor, separation and collection to be demonstrated in an industrial environment providing a key step in the commercialisation of the technology.

Key Results

The ILLUMINATE project has delivered efficient technologies capable of significantly and measurably reducing the ecological impact of material contamination during waste lighting collection and storage. Early implementation of the developed sorting technology should increase purity of end fractions and will help to shield European recycling SMEs from the threat of tightening regulations. The technology will additionally deliver substantial improvements in resource efficiency and benefit a large number of SME’s throughout the supply chain. It is the view of the participants that to achieve only recycling of lamps to low value outlets is not ambitious, and that the industry should be seeking to recover valuable materials back to a high value use. The partners in the ILLUMINATE project have developed technology that can be exploited by the SMEs involved creating opportunities for growth within these companies. This growth will be derived from the manufacture of the high tech systems required to sort and recycle the vastly increasing number of waste lamps, resulting in skilled jobs in the manufacture, operation and maintenance of these systems.
LED lamps are naturally possible to handle in the ILLUMINATE sorting system. The main problem is that LEDs do not need to be processed in an expansive Hg treatment process and when LEDs increase in the waste stream, they will clog the Hg treatment machines and therefore increase the costs. With the Illuminate project we are able to lift out the LED lamps and create cheap and adequate recycling processes for these lamps without allocating the expansive Hg treatment processes.

Project Context and Objectives:
The ILLUMINATE project aimed at developing a concept for the collection and sorting of lighting waste to enable a higher materials recovery rate and value, reduced mercury emissions, improved working conditions and waste statistics. The concept enabled enhanced compliance with the European Community directive 2002/96/EC on waste electrical and electronic equipment (the WEEE Directive) which, together with the RoHS Directive 2002/95/EC, became European Law in February 2003, setting collection, recycling and recovery targets for all types of electrical goods including lamps.

The ILLUMINATE target concept included:
i) Versatile, robust and easily transportable collection containers that prevents a) the breakage of lamps; b) the intrusion of snow/rain; c) emissions of mercury; and d) that, for safe and easy discharging, can be directly docked with
ii) An automated lamp sorting equipment which recognises and sorts the mixed waste lamp stream into relevant lamp streams for subsequent dedicated pre-processing, registers the number of mercury containing lamps sorted, and can be seamlessly integrated with existing lamp pre-processing equipment. The proposed technology will be focused on identifying hazardous lamps based on recognition of certain elements within the lamps and on characterizing shapes and sizes.
iii) A tool for generating and supporting statistics on Hg-waste lamps, including numbers and materials content, and detect trends in waste flow composition over time. The original plan was to also register and gather statistics on different manufacturers and brands but because of unforeseen changes in the consortium this will probably not be possible (The original DoW included Optisort´s detection technology based visual detection technology. Due to the bankruptcy of Optisort, another technological approach was proposed by CIT. The used technology based on specific sensors responding to materials/sizes does provide access to mapping certain lamp types but cannot distinguish between different manufacturers and brands.)

ILLUMINATE brings together leading experts in lighting recycling technology and sensor systems with a group of innovative SMEs in a multi-disciplinary project able to deliver significant technical advances. The key technical roles of the SMEs as either equipment manufacturers (MRT) or direct end users of the systems/processes (MER, REL, NOR, ELK) will focus the project towards commercial outputs that will result in excellent exploitation opportunities for the developed processes. The participants represent the value chain from collection and transportation to pre-processing and recycling of waste lamps. The overall goal is a comprehensive end-of-life solution for waste lamps ensuring the increase in quantity/quality of materials for reuse, reducing the impact on the environmental and human health, profitable recycling operations and improved working environment.
To achieve the ILLUMINATE project goal, key advances beyond the state of the art are required in technology for mechanical sorting and sensing. In addition to this a step change in thinking and education are required towards storage and collection of lighting waste.

Project Results:
Detailed studies on different types of lighting products, including market studies, literature studies and material studies were performed in WP1. This work provided the technical and statistical background to define critical process/sorting criteria for the processes developed in WP2 and WP3. The current collection processes for lamp waste in the EU were reviewed. The recycling SME partners of ILLUMINATE provided further data from their production lines to supplement the statistical data obtained from official sources. An extensive survey on lamp technologies and used materials was performed. Lamp waste composition and distribution of different lamps types in the waste stream were reviewed in detail. Lamp recycling processes were presented on a general level in ILLUMINATE.

The work on novel sensor technologies for detection/separation of mixed lamp waste streams were carried out in WP3 by CIT. In an initial study the material characteristics of different lamp types were evaluated and possible sensor types for detection were proposed. Positive detection results were obtained and a first prototype for the detection of CFLs was achieved in CITs labs. The prototype sensor system was integrated on a moving conveyor belt system running at industrial speed and during period 2 these sensors were integrated with a suitable feeding system for testing in an industrial environment. Relight SRL housed this prototype and ran the system with representative waste streams. The prototype and results obtained were showcased to a group of industry experts and the project has received excellent feedback.

The current collection processes for lamp waste in the EU were reviewed in detail by MER, NOR, REL, UCLAN and CIT. In the first WP of the project a questionnaire was developed by CIT in collaboration with UCLAN to collect data on the collection methods, storage and current waste processing methods. The questionnaire was sent out to SME partners (MER, NOR, REL) which provided the necessary data and statistics. CIT and UCLAN evaluated and validated the data. The collected data in this questionnaire provided the foundation for further discussions on defining sorting criteria for further work packages and process requirements for the automated sorting process to be developed in ILLUMINATE.
The collected data of the questionnaire was included in the deliverable reports D1.1 and D1.3. Details about collection, storage and sorting were furthermore provided by MER, REL and NOR. Three onsite-visits at MER, NOR and REL were conducted to facilitate understanding of current collection and recycling practice within the consortium. MER, REL and NOR also provided descriptions of their recycling process lines (flowcharts and pictures). Best practices and report leading examples (as well as bad examples) of collection methodologies and processing routes were reviewed. The necessity of adaptations of the sorting steps and process streams with variations in the lamp waste were discussed within the consortium
Statistical numbers on EU-production of lamps, imports into the European Union and exports out of the European Union were derived by CIT from EUROSTAT for the years 2007-2012 for different lamp types (except LED-lamps, which are currently not recorded in EUROSTAT). From these statistical numbers the number of lamps that entered the European market could by well estimated and trends on the lamp market for different lamp types were derived for this period. Other sources, such as market studies to predict trends in lighting were included in report D1.2 in particular for LED-lamps. The amounts of lamps that entered the market were compared with the collected amounts (households, commercial) in the EU (data from EUROSTAT, WEEE-forum). The statistical data from the official sources was compared with the statistical numbers on lamp waste processed at REL, MER and NOR. Specific tests of sorting municipal material and collection of statistics from these data were performed monthly by REL, providing data for WP1, WP2 and WP3. Mercury Recycling conducted extensive, monthly material tests and provided statistical data on their lamp waste processed. El Kretsen and NOR contributed to WP1 with very detailed statistical numbers for collected and processed lamp waste for Sweden and Norway for the years 2014 and 2015. Statistical irregularities and variations were discussed in detail between the recycling SMEs and CIT. The outcomes were summarized in report D1.2.
A major objective of the project was to provide a detailed description and quantification of the materials (metals, polymers, glass, other) faced in different lamp types. This information provides the necessary background for further technical (e.g. choice of separation processes, possible sensor technologies) and economic decisions in the forthcoming work-packages of the ILLUMINATE-project. The current EU legislation (e.g. ROHS-directive) and permitted amounts of mercury in lamps were summarized and compared with reported values from different sources. Mercury concentrations in lamps were not further analysed in detail, due to reported variations in the manufacturing processes and analytical methods, sample preparation and changes during the lamp lifetime. A comprehensive summary with reported values was found to be more accurate for the required decisions in ILLUMINATE. CIT/CTECH dismantled a number of lamps of different type and classified the materials into different fractions (plastics, metals, glass, electronic parts) in their laboratories. Certain parts were further analysed by CIT in detail (with a focus on CFL and LED-lamps) by different analytical methods (FT-IR, XRD-SEM).
Relight (REL) took responsibility for the ‘Definition of sorting criteria and performance’ and led the discussions in WP1 on defining sorting criteria and performance criteria for new processes and collection strategies to be developed within ILLUMINATE. The results were summarized in Deliverable report D1.1. but furthermore included in the reports D1.2. and D1.3. REL, NOR and MER provided insight on their current collection and their lamp sorting practice. The information was distributed among the project partners and discussed at SME onsite visits, project meetings and via teleconferences. The onsite visits complemented the data collected in the questionnaire prepared by CIT. It was found that collection strongly influences the complexity of current forthcoming process steps and future automated sorting processes. Percentages of additional, unwanted scrap and broken lamps that come along with the lamp waste were quantified by REL, NOR and MER. Furthermore, data on the current logistics and used collection containers were provided. Based on these numbers pre-sorting requirements for the automated sorting process in ILLUMINATE were discussed and appropriate target fractions after the sorting process were defined. The target fractions often do not tolerate impurities by other lamp types, thus additional purity criteria were defined for the target fractions. The minimum throughput requirements for economic processing of current amounts of lamp waste at the different recycling facilities were determined.
Most significant results
- Data on lamp materials for further R&D work packages was provided
- Statistical data on waste collection and waste processing was gathered from databases and the SME partners
- Existing recycling technologies were reviewed
- Current issues and challenges for the recycling SMEs were pointed out
- Requirements for integration of a new sorting process were determined
- Sorting criteria and performance criteria for the automated sorting process were defined by the consortium
- Requirements for additional manual/automatic pre-sorting were proposed

Work was carried out by MER throughout the project to identify the waste stream composition and the ongoing composition trends.

At the collection sites in Sweden, UK, Italy and most other countries lamps are sorted into linear and non-linear lamps, but not effectively these days. A common situation faced at the sites visited was that long linear fluorescent lamps were collected in a separate container, while CFLs, integrated CFLs, short linear tubes, circular fluorescent tubes, filament, halogen, etc. are collected in a separate smaller container. In many cases non-linear and linear lamps end-up mixed with each other in the containers. As a consequence, the lamp waste often has to be sorted manually to remove packaging odd-shaped lamps prior to the final decontamination and materials recycling process.

From the different site visits performed in the project, it has been concluded that the composition of the lamp waste in Figure 1 could be seen as representative for waste from household collection points throughout Europe.

The ILLUMINATE lamp recyclers’ (Relight, Mercury, Nordic) experience is that storing linear fluorescent and large circular fluorescent tubes with non-linear lamps leads to increased breakages. Based on the material and technology used in short linear fluorescent lamps, an option is to process these together with longer linear fluorescent lamps. In this case it will be beneficial to separate and store both the long and short linear lamps in the same container. The Illuminate concept looked at ways of reducing breakage and the need for manual pre-sorting at the collection and treatment facilities.

Figure 1: Presentation of waste across different EC sites.

The Illuminate project has demonstrated that the amount of information related to deposition of lamp waste at DCFs is currently limited to general signage and signage attached to the lamp waste containers (Sweden, Italy and UK). The appearance of a mixed lamp waste stream in terms of the collection system operated by MER is currently unavoidable. However, the levels of contamination and lamp breakages appearing in the lamp waste suggests that the public/DCF operators depositing lamps are unaware of the difficulties faced by the treatment facilities during sorting and segregation.
From the design requirements of the proposed trials, a public recycling centre pre-sorted lamp waste stream would benefit best practice in the handling and storage of this waste stream in:

• increasing the homogeneity of the lamp waste stream thereby avoiding lamp breakages through mixing of lamp types and shapes in the containers;
• reducing operator intervention to remove contamination during material quality verification;
• streamlining the segregation process for mercury bearing and non-mercury bearing lamps at treatment facilities;
• increasing the level of staff training at recycling centres;
• engaging the public in the proper and correct deposition of lamp waste through direction and education;
• encouraging the safe handling and deposition of intact lamps into lamp waste containers;
• placing broken lamps in a designated broken lamp container.

Figure 2 Flow diagram of lamp container collection trial to treatment facility processing

The introduction of public/DCF operative awareness into the importance of disposing of lamp waste correctly at DCFs has benefits not only for the downstream treatment facilities but meets some of the central pillars of European Environment Policy in protecting human health, the wider environment and contributing to resource efficiency.

The waste stream composition data detailed the container type selected, informed on the best implementation for the infograms and container layouts to ensure the correct deployment of the containers in the shelter.

Figure 3 Variance of lamp types.

The container decanting data from MER, REL and NOR/ELK after being assessed by MER was provided as further information on the contamination which could cause issues in the feed system integration as well as the container integration into the feed system.
The shelter and container trials and the collection of the full containers, analysing their contents and weight distribution provided much needed data for the feed system.
The current storage methods of lamps at collection sites has been reviewed. The conclusion is that there are a wide range of variables both in types of lamps and in type of containers. After evaluating the different type of containers identified, the Swedish type has been selected to be the design base for the sorting unit.

MER spent time to adapt the collection system layout and container (figure 4) provision to meet the parameters for efficient interface/system integration, by use of infograms (figure 5), other waste containers and direction signs in the shelters.

Figure 4: Container and Shelter.

Figure 5: Shelter Infograms
A requirement of the Illuminate concept is the minimisation of broken lamps entering the treatment sites. Data obtained from the partner SMEs reported in Illuminate 2.2 ’Report on the best methods for the handling and storage of this waste stream’ section 3.4 – mercury emissions from broken lamps in the containers’ has shown that a substantial number of containers arriving at the treatment contain broken lamps Figure 6.

Figure 6 Percentage of containers arriving with broken lamps
The literature has shown that gas discharge lamp breakages are a source of mercury emissions to the environment and wherever possible lamp breakages should be avoided in uncontrolled conditions (McDonnell and Williams, 2010).
The separation of lamps and tubes helps to remove mechanical stress presented when a layered mix of tubes and round lamps or bulbs are present. Increasing the number of containers that are specific to the lamp type will aid the reduction in breakages due to mechanical stress on the glass envelopes of the lamps.
Awareness amongst the public and operatives in the best methods of handling waste lamps during deposition is also a factor in reducing the lamp breakages. Visual instruction to place lamps in the containers avoiding dropping lamps is a method to reduce lamp breakages in the containers.

Figure 7 Extract of infogram instruction ”don’t drop” lamps but ”place with care”
Figure 7 shows the person not to drop lamps into the container but place them carefully – the result being self-explanatory in reducing lamp breakage.
However, to cater for members of the public depositing broken lamps, separately labelled containers for broken lamps would help to keep the main lamp container free of broken lamp debris. An example of a container used in the UK lamp recycling shed trials is shown in Figure 8.

Figure 8 Broken lamp waste container used in the UK lamp recycling shed trial
The infogram above and affixed to the container clearly instructs the depositor as to its purpose. The swing lid on the container aids mitigation of mercury emission from broken lamp debris.
An assessment of mercury emission concentrations from broken lamps
As part of the task schedule 2.4 an assessment of procedures or design to minimise mercury emissions is described. The following section investigates methods of reducing mercury emissions from containers holding lamp waste. In overall terms the best solution to the prevention of mercury release from lamp waste to the wider environment is the preservation of the glass envelope of the lamp thereby keeping the mercury contained. Environmental education of the public and the operatives as to the negative environmental impact of mercury release is part of the reduction process. This is also emphasised on the infograms instructing the depositor to carefully place the lamp into the containers thereby avoiding lamp breakage. The results of the UK lamp waste recycling trials in terms of lamp breakage and contamination is discussed in the section 4 ’Effectiveness of lamp waste recycling trials’.
However, lamp breakages are inevitable and receptacles to collect broken lamp waste have been installed in the trial lamp waste recycling sheds. To assess the levels of mercury concentration that might be present in enclosed containers holding smashed lamps a laboratory experiment were performed to quantify the levels present.
The experiment seeked to quantify the mercury concentration level attained in a standard industry container when a single new 600mm T8 tube was intentionally smashed inside the container.
The industry container was a standard lamp coffin designed for the collection of linear tubes up to 600mm in length. The overall length was 60 cm x 30 cm x 30 cm giving 54000 cc or 54 litre volume box. The experimental setup is shown in Figure 10. A branded new fluorescent tube referred to as two foot (2ft) was placed in the container. The lid was replaced with two magnets A & B on either side of the lid surface. The magnetic attraction held the larger magnet B (approx 1 kilogram) in position. The mercury monitor was a portable cold vapour atomic absorbtion spectrometer (CVAAS) with a 0-2000 µg/m3 range. The CVAAS had a sampling pump rate of 1 litre per minute and the temperature was ambient. For the first experiment the extraction port was sealed. Magnet A was removed allowing magnet B to drop and smash the fluorescent tube lying in the base of the container.

Figure 9 Experimental setup to assess mercury emissions from a broken tube in an enclosed container
The mercury monitor output was fed into a data logger for the sample period. The experimental setup is shown in Figure 9. The lamp coffin and mercury monitor are shown in the background and foreground respectively.

Figure 10 Experimental setup for assessing mercury vapour concentrations in broken lamp containers
Figure 11 shows the inside of the container with the lid removed following the breakage of the lamp by magnet B. The glass shards and internal fluorescent coating of the two can clearly be seen in the base of the container.

Figure 11 The smashed lamp in the container together with magnet B

Results and Discussion of the smashed lamp experiment
The mercury monitor was set to record the mercury concentration in the container over a five minute period. Figure 12 shows the concentration of mercury measured within the container for the duration of the monitor period.

Figure 12 Mercury concentration versus time period after tube is amshed in container
The mercury vapour concentration shows a steady increase for the first 60 seconds to reach a concentration level of 500 µg/m3 which subsequently falls to a mean level of 400 µg/m3 over the five minute monitoring period. It is clear that the concentration levels are high in this confined volume of 54 litres and would dissipate quickly when the lid is removed. However it is clear that large volumes of broken lamps would be point sources of mercury inside a relatively small volume such as a container. To reduce these levels before opening a container at a treatment facility the use a vacuum port directly connected from the container into the facilities mercury abatement system would reduce the mercury vapour concentration in the container and thereby reduce any emissions to the wider environment. This could be another example for future best practice.

WP3 focussed on the Identification of suitable sensor technology and material separation techniques. To this end CIT proposed a multisensory approach based on detection of material characteristics complemented with geometry measurement to differentiate and separate between mercury and non-mercury lamps. Figure 13 shows the general schematic of the proposed multi-sensor method.

Figure 13 Original schematic of the setup proposed for differentiating mercury from non-mercury lamps.
After confirmation from the EU, that a continuation of ILLUMINATE would be possible with CIT responsible for developing the sensor/separation unit, CIT started with the development of the proposed system at the end of 2014.

CIT has since developed a multi-sensor system that both characterizes the specific material content of the lamps and uses shape recognition. To do this CIT developed deigned and fabricated specialised sensor unit. Necessary software with algorithms specific for data acquisition and data treatment was developed at CIT to enable correct detection of the lamps (see Fig 14). As IP status and strategy questions remains we refrain from more in-depth and detailed description surrounding the developed sensor system.

Fig. 14: Above figure shows the interface of the developed software.

The final task in WP3 concentrates on the design of the mechanical parts of the system for separation which includes the conveyer belt unit and a mechanical separation unit. Different sorting techniques have been considered and reviewed in details. The most robust, fast and practical method was decided to be an electrical motor, controlled by microcontrollers programmed at CIT. Tests showed that as many as 3 lamps per second could be sorted with the developed sorting arm. The control of the sorting arm was through precisely timed communication between the detection software developed and the microcontroller.

Fig. 15, The sorting arm.

To achieve a stable and effective unit able to handle a wide range of containers and processing plants, the solution is based on a turning device feeding the sorting unit. The turning device could be modified for different type of containers, but will in the demonstration be designed based on El-kretsen containers. The output from the sorting unit will be collected in 140L-wheelie-bins, but could be modified to different type of containers.

Based on research out at MRT a conceptual design of the feeding/sorting system has been completed. After procurement, the feeding/sorting prototype was assembled. To verify functionality before installation and integration, the prototype was tested at MRT. Testing was also carried out with CiT (WP3) with WP3 in order to verify the integration of the sorting and pre-sorting units. After initial testing and risk assessment, safety system was updated and integrated in prototype. This prototype was then installed and commissioned at Relight for demonstration.

Fig. 16: Conceptual design of sorting unit.

The basic concept of the design is a turning device followed by conveyor belts with sequentially increased belt speeds in order to separate the lamps from each other. The concept also includes a sieve unit where the lamp volume is separated into two main fractions and a scrap fraction.

Conceptual process description:
1. A lamp container is loaded on to a roller conveyor.
2. The container is pushed in to the turning device and the container is locked in position.
3. An electric lifting chain is used to turn over the container.
4. Lamps fall on to a conveyor belt or vibration feeder (Speed A).
5. A rotating brush will secure the flow of the lamps and avoid over feeding.
6. The lamps fall on to a second feeding conveyor belt with increased speed (Speed B=2xA) in order to initiate the separation of lamps.
7. A vibration sieve separates the lamps in 3 fraction (>200mm, <200mm, <20mm)
8. The main flow falls on to a third, angled conveyor belt or vibration feeder. Lamps will be put in a straight line by gravity. (Speed C = B) 35m/min
9. A small and fast fourth conveyor belt will finalize the lamp separation (Speed D=2xB).
10. Lamps will, one by one, be transported by a conveyor belt on to the sensor table. (Speed E=B).

Fig. 17: Final design of sorting unit.

Final process description:
1. A lamp container is loaded on to a roller conveyor.
2. The container is pushed in to the turning device and the container is locked in position.
3. An electric lifting chain is used to turn over the container.
4. Lamps fall on to a conveyor belt (Speed A).
5. A rotating brush will secure the flow of the lamps and avoid over feeding.
6. The lamps fall on to a second feeding conveyor belt with increased speed (Speed B=2xA) in order to initiate the separation of lamps.
7. A vibration sieve separates the lamps in 3 fraction (>200mm, <200mm, <20mm)
8. The main flow falls on to an angled vibration feeder. Lamps will be put in a straight line by gravity. (Speed C = B) 35m/min
9. Main flow delivered into WP3

On July 11th 2016 the prototype was put into operation and Relight workers started the validation trials to test the performance of the sorting unit. The integrated system was validated against the key criteria, for collection efficiency (breakages), sorting efficiency, processing rate and environmental impact.

Integrated Trials: 29th August - 2nd September
At the beginning Relight tried to feed a small quantity of lamps for the first integrated trials. The speed of the conveyor was very slow so we needed a lot of time to feed the sorting system the first day. Lamps should come ideally in one row on the shaking table before entering the detection process, with a small gap between them. The following day we constantly increase the speed of the conveyors and also the vibration of the sieving system in order to get used to the feeding system. The operators were also informed about the risks and the other operative instructions related to the feeding part.

Integrated Trials: 5th August – 7th September
During this period Relight performed some tests with a higher speed of the conveyors in order to reach the target speed for the sensor.
We saw that if the conveyors are fed with a big amount of lamps, even with an high speed of the conveyors, sometimes some lamps can exit the feeding system from the edge and go under the prototype. This is not a big issue because it can be solved only putting a containment barrier.
We also observed that some compact fluorescent plug-in bulbs with particular base type (PL, Double or Triple Tube as shown in the picture below) are blocked into the sieving system because of the shape of the base.
In that case also other lamps blocked just behind them forming a sort of cup. When all that lamps are released we have a pick of lamps feeding the sorting system and we need to stop the prototype otherwise this big quantity of lamps can damage the sensor unit.
Finally we found some small linear lamps in the broken lamps container however this is not a problem for Relight because at the end this container is sent to the Hg lamps treatment plant.

Maintenance of the sensor: 8th and 9th September
During those two days CIT came in Relight in order to check the functionality of the sensor and to prepare the prototype for the Illuminate Demo Day.

Illuminate Demo Day and Final Meeting: 12th and 13th September
The ILLUMINATE Demo Day and the final general meeting of the project were arranged in Relight site on 12th and 13th September respectively. During those days the partners of the project and also the attendees were able to see the prototype running during the integrated trials. No data about the speed and precision of the sorting were taken in that occasion.

Integrated Trials: 14th – 23rd September
During this period some integrated trials were performed in order to collect data information about the feeding and sorting of mixed lamps directly from the containers coming from the collection point. All the trials were performed in 6 hours per day only weighing the INPUT and OUTPUT of the process. On average 10 containers of lamps were treated every day.
INPUT 15/09/2016 16/09/2016 19/09/2016 20/09/2016 21/09/2016 22/09/2016
Weigth [kg] 830 920 875 890 800 840
Broken lamps [kg] 47 60 66 65 38 42
Big lamps [kg] 19 14 5 10 23 4
Hg fraction 687 742 729 715 670 714
Non-Hg fraction 77 104 75 100 69 80

Because of the high quantity of lamps and the presence of several object different from lamps we only evaluated the weight in kilograms of the different fractions.
The total amount fed in the prototype during the selected period was 5155 kg of mixed lamps with the following final distribution:
• 318 kg broken lamps
• 75 kg big lamps
• 4257 kg Hg lamps container
• 505 kg non-Hg lamps container
We found 17 kg of Hg lamps and other objects in the non-Hg fraction and 270 kg of non-Hg lamps and other object in Hg fraction. So the accuracy in non-Hg fraction is about 97% considering that the errors are not only Hg lamps but also other different objects.

Second Demo Day: 27th September
Another Open Day was organized on the 27th of September. Relight had the great opportunity to host a collective schemes’ meeting so the key players in the sector of lamps recycle were able to see our prototype running that day.

The generated sorting statistics don’t identify every individual lamp, with info about what type of lamp and who was the producer, i.e. Philips, Osram etc. because the quick change in product range make that kind of data obsolete rapidly. Lamps are divided into two fractions for mercury and non-mercury first and foremost.

The original plan was to also register and gather statistics on different manufacturers and brands but because of unforeseen changes in the consortium this will probably not be possible.

Since the project lost an important partner and also much needed means that would convey the technology and skills to obtain data that could be used to identify, for example manufacturers, this task cannot be accomplished.

To get statistics on the manufacturer which could provide valuable information to be processed for business use a technology that can identify brand or in detail more characteristics would be necessary. Right now there is no direct benefits, from El-Kretsen perspective, to identify manufacturers beyond to see who stands for the collected volume and also what type of lamp that has been collected. Looking at the impact of this information on the final result of the project, i.e. to sort out bulbs with mercury has been very important and, above all, it can lead to cost savings if the technology is refined, we cannot identify that the lost potential statistics on manufacturers would generate major progress in relation to the information we otherwise have developed. To obtain the required statistics we could use manually sorted batches and utilize mathematical calculations instead at the cost of direct data and quality, however, we would still get a benchmark of the desired data that is probably good enough.

The benefits are mainly logistical, environmental as well as economic. For sorting this means less mercury handled volumes and less wear and equipment. El-Kretsen will primarily take the information and knowledge to further develop it with the focus on logistics savings. We will work with information from this project in order to increase the take back rate of light sources. We will hopefully in the near future focus more on streamlining and improving the collection of light bulbs with mercury but also LED.

El-Kretsen will participate in further projects that will fine tune and take these ideas longer to develop a more effective collection as well as to simplify the recycling process regarding recycling. The goal is to reduce transport and to reduce the number of units processed in the mercury facility.

• Highlights of most significant results
On July 11th the prototype was put into operation and Relight workers started the validation trials to test the performance of the sorting unit (CIT part).
During the demonstration phase the characteristics of the lamps not well sorted were updated into the database and then re-fed into the Illuminate sorting process.
After the first phase we demonstrated how the sorting process can be integrated in existing recycling processes in Relight with the help of a specific feeding system (MRT part).
Thanks to the collaboration with MRT and CIT, the prototype was optimized. Relight performed other combined tests on the integrated system, feeding system and sorting unit.
Lamp and luminaire technologies are converging with an increasing emergence of luminaires incorporating LED modules. The move towards LED lamps means that a bigger effort is needed to rich an high accuracy in the identification of this kind of lamps.

The main target of the project is to reduce the mercury contamination/spillage from collected waste bulbs, in order to enhance the working environment and environmental performance of the collection-sorting system, but also to prevent the mercury contamination of the non-mercury containing lamps fraction of the waste stream, thus enabling an appropriate identification and separation of both mercury containing and non-mercury containing waste fractions. This is achieved with an improved collection of the waste streams (collection points with new designs and signage, use of new containers) and a better sorting of the waste (automated sorting unit based on a sensor system). This results in a reduction of the breakage of that lamps and in a reduction containers’ trash content, whose effect on the environmental performance of the collection-sorting system has been assessed in previous sections.
At this point, it can be stated that applying Illuminate innovations into the lamp waste collection and sorting processes brings a more environmentally respectful solution, according to the values obtained in four of the five selected impact categories (Figure 5). In the three toxicity categories (Human Toxicity Potential, Terrestric Ecotoxicity and Freshwater Aquatic Ecotoxicity) largely related to mercury emissions, which is, indeed, an important problem in current procedures of spent CFLs treatment (Ramroth 2008), a significant reduction has been achieved. In addition, the impact assessment of these three categories has been identified as the most accurate assessment thanks to the verified data that has been the basis for the calculation of mercury emissions, being therefore, the key driver for comparing the environmental performance of the current scenario with Illuminate application scenario.
The impact category in which an increase is expected is the Abiotic Depletion Potential (131% higher). This impact category measures the depletion of non-renewable and non-biological resources that are not generated by nature within a very long period, such as minerals and fossils, in kg antimony-equivalents (Baumann & Tillman 2004). Thus, as has been repeated throughout this report, that increase is compressible as Illuminate scenario has included some capital equipment (sorting device and shed) that has not been encompassed in the Current scenario, and the manufacturing of those capital equipment constitute the main process within assessment’s boundaries whose target is to manufacture something, and therefore, to consume directly resources (i.e. to deplete abiotic elements).
Besides the toxicity impact categories related mercury emissions, whose improvement has been corroborated and stablished as the key output of this study, Abiotic Depletion could be considered as an important driver too, since this study analyses the initial stages of a recycling process, i.e. processes whose target is the material recovery that avoids direct material extraction from nature. Therefore, this impact category could be a good indicator of the improvements on the material recovery rates that Illuminate sorting will potentially lead into the whole End-of-Life phase of lamps. The whole EoL phase is out of this study’s scope (see Illuminate LCA Methodology 2.3 section), but it has been found reasonable to expand the scope of the assessment in the next paragraphs, in order to put into context the magnitude of Abiotic Depletion increase due to Illuminate capital equipment use in the collection-sorting.
Illuminate technology application, apart from the mentioned trash and lamp breakage reduction, will bring two other improvements that specifically are related to Abiotic Depletion issues: on one side, potentially more lamps are going to be led into the scheme of collection and subsequent recycling, thanks largely to the developments focused on promote correct habits of the collection points users; on the other hand, the post-sorting recycling treatments are going to be more efficient, as the correctly separated waste streams can then be handled by the appropriate processing steps in order to effectively recycle the waste bulbs. This two improvements can be defined in a more simple way as an increase of lamps that are recycled instead of be landfilled, and as a maximization of the recovery rates of the recycling process. As mentioned above, both improvements are out of the scope of this LCA, but shall be considered in order to have a broad understanding of Illuminate application effect on environmental burdens.
In what refers to the first improvement, i.e. increase of lamps that are recycled instead of be landfilled, some straightforward estimations can be done. Simply assuming that the amount of lamps that is not broken in the collection-sorting thanks to the Illuminate technology application is recycled instead of being landfilled, the environmental improvement that this fact supposes can be calculated for the Functional Unit used in this report, by means of application of values got from literature. Assuming that principle, Illuminate Scenario supposes -0.003 kg Sb-Equiv. less than the Current Scenario, for the ADP impact category. To that value, if the expected increase due to Illuminate capital equipment is added (+ 3.27E-06 kg Sb-Equiv.), finally an overall value of -0.00299 kg Sb-Equiv. is achieved. Therefore, it can be stated that the increase of ADP that Illuminate application supposes in the collection-sorting is insignificant, compared with the influence of that application in the following phases of the recycling of lamps.

Figure 18. ADP values per FU, encompassing the environmental burdens of the variation in the rate of lamps that are recycled instead of landfilled thanks to ILLUMINATE technology application.

The second improvement, i.e. enhanced recycling process thanks to the Illuminate application technology in the collection-sorting, is difficult to measure at this point. However, is expected to reduce the processing time largely and increase consequently the capacity of facilities. Thus, more efficient processing steps would be achieved, and energetic requirements, for instance, are likely to be lower than in the current state of art. Nevertheless, if the whole lifetime of a lamp is taken into account, the contribution of the end-of-life phase to the life cycle energy use can be considered negligible (<1%) comparing with the contribution of the use phase of the lamps (Quirk 2009). Focusing only on material recovery itself, purer material fractions such as aluminium and printed circuit boards are expected to be achieved, increasing the amount of elements that are reintroduced in the supply chain. That would result in a lower Abiotic Depletion impact on the recycling process after an Illuminate collection-sorting, compared with the recycling after a conventional collection and sorting, and therefore, potentially that value would be below the one displayed on Figure 19 for the recycling of CFLs and LFLs.

Figure 19. Abiotic Depletion Potential values per kg of lamp. Estimations based on data from literature and GaBi database.

We can conclude, it can be stated that within the collection-sorting of waste lamps, Illuminate technology processes will be a greener solution than the current processes, with special mention to the toxicity decrease related to mercury emissions. In the same way, the application of Illuminate technology during the collection-sorting is expected to bring significant environmental improvements during the post-sorting recycling processes, especially in material recovery terms. If the environmental improvement expectations are extrapolated to the approximately 1,000,000 tonnes of bulbs that have been estimated to appear in the European waste stream over the period 2011-2018, the sustainability and environmental credentials of the Illuminate application scenario are encouraging.

Potential Impact:
The ILLUMINATE project will deliver efficient technologies capable of significantly and measurably reducing the ecological impact of material contamination in turn massively increasing their market potential. Early implementation of the developed sorting technology will increase purity of end fractions and will help European recycling SMEs define and push legislation forward. The technology will additionally deliver substantial improvements in resource efficiency and benefit a large number of SME’s throughout the supply chain.

The main environmental issues from lighting waste are associated with the release of mercury by means of breakages and incorrect recycling methods. ILLUMINATE will reduce these effects by the development of correct handling procedures and the education of those required and by the segregation of non-mercury containing waste to significantly lower costs and environmental impact of this waste stream. A key part of workpackage two is to reduce the level of breakages at collection point by education of both the public and operators. This will reduce the escape of mercury from the broken tubes thereby lowering the environmental burden of this toxic substance. Studies carried out by researchers at UCLan have shown that mercury can be released from gas discharge lamps contained in WEEE in a number of forms (WRAP, 2010). Therefore UCLan is building on this knowledge base to assist the partner SME’s in their quest to improve their education of the players upstream in the recycling heirarchy. The outputs from this education will be quantified in monitoring the quanitities of broken bulbs collected in the waste stream
Illuminate will focus on the ‘separation’ and ‘release by breakage’ to significantly reduce the environmental impacts of lighting waste disposal. The education of EU collectors during the project will leave a legacy in the market which will be improved upon year by year after the project completion.

The project aims at developing a sorting line for handling discarded lighting products. The global production of lamps in Europe introduces close to 1.5 billion units per year. The volume of bulbs appearing in the European waste stream is expected to be approx. 800 million units over the period 2011-2018 . The partners in the ILLUMINATE project will hope to develop technology that can be exploited by the SMEs involved creating opportunities for growth within these companies. This growth will be derived from the manufacture of the high tech systems required to sort and recycle the vastly increasing number of waste lamps, resulting in skilled jobs in the manufacture, operation and maintenance of these systems. The ILLUMINATE proposal has the potential to open new markets and improved competitiveness for the SMEs and other supporting industries based on creating knowledge and innovations arising from this research. The technology will contribute to an improved working environment (lower emissions) as well as to the elimination of manual work in direct contact with mercury/contaminated material. This will greatly affect the working styles and processes used in lamp handling companies across Europe.

Economic impacts of the ILLUMINATE concept
The technology has clear market potential and will have a strong impact on the economic prospects of the participants via the following routes:
• The SME participants will use the technology directly in their own manufacturing operations and/or directly in the services they provide.
• MRT will manufacture and market process equipment and complete process lines based on the technology. The sorting unit will become a USP for MRT and will facilitate sales of an extra 5 process units per year for the first 3-4 years generating a maximum of €2500k.
• CIT will gain detailed knowledge in the design of novel, combined sensor systems, including novel hardware and software. The obtained IP will be examined in detail and if necessary secured by CIT and the consortium. The granting of licenses may provide a source of income to CIT after the project as the results can most likely be applied for the detection of other end-of-life products.

Gate fees are currently out of recyclers control and differ from market to market. What is a clear fact is that all collection schemes strive to reduce the gate-fee as much as possible.

Today we have a gate-fee in North Europe in the range of 300- 400 €/ton and in south Europe 400-600 €/ton. If we could lift out Hg lamps from none-Hg lamps, we could keep or even increase gate fee for Hg lamps and reduce for non-Hg lamps. This is the only way to keep a good recycling for Hg lamps. Today LED lamps is only a few % of the lamp waste, but over time it will be up to 50%. To recycle an LED lamps is completely different from recycling a Hg lamps and the process of handling Hg lamps is much more expensive.

In the EU all recyclers will be affected by the increasing amount of LEDs. The collection schemes are showing a high interest and they will most likely require separate streams (as indicated in the project). Since Hg lamps will be in the material stream for a foreseeable future, the commercial life expectation is at least 20 years. Technically the units should be something similar.

Using the technology in these ways will provide the following economic impacts for the ILLUMINATE participants:
• Mercury will reduce the processing time for a contaminated batch of lighting waste 12 fold (from 3 hours to 15 minutes) and the associated costs will be reduced 3.5 fold (average time it takes to remove 0.25 tonnes of mixed lamps from a container took 195 minutes. It takes 20 minutes to recycle neatly presented lamps through the treatment process, therefore having the lamps presented neatly without packaging or mixed lamp types would reduce down the time by 12 fold). The introduction of these sorting lines will provide similar economic impact for Elk and Nor. This is turn will lead to an increased capacity allowing for the handling of larger loads.
• REL will reduce sorting time of 1/3, and save 50% costs on personal sorting personnel.
• Mercury, Nor and Rel may achieve purer metal fractions such as aluminium and printed circuit boards (sorted material will be run through existing processes during the project).
• Mer and Nor can track trends in the use of lamps, such as phase-out of certain types of lamps.
Technological Impacts:
• New applications for the use of sensor based sorting in a growing market.
• Novel equipment (prototypes and demonstrators) and methods for sorting bulbs and recording data to allow increased manufacturer responsibility and decrease processor downtime.
• Novel developed collection processes to reduce breakages/foreign object incursion significantly reducing environmental impacts and increasing work place health and safety. The technology will be located at RELs facility and trialled during the project. A significant amount of dissemination and exploitation (including open days) will educate and inform other recyclers across Europe.
• Advanced analytical techniques for the assessment of material contamination leading to better standards of end fractions opening new market possibilities.
The participant group will provide the expertise and capabilities to implement the planned S&T innovations:
• Key scientific expertise covering the recycling and handling of waste lighting products will come from the vastly experienced Nordic Recycling, Mercury and El-Kretsen. The inclusion of these 3 end users from 3 EU countries gives a broad background base to develop process applicable to all EU members.
• The state of the art analytical facilities and capabilities provided by CIT, CTECH and MRT will enable advanced analytical methods to be applied to the assessment of contamination and to the development of Standard Methodologies and Codes of Practice
• High level expertise in the key design and construction activities will be provided by MRT, Relight and CIT to implement the innovative sorting system design and integration which is central to the achievement of the Work Programme impacts

List of Websites: