Refinery improves efficiency using smart data
The process control system at Qingdao refinery already ensures that the operators have full control over the plant at all times and provides direct access to all relevant plant and process data. This means that the owner, Sinopec, can easily adapt plant operation to new requirements. As a result, the plant isn’t just efficient, it’s also highly flexible.
SPS 2021 in Nuremberg – There for you
Come to our booth at the SPS in Nuremberg from November 23 to 25, 2021, where you can network with others in person, check out exhibits, and find solutions to problems. All of this will take place, of course, in line with a safety and hygiene concept that complies with the latest regulations.
If you can’t attend in person, simply participate online and take a virtual tour of our 3D showroom, attend lectures and presentations, and talk to our experts.
Typical petrochemical plant
During daily operations, availability is the key requirement for Sinopec. “Our refinery is a typical petrochemical plant that uses a single-series process, so our equipment needs to run with high reliability and a low failure rate for a long production period,” says Sinopec engineer Liu Yan Chang.
High plant availability
“To further improve availability, we’ve worked with Siemens to apply advanced process control (APC) methods using model-predictive control in the sulfur recovery unit, which has helped cut fuel gas consumption there, and it also helps streamline asset management and alarm handling,” he explains.
The results have been impressive: The daily alarm volume has been cut by 80 percent and the improved condition monitoring provided by Simatic PCS 7 facilitates the early detection of equipment damage and abnormal situations. “This preventive maintenance approach helps us reduce maintenance costs and avoid risks and unplanned facility downtime,” says Liu.
New system sections smoothly incorporated
Liu also highlights the reduced need for spare parts in addition to reliable performance and low failure rates. The system’s well-structured interfaces and excellent extensibility also made it easier to incorporate new units. On top of that, Siemens provided high-quality service.
“The preventive maintenance approach helps us reduce maintenance costs and avoid risks and unplanned facility downtime”.Liu Yan Chang, engineer at Sinopec
The refinery has been operating for the last ten years with no major incidents, and the redundant system design of the DCS combined with rugged components and advanced functions for asset management, diagnostics, and preventative maintenance is recognized by Sinopec as a key factor. And the plant isn’t just operating safely, it’s also more secure thanks to a defense-in-depth industrial security solution that provides protection from potential threats and attacks from outside.
New digital concept
“In the last ten years, we’ve upgraded our systems in several areas, including cybersecurity, engineering, and operation,” says Chen Xin, I&A Manager at Qingdao refinery. “In the next step, we’ll be introducing the latest digital concepts into our facility.”
Siemens will support Sinopec with an integrated engineering and operation platform that combines its Simatic PCS 7 process control system with Simit simulation software for operator training (OTS) and restoring historic data; other components include Comos an integrated engineering system for greenfield plants and integrated operation for brownfield plant assets; and Comos Walkinside for 3D visualization.
From big data to smart data
Most recently, Siemens has deployed an innovative solution for big data analysis to monitor equipment condition and support data-driven predictive maintenance for critical assets.
“In the last ten years, we’ve upgraded our systems in several areas. In the next step, we’ll be introducing the latest digital concepts into our facility.”Chen Xin, I&A Manager at Qingdao refinery
Machine learning using AI technologies
This new concept based on Siemens Predictive Analytics (SIEPA) joins plant and equipment data with the latest data analysis capabilities. Here, advanced features like machine learning and visual analytics are used.
Siemens Predictive Analytics is a key factor to become a smart plant and the pioneer of real usage of AI technologies in the process industry. As required by Sinopec, SIEPA (formerly EPA V2.0) is now an official module for Asset Management and part of Sinopec’s "Intelligent Plant Project 2.0" for the Qingdao refinery.
The application consists of two major functionalities: Risk prediction/pre-alert based on machine learning and advanced diagnosis based on natural language processing. Integrating SIEPA currently helps Sinopec increase efficiency and reliability for their operation as well as improve intelligence level of the Sinopec plant continuously.
Furthermore, this allows Sinopec central access to all relevant data on a piece of equipment or a plant section and can also utilize the expertise and experience of its operators in the Qingdao refinery through a flexible and intuitive interface. Based on historical data and correlations between sensors, the app can generate equipment or plant section DNA that’s saved in a Status Reference Library.
This data is processed into predictive models for anomaly identification and risk assessment that provide insights into plant status and potential issues. The app is used to improve availability by preventing unplanned downtime.
Qingdao Refining & Chemical Co., Ltd, owned by Sinopec, is one of China’s largest refinery plants. In operation since 2008, the plant covers an area of 2.65 square kilometers, has 22 sets of processing units, and can process 12 million tons of crude oil per year.
The Simatic PCS 7 process control system integrates all of the plant’s automation systems – from field devices, operational systems, and safety instrumented systems through plant information management to long-term archiving servers – in a single platform.
Subscribe to our Newsletter
Stay up to date at all times: everything you need to know about electrification, automation, and digitalization.