Cemtech Live Webinar: Quality Control December 2025

Hello and welcome to this webinar. my name's Thomas Armstrong, managing editor of International Cement Review. and it's a pleasure to welcome you to this edition of Cemtech Live the cement industry's most viewed webinar. and we are here every month and it's amazing to, to be here now in December at the end of, of a year that's gone past very quickly for us all. these webinars are designed to, to place a spotlight on the latest technological advances, as well as to promote best available technologies, helping producers like you take practical steps towards manufacturing excellence and sustainability.

So a little introduction for those of you who haven't been here before, but these events are organized by International Cement Review. The cement industry's leading monthly publication for 35 years. each issue offers deep insights into manufacturing technologies and their application across the entire cement production process. So if you are in the cement sector attending Cemtech webinars, you should be reading ICR on a monthly basis, and that's easy to do. Just log on to simnet.com, subscribe and enjoy instant access to the latest issue, as well as an archive stretching back many years.

and don't forget, you'll also receive a copy of our Cement Plant Operations Handbook, for those of you who need to know about the markets, the location of cement, plants, volumes market intelligence that's all delivered to you very conveniently in our global cement market outlook and online platform released earlier this year. It compliments the global cement reports something that many of you're familiar with. do go back onto simnet.com if this is something that can help you in your work. When we're not doing webinars we are doing live conferences, and our next one will be in Riyadh, Saudi Arabia in February.

it's our first time to Saudi Arabia, and I have no doubt that many of you have been following its progress over the last few years. but it is poised to become the largest construction market in the world, and it's driven by the country's Vision 2030 development program. and as well as gearing up for rising cement demand. the industry is modernizing, it's transforming itself through the adoption of new technologies, low carbon technologies like the rest of the world. and it's doing this. right now we are offering the opportunity to go and interact with the industry in Saudi Arabia and the whole middle East and Africa cement industry. So that's up first of 4th of February.

Again, go onto CNET and you'll find all the details there. So that brings us to the matter in hand quality control for cement plants. And we have three fantastic presentations. really delighted to be able to introduce these speakers, these companies that are all very well known leaders in the sector in different areas. we're gonna learn how intelligent automated materials testing solutions can help manufacturers like yourselves achieve accurate compliant results and enhanced quality control.

We'll hear about leveraging XRD and REIT failed analysis to achieve greater quality and process control in cement production, and we'll learn how AI has been used to optimize cement strength as well as addressing other daily plant challenges. we'll have a live demonstration here of some very interesting software AI driven solution in a moment. but to start I'm gonna free up the screen and ask our first speaker Shiam from Tony, NIC in, in Germany based in, in Berlin. And he's gonna talk to us about some technology that actually much every lab in the, in the cement industry will, will know of, if not own some equipment produced by Tony Tech Nick. and a little introduction for Shiam.

he's a seasoned business professional with deep experience in international sales business development who holds an MBA in international business from HWG Ludwig Chafin and brings hand hands-on quality control experience from the ready mix concrete industry in the Middle East. He's also a certified concrete technologist with the N-R-M-C-A and holds a bachelor's degree in civil engineering from Hamley University in India. Sheam began his career in Germany as a product management intern at BASF, now Master Builders Solutions where he focused on admixture systems for Europe Today.

He leads business development and channel partner management across Middle East, Africa, India for Tony Tnic, which is part of the a swic role group. a brand that is widely recognized for its innovative building materials, testing systems, so bringing together strong technical foundations and strategic co commercial insight. please sham, take over and give us your presentation. Thanks, Thomas, for this introduction. Am I audible? You everyone? Yep. Yes, we can hear. That's great. And we can see, okay, so you are all set. Please, please go ahead. One second. Okay. thanks, Pam. The right direction.

the topic that we have today is ensuring cement quality, and thanks for thinking about Tony JK when we thought about cement quality. And we are having the topic about the intelligent and digitalized testing solutions from Tony Nik, and this time having a more focus on traceability part of physical testing of cement myself, Shamali, thank you for the introduction. Joining with me my colleague, Clements Hoberg, who will be supporting me for the q and e from the customers. Nice. Yeah, sorry. So the agenda for the meeting is basically about four different stages of the lifecycle of a cement sample that is from the sample preparation that is mixing and compaction and the storage of samples.

And finally, the strength testing that is both comprehensive strength as well as the flexural strength. The focus will be on the full life lifecycle traceability, which is used by the customers, which can be used by the customers for back tracing and verification. We are also adding two more extra products into this webinar, which is our series. That is our 20 set series for t setting time measurement, and our T perm series for blame finance measurement, which enables smart testing with the highest traceability and digital data export.

The first step of sample preparation is mixing, and we introduce you to Mix standard, which is a complete fully automatic micro control mixer, which is completely automatic and having all the mixing programs as per the standards ENAT completely programed the mixer will release the user from complete user supervision. Comes with a digital display that shows the mixing speed as well as the status of mixing the mixer comes with a option either to have a automatic sand dosing and water dosing, but the water dosing is kept optional. The customers can decide if they want to go for complete automation or half automation. What we want to focus here is about the data export possibility of a mixer.

And the mixer comes with a process monitoring. That means the mixer will analyze himself if he's able to do it properly, if he's correctly mixing the motor mix with respect to the so-called standards. So if there is any non-compliance with the process, the mixer will itself identify it and will let the user know. So this shows the traceability part of our motor mixer. The mixer will also come with a future condition monitoring, which is planned for the end 2026, which will be a software upgrade for the customers who are already having our CEB central electronic board mixers. The new mixers will be able to upgrade it as an extra package later. What you can see on the right is our protocol.

So for the mixing, this is not for this particular mixer. What you see, see on the right on the left, and this particular protocol is for a much advanced mixer, which shows the correct time of mixing the amount of water, the timing of the wait timings in between and when exactly there is the ingredients added and how for how much time it is being mixed along. Coming to an extended portfolio, we have Toni Mix expert, the mixer with some advanced options for r and t customizable, and it is having a self programmable mixing sequence. So apart from the standards, you can define what you want the mixer to do.

Basically, it is programmable from different speeds as well as different variable water dosages. We have one extra item added to the expert mixer, which is called Visco expert, where the mixer is enabled with the top measurement. That is a tox sensor for viscosity measurement. The mixer can generate a graph that is new to meter versus time, and it comes with an advanced process monitoring. So here with the tox sensor, the mixer can even identify if it was properly cleaned after the first cycle. This can be completely identified and protocol by the mixer. We have one more, more advanced mixer, which was manufactured, which was designed for a particular customer called Tomic Expo.

Tomic Pro, sorry, with the integrated temperature measurement of the mix, which is used to simulate the real concrete transport conditions. The graphs that you see is about the Newton meter per time graph. So you can see both the graphs of water as well as viscosity in a single graph. This can be exported and stored, which can be back trace. So when there is a problem with the, with the final results, this can be completely back trace and it could be used to understand where, what went wrong. Coming to the second part of simul reparation is compaction, and we are very proud to introduce our closed loop control vi table called Tony Whip.

And we are very proud to say this, the, we are the only manufacturer having a close to control vibrating table where the amplitude variations caused by any weight fluctuations is completely held constant. The amplitude will be always compensated and held constant by the intelligence of the, of the combating table. It comes with a digital display, again, very rare in the market, and the whole filling process is supported by a time bar. The mixer is adaptable for molds from other manufacturers. It is very silent and it is portable as well as plug and play. So you don't need to have a foundation like for a jotting table to operate a in your laboratory.

The results are completely reproducible and independent of the operator, and the data can be completely exported and documented. So that shows the traceability of the compaction process from mixer and compaction. You're having a proper data about your mixing process as well as compaction process. This vibrating table also comes with a process monitoring, which himself, he will analyze if it was properly done and the report is being generated. The last one is about the condition monitoring.

As I explained in for the mixer, it comes it is currently not there, but we are planning to have it for our violating tables by end of 2026, which will be provided to the customers, existing customers as an upgrade. This is what I meant about the protocol generation. So you can see that the amplitude should be 0.75 plus or minus 0.05, and if there is any kind of fluctuation, if there is anything that is not under this tolerance limit, there will be a orange signal indication from the vibrating table saying that there is something wrong. You have to, you have to rectify it.

This is something that is very new as well as we are the only only manufacturer having this, this, this particular traceability for the sample preparation part, coming to the sample storage part, which is not really digitalized. Basically it's a data logger used by the customers for its purpose. We have got mo curing tanks where the humidity is maintained by any evaporating by evaporating water from the bottom. There is a fan for the equal distribution of and humidity. And there are vents for the, for the sample, for the temperature, you know, because this is an exo process, the, the curing process.

And the particular model is not really ambient temperature controlled, sorry, it's not temperature controlled. It works on ambient temperature and has got a capacity of 36 samples. So basically 12 Ian moles with the moles. Coming to the next part is the queing part, where we have two different models. One is the model that you see on the screen having a capacity of 216 Ian Prisms and 3 36 A STM cubes. This particular one comes with sample containers, and it is completely temperature controlled with the moat. The last model that we have is a three story or three level water storage with the higher capacity.

If it is kept vertically, the Ian samples, you can, you can store up to four 50 samples and six 30 samples that it is kept in a vertical vein. It's completely temperature controlled with three levels, having three separate thermostats coming to the strength testing part. The flexural and comprehensive strength, as you all know, it's one of the most critical parameter for, for a manufacturer of cement. There should be a reliable strength testing machine to minimize the variance and to reduce the human error. We are introducing our two different machines, which is normally used offered to the cement industry. That is Tony Zim, which is a single purpose machine, and we have to pra on the right.

That is a combined compression or and strength testing machine. We call it also very futuristic because to p prax comes with an extra capability of adapting and extra load frame. In the future for concrete testing, up to 3000 ton, our machines completely comes with fully automatic processor control machines, which is completely civil, hydraulic deformation displacement and load controlled. The machine is having a digital controller. Our on developed to control digital controller with a data export possibility with our software called Tunnel Link. We assure accuracy, repeatability, reproducibility, and traceability. That is a RRT from our mother company's.

We crawl the machines come with, it comes with an oil filled ball seat with strain cylinder principle, which is something that is a mandatory thing for European standards for concrete testing. But we have adapted, and I would say we are the only manufacturer with an oil filled ball seat with strain cylinder principle adapted for motor testing, motor compression testing. And what is it? It's basically, it ensures that the consistent force is being transmitted over the entire area of the specimen. You can see a picture that was kind of a test that was done by one of our customer comparing us with one of our competitor. I'm not naming the competitor.

And you can see the equal distribution of the force with istic machine in comparison to the middle one that is from our competitor. It ensures reduced variance and friction free equal distribution of load. I repeat the second machine. The combined compression and flexion machine is convertible to a 10 kilo Newton soft sample compression, as well as there is a possibility of having an extra load frame for concrete in the future, which you can adapt later.

Coming to the traceability part, as you can see, we come with our software that is from our mother company called Test Expert three, that ensures full control of the machine with complete data traceability and export test Expert three is a state of art software, which is so easy to use. What you see on the left is just four different items. That is basically the setup thing of the system. Configuring, run and view cells starts. Test expert. Three comes with high level of user management and control, where you can define when, who, what, why, and responsible for each and every test.

For example, on the right, what you see is a view for a tester for end user who is basically running the test and doing the results. Being a laboratory manager, you can have more rights and that is what you see on the list. Again, some more details about our data traceability. you can configure what exactly you want to have in your result page, and on the right is a result page that you see that is being generated from the test expert software Coming to the next measurement, which is the setting time measurement. And as you all know, it's a test that influences the complete workability.

That is the initial setting time, and for the final setting time, it completely influences the speed of construction. Introducing the Tiny Set Series. So as you all know, that manual method is having a high dependency on skilled labor, also disciplined technicians who is ready to come and do the measurement in five to 10 seconds. And in 10 to 15 minutes, sorry, test must be performed when it is manual during the working hours and working days, and there is a huge chance of inaccuracy, inconsistency. And whenever the person is writing it on a paper and transferring to a Excel sheet, there is a huge chance of errors here and the results are not traceable.

Our tiny set series can work with one laboratory technician with minimum dependency. There's a time flexibility with our measurement, and we are using our incremental measuring system. That is the digitalized testing with 0.05 millimeter accuracy, highly consistent, repeatable and reliable. And with our own developed software called wdb, it comes with data storage and traceability. Four different equipments that we have in our portfolio, which is categorized depending on the number of test stations. We have it from 2, 6, 8, and 12, depending on the flexibility with the standards, which I'll be explaining the next slide, and here comes the differences.

The Tony said to an expert, that is two station and 12 station comes with the highest flexibility. It can test Ian, a STM, sim and gypsum and motor if you buy some extra accessories. But the Tony said compact, that is a six station and classic. The eight station comes with a bit of limitations. They cannot test motor and Tony said, classic is only meant for Ian. There are also certain other differences in terms of the footprint stand around capability and requirement of a thermostat coming to the traceability part of 26 series. That is the graph that you see on the right, which is being generated. That is the depth versus time graph.

The storage and reproducibility of whistle can be done as tables, as well as graphical formats, which can be exported with USP as well as limbs. Some other features, there is automatic cleaning and dry mechanism with our set devices, and it makes sure that the, is a precise positioning from the ring as well as from each and every penetrations operation is without any needle change for EN for initial, as well as final test because our mission is having our occupant is having a 0.05 millimeter accuracy. You don't need to have a separate extra needle for the, for the final testing of cement. An interesting point about our equipment. It comes with two different options.

Again, this is something that is unique in the market, but we have two different options for our Tony set. One is a compact option and export option. So the compact option completely is according to the standard. So as per Ian, it maintains almost 24 penetrations. It makes sure that it is eight millimeter from the sides as well as five millimeter from each other. But if you don't know anything about the cement, and if you want to have a initial test just to understand the range where the setting time lies, you can go for the expert mode where it makes around 2 54 penetrations into a single sample, and it'll give you a very good idea about where the setting time range lies.

The last equipment is about a blade fineness. That is a air deliverability tester. Again, one of the most important parameter added as it is directly proportional to the strength of the cement, which has to be measured with highest accuracy and least human errors. The method that we use, like every other one is the Air ty method, where a certain quantity of air is passed through a compacted cement bed and the time taken by the AR is being calculated automatically. Introducing tny PERM standard, which is with the name says it's standard compliant, fully automatic measurement, which is used for both quality control and production monitoring. Tny perm comes with a range of 2000 until 30,000.

So it is used in different industries like pharma, metal, cement, food, wherever. There's a powder that is 20 per here. it's single station measurement and I count the standard. It's around 1.9 centimeter cube volume to per comes with the integrated temperature sensor, and it comes with the integrated database which shows different densities and porosities the machine. The equipment can calculate the sample weight himself if you provide the density and porosity. So that's automatic measurement and it can store different K factors, calibration factors with the help of your calibration sand.

There is definitely here the same thing what I explained, a integrated result database, and it can be exported via USP and lis. This equipment has got the extra capability to be connected to our software called Test Expert because of data traceability and FDA CFR 21 requirements by pharma. So it it has the highest user control and user management capabilities on the right. What you see is about the the graph, sorry, the, the display that you have, which shows average time and the temperature and all required data. One extra equipment that we have as a excellent portfolio is to perm two. That is system which is used for fast, faster, what do you call it, like practice oriented alternative.

This equipment the size of a bigger measuring cell is currently accepted, or let me call it tolerated by A STM. And it comes with an extra dial gauge that compensates for the deviation and the compaction process here, the volume is a bit different. It's, it's almost a hundred gram material you can put in depending on the density, definitely. the volume is something around 74 centimeter cube. It comes with automatic measurement, and the number of test stations are two. You cannot operate them simultaneously, but you can get it prepared and you can do the measurement one after the other.

Coming to the traceability part, again here in the results, you can, you can define what data you want to have in the final report. And on below you see the tunnel link software, CEB, that is used for the data. Export results can be exported as Excel file CSV raw data as well as a PDF protocol, which can be easily stored and it's traceable by the customers. Thank you so much for listening and thank you again everyone for Thank you also Global Cement simnet for considering us in this webinar that is about quality control. Thank you so much. If you have any questions, you can contact us directly@inferto.com or else you can scan me directly. you can reach out to me with this QR code.

Thank you so much. Thank you very, very much. sham wonderful overview of an amazing array of equipment there. and if I can bring in Clements now. Nice to have you with us. area sales manager for the home market, and also someone with a, a deep understanding of the company's products, having worked there for many years. So just looking through the chat, it's great to see there are a a number of companies using the equipment using your, your different products in their, in their labs. So it's good to have them on here with us. a couple of questions I'll try and I'll try and bring you one or two now.

it's one that we often have when talking about lab equipment is how often do we need to calibrate these these equipment? what are the signs of deviations from real values? are, can you tell us about your service after sales service and if you have that online after sales service? So maybe that's a, a good question for, for each of you. the, how often you have to calibrate the machine, it depends from the area where you stay with your equipment. Normally in central Europe, you have to calibrate the machines every two years Yeah. with ducks or some other organizations. And we always said that bigger service on our server hydraulic controlled machines must be roundabout after 60 years.

It means change oil and check some parts between it's nice to check the machine after a checklist. We delivered also together with the machine. Therefore, you can do many things yourself. In central Europe, it's used normally that we make a service and calibration every two years, but in more other areas, it could be also be useful to make the service a little bit more less. It means after six years, it's it's, it's a good tact for, for our machines. Very good, thank you. And and something about the after service, and there are a couple of questions actually. do you service in East Africa as one, do you service in India? tell us tell us about your, your, your service.

We do have for example, okay, it, it, it is quite different with different countries. You know, for example, in Africa, we do have our direct partners in South Africa, west Africa, and North Africa. And our South African partners also taking care of a couple of customers in East Africa. So we have local partners who, who are responsible for our after sale service, who are long-term partners capable of doing service for team. Okay. For India. We do have a new partner since last three years based out in Mumbai called IR Technology, and they're the ones who are currently taking care of our business for cement industry. Very good, thank you. Okay.

there's a, there's a question about comparing the vibrating table versus jolting table. is that a question that you get? a lot. It's trying to tr there's someone who's has the experience that both of them are quite similar, but do you have an opinion Compassion charting table? it's not so easy to answer this question because the, the datas from the, from these both methods must be checked by the customer. We are a manufacturer. Yeah. What we know is that 20 years ago, the 40 set in Berlin make some tests where they want to make a verification of both metals.

And this time it is long ago, we know that cement sorts with a very low cement water factor did not get the right, the same results with both metal metals. Mm-hmm. But for standard cements, there is no different, normally it's in very extreme possibilities for for mixtures that they can have difference. You know, it's not so easy to answer, therefore I think it's a better way that we go in touch with the customer. They want to know more in detail and then we can discuss it later. Yeah. Okay. So it depends on quite specific conditions. Yes. is your closed loop controlled vibrating table, the Tony VIB is that used by a STM customers as a substitute for manual hand tampering?

Therefore, we now have the, the competed verification process as be done by cmex in technical center in Riverview, Florida. It's really actual this standard committee I think they meet together next or two weeks in the future, and then they want to present the results of this verification process. That means all the RSTM customer can use our vibration table in the future. Okay. Excellent. So I hope to hear more on that in a couple of weeks then. maybe maybe we can we can publish that in our news. Okay. Yeah, sure. Well, there was a lot in that presentation. Very interesting.

There were a few questions in the q and a but I'm sure once people have the presentation, they will get in touch with any further questions. So but Shiam and Clements, thank you very much for your time. Thank you. And great presentation. Thank you. Thank you. That's that's wonderful. Well as as we get ready for our next speaker, it's, it is great to be reading through the chat and seeing where everyone is from, and it really is amazing from reek Cement in Indonesia using Tony Nic tools. we have a number from Saudi Arabia. we are oja cement in Dango Dangote in Nigeria. Welcome sim and Cameroon, new Cal sign Clay plant there. Maybe they're testing their castine clays as we speak here in the uk.

Tilbury well wholesome uk building a new vertical roller mill at Tilbury. as we speak. Again, that's coming online next year. so lots of people and even over, we're going west Mexico GCC Chihuahua, someone up early, so it's great. great coverage of the world. Welcome to everyone who's here. I'm gonna introduce our next speaker. and I'm really pleased to have a company that we've been keeping in touch with since around 20 19, 20 18. I think we first met Sami. and it's great to see the development. And today we have Anne Catherine Holka, who's going to talk to us and, and show us a a bit about their, their products and solutions. Anne is head of customer success at Alami.

She joined the company as one of its first 20 employees and has been building its customer facing functions from the ground up with a strong technical background and deep industry knowledge. She now leads a team supporting more than 40 cement plants across 16 countries. Her work focuses on delivering measurable value for customers, managing change in diverse operational environments, and ensuring successful adoption of alchemy's machine learning solutions. These tools help cement producers reduce CO2 emissions while maintaining consistently high product quality.

Over the six years, Ann has built long-term partnerships with clients worldwide, bridging the worlds of machine learning, change management and cement production. So great to have you here with us, Anne, and over to you for your presentation. Thank you. Thank you so much for the introduction. Yes. So today I wanted to present some more details about SE for cement. You might have seen in some other webinars before, generally what SM e is about. and today I want to focus and also show a little bit on the actual web application that we have.

I will still give some background first on the company itself and how our cement setup works and the results we are working towards with the plants before we do a little deep dive. So let me jump in. So in case it, me as a company is new to you, we are offering a predictive quality control software for both cement and concrete. And yeah, that means we tie into the lab equipment, for example, that we just heard in the webinar before and presentation before and predict quality parameters during production.

And the goal is to use those to be more profitable with less CO2 and with still keeping the same quality because we leverage the additional control we have for yeah, the better savings in the ongoing production. A few details about the company. So we are currently around 50 colleagues and we are venture capital funded. that means we have external support from mainly green tech investors currently in cement. We work with around 40 plants and concrete around 120 in ready mix. And I'm at least very proud to say that so far we have not lost a single plant we've been working with in the last up to six years with the different yeah people.

in terms of payback or licensing cost with the different value levers that we have, we usually see the yeah, return on invest already within the first year as and, and we heard an introduction. We are currently in 16 different countries. We are operating from Germany, however, have also our first employee in North America where we yeah, stay close with the plants over time. We are not a company to kind of set up our system and go away, but we need to make sure that what we deliver, the insights that we create can really drive value and are usable in practice. Yeah, we are, it's very important to us to stay very close to the industry.

And so therefore we are happy to have board advisors for example through Dr. Martin Schneider, CEO from the VD set and Profe professor Karen Scrivener, that really help us on the technical industry side. And as I said before, we are at the same time also tied very much in the yeah, CO2 and climate funds. So for example, we have Secretary John Kerry as one of our supporters so far about ME Let's look at how the system works. I said earlier, most important to us is the lab values that are measured about for the finished cement. The model that we want to create to predict during production is based on the historical data.

So the information we have from the past, which if in an ideal case, we would like to look at the mine mineralogy, the fineness and the chemistry in as many, in as much detail as we can. So we know how to quantify and to identify the cement properties. And we use that then to predict the compressive strength, so the input R-P-S-D-X 30 XR F values and predict it as, for example, the strength in megapascal PSI, whichever measurement is necessary. And that kind of describes the model that is kind of the little core of what we are doing. However, to make that work really, and production, a lot more goes in and behind the hood on what we're doing. That means production data can be messy.

We need data cleaning, we need to impute values. If they're missing, the models need to stay retrained. So we can keep up with also measurement that are not measured, that values that are not measured. For example, some parts reactivity of slack or clinker we can catch if we stay very close with updating the model. So there's a lot of background that today we will not go into detail but that makes really the control run in practice. So let me jump in how the model that we just described is used in the finished mail process.

So we, looking here at the setup where cement is being produced, let's say the target strength would be after 28 days of 50 megapascal, the first production sample is taken and measured in the lab as usual. What happens then is that the, a simple lab export from the limb system or the spreadsheet that is used is sent automatically to omi. And we are predicting based on the input measurements, what the compressive strength will be. So after 28 days, for example, the prediction could be 52 megapascal for the current production sample. We will then use that knowledge and compare it to the target.

So if we are here now around two megapascal above target, we directly also give feedback on the best reaction to that. That can be in fineness. we also have some setups where we use proxies for clinker, like C3 s, so similar. So let's say we have here the two megapascal higher than the actual target that we need. So the feedback that would be given back to either the control room operator or the expert system would be please grind a micrometer, Corsa in a deep prime value, for example. And with that, with every sample that is measured in production, we directly know or use the prediction of the compressor strength and directly react to it.

And what we want to see is that the fluctuations decrease with the fact that we react to any changes in material directly. Now, I described somewhat an ideal setup, but it doesn't mean this is all that we need to start with. Usually the corporations with the different plants grow over time and we look at what is currently available and can we create a setup that already improves or supports the plant in their status quo. So if, if we look at lab data, for example, we can also look at set settings where we only have C or XRF, where we would then potentially take the early strength measurements on top to predict the 28 days and still react vastly quicker than waiting the time.

We can work with different systems to transfer the data. And if there's very little data historically we can evaluate how, and we could start anyway in the beginning. There's other different, a few details about it, but I want to say it's very important to look at the whole concept, what is available and what are the goals to create a solution that is really usable in practice. And like I said, we don't like set it up once and go away, but that will be a continuous discussion, especially in the beginning to make sure everybody understands, gets that machine learning, understanding and trust. So it's actually not yeah, tossed out because of uncertainty.

What we then want to see is that fluctuations reduce as, as simple as that here pretty, yeah, straightforward example on some results from the past. And what we then have is an area where in the past we had two strong cement that was higher in strength and really required. So if we can reduce that, that is one of the savings that I was talking about earlier, either in fineness and throughput or if the clinker factor can be adapted also in clink factor, then we have the less lower strength results.

So question is, is it possible to reduce the target or at least the target that production is using Slightly, is the marking market allowing that or anything else that we can make use of the fact that there's less very low results? And lastly, of course I think concrete is also very happy if the quality stays more stable than before. So also that has its value, of course, with that, here are a few examples what we achieved with different customers in the past, as I said, dependent on what we're aiming for and what the yeah, recipes of the cements allow. All right, I'll jump into the software. Let me just switch my screen real quick to show some more real life examples.

So this is how the web app looks like. It's just started like you open it in the browser with your access credentials. So nothing needs to be installed locally. You can access it wherever you have internet connection. What we see here now is the analysis view. That's the detailed view on what's happening used especially from the quality team or production team. And in the, in the top we see some settings on what te cement, cement type we wanna see in timeframe. So here it's a limestone, limestone based cement, and we have these top paints here. The first one is showing the compressive strength predictions after 28 days.

So each of the dots that we see here is one sample in production that is predicted in strength Below. We have the prime that's here in the example, this during parameter to react to any fluctuations. And we have a table to show more details. Let's check out some, some, some examples. So if we look at the first batch here, we can see that is higher than the target, which is here currently 52 mega pascal after 28 days. If we look at the lower graph here, that the prime value, we have one, the measurement, which is the field dots and the set points. So these are the targets as I explained earlier, that are feedback to the plant. So based on all the lab values, the strength is higher.

So we can see the recommendation of fineness is to grind Corsa than the measured set points at the moment in practice. So with changing here, defined as over time, we can see here it was not changed. however, with each of these samples coming in, the reactions possible to slightly lower here. The strength arm, if we look at the second one here, the second batch also similar situation. However, we can see here that the set points actually are limited because there are a lot of, yeah, safety setups to make sure that only targets that really are in in best setup are handed over.

So here the maximum limit of fines, 25 microns might be due to the mill concrete properties or whichever reason that this should not go cossa. We can also switch to see how coarse would we go, but it will never be handed to control room of expert system more than in an allowed range. Each value, each set point is smoothed over time to make sure we don't have sudden jumps or similar and limited in how yeah, what is is handed over. Now if we wanna look at one example here where it looks like it's a bit lower to show some other functionalities. So we can use the select function here and zoom, not zoom, but select these areas.

What we can do then is go to the table and learn more about the values that have been measured for this cement. the color coding that we see is showing especially low or high values in comparison to each of the, the measurement itself. If we here take a look, there's a few that seem to be a bit unregular. one culprit we can see pretty easy here is a C3 S value. That seems to be especially low as we know the predictions take all values into account, but sometimes there's also pretty easy way to see where specifically low or high values could come from. What we then can do is also add these as a graph on top to understand what's happening here. We wanna see everything.

So it's a min-max set up, and we can see usually the C3 s has been higher and this batch it has been lower. Might be that some, some recipe change has happened, or actually the clinker might not be as yeah, as good as it was before. So what we want to see and what we can see here is that we directly react to that as soon as it comes up. So here, the starting point and find us was similar, but directly we can see that we go lower to balance out the required strength. Yeah, so that's kind of the base system, how you would analyze using this analysis view. Now, if we think about practice of ongoing, you would not always of course deep dive into each sample.

It's possible of course, but we usually want to know three things. One, are we set up that the predictions are staying good or best setup for predictions? Is the finest target or the steering target being properly followed? So we react as we need to and what are the results? And for that, I would like to take a quick look into the overview first. So first question about are the predictions as good as they can be? We have a section here, data health, because most important for good predictions is that we have consistent data and we learn ongoingly here on shipment samples. So we want to see that they're always fully measured and we can learn from them.

So we can get an overview here if that works. If we then use them on the production samples, of course we need to be sure that we can make predictions. So they need to be measured and in case single singular values are not available, they will be interpolated. So we will still make a prediction where possible, but of course that is less accurate or can be less accurate than the full measurement. So here's a good overview to keep in mind what could influence the model. Then for steering, we have a setup where we can get more insights into all of the cements, don't have to go through all the details to understand is the current production on target.

So that's the comparison between fineness the measured fineness and the set point. How we cau of finer with a color coding. Green means it's pretty close red, very far away. And that is useful not only for yeah, control if the mills manually operated, but also for experts systems. So we can make sure the moving and the the wider set points are transferred actually usable. Or also if the expert system settings need to be tweaked in any way. And also like what we just see saw, the restrictions that are important are also shown here. So if we run into some upper or lower limits on what is allowed, this is highlighted here.

And in this case might be that if we can't go corer, the question could come up. Is there the potential to take out half a percent of clean care or similar to balance out here? Because as soon as the recipe is changed, it will be reflected in the lab measurements. The steering will react accordingly and we will go back further fin than what's necessary. And that's kind of how it's an ongoing process of using the set points really in a closed loop to get to the results that we're looking for. Now, let me switch back to the other screen to my final remarks before I run out of time. The last one I haven't shown yet is then how do we see the actual results?

So currently in, in the setting that we looked at production samples, there were no actual strength measurements here we can see how that then looks like in the shipment view. So same as before, we see the predictions here for each silo sample and if available, the measured strength in comparison to it. Here we have some values where there's only predictions available because which each sample, we do give predictions back here after 28 days.

But as there's a, these are silo samples, there's no direct action associated with it, it is to be used can be, for example, used for concrete to already give a heads up on how the strength will develop and to learn generally on what the current quality in the silo is. That will be shipped out then for any that are interested in what the control room will get. So they wouldn't use the detailed view that we looked at, but one summary where mainly the information of the reaction will be transferred saying, Hey, please grind finer, please find Corsa or we are on target to give that. yeah, as a, as a direct insight. Yeah, that's all from my side.

Thank you so much for your time and happy for any questions. And yeah, please reach out in case you have would like any more detailed information or yeah, more demos. Fantastic, thank you very much. Ann Rudy good to see actually demonstrated on the interface that you, that you're using that everyone has and some incredible results there. It's I was wondering when you're changing a recipe product, how how long does it take for the system to respond to is if it's a new product? I mean, I, I guess if it's an existing one, there's a history, but if you're, if you're starting with a new product, is that something that needs to be kind of learned and then built up?

and then after time it will be able to respond in a, in a, in a more accurate way? So it depends. So all the models are trained on all the cements and all the models, all we use the whole cement base, the whole cement history of that specific plant to predict the cement. So there's some base model and it's optimized and per type. So if you create a new cement type, for example, higher limestone base and before you already had limestone cements, then that would be fairly easy for the system to adapt. So we wouldn't need many samples to get already some, some good insights and results on the predictions.

Now if we talk about a completely new material, for example, a slack has never been used and now the first time a slack cement is introduced, we need some more back some more content, some more samples to see how well the model performs before we could start. So we would kind of go into the conversation, see if you already from lab experiments or similar, you have samples, set it up and we can already create it and not use it yet in case that's new. And so mm-hmm. As soon as we see it that the model performs as we would like to be, we can switch and use it. Yeah. Okay. Very good. There's a, a couple of questions that relate to what I've just asked.

I can see now there there's also some, some obvious questions around return on investment and how much time do you spend with customers per week? Could you talk about those two aspects? So how much time do we spend in the beginning? We would like to talk at least weekly to make sure everybody understands what we see. It can be a bit much in the beginning and we can discuss how is that actually used and in practice, how do the processes change? But it's not required to stay in touch that often, but we usually would like to then ongoingly, we usually meet once per month just a short check in. but again, it is more for us to kind of keep in touch.

We can reduce it if that is what we would like to, but it's helpful to understand what are the goals and the current challenges, and we can advise in which way we can use this, this system best. And the second part was for the return on invest. Mm-hmm. so the, the time and volume is of course, again, it depends. That's I guess the favorite answer. So for some plants, the already the, the biggest impact is the stabilization of the quality. We say we had a very high strength before we directly get into that and can see that fairly quickly.

If we have now a plan that has very low data set up and we need to get some time to kind of get the model running in a way that we want to steer, we need some more time to gain trust, it can take a little bit longer, but usually within the first year, we definitely want to see the, some proofs and some calculations on how that will work out long term. Okay, thank you. and there are quite a few questions around kind of minimum requirements of the, does the system integrate into the laboratory information management system and what is the, what are the minimum amounts of data that you require?

But I think you've kind of covered that on that slide where you, you you kind of gave the range of setup mm-hmm. From the minimum, which is really quite basic actually you can, you can still operate off quite a, a simple bunch of data inputs. Yes. But of course the, the model quality will reflect that there is less inputs, but always the question, does it improve on what the plan is currently working with? Mm-hmm. Maybe just to touch shortly on how to integrate. So if, if there's a limp system, we would look at that. There's often a way to schedule a report like a CSV file every 15 minutes and there's a little script running in the background that upflow uploads it to the cloud to us.

So that is what would be, what would be installed kind of that that little li setup and upload. So it's should be fairly easy usually, but it really, if there's the database system already running or if there's just some manuals, spreadsheets, we will take a look and understand how can the IB centrals. And you shared the example of of focusing on strength as the target. Mm-hmm. can you control initial set time and can you, what Currently Areas can you, can you set the system up for? Yeah, so we use compressive strength as targets.

We can add limits, for example, also for bla or similar, we have some settings where we pre kind of predict with the focus on slag reactivity that some special setups, let's say, not our standard way of doing it. If we talk about setting time or concrete properties, we did look into it, but so far we haven't had results that we are happy with. So we would not be able to give that currently as, as limits or, or targets. Okay. Okay. Yeah, something for the future. Yeah. Well thank you very much. the questions are kind of getting out of hand now that in the chat and then the q and a, so you might want to just move over to those.

And I've noticed that David Shepherd your colleague is there as well and supporting you. So hopefully you can you can get through some of those, but obviously we are sending out the recordings of this presentation along with the slides contact details. So if you need to get in touch with Anne or David, then you, you'll have the ability to do that. So once again, thank you very much. Thank you. And I'll answer very good. So we're gonna, we're gonna move on to our, our third and final speaker now. And I'm gonna invite Simon, Simon Wels Miller from Thermo Fisher Scientific to share his slides.

and Simon is the global application specialist for x-ray diffraction at Thermo Fisher Scientific, where he leads the global XRD applications team. He holds a PhD in solid state chemistry and crystallography and bring more, it brings more than a decade of hands-on experience across a wide range of XRD methods and material systems. Since joining Thermo Fisher in 2017, Simon supported scientists and industries across the world in applying XRD to solve real analytical challenges. His work spans mining minerals, semiconductors and cement, as well as sustainable materials such as activated clays for CO two reduced cements.

So right on target and with a really good fundamental look at optimizing cement quality and process control through REIT failed analysis. Simon, over to you. Thanks for the introduction. So welcome everyone. So in this presentation I will show you where some of the data which is used in this very nice model is actually coming from. So we are talking about XAD analysis and we felt analysis. So first I want to give a brief introduction into our company Summer Fisher. So we are quite big. We have more than 125,000 colleagues globally and huge investment in r and d and also very huge revenue of, it's currently more than 40 billion.

we have a broad variety of different products for element and phase analysis of part materials. So I think a lot of you are familiar with a lot of these products. So we have from crossbar analyzers especially, this one is used in cement industry over spark OES, which is mostly used in metals industry to XOF spectrometers. So we have a series of ED XOF and W-D-X-O-F-X-F spectrometers. I also want to draw your attention on the recently released new X 900 XF specter, which is especially designed for the cement industry. And more recently we also launched a new XA Deep product, which is called X Extra Companion, which is this one here which is a benchtop iadi, a routine analysis.

very recently we also started this lali to, which is a very highend instrument for also elemental analysis. So now a little bit more details about the our XRD instrument and also new release, which is which is, we, we are the first audience to really share it. So the extra companion is a bano CC instrument has a quite wide measurement range, which is not really necessary for cement application. We use a very modern time. Pick three, 2D solid, say detector. very nice feature, which is good for, for this type of analysis, is a motorized beam knife, which is a default option for us. The resolution and accuracy of the in, of the instrument are very good.

You can see, well, you'll see it very much in line with with requirements. We have an integrated chiller, a sample spinner six position auto sampler. we support one click click analysis based on, on wheat felt analysis. And of course, you can connect the instrument to a limbs and have direct result transmission. Now, our new release is that we also have a 1.2 kilowatt version now available. So if you, if you have higher samples, throughput you can actually half your measurement time with this new version. So now let's come, let's come to the science.

so first I wanted to tell you what you can actually learn from an XRD analysis, because there are different parameters you will, you will get from it. So first we have the peak position which actually is a indicator of the face. So this would allow you to identify, for example, allied daylight, gypsum, whatever. So this is mostly based on the peak location, on the peak position. Then we have the peak intensities. So basically how high the peak is. and this will is used for the quantification. So these are for the quantities of C3 s, C two s, whatever. So this is used in the refinement. Then thirdly there is the peak broadening.

So you see, you can have different peak shapes and different peak widths. this is an indication of crystal size and micro strain. So as a rule of sump, you can say if the peaks up water, you have finer crystals and basically of a more reactive clinker, for example. there you not to confuse with actual particle size because with IX a D, we measure the size of the scattering domain, which might not a hundred percent reflect the size of a particle because this could be just built from multiple crystals, which are just stuck together and form a particle, which you would measure, measure, for example, by laser diffraction. And then we have an amorphous hump like this one here.

This is for the glossy or non crystalline content. This, for example, is necessary for supplementary cementitious materials like clays or also ashes. These are very amorphous, and this information can also be derived from XAD analysis. so now how can you actually use this type analysis of analysis in the cement production process? Here's just a little workflow. So I have some selected examples for you from some of the steps. let's start at the beginning. So for the analysis of raw materials one of the major raw materials is limestone. So the, the, the information you can get from this analysis is basically, basically the face makeup.

In this case, the sample contains mostly cy with a little bit of dolomite in quarts. From that, you can actually calculate it as the QMetry which you would see here. And you, there you can actually yeah, get the quality of the, of this raw material from this analysis. So it has very high calcium carbonate, lower silicon aluminum iron oxide. So this is really a very excellent lime source. It's, it's very pure and which allows you to have a very stable hen feed and which improves the reproducibility. And it's also very energy efficient. what also would allow this analysis would allow you to detect like other phases.

for example, if there would be a clay or whatever else some felt bars, which could be in the, in the, in your, in your limestone and by XID you would be actually able to detect them by XOF. You would only see the elemental composition and, but not where these elements are actually bound. So and there are, there is always some monetary and, and, and, and environmental impact. So you can actually save some fuel. You have lower variability in, in your production, and that's very good for sustainability because you can re reduce the CO2 emissions. So then the next step would be the raw meal, which you can also analyze by.

So d so it's basically the mixture at the start of before the before the calcination. And this sample here is mostly, mostly consists of, of calcite and dolomite. So this is some limestone, which was mixed with a quad source. And here we also have some some other some other materials like clays and spars. There was also iron oxides in the mixture. This is here mostly hematite and ide. There were no clinker phases yet because it's the raw meal, so there is no calculation. And so this can help you to optimize your fuel use and and actually yeah, allow to, to reduce the, the amount of decomposed CO2. this will increase the clean stability and also the quality predictability of your materials.

So here are some, some, some values. So you can actually save a lot of money if you, if you optimize your, your war mix control, which also would use the CO2 emission and you have less waste. So if you have a better product, a better product quality, you also have less waste. So then the next step is in the calcination precum stage. I mean, most of you would maybe ex expect a hot meal here. I'm try to show you something else, which is more of a carine product which is used for cement. And this is a calcine clay here. And I wanted to show you that there are different things which are called calcine clays.

And this is an example which is not particularly useful in a cement because it consists mostly of moon light, which is a high temperature product during the calculation of clays. but it's in earth, so it'll not contribute to the cement strengths. There is also a lot of unsigned caite here and also compound ide and ide, which are aluminum hydro cytes. and there is a little bit of amorphous content here. You see, see this hump here. but the content is not so high. So this means we have a transformation of caite to a cal and clay, but it's not complete. And the, and the mu light indicates a very high temperature, but not a complete transformation. And which has a very low activity.

And also the GCI and the mite, which are, which we can find, indicate this is some kind of blended material where different different mixture. It was basically mixed and it's, it's not a good, good material for, for cement using a caran clay. So this is a, a simple, which is actually available. So you can buy it on the market, but I would say it's not very useful. So, and what you actually want to do for these is you you want to optimize the temperature for the calcination because yeah, this is really money you want to avoid having an under burned material because the koli night, which is not ign properly, will not contribute to the salmon spans.

And you want to also have a high amorphous content. In this case it was 7%, but it should be more in the 30 to 50% range. So now back to more classical app applications. and here we have a clinker sample. so there are different compounds here components of this, of this clinical. So there is C3 s and C two s, which is like a very, a nice example here. It has high C3 s complementary C two s. There is C3 A, which is very much in the optimal range. We find also C four af. we have kind of hyper peric place, but still it's still in the okay range, low free lime. So this means we have a complete burn and a good Lyme situation factor. There are some sulfur phases, but it's these are also minor.

we have a little bit of Portland diet, which shows a little bit of hydration. And also the quad is quite low. So this is also from, from the, from the starting materials. So this is a very well burned, very balanced clinker. And actually if I wanted to show you what you should watch out for, if you do this analysis on clinkers. So there is the free lime. So if the free lime is very high, this indicates that it's under burned or you have a there is something wrong with the LSF or the fetus to course. so you, you can take some correctional actions. I indicated a couple of them here. I'm not going to read out all of those.

then another case would be if C3 s is, is low and C two s is high, this is an indication for a low temperature or a poor vulnerability, then you should adjust your, the parameters of your, of your warm mill. If there is the gamma phase of bill light detected, this indicates that something is wrong with the cooling, so it's too slow because then you have a transformation of better to the gamma face of be light. And the problem with the gamma face is, is that it's in earth, so it'll not contribute to the cement strengths. So you really want to avoid having a lot of C two as gamma.

So yeah, you should, should adjust your cooling weight and to avoid having like a big amount of this gamma face if your peric clay is, is very high, this, this could indicate that you have high dolomite in the feed or slow cooling. And because this causes expansion during the setting, you want to reduce the, the peric clay amount. So also if the quarts exceeds one or 2% yeah, this means your silica is not fully reacted, and you should adjust your parameters accordingly. And you could also have a very high C3 A, this is the aluminum ratio in your raw meal.

And yeah, you should adjust it accordingly because this also has an influence on the, on the, on the setting properties of the, of the cement. Yeah. And also, yeah, if the ali sulfates could also be too high, and there is a risk of coating, so you should also balance this. There are a lot more parameters to control, but this is just an like a, a little list to start with. And of course, there is also a monetary and environmental impact here. So as usual, if you optimize your parameters, you can save a lot of fuel which will reduce the CO2 emissions per ton of, of, of clinker.

And you can also increase the, the, the lifetime of your kiln if, if you have a, if you have a, a homogeneous reaction. Yeah. And overall you will just reduce your CO2 footprint and improve your energy KPIs. So the last step is a cement here. It's hydrated in bracket. So it's a little hydrated just to show you that actually the age of a cement has an impact on the analysis. So the makeup of this sample is as you would expect, so there is there is a good amount of C3 s and C two s with an adequate amount of C3 A. We have a typical amount of C four F, the, there are also calcium sulfates here because it's a cement, which is mostly ignite and gypsum.

So the, you, you usually want to mix the two to moderate the setting. We have a low free lime, so we have complete burning. It's a, it's a well-acted clinker. we find around 2% of Portland diet. So this is a indication of dehydration. So this means this could mo is most likely due to storage or handling of the cement. when these samples reach our lab, they're usually not fresh, so we always see some hydration. And then this cement also contains carbonate. So there is dolomite and calcite in the sample. This could either be to exposure, air moisture, but more likely it's a intentional limestone addition to the cement. So actually it's a, it's a STEM two, so it contains some, some limestone.

So, and what, what impact would this have? So so the product reactivity, so if the, if the, if the, if the cement is too much pred this reduces reactivity. But more importantly, you want to balance the different hydrates of calcium sulfate, anhydrate, bass, and gypsum, because this has an influence on the cement setting. also if, if it hydrates you, it's, this is not good for the grinding, and if it hydrates too much you might have to dispose it or you have to rework the, the, the cement. So you want to avoid this. So now my final remarks. So what is important about this?

So you want to quantify every phase at every stage from raw materials to final cement, because by doing that, you can control the chemistry, the process, and the product performance. So this is really important. you want to optimize your kiln in calcination parameters. So you can, in real time track faces C3 s, C two s, the balance you see the, the, for, for example, for cal and Clay the ratio of CARite versus mo light versus amorphous pop, which will give you a good idea of the quality of the carine clay. So you can detect hydration, carbonation, and also different polymorph changes which is important for the reactivity and in consistent cementing cement quality.

And also for sustainability, you of course want to use as little fuel as possible to reduce your CL two emissions and also save a lot of money, and it's, it's for me. So thank you for your attention. I have some contact details here, so if you have any questions, please feel free to reach out to anyone on this list. And oh yeah, I have to mention that I forgot to mention it in the beginning, actually, the announcement was for my colleague Kata to give this presentation, but unfortunately she wasn't available, so I had to cover her quickly, and that's the reason why I'm giving this presentation today. So Simon, thank you very much.

that's a, a great, a great presentation, very comprehensive, and thanks also to Carlotta for her efforts in, in putting it all together, and hopefully we see her another time. really a really fundamental piece of equipment and you've, you've really demonstrated the applications across the cement production and process. it's it's interesting to see you interpret these results as well. is that, I mean, that's part of the, the process of understanding the, the the possibilities of your, your equipment, I guess. but do, I mean, are you do you have a, is there an option for users to, to get in touch with you and to ask questions? Is that part of the service offer offering?

I mean what we offer as a service is a method creation. So if if a customer would like to obtain equipment, we offer this one click analysis, which usually includes to set up a method which consists of sample preparation and all the measurement parameters, and also the regional parameters. So the operator there is zero analysis involved. It's really just you push one button and at the end you get the full wheat for quantification for the interpretation of the results. I have to say, I have to rely on sources I can find in the internet because I'm not working in a cement factory. Mm-hmm.

So I'm not so much up to speed with, with the ranges of each of these parameters and what how to actually steer them. This is really more up to the, to the customer. I can give some guidance, but on the details, I'm, this is above my knowledge. Yeah, sure. just looking at the the package in, in general in terms of the, the throughput and automation opportunities what, what's the sample ca capacity of these machines and how, how can you integrate them into, into kind of robotic labs and that, that kind of thing? I mean our current product is for manual operation. So it's a, it's a benchtop instrument, which is for manual operation.

samples, throughput with the 600 watt version is around five samples per hour. because measurement time is usually around 10 minutes, so you can handle five samples per hour if you need more. That 1.2 kilowatt version can basically handle double that because you can half the measurement time. So then it's around 10 samples, I would say per hour. And yeah, for automation at the moment we only have the X 900 product, which has a integrated compact XAD, which has some capabilities for which are not as broad as the ones for the full XAD. So you can get information on the main phases, but you cannot measure like all of the details of all of the polymorphous slinker, for example.

But this is fully automatable and integrated in our XID in our XOF instrument that's available. And for add the additional instrument just follow us on, on social media on LinkedIn to keep up to date with, with our new developments, so then you don't miss out on it. That's, that's right. Or, or read international cement review. Oh, Yeah. Or come here to the, to the SIMTECH webinar. yeah, no, that's, that's fantastic. there are lots of questions, but one, one interesting one is how accurate is XRD in quantifying low amounts of B light? that's from one, one of the, one of the questions.

my question arises because sometimes B light isn't visible under a microscope, but XRD does give me a reading for this mineral. I mean, so as a rule of thumb, the limit of of detection, limit of quantification is around 1% per face. for wheat 12, if it goes below that, there are other methods like calibration curves you can use. But if it's up, if it's above 1% it's usually quantifiable. And, but I, I know this is already very low for pite in, in, in clinker or in cement. I mean, for by iad for with IAD d you, you you analyze a different atomic structure of the material.

So it will clearly see the difference between allied and B light because it has a completely different structure, different arrangement of the atoms. So yeah, it's actually the best method to do this kind of, of quantification. Very good. and maybe one last question just around maintenance and, you know, what are the requirements for the x-ray tube and what's the expected lifetime? You know, in a typical cement lab, I mean so our instrument has an integrated chillo, so you don't need any external water chillo. The instrument runs on the standard 220 volt block, so you don't need high, high voltage or anything. in terms of maintenance, we offer maintenance contracts.

This is mostly some air filters, and there the, the system of course has some cooling water, which should be checked occasionally, but there is not a lot of, of like extensive maintenance for this type of equipment and tube lifetime. I mean suppliers of these instruments usually give, don't give any, any guarantees for tube lifetime because it's a consumable, but I can say with the 600 watt version, we are using a 1500 watt tube. So we are actually tickling it because it's really running much below the, the maximum power, which should increase the lifetime dramatically. I can tell you if you want a tube like at its maximum power, you will lose like maybe 50% of intensity in a year.

But because we low, we want it so low it's so, it's so, so much below its maximum power. Usually you can expect a couple of years of two lifetime. So that's what I would as a rule of thumb. Very good. Okay, well that's that's a great presentation. Thanks for stepping in for Carlotta. Thanks for No problem asking questions. that's all we have time for today. thanks to all our presenters from Tony Technik, SMY and Thermo Fisher. I think we've had a, a fascinating discussion and lots of good ideas about the technologies that are available in cement plants and something great to end the year on. so thanks to all of you. we'll be back in January, 2026. It sounds amazing to say it.

we'll be back then for our webinar on Wednesday, the 7th of January. we'll be looking at alternative fuels. and we have four great presenters lined up already to get us underway as soon as the year starts. But for now I'd like to, to wish you all the best for the rest of the year, the last few weeks in the runup to Christmas and the new year. thank you all for watching and look out for your email, which will give you the opportunity to download a recording and all the presentations that have been made today. Thank you very much, and see you in January. Bye. Thank you. Bye-bye. Thank you.