Reduce Production Costs with Data Analytics Software
Achieving a high rate of pulp and paper production while still maintaining consistency and quality can be a challenge. The pursuit of higher productivity puts a lot of stress on the equipment. Interruptions can be very costly, and a high variability is bad for quality. Thus, any measures to keep production consistent – or even improve it – without exhausting or damaging the hardware can mean big savings.
Data analytics software offers an alternative way for pulp and paper mills to reduce costs and improve quality without purchasing expensive new equipment. Using software to achieve process optimization and control can bring big returns – including reduced waste, improved efficiency and prolonged equipment lifespans.
SIMCA® Multivariate Data Analysis (MVDA) software provides a way to adjust production processes in order use raw materials more efficiently, increase output, and reduce wear and tear on equipment. SIMCA® uses historical process data such as yield, output and quality parameters to create a model of the production process at optimal performance. It can also help optimize steam and power generation, including recovery boilers, resulting in additional savings on fuel, as well as keeping emissions under control.
Production success in pulp and paper mills is governed by many factors, as processes are complex, and many pieces of equipment including paper machines and auxiliary devices are used. Sartorius’ multivariate data analytics software SIMCA® provides a comprehensive, yet easy-to-use toolset to analyze historical production data to determine the most influential factors driving final quality, yield, sustainability and many other important optimization goals you may have. The ultimate goal of such an optimization process can be full visibility and quality prediction from raw material to final product.
Coefficient plot for a multivariate variable, a point in a score plot, describing the operating state of a boiler / steam turbine power and steam generating unit with respect to NOx production. This plot shows the influences of the original variables on the point ranked from strongest positive (red) to the strongest negative (blue). This shows at a glance which factors should be strengthened and which need to be reduced to optimize NOx production.
One of the most prominent advantages of multivariate process data analysis is the reduction of information depth. Years of historical process data with hundreds of variables can be displayed in a single graph. Nevertheless, the whole process is lossless and tools like the above coefficient graph allow a drilldown at any point. This shows what variables have the strongest influence at each point and drive the process. It also uncovers hidden correlations as well as cause-and-effect relationships that provide additional insight in processes.
This makes multivariate data analysis an invaluable tool for training. New operators and will be able to understand the function of the mill much better by seeing how different parts interact and which variables drive the interaction. Additionally, it also verifies the knowledge of seasoned employees providing the added benefit that it is not lost with these operators going into retirement.
Typical questions around production optimization answered by using data analytics:
One of the most effective ways to ensure your processes stay within their ideal parameters is with real-time process monitoring and control. Being able to take corrective action immediately helps reduce waste and improve efficiency. With SIMCA®-online, you’ll gain confidence in your production processes and achieve more consistent product quality.
The ability to monitor and control operations in pulp and paper – like in any other process – requires the control program to be connected to the distributed control system (DCS) or the manufacturing execution system (MES), depending on layout of the plant. The control software by Sartorius – SIMCA®-online – communicates with the DCS or MES and does not replace them, therefore supervisory control.
Modelling with data analytics can be used to combine many control charts for single variables into one multivariate control chart. As this reduction of information depth is lossless the drilldown capabilities can be used at any point in the control chart to detect the root cause for a deviation and to take corrective action.
Setting up multivariate process control is quite uncomplicated. The following is a typical scenario, in detail there will be adaptations depending on conditions and customer priorities:
Provided that according access to IT installations is possible, all these steps, including consulting and training, can be done remotely without any external persons on the mill premises.
Learn More About SIMCA
Learn More About SIMCA-online
Another advantage of real-time monitoring is predictive control. SIMCA®-online is able to predict how a process is developing and determine whether it’s deviating from an optimal model. This provides an opportunity for prescriptive action, changing the variables early on, which can be done automatically in the form of a closed loop, or manually by an operator.
Schematic of process control using SIMCA®-online. The program connects the DCS or MES via several different databases which in turn connect to a data network. A paper machine and one or more auxiliary devices of all kinds hook up to the same network. The SIMCA®-online server reads data from these databases and can also write back to the historian or batch database (if available) and if desired to the DCS / MES.
A setup, as shown in the schematic, can be used for process control, ranging from simple monitoring to closed loop control. In case of monitoring SIMCA®-online compares incoming process data to the model of the ideal process and plots it in a single multivariate control chart, with according 2 σ and 3 σ deviation limits. Operators can so easily see when a process is moving too far away from ideal.
Predictive control is the next step up on the sophistication scale. In this case the software predicts the development of the current process based on knowledge gained from the analysis of historical data. After a certain process run time it will therefore be able to predict process outcome (yield, quality parameters) or if it is likely to go out of control in the future. In the latter case this can be used to alert operator far earlier that the human eye sees the deviation happen.
Most advanced is prescriptive control. It adds the option to advise on changes to important settings to prescriptive control. This means operators know the parameters to reset to new values to stop processes from deviating. We refer to this as advised future. If desired this can also be taken to closed loop control by writing the changes back to the DCS / MES automatically.
Data analytics software costs far less than investments in capital equipment, typically doesn’t require permits and is quick to implement compared to installing new equipment.
Reduce Emissions: Optimize Your Heat and Power Units Using Data Analytics Try our Boiler Benchmark Survey >
The return on investment (ROI) for pulp and paper mills that have implemented SIMCA® and SIMCA®-online for process optimization and control is six times to nine times or more, per year. In short, companies can save millions of dollars in just a few months with the software.
The goal for one pulp and paper company was to improve recovery boiler efficiency and save fuel cost.
Two paper specialty companies implemented SIMCA® and SIMCA®-online for real-time assessment of product quality.
A customer reduced soda waste and lowered material costs using SIMCA®-online
Optimize Your Heat and Power Units Using Data Analytics Try our Boiler Benchmark Survey >
Optimizing power production and emissions control using software is far less costly than purchasing new equipment, and it avoids permitting hassles.Read more
Data analytics software helps you optimize your processes and keep them under control during manufacturing.
You can use data coming from your instruments and analyzed in a multivariate way to manage the long-term operational health of production equipment, reduce unscheduled downtime, and prevent breakdowns. With a statistical analysis of all the factors that lead to wear and tear, you can get insights into the operational settings or processes most likely to extend the life of equipment or shorten downtimes between processes, and even implement predictive maintenance.
Overall equipment efficiency (OEE) is the gold standard and a best practice for measuring manufacturing productivity. It provides an objective measurement for the percentage of manufacturing time that is truly productive. By measuring OEE and analyzing underlying losses and bottlenecks, you will gain important insights on how to systematically improve your manufacturing process.
The right data analytics tool can help you assess the composition of chemicals, minerals and other raw materials to make sure they meet production requirements. This can help prevent inferior raw materials from being used that could affect the quality of your final product or ruin batches. It can also help you find the right volume, temperature, composition or any other relevant property of raw materials to optimize your process and produce the best quality product with the least waste.
Read more: How Manufacturers Are Using Big Data Analytics to Improve Processes
With typical process monitoring applications, your operators will have a number of different control charts they need to watch and monitor for deviations. Combining multiple charts into a single control chart to monitor all variables simultaneously and get alarms when a process starts deviating from the optimal path can be a huge benefit, especially when you are operating with less than normal staff. Drill-down capabilities will allow you to identify root causes for deviations immediately.
With statistical process monitoring in place, you’ll be able to more quickly identify any defects in your products, materials, equipment or processes. Ultimately, that will mean an improvement in your overall production quality, reduction in process deviations and fewer ruined batches.
The multivariate analysis model provides a basis for predicting quality parameters over time. It lets you predict the final critical quality attributes with a high degree of confidence. Manufacturers can use advanced data analytics to compare and measure the effect of various production inputs – often finding surprising and unexpected dependencies that are impacting output.
Watch this recorded webinar: Predictive maintenance – monitoring and forecasting the state and performance of machines and equipment
When it comes to creating an optimal manufacturing process that limits variation and conserves energy or resources, or a developing a new formula that...
You may have heard the term Six Sigma used in conjunction with lean manufacturing, a Kaizen approach or continuous quality improvement. Perhaps you th...
Please select your country so we can show you products that are available for you.
The content of our website is always available in English and partly in other languages. Choose your preferred language and we will show you the content in that language, if available.