S. R. Subramani


ABB Industry, Singapore



asset optimization, benchmark, controls, paper quality


This paper describes the practical experience on Asset Optimization on the Paper Machines. It describes the identification of process and instrumentation issues, which affect the productivity, and also resolution of the issues. The optimization audit performed on two new high-speed Newsprint and Writing paper machines in Asia, resulted in enormous benefits in terms of quality and productivity and savings. The optimization audit identifies the bottleneck and limitation on the equipment, instrumentation and improves the process quality and productivity.

The Data Analysis tool is used to analyze the measurements. The objective of this paper is to look at the practical experiences and also how it has contributed in improving the process quality and productivity. The analysis methods and the other techniques are not explained in this paper because of the limitations.


Pulp and paper making is a dynamic process, and the process condition change every day. Normally the controls are tuned during commissioning. Then after that the process condition changes and equipment wears out, which deteriorate the efficiency of the process; in many cases the unmatched control tuning makes things worse, and eventually the process will have its own behavior which is uncontrolled. In most cases the operator learns to live with the limitation. This results in down graded performance of the process and controls. These issues degrade the efficiency of the process and affect the quality and loss of raw material and energy. The objective is to get the MOST out of existing equipment and identify the areas for change and modification with proof.

Asset Optimization has been done on a one-year and three-year-old paper machine, which run at 1500mpm and also few other smaller paper machines. To everyone's surprise, it resulted in enormous benefit with in few weeks. When new paper machines achieve such benefits, the opportunities are many folds with old paper machines.

These studies were done by experienced consultants with good process and control expertise and assisted by data analysis tools. The control work is done by process identification, simulation and current control tuning practices.


First task is to benchmark the current variability on the Basis Weight and Moisture measurements. It is important to identify the status of the machine quality and performance by collecting the data from the Quality Control System (QCS) before making changes. The Machine Direction (MD) and Cross Direction (CD) data is collected using a data collection tool, which can collect high frequency measurement data up to 2000 samples/second. The high frequency data analysis can reveal if there is any high frequency variability like vibration, which cause variability on Basis Weight and Moisture. The low frequency machine direction and process measurement data is also collected from the QCS/DCS system. The data is used to benchmark the variability on Basis Weight and Moisture measurements. The result is a window to view the variability of the whole paper machine, since the variability in the paper machine will be reflected in the final quality.

After the improvements are done, then data is collected again and the variability comparison is performed. Figure 1 shows an example of the comparison of before and after an audit. After extensive control and process work, the overall variability was cut in half. The cross direction is still high, but this mill only had manual control of the actuators. The improvement that was made was a result of manual adjustments made to the actuators.  This mill would be a good candidate for automatic CD control. 

Figure 1

Figure 1. The improvement in CD and decade 1, 2 and 3 are the result of the optimization


The data analysis is performed on each measurement using the collected data, through which it is possible to identify the status of the measurement and control. On the measurement it is possible to identify whether the measurement is noisy or filtered. By using various analysis methods like power spectrum and auto correlation it is possible to identify various problems with the process measurement or control. It is also possible to identify the different dominant frequencies in that measurement. Basis Weight measurement can be analyzed with stock flow, consistency, retention aid, etc, the analysis pin point which measurement is contributing variation and power of the variation to critical process parameters.


By collecting the data from the QCS and analyzing it, verifying measurement integrity is possible. Figure 2 shows the Basis Weight measurement is oscillating for each scan by about 1gsm. This was caused by poor alignment of an old scanner head. The oscillation is eliminated after scanner head was aligned properly.

Figure 2

Figure 2. Basis Weight measurement stepping with scanner misalignment


When process data is collected and analyzed it is possible to identify the status of the measurement and the controls. The following is an example of a measurement, with high filtering. The flow measurements were dampened with filters, when the filtering was removed; the problem with the control was exposed. The sticky control valve causes pulsation as shown in the measurements in Figure 3. Servicing control valve positioner and providing additional filters in the instrument airline eliminate this problem.

Figure 3

Figure 3. Flow measurement without filtering exposes the sticky valve problem

In Figure 4, the noise level on the flow measurement is about 150lpm on a range of 0 to 350lpm (>40% noise). This noise is caused by bad installation of the flow tube. The flow tube is of DC type and is installed on the downward piping and also mounted close to the suction of the pump. All the combined issues cause the flow measurements to be noisy. High filter has been introduced to get reasonable measurement for control. Because of this, the control has to be de-tuned to make it slow. De-tuning degrades the performance of controllers. In this case the transmitters have to be relocated and even changed to an AC type flow meter, which is appropriate for the application.

Figure 4

Figure 4. Noise on the flow measurement due to bad installation, with filter removed and then with filter introduced.


The Machine chest consistency measurement, which is of microwave type, has been erratic and the variation range is about 0.8% consistency. Dry stock compensation uses the Machine chest consistency to compensate the stock flow to maintain total dry fiber flow. Since the machine chest consistency is varying, this causes a change in stock flow but most of the time results in a wrong correction to the stock flow. These incorrect corrections cause variation in the final Basis Weight and Moisture. The scanner senses the change in Basis weight and Moisture and takes corrective action, resulting in variation in Basis Weight, Moisture and other paper properties. The mixing chest consistency, which is of rotary type, is stable during the machine chest consistency variation as shown in Figure 5. The dry stock control is turned off till the measurement problem is resolved.

Figure 5

Figure 5. Mixing chest and Machine chest consistency measurement on the same stock flow


Consistency unstable because of insufficient dilution water pressure
Figure 6 shows the impact of the dilution on two of the consistencies. Since the dilution pressure is not constant, it is a disturbance to the consistency controllers. If the loops are properly tuned, then their outputs should reflect this dilution pressure disturbance. Notice how similar the outputs of these two control loops are.  Also notice how the consistency was worse for both at approximately the same time. This is also when the dilution pressure was changing.  As a result of these interactions and the inability to get the dilution pressure into a controllable range, each of the consistency loops was slowed down.  This reduces the demand on the dilution pressure. Since this is the same dilution water that supplies the thick stock consistency, making the inlet consistency controls slow to stabilize pressure for the thick stock was the only alternative. The permanent solution should be to improve the demand and stability of dilution water pressure header.

Figure 6

Figure 6. Two consistency measurements and valve out puts, operations of one affect the other consistency

Ash Control:
It was noticed that the filler flow dropped by about 100lpm while pump output was constant. This causes the Ash measurement to drop as shown in Figure 7 and also causes sheet break. This was not noticed since the dip in flow last only for few seconds. This has been a result of the switching filter on the filler line for cleaning. Operating procedure has to be improved to avoid the discontinuity in the filler flow.

Figure 7

Figure 7. Filler flow variation and effect on Ash measurement

Identification of the source of disturbances
The paper machine has many process variables, which can cause variation in the final quality. The best way to identify the source of the variation is with the analysis of the final processes variable and the other process measurements, which has influence on the final process variable.

The following example shows a dominant variation at 100sec on basis weight measurement. The Figure 8 shows the actual measurements of basis weigh and headbox pressure. Just looking at the actual measurements cannot identify these problems. The Figure 9 shows the power spectrum of basis weight and headbox pressure. When the all the process variables that can influence the variability on the basis weight was analyzed, it was identified that the variability was caused by the 100sec cyclic variation of the headbox pressure. Then when the headbox pressure variation was eliminated, the basis weight variation has been controlled.

 Figure 8

Figure 8. The figure shows cyclic variation on basis weight and headbox measurement

Figure 9

Figure 9. Power spectrums of basis weight and headbox helps to pinpoint the cause of variation

In the process of these optimization audits we learned that the major issue for the instability of the paper machine is identified as wet end variations, which was causing sheet breaks and also taking longer time to stabilize after sheet break. Identifying and resolving instrumentation problems, tuning the controls, modifying the control strategy and also training the operator for best practices, have stabilized the wet end.

In these audits we have also optimized and stabilized the other control in the paper machine wet end area, steam section and QCS controls. The total audit involves analysis of at least a few hundred loops but only a few issues have been presented in this report. Apart from improving the control the operators have been trained to use the controls in a best way and also look for any limitation of the controls and take corrective action.

Wherever required, advance tie-up and multivariable controls have been implemented to automate and eliminate the operator actions. These controls monitor the limitation on the controls and the process and make correction on a different control and process area, which in-turn improves the previous measurement and stabilise the process.

Only paper machine examples have been presented in this paper. These audits are also have been done in the Recycle fiber line, Recausticizer, Fiber line, Recovery Boiler and Power Boiler plants. 


Mill Runability data
The monthly sheet beak data was collected before and after the asset optimization audit. The Figure 9 shows the reduction of sheet break duration by 29% and reduction on the sheet break numbers by 25%.

Figure 10

Figure 10. The above improvement on sheet break time and number of sheet break reduction contributed a productivity increase of 835 tons/month

Reel reports sheet break details are collected from the reports during July and October 2001 on the same grade run. There is a clear improvement from Figure 10 on the reduction on number of sheet break and also reduction on sheet beak duration.

Figure 11

Figure 11. Significant reduction of reel to reel sheet break time and number of sheet break on the same grade run.

Quality data analysis
The reel report is an excellent source of statistical quality information about each reel of paper that is being made. Every reel that is produced causes a reel reports to be generated .  These reports were automatically collected continuously and stored in a database. This information can then be used to trend any reel report information. The reel reports on the same grade are collected for quality comparison before and after the asset optimization audit. The results in 2 Sigma variations are presented in Figure 11.

Figure 12

Figure 12. Improvement on the paper quality is obtained as a result of the wet end stability after the optimization audit.

Result Summary

Sheet break reduction between July and October 2001.

Sheet break duration reduction 29% (18.5 hours extra productive hours in a month)
Number of sheet break reduction 25% (reduction of 58 breaks)

The following are the percentage reduction on long-term variability of 32 reels (about 1500tons) comparison between July and October 2001, ie before and after the Audit

Basis weight   53%
Bone dry weight  51%
Moisture 60%
Ash  80%

These results show that the control and process optimization result in improved machine performance and reduced variability in the paper quality.  There are also big reductions in maximum values of each quality parameter, which would have caused bad quality paper and rejects. Apart from the above projected improvements, There is other benefits like faster to achieve target after sheet break and during grade change, less bad quality rejects, savings on raw materials and steam. The results are consistent or even better for three months, when this article was written. Projected increase profit on this machine is above US $2,000 ,000; which is achieved with few weeks of this asset optimization audit and also with no capital investment.


Consultants with high level of control expertise, automation system knowledge and process knowledge perform the optimizations audit. The key factor in these audits is the analysis tools, which helps the consultant to identify the issues by using statistical and time series analysis and also identification of process model for tuning.

Now the challenge is to sustain the results achieved after these audits, which require commitment from the mill to monitor continuously and also take action as soon as some controls and process condition deteriorate or any change in process condition. One way to sustain the results is by performing periodical audit and evaluating the performance and make improvement. Another way is with advance supervisory controls, were the process models and expertise is built into the system. This advance control can monitor the status continuously and take corrective action or prompt the operator to take action.


Kevin Starr, ABB Process Automation Inc., USA. : Asset Optimization through Data Variation Analysis, TAPPI, August 2000.

Johan Jansson, Malardalen University, Erik Dahlquist Malardalen University and ABB Process Industries Sweden, S.R. Subramani, ABB Industry, Singapore, : Applications of physical models for optimization and control in pulp and paper industry, Asia Paper 2002.


[Home] [Title] [Author] [Organisation] [Keywords]