Five Must-Have Manufacturing Metrics for Operational Excellence

Establishing appropriate metrics provides a foundation for continuous performance improvements.  Quarterly performance reports typically contain metrics such as unscheduled downtime, production-per-unit-cost, quality-per-unit, or energy consumed-per-unit-produced.  Unfortunately, these metrics are after-the-fact averages.  Using them can be compared to trying to drive an automobile while looking in the rear-view mirror. 

Benchmarks Provide Basis for Metrics for Operational Excellence

Using metrics based on overall averages to understand where you have been certainly has value.  The ARC Benchmarking Consortium was been providing the benchmarks that members have used for their Metrics for Operational Excellence for more than 10 years.  Real-time industrial manufacturing and production environments, operators, engineers, and other plant personnel must pay attention to metrics in real- or near-real time so that they can make preemptive corrections to support the ongoing journey to operational excellence.  

Effective automation of a plant or other production facility requires hundreds, if not thousands, of real-time measurements.  A few of these measurements also often become the basis for performance metrics.  The following five metric groups -- energy management, alarm management, operator loading, control performance, and safety -- are important in the real-time world of industrial operations.  Many of the individual metrics within these five groups depend heavily on real-time data collected from the plant automation systems.  The metrics can be used internally to compare different production sites within an organization, and/or to benchmark against peer companies to support continuous improvement.

Energy Management Metrics

Less than half the respondents to an ARC survey indicated that their operators have energy-related Key Performance Indicators (KPIs).  Furthermore, 53 percent indicated that just 20 percent (or less) of the required energy-related measurements are on line.  This indicates that far too many industrial companies do not have the necessary metrics in place to improve their energy management.

Important metrics in this group include the percent of energy-related processes that are under automatic control and the type of control deployed: manual, automatic, or some type of advanced process control (APC).  Of course, when comparing energy metrics across different plants, it is important to understand the complexity of the specific energy management situation.  An "energy footprint," or energy complexity index can be used to allow to benchmark different plants energy management programs.

Alarm Management Metrics

In theory, alarms inform operating personnel when something in the process is not correct and they need to take action.   In practice, if not designed and managed properly, alarms can actually be the source of problems.  The classic example is "alarm showers," during which so many alarms occur at once that the operator does not know what to do.  This typically results in lost production or out-of-specification product, or in the worst case, in environmental releases and even loss of life.   To manage alarms, a combination of real-time measurements and analysis is required. 

First, the real-time measures need to be represented in a family of metrics. Alarms-per-operator, peak-alarms-per-operator, and standing-alarms-per operator are key indicators that you can use to determine if the alarm settings in combination with control operator loading is correct.

The emergence of predictive analytics can help if the application can remove the root cause for the impending alarm, but if the analytics application is set up to only warn the operator of an event that will happen, it only adds to the potential of more alarms for the operator to handle.

Operator Loop Loading Metrics

How many control loops can an operator manage?  The answer is, "it depends." According to ARC's Benchmarking Consortium, while some process plants expect each of their operators to be able to manage as many as 400 control loops, the overall average in the process industries is about 140 control loops per operator.

This important metric must be viewed in context of your specific processes and operations.  If you are trying to push the loading of your operators to 400 loops and you are having incidents related to too many alarms per operator or too many unscheduled shut downs, then you need to adjust this goal to a more reasonable target.

Control Performance Metrics

If your controllers are not tuned correctly, or are in the wrong mode, it is difficult, if not impossible to achieve optimum process performance.  Metrics such as “percentage of time control loops are at limit”, or “percent of loops in incorrect mode” can help you determine if the operations have the correct fundamentals to reach your performance goals.

Improperly tuned controllers or controllers that have been put in manual when they should be in automatic can lead to disappointing results in the quarterly plant performance reports. 

In addition, many processes depend on APC to operate at or near optimum.  In a survey on the value of automation several years ago, 22 percent of respondents indicated they would lose revenue without APC, and 4.5 percent said they could not operate at all without APC.  End users have learned that peak performance of APC requires an on-going effort to maintain the controller models.  While a metric showing the APC service factor is helpful, it is even more important to have a metric that indicates if the APC is performing effectively. 

Safety Metrics

Unfortunately, many recent headlines focus on industrial safety, or lack thereof.  While most companies track safety-related metrics such as lost time accidents, many do not know if they are doing enough to prevent incidents.  The question is, "Are your safety business practices as good as or better than peers in your industry?"

Understandably, owner/operators are reluctant to share information about safety incidents beyond the severe ones regulations require them to report.  However, keeping metrics on near or minor incidents can help companies in their quest to achieve zero incidents.  According to the AICHE’s Center for Chemical Process Safety (CCPS), three important metrics are process safety incidents count (PSI), process safety incident rate (PSR), and process safety severity rate (PSSR).  Chemical, pharmaceutical, and petroleum companies can register and submit metrics on the site. 

Benchmarking Consortium Can Help You Know Where You Need to Focus

Obviously, you can use many different metrics to help you measure your progress toward your company's operational excellence goals.  The point here is that the measurement data for many performance metrics should be measured in real time and effectively presented to operating personnel and support staff in a timely manner to enable them to take appropriate corrective actions.

Sharing metrics internally within your company is better than not having metrics at all.  However, we feel that companies will benefit most when they share metrics results with their peers for benchmarking purposes.  Sharing fundamental metrics in these 5 must-have groups will allow each participant to know where they are starting from.  The real-time monitoring of the metric KPIs will allow them to continue the journey to Operational Excellence.

For more information on benchmarking, readers can visit https://www.arcweb.com/performance-benchmarking.