Brüel & Kjær Vibro (B&K Vibro) has been monitoring in the cement industry for more than 35 years and has gained a lot of experience since then. This has fortunately resulted in a lot of success over the years but there also have been a fair share of challenges. Hugues Delannoy, Channel Partner Manager for EMEA at B&K Vibro, is one of our specialists who has been intimately connected to the cement industry for many years. He shares some of his experience in this interview.
You have been personally involved in several cement plants over the years. When did it start and what kind of plants have you visited?
I have already been working for 35 years in the company, so I’ve seen a lot of different industries and machines. Cement is probably one of the first plants I visited because at that time there were only a few industries using machine condition monitoring, such as Oil & Gas and Process Industry. At the time the cement industry was also important. The monitoring techniques and cement industry were also important. The monitoring techniques and technology were quite new and expensive.
I started with France but I also visited plants in Spain. Then I went to North Africa and sub-Saharan Africa. I also visited many plants in Europe. One of the things you can say about the cement industry is that you can decide to install a cement plant anywhere in the world. Limestone is found everywhere and the same technology is used everywhere, which makes it easy to build a plant. This is different for the oil and gas industry since there are only a few places where you find oil. And the technology to produce and refine oil and gas is different for different regions and products. Although cement production technology is not very difficult, operation and maintenance of the machinery can be more tricky.
What kind of need has there been for machine protection and condition monitoring in this industry and is it the same for all cement plants or is it very different?
It’s basically the same everywhere. The reason why they started with machine protection many years ago is that the critical machines are often very large. A catastrophic failure, therefore, can be expensive due to both downtime for the extended repair time and the cost of the repair itself. Later on, condition monitoring became important to reduce expensive production downtime that would be due to an unexpected machine protection shutdown.
From a process point of view, all cement plants are similar. The main difference is the size of the plant. You can find very small plants in some of parts of North African and southern Germany where there can be eight production sites, where each has two or three kilns. That’s a really big difference. And as the plants get larger in size, there is more money involved and the need to monitor the machines becomes more important, partly because there is less tolerance for downtime.
Are the larger cement plants more complex?
Of course, because there’s a need for a lot of raw material and it is possible there is not enough around the plant. Even if you can find raw materials elsewhere, there is a balance point where it may be more economical to build another plant where it would be easier to find raw material.
For those plants that had a need for a Condition Monitoring System, what kind of maintenance history did they have, and what were the consequences of failures?
Condition Monitoring started to become important around 1995 because the machine protection systems were protecting machines but there was significant production loss due to the unexpected shutdowns. There emerged a need to get better information on the machine health. The technology at that point was not ready to automatically provide the information needed. So, the first condition monitoring systems that came out in the nineties were portable data collectors. And that’s the technology that was available when condition monitoring became important. The portable instruments provided vibration data but, the real need was to better know how the machine will behave and when it has to be repaired. We have to make sure that you get fault detection and diagnostic information early enough for planning, machine stops, repairs, and spares.
One of the problems they have in the cement industry is that the machines are very often unique. They don’t have a spare. As for example, the large mills and the kilns. Often, they have only one kiln. So, it’s necessary to keep the machine running and avoid unplanned machine stops. If the kiln or one of the other large machines fail, then the plant can be completely stopped for months.
A cement plant is similar to other process plants; it’s a continuous process. The machines are running 24/7. And that’s something you have to take into consideration. You need to have a good overview of the health of the machine to accurately plan a service shutdown with good lead-time and with minimal impact on the production. At the same time of course, you want to avoid unexpected shutdowns.
What would be a typical monitoring strategy for these plants, if there is one? And is it different from other industries?
It’s not very different from one industry to another, but one of the biggest challenges we have is the operational environment of the machines. There is a lot of dust in the air and it finds its way in the lubrication systems also. Another challenge is the machine type and configuration. There are a lot of fans and many of the machines are belt driven. Other machines like the kilns or mills, are unique to the cement industry. Until recently it was quite difficult to automate fault detection and diagnostics for a kiln for condition monitoring. We recently developed a new product called VCM-3, and this product fits the kiln monitoring application very well. It has recently been adopted as a standard OEM for the biggest kiln manufacturer in Europe, FLSmidth. It’s a really good step towards the future. The step forward in this industry isn’t coming from the evolution of the cement process, which is still more or less the same. But the technology of machine measurement, the technology of storing data is evolving for us and this allows us to go forward into the application.
I remember reading about AI and its development. Is AI helping with being able to do condition monitoring for these types of machines?
Not yet in cement. We have a lot of experience with wind turbine monitoring and have implemented some AI functionality to facilitate diagnostics. This is important for this particular industry since we are monitoring more than 30,000 wind turbines. We will extend this to other industries in good time. It is important to say, however, no matter how much AI we implement in our monitoring solutions, it is still important to understand that this functionality reduces the work load of the diagnostic specialist but does not replace that person.
So, whether AI or not, there’s always a human element involved to mange it?
When I started my career years ago, I did machine vibration monitoring with a tape recorder system. And at the end of the day the guy in charge of the plant would ask the same questions. How is the machine doing? Do I have to stop it or can I continue my process? It was not always so easy to answer these questions 35 years ago, but now AI and ML is making it possible to answer these questions more reliably and quickly.
The last question I have is has the industry changed over the years and what does it look like for the future?
The change is not really linked to the process itself. The industry is producing cement the same way it was produced many years ago. Perhaps the machines are getting a bit bigger, a bit better, but there is not so much change in the process. The main change, in my opinion, is really the change of paradigm of the cement market. Nowadays you have fewer players in the market. The cement industry is a global industry. The industry recently had a merger of two of the biggest players in the world. Therefore we have an enormous company in front of us. The small cement plant remotely located in the Sahara desert are more or less disappearing because they are being bought by one of the big companies. The big companies have many units and plants around the world. They need to control it more and more. That’s the first challenge they have. The second challenge is that we have fewer and fewer people on site. Moreover, more things are automated on-site. The guy that use to run around the plant placing things around the machines, collecting data, then going back to his desk using software to make trends and diagnostics is disappearing.
The companies are getting bigger and they need to manage their big data. They need to get all the data about their machine behavior from within the cloud. And with AI and machine learning, they can do a lot of analytics. Up until now, it was difficult to do that with condition monitoring because vibration is something that is very specific but now with the new product we were talking about a few minutes ago, the VCM-3, we are able to fit into this digital transformation.