Metals makers press to connect islands of automation as, in the words of one technology expert, ‘everyone talks about Big Data, but data is nothing. What matters is intelligence.’
Although the relationship between metals company strategic planning and the tactical shift-by-shift, tap-by-tap operations on the shop floor has rarely been as hostile as management vs. labor, there have been many technical impediments to the smooth flow of information both ways. Those impediments are now falling away thanks to technological advances, as well as realizations from both camps that they can make common cause. New approaches are being used to integrate the business side and mill floor processes in metals.
It is still the case that there are islands of automation in the steel industry, according to Stefan Koch, global lead for metals at process-control giant SAP SE, Walldorf, Germany. That is not just the case country to country, or company to company, but in some cases plant to plant, he said. Even facilities next door to each other it is common to see.
Beyond the linkage within production, there is also the linkage of production to sales, supply chain, even human resources. That is not just a beautiful dream. As we look to our customer base in metals, Koch said, especially the large-footprint steel and aluminum companies, they are all looking at how to get away from those islands of automation. They are even looking beyond their own fence lines.
Metals makers have made good progress in some areas, notably enterprise resource planning (ERP), as well as on connecting their rail and highway transportation systems, according to Koch. But they have not made good progress connecting any of that to the shop floor. Everyone talks about Big Data, but data is really nothing. What maters is intelligence.
Among larger facilities Koch relates that the intelligence they seek is to be applied predictively for quality and ultimately for profitability. In the end you get more money for better metal. The first step is often predictive maintenance. Many companies have partial solutions in place. Each one is good on its own, but they can be difficult to organize, difficult to harmonize.
The prevalence of those patchworks of independent systems is a legacy of both the age and the consolidation of the industry. There are furnaces and mills next to each other or even upstream and downstream of each other that were built by different engineering firms for owners in different decades.
There is also another legacy in the metals industry: independence, not just among competitors, but between furnaces, lines, even shifts within the plant. That can hamper sharing of best practices. Often an opportunity for improvement in the system can be traced back to a customer complaint, Koch noted ruefully.
But whatever the source, it is essential, he added, for metals makers to recognize patterns and take actions. They have seen the benefits of continuous improvement in isolated cases, and now they are coming to realize the broader solutions. I think we are on the edge of major advances in this area.
Examples abound, and are almost too easy to cite. But Koch offers a short, compelling one. One of our customers was able to shorten the start time for a coating line by just 10 meters per coil. Just 10 meters. Not a big deal, given the length of a coil. But over all the coils coated, the company realized an annual savings that ran to euros in the seven digits.
Far from criticizing the inroads made through process control and analysis, Koch believes that was a good point of departure. The devices are smart, and people in the metals industry are techies. They like that kind of stuff. It is all continuous improvement. The speed of reaction to exceptions is an ongoing game. In the hot mill it is one thing; in the cold mill it is something different. That is fine. The challenge then becomes to harmonize the data models. That is what enables the intelligence. The frontrunners are already looking into that.
Given the high capital cost of equipment in a metals mill, and the extreme operating conditions, there is little surprise that maintenance and operational reliability were early applications for optimization and continuous improvement.
When we talk to metals companies, they are very focused, but also very siloed, according to Simon Davidoff, business manager of plant data systems for the digital factory division of German multinational systems company Siemens AG. They say they want to make this blast furnace or that rolling mill more efficient. They ask what instrumentation and what software do they need to draw out the data they need to draw out those efficiencies.
There are also concerns about cyber threats. Those perils are ubiquitous, but they are very real in the industry, given the damage done to a blast furnace and mill in Germany in December 2014 after an electronic security breach. Attackers used booby-trapped emails to steal logins that gave them access to the mills control systems, according to the initial news report by the BBC. The steel mill was not identified. The hack led to parts of the plant failing and meant a blast furnace could not be shut down as normal. The unscheduled shutdown of the furnace caused the damage, the report said, according to the BBC.
We have guys out at a steel plant now conducting a risk and vulnerability survey. In this one area progress in the metals sector has been rudimentary. There is so much scope for improvement. Basic stuff like dont plug in flash drives, make sure things are locked, make sure firewalls are properly configured and current.
At higher levels, Davidoff noted that maintenance and operations figures large in the management and budgeting around assets in production. Reducing downtime is very important. We are talking about massive costs per hour out of operation, and also all the headaches and expense from running around for spare parts.
There is little surprise that with high-capital assets, long supply chains upstream and down, and tight tolerances for quality, that unplanned outages are the bane of the metal maker. Maintenance is not glamorous, but when technology can start to make it predictive, it enters the realm of continuous improvement in both process and in operations.
Whats new at this level is the need to be flexible in the use of energy, Davidoff said. By that he means not just energy-efficiency in terms of lower consumption, but really smart use. He explains that certain pumps, reactors, motors and other equipment are running all the time, some at constant rates, some at variable rates. Other equipment is only operational sometimes, again at constant or variable rates.
Each piece of equipment has its ideal operating and resting rates and times. As individual assets, those optimum states are probably specified by the manufacturer. In use, however, factors like wear and ambient conditions, as well as the rates and times that related equipment is operating, with change the in-service optimums for each component. Adding the fourth dimension, those optimums will change over time. Data can track those changes. Big Data can integrate those changing optimums, and optimize the whole line or process, even the whole plant.
When certain assets are running at certain times, Davidoff said, you may be able to turn one up or down just a notch. We have found in other industries, such as automobile manufacture, that distributed control through the system has enabled better adjustments. More data makes for more connections, but also is more challenging to the whole system.
The ultimate goal, Davidoff noted, will be to predict the failure of a component, and send a flag not just at the maintenance level, but up to operations and management saying watch this bearing or motor and have a spare ready. That will enable companies to avoid production outages and can be tied all the way up to the order book.
One of the meta challenges of continuous improvement and the integrations of management systems with the shop floor is that historically it has been imposed from above or outside. The guys making steel each day dont necessarily like the business-school people looking over their shoulders, according to Mike Falk, president and chief executive officer at Falk PLI Engineering & Surveying, Portage, Ind. The big mills have tended to hire the big consulting firms to bring in Big Data, and that has challenged a lot of conventions in the industry.
Having worked for an engineering firm previously, Falk can relate to that resistance. But he has also seen firsthand that more and more operational folks are having those dialogues internally and even externally. So, yes, there can be continuous improvement in metals making, and yes, it can come from within.
Several sources suggest that two trends are coming together. After decades of rationalization, management around the industry is realizing it is impossible to cut your way to prosperity. At the same time, operational and technical personnel are realizing that working conditions as well a profitability are most easily improved through optimization.
We get to see hundreds of sick furnaces and casters every year, Falk said. Even healthy equipment and operations can be improved. Even if you are already getting to 100 percent on spec material, you can get to it faster.
Deana K. Lecy, director of sales and marketing for Falk PLI, concurs. I have been with the company for 10 years, and was in industry before that. There has been a complete change of how people share information. When I came in, it was only possible to benchmark against your own similar operations. If there were a problem at the hot mill, they would point back to the caster, but could not suggest how to solve the problem.
And those were the best situations. We have worked at facilities where there were two furnaces and those guys would not even talk to each other, Lecy recalls. Now our company has had 200 caster campaigns. We are able to see issues both within companies and across the industry.
One of the challenges across industry, Falk said, is that as companies have downsized, they have reduced or outsourced engineering. So they dont even have the capability or capacity to have that internal dialogue about process improvement or optimization. Lecy said, If you are down in manpower you are down in experience. That is where a company like ours comes in.
She noted that billions of data points make for better decisions. Executives are asking what else they can do, because they have been focused on capital and equipment. We are focused on process improvement.
Falk said that for measureable results metals makers have to have a process on how to gather that data and how to use it to make decisions. The majority of operational people are metallurgists. The science of the mix is what is important to them, the temperature and the solidification. They also understand mechanical issues, wear and alignment.
But, he added, the first time he proposed quality improvements on a caster that were based on a holistic look, the initial response was skeptical. I suggested tightening one variable to open others. We have lots of ideas on how to use Big Data on an engineering level. That also filters up to management and back office, but operations are the first focus.
The principles apply beyond the big integrated operations as well. For second-tier processors, they may be buying substrate from just one or another supplier, Falk said. If they better understand their process and their markets, that can open their options on the supply side.
This leads Falk back to a universal view of continuous improvement. Think of the number of times that your material does not reach first-time prime. What is the cost of that to your client, as in missed schedules, lost orders, rework? It is always cheaper to get it right the first time, looking not just at remanufacturing costs but also customer and public relations costs.
Continuous improvement is a big issue in steelmaking, especially process analysis and process automation, Rebecca Dettloff, marketing manager for SensoTech GmbH. The emphasis now is on in-line analysis providing real-time data online and replacing off-line procedures, that require manual sampling and provide delayed results. The company is based in Magdeburg-Barleben, Germany, and has an office in Wayne, N.J.
The real-time monitoring of critical parameters has a big benefit in improved quality assurance, higher resource efficiency and lower costs, Dettloff said. This is why the market has a high potential, but not every steelmaker has the understanding of what greater benefits they can. It is still a challenge.
Using in-line process analyzers, enables online information on the process status and alarms for when the process is above or below limits so that countermeasures can be taken on time. For example, the acidity in a pickling bath or the oil conditions in a cold-rolling mill. Integrating the measurement data into the process automation is simple, Dettloff said, so the next step for many steelmakers is to connect in-line analysis systems to an automated level of process control.
To foster demand, equipment companies can do well by doing good. We will have a growing demand, Dettloff noted, if we are able to inform more steelmakers about the great benefits (of advanced analysis and integration), because the positive feedback we get from our customers effectively helps to optimize their steelmaking processes.
She said that there are still steelmakers that take manual samples and make an offline laboratory analysis. That is time-consuming and does not provide real-time data to intervene in the process in time in cases of deviations from production reference values.
At the most elemental level, continuous improvement can be as simple as real-time recalibration. In principle, our analysis system doesnt need a calibration, Dettloff said. If the process conditions make a calibration necessary for effective measurement, it can be done, because the controller, part of the analysis system, includes calibration functions.
Using sensors installed in the pipe, vessel, or bath enhances quality, Dettloff said, because critical parameters are monitored constantly and in real time. There is no need of manual sampling and lab analysis. The data are provided online, and for process automation the data can be sent to the process-control system.
In the event of deviations from the production reference value, a signal will be sent immediately, so the manufacturer can take countermeasures on time. Consequently, the focus is on a preventive quality management avoiding failed batches and costs. That helps to optimize production efficiency, Dettloff said.
Also, because of knowing the values of critical parameters, the manufacturer can control the process in a targeted, efficient way. For instance, how much of a certain liquid needs to be added within a certain process to ensure on-spec results and minimum waste. The focus is on saving resources and costs, and assuring quality, Dettloff said. To have detailed information on the process parameters increases the safety and utilization.
She summarizes that metals makers are looking for process analytical technology providing them real-time information on critical parameters that have an important impact on the quality and resource efficiency. Such technology contributes significantly to the optimization of the production process. I feel the challenge is to inform manufacturers of the technical solutions and the benefits.