Abstract: With the development of artificial intelligence, it is increasingly being applied in factories to achieve automated control, reducing heavy manual labor and bringing considerable economic benefits. This paper introduces the successful application of iMES control in the coking gas compressor control system, which achieves precise control of the fractionation pressure in the coking tower by adjusting the load of the rich gas compressor.
Introduction: Coking tower cutting is a challenging aspect of coking process operation, significantly impacting the coking process, dry gas hydrogen production, and even the safety and environmental protection of the entire plant. The rich gas compressor unit in the coking plant compresses the rich gas produced by the unit and sends it to the absorption stabilization system. During tower cutting, process fluctuations are significant, requiring operators to manually operate for two to three hours to restore stability. This is labor-intensive, risky, and wasteful of steam. Jinan Refining & Chemical Co., Ltd., as the first "city refinery" in China, considers safety and environmental protection its lifeline. Independent innovation and leveraging technological advantages to support safe production are imperative. In 2017, Jinan Refining & Chemical Co., Ltd. implemented a fully automated tower cutting system for its coking gas compressor unit using the iMES control system, achieving undisturbed operation throughout the process. This not only frees up labor but also generates substantial economic benefits.
I. Existing Control Methods and Problems in Coking
The 1.2 million tons/year delayed coking unit, an expansion of the existing facility completed in 2009, plays a crucial role in the plant's residue-oil balance. It experiences significant fluctuations in processing load and frequent process variations. The rich gas compressor unit compresses the rich gas produced by the unit and sends it to the absorption stabilization zone. Due to the large fluctuations in operating conditions, the original unit's speed governor control system could not meet the requirements for automatic control, requiring manual operation by personnel, resulting in high labor intensity, high risk, and energy waste. To achieve automated operation of the unit, the control system underwent optimization and upgrades in 2009. A control system from a certain international brand was used to specifically modify the unit's anti-surge control and speed regulation components. After several years of operation following the upgrade, due to insufficient technical support and poor hardware reliability of the control system, the control quality deteriorated. The anti-surge control and speed regulation system could no longer meet the requirements for maintaining automatic control under changing process loads and operating conditions, posing a potential safety hazard. Jinan Petrochemical is an "urban refinery" facing immense environmental pressures. Its coking units cannot be automatically controlled, requiring several operators to coordinate manual operations during venting and tower switching, resulting in significant operational intensity and risk. Furthermore, the manual operation of the air compressor units is incompatible with the requirements of advanced control (APC) software.
II. Drawbacks of the current control methods
Distillation column switching is a challenging aspect of the coking process, as the pressure in the distillation column fluctuates significantly with each switching operation. Operators maintain pressure stability by adjusting the opening of the anti-surge valve, inevitably resulting in a large margin of error for the compressor; typically, the anti-surge valve opening is around 30%. This valve opening leads to substantial steam waste. The switching process requires frequent manual operation by multiple operators, resulting in significant mental stress due to the high level of concentration required. Furthermore, manual switching causes substantial pressure fluctuations in the distillation column, as shown in Figure 1.
Figure 1. Operating curves of the iMES control system before its commissioning
As can be seen from the curves, during coking tower switching, the turbine speed remains high, the anti-surge valve opening is large, and the pressure at the top of the fractionation tower (green curve) fluctuates significantly. Operators frequently adjust the anti-surge valve to stabilize the pressure, consuming considerable effort. Furthermore, due to the large opening of the anti-surge valve, a large amount of compressed rich gas is backflowed, resulting in wasted compression work and steam loss.
III. Improvement of Control Scheme
The core of the iMES control system for the gas compressor unit is the cascade control of fractionation tower pressure and turbine speed. It adds decoupled optimization control functions for fractionation tower top pressure control and anti-surge. These two functions are independent yet interconnected, allowing for both individual control and synchronous automatic adjustment. When both functions are in automatic mode, coupling between surge control and fractionation tower top pressure control is prevented. When the anti-surge valve must open, the iMES control system automatically tracks and adjusts the turbine speed to prevent large fluctuations in fractionation tower top pressure. Because the unit's operating point is always near the operating curve, the unit operates at its highest efficiency, resulting in significant steam savings and greatly improved economic benefits.
3.1 Hardware Configuration
The system hardware uses TRICON series products, and the power supply and control cards are designed with redundancy and fault tolerance. The key control parameters are interlocked with a 2-out-of-3 ratio, resulting in high system reliability and fast processing speed.
3.2 Software Functions
The core of iMES control is the top pressure of the distillation tower and the speed of the cascade steam turbine. The anti-surge valve, in conjunction with the decoupled control of the turbine speed, enables "edge-locking" operation of the turbine, keeping it near its operating curve, as shown in Figure 2. This maximizes efficiency while preventing turbine surge through anti-surge valve control, achieving safe, stable, and economical operation. Furthermore, the system's computational operations allow artificial intelligence to liberate operators, achieving unmanned operation.
Figure 2. Anti-surge line for steam turbine
The intelligent control process for downsizing column pressure control in iMES is as follows:
1. Determine if the anti-surge operating point is in the safe zone. If it is, close the anti-surge valve first. When the anti-surge valve is fully automatic, it can be closed directly until the pressure at the top of the distillation column is met. If it is closed to a certain extent and the pressure at the top of the distillation column is met, and the operating point is still far from the control line, then start reducing the turbine speed at a rate of 60 RPM. At this time, the reflux valve is still in automatic adjustment mode. Start slowly closing the reflux valve until the reflux valve is completely closed or the speed is reduced to the minimum.
2. Determine if the anti-surge operating point is within the safe zone. If it is, first close the anti-surge valve. If the return valve is completely closed and the set value has not yet been reached, start increasing the speed until the set value is reached.
3. Determine if the anti-surge operating point is in the safe zone. If it is in the safe zone, first close the anti-surge valve. If the return valve is in semi-automatic mode, close it to the semi-automatic setting value. If the setting value has not yet been reached, start increasing the speed until the setting value is reached.
4. Determine if the anti-surge operating point is within the safe zone. If it is, close the anti-surge valve. If the anti-surge operating point is within 1% of the control line but the set value has not yet been reached, start increasing the rotational speed. If, during speed increase, the operating point reaches more than 3% of the control line but the set value has not yet been reached, stop increasing the rotational speed and continue closing the anti-surge valve. When the operating point is between 1% and 3% of the control line and the set value is reached, maintain the current state and adjust the anti-surge valve to regulate the distillation column top pressure. Otherwise, continue the previous steps until the set value is reached. The control curve is shown in Figure 3.
Figure 3. Parameter curves under iMES intelligent software control
The pressure control process for ascending fractionation towers is similar to that for descending fractionation towers.
As can be seen from the control curves, after implementing the iMES intelligent control software, the control curves changed significantly. The entire control process requires no manual intervention. The process parameters for venting and tower switching cycles smoothly track changes in control requirements, and the fluctuation range of the fractionation tower top pressure (green curve) is within 3 kPa. The turbine speed (yellow curve) automatically adjusts in response to changes in the fractionation tower top pressure. The operation of the gas compressor unit is fully automated, truly achieving "unmanned operation" of the unit control.
Functionally, this system optimizes the decoupling of distillation column top pressure control and anti-surge control. When both functions are in automatic control mode, it prevents coupling between anti-surge control and distillation column top pressure control. When the anti-surge valve must open, the unit speed automatically tracks to prevent large fluctuations in the distillation column top pressure. The decoupling control employs differential control; operators can set a dead zone based on the pressure fluctuation range at the distillation column top. Adjustment mode is only activated when the measured value exceeds the deviation from the set value and this dead zone. Adjustment automatically stops when the value is adjusted within the set deviation range, preventing frequent speed adjustments to the surge valve. The specific implementation principle of the decoupling function between distillation column top pressure control and anti-surge is shown in Figure 4.
Figure 4. Turbine speed versus anti-surge valve decoupling control curve
IV. Operational Performance Analysis
The APC software at Jinan Petrochemical Coking Plant went online at the end of 2016, using intelligent adjustment functions to optimize the material transfer balance between various parts of the unit; it also optimized the operation of the distillation section, achieving an average increase in light gas yield of 1.7% and a reduction in energy consumption of 1.8%. However, before the unit control optimization, the APC could not seamlessly connect with the rich gas compressor unit and the top pressure of the fractionation tower, creating a bottleneck for further energy saving and consumption reduction.
With the commissioning of the iMES (Intelligent Management System) for the optimized control of the rich gas compressor, precise control of the top pressure of the fractionation tower can be achieved, which can further reduce the coke yield, increase the light gas yield, and increase the wax oil yield, thus truly realizing intelligent operation.
The actual efficiency gains from this renovation are:
3.36 million (steam saving) + 500,000 (reduced air cooling load) totaling 3.86 million yuan/year.
V. Conclusion
The intelligent software of the Jinan Refining & Chemical iMES control system is identical to the black-screen operation of Yanshan Petrochemical. The application of intelligent software has a profound impact on control methods, operational efficiency, industrial structure, and even safety and environmental protection. It is believed that with the development of computer technology, instrumentation, and artificial intelligence, the application of these high technologies will bring about entirely new changes to the hazardous chemical industry. New intelligent and automated production enterprises will no longer be limited by the geographical location of the factory. Safety, environmental protection, energy conservation, and high efficiency will all be achieved under the impetus of artificial intelligence. The trend of fully automated, driverless operation replacing manual labor in industrial production is inevitable.
