The most modern technologies are there for the mechanic. And not the other way around.
On January 25, 2024, FHI will organize the next edition of the Production Process Automation (PPA) event. I am speaking with Perry Jaspers, one of the plenary speakers.
By: Hans Risseeuw
Perry Jaspers, Technical Process Owner (TPO) Electrical & Automation at Yara, gives a plenary lecture during PPA about Digital Production Platform (DPP), asset management and realization.
Six years ago, Yara started its digital journey, and Perry immediately admits that Yara made exactly the same mistakes as the other big end users: software as an end, not a means. Perry already indicated six years ago that Yara does not need new software, but a new approach: “Can we tackle problems that we cannot solve in a traditional way in a different way?”
Data is the first step
“The first step is to unlock data and make it easily accessible. But if there is one truth within the process industry, it is: the data is available, but very inaccessible.”
Worldwide Yara has connected all control systems to AWS, Amazon Cloud. In principle all process data is available for data analysis. This is the Digital Production Platform (DPP). The next step is to create value from the data.
The biggest challenge was the development of the Digital Production Platform (DPP), mainly because the different OEMs (Original Equipment Manufacturer) communicate with the Cloud in different ways. “For example: a system that Yara used in India did not communicate with the same system that was used in Europe.” Now, all data is available in a standardized way in the DPP. This makes further development much easier.
All data is available. Now what?
One of the advantages, says Perry, is that he can now conduct pilots himself to see whether they can also create value with this. He starts by asking the question: where is the greatest value creation theoretically possible? This is within the group of assets where the greatest losses occur. In the case of Yara, you quickly end up with heavy rotating machinery. But these are also the extremely expensive solutions. “This is often where the initiatives stop immediately.” He is talking about very expensive heavy turbine-driven compressors. And in addition, these compressors are quite sensitive to failure. He mentions the nitric acid compressors as an example. These are so complex that there is only one manufacturer worldwide that offers these compressors. “These are very complex systems that very few specialists can work with. In theory, you could tinker with this, but because it is so complex and expensive, the decision is quickly made: it works, we will keep it this way.”
Yara then introduced a value potential: “What is the ratio of value over cost.” Perry thinks of all kinds of condition monitoring systems that are already installed, but not (fully) used yet. For relatively low costs, Yara could use these monitoring systems better – in theory. Now Perry prefers to build trust with simple solutions. Small steps.
One of the steps that Yara has taken is the establishment of its own digital group. This group of specialists has developed various solutions on the Cloud platform. Such as, on/of valve monitoring. On/of valves always have a safety function; this mainly concerns the time span between command and end position of the valve.
The journey to standardization
The big problem with systems developed by the big OEMs is that they develop systems with lots of alarms that are useless to Yara. These complex systems do not provide any value to Yara, except a lot of effort. Simple solutions, like the on/off valve – set point and position of the valve – provide direct value to Yara.
The next question Yara asked itself was whether it makes sense to set up a 'remote monitoring centre'. A central point where a number of experts - whether or not in a physical location - provide support to the various sites. Perry himself is a great supporter of this solution. Because he sees this as the only way to implement standardisation in the various solutions.
“We have a sea of data, we have different pilots, which are available for all sites and now we are going to see if we can centralize this, physically or otherwise.” Experts who monitor the different sites and do condition-based predictive maintenance. On the operations side, they also focus on energy use and efficiency of applications on static equipment.
“Predicting errors is unnecessary. Indicating that a machine is running well, that is where the value lies.”
The key for Yara is in the valuation. “Every year you have to make a budget plan. And of course we all have a long-term plan. But even if you have a perfect system, without a 'technician' you are nowhere. What do you do where? The goal is not to generate data.”
For example, he says about predictive maintenance: “When you talk about predictive maintenance, the expectation is that the system will tell you exactly what is wrong with the machine and when. But this is fiction and is still far in the future. What we have now, and what we only want to know now: okay, the machine is running within predetermined parameters, you don't have to look at this now. Especially in light of a general shortage of resources and people, this approach offers a huge advantage. Predictive maintenance can very well indicate what you don't have to look at. The machine is running well.” Perry states that this approach was an 'eye opener' for Yara. “We don't need the latest of the latest, and we don't have to measure everything and know everything. We want to know whether a machine is running 'okay' within a predetermined framework. As already mentioned, an additional advantage is that you don't need the top model for this. We can already get there with fairly simple solutions.”
Data could therefore be used to deploy resources where they are needed. And not where we think they are needed. Yara wants to deploy data in this way. Concretely applicable, and value-driven.
You write here for a free visit to the PPA event. Be surprised by the participating companies and explore the extensive seminar program in depth.