The Chatbot Interview

Currently, Chatbots are on everyone’s lips and most famous is Chat GPT. Chat GPT stands for Generative Pre-trained Transformer. For those of you who are not familiar with the topic, a chatbot is a text-based dialogue system that is able to discuss a certain range of topics or questions with users. These systems are based on machine learning and artificial intelligence. The Chat GPT can interpret pictures and tell you jokes, but it also can give you the pros and cons on a variety of topics.

Reason enough for our technical project coordinator Alexander Leitner  to ask Chat GPT about our project.

Alex:
Why should we apply for Horizon Europe on the topic of refractory recycling?

Chat GPT:
There are several reasons why applying for Horizon Europe funding on the topic of refractory recycling may be beneficial:
1. Environmental benefits: Refractory recycling can reduce the environmental impact of mining and processing raw materials for refractory production.
2. Economic benefits: Recycling refractory materials can reduce costs for industry and potentially create new business opportunities.
3. Resource efficiency: Recycling refractory materials can help conserve natural resources and reduce dependence on imports.
4. Innovation: Developing new technologies and processes for recycling refractory materials can drive innovation and lead to new products and applications.
5. Waste reduction: Refractory recycling can help reduce the amount of waste generated by industry, and decrease the amount of materials sent to landfills.
6. Climate Change: recycling refractory materials can reduce the greenhouse gas emissions associated with the production of new materials, thus helping to mitigate the impact of climate change.

Alex:
Do you think science communication would be important on this topic?

Chat GPT:
Science communication is important for a wide range of topics, including refractory recycling. Here are a few reasons why:
1. Public awareness: Science communication can help raise awareness about the importance of refractory recycling and the benefits it can provide. This can help to generate support for research and development in this area.
2. Stakeholder engagement: Science communication can help to engage stakeholders, including industry, government, and the general public, in the research and development process. This can help to ensure that the research is relevant and responsive to the needs of different groups.
3. Dissemination of results: Science communication can help to disseminate the results of research to a wide range of audiences, including researchers, policymakers, and the general public. This can help to ensure that the research has a positive impact on society.
4. Public understanding of the science: Science communication can help to increase the public understanding of the science behind refractory recycling, and help to address misconceptions or misunderstandings.
5. Increase funding: Effective science communication can increase public awareness and interest in the topic and thus increase the likelihood of securing funding from various sources, including Horizon Europe.
Overall, science communication can play a key role in promoting the development and implementation of refractory recycling technologies and in ensuring that the benefits of this research are widely understood and appreciated.

Alex:
How can I explain refractory recycling to my six year old daughter?

Chat GPT:
Refractory recycling can be a complex topic, but it can be explained in a way that a six-year-old can understand. Here is a possible explanation:
“Imagine that you have a toy that you don’t play with anymore. Instead of throwing it away, we can recycle it. We can take it apart, clean it, and make it into a new toy for someone else to play with.
Refractory recycling is like that, but instead of toys, we’re talking about materials that are used to make things like ovens and fireplaces. These materials are called refractory, and when they’re used up, they can be recycled and turned into new materials. This way, we don’t have to use new materials from the earth to make new ovens and fireplaces, we can use the old materials again. This way we can save the earth’s resources and also reduce the amount of waste we make.

Alex:
Could you suggest a poem about this topic?

Chat GPT:
Sure, here’s a poem about refractory recycling:

Refractory, refractory,
Used to make things hot and fiery,
Ovens, fireplaces, and more,
But when they’re used, they’re not a bore.

We recycle them, we don’t discard,
We take them apart, it’s not that hard,
We clean and melt, we shape and mold,
A new material, as good as gold.

We save resources, we reduce waste,
It’s an environmental feast,
Refractory, refractory,
Used again, it’s not a mystery.

With this new material, we can build,
Ovens and fireplaces, as strong as steel,
Refractory recycling, it’s the way,
To a brighter future, today.

Keep in mind that poetry is a form of creative expression, this is just an example and might not be perfect, but I hope it gives you an idea.

The ReSoURCE Team wishes you a happy April Fools’ Day – and don’t be mistaken, we usually take the topic of artificial intelligence very seriously 😉

As the project leader of the ReSoURCE team, I am thrilled to let you know that we have submitted our 4th deliverable one day ahead of the deadline. This deliverable is critical to our project, as it contains precise material specifications and internal process information that cannot be shared with the public. However, there are a few non-confidential elements that I would like to share with you.

One of the most crucial aspects of this deliverable is our process for collecting samples of feedstock piles that are as high as 200 tons while ensuring that the samples are still representative. Imagine you’re a gardener and you want to test the soil quality in your garden. You have a large garden with different types of soil and plants growing in different areas. To get an accurate picture of the soil quality, you need to take samples from different parts of the garden. You wouldn’t want to just take a sample from one small area and assume that it represents the whole garden.

Similarly, in our project, we needed to take representative samples of the feedstock piles to get an accurate picture of the material properties. Just like a gardener needs to take samples from different parts of the garden, we needed to take samples from different areas of the feedstock piles to ensure they were representative of the entire batch.

After collecting the samples, we had to determine who would need which type of samples and how much at which time. Let me use another parable here: It’s like organizing a dinner party – you want to make sure you have enough food for everyone, but you don’t want to have too much left over. In the end, we ended up with more than 60 individual sample sets with a total sample weight of 9500kg, which is quite different from what you would expect in a typical lab study.

But that’s not all – we also had to review the currently existing sorting classes of the spent refractories, which we use for manually sorted material. This review made it clear that the classification is not directly based on the chemistry but on secondary properties, such as color or certain material patterns that are visually recognizable. This means that there can be significant overlap between the chemical specifications of different sorting classes, and we had to be aware of that as we moved forward with our project.

Now, let’s talk about the benefits of our automated sensor-based sorting compared to manual sorting. We used a theoretical case example for a specific steel casting ladle to demonstrate this. In one scenario, we used the currently existing approach for manual sorting, and in the other scenario, we implemented the automated sensor-based sorting. The results were striking – in the manual sorting scenario, 94% of the total breakout material was lost for reuse in refractory products and ended up being either a metallurgical additive or landfill. In contrast, in the automated sorting scenario, the proportion of the presently “unsortable” material was significantly reduced to 20% of the total breakout material.

So, what’s next for the ReSoURCE team? We are eager to continue our work and push the boundaries of what’s possible. Stay tuned for more updates on our exciting journey towards a more sustainable future.

Since we already sampled 3 suitable feedstock materials with the same sampling procedure, the question arises how we could achieve similar results regarding accuracy and validity while decreasing the effort of sampling. The current, rather complex sample procedure involves a wheel-loader, a spare area to spread out the material, time, and a few slightly crazy people which are willing to get dirty, are quickly available, ready to get out regardless of the weather condition and are ready to move around a few tonnes of material. As one can think of, this type of sample procedure is not suitable, let alone difficult to implement when sampling new and possibly appropriate recycle material.

For this reason, we tried something different while sampling the latest materials of a cement rotary kiln additionally to the above described complex way of sampling. So, we chipped off 20 edges of 20 randomly selected spent refractories bricks. These 20 edges will be crushed and mixed together to create a combined sample which we will analyse the same way we analyse the material we sampled with the more complex way.

Afterwards we will compare the two results. If we assume that the representative, complexly taken sample represents 100% accuracy and we compare the results of the combined sample of those 20 edges, we can estimate the accuracy and the applicability of this attempt.

As the year goes by and the season changes, it is about time for cement producers to break out their old, worn-out kilns and renew their refractories to prepare themselves for a new and more productive season. Since breakout material from cement rotary kilns (CRK) is considered as valuable recycling material, recyclers all over the world swarm out to get their share. Luckily our team members at RHI Magnesita are always on the lookout and again chose suitable feedstock material for our project. As the 15th day of February dawned, the third representative sampling campaign began. The weather was not in our favour at the beginning, but it finally cleared up, and the sun showed itself benevolent to our project.

In contrast to the last CRK refractories, the encountered material was the same size, but many of the stones were still cemented together and much more fines (grain size < 5 mm) could be found. The sampling procedure was exactly the same as last time. The reason for this decision was the greatest possible comparability with the other CRK material. Since sampling significantly impacts analytical results, comparability can be ensured if the sampling procedure remains the same.

So again, 8 wheel-loader buckets of 4 tonnes each were dumped on a separate area (this time, it was the next roofed outdoor storage compartment) and subdivided into 4 subsets. The last two sample piles were located in the remaining pile. A total of 40 samples á 25-30 kg were taken, summing up to approximately 1.2 t of sample material.

The increased amount of fines also benefited our sampling because after tipping the sample material, all the stones were covered with fines and could no longer be distinguished from each other, which is very conducive to random sampling.
The next step will again involve hand sorting of chosen sample bags and analytical investigations like X-ray fluorescence spectrometer to identify the chemical composition or X-ray diffraction analysis to display the mineral content and, of course, the comparison with the results of the firstly sampled cement rotary kiln.

Today, we finally can present the construction of our sorting unit. In the picture you can see the design of the plant.

Container 1, the left part in the illustration, houses the material feed and the supply unit. The supply unit includes components such as the dust extraction system and the compressor components. The so-called feeding container also has an opening in the roof through which the material can be fed into the bunker by a wheel loader. To protect the opening against environmental influences such as rain, a folding roof is provided.

The material, lying in the bunker, is singularized and transported further by an adjustable vibratory conveyor. Singularization is an essential task in this context because subsequent measurement is only possible when particles do not overlap or touch each other. After this is done, the material is conveyed to the second container, the sorting container, via a conveyor belt.

In the second container, the material is moved over a large conveyor belt. In the left part of the container are the sensors which are needed for the geometric measurement on the one hand and for the analysis of the particles on the other hand. For the geometric measurement a 3D camera scans the bricks and send the position information to a LIBS sensor, that focuses on the surface of the bricks and measures there the material composition.

As soon as the brick fragments have been measured, an AI classifies them into categories. After classification, they are discharged by robots into bins or blown out by compressed air by a blow-out unit. The robots eject larger particles and the air blast unit ejects the smaller ones.

The fractions can be discharged individually via a system of conveyor belts and collected in large bins. These bins can then be emptied using a forklift or front wheel loader.

We know, it is still a while, but we are already looking forward to seeing it in action.

In order to get a better understanding of the material and its condition, we are visiting our customer industries in the course of WP1 T1.3. We want to gain experience of how the demolition process works, i.e. the manual activity, the storage and also the waste management, in order to find recycling opportunities here as well. One of RHI Magnesita’s customer industries is the lime industry.

Alexander Leitner, Thomas Zoescher, Sandra Koenigshofer and I had the rare opportunity to visit an annular shaft kiln during repair time and evaluate possible recycling opportunities. The plant manager was so kind to give us a detailed explanation of the kiln and the processes. He also informed us about the different product types in the lining, before we went to the aggregate.

This kiln is only shut down every 4 to 5 years and checked for defective spots. If they find a damage, only these areas are replaced or repaired. On site, we went 34 m up to one of the combustion chambers. That was quite impressive. We could join the lining evaluation and also the manual activities during the breakout and repairing the damaged areas.

We must admit that we were very pleased to see how good the condition of our products, which were installed in this kiln, was after 5 years of continuous operation. Especially, when you consider that they have to withstand temperatures of around 1,300 °C on a daily basis.

Back on the ground, we then had the opportunity to sample the excavated material and assess the condition and texture of the material. Since the excavated material was not pre-sorted, we sampled what we considered to be interesting pieces from different furnace areas.

This valuable experience will certainly benefit us in the further course of the project.

When it comes to our sorting unit, it almost makes one feel sentimental to look back at the beginning. It all started in 2020 with a LinkedIn post from RHI Magnesita asking the community to analyze fired carbon bricks. So, you could say, an open innovation regarding scientific challenges was where it all began. However, LSA accepted this challenge. It took us a while, but after some initial analyses in LSA’s laboratories in Aachen, Germany, we quickly were able to run a pre-test study with RHI employees on site in Germany.

The results were very promising and further cooperation was set. The goal was clear: the development and construction of an automated sorting plant that can be used on a mobile basis. First, rough requirements were defined. The mobile application is a central point as RHI Magnesita also wants to set up the plant at customers` sites in the future to sort on site.

Also, the technical requirements are important. Next to the general throughput and the grain size, the number of different fractions were defined. With these main requirements the development process started. LSA created a first design with the CAD-Software Autodesk Inventor.

To fulfill the basic requirement of mobility, a 40” container was chosen for the base. On the one hand it offers a lot of space and design freedom and on the other hand it is easy to transport. To put some life into the container, LSA created dummies of the necessary components and placed them inside the container. With the first drawing, LSA had a larger meeting with RHI Magnesita to discuss the first draft drawing. All in all, the draft drawing was accepted. Nevertheless, new requirements for the installation site in Austria were identified, that are at some details different to Germany. The main result of the meeting was: We need more space. Therefore, we added a second 40” Container to the design.

The feeding for the plant were set into container 1 and also the parts for the basic supply like compressed air and the dust filter system. The heart of the plant, the actual analyzing and sorting were housed in container 2. Parallel with the design, LSA started to find suppliers for the core components and prepared pre-tests. For example, several buckets full of different bricks were send to a company that builds customized blow-out units. The company tested, if their blow-out units are suitable for bricks and which grain size they can sort.

Further tests have been completed at various points in the project or are still pending. If a supplier company could fulfil the requirements and was convincing, a highly detailed CAD model was created. This detailed model then replaced the dummy in the construction. In this way all dummies were replaced by real representing CAD models and the whole construction got more and more detailed. The development process of such a system naturally requires several iteration loops. Therefore, in bi-weekly meetings with RHI Magnesita the design was discussed and further refined until the final construction. We will present the result of our efforts in one of the next postings.

The ReSoURCE Project makes an important contribution to the reduction of CO2 and considering the size of the consortium with 9 partners and a budget of 8.5 million Euros, it is by no means a small project. Independently of the funding project, RHI Magnesita ordered an automated sorting plant, which is essential for the project ReSoURCE. As this plant had to be designed highly adapted to the requirements and needs of RHI Magnesita, the customization of the overall plant claims specialized experience in measurement technology, interfacing, and automation.

As you can imagine, handling such an order is not a simple process. It all started with our biweekly meetings. In these biweekly meetings, the actual status regarding engineering and design was presented by LSA. A preselected team of RHI Magnesita experts was involved to optimize the process of development. Combining RHI Magnesita’s knowledge and LSA’s process experience, we were able to develop a compelling design that met all requirements and will enable ReSoURCE to be a great and successful project.

In addition to the design itself, we wrote a technical documentation showing the current state of all components and functionalities. It was deliberately intended to be an open process involving experts from different areas of RHI Magnesita. We wanted to make sure, that they were able to see the design of the system at an early stage in order to comment or identify potential challenges as early as possible. As a result, more and more detailed designs were presented during the biweekly meetings and each iteration took the design closer to the final version.

As you might have learned from our first post A Legal Issue, also contractual matters must be clarified in such processes. In addition to general contract conditions, such as delivery times and delivery terms, acceptance criteria for the entire plant must also be defined (performance values). Performance values, which are always defined in advance of such projects, are measurable acceptance criteria. Such criteria can be, for example, the throughput in tons per hour or the sorting accuracy in % that the system should achieve under defined conditions.

Once the concept was agreed and accepted, the purchase price was negotiated, and the order was placed.

Both, the development of machines in the field of laser-based analytics determines the daily business of LSA GmbH, as well as the implementation of international research projects with various project partners. In this case, the system and the project are very special in more ways than one. On the one hand, it is a very labor and budget intensive project, which is of course very exciting and interesting, but at the same time requires a lot of discipline and strict management. On the other hand, we want to make a green contribution to the industry with our systems, so this is not only very interesting for us in terms of scaling and application, but also a personal concern.

For the Laser Induced Breakdown Spectroscopy (LIBS) measurement to be performed for the analysis of refractory materials, a laser source is needed. Readers who are not familiar with the field of lasers may ask themselves a very simple question while reading these lines: “What is a laser?”. Therefore, a brief introduction makes sense:

Simply put, a laser is a monochromatic, directional, and coherent source of light. As a comparison let’s take a look at an ordinary household light bulb: it shines a white light in every direction. What we call “white light” is actually a spectrum composed of a lot of different wavelengths, roughly from 400nm to 700nm, corresponding to the visible spectrum our eye can see. Our laser light will have one very specific infrared wavelength, in our case, 1064nm emitted by the Nd:YAG laser emitter. Since we only have one wavelength, all our laser light is in phase: imagine water waves of the same height going in the same direction in a narrow canal with the same period. A light bulb, on the other hand, can be described as chaotic waves in a storm going in every direction with varying heights and periods. This ordered detail makes the laser source coherent and directional. Lasers are very common in our everyday lives. We find them at the supermarket cash register where they scan the bar codes of products. They are in our remote control when we watch TV and many more. Try and count how many times you encounter the usage of (very weak) lasers in a typical day!

Depending on the desired specifications, there are different types of lasers and different components used within. In the ReSoURCE project, it is a “diode pumped solid state” laser where diodes are used as an excitation medium. These are powered by a current and excite the Nd:YAG laser rod which then emits the first laser light. The light bounces multiple times between two mirrors on each side of the lasing material, by passing the laser rod, again and again, stimulated emission occurs and amplifies the light in the so-called “optical cavity”. To access the amplified light, one of the mirrors is actually partially transparent. Therefore, a portion of the laser light exits the cavity and can be used as desired.

Important to mention, is also the difference between a continuous wave laser and pulsed laser. In the case of a continuous wave laser, the laser light is steady, and output is present at any time. For a pulsed laser, which is used in the ReSoURCE project, the output is not continuous but rather composed of pulses having a time duration of a few nanoseconds. By having pulses rather than a continuous wave, very high peak power is achievable.

Of course, the laser system for ReSoURCE is being built with strict specifications in mind to ensure the flawless functioning needed for the LIBS measurements.

More specifically, two different laser systems will be developed for the ReSoURCE project, each having its own purpose:
The first laser system will operate at a repetition rate of 400Hz (that is 400 pulses every second!) with the ability to monitor any energy fluctuations which may be caused, for example, by the accumulation of dust in the sorting system over time. The high repetition rate allows to analyse a large number of materials on the conveyor belt, for example, so-called “fines”.

The second laser system will have the particularity of being able to send two pulses within a short time span. The two pulses have a very specific function, the first one ablates any surface impurities present on the refractory material and the second generates a plasma which will be analysed to determine the composition of the brick. Furthermore, an additional function will ensure it only sends pulses when desired. For example, if no refractory materials are present on the conveyor belt at a certain time, no pulses will be sent by the laser system.

More about the LIBS technology can be read in the previous blog post Let Them Shine in the Light of the Laser from our partners at Fraunhofer ILT.

CPI, the Centre for Process Innovation, is a leading independent technology innovation centre and a founding member of the UK Government’s High-Value Manufacturing Catapult. Bringing together academia, businesses, government and investors to translate bright ideas and research into commercial products, is our main focus. In the project ReSoURCE, we collaborate with RHI Magnesita and other partners within the consortium to generate a model of the complete Secondary Raw Materials (SRM) refractory recycling process. Our goal is to determine the economic viability and environmental impact of the SRM recycling process, this is called a Life Cycle Assessment and Techno-economic Analysis.

A life cycle assessment is an environmental impact assessment that considers the entire life cycle of a product or service from raw material extraction all the way through to its end of life. This often means companies must look both upstream and downstream of their own processes. As an environmental study, there are multiple impacts that ought to be considered regarding refractory brick recycling, including climate change (or carbon footprint), resource use, land use, eutrophication, ecotoxicity and many more. Such analysis allows the hot spotting of stages on a product or process’ life cycle for high environmental impact and comparison between different processes to ensure new methods are put in place sustainably.

To understand the economic impact of the new SRM process, we – as you might expect from us Brits – use TEA. In this case though, TEA is an abbreviation for Techno Economic analysis. This is a review of a manufacturing process looking at the costs involved including raw materials, equipment purchases and overall operational costs such as energy requirement and staffing levels. It allows us to identify the parts of the process that contribute the most to the overall costs or inefficiencies of running the recycling process and helps these areas to be focused on for improvement.

As other blog posts have pointed out, there are a vast array of refractory materials used each with different compositions, impurities and rates of degradation or change based on their use. This means that when these materials are considered for recycling the required processing steps will not always be the same. Initial composition, type and level of impurity as well as material condition when they arrive at the recycling plant all contribute to these processing requirements. Initial analysis of the material is used to identify what must be done in order to turn the spent materials into useful secondary raw materials.

We are currently working closely with other consortium partners, particularly RHI Magnesita, and researching existing literature to understand the range and composition of spent refractory feed stocks and the differing processing steps that each will require. This will allow a simulation to be generated that maps the existing recycling process to give a base-line which future project work can be compared against.

Author’s Portrait

James Condliffe

Mr James Condliffe graduated from the University of Nottingham with a Masters in Chemical Engineering. He is a Research Engineer at CPI and part of the project ReSoURCE WP2 team to lead the Life-cycle Assessment (LCA) delivery.

Author’s Portrait

David Rae

Mr David Rae is a Senior Process Engineer at CPI, leads and oversees the technical delivery for the project ReSoURCE WP2 and focuses on the Techno-economic Analysis delivery.

Author’s Portrait

Baijia Tang

Dr. Baijia Tang graduated from the University of Hull with a PhD in chemistry. He is a Project Manager at CPI, oversees the operational delivery and coordinates the part of project ReSoURCE assigned to CPI with RHI Magnesita and other partners.