Profile
How can energy-intensive petrochemical plants promote industrial sustainability through AI? The petrochemical industry is a high-capital and high-tech-intensive industry. There are tens of thousands of equipment in a single petrochemical plant, which relies on a high degree of automation. The center is dispatching and commanding, so there are not many people actually operating on site. However, the petrochemical industry is very similar to the aviation and nuclear energy industries. The requirements for equipment reliability are very high, and mistakes are not allowed. For this reason, the early practice was to buy the same equipment as a backup. However, not all equipment is suitable for this method. In addition, the spare machine is easy to break if it is not used for a long time. Therefore, there are "regular maintenance" and "Predictive maintenance" using sensor detection. Until later, concepts such as reliability maintenance and risk management began to appear in foreign countriesa. Formosa Plastics gradually introduced smart maintenance, grasped the risks of equipment through AI, and provided on-site personnel with equipment management countermeasures to deal with these risks. Smart maintenance is actually a very shocking concept for petrochemical plants. Zheng Qicong said that the traditional concept is that the maintenance benefits are not high, and equipment failures can be repaired by outsourcing companies. However, it was later discovered that this model is not suitable for Taiwan. One of the reasons is that Taiwan Small and medium-sized enterprises are small and cannot afford expensive equipment, so outsourcing factories may not be familiar with these equipments; secondly, technical personnel for high-end equipment are not easy to find. Therefore, Formosa Plastics established its own training center and improved maintenance efficiency by introducing AI. Lubricating oil is changed every time 5 million AI assists in judging the timing of oil change What benefits can the introduction of AI bring to the petrochemical industry? Zheng Qicong believes that what AI can do at this stage is mainly "human-machine collaboration" to assist on-site personnel to work more easily and efficiently. "A petrochemical plant has at least 30,000 to 40,000 pieces of equipment, and there are only 30 equipment engineers and supervisors in charge of maintenance. The number of people may be further reduced in the future, but the machines will be more complex. Only through digital transformation and the introduction of AI can management tens of thousands parts." He gave an example. The purchase of equipment assets in petrochemical plants starts at hundreds of millions, or even more than one billion. It takes five million yuan to change the lubricating oil, which is about every two years. This is far from the concept of changing the oil in a general car. It is generally believed that lubricating oil is a consumable and cheap, no matter whether it is good or bad, it does not matter to replace it, but the oil in petrochemical plants is different, because the oil needs to be chemical resistant, high temperature resistant, and resistant to consumption, and many special raw materials must be added, so it cannot be disposed of casually. . The replacement time can usually only follow the original manufacturer’s suggestion, but the cost is relatively high. There are also manufacturers who claim to be similar to the original manufacturer and want to provide oil products, but the expensive machine cannot afford the risk. At that time, I began to think, when is the best time to change the oil? How to make the oil stable and long-lasting? The oil regeneration system was born for this purpose. It can detect whether the oil has moisture and cause deterioration, and further check for leaks, check whether the process is mixed with the oil, and find defects immediately. Zheng Qicong mentioned that just one oil product requires many considerations. If you replace it according to the original manufacturer's suggestion, although the equipment can be kept stable, it is not friendly to the environment.
Forum Role: Participant
Topics Started: 0
Replies Created: 0