Real-time control system unlocks real-time optimization

Process optimization covers a lot of ground these days. There is a growing range of tools and techniques for optimizing labor, materials, production, the supply chain and perspectives, as well as many other things. But how can anyone bring it all together?

By Stan DeVries and Maryanne Steidinger, Invensys Operations Management September 23, 2010

Second of Two Parts. Read Part 1 here.

Process optimization covers a lot of ground these days. There is a growing range of tools and techniques for optimizing labor, materials, production, the supply chain and perspectives, as well as many other things. Data collected and monitored within each function contributes to the overall tuning and optimization of the plant, and the accelerating speed of business makes it increasingly critical to view all optimization solutions as parts of an integrated whole rather than as stand-alone point solution.

But how can anyone bring it all together? Optimizing any single function is challenging enough, how can you possibly approach it at an enterprise level? The short answer is that you shouldn’t start any optimization initiative unless you can justify it within the context of your company’s business strategy at its highest level. But there is much more to it than that.

Real-time Operations Control

In the control function, there is almost always an automation mechanism, which may even have been the result of a previous optimization cycle. This is not necessarily the type of process control often associated with the process industries, i.e. combined regulatory control, batch and sequential control, and multivariable control. It is still control, but often is implemented in forms such as:

•              Prevention of unplanned shutdowns by augmenting safety systems to incorporate site-wide techniques to absorb energy and mass

•              Prevention of unplanned shutdowns by stabilizing steam headers, by counteracting major reductions in steam use, and balancing letdown flows

•              Increased automation of unit and machinery start-up sequences to minimize restarts and lost energy and production limits

Real-time control provides operational controls in the context of overall performance, including the impact on business measures. This is a more holistic approach, and considers trade-offs between individual decisions to make the right decisions for the overall operation, in financial terms and within the proper timeframe.

Site-wide stabilization is achieved beyond traditional unit advanced process control using techniques, which contain energy and mass transfer to help prevent most unplanned shutdowns. Other techniques might include using formula management to adjust key equipment operating rates to support changes in product characteristics and site targets, and steam distribution stabilization by countering the tendency to starve high-pressure steam and flood low-pressure steam users.

Real-time Operations Optimization

A careful look at whether or not you are achieving your metrics will suggest whether and what you need to optimize. If you are making your numbers, it could mean that your empowerment and control systems are effective. If so, this may be the time to raise the bar, set new metrics on yield, quality, sustainability or whatever else might be at the center of your business strategy. Then, just as you would do if you were not achieving the measures – and you were comfortable that you would have the right measures – you revisit your empowerment and control systems to see what might be done to help hit the targets. Interventions range from high end quantitatively driven metrics to operator training.

This optimization considers the key business measures that have been identified at the measurement phase. The goal is to improve business performance, not just operating efficiency. It is still optimization, but often implemented in such forms as:

• Maintenance-production schedule balancing

• Short-term vs. long term maintenance schedule balancing

• Energy and quality optimization (balancing with production)

• Guiding a fleet of like operations through demand changes for maximum efficiency

• Balancing production with the availability of the asset to optimize the overall business value

A key element of this optimization technique is that it embraces the dynamic nature of true decision-making and takes into consideration the balancing of asset availability and asset utilization to drive business value. When approached in a more holistic manner, better business value can be achieved.

Decisions of managing more critical assets (such as energy conversion, rotating equipment, instrumentation, and reactors) enable tradeoffs between current commitments, longer-term major maintenance activities, and production scheduling. Using heat exchanger fouling to forecast energy transfer efficiency and thus justify changing the maintenance schedule is a good example.

Such decisions, however, are also balanced by information about supply chain opportunity costs and real-time risk assessment. This approach uses real-time calculations of supply chain impact, maintenance impact, and operations risk impact. The same approach is used for reactor catalyst activity, motor efficiency, pump capacity, and seal life. T he optimization cycle extends to process engineering to evaluate what-if scenarios; training sessions to practice for upcoming transitions; maintenance schedule change scenarios; and production.

The measurement system must first be put into place to make the optimum decisions for the business and be able to determine the value that is being created. This approach focuses on the dynamic business environment and includes the rapid changes in demand and supply, and offers the decision information to exploit these.

The optimization process begins at any point of the strategic alignment chain that executives, managers, operators or others with control over their metrics determine that they are not meeting their metrics. Intervention then, is whatever it takes to meet those metrics. As the following examples illustrate, optimization can, but does not always require, a calculus based engine.

Optimization through improved workflow

A large aerospace manufacturer had 10 manufacturing units spread over multiple locations, together with R&D units supporting them. Tracking the supply chain at a central point for increased efficiency was identified as a concern area to meet the organization’s metrics for prompt deliveries and competitive pricing. The company examined existing processes and move to the next level through collaboration with global aeronautics parts manufacturers for mutual.

The organization called upon BPM solution providers to technically architect expert reports for a collaborative BPM platform that integrated the various supply chains for process monitoring through the entire lifecycle of the event. Skelta business process management software from Invensys) provided a collaborative platform that connected with external vendor systems to enable unified reporting.

By automating administrative tasks associated with process flow, the company optimized resource utilization, thus freeing human capital for strategy related and other people centric functions. The agility resulting from this collaborative platform reduced decision time by 30% and costs by 35%.

Optimization through mobility solutions

Empowerment intervention does not always need to be completely automated. Shell Oil for example, empowers maintenance workers with a handheld computer or PDA running software tools and applications that process and transmit device-specific information to data historians, operators or enterprise decision-support systems.

This enables workflow, procedural and general task-management capabilities typically involving plant operations, maintenance management, production tracking and compliance applications. In Chevron, when some maintenance workers start their rounds they might grab a PDA instead of a clipboard. As they check equipment, they enter data which is transmitted into process historian systems when the PDA is docked in a cradle, from which the Wonderware mobile workforce and decision support system from Invensys synchs the data with real-time process data from the control systems for viewing by operators and engineering.

If a piece of equipment is not operating correctly or efficiently, field workers can collaborate with other personnel to take the optimal corrective action then and there.

Model predictive control

For the NRG Huntley Power Station in Tonawonda, NY, for example, success metrics were heat rate and dispatch rate, which they wanted to reduce while at the same time lowering NOx emissions in compliance with regulatory requirements. A key unit to optimize was their boiler system. To do this they installed a SimSci-Esscor model predictive and neural network based controller from Invensys.

Models of the important variables for its twin furnaces were derived from a brief testing conducted in the initial phase of the project. The twin furnace design with physically separate furnaces for the superheat and reheat functions increase the number of controlled and manipulated variables and complicates the control through the highly interactive effects. Through the models designed within the controller, the effects of load changes on the process are anticipated, providing a smooth response to the optimum operating points. Implementation of the model predictive controller allows the units to operate at reduced excess O2 rates, which not only decreases NOx emissions, but also helps improve heat rate by reducing dry gas losses. The system contributed to lowered NOx emissions by more than 10% while improving heat rate by more than 0.5% and increasing the dispatch rate by a factor of 2.5.

Unlocking the optimization potential

Real-time process automation is as feasible today as it is necessary. Thanks to continually decreasing software costs and increasingly easier access to open, powerful software, technology is not a barrier. Applying the measurement, empowerment, control and optimization concepts we have been discussing helps marshal advanced technology on behalf of enterprise optimization, but it does raise new organization and cultural challenges.

Those responsible for defining the measures, implementing empowerment and designing control strategies are often distributed throughout multiple islands of organization, each with its own definition of optimization or business improvement. Surely improved collaboration tools and strategies, will make it easier for people to communicate and collaborate across various segments of the enterprise and integration standards like ISA-95 will provide some guidance in structuring the flow of enterprise information, but in any company, the move to enterprise optimization is much more evolutionary than revolution.

It will take the emergence of champions at every level of the company, people who see how it can benefit their sphere of operations and those that they touch. You can begin the process by asking yourself three questions: “What is my company’s business strategy?” “What metrics define successful execution?” and “What is my role in making that happen?” Then implementation can use an effective vocabulary which separates the broader meanings of measurement, empowerment, control and optimization.

Culture change will often lead asking questions about how to pilot and roll out the change, such as piloting the full implementation of techniques for a smaller area of the business, instead of piloting and rolling out one technique or technology across multiple areas. Your answers to those questions will set you – and hopefully your company — on a path to successful, sustainable enterprise optimization.