InTech

MAY-JUN 2018

Issue link: http://intechdigitalxp.isa.org/i/989516

Contents of this Issue

Navigation

Page 16 of 53

PROCESS AUTOMATION INTECH MAY/JUNE 2018 17 fied to some extent by using the big data stored in distributed control systems (DCSs) and plant historians to create better KPIs, but this requires careful analysis to produce the desired results. What if KPIs could be structured systemati- cally across an organization from management through engineering to operations to unify all the employees' capabilities, motivate them to do better, and cultivate the knowledge that is in the DNA of the organization? This can be done, but it requires alignment of KPIs among all plant personnel, starting with optimization of related process variables in operations. Creating systematic SPIs Domain knowledge and the big data found in the plant DCS can help process industry companies select, monitor, and optimize key process vari- ables and set points for operations, such as fuel gas consumption, furnace efficiency, and main fractionator pressure. Given the right informa- tion, control room operators can accomplish these tasks by interacting with the DCS. Success- ful implementation will not only result in safe and stable operation, but also improve profit- ability, energy conservation, and asset reliability. Process variables are optimized by setting an optimal range for each critical set point, which can be used to systematically structure a plant's KPIs and create an improved set of metrics, henceforth referred to as synaptic performance indicators (SPIs). Process historians are electronic databases, typically used to store and display data. In a pro- cess plant, common data points include temperatures, flow rates, pressures, levels, and other types of an- alog data. Increasingly common is the "his- torization" of digital data, such as the out- put or feedback states associated with valves, pumps, and other dis- crete control devices. Analysis of histori- cal data by process engineers and experts with extensive domain knowledge can deter- mine the correct range for each set point, a complex task because many control loops interact with each other. Further com- plexity is added by also considering the effects of set points on the supply chain. All this data can be combined to provide deep insight into the behavior of a process at any point in time. These insights can be used to optimize set points for improved operations. SPIs for each area can be defined by using domain knowledge and historian information. SPIs help operators set an ideal range for each variable with an alerting function. They provide expert guidance in the form of messages when a variable goes outside of the ideal range. Other functions in a DCS can guide operators to opti- mize related process variables for more profit- able and reliable production. These key process variables are systematically connected to engineering and C-level KPIs, so C-level personnel can drill down on KPI issues related to engineering or operations. Operations can also contribute to C-level KPIs. SPIs in action At the operations level, a typical midsized re- finery has about 377 SPIs, and a typical ethyl- ene plant has about 77 SPIs. Approximately 60 percent of SPIs indirectly correspond to control variables, and 40 percent directly correspond to FAST FORWARD l KPIs are often not aligned, producing suboptimal process plant performance. l Advanced process control metrics referred to as synaptic performance indicators can be used to align KPIs. l The result is improved performance, primarily driven by control room operators supplied with SPI information and suggested actions. SPIs based on domain knowledge and DCS big data Drill down on SPI issues related to engineering or operations Contribute to management SPIs Synaptic performance indicators (SPIs) unify vertical organizations 93 Daily SPI 357 Hourly SPI Production cost Margin Energy cost Maintenance cost Yield or Quality Optimizer performance Energy consumption Availability ESD triggers Incidents Production Profit Reliability Energy Safety Control performance Process performance Furnace efficiency Asset performance Critical alarm 377 Real-time SPI Operations Management Engineering 12 Daily SPI 32 Hourly SPI 77 Real-time SPI Refinery Ethylene on DCS on cloud on cloud Source: Yokogawa Figure 1. The term synaptic performance indicator (SPI) identifies an improved set of key performance indicators that can be used to optimize plant operations.

Articles in this issue

Archives of this issue

view archives of InTech - MAY-JUN 2018