Get best of the week

How our sensors spent their first winter

Hello! In one of the previous articles, I wrote that we decided to make our own explosion-proof sensors for our facilities, and not to use third-party products. So, it is said - done. Now we have a pilot batch, 190 pieces, which we have been testing for six months. The start took place in November 2019, when we installed a sufficiently large number of sensors in Tobolsk to see how they work in combat conditions and collect complete statistics. Six months later, we can say for sure that the idea of ​​doing everything ourselves was correct and was justified - they normally went through baptism of fire in winter (and winter in the regions where our facilities are located is a very, very serious thing). So, we will release such sensors in circulation.





What are these sensors and why are they needed


First of all, these are sensors that measure both the temperature of work surfaces and the temperature of objects. Therefore, the sensors should work in a plug of temperature conditions from -56 to +50 degrees Celsius. We put them into the production of ZapSibNeftekhim pyrolysis in order to protect the pipes of the heat satellites (production facilities of secondary importance) from freezing. Prior to the introduction of the sensor system, this control was carried out by people who walked and touched these pipes with their hands for the subject “frozen - not frozen”.

At the end of 2018, while the plant was not yet commissioned, we delivered a batch of non-explosion-proof sensors there just to check whether they would work or fall off in the cold. Then it was still possible, there were no explosive zones. And in the summer of 2019, after the start of the plant, such zones were expected to appear, and the sensors had to be replaced before the winter.



On these lines, a number of people have thoughts - yeah, they put sensors, and people had nothing to do, everything, robots are crowding out people from work, it started. In fact, everything is not so clear. ZapSibNeftekhim, in principle, was planned as an object with a fairly high level of automation, and the installation of these sensors simply supported this strategy. In our country, the specific number of personnel is, in principle, less than in older factories.

Problems


There were no problems with the sensors, for the most part thanks to a rather harsh approach to testing (this was explained by my colleague Justina Tsiga, engineer of Industry 4.0). We have strict criteria for checking equipment, therefore, if some sensor could have cluttered in the winter, it would certainly have shown its thermophilic nature even during testing. And so all winter everyone worked out with a bang.

What the problems were with is the batteries. Firstly, the first batch of batteries for sensors with a high percentage of rejects came to us corny. About 30% of the batteries are dead before the declared time, their actual capacity was significantly different from the one declared by the manufacturer. The remaining 70% is still normal and continue to work, maintaining the nominal capacity. We solved this problem quite simply by changing the supplier and starting to buy batteries in Europe instead of China.

Speaking of batteries, these are lithium-thionyl chloride power sources. The sensor is organized in such a way that it sends a signal via the LoRaWAn protocol not constantly, but once every 20 minutes. Therefore, on one battery, it can fully work for 5 years.

The second problem was that the staff was a little unprepared for the process of installing sensors. It seems to be from the side, it seems nothing complicated, took the sensor, put it where necessary, and went on to work. In fact, often during installation they simply cut off a number of wires. We decided this by the operational release of detailed instructions - how to install these sensors, what to look for and so on. Now a special training class has been equipped for this in Tobolsk based on SIBUR Inteha, where our sensors are located in the mechanics class and training is being carried out on installation.



Production, work, plans


It was possible to achieve a situation in which we can say for sure that the sensors are a Russian development. Because the engineers of two companies took an equal part in the development process: SIBUR and our contractors, the M-Plata company. These are great guys, we paid them development services and are owners of intellectual property for the result. And M-Plata has its own electronics factory right in Moscow, from where the sensors came from.



This is a very big plus, not only from the point of view of import substitution and logistics, but also from a purely practical one: it turns out that such production in our country is largely detached from the foreign exchange rate. Enclosures and sensors are also manufactured in the Russian Federation, and not in China. Of course, we continue to order a number of electronic components (and the batteries mentioned above) from abroad.

As I wrote above, 190 sensors were taken as a pilot batch. In fact, there will be more of them, because we continue to equip the remaining sections of ZapSibNeftekhim and other holding plants. We were completely satisfied with the pilot, we will continue to actively integrate sensors with the rest of the enterprise management systems and build predictive models and analytics based on data.

Yes, the sensors work in close conjunction with the SIBUR IoT platform specially created for such needs, which I will discuss in more detail in a separate post.

Vasily Ezhov
IoT product owner in SIBUR

More articles:


All Articles