Comment

1 July 2006

Taking it to the boardroom

Use performance supervision information for higher-level management decisions

By Chris McAnarney and George Buckbee

Columbian Chemicals Corporation hit the automation trifecta. The company has reduced energy costs, improved product quality, and increased production using a performance supervision system. 

By sharing real-time plant performance data across all levels of the organization, Columbian has gained insight into process, equipment, and controls performance. Plants have benefited greatly by tapping into company experts. 

At plants worldwide, daily operations and maintenance tasks are now closely coordinated with corporate plans. Capital project planning and production scheduling is closely tied to actual plant performance data.

Columbian’s critical approach to performance supervision included:

• Gathering dynamic performance data
• Sharing corporate data worldwide
• Coordinating plant- and corporate-level decisions

In 2004, as energy prices rose, Columbian Chemicals also felt increasing pressures to meet ever-tightening quality standards. Columbian set out to find new ways to reduce product variability and save on energy costs at plants worldwide. 

The company found a performance supervision system that not only reduced variability, but also delivered savings in energy, increased production rates, and streamlined communications from the plant floor to the corporate boardroom.

Columbian is a global provider of high-quality carbon black additives for rubber, plastic, and liquid products. Columbian’s customers demand each grade of carbon black produced meets high standards for quality. 

This includes meeting quality standards for structure, particle size, surface activity, and porosity. To achieve these quality requirements, the process, equipment, and controls must all perform at their best.

Seeing the bigger picture

In 2004, Columbian Chemicals initiated a six-sigma project to find ways to reduce variability. Efforts quickly focused on control loop performance monitoring/performance supervision systems. 

Performance supervision systems monitor the plant, 24 hours a day, assessing the performance of every control loop. The system then finds and prioritizes problems, focusing efforts where the largest economic impact can happen.

Michael Kennedy, global manager of process control engineering, was the leader of this project. Kennedy said, “We researched many different organizations and companies that produced similar products. We eventually decided on Plant Triage.”

For Columbian, some of the key measures of performance included:

• Variability and variance
• Time in normal mode
• Oscillation, period, and magnitude 
• Robustness of controller tuning

By combining these measures with economic factors, the system formed a composite number, referred to as loop health. This helped to focus attention on those control loops with the greatest potential to improve the process performance.

A short trial period in one plant confirmed the expected results, and Columbian started rolling out performance supervision to several more plants.

The leadership team at Columbian realized there could be great benefit to gaining remote access to the data from all plants worldwide. Working with the corporate IT department, the process-control engineering group established a secure network to share live plant data via the performance supervision systems, into a central control room at headquarters in Marietta, Ga.

Using the new corporate control room as a base of operations, corporate process experts were able to see and interact with real-time plant information for the first time. 

Columbian began addressing the most pressing loop health issues at each plant. They used the key performance indicators to triage, or prioritize, which loops required attention first. 

Using this triage process, plant performance came across in a new way. According to Kennedy, “Because of the information available on a normal DCS system, you are just unable to push these things, to bring them to the surface, to look at them, to analyze them and examine them, and to make the proper changes on them; it’s just impossible. But with this system, as well as our operations center in Atlanta, we were able to get a bird’s eye view of the entire process, make the changes necessary to improve, and overall just had a totally different concept than even we had expected at the very beginning.”

Coordinating activities 

Of course, finding the biggest opportunities is only the first part of the problem. Resolving process and control problems from a remote location was breaking new ground.

Members of the plant and corporate teams met face-to-face and developed a work process to ensure safe, yet efficient ways to optimize the process remotely. Fundamentally, the actual controller makes changes in the plant itself at the behest of the leader of the process-control optimization team who is at the remote location where all data assembles.

With this work process in place, corporate experts can work hand-in-hand with plant personnel to identify issues and recommend improvements. These remote optimization efforts include equipment repairs, process operational adjustments, and controller tuning.

Collaborating remotely was challenging at first. It took a little while to build up trust in the process and trust in each other. 

Russell Webb, general manager of Columbian’s North Bend Plant said, “When we first heard about remote access/ remote tuning, I think there was some fear on some people’s part, out in the plant. But what we’ve been able to see is it allows us to get additional expertise from the outside, to monitor plant performance, and bring additional resources to bear on plant problems.”

Now, it is routine for the plant and corporate personnel to work together to address plant performance issues on a daily basis.

A recent example of this came when a dryer in England showed a sudden increase in demand for natural gas. From his office in the U.S., McAnarney was able to use data from the performance supervision system to immediately identify the root cause of the upset. Coordinating his efforts with plant personnel, they were able to resolve the problem and put a permanent fix in place. A problem, which could have lingered for days or weeks, was resolved in a matter of hours. 

Measuring results

Columbian personnel started with an initial focus on variability reduction. Some variation was cyclical in nature, due to interactions between controllers. 

Variability reduction may seem an abstract way to reduce costs. However, it is quite common in energy intensive processes to see efficiency improvements when you reduce oscillation and variation. 

In many ways, it is similar to the way the fuel efficiency of your car suffers if you drive erratically. By stabilizing the plant, energy usage has dropped significantly. 

In many cases, process improvements can translate directly to the bottom line. Improvements to temperature control that in turn drove a reduction in energy usage and an increase in product throughput successfully transferred from one plant’s operations to another’s.

Webb confirmed the impact on the company’s bottom line: “The system has really helped us, improving the performance of our plant. It has helped us to zero in on areas where we can run more efficiently and optimize the overall process. It has helped us in reducing natural gas usage on the dryers.” He said, “We’re really utilizing it to fine-tune the operation, and also to zero-in on problem areas and help us to take corrective actions to prevent them from happening again in the future.”

While they had expected the process to become more effective, Columbian personnel were happy to see other, unexpected benefits. The combination of an effective work process and remote access has streamlined process optimization work. 

“In the past, we may have been able to tune 3 or 4 loops in one day, after traveling to a plant. Now, we are tuning 10 to 20 loops in one day, with no travel at all,” McAnarney said.

Travel costs have dropped, and personnel are spending less time on airplanes and more time optimizing the process. And when asked how remote access has affected his personal life, McAnarney quips, “I get to be home every night.”

Information to management

When you save $1 million, it is easy to get the attention of management.  The management team recognized the opportunity to start using performance supervision information for higher-level management decisions. Management now uses plant performance data to make decisions for production planning and capital allocation.

The real-time data provides insight into process capability, efficiency, and utilization. According to McAnarney, this has sped-up the decision-making process. “I can supply the information to management now, and they can make a decision now.”

The results have allowed Columbian Chemicals to reduce energy costs, improve product quality, and increase production using a performance supervision system. 

At plants worldwide, daily operations and maintenance tasks are now closely coordinated with corporate plans. Capital project planning and production scheduling is closely tied to actual plant performance data.

ABOUT THE AUTHORS 

Chris McAnarney (wmcanarn@phelps dodge.com) is leader of process control optimization at Columbian Chemicals. George Buckbee (george.buckbee@ expertune.com) is a registered PE and works as director of product development at ExperTune, Inc.