The most effective, least expensive and fastest way to apply green technology to plant systems is to get them under control and operating at maximum efficiency. This cannot be done without a well-thought-out, comprehensive, active monitoring program.
Information is routinely collected around equipment operations within a system. This can be collected automatically, by dataloggers attached to controllers, or manually, by operators. All too often though, these valuable bits of operating information are filed away and forgotten. This type of “passive monitoring” may be enough to keep a plant going from one turnaround to another, but it cannot point to the types of changes that can lead to improved process efficiency and resulting lower carbon footprint.
The alternative, “active monitoring,” requires someone to analyze the collected data and recommend actions that can improve operational efficiency. While there are a number of tools that can help with the analysis, first and foremost, a comprehensive monitoring program requires a person, or persons, familiar with the process. The personnel should determine which parameters to measure, how often they should be measured and the appropriate types of analysis to apply.
Asserting the value of active monitoring is not intended to diminish the importance of passive monitoring. Controllers can be critical to a plant’s ongoing operation and allow systems to be maintained within smaller tolerances than would otherwise be possible. However, they cannot point to the changes that can lead to areas of process improvement. Only an active monitoring program can do this.
Some of the considerations for developing such a monitoring program, as well as some of the analyses and analytical tools that can be used, will be discussed.
What Parameters Do You Measure?It is important to understand that there is a cost associated with every parameter that is included in a monitoring program. Specialized testing equipment may be required. Someone must take the reading and record it in a way that facilitates its analysis. And someone must perform that analysis. So, there should be a specific reason to acquire each measurement.
Some measurements are taken to ensure that processes are operating safely and within the values specified by the equipment manufacturers, treatment suppliers or regulatory bodies. Other measurements are taken to be used in interpretive calculations such as mass balances, costs and unit efficiencies.
Locked up in each set of measurements is a wealth of information about the process condition. It is a wide-angle snapshot of the entire system at the point in time in which the readings were taken. However, if you string these sets together, you can better understand the operation over time, note trends and apply statistical analyses. This information can help you spot areas where improvements can be made.
Available Tools for AnalysisHistorically, test results and instrument readings taken manually have been entered on paper log sheets and filed away after review. Readings also have been entered into spreadsheets, where it is easier to generate graphs. To make full use of a spreadsheet for the purposes of analysis, however, usually requires a degree of customization that is difficult for the average user to implement.
Starting in the early 1990s, software became available that makes it easier to configure the parameters involved in a specific monitoring program. Many had built-in tools for generating calculated results based on input, producing high-quality graphs, trends, various statistical analyses and relevant reports. Recently, some of these packages have moved online, where data can be accessed from any computer connected to the Internet. This makes it faster, easier and more convenient to review process data.
Choosing the right tool is important as it can help determine the level of success that a monitoring program will have. Ease of configuration and use, range of analytical capabilities, and access to technical support and training - all should be considered.
SPC: A Key to Process ImprovementThe use of calculated values, graphs and various statistical methods helps stabilize operations. However, one approach - statistical process control (SPC) - can point directly to areas where process improvement can be realized. Better efficiency means lower costs and, of course, a smaller carbon footprint. This is the basis of most process improvement programs introduced over the last 50 years.
Unfortunately, SPC is not well understood; therefore, it is not implemented as often as it could be. A summary of SPC concepts may help clarify the significance of the standard charts used to implement such a program.
Process Variations. No matter how tightly controlled and well run a process is, variations will be found in the quality of the resulting product, and in all of the intermediate operations that contribute to the product. It is virtually impossible to operate in a steady-state condition where quality results are constant.
Some variation is inherent in the process itself. A continuously recording device typically will show a sawtooth pattern to almost any measurement taken around a process. Other variation can be attributed to conditions, called “special causes,” that result from specific events outside of the process itself. These might include improperly calibrated instruments, insufficient operator training or outdated reagents, for instance.
Instead of continuous monitoring, SPC techniques use random sampling and statistical analysis to determine whether a noted variation is due to the process itself or to special causes. As special causes are detected, procedures can be developed and implemented to ensure that they do not continue.
X-Bar and Range Charts. The primary SPC presentations are X-bar and range charts. Each pair of charts is derived from the same set of data. Randomly acquired measurements are organized in groups, the size of which is normally defined by a statistician when a formal SPC program is implemented in a facility. The data included in the groups should cover a time period during which all normal process variations will occur. So, for measurements taken hourly, the grouping typically might be 24 (a full day) or eight (each shift).
Once collected, each set yields two values:
- The average, which is obtained by adding the values together and dividing the sum by the number of values in the set.
- The range, which is the difference between the largest and smallest measurements in the set.
- X-bar average.
- Upper control limit (UCL) at +3 standard deviations.
- Lower control limit (LCL) at -3 standard deviations.
- Range average.
- UCL at +3 standard deviations.
- LCL at -3 standard deviations.
By eliminating special causes, the range is narrowed between the upper and lower control limits. Each time this happens, the process itself is improved and made more efficient.
At times, the special cause can be a bottleneck within the process itself. It then is possible to evaluate the benefits of modifying the process to narrow these limits.
Process Capability Charts. While the X-bar and range charts describe whether or not a process is in statistical control, a process capability chart is used to determine whether or not the process is “capable.” In other words, as defined by the upper and lower specification limits, is the product of the process of acceptable quality? The statistical result of this analysis is the process capability index, or Cpk.
A Cpk reading of 1 means that the process variation meets the specification; however, the slightest upset would produce unsatisfactory results. For this reason, a Cpk of 1.33 or better is desired for a process to be considered capable of consistently meeting the requirements imposed on it by its internal and external customers.
Process capability is a predictive tool, and it requires that the process itself be statistically stable. This only can be ensured by consistently using X-bar and range charts.
The process capability chart is a histogram that presents the data in six blocks, ranging from -3 to +3 standard deviations. Vertical lines are drawn at the appropriate locations, and a “normalized” bell-shaped curve is superimposed on the bars. Please note that this is merely an indicator of the distribution of the results. The indicator of the capability of the process is the Cpk.
Control Limits. Operating processes that use SPC may have to deal with as many as three types of control limits.
- Operational control limits are used to ensure that safety and/or economic factors are not exceeded. These limits could relate to temperature, pressure, chemical residuals, inventory or other process variables. Frequently, corrective actions are associated with these limits. If the limits are exceeded, implementing a corrective action can bring the process back into control.
- Upper/lower control limits are calculated from X-bar and range values. These limits are ±3 standard deviations from the calculated average and represent a 99.97 percent confidence factor that any reading that falls between them is the result of normal process variations.
- Specification limits are used to describe the minimum and maximum acceptable levels of some process characteristic. In this case, acceptance is used to describe the range of variation that would be acceptable to a customer. These limits may or may not be related to other types of limits.