Autonomous operation

Increasing customer demands, stiffer competition, and an environment undergoing ever more rapid change – all these are challenges that modern industrial production operations today have to overcome.

Digitalization and networking of entire value chains make a decisive contribution towards companies in this environment being able to achieve their efficiency and profitability goals more successfully and more effectively. The ANDRITZ answer to this digital transformation is: Optimization of Process Performance (OPP).

Metris OPP comprises a range of digitally supported tools based on big data analysis to improve industrial processes. This system developed by ANDRITZ has been used successfully for over a decade. More than two dozen long-term OPP contracts have been concluded with customers, mainly pulp mills, but also with some steel and chemical works. The customers’ facilities include both older and some very modern mills and plants.

OPP can make both types of facility much more efficient, for example by avoiding sheet breaks in dryers or paper machines, reducing chemical consumption in bleach plants, selling surplus electricity to the grid thanks to a better steam-energy-balance, or preventing downtime by carrying out predictive maintenance.

Eldorado Celulose, one of the largest pulp manufacturers in the world, relies on Metris OPP with great success. Production and operating reliability have been increased significantly without incurring further investment costs.

1. INSTALLATION

Intelligent sensors are installed wherever the customer wants to have them: in the plant itself – and also outside it at trouble spots in the process chain. They can be installed easily and flexibly in new and in older plants alike.

A large number of sensors send data independently via a wireless connec - tion and with data protection to edge devices that are networked with the overall system.

2. MEASUREMENT

The sensor data measured and transmitted are verified in the system and stored for further processing.

Industrial plants produce vast amounts of data. The Metris OPP system gathers and compares them and also pieces them together. In reality, this is a very complex process that takes place at lightning speed. In a pulp mill, for example, several thousand process data are produced and processed in a single second.

3. DETECTING PATTERNS

OPP helps decrypt the DNA of an industrial plant and understand the characteristic way in which it operates and functions. Various peculiarities and anomalies come to light during this process that can then be ironed out. Plant operation becomes more stable and consistent.

4. OPTIMIZATION

Every aspect of the production process monitored by OPP can be analyzed, regulated, and improved as desired and as necessary. Faults are reduced, and production is more stable, safe, and efficient.

MEASURING AND OPTIMIZING PERFORMANCE

Metris OPP analyzes huge data volumes collected at numerous points in a plant’s production systems in order to use them as a basis on which to improve performance. This sounds easy, but it is in fact complex because plant operation has many facets. The control system controls the plant, and the operating personnel take appropriate action. Fluctuating raw material quality and environmental factors can lead to a slow but steady deviation of performance from the target set. In order to identify the causes in such a complex system, it is important to analyze both the structured data from the control system and also the unstructured data produced when the operating personnel take action. Interactions and control models can be derived from this and then applied in order to stabilize or increase the throughput. In addition, savings can be made on resources.

BIG DATA AND EXTENSIVE KNOW-HOW

Metris OPP masters this complexity and provides effective support to ANDRITZ customers. The system operates with 5,000 to 125,000 real-time process variables. These variables are collected via numerous sensors and the control systems in the plant and then analyzed statistically by the OPP software in order to detect any anomalies or potentials for improvement. At the same time, a team of 50 ANDRITZ engineers helps out the plant technicians on the site and continues developing OPP further. The team consists of control system experts, electrical engineers, electronics engineers, computer scientists, and specialists in chemistry, mechatronics, and mechanics. In order to exchange knowledge and ideas and also address new challenges together, ANDRITZ has developed an online tool called Logbook, where experts can exchange views on the topics in hand.

A SYSTEM WITH A FUTURE

ANDRITZ is extending and refining OPP on a continuous basis, for example in the deep learning sector, a machine learning method based on the way in which the human brain functions. In addition, so-called Bots are used – computer programs that act as personal assistants when using the OPP software. There are already almost 40 apps in use, and many more new and mainly mobile apps are to be launched in 2018. In addition, ANDRITZ is closely reviewing a list of around 500 possible features, most of which have been suggested by customers. And last, but not least, ANDRITZ has started an initiative to take a closer look at sensor technologies for machinery and plants. There are already very small wireless instruments available today that can be used to obtain more data from customer plants and, for example, to detect wear at an earlier stage. In this way, users are laying a solid foundation for predictive maintenance. Augmented reality is another focus. With the aid of modern mobile equipment like data glasses, information is provided exactly where it is needed.

WE ARE FOCUSED ON RESULTS

Leonardo Pimenta is Technical Control Manager at Eldorado Celulose in Brazil. He manages process engineering, IT solutions, and quality control in the mill. There are over 40 people in his team. Eldorado has been using Metris OPP from ANDRITZ since the end of 2016. Has it proved successful?

Mr. Pimenta, why did you establish a partnership with ANDRITZ to use OPP services?

The main reason we started this service was a proposal from ANDRITZ for the launch of a pilot project to improve our fiberline performance. ANDRITZ came back with some clear targets to control all the main processes in the fiberline by using APC (Advanced Process Control) and to ensure that these control loops would stay in automatic mode 90 percent of the time. We know from our own experience that the mill achieves better performance using automatic control systems than it does in manual operation. In addition, we felt that if we could achieve those targets, our operational stability would increase and our variable costs would be reduced.

Leonardo Pimenta, Technical Control Manager, meets with the OPP team to discuss progress. The team consists of Eldorado process engineers, reliability engineers, and ANDRITZ OPP analysts working side-by-side.

How did you then proceed?

We deliberately applied all the concepts and technologies that OPP offers all at the same time and fast. That’s the Eldorado way. From the very beginning, we formed a joint team of ANDRITZ OPP analysts, Eldorado process engineers, and Eldorado maintenance engineers who collaborate constantly. The front-end work included analyzing and tuning all the control loops in the fiberline. Then we set up over 40 projects together using combinations of smart sensors, APC, loop tuning, and data mining.

That sounds rather complex. Where specifically did you start?

We started with APC for the fiberline and have since added the recausticizing and the drying plant to this process. Our fiberline progressed from basic control to “hands off” and even “eyes off”, which left our operators free to work on other, more productive tasks.

What Key Performance Indicators (KPIs) have you defined, and what results have you obtained so far?

The results in the first year have been impressive. Operating efficiency was increased from 89.2 to 93 percent. Variable costs were reduced by five percent, production was 38,000 tons over the budgeted amount, and APC controls ran in automatic mode for 95 percent of the operating time.

What effect did this have on the mill’s productivity?

As already mentioned, we have achieved some important milestones in less than one year, including various production records. Every control loop that we operate with the aid of APC makes us money. Our 3.8 percent increase in operating efficiency has an enormous effect. The mill was designed to produce 1.5 million tons per year, and we are currently producing 1.7 million tons per year. That amounts to a revenue increase of many million Brazilian reals. And we have achieved this improvement with virtually no additional capital investment.

Is mill operation also more stable thanks to the use of OPP?

Yes. We were having issues with the hydraulic balances, which caused us to have to backwash a screen in the digester every three days. This cost us time and production. Ever since OPP was implemented and an APC routine was created, our digester has been running with excellent stability. We have not had to purge the screen in a year.

What about operating reliability?

We have three reliability engineers working on the OPP team who support the project to promote our risk-based maintenance program. In addition, the OPP analysts from ANDRITZ have tapped into our SAP maintenance planning system to obtain vital information from the 23,000 assets in this database. We have combined this data with info from our distributed control system. One glance at a computer screen shows us where the critical risks are. This enables us to avoid unnecessary shutdowns. All in all, we are operating at 95 percent equipment availability. That is excellent.

How difficult is it for you and your team to actually work with OPP?

Although there is really a lot to do, it is not complicated to work with OPP. ANDRITZ provides the programming that runs in the background. Our focus is on clearly defining our objectives – an increase in capacity or a reduction in variability for example. Then we define the variables that impact the outcome. Sometimes we have to add a new instrument so that we have reliable measurements for control purposes.

What plans do you have when it comes to Industrial Internet of Things?

We are very much focused on results. For example, we have an OPP project underway to assist our supervisors in adjusting the mill production in real time with the aid of computer-assisted algorithms. In addition, we are working with ANDRITZ on an augmented reality project. It will help maintenance staff obtain information in the field simply with the aid of augmented reality glasses. I am absolutely convinced that digitalization offers us many more opportunities for improvement.