Intelligent Technology Helps Save a Plant
Take a plant that has been in existence since the turn of the century; a plant whose well-focused, small set of products has a strong demand in the marketplace; and one that has a few important, strategic advantages in the marketplace, such as a 100-MW hydroelectric plant. A natural assumption would be this is a plant that has little difficulty being profitable.
Take a plant that has been in existence since the turn of the century; a plant whose well-focused, small set of products has a strong demand in the marketplace; and one that has a few important, strategic advantages in the marketplace, such as a 100-MW hydroelectric plant. A natural assumption would be this is a plant that has little difficulty being profitable. Not quite so.
The plant is the Elkem Metals Co. Plant in Alloy, West Virginia. It produces high-quality silicon metal for various industries, including steel, iron, aluminum, silicones, and electronics. Not much has changed at this plant for many years. The assets (e.g., reactors), the availability of raw materials, and cost of raw materials have changed little. Neither has the availability of a work force nor the process used to make the product line change significantly. If anything, demand for the products increased.
So why would there be a problem in staying profitable? The answer is related to market price. Although global supply of the product did not increase substantively, availability of the product in the free world changed, as the former USSR changed and started to sell its products. The effect was a net decrease in price.
The plant's profitability hit a low, and the company's home office considered shutting the plant down. The plant manager and his team knew they had to do something effective and quick. Their solution included development of new process control systems using intelligent technologies.
The basic process
Silicon products are made from U.S.-based quartzite rock, which is basically silicon dioxide. Crushed quartzite goes by the common name of sand, like that which is found on many beaches.
The quartzite raw material is mixed with wood chips and coal. The quartzite is reduced by the coal at very high temperatures to produce the silicon metal products.
Silicon products are produced in submerged-arc furnaces that stand over three stories high. Raw materials are continuously fed into the top of the furnaces, then pass down through the furnaces' hearths where they are heated by an electric arc passing between three carbon electrodes. When the materials' temperature passes 3,000 8F, the coal reduces the silicon dioxide to silicon metal.
When the silicon metal is ready to leave the furnaces, operators open tap holes in the lower section of the furnaces. A white-hot metal stream flows from the tap hole into ladles positioned below.
As the exiting metal fills a ladle, refining techniques remove impurities found in the original raw materials. It is in this refining process that intelligent technologies were introduced.
The Silicon Refining Assistant
The Silicon Refining Assistant (SRA) was built using off-the-shelf developmental tools, and IBM personal computer (PC) standard architecture that uses Intel processors. For ease-of-use in a manufacturing environment that includes necessary use of gloves, the user system interface included a touch screen setup for both user input to the system and system output to the user.
The original SRA employs 21,000 lines of code that include over 1000 rules. Business rules and process rules relating to refining are encoded in the SRA. Process rules have the form as illustrated in the following example:
RULE process rule 123: IF weight is less than or equal to X pounds AND aluminum is less than Y pounds THEN a process change is needed now whereby Z pounds of Refining Additive A needs to be added.
SRA presents questions and answers, process checks, and process guidance for operators. The system presents questions and possible answers to operators. The operators respond to the system using the touchscreen interface. Between this question/answer communication model, data provided to the system from sensors, and data provided by laboratory tests, the system becomes familiar with the refining process as it operates in real-time. As a refining cycle progresses, the system analyzes the data it receives and determines which advice, in the form of checks and checklists, should be presented to the refining operator. When the system determines a need, it presents process guidance to the operators.
SRA includes a process database that, as refining events occur, stores, uses, and archives the process data. This historical process data provides multiple benefits, such as allowing the system to notice and react to process trends, allowing process engineers to study the data off-line, and providing process status reports to management.
The system has been subject to many changes and updates. Its use created an immediate consistency in operation. This consistency allowed the company to better see the results of its process rules. Where rules produced desired results, they were retained; where rules did not produce desired results, they were changed or deleted. Use of the system assisted the company in focusing and highlighting its continuous improvement program, which spilled over to other components of the refining process, including changing the method of introducing additives.
Dramatic quality improvement
In producing silicon metal, the company uses the adjective "in-grade" to mean the product is within the required quality specifications.
The use of the Assistant, together with other process improvements highlighted by the Assistant (e.g., method of adding refining additives) and other research and engineering, resulted in near 100% improvement of the average in-grade production of silicon metal products at the plant. Process operation became more consistent, giving metallurgists and process engineers opportunity to determine effective process improvements. When consistent processes were noted, process changes were tried one at a time. Effective changes were retained.
Permanent, effective changes were continually identified. The refining process kept improving. Today, the process is run at near 100% in-grade production.
Dramatic financial improvement
Although all products can be sold, the in-grade products sell for five times the price of the out-of-grade. Production of in-grade and out-of-grade requires the same plant assets, the same labor, and the same raw materials costs. When cost factors stay constant, and price obtained in the marketplace of a large proportion of production goes up by a factor of five, the profitability of the plant dramatically improves.
The total project cost was just under $200,000, which included multiple copies of the hardware suite and software licenses, custom development, and travel expenses for the developers. The return on investment took only two months. The system increased dollars to the bottom line. This financial effect came from increasing income, in the millions of dollars.
For more information from OXKO, visit www.controleng.com/info :
Steven W. Oxman is the president of the OXKO Corp. OXKO was responsible for the development of the SRA, working with many process experts from Elkem.
|Search the online Automation Integrator Guide|
Case Study Database
Get more exposure for your case study by uploading it to the Control Engineering case study database, where end-users can identify relevant solutions and explore what the experts are doing to effectively implement a variety of technology and productivity related projects.
These case studies provide examples of how knowledgeable solution providers have used technology, processes and people to create effective and successful implementations in real-world situations. Case studies can be completed by filling out a simple online form where you can outline the project title, abstract, and full story in 1500 words or less; upload photos, videos and a logo.
Click here to visit the Case Study Database and upload your case study.