Visions of Tomorrow – Engineered Today
Search

Selection and refinement in developing production wide performance monitoring

Author Jussi Parvianen
Posted on

Reducing thousands of measurements into a handful of key parameters

Industrial digitalisation has left factories with copious amounts of measurement data. A single production line can have thousands of measurement tags, each capable of producing a nearly continuous stream of information. This data is utilised, above all, in operating the process, but can it also be combined with smaller entities to aid in determining the state of the whole process?  

Typically, in graphical user interfaces of process automation systems, measurements are grouped in frames according to what part of the process they are positioned in. Usually, these measurements are placed on top of a crude drawing of the part of the process to illustrate where the measurements are physically positioned. Depending on the process, several frames might be needed to cover a whole production line.

On the one hand, this kind of representation is easy to interpret and serves the needs of process operators. On the other hand, critical information can be sparse and hard to find. Locating bottlenecks in the process can be difficult in this kind of system. Tracing the effects of a change in the quality of a raw material on the process can be even more cumbersome.

Obtaining critical parameters for performance evaluation

It is obvious that, in determining the state of a process, some information is more critical than the rest. For example, a drying process that utilises steam for heat can be seen as a bottleneck of the process when the steam valve is open wide enough so that the steam flow has plateaued, indicating that the dryer has reached its full capacity. Thus, this simple limit value of the steam valve at the start of the plateau can indicate the state of the drying process as a whole.

Critical process information can also be a combination of measurement data. In a wire filtering process, for example, the mass of a filtrate cake per area on the wire filter is a critical parameter. Assuming constant mass across the width of the filter, the mass of the filtrate cake per area can be calculated from the:

  • Width of the wire
  • Diameter of the drawing roll
  • Rotational speed of the drawing roll
  • Mass of solids in the feed per volume
  • Flow rate of the feed

As can be seen in this example, several data points can be combined into a single and easy-to-understand figure, thus lowering the amount of figures to be monitored. This information-combining effect can also be achieved when efficiencies, specific energy consumptions and other key figures are calculated for different parts of the process, e.g. heat exchangers, grinders and dryers.

It is easy to see that finding the parameters that best describe the performance of the process requires insight into the process at hand. Therefore, discussions with factory staff are crucial in choosing the parameters for performance monitoring.

After the parameters for performance monitoring have been selected, they can be visualised in a performance monitoring display (See Figure 1 below). This way, the crucial parameters that indicate the efficiency of production are all in the same place and the state of the whole production line can be easily interpreted.

Figure-1

Figure 1. An example of a performance-monitoring display. 

Matching contents with needs

As described above, measurements can be used as is, or in combination with other measurements to indicate the state of a process. In addition to these parameters, a performance monitoring display may consist of:

  • Waiting times and delays
  • Rate of production in different parts of the process converted into the same unit, e.g. metric tons of product per day
  • Current production rate in relation to the theoretical maximum
  • Energy of raw material consumption in relation to production rate
  • Cost of non-optimal operation
  • Other key indicators, e.g. Overall Equipment Effectiveness (OEE)

These figures can be displayed either as their current values, or as a short trend. They can also be visualised as gauges or traffic lights to ease interpretation.

The content of a performance monitoring display should be designed according to the needs of both operators and engineers. Operators can use it to evaluate the best ways to drive the process and engineers can use it to find the root causes of malfunctions. A well-executed performance monitoring display may serve several purposes, depending on the needs it was built.

Use of a performance monitoring display

A performance monitoring display can help or even add communication between shifts. At the start of their shift, operators can easily see the overall status of the process. It also helps to see how previous shifts have played out. This is useful especially when carrying out long operations that can span over multiple shifts, e.g. the ramp-up after a stoppage. Bringing in versatile methods for monitoring process performance can also help paint a clearer picture for shift supervisors and other staff of what is actually taking place in the process.

From an engineer’s perspective, having all the key parameters in one place makes it easier to dive into the root causes of factors and events that hinder production. This in turn speeds up problem solving and reduces downtime. Having all the crucial figures in one place also facilitates reporting.

Refining existing resources

Many production facilities already have the prerequisites for the contents of a performance monitoring display. Measurement positions are abundant and the data they produce is digital and is stored on servers. The existing automation system can be used as a platform for execution but ideally, there is some other software for data analysis and visualisation, such as Wedge, that is more suitable for execution.

The execution of a performance-monitoring project is presented in Fig. 2 as a continuous cycle of improvement. The process starts with understanding the needs of the client, after which an initial draft of the display is made. The draft is then improved with the changes decided upon with the client until it is implemented. After implementation, employees that make use of the display are asked for feedback. The feedback is then compared with the needs of the client and the display is further improved accordingly.

Figure-2_performance-monitoring

Figure 2. Execution of a performance-monitoring project.

Conclusion

Building a performance-monitoring display on a production line is a simple and effective way to highlight the efficiency of production. It is a way of compressing online measurements into an information dense presentation. It also enables shedding light on the most important factors that have to be kept in check in order to ensure safe, reliable and profitable production.

Performance monitoring must be executed in close cooperation with the client to ensure that the needs are met. Choosing the parameters to be displayed also requires a deep understanding of the process, a fact that further underlines the role of the client.

Enhancing performance monitoring in production is a natural first step towards more efficient production.

Leave a Reply

Your email address will not be published. Required fields are marked *

Jussi Parvianen

Jussi Parviainen graduated from the University of Oulu in 2018, majoring in process engineering with a focus on process automation. He has gained experience in process efficiency monitoring and data analysis since he joined Elomatic in 2017. Jussi currently works as a Senior Design Engineer at Elomatic’s Jyväskylä office.

Intelligent Engineering

Latest post

14/01/2021

With wind power towards zero-emission maritime traffic

Kirjoittanut By Antti Yrjänäinen

Ever-tightening environmental regulations, combined with the pressure to lower fuel costs, have made wind power an attractive option as transport energy for ships. Many sail applications have been developed, but no real breakthrough has been...

Read more » Lue lisää »
Top Engineer

Copyright © 2020 – Elomatic, All rights reserved