Optimize Your Loops

Control-loop performance monitoring can be applied throughout an organization’s operational assets to realize high-level efficiencies.

By Robert Rice and Bryan King, Control Station

Process manufacturers understand the value of well-behaved control loops because these lead to high-quality products, minimize energy usage and equipment wear, and generally make life easier for operations personnel. It stands to reason that any loops causing trouble in these areas must be dealt with swiftly as performance levels have already been diminished. As the adage goes, the squeaky wheel—or in this case maybe the squeaky modulating valve—gets the grease, but wouldn’t it be better if squeaking could be avoided altogether?

While rapid response to poor control is positive and will provide some improved outcomes, many users are realizing it is nonetheless reactionary in nature, and that a more proactive effort could yield greater benefits. Large companies, and even smaller companies with many similar equipment assets, are exploring how they can move beyond a troubleshooting methodology for loop tuning with the help of control loop performance monitoring (CLPM) software. CLPM can advance a manufacturer’s optimization effort toward a plant- or organization-wide evaluation of all assets. They can improve efficiency of their highest-value and most problematic loops, while also reaching for savings across the entire range of assets.

Beyond one loop at a time

The activity of optimizing a single PID controller is itself an endless loop because processes and equipment are constantly changing. Users gather data, analyze it, adjust, and repeat to their satisfaction. While data gathering and analysis could be performed manually, CLPM software makes the task orders of magnitude easier, while providing a variety of solutions and documented results. 

Optimization is often initially focused on difficult loops that have a large impact on cost, and for those cases where there are strong interactions with other loops that can lead to even greater cascaded issues. While some organizations may encourage top-down loop-tuning initiatives, it’s much more common for a bottom-up approach, where difficult loops are addressed on an as-needed basis.

When an operation has multiple processing trains, or a quantity of skids or units with similarities, it is often possible to copy/paste tuning values from one asset to the next and obtain reasonable results. However, sometimes there are conditions, such as mechanical differences, equipment fouling, or other influences, that may cause one or more loops to underperform. One key to recognizing these issues is to use CLPM software that is capable of continuously monitoring control loops, as well as understanding when the PID controller might be in an idle state.

Manually tuning individual PID control loops is giving way to continuous, plant- or enterprise-wide process optimization. PlantESP’s TuneVue automatically isolates output changes, calculates process models, and recommends tuning parameters. Courtesy Control Station

Next-level optimization

Basic CLPM software already has some degree of data connectivity and analytics built into it to provide typical loop-tuning functionality. However, a platform designed to operate at the enterprise level for all plant assets or spanning multiple sites needs more capabilities.

One important example is the means to leverage enterprise data services. The software must be able to readily access process historians, databases, and similar resources, wherever they are hosted, and use many different protocols. 

Next, the platform needs to be structured with deployment and scalability in mind. The software architecture becomes one of an analytical engine—instead of a standalone application—capable of accessing data sources and integrating with other reporting dashboards. Although it is used as an operational technology (OT) application, it needs to be IT-friendly so it can be deployed on-premises and/or in the cloud and associated with appropriate computing and storage resources. A clearly documented REST API, with characteristics specific for control-loop monitoring, is one of the best ways to ensure CLPM software is able to integrate with the enterprise.

There are many reporting and visualization options used by companies to uncover insights. The Microsoft Power BI platform is one popular example, but an open CLPM platform lets users apply familiar tools to perform the evaluation. Higher-level analysis also lets users incorporate more business context, such as costing and quality, or other associated outside influences, such as weather or raw-material characteristics.

Finally, the platform requires provisions for benchmarking individual loops, units, areas, and plants so that the performance of like assets can be effectively compared. Benchmarks can be specific to timeframes and can be reset after planned maintenance or interventions, so that limits are kept fresh and realistic.

Enterprise-grade CLPM software typically yields these benefits:

• centralized access
• consistent metrics
• performance monitoring.

Centralized access involves two perspectives. First, even though data is sourced from many repositories and results will be applied in many locations, it’s good to keep all the information at a single site for best processing efficiency and management. This site could be a cloud resource, with appropriate data backups and redundancy.

Second, from a user standpoint, any credentialed user should have access to the information wherever they are located. This is especially important when an organization operates multiple sites.

Consistent metrics are the only way to ensure users and evaluators are comparing like conditions. For loop-tuning purposes, even slight differences in data gathering or processing can introduce significant differences. Identical equipment installed at multiple sites could have varying upstream/downstream influences, or maybe the personnel at each site evaluate performance using different methods. A centralized system with consistent metrics is necessary to normalize loop optimization on an enterprise-wide scale. Or, if companies are performing comparative evaluations of new instrumentation or equipment for a project, the availability of identical metrics and calculations gives the results a feel of impartiality and could lend credibility to successful implementations. 

Performance monitoring verifies the results of loop tuning, in a real-time manner. CLPM software, applied at the local or enterprise level, allows users to develop optimized loop tuning, but goes far beyond that in many ways. Continuous performance monitoring is the only way to ensure that tuned loops are living up to their potential. It also can be used to indicate a decay in performance or to help uncover issues before they escalate into major problems.

With the help of centralized access and consistent metrics, company personnel can fully take advantage of performance monitoring and they can pursue optimization efforts together using a common basis. Deployments of CLPM spanning dozens or more unique sites across numerous countries are increasingly common among large, multinational manufacturers. Beyond maintaining well-tuned PID loops, these manufacturers are leveraging CLPM to avoid equipment failures and to squeeze out manual operations so that their multi-million-dollar automation investments can deliver maximum ROI. One such user from the food & beverage sector cited a 6% increase in throughput in its milling operations, combined with a 14% to 23% increase in steam efficiency across evaporator units.

Asset optimization is a continuous journey. CLPM software has become an essential tool for companies to monitor and optimize process control loops. Individual loop tuning, especially for the most critical applications, provides immediate payback in terms of improved product quality, minimized equipment wear, reduced energy costs, and an overall streamlining and simplification of operations. EP

Bryan King is Senior Application Engineer at Control Station, Manchester, CT (controlstation.com), where he is responsible for the deployment, use, and support of the company’s portfolio of process diagnostic and optimization solutions. Robert Rice, PhD, is Vice President of Engineering at Control Station. He has published extensively on topics associated with automatic process control, including multi-variable process control and model predictive control.